Advances in Immunology Volume 15

ADVANCES IN

Immunology E D I T E D BY

F. J. DIXON

HENRY G. KUNKEL

Division of Experimental Pathology Scrippr Clinic and Research Foundation La Jolla, California

The Rockefeller University New York, New Yo&

VOLUME 15 1972

ACADEMIC PRESS

New York and London

COPYRIGHT 0 1972, BY ACADEMIC PRESS, INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC. 111 Fifth Avenue, New

York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD. 24/28 Oval Road, London NWI

LIBRARY OF CONGRESS CATALOG CARDNUMBER:61-17057

PRINTED IN THE UNITED STATES

OF

AMERICA

PREFACE Despite our very considerable information about antibodies and the immune process a number of striking gaps in our knowledge certainly remain. Perhaps foremost among these is the mechanism by which the diversity of antibodies is achieved. Immunologists remain divided into two highly divergent camps on this issue and those espousing the germ line theory appear almost as numerous as the so-called somatocists. The very significant developments regarding the structure of antibodies have failed to resolve the dispute. Many workers in the field now feel that the answer will come only through concerted studies at the cellular level. An understanding of the different types of lymphocytes and their differentiation appears essential. Volume 15 is replete with new information in this area. In the first chapter Dr. David Katz and Dr. Baruj Benacerraf have summarized in a superb fashion the many lines of evidence concerning the interaction of B and T cells in the immune response. The regulatory effect of the T cells on B cell activity has clearly emerged as a dominant principle. The cooperative interaction between specific B and T cells and antigen is discussed at length and the evidence for distinct factors secreted by T cells which are capable of affecting B cell function is presented. The advantages of such a two-cell mechanism in terms of host defense mechanisms become apparent. In the second chapter Dr. Emil Unanue deals with the other important cell involved in the immune response, the macrophage. He has contributed very significantly to our understanding of the role of this cell in the removal, processing, and presentation of antigen to the responding lymphoid cell. Various in vitro systems for studying the immune response are considered in detail and the significant role of the macrophage in most of these is apparent. Variations and contradictions of past studies, primarily because of the widely different antigens employed, are brought together so that a coherent picture emerges. The chapter by Dr. Joseph Feldman covers a most timely subject, immunological enhancement. It is a phenomenon that has long been known but which has suddenly come to the fore with the appreciation that it has broad immunological significance ranging .from effects on tumor growth to the privileged position of the fetus in the maternal environment. Dr. Feldman prefers the term “immunological blockade” in this broader setting and has given it specific meaning in terms of known ix

X

PREFACE

antibodies and their effects on lymphocytes. Although considerable gaps remain in our knowledge of the exact types of antibodies involved and their specificity, the author has done much to dispel the mysterious aura that for so long has surrounded the enhancement phenomenon. The fourth contribution, by Dr. David Gasser and Dr. Willys Silvers, presents in detail the subject of sex-linked or presumed sex-linked antigens. The intriguing transplantation phenomenon of specific rejection of certain male to female skin receives special consideration. Evidence is cited which supports the concept that this Y antigen is determined by a gene on the Y chromosome. However, it has not reached the status of the more numerous X-linked antigens which are also discussed. The authors avoid the intricate language that characterizes many reviews in transplantation immunology; this chapter should prove broadly enlightening. In the final chapter Dr. Edward Franklin and Dr. Dorothea ZuckerFranklin present a very thorough review of the problem of amyloid and the recent exciting developments concerning the nature of the deposits. This very considerable clinical problem has intrigued immunologists for many years but always proved uniquely resistant to the many investigative efforts. It now appears that there are two distinct types of deposit. One of these clearly involves the variable region of immunoglobulin light chains. The other type is less well defined but appears to involve a totally different protein. Much of the controversy and confusion that have troubled most outside observers is resolved in this very timely review. The constant cooperation and assistance of the publishers in the production of Volume 15 are most gratefully acknowledged. HENRYG. KUNKEL FRANK J. DIXON

LIST OF CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors’ contributions begin.

BARU J BENACERRAF, Department of Pathology, Harvard Medical School, Boston, Massachusetts ( 1) D. FELDMAN, Department of Experimental Pathology, Scripps Clinic and Research Foundation, La Jolla, California (167)

JOSEPH

EDWARD C. FRANKLIN, Rheumatic Diseases Study Group, Department of Medicine, New York University Medical Center, New York, New York (249) DAVIDL. GASSER,Immunobiology Research Unit, Departments of Medical Genetics and Pathology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania ( 215) DAVIDH. KATZ, Department of Pathology, Harvard Medical School, Boston, Massachusetts ( 1)

WILLYSK. SILVERS,Immunobiology Research Unit, Departmnts of Medical Genetics and Pathology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania ( 215)

E. R. UNANUE,Department of Pathology, Harvard Medical School, Boston, Massachusetts (95) DOROTHEA ZUCKER-FRANKLIN, Rheumatic Diseases Study Group, Department of Medicine, New York University Medical Center, New York, New York (249)

vii

The Regulatory Influence of Activated T Cells on B Cell Responses to Antigen

.

DAVID H KATZ AND BARUJ BENACERRAF Deportmenf of Pofhology. Harvord Medical School. Borfon. Morrochuretts

. . . . . . . . . . . . . . . . . . . . .

I . Introduction I1 Specific Cells of the Immune System I11. Requirement of Two Distinct Lymphoid Cell Types in the Development of Humoral Immune Responses . . . . . . . . . A . Response to Foreign Erythrocyte and Protein Antigens B. “Carrier Effect” and Cooperative Interactions Specific for Different Determinants on the Same Antigen IV. Nature of the Regulatory Influence of Activated T Cells on Antibody Responses by B Cells A . Stimulation of B Cells in the Absence of T-cell Regulation . . B. Effect of T-cell Activity on the Class of Immunoglobulin Synthesized C. Role of T-cell Regulation in the Selective Pressure by Antigen on B Cells . . . . . . . . . . . . . V. Immunological Specificity and Properties of T and B Cells Concerned . . . . . . . . with Cooperation Phenomena . A . Immunological Specificity of T and B Cells B. Antigen Receptors on T and B Cells C. Recognition of Hapten and Carrier Determinants by T and B Cells . D Sensitivity and Resistance of T and B Cell Function to X-Irradiation and Corticosteroids VI. Mechanism of Regulation of B Cell Function by T Cells A . Transfer of Genetic Information B. Antigen Presentation and Concentration . . . . . . C Regulation of B Cell Function in Antibody Production by Mediators Produced and Secreted by T Cells VII . Suppressive Effects of T Cells on Antibody Synthesis . . . . A . Enhancement of Immune Responses by Depletion of T Cells . . B. Suppression of Antibody Responses by the Administration of More . . . . Than One Antigen (Antigenic Competition) . VIII. Functions of T and B Lymphocytes in Various Immunological Phenomena A . Immunological Tolerance . . . . . . . . . B. Immunological Memory . . . . . . . . . . C Immunological Adjuvants . . . . . . . . . D . Cell-Mediated Immunity . . . . . . . . . . IX . Biological and Pathophysiological Significance of the Regulatory In. . . . . . fluence of T Cells on Antibody Production . References

.

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

.

. . . . . . . . . . . . . . . . . . . . . . .

.

. . . . . . . .

. . . . . . . . . . . . .

.

. . . . . . . . . . . . . 1

2 3 4 4

14 23 23 25 26 28 28 32 33 37 42 42 43 47 62 62 65 67 67 75 79 81 82 85

2

DAVID H.KATZ AND BARU J BENACERRAF

I. Introduction

The clonal selection theory of Burnet and his postulate that antigenreactive precursors of antibody-secreting cells bear antibody receptors of unique specificity (1-3) have been largely substantiated in the past decade ( 4 3 ) . Another major advance in immunobiology has been the recognition of two pathways for the differentiation of antigen-reactive cells. It is generally accepted that a class of bone marrow lymphocytes migrates to the thymus where the cells develop the ability to respond to antigen. These thymus-derived lymphocytes, generally referred to as T cells, are responsible for the various phenomena of cell-mediated immunity: delayed sensitivity, homograft, and graft-versus-host reactions. The second lymphocyte cell type arises also in the bone marrow and settles ultimately in distinct anatomical sites in peripheral lymphoid tissues where these cells give rise to the precursors of antibody-secreting cells, B cells (9-11). The most recent discoveries in immunobiology concern the realization that the differentiation of antigen-stimulated specific B cells into antibody-secreting cells depends, for most antigens, on the concomitant activity of specifically stimulated T cells. The original observations established the requirements of specifically activated T cells for the antibody response by B cells to antigen in vivo and in witro in various systems and clarified the relationship between hapten determinants and carrier function originally introduced by Landsteiner ( 1 2 ) . It was later recognized that the effect of stimulated T cells on the response of B cells to antigen is more complex and affects also ( a ) the switch from the production of IgM to IgG antibodies and ( b ) the rate of selection of specific cells by antigen in the immune response as reflected in the change in affinity of humoral antibody with time. It was further shown that the activity of histocompatibility-linked, specific, immune response ( Ir ) genes in T cells is essential for all these phenomena triggered by antigen. More recently, it is becoming apparent also that regulatory effects of activated T cells on antibody responses by B cells may be suppressive under certain conditions, whereas under other conditions, as stated above, they are stimulatory, which may explain the well-known phenomenon of antigenic competition. In fact, what appeared at first as an important and essential cooperation phenomenon between two specific cell types and antigen to trigger effective antibody responses is now more appropriately interpreted as the expression of a fundamental regulatory function of activated T cells on B cell responses. The present review first describes the experimental data on which these statements are based and relates how insight into these fundamen-

REGULATORY INFLUENCE OF ACTIVATED T C E L L S

3

tal and fascinating phenomena was achieved. The topics discussed also include intimate mechanisms of the regulation of antibody responses by T cells, the signifbance of these phenomena for the regulatory processes of the immune system, and their possible implication for the pathogenesis of various immunopathological states. II. Specific Cells of the Immune System

The immunocompetent lymphocytes can be divided into two general types on the basis of functional differences: (1) T cells-small lymphocytes that have adapted to certain specific immune functions by virtue of some as yet undefined influence of the thymus (thymus-derived); and ( 2 ) B cells-small lymphocytes that have not been directly influenced by the thymus and which are the progenitors of mature antibody-producing plasma cells. Experimental evidence weighs heavily in favor of the concept that unipotential cells which populate the various hematopoietic tissues of fully developed individuals are derived from common pluripotential stem cells (for review, see 9). Ontogenically, stem cells originate in the embryonic yolk sac and primitive blood islands, migrating later to hematopoietic colonies in fetal liver and bone marrow. Further migration occurs via the bloodstream to various tissues of the hematopoietic system where further differentiation occurs ( 13, 1 4 ) . Differentiation to unipotential progenitor cells of either lymphoid or myeloid lines is signaled by inductive factors, presumably existing in the microenvironment of the different hematopoietic organs of the individual ( 1 5 ) . We shall limit our considerations here to the lymphoid cell lines. Stem cells differentiate into unipotential progenitor lymphoid cells under the microenvironmental influences of the primary lymphoid organs ( 1 4 ) . Avian lymphoid systems have been shown to consist of two distinct primary lymphoid organs-the bursa of Fabricius and the thymus-the influence of which on the differentiation of the stem cells that have migrated to them is clearly distinguishable on the basis of the functional differences of such cells in the immune system (16-20). Surgical extirpation of the bursa from a newly hatched chick results in depression of serum immunoglobulin levels and marked diminution in the capacity to develop humoral antibody responses to antigen stimulation, but has little effect on the ability to reject tissue allografts (20-22); in contrast, early removal of the thymus diminishes the capacity to develop delayed hypersensitivity and impairs allograft rejection (20-25). In mammals, it is now also clearly established that there exists two distinct lymphoid systems responsible for differentiation of immunocompetent cells. One is clearly thymus-influenced, but the other is not. Hence, neonatal thymec-

4

DAVID H. KATZ AND BARUJ BENACERRAF

tomy of rats or mice virtually abrogates their capacity to reject tissue allografts specifically (26-29), and humans with congenital absence of the thymus are markedly impaired in their ability to develop delayed hypersensitivity reactions or manifest allograft immunity ( 30734). Such individuals have normal levels of serum immunoglobulins and can respond to certain antigens with production of humoral antibody. The analog of the bursa of Fabricius in mammals has not been discovered in any discrete lymphoid organ, and it is quite likely that it does not exist as such. In this review, we adhere to popular terminology to refer to the two distinct classes of immunocompetent lymphocytes. (1) The T cells are lymphocytes that have differentiated under the influence of the thymus and are responsible for mediating cellular immune reactions such as delayed hypersensitivity and transplantation reactions. These cells participate in the development of humoral immunity, as will be detailed below, but are not capable of secreting humoral antibodies. Since T cells stimulated by antigen respond, on the one hand, by a clonal expansion and differentiation, and, on the other hand, by being activated to perform their specific function (i.e., helper cells, target cell killers, etc.), we have elected to refer to the former as “educated T cells, and the latter as “activated T cells. As will be discussed in this review, activated T cells may result from stimulation by agents other than specific antigen, ( 2 ) The B cells, are lymphocytes that have differentiated under the influence of the bursa or its analog in mammals and ultimately become the effector cells in humoral immunity by virtue of synthesizing and secreting immunoglobulin antibodies. Ill. Requirement of Two Distinct Lymphoid Cell Types in the Development of Humoral Immune Responses

A. RESPONSETO FOREIGN ERYTHROCYTE AND PROTEINANTIGENS

1. In Vivo The compartmentalization of immunocompetent lymphocytes into T and B classes was recognized as an efficient product of evolution in the development of more sophisticated and complex immune systems in higher animals. In recent years, however, it has been realized and now firmly established that these distinct lymphoid cell lines not only perform different roles in the generation of different forms of immunity but also interact with one another in the development of certain immune responses, notably the humoral immune response to various antigens.

REGULATORY INFLUENCE OF ACTIVATED T CELLS

5

In retrospect, the requirement of T and B cell interaction in eliciting antibody responses is suggested by results obtained from studies in neonatally thymectomized mice 10 years ago (26, 35-37). These studies demonstrated that neonatal thymectomy in mice prevented the normal development of immune responsiveness. The effect was particularly marked with respect to cellular immunity exemplified by the homograft reaction. It was also noted, however, that the ability of such mice to develop antibody responses to certain antigens, such as sheep erythrocytes or foreign serum proteins, was diminished or absent (26, 3 5 ) , but the response to some bacterial and viral antigens was normal. Moreover, the defect of neonatally thymectomized animals could be partially restored to normal by implantation of a thymus graft (26, 35, 3 8 4 0 ) . Furthermore, it was found that lethally irradiated C57B1 mice which were reconstituted with syngeneic bone marrow cells recovered full inimunoconipetence only if the thymus was present (41 ) . The initial understanding of thymic function in the immune system was based on the findings of Osoba and Miller ( 4 2 ) that the immunological competence of neonatally thymectomized mice could be partially restored by implantation of thymus grafts enclosed in Millipore chambers. This observation suggested that the thymus can function as an endocrine organ liberating a humoral factor which can potentiate antibody responses. Arguments against this interpretation have been based on the fact that the level of immunological restoration by chamberenclosed thymus grafts was less than that obtained with nonencapsulated grafts. Moreover, full restoration of peripheral blood lymphocyte counts was not achieved with chamber-enclosed thymus grafts. However, it does not seem likely that enclosing a thymus graft in such chambers would create a favorable milieu for full functional expression. Interest in thymic function soon shifted from considerations of endocrine-like activity to more careful scrutiny of the behavior of cells derived from this organ. Using an irradiation-induced chromosome abnormality as a cytological marker ( T6T6in CBA/H mice), it was found that in neonatally (35, 43) or adult ( 4 0 ) thymectomized mice reconstituted with T6-marked thymus grafts, a small number of the T, thymus cells underwent mitosis in the host spleen. Although the significance of these dividing cells was not at first appreciated ( 4 3 ) ,it was subsequently demonstrated that they were sensitive to antigen-induced mitosis following intraperitoneal administration of sheep erythrocytes ( 4447). Thus, Davies et al. ( 4 5 ) used T6T, thymus grafts to reconstitute thymectomized syngeneic radiation chimeras and demonstrated a well-defined mitotic pattern of response of the thymus cells to either sheep red blood cells (SRBC) or skin homografts.

6

DAVID H. KATZ AND BARUJ BENACERRAF

Another interesting observation which, retrospectively, illustrates the requirement of T and B cell interaction was described by Kennedy et aZ. ( 48). Their experiments were designed to enumerate cells in normal mouse spleen which were specifically committed to recognize and respond to sheep erythrocyte antigens. Lethally irradiated mice were injected intravenously with normal mouse lymphoid cells and SRBC. Eight days later, thin serial slices of recipients’ spleen were cut and embedded in agar. Incubation of the spleen slices with SRBC in the presence of complement at 37°C permitted the development of hemolytic foci or discrete clusters of plaque-forming cells (PFC), the number of which were linearly related to the number of cells in the orginal donor inoculum. The latter point suggested that each cluster of PFC reflected the activity of a single SRBC-sensitive cell in the original inoculum which had lodged in the recipient spleen and proliferated and differentiated in response to antigenic stimulation. The development of such hemolytic foci occurred regularly when normal mouse spleen or lymph node cells were used as the original inoculum, but neither thymus cells alone nor bone marrow cells alone were capable of initiating hemolytic foci in recipient spleens. Several other observations indirectly suggested the requirement of cell interactions in antibody responses. Thus, Celada (49) examined the relation between the transfer secondary response to human serum albumin (HSA) and the number of syngeneic spleen cells transferred and found that over a restricted range of cell numbers the response increased disproportionately to the increase in cell dose. He referred to this phenomenon as the “premium effect.” Later, Gregory and Lajtha (50) compared the numbers of hemolytic foci and PFC developing in spleens of irradiated mice injected with syngeneic spleen cells and SRBC. They found that, although the number of foci was linearly related to number of spleen cells transferred (slope of l), the number of PFC increased disproportionately (slope of 2 ) . Such results can be explained if the development of a hemolytic focus depends upon the presence of only one cell type, whereas the development of antibody-secreting PFC requires an interaction between two cells neither of which is present in overwhelming excess. Bussard and Lurie ( 5 1 ) observed a similar effect in in vitro cultures of peritoneal cells in presence of SRBC. The first direct evidence for T and B cell interaction was provided by Claman and co-workers in studies of the humoral response to SRBC in mice (52). These investigations were based on a very simple design to test for the existence of potentially immunocompetent cells in the adult mouse thymus using the hemolytic focus assay system (48,53). Thus, lethally irradiated mice ( 650-750 R) were injected intravenously with varying numbers of either spleen, thymus, bone marrow, or thymus

REGULATORY INFLUENCE OF ACTIVATED T CELLS

7

plus bone marrow cells from normal or immune syngeneic donors. The cell transfer was followed by antigenic challenge with SRBC. Spleens of recipients were assayed for hemolytic focus activity at various times after cell transfer. The results were strikingly clear in showing the dissociation of immune responsiveness among the various cell populations. Thus, neither normal marrow nor normal or SRBC-immune thymus contained cells which alone could give a hemolysin response. However, when thymus and marrow cells were combined in the same recipient a marked synergy, which was linearly related to number of thymus and marrow cells in the mixture, was observed ( 5 2 ) . The authors postulated from these results that the marrow population contained “effector cells’’ capable of producing antibody, but only in the presence of “auxiliary cells” present in the thymus population. Support for this interpretation was forthcoming from subsequent studies carried out in two independent laboratories. Davies et al. ( 5 4 ) transferred spleen cells from donors, which had been immunized with SRBC 1 month after irradiation and thymus-grafting, to irradiated recipients which had been presensitized against either the thymus-derived cells or the marrow-derived cells. If the donor marrow cells were rejected, the secondary response to SRBC was abolished, whereas rejection of the thymus cells only diminished the response, suggesting that antibody production was by marrow cells and not by thymus cells, although the latter cells made a most vigorous mitotic response to antigen. Moreover, the highest antibody response occurred when both cell populations were allowed to react to antigen. Miller and Mitchell ( 5 5 ) reported that bone marrow of neonatally thymectomized mice was as effective as that of normal mice in restoring immunological responsiveness to heavily irradiated mice with intact thymuses. Mitchell and Miller ( 5 6 ) further showed that in neonatally thymectomized mice given allogeneic thoracic duct or thymus lymphocytes to restore the response to SRBC, the PFC could be inhibited with anti-H-2 sera against host cell antigens but not with anti-H-2 sera against donor thoracic duct or thymus cells. Moreover, in an adoptive transfer system in irradiated hosts in which thymus cells were temporarily separated from bone marrow cells, it was shown that the thymus lymphocytes had to react first with the specific antigen before interaction with bone marrow cells could produce a significant anti-SRBC response ( 5 6 ) . These observations established that the antibody-forming cell precursor is marrow-derived and that thymus cells recognize and react specifically with antigen. In 1968, a series of elegant experiments were published by Miller, Mitchell, and associates from the Walter and Eliza Hall Institute (57-59). Their results provided the first clear elucidation of the occurrence of

8

DAVID H. KATZ AND BARUJ BENACERRAF

specific cell interactions in the humoral immune response to sheep erythrocytes in mice and deserve, therefore, detailed analysis. The first series of experiments (57) examined the capacity of various cell types to restore the yM hemolysin response of neonatally thymectomized CBA mice and these results are summarized in Table I. In the absence of any treatment, such mice gave a very meager response to SRBC, as measured by the number of PFC in their spleens (2300) compared to that of sham-operated littermates (32,000 PFC/spleen). Transfer of 10 million viable syngeneic thymus or thoracic duct cells increased the PFC response of neonatally thymectomized mice to 20,000, and transfer of 50 million viable thymus cells restored their response to normal ( >32,000 PFC). Furthermore, 10 million viable semiallogeneic ( CBA X C57B1)F, thoracic duct or thymus cells or allogeneic C57B1 thoracic duct cells were equivalent to syngeneic cells in restorative capacity. Allogeneic thymus cells could restore but were slightly less effective. In contrast, these authors observed no restoration following transfer of syngeneic bone marrow cells, thymus, or thoracic duct cells which had been exposed to 1000 R in vitro X-irradiation or to thymus extracts. Thoracic duct cells obtained from syngeneic donors made tolerant to SRBC were markedly reduced in their capacity to restore immunocompetence in neonatally thymectomized mice. The most important and crucial denionstration was that the hemolysin response of neonatally thymectomized mice given thymus or thoracic duct cells reflected antibody production by cells of the host. This was accomplished by Miller and Mitchell by employing anti-H-2 isoantisera (57) and confirmed by Nossal et al. (59) using the TBchromosome marker. Thus, in neonatally thymectomized CBA recipients of semiallogeneic F, or allogeneic C57B1 thymus or thoracic duct cells plus SRBC, the anti-SRBC PFC could be specifically reduced by more than 90%following treatment of their spleen cells with anti-CBA serum in the presence of complement but not by treatment with anti-C57B1 serum. Similarly, by the use of neonatally thymectomized CBA mice possessing the T, chromosome marker as hosts for syngeneic thymus or thoracic duct cells, it was possible to demonstrate that all of the PFC in mitosis were of host rather than donor origin (59). In the second series of experiments ( 5 8 ) , a study was made of the capacity of various cell types to restore the yM hemolysin response to SRBC of irradiated or thymectomized and irradiated adult mice. In lethally irradiated mice, a synergistic effect in the response to SRBC was obtained when both syngeneic thoracic duct and.bone marrow cells were used for reconstitutions. In adult thymectomized, lethally irradiated mice protected with syngeneic bone marrow, a similar, although less

TABLE I DIRECTPLAQUE-FORMING CELLSPRODUCED IN SPLEENSOF NEONATALLY THYMECTOMIZED MICE AFTER INJECTION OF SHEEPERYTHROCYTES AND THYMUS OR THORACIC DUCTLYMPHOCYTES FROM SYNGENEIC, SEMIALLOGENEIC OR ALLOGENEIC DONOR SO^^ Inhibition of P F C by anti-H-2 serum (yo)

Avg. No. P F C per spleen ( fSE) Anti-C6,Bl

Cells inoculated A. Sham-thymectomized recipients None SRBC B. Thymectomized recipients SRBC SRBC’ 50 x 106 CBA thymocytes 10 x 106 CBA thymocytes SRBC 10 x 106 CBA TDL SRBC 10 x 108 F1 thymocytes SRBC 10 X 106 F1 T D L SRBC 50 X 106 C57B1 thymocytes SRBC 10 X 108 C57B1 TDL 50 x 106 X-irradiated CBA SRBC thymocytes 10 x 106 X-irradiated CBA TDL SRBC 10 x 106 X-irradiated F1 TDL SRBC Thymus extract SRBC 10 x 106 T D L from control cyclophosphamide-treated CBA donors SRBC 10 X 106 TDL from SRBC-tolerant CBA donors (treated with SRBC and cyclophosphamide)

+

+

+ +

+

+

+

+

+

+

+

~~~~

~

Anti-CBA

123 f 29 32,177 f 3550 2,356 f 537 38,855 f 7448 19,160 f 3840 20,254 f 3646 22,380 f 6285 24,537 f 5519 17,448 f 3907 30,120 f 7791 1,160 f 615

-

-

6 0-17 8-12

97 89-100 90-92

0-1

86-96

-

-

3,776 f 979 4,344 f 2540 1,802 f 1155 41,600 f 13,436 7,331 f 2276

~

We thank Dr. J. F. A. P. Miller for permission to reproduce some pertinent data from Miller and Mitchell (67). b This table shows the antibody response to sheep erythrocytes (SRBC) of neonatally thymectomized CBA mice reconstituted intravenously with varying numbers of syngeneic, semiallogeneic (CBA X C57Bl)E’I, or allogeneic C57Bl thymus or thoracic duct lymphocytes (TDL). Results obtained in sham-thymectomized controls are included for comparison. The data are expressed as the average number of plaque-forming cells (PFC) per spleen detected on days 4-5 after inoculation with donor cells and SRBC. Recipient groups ranged from 4 to 36 mice each. The data obtained in studies of the source of P F C by incubation of spleen cells with anti-H-2 serum plus complement are included in the far right columns expressed as percent inhibition. X-Irradiated donor cells were exposed to 1000 R in vitro immediately prior to transfer. Thymus extract equivalent to 100 x lo6 thymus cells per recipient WBS injected intravenously with SRBC. Mice were rendered tolerant to SRBC by treatment with cyclophosphamide and SRBC and used BS donors of TDL 23-26 days after completion of tolerance induction. a

10

DAVID H. KATZ AND BARUJ BENACERRAF

marked, synergism was observed in the response to SRBC following transfer of F, thoracic duct cells provided the latter were not administered until 2 weeks after the syngeneic bone marrow cells. Employing anti-H-2 sera (58) or the T, chromosome marker ( 5 9 ) , it was firmly established that the source of the PFC in this system was the bone marrow lymphocyte population. Such studies made it clear, therefore, that the unicellular concept of antigen recognition and antibody formation did not reflect the true state of affairs, at least with respect to the sheep erythrocyte antigen. Thus, at least two functionally distinct but morphologically indistinguishable lymphoid cells apparently must engage in some interplay to effect in wivo antibody production. One cell, derived from the thymus and also present in thoracic duct lymph, thereby placing it in the class of lymphocytes comprising the recirculating lymphocyte pool ( 60), is stimulated by antigen to undergo mitosis (44-47) but is incapable of secreting antibodies (48, 52, 54, 5 6 5 9 ) ; the other cell, derived from bone marrow, is the precursor of the antibody-secreting cell but requires an interaction of some sort with the thymus-derived cell before it can perform its function (9-11,52, 5 7 5 9 ) . The development of these new concepts concerning the immune system not only stimulated a reappraisal of some earlier phenomena of immunology and their explanations, but also resulted in a virtual avalanche of experiments designed to probe the nature of cell interactions and the cell types involved, and their significance in normal and abnormal immunc reactions. Indeed, since the time of the initial studies demonstrating the requirement for cell interactions in the sheep erythrocyte system, analogous interactions have been demonstrated in the in vivo humoral response to protein antigens (61, 62) and to haptencarrier conjugates (63-74), which will be reviewed in detail in Section II1,B. Other studies illustrated some of the properties of the interacting T and B cells. First, Claman and co-workers (75, 76) showed that, in their single transfer system in lethally irradiated recipient mice, only viable syngeneic thymus cells could interact synergistically with marrow cells in the response to SRBC. Hence, no cooperative interaction was obtained with sonicated or minced thymus cells. Second, irradiated thymus cells or thymus cells from xenogeneic or semiallogeneic donors did not collaborate with marrow cells. Third, the T cells had specificity; using a double transfer system similar to that used earlier by Mitchell and Miller (56), Claman and Chaperon ( 1 1 ) attempted to dissect further their model of synergism. The experimental design consisted of transferring thymus cells with or without SRBC to a lethally irradiated first host and after 6 to 8 days transferring the spleen cells of this first

REGULATORY INFLUENCE OF ACTIVATED T CELLS

11

host to a second irradiated syngeneic recipient together with syngeneic marrow cells. Six days later the spleen of the second recipient were assayed for y M anti-SRBC PFC. They found that thymus-marrow synergy in the second host occurred only if the original thymus cell inoculum was exposed to SRBC antigen in the first host for more than 2 days, confirming the earlier finding of Mitchell and Miller (56). Moreover, exposure of the transferred thymus cells to lOOOR X-irradiation in situ, immediately prior to transfer to the second host, abrogated their capacity to cooperate synergistically with marrow cells, thus confirming earlier observations from their laboratory (77) in their single transfer system. Shortly after the initial studies with erythrocyte antigens were reported, analogous observations were made in responses of mice to protein antigens. Thus, Taylor ( 6 1 ) and Chiller et d. (62) demonstrated synergism between thymus and bone marrow lymphocytes in antibody responses to bovine serum albumin (BSA) and human y-globulin (HGG), respectively. Similar findings have been reported by Miller and Sprent (78) in the response of mice to fowl 7-globulin ( FyG). A crucial question naturally is raised in the light of two-cell models for hunioral immunity, namely, What relevance do these required interactions have in the context of Burnet’s dogma of clonal selection? Now we are faced with the problem of not just one cell type precommitted in its antigen specificity, but at least two. In this sense, a serious threat to any theory of clonal selection is readily seen when some thought is given to the probability of two functionally distinct cells with the same specificity characteristics meeting at random to carry out a necessary interaction. The cells involved are functionally distinct and may be distinct in their determinant specificities as well; this latter point is exemplified by models of cooperative cell interactions in hapten-carrier systems to be discussed. The distinction in determinant specificities, we feel, probably exists in the erythrocyte antigen system also but the very nature of the antigen makes this difficult to appreciate. There is, nevertheless, a body of evidence in the SRBC system which supports clonal selection in a two-cell model. Shearer et al. (79, 80) transferred unprimed spleen cells as precursors of immunocompetent cells with SRBC to irradiated syngeneic recipients and measured the development of three types of antibody-producing cells-IgM and IgG hemolysin ( PFC ) , and hemagglutinin-producing cluster-forming cells ( CFC )-in the recipients’ spleens. By employing graded numbers of transferred donor spleen cells, they could limit to one or a few the number of precursor cells or antigen-sensitive units ( ASU) reaching the recipient spleen.

12

DAVID H. KATZ AND BARU J BENACERRAF

By applying this model to thymus-marrow cell interactions, Shearer and Cudkowicz (81) demonstrated that when graded numbers of bone marrow cells ( spleen cells of bone marrow-reconstituted X-irradiated mice ) were transferred to an irradiated recipient with constant numbers of thymus cells (from intact donors) plus SRBC, the production of IgM or IgG PFC or CFC (agglutinins) varied independently of each other but in relation to the number of grafted marrow cells. This suggests that there is specialization with respect to class and type of antibody produced within the precursors in the marrow population prior to antigenic stimulation. The question of whether thymus cells restrict the potential antibody class of ASUs was investigated by transferring thymocytes in limiting dilutions with constant numbers of marrow cells (plus SRBC) (82). Under these conditions, the frequency of formation of IgM and IgG PFC were not independent, thus indicating that the thymic-reactive cells were not themselves specialized nor did they determine directly the molecular class of antibody produced after interaction with marrow cells. This is consistent with our own findings (66) and of Mitchison et al. (65) in the hapten-carrier models. 2. In Vitro The development of methodology by Mishell and Dutton (83, 8 4 ) and by Marbrook (85) to obtain in vitro immune responses has offered another highly useful approach to the study of specific cell interactions in responses to erythrocyte antigens. Mosier and Coppleson ( 86) separated mouse spleen cells into nonadherent and adherent ( macrophage ) populations and cultured serial dilutions of one population in the presence of an excess of the other. The order of cell interactions required to produce the primary anti-SRBC response was predicted from regression line slopes derived by plotting the log of the limiting cell dose against the log of the antibody-forming cell response. They suggested from such data that one adherent and two nonadherent cells must interact for the immune response to develop and that the likely candidates for the two interacting nonadherent cells are T and B cells. The latter point was confirmed in subsequent studies using spleen cells from thymectomized mice ( 87). Hence, spleen cells from adult thymectomized, lethally irradiated mice protected with syngeneic bone marrow gave no primary in vitro response to SRBC. However, when such mice were reconstituted with a thymus graft, the in vitro response was normal. The adherent cells of spleen of thymus-deprived mice were comparable to those of thymusreconstituted mice in supporting the in vitro response of nonadherent spleen cells from normal mice. However, nonadherent spleen cells of thymus-deprived mice, were unable to respond in the presence of adherent normal spleen cells, whereas nonadherent cells from thymus-

REGULATORY INFLUENCE OF ACTIVATED T CELLS

13

reconstituted mice could respond. These observations clearly placed the cellular deficit in thymectomized mice in the nonadherent populations of lymphocytes but did not firmly establish the cell type involved. Working independently at the same time, Munro and Hunter (88) restored to normal the in vitro primary response to SRBC of spleen cells from adult thymectomized, irradiated, bone marrow-protected CBA mice by the addition of small numbers of normal BALB/C spleen cells. Using anti-H-2 sera, they showed that it was, indeed, the CBA cells, not the BALB/ C cells, which produced the anti-SRBC antibody. Irradiation of the BALB/C donor with 1000 R immediately before sacrifice diminished, but did not abrogate, the capacity of such cells to restore the response of the thymus-deprived CBA cells. It was not clear, however, whether or not this reflected radiation resistance in the macrophage population which was a documented property of such cells (89). Hirst and Dutton (90) restored the primary in vitro anti-SRBC response of spleen cells from neonatally thymectomized mice with subpopulations of spleen cells from normal mice. They showed that small numbers of nonadherent normal spleen cells irradiated with 1100 R in vitro restored or enhanced the in vitro response and that this enhancement was most marked when the nonadherent cells came from allogeneic donors. Furthermore, the primary in vitro response of spleen cells from normal mice could be enhanced by the addition of small numbers of irradiated allogeneic nonadherent cells. This observation extended to the SRBC system the finding of Katz et al. (68) that the “helper” activity of the T cell in immune responses to hapten-carrier conjugates is radiation-resistant in contrast to the earlier conclusions of Clainan and co-workers ( 1 1 , 77) and Miller and Mitchell (9, 57). The studies cited above clearly infer the required role of the T cell in the primary in vitro anti-SRBC response. More direct evidence, however, comes from studies in which thymus cells themselves are employed to restore the response or when specific elimination of such cells by treating a normal spleen cell population with anti4 serum and complement abrogates the response. Despite the many investigations, the first approach has not been as easy to accomplish as one might expect (88, 90-94). Indeed, there is only one report by Doria et al. (95) in which spleen cells from neonatally or adult thymectomized, irradiated bone marrow chimeras could be restored by the addition of normal syngeneic thymus cells to respond to SRBC in vitro. Recently, Schimpl and Wecker ( 96) have reported that thymocytes from cortisone-treated syngeneic donors could restore the primary in uitro anti-SRBC response of anti-6’ serum-treated mouse spleen cells. In general, however, thymus cells must first be injected into irradiated recipients with SRBC and then recovered from the spleens of such hosts before they will optimally re-

14

DAVID H. KATZ AND BARUJ BENACERRAF

store the in vitro response (91, 92). This is exemplified by Hartmann (91 ), who studied the in vitro response of bone marrow-derived cells, obtained from spleens of thymectomized, irradiated recipients %4 weeks after injection of syngeneic bone marrow. Such cells, which alone were unable to develop an in vitro primary response could be restored to respond by addition of thymus-derived cells obtained from spleens of irradiated recipients 8 days after transfer of thymocytes. The T cells had to be “educated,” (that is, the SRBC antigen had to be given to the irradiated host) in order to cooperate with the B cells in culture. Uneducated T cells did not support the response, a finding which corroborates the observations of Mitchell and Miller ( 5 6 ) and Claman and Chaperon (11) in their in vivo systems. When synergy was obtained with allogeneic mixtures of B cells and educated T cells, use of anti-H-2 sera showed that the PFC were derived from the B cell population. Moreover, the synergy was dependent on specific T cell education, i.e., SRBC-educated T cells cooperated with B cells in response to SRBC but not to horse red blood cells (HRBC) used as the in vitro immunogen. The second approach, employing the effects of treatment with antiserum and complement on the capacity of spleen cells to respond in uitro, has been studied by several investigators. Schimpl and Wecker (93) demonstrated that treatment of spleen cells from normal mice with anti-9 antibodies in the presence of complement regularly reduced the primary in vitro anti-SRBC response of such cells by 80%. Chan et aZ. (92) found that treatment of spleen cells from normal or SRBC-primed mice with anti-9 serum plus complement abrogated the primary yM and secondary yG in uitro responses, respectively, These responses could be restored by adding, to anti-&treated cells, educated T cells from spleen of irradiated thymus-infused syngeneic or semisyngeneic donors. In the latter case, the precursors of PFC were shown by anti-H-2 antisera to derive from the anti-&treated cell population. The educated T cells were likewise shown to be sensitive to anti4 antibody and complement. The major points to come out of such studies on the in vitro antiSRBC immune response can be summarized, therefore, as follows: (1) the requirement for cell interactions between T and B cells is mirrored in both in vivo and in vitro systems; and ( 2 ) the T cell component appears to become optimally functional after it has peripheralized to secondary lymphoid organs and undergone specific antigenic stimulation.

EFFECT” AND COOPERATIVE B. “CARRIER INTERACTIONS SPECIFICFOR DIFFERENT DETEFLMINANTS ON THE SAMEANTIGEN The introduction of defined haptenic determinants onto immunogenic carriers by Landsteiner ( 1 2 ) has provided a powerful tool for the analysis

REGULATORY INFLUENCE OF ACTIVATED T CELLS

15

of specific interactions between antigens and specific cells of the immune system. Considerable evidence has been obtained that cellular immune reactions to hapten-protein conjugates [delayed sensitivity (97-101 ), stimulation of deoxyribonucleic acid (DNA) synthesis by antigen (102, 103), and hapten-specific secondary responses (104106)ldisplay a significant although variable degree of carrier specificity. Such carrier specificity of hapten-specific cellular reactions was interpreted to reflect the partial specificity of the antigen-binding receptors of specific cells for the carrier molecule (107-111), exceeding the real but modest contribution in energetic terms of the carrier to the specificity of antihapten humoral antibodies (111-113). This interpretation of carrier function, however, is not able to explain several essential characteristics of hapten-specific humoral immune responses: 1. Hapten conjugates of immunogenic molecules are required to elicit strong antihapten antibody responses; nonimmunogenic substances serve only poorly, or not at all, as carriers for haptens (114116). 2. Optimal hapten-specific secondary responses require challenge with the hapten-carrier conjugate used for primary immunization (104). 3. Induction of immunological unresponsiveness to the carrier molecule results in partial or total suppression of the responses to haptens on the tolerated proteins (117-122). Assuming that the specificity of serum antibody accurately expresses the specificity of the antigen-binding receptor molecules on the precursors of antibody-forming cells, then these findings suggest the operation of an additional recognition mechanism for the carrier molecule, This interpretation finds validity in recent demonstrations that cooperative interactions between carrier-specific and hapten-specific lymphoid cells are essential for the development of antihapten immune responses. Two basic in vivo experimental models have been employed to establish this latter point: (1) the adoptive secondary antihapten response following transfer of hapten-primed and carrier-primed cells into irradiated recipient mice; and ( 2 ) the use of preimmunization or supplemental immunization with free carrier to enhance primary and secondary antihapten antibody responses in guinea pigs and rabbits. The phenomenon has also recently been investigated in in vitro systems as will be discussed.

1. In Vim Models of Cooperative Interactions in Responses to Hapten-Carrier Conjugates Even before direct evidence for cooperative interactions in immune responses to hapten carriers was obtained, there were several observa-

16

DAVID H. KATZ AND BARUJ BENACERRAF

tions which clearly, in retrospect, hinted that this be the case. These observations derived from genetic differences among certain animals in their capacity to develop immune responses to distinct antigenic determinants. Thus, one strain of inbred guinea pigs, strain 2, is genetically capable of responding to poly-L-lysine (PLL), whereas strain 13 is not ( 123). These latter “nonresponders’’ also fail to make an anti-2,4-dinitrophenyl (DNP) humoral response to D’NP-PLL under normal circumstances, However, when DNP-PLL is electrostatically complexed to methylated BSA, the nonresponder animals develop hapten-specific antibody responses, provided they have not previously been made tolerant to BSA. The second example is the response of rabbits to the tetrameric isoenzymes of lactic dehydrogenase ( LDH) (120). Some rabbits respond perfectly well to both Type I and Type V isoenzymes, whereas others respond poorly to Type I. If a hybrid molecule consisting of both Type I and V subunits is used to immunize the poor responder rabbits, they develop antibodies specific for both subunit types, suggesting that one subunit has served as a carrier for the other to which they are normally unresponsive. Both of these genetic models clearly indicated that recognition of both hapten and carrier determinants must occur before an antihapten response will develop. A somewhat analogous interpretation can explain the observation of Schierman and McBride (119) in which chickens developed enhanced primary antibody responses to weak erythrocyte antigens when highly immunogenic isoantigens were also present on the same erythrocyte. It was Mitchison ( 6 3 ) , however, who, several years ago, obtained the first direct evidence for cooperative participation of two cells with distinct determinant specificities in the humoral response to haptencarrier conjugates. By employing an adoptive transfer system in irradiated recipient mice, he made the following observations. Spleen cells from syngeneic donor mice, which had been immunized with 4-hydroxy5-iodo-3-nitrophenacetyl ( NIP )-ovalbumin ( OVA ) , injected into irradiated recipients made a secondary anti-NIP response following challenge with the homologous conjugate, NIP-OVA, but, as expected in the light of the earlier findings of Ovary and Benacerraf (104), not to a heterologous conjugate, NIP-BSA. However, when spleen cells from donors immunized with NIP-OVA were injected together with spleen cells from donors immunized with BSA, a perfectly good secondary response was made to the heterologous conjugate NIP-BSA. Hence, addition of cells specific for the heterologous carrier, BSA, permits the haptenprimed cells to make a secondary anti-NIP response to NIP-BSA. This basic in vivo observation has been confirmed by many investigators (65, 69, 70, 74, 124-127). This experiment demonstrates that in antihapten antibody responses

17

RECULATORY INFLUENCE OF ACTIVATED T CELLS

an interaction of carrier-specific cells with the hapten-carrier conjugate is required for maximal stimulation of the precursors of antihapten antibody-producing cells. The analogy to thymus-marrow cells cooperation in the response to sheep erythrocytes in mice is obvious and is now firmly established. Raff ( 1 2 6 ) showed that the carrier-specific cooperating cells, or “helper” cells are, indeed, thymus-derived, whereas the antihapten antibody-forming cell precursors are not. The results of this experiment are summarized in Table 11. By using the adoptive transfer system in mice, he found that treatment of spleen cells from donors primed to the second carrier (BSA) with anti-6’ antiserum and complement abrogated the capacity of such cells to cooperate with haptenprimed [NIP-chicken y-globulin (CGG)] spleen cells in the adoptive secondary response to NIP-BSA. Treatment of the NIP-CGG-primed spleen cells with anti-0 antibody, however, did not affect the capacity TABLE I1 EFFECT OF ANTI-BSERAO N CARRIER-PHIMED (BSA)

A N D HAPTEN-PRIMED SECONDARY (NIP-CGG) SPLEENCELLSIN COOPERATIVE ANTIHAPTEN RESPONSETO NIP-BSAasb

Boost NIPwith CGGBSANIPprimed (In vitro primed (Invitro BSA cells treatment) cells treatment) (100 pg.)

++ 0 + + +

+

+

+0 + +

0

0 0 0 0

+

+

0

+

Anti-8 GPC NMS+ GPC

+

0 0 0 0 Anti-8 GPC NMS+ GPC

+

0

+

0 ~~

+0 + ++

+ + +

Anti-NIP response on day 10 Exp. 1

Exp. 2

*

-0.91 f 0.20 -0.78 0.32 0.45 f 0.25 0.27 f 0.48 -0.77 f 0.33 Not done Notdone 1.0 f 0.15 -0.27 f 0.54 0.7 f 0.29

* 0.25

1.05 f 0.42

1.79

0.99 f 0.43

1.78 f 0.18

1.08 f 0.16

1.92 f 0.12

~~~~

~

We thank Dr. Raff for permission to use these data from Ref. (126).BSA, bovine serum albumin; NIP, 4-hydroxy-5-iodo-3-nitrophenacetyl; CGG, chicken 7-globulin. * The protocol of these experiments consists of adoptive intraperitoneal transfer of spleen cells from mice primed with NIP-CGG either alone or mixed with spleen cells from mice primed with BSA in irradiated syngeneic recipients. The respective cell populations were exposed prior to transfer to in uitro treatment with either: ( 1 ) nothing, ( 2 ) anti4 serum plus guinea pig complement (GPC), or ( 3 )normal mouse serum (NMS) plus GPC. Secondary boost with 100 pg. of NIP-BSA was administered intraperitoneally 1 day after cell transfer, and recipients were bled 10 days later. The anti-NIP antibody response is expressed as the loglo molar binding capacity ( x 10-8 M ) .

18

DAVID H. KATZ AND BARUJ BENACERRAF

of such cells to produce anti-NIP antibodies when transferred together with BSA-primed spleen cells and challenged with NIP-BSA ( 126). The same cooperation phenomenon between carrier-specific and hapten-specific cells has been demonstrated in guinea pigs and rabbits by immunization with free carrier. Rajewsky et al. ( 6 4 ) showed that rabbits immunized with a p-azobenzenesulfonic acid ( sulfanil) derivative of BSA made significant secondary antisulfanil antibody responses to sulfanil-HGG if they had received a supplemental intervening immunization with the free unconjugated carrier HGG. These observations have been confirmed and extended in our own laboratory (66, 67) in both rabbits and inbred guinea pigs. Guinea pigs or rabbits which have been primed with DNP-OVA fail to respond to a secondary immunization with a heterologous conjugate, DNP-BGG. However, if DNP-OVAprimed animals receive an intervening supplemental immunization with unconjugated BGG they not only develop secondary anti-DNP antibody responses to DNP-BGG, but the magnitude of such responses may be significantly greater than those elicited by secondary challenge with the original immunizing conjugate DNP-OVA (Fig. 1). Moreover, this phenomenon is not restricted to secondary responses (66). Under appropriate conditions of dose and timing, guinea pigs or rabbits that have been preimmunized with free carrier BGG manifest enhanced primary anti-DNP antibody responses following primary immunization with DNP-BGG. The kinetics as well as the magnitude of anti-DNP antibody production are sharply augmented under such conditions. Additional examples of augmentation of a response to a given antigenic determinant as a result of a concomitant, or prior, immune response to another determinant on the same antigen have been described (97,119,124,128,129). The failure of several other investigators (130-133) to obtain evidence of enhanced primary antihapten responses as a consequence of carrier preimmunization most likely reflects the conditions employed for immunization. Such studies clearly establish the operation of distinct recognition units for carrier and haptenic determinants. The carrier recognition unit is clearly not classic serum antibody. Thus, the capacity to enhance primary or secondary anti-DNP antibody responses in guinea pigs and rabbits by carrier preimmunization or supplemental immunization could not be supplanted by the intravenous infusion of small or large quantities of homologous anticarrier serum either of low or high affinity (66). Similar findings have been made in the adoptive transfer system of Mitchison in mice (65, 7 0 ) and in the supplemental carrier immunization model of Rajewsky in rabbits (65, 134). Moreover, such observations are in accord with the findings of Mitchison (70) and Kontiainen (124)

REGULATORY INFLUENCE OF ACTIVATED T CELLS

STRAIN 2 GUINEA PIGS

700 -

a

19

600 -

W

E

n

z a

2 400 z

0

a I-

W

5

8 0

4200 3

a

DNP-OVA BGG 5 0 p g

+ E

* >

DNP-BGG

\

m

0

0

I

1

50F-0:r O

I

I

I

,

I

,

DNP-OVA ;~

0 1 2 3 4 WEEKS AFTER PRIMARY

/ "'

0 4 7 DAYS AFTER 2* CHALLENGE

FIG. 1. Enhancement of hapten-specific anamnestic responses by carrier preimmunization in guinea pigs. Primary immunization with 3.0 mg of 2,4-dinitrophenyl ( DNP ) ?-ovalbumin ( OVA), administered intraperitoneally in saline, was performed at week 0. One week later supplemental immunization with either 50 pg. of bovine y-globulin (BGG) emulsified in CFA or with a saline-CFA emulsion was carried out. Four weeks after primary immunization, the animals were challenged with 1.0 mg. of either DNPZS-BGG or DNPrOVA in saline. Serum anti-DNP antibody concentration just prior to challenge and on days 4 and 7 are illustrated. The numbers in parentheses refer to the numbers of animals in the given groups. The lowermost panel illustrates the normal secondary response of DNP-OVA-primed animals to DNP-OVA challenge; the middle panel shows the absence of a secondary response to DNP-BGG in DNP-OVA-primed animals and furthermore demonstrates the failure of transfused anti-BGG serum to stimulate a response. The uppermost panel presents the enhancement of the secondary response to DNP-BGG in DNP-OVA-primed animals which have been supplementally immunized with BGG. [These data from our laboratory appeared in 1. Exp. Med. 132, 261, 1970 (reference 66).]

20

DAVID H. KATZ AND BARU J BENACERRAF

CFA immune ce/ls 0-- 4 BGG immune ce/ls

l.0-/.6x/O9 LYMPH NODE CELLS

1

DNP-OVA

0

,

I

IDNP+;

I 1 2 3 4 WEEKS AFTER PRIMARY

I

I I I

I

I

I

I

I

I

I

I

I

I

I ( 4 7 DAYS AFTER SECONDARY

0

FIG.2. Ability of bovine y-globulin (BGG)-specific lymphoid cells to prepare 2,4-dinitrophenyl ( DNP)-ovalbumin ( OVA) -immunized guinea pigs for an enhanced response to DNP-BGG. Primary immunization of recipients was performed at week 0 with 3.0 mg. of DNPT-OVA administered intraperitoneally in saline. Three weeks later the guinea pigs were transfused with 1.0-1.6 x los lymphoid cells from syngeneic donors which had been immunized with either 50 pg. of BGG in CFA or saline in CFA 3 weeks earlier. Six days after cell transfer, the recipients were boosted with 1.0 mg. of DNP2,BGG in saline. Serum anti-DNP antibody concentration just prior to challenge and on days 4 and 7 are illustrated. [These data from our laboratory appeared in J. E q . Med. 132, 283, 1970 (reference 67) .]

that the cooperative function of carrier-primed spleen cells reaches a maximum earlier after immunization than their capacity to produce carrier-specific humoral antibodies. Contrasting with the inability of antibodies to do so, lymphoid cells from syngeneic animals immunized to the second heterologous carrier will transfer enhanced antihapten responsiveness to recipient animals which themselves have not been exposed to supplemental carrier immunization. This has been demonstrated by the studies of Mitchison (63, 65, 7 0 ) in mice and of Paul et al. (67) in inbred guinea pigs, as shown in Fig. 2. 2. In Vitro Models of Cooperatiw Interactions in Antibody Responses to HaptenXarrier Conjugates The development of a system for obtaining immune responses in oitro (83, 8 4 ) has offered certain advantages over in vioo experiments in

REGULATORY INFLUENCE OF ACTIVATED T CELLS

21

the analysis of cellular events in the immune response. Since such analyses are less restricted with studies employing well-defined antigenic determinants in the form of hapten-carrier complexes, it was a logical extension of the system to develop methodology to obtain in vitro immune responses to such immunogens. This was greatly facilitated by the discovery of Rittenberg and Pratt ( 1 3 5 ) of a rapid and simple method for coupling the chemical hapten trinitrophenyl ( TNP) directly to erythrocyte membranes. Based on this method with some original modifications, Kettman and Dutton (136) were successful in obtaining a TNPspecific, primary, in vitro response of mouse spleen cells cultured with TNP-coupled erythrocytes. Shortly thereafter, investigators in three different laboratories, working independently, presented evidence for cooperation between carrierspecific and hapten-specific cells in the in vitro primary immune response. Katz et al. ( 1 3 7 ) found that spleen cells from mice which had been primed several days earlier with free carrier (either burro erythrocytes or ax174 phage) developed enhanced primary anti-TNP responses in vitro when cultured with a TNP conjugate of the carrier used for priming (Table 111). Essentially identical observations in the primary in vitro response to NIP-coupled erythrocytes were reported at the same time by Trowbridge et al. ( 1 3 8 ) . The phenomenon was studied in considerably greater detail by Dutton and his colleagues (139, 140) who also obtained enhanced primary in vitro anti-TNP antibody responses in cultures of carrier-primed spleen cells. In addition, they demonstrated that the primary anti-TNP response of normal spleen cells could be enhanced by the addition of carrier-primed spleen cells which had been exposed to in vitro X-irradiation ( 139, 140). The radioresistant cell functioning in this helper capacity is, indeed, a thymus-derived cell as shown by the following observations (139): ( 1) the activity of the irradiated, carrier-primed spleen cells can be abolished by treatment with anti-8 serum and complement, and anti-8 serum-treated normal spleen cells can be restored by the addition of irradiated, carrier-primed spleen cells; ( 2 ) the helper effect of irradiated, carrier-primed spleen cells can be replaced by thymus-derived cells obtained from spleens of irradiated mice which had been injected with thymocytes plus SRBC. By employing a different in vitro culture system, based on a Millipore filter well technique for spleen organ fragments (141 ), Kunin et al. ( 142) corroborated these findings. The phenomenon of cooperative cell interactions in the secondary response to hapten-carrier conjugates in vitro has been less well studied. One such study has been recently reported by Cheers et aE. (143) and confirms the essential features of the phenomenon which occurs in vivo.

22

DAVID H. KATZ AND BARUJ BENACERRAF

TABLE 111 AUGMENTED in Vitro PRIMARY ANTITRINITROPHENYL RESPONSESOF @ X ~ ~ ~ - P R I M E D SPLEENCELLSSTIMULATED WITH TRINITROPHENYL-CONJUGATED ax1740 Direct PFC/IOBrecovered spleen cells Anti-TNP SRBCd

Protocol An tigenb TNP4X174 (1 x 108) ax174 (1 x 109

BALB/c spleen cellsc Normal ax174 7-day primed Normal ax174 7-day primed

Total

Above unstimulated

555 1163

25 1 876

304 287

-

-

Data taken from Katz et al. (137). Numbers in parentheses refer to quantity in plaque-forming units of trinitrophenyl (TNP)-substituted phage added to each culture. Cell density, 10 x 106 cells/ml. Primed mice received 2 X looplaque-forming units of +X174 intraperitoneally 7 days before spleen cells were cultured. The number of anti-TNP plaque-forming cells (PFC) presented have been corrected for background plaques against unconjugat,ed sheep red blood cells (SRBC). The column headed “above unstimulated” represents the number of anti-TNP PFC attributable to in vitro antigen stimulation (Le., total PFC in hapten-stimulated cultures less total PFC in hapten-unstimulated cultures of the same cells).

The hapten employed was 3,5-dinitro-4-hydroxyphenylacetic acid ( NNP ) coupled to either FyG or OVA as carriers. The secondary in vitro antiNNP response of spleen cells from mice primed with NNP carrier was optimal when the homologous NNP carrier was used for the in vitro immunogen. However, cells from an NNP-OVA-primed donor would respond in vitro to NNP-FyG by adding to the culture spleen cells from other mice primed with the carrier FyG. When semiallogeneic carrierprimed cells were used to enhance the secondary anti-NNP response it was shown by anti-H-2 serum that the majority of anti-NNP PFC were derived from the hapten-primed cell population (not the carrier-primed cells). Furthermore, activated thymus cells obtained from spleens of irradiated mice injected with thymus cells plus FyG could substitute for whole spleen cells from FyG-primed mice in enhancing the in uitm secondary anti-NNP response of NNP-OVA primed cells. The effect was specific since thymus cells activated to FyG enhanced the response to NNP-FyG, whereas thymus cells activated to BSA did not.

REGULATORY INFLUENCE OF ACTIVATED T CELLS

23

IV. Nature of the Regulatory Influence of Activated T Cells on Antibody Responses by B Cells

In an assessment of the physiological significance of T cell activity in the regulation of antibody responses by B cells, an important fact must be recognized initially: T cell activity is not an absolute requirement for the induction of antibody synthesis. There is a class of antigen which is considered to be thymus-independent. Moreover, even thymusdependent antigens in a narrow concentration range can elicit low level antibody synthesis, generally confined to the IgM class in the absence of T cell activity. It seems, therefore, appropriate to view the role of activated cells in antibody responses to be, in large part, regulatory in character in many aspects of B cell function. The evolutionary advantages of such a mechanism are readily apparent. Since the two lymphocyte types are both specific and need to be activated by antigen for antibody synthesis by B cells to develop opitmally, important control mechanisms, concerning antigen recognition and tolerance, should have evolved primarily in the T cell population. This is consistent with the body of evidence identifying T cells as the site where specific histocompatibility linked Ir genes are expressed, immunogenicity is recognized, and tolerance most readily induced. In this section, we shall discuss the evidence for the general regulatory role of T cell activity on various aspects of the antibody response. A major consequence of T cell activity, as will be shown, is the facilitation of the selective pressure exerted by antigen on the proliferation and differentiation of B lymphocytes and, thereby, on the emergence of adequate populations of memory cells. In addition, the most dramatic effects of T cell activity on antibody production by B cells concern the synthesis of immunoglobulin other than IgM; in fact, there is increasing evidence from experiments in genetically controlled systems that the switch from IgM to IgG is largely dependent on the regulatory effect of appropriately activated T cells. A. STIMULATION OF B CELLS IN

ABSENCEOF T-CELLREGULATION It is now well established that optimal specific humoral immune responses to certain antigens can develop in the absence of T cell participation. Humphrey et ul. (144) some years ago showed that neonatally thymectomized mice were capable of developing normal antibody responses to pneumococcal polysaccharide-an observation confirmed recently by others (145,146). Armstrong et ul. ( 147) found that bone marrow-reconstituted, lethally irradiated mice could develop normal humoral THE

24

DAVID H. KATZ AND BARU J BENACERRAF

responses to purified polymerized flagellin (POL) of SuZmoneZZu u&Zui& without the addition of thymic lymphocytes. (Indeed, the addition of thymocytes in this situation depressed the response to POL, a finding not explained by the authors that we shall consider in a later section.) Similar T cell independence has been observed with other antigens such as Escherichiu coZi polysaccharide ( 148, 149), polyvinylpyrrolidone (PVP) (149), and MS2 phage (150). Two points of considerable importance with respect to thymus-independent antigens are ( I ) the nature of their physicochemical structure and ( 2 ) the nature of the classrestricted antibody response that they induce, i.e., predominantly yM antibody responses. These features are discussed below. An important feature shared in common by the antigens discussed above is their structure which consists of identical units arranged in a more-or-less linear repetitive sequence. Their unique three-dimensional structural characteristics appear to favor a positive immunogenic signal upon direct interaction with specific receptor sites on B cells. This is to be contrasted with molecules not possessing these structural features which, in the absence of T cell function, generally result in either transient stimulation, no stimulation, or even a negative tolerogenic signal upon direct interaction with B cell receptors. The latter point will be developed in more detail in the later section on tolerance. The question is raised, then, about the physicochemical properties of receptor molecules on B cells which create this marked distinction between various antigenic structures insofar as their capacity to trigger indirectly or directly the specific immune response of these cells. There is no hard evidence from which to draw conclusions concerning this question. However, it is clear that a crucial relationship exists between the structural presentation of antigen and the ability to trigger B cells. This is exemplified by recent studies of Feldmann and Basten (151) . These investigators studied the primary in vitro antibody response of spleen cells from thymus-deprived mice to antigens of various physical forms and showed that the ability of such cells to develop a primary anti-DNP response in culture was related to the thymus-dependence of the carrier molecule employed. Thus, when POL was used as carrier, a normal anti-DNP response could be obtained, whereas little, if any, response was obtained with a conjugate of DNP-erythrocytes in which case the carrier is T cell-dependent. This observation has been interpreted as evidence for the notion that the predominant T cell function is related to modulating, in some way, the presentation of antigenic determinants to B cell receptors for appropriate immune induction to occur (74, 151). A word of caution is appropriate in this regard, although a T cellindependent antigen, such as DNP-POL, will induce an antibody re-

REGULATORY INFLUENCE OF ACTIVATED T CELLS

25

sponse in one dose range, the very same DNP-POL readily induces DNPspecific tolerance in vitro in slightly higher doses ( 1 5 2 ) . Taken together with other recent observations concerning hapten-specific tolerance induction in vivo, where it has been found that this can be readily obtained with hapten-conjugated to nonimmunogenic carriers (for which presumably no or few specific T cells exist) (153-155), the implication is clear that direct interaction of antigen with B cell receptors may favor a tolerogenic signal, even where the antigen is clearly capable of providing an immunogenic signal, in the absence of T cells in the appropriate concentration. We feel that these points suggest the operation of more than just presentation of antigenic determinants as the major T cell function in the regulation of B cell responses to antigen (see Section V1,C). In our opinion, it is more appropriate to consider that the stimulation of specific B cells by thymus-independent polymeric antigens in the narrow dose range where it can be achieved represents the limited response which B cells can display in the absence of T cell function. This interpretation is strengthened by the consideration that thymus-independent antigens elicit antibody responses generally restricted to the IgM class as will be discussed in Section IV,B.

B. EFFECTOF T-CELLACTIVITY ON SYNTHESIZED

THE

CLASSOF IMMUNOGLOBULIN

Thymus-independent antigens elicit antibody responses predominantly, if not solely, of the IgM class. This is a well-known feature of the immune response to pneumococcal polysaccharide (156) and has also been shown to be the case with Escherichia coli lipopolysaccharide ( 1 5 7 ) and PVP (149). This is particularly relevant in light of reported observations that even thymus-dependent antigens can elicit, under appropriate conditions in thymus-deprived mice, low antibody responses of the IgM class. Thus, Taylor and Wortis (158) have shown that IgM antibody production was less thymus-dependent than the production of IgG, when thymectomized, irradiated mice were given increasing doses of SRBC. Aird (159) has recently reported that thymus-deprived mice could be made to develop anti-NIP antibody responses with high doses of a polyvalent conjugate of NIP-BSA. The antibodies produced were exclusively of the IgM type. Moreover, monovalent NIP-BSA failed to elicit any response even in very high doses. This observation confirmed the findings of Makela ( 160) that polyvalent hapten-carrier conjugates give rise to higher proportions of IgM antibodies than monovalent or mainly monovalent conjugates. Similarly, Hamaoka ( 161 ) has recently observed that the antihapten response elicited in primed mice with high

26

DAVID H. KATZ AND BARU J BENACERRAF

doses of a heterologous carrier-hapten conjugate consists exclusively of IgM antibodies. The influence of T cells on production of IgM versus IgG antibodies is particularly apparent in the genetically controlled, H-Zlinked responses of mice to (T,G,)-A--L (162). It has recently been demonstrated by Grumet (163) that nonresponder mice develop early IgM antibody responses to 10 pg. of aqueous (T,G,)-A--L comparable in titer to those of responder mice. Extending this observation, Mitchell et al. (164) compared the antibody responses to ( T,G, ) -A--L of thymectomized responder and nonresponder mice to those of control mice of each type. Early IgM antibody titers were comparable among all the groups, whereas additional antigenic challenge induced IgG antibody production in only the group of responder mice which had not been thymectomized. These observations indicate that one of the requirements for a T cell regulatory influence in antibody production is related to the development of IgG antibodies more so than IgM. Thus, it may well be that the nature of the cell receptors together with physicochemical properties of the antigenic determinants, such as valency and size, are major determining factors as to whether or not B cells can be triggered to antibody production in the absence of T cell function. Moreover, it appears that where T cells participate in antibody responses, a definite consequence of their activity is related to selective forces relevant to the switch from IgM antibody synthesis to synthesis of IgG antibodies. C. ROLEOF T-CELLREGULATION IN THE SELECTIVE PRESSURE BY ANTIGEN ON B CELLS One of the regulatory functions of T cells in antibody responses appears to be the exertion of some selective pressure either directly or indirectly on the precursor B cell population. The absence of such effects most likely explains the occurrence of predominantly yM antibody responses to thymus-independent antigens ( see Section IV,B ). Our own attempts to demonstrate a selective pressure with respect to antibody class by carrier-prinicd T cells on hapten-specific B cell precursors in guinea pigs were complicated by the fact that this species makes little, if any, yM antibody to hapten-carrier conjugates (66). Others (65, 82, 1 3 4 ) , however, have failed to observe an effect of T cells on B cells as concerns the class of antibody produced in rabbits and mice. Recently, however, Miller et al. ( 7 4 ) have reported that they observed a marked shift in the class of anti-NNP antibodies produced in response to NNPHRBC by mice preimmunized to the carrier, HRBC, alone. Thus, in contrast to noninmuiie mice which developed predominantly yM antiNNP antibodies within 4 days aftcr immunization with NNP-HRBC,

REGULATORY INFLUENCE OF ACTIVATED T CELLS

27

mice preimmunized to the carrier 4 days before primary immunization with NNP-HRBC developed predominantly yG antibodies in the same time interval. It is noteworthy that carrier-primed mice produced equivalent yM responses to the nonprimed group which indicates that carrier priming did not truly “shift” the antibody class response but rather enhanced the kinetics of yG antibody-forming cell expression. The latter point is relevant when considering the failure of others to observe such an effect in different systems (65, 82, 134) where the factor of timing in the experimental protocol may be crucial. Thus, a selective pressure by T cells on antibody class might not be expected to be readily obvious at a time after immunization when yM and yG antibodies are both being produced. Perhaps the best illustrations of the participation of T cell function in the exertion of selective pressures on B cells derives from the following two recent studies. The first is the experiment of Mitchell et al. (164) on the effects of thymus deprivation on antibody responses of nonresponder and responder mice to (T,G)-A--L which we already described (see Section IV,B ) . The second is a recent study of the effect of T cells on affinity of antihapten antibodies ( 165). The experimental model employed adult thymectomized, lethally irradiated, bone marrow-reconstituted mice as thymus-deprived recipients of syngeneic T cells administered in various quantities. Such mice were then immunized with DNP-BSA or DNPkeyhole limpet hemocyanin (KLH), and the affinities of the anti-DNP antibodies produced were determined. In this system, Gershon and Paul (165) found that the affinity of anti-DNP antibody produced depended on the nature of the carrier molecule and on the number of T cells possessed by the immunized animal. Thus, the affinity as well as amount of antibody produced by thymus-deprived mice challenged with DNPKLH could be restored to normal with 0.33 X lo5 syngeneic T cells, whereas the response to DNP-BSA was not fully restored with even 1 x lo5T cells. The difference in numbers of T cells required for restoration of antibody probably reflects a difference in numbers of cells specific for KLH and BSA in nonimmunized T lymphocyte populations. These observations point clearly to an important role for T cells in the regulation of precursor B cells with respect to emergence of cells bearing highaffinity receptors. Moreover, the data argue somewhat against a restricted role for T cells of antigen presentation or concentration by virtue of the following reasoning. If T cell function were limited to presenting antigen to B cells in an effective concentration for stimulation to occur, then the diminution or absence of specific T cells would be likely to favor production of high-affinity antibodies, since only those B cells bearing high-

28

DAVID H. KATZ AND BARU J BENACERRAF

affinity receptors for the hapten would be expected to be stimulated. If one assumes, as we do, a more sophisticated role for T cells in the regulation of B cell function, it is not at all surprising that Gershon and Paul (165) found low-affinity antibody production by B cell descendants in the relative absence of T cells. The manner in which T cells exert selectional pressures on B cells could be either by direct or by indirect means. Inasmuch as a direct effect implies that T cells may regulate selection of B cell precursors of antibody-forming cells irrespective of an antigen-driven mechanism, we favor the concept of an indirect role of T cells in this regard. Thus, T cells may indirectly influence B cell selection by increasing the rate of proliferation, induced by antigen, of specific B cell precursors of antibody-forming cells. In this situation, one can easily imagine a more rapid change in the B cell population upon which selective pressure by antigen is being exerted, thus leading, through differentiative events, to more rapid appearance of B cells bearing high-affinity receptors and ultimately to their progeny synthesizing and secreting high-affinity antibodies. V. Immunological Specificity and Properties of T and B Cells Concerned with Cooperation Phenomena

The establishment of the requirement for the participation of two distinct populations of lymphocytes in the induction of immune responses to various antigens raised questions concerning the properties of such cells. Foremost among these are questions concerning the immunological specificity of each cell type. Inherent in such considerations is the question of receptor specificity among T and B lymphocytes. In this section we shall deal with these questions and consider also certain distinctive features of T and B cells. We shall describe the available evidence showing that ( 1 ) immunological specificity is a property of both B and T cells; ( 2 ) receptor specificity exists among both B and T cells; ( 3 ) B and T cells manifest distinct differences in their capacity to recognize hapten and carrier determinants; and ( 4 ) B and T cells are functionally distinguishable in their sensitivity to X-irradiation and corticosteroids. The major distinguishing features of T and B lymphocytes are summarized in Table IV.

A. IMMUNOLOGICAL SPECIFICITYOF T AND B CELLS There is little doubt that both T and B cells manifest immunological specificity. This is exemplified by the fact that specificity exists in immune responses characteristic of each cell population. Thus, cell-mediated immunity, such as delayed hypersensitivity which is predominantly a function of T cells, and antibody production which is mediated by B cells,

REGULATORY INFLUENCE OF ACTIVATED T CELLS

29

are specific events in the immune system. In immune responses to T cellindependent antigens (see Section IV,A) in which B cells alone develop antibody responses, such cells truly reflect specificity for the antigens involved. The situation becomes more complex in humoral immune responses to antigens where T and B cell cooperative interactions are required. The question arises as to which of these cells has the predominant role in dictating the specificity of the response. The existence of specificity among T cells has been demonstrated on the basis of (1) classic cell-mediated immunity experiments (Section V,C); ( 2 ) the ability to induce specific immunological tolerance in T lymphocytes (Section VII1,A); and ( 3 ) the capacity of T cells to be specifically activated by antigen as shown by several investigators. Thus, in the studies of Mitchell and Miller (56) using a double transfer system in mice, the ability of thymus-derived cells from the first host to cooperate with bone marrow cells in response to SRBC in a second irradiated host required specific antigen activation of the thymus cells in the first host. Activation of thymus cells in the first host by non-crossreacting erythrocytes, such as rabbit or horse, did not permit such cells to cooperate with bone marrow cells in response to SRBC in the second host. The requirement for exposure of thymus cells to antigen in the first host of the double transfer system was confirmed by Claman and Chaperon ( 1 1 ) and by Shearer and Cudkowicz (166). Similar observations have been made by a number of investigators (91, 92, 9 4 ) studying the reconstitution of in vitro responses to SRBC as pointed out in a previous section. Recently, Miller et al. ( 7 4 ) have extended their studies of specificity among T cells. Again employing a double transfer system they have found that T cells activated in a first irradiated host with FyG cooperate specifically with bone marrow cells in a second irradiated host in the response to FyG. The T cells activated by BSA in the first host failed to cooperate with bone marrow in the response to FyG. Unfortunately, these studies lack a very crucial additional control, namely, the response to FyG in the second host which received BSA-activated T cells if BSA was given as well. The significance of this observation would have a direct bearing on the specificity requirements of the actual cooperative interaction between T and B cells and will be considered in detail in a later section. The existence of specificity among B cells is most readily exemplified by the fact that it is this population, and not the T cell population, which gives rise to cells ultimately synthesizing and secreting specific antibodies. Furthermore, B cells possess immunoglobulin receptor molecules on their surface with antigen-specific binding properties both in nonimmune

w

0

TABLE IV DISTINGUISHING FEATURES OF B AND T LYMPHOCYTES Parameters

B Lymphocytes (Refs.p

T Lymphocytes (Refs.p

1. Surface antigenic markers 2. Surface immunoglobulin determinants

Mouse B lymphocyte antigen (1) High density of immunoglobulin (106 molecules) (4-8)

3. Surface recept.ors

Specific antigen receptors (9-15) C3 receptor (16) Receptors for immune complexes

0 Antigen (2,3) Not readily detect.able as classic Ig either because of low density (103 molecules or less) or absence (4-7) Specific antigen receptors (18-20) No C3 receptors (16) No receptors for immune complexes (17)

4. Peripheral localization (as % of lymphocytes) (a) Blood (b) Thoracic duct (c) Lymphnode (d) Spleen 5. Functional sensitivity to X-irradiation and corticosteroids

(17) 10% or less (3) 15-20% (3) 25% (3) 6 0 4 5 % (3) Sensitive to X-irradiation (21, 22); sensitive to corticosteroids after peripheralization, resistant before leaving marrow (28, 29)

90% or more (3) 8 0 4 5 % (3) 75% (3) 3 5 4 0 % (3) Functionally resistant to Xirradiation with respect to helper activity (22-24), cytotoxic activity (25) and delayed hypersensitivity (26, 27); corticosteroids distinguish 2 populations of T cells-a sensitive population (95%) located in cortex of thymus and a resistant population (5%) in the medulla (30); cortisone-resistant T cells are capable of performing all T cell functions (29, 31-35)

t!

"3 *s

-E U

c

z W

P q

6. Response to ant,igen (see text)

a

Key to references: 1. Raff et al. (3.45) 2. Reif and Allen (346) 3. Raff and Wortis (347) 4. Unanue et al. (174) 5 . Rabellino et a/. (173) 6. Raff et al. (171) 7. Pernis et al. (172) 8. Coombs et al. (176) 9. Naor and Suliteeanu (167) 10. Humphrey and Keller (170) 11. Davie and Paul (181)

Differentiate into antibodyproducing cells; usually require T cell influence to accomplish this. Not specifically involved in cell-mediated immune reactions. Establish immunological memory and can be rendered specifically tolerant

Recognize and bind antigen; undergo mitotic proliferation; exert regulatory influence on B cells, but do not synthesize and secrete classic antibodies. Specifically involved in cell-mediated immune reactions. Establish immunological memory and can be rendered specifically tolerant

5 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Byrt and Ada (177) Dwyer and Mackay (178) Unanue (180) Warner et a / . (173) Nussenzweig et al. (348) Basten et al. (349) Basten et al. (186) Roelants and Askonas (125) Engers and Unanue (350) Claman and Chaperon (11) Kate et al. (68) Dutton et al. (139)

24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.

Kettman and I h t t o n (140) Moller and Moller (201) Asherson and Loewi (202) Feldman (203) Levine and Claman (211) Cohen and Claman (212) Ishidate and Metcalf (207) Warner (19) Blomgren and Andersson (209) Cohen et al. (210) Andersson and Blomgren (213) Cohen and Claman (214)

8 4

a

8 4

8 4

H

M

U

H

$3r

c

32

DAVID H. KATZ AND BARUJ BENACERRAF

and specifically immunized animals (167-185). The fact that B cells express specific immunological memory and can be rendered specifically tolerant is further confirmation of their specificity; these will be dealt with in subsequent sections.

B. ANTIGENRECEPTORSON T

AND

B CELLS

In recent years a considerable body of evidence has accumulated demonstrating the existence of specific receptors for antigen on lymphocytes. Now it is unequivocally established by direct methodology that the receptors on B lymphocytes are immunoglobulin in nature (for review, see ref. 8 ) . Coombs et al. (176) have shown that immunoglobulincoated erythrocytes will form specific rosettes with normal lymphocytes in the presence of anti-immunoglobulin antiserum. Using refined techniques of immunofluorescence microscopy, several investigators have directly demonstrated the presence of immunoglobulin determinants on the surface of live B lymphocytes (171-175). The receptor function of these immunoglobulin determinants is based on radioautographic observations of specific binding of highly .radioactive antigen to B lymphocyte surface membranes (167,170,177-182) and the inhibition of such specific antigen-binding by preincubation of the cells with anti-immunoglobulin antisera (177, 178, 180-182). Finally, it has been well demonstrated that antigen-specific B lymphocyte precursors of antibody-forming cells can be selectively depleted from a heterogeneous lymphocyte population by passage through antigen-coated bead columns (169, 181, 183, 184) or by antigen-induced suicide which follows specific binding of radioactivelabeled antigen of very high specific activity (168, 170, 180, 185, 186). Direct evidence for the presence of immunoglobulin receptors on T cells has been more difficult to obtain, possibly because the quantity of T cell receptors is significantly lower than the sensitivity which present technology will permit to observe in a reproducible manner. In fact, the number of Ig receptors on B cells has been determined to be approximately lo5,whereas if they exist on the T cell they could not exceed lo3 and remain undetectable by direct techniques (174). Certain observations demonstrating the inhibition of some T cell functions by antibodies directed against immunoglobulins or their subunits have been interpreted to indicate that T cell receptors consist, at least, of subunits of classically defined immunoglobulin structures ( 186-191 ) , However, these observations are not without some controversy. Other investigators have tried and failed to confirm some of these experimental results. Indeed, a very critical source of error and variability in such studies, which must be taken into account, concerns the quality and specificity of the anti-immunoglobulin reagents employed by different

REGULATORY INFLUENCE OF ACTIVATED T CELLS

33

investigators, Furthermore, interpretations based on inhibition studies are subject to several alternative explanations. Thus, the inhibitory effects of anti-K L-chain reagents could also obtain if such determinants were, indeed, present on the surface of T cells but bore no functional relation to the true antigen-specific receptor, Inhibition may then occur as a result of steric considerations of the L chain determinants with respect to the actual receptor. This would offer one explanation for the failure to observe inhibitory effects with class-specific anti-H chain sera. The alternative consideration is that T cell receptors either consist solely of L polypeptide chains or that the L chains are present in association with a unique class of H chain (IgX) not found in secreted immunoglobulin (188). Thus, it is established that B lymphocytes possess antigen-specific receptors which are immunoglobulin in nature. The T lymphocytes clearly appear to possess antigen-specific receptors, but it is not yet definitely established whether or not these are immunoglobulin. TWO important points concerning T and B cell receptors are germane to developing an understanding of T and B cell interaction. a. Both T and B cells possess antigen-specific receptors which bear a functionally important relationship to the immunological activity of such cells. This is well exemplified by studies of Basten et al. ( 1 8 6 ) using the technique of radioactive antigen-induced suicide. Under appropriate conditions, incubation of either thymus lymphocytes or bone marrowderived splenic lymphocytes with high specific activity '"I-FyG abrogated the cooperative adoptive transfer response to FyG. The specificity of the suicide was shown by the ability of such treated cells to cooperate in response to another unrelated antigen, HRBC. Roelants and Askonas (125) have made similar observations in the adoptive transfer haptencarrier system in mice. Thus, incubation of Maiu squinado hemocyanin ( MSH) -primed spleen cells with high specific activity 1251-MSH abolished their helper effect with DNP-OVA-primed spleen cells in the adoptive secondary response to DNP-MSH. These data very clearly illustrate not only the receptor specificity of both T and B cells but the functional relevance of such receptors as well. b. Although both T and B cells possess receptors, the specificity restrictions of such receptors may manifest considerable differences in T cells compared to B cells, as discussed below. C. RECOGNITION OF HAPTEN AND CARRIER DETERMINANT^ BY T B CELLS

AND

Studies on the specificities of T and B cells have been best carried out with responses to hapten-carrier conjugates and have been elegantly

34

DAVID H. KATZ AND BARU J BENACERRAF

reviewed recently by Paul (8). Indications from such studies, in general terms, are that T cells participating in cellular immune reactions to hapten-carrier conjugates have specificity characteristics different from that of antihapten antibody and, therefore, by extension, different from the specificity characteristics of the hapten-specific precursor B cells. In this context, it is relevant to note that in responses to hapten-carrier conjugates, haptens do not, in themselves, constitute determinants capable of being recognized by cells concerned with carrier function or, alternatively, of stimulating such cells to perform this function under usual conditions of immunization. Haptens may, however, contribute to such determinants. Evidence for this statement derives from the following studies carried out in our laboratory (67). Sensitization of animals to haptens by preimmunization with haptenprotein conjugates failed to prepare them for enhanced primary or secondary responses to other determinants associated with that hapten on a different carrier (67). Thus, guinea pigs immunized with DNPBGG in complete Freund's adjuvant did not manifest an enhanced primary anti-OVA response upon challenge with DNP-OVA. Similarly, DNP-OVA-primed rabbits which received a supplemental immunization with sulfanil-azo ( SULF) -BGG in complete Freunds adjuvant failed to display enhanced secondary anti-DNP responses upon challenge with the double hapten conjugate of the protein carrier glucose oxidase ( GO), DNP-SULF-GO. Finally, SULF-BGG-primed guinea pigs which received a supplemental immunization with DNP-guinea pig albumin (GPA) in complete Freund's adjuvant did not make enhanced secondary anti-SULF antibody responses to a challenge with SULF-DNP-GO. However, the finding that increased anti-SULF titers resulted from secondary challenge of such animals with SULF-DNP-GPA but not with SULF-GPA indicates that the DNP determinant, indeed, contributes to the carrier determinants recognized by helper T cells. It may be argued that the latter finding may result simply from a general alteration in GPA imposed by substituting many groups on the lysine side chains, However, the failure of SULF-dimethylaminonaphtalene sulfonyl ( DANSYL )GPA to elicit a secondary anti-SULF response in the above experiment makes this explanation unlikely ( since DANSYL also binds covalently to lysines). It may likewise be argued that the capacity to enhance secondary antihapten responses by preimmunization with a protein carrier but the failure to prepare for enhanced responsiveness by preimmunization with a hapten, conjugated to a protein not used again for immunization, could be explained by the great diversity of antigenic determinants on proteins in contradistinction to the large number of similar haptenic determinants

REGULATORY INFLUENCE OF ACTIVATED T CELLS

35

in hapten conjugates. If this were the case, then it may be expected that relatively simple immunogenic polypeptides, which are limited in antigenic heterogeneity, would be marginal in their capacity to function as cooperating carrier determinants. In order to test this hypothesis, we compared the relative efficacy of a hapten, DANSYL, and a very simple carrier immunogen, PLL, in enhancing secondary anti-DNP antibody responses (67). Strain 2 guinea pigs were primed with DNP-OVA and then given a supplemental immunization of DANSYGPLL in complete Freunds adjuvant. The effectiveness of PLL as a supplement was tested by challenging one group of animals with DNP-PLL; another group was challenged with a double hapten conjugate of poly-D-lysine (PDL), DNP-DANSYGPDL, to test the effectiveness of DANSYL as a supplement. We employed PDL as a nonimmunogenic or weakly immunogenic carrier for the two haptens because its physical properties are similar to those of PLL. Whereas the animals challenged with DNP-PLL dispIayed enhanced secondary anti-DNP responses, no secondary anti-DNP responses were observed in those challenged with DNP-DANSYL-PDL despite the fact that it was considerably more heavily substituted with respect to both DNP (174 groups/mole) and DANSYL (81 groups/ mole) than the corresponding L-polymer ( 35 DNP groups/mole), which might have been expected to favor its binding by hapten-specific cells. Thus, even a molecule with restricted determinant heterogeneity, such as PLL, can function quite effectively as a carrier or helper molecule, provided it can stimulate T cells, whereas a univalent haptenic determinant, such as DANSYL, either fails to do so or does so very poorly. This evidence argues against the notion that properties simply related to possession of multiple determinants on the same backbone are relevant in determining the capacity of molecules to express carrier function. Indeed, it has recently been shown that even a nona-L-lysine can effectively exert a helper function in anti-DNP responses ( 1 8 7 ~ )We . should stress again, however, that carrier-reactive helper T cells can recognize the existence of haptenic determinants on a functional carrier. This was illustrated in the experiments employing DNP-GPA described earlier and was also observed, even more strikingly, in the later-described studies using DANSYL-PLL. Hence, although enhanced secondary antiDNP responses were obtained upon challenge with DNP-PLL, secondary challenge with a double hapten conjugate of PLL, DNP-DANSYL-PLL, resulted in considerably greater ( twelve-fold ) secondary anti-DNP antibody responses. This implies, therefore, that the relevant T lymphocytes can distinguish between PLL and DANSYL-PLL, illustrating that carrier-specific helper cells have equivalent specificities to cells mediating cellular immune reactions.

36

DAVID H. KATZ AND BARUJ BENACERRAF

We wish to make it perfectly clear that the observations described above are not interpreted by us to rule out definitively the existence of T cells with hapten specificity. It is, indeed, possible that such cells exist. Nevertheless, the existence of hapten-specific T cells need not imply that they manifest an analogous role to that of carrier-reactive helper cells, or, alternatively, if they do, their functional effectiveness may be marginal under ordinary circumstances. Mitchison ( 65, 73, 192) has reported marginal enhancement in a hapten preimmunization system in mice which may argue in favor of the existence of hapten-specific T cells expressing helper function. Thus, the adoptive secondary response of spleen cells obtained from donor mice primed with a protein BSA could be suppressed by treatment of the donors with antilymphocyte serum (ALS). However, the antibody response to BSA could be partially restored by transferring also cells from donors primed with DNP-CGG and then challenging with DNP-BSA ( 73). Mitchison interprets the observation as evidence for T cells specific for DNP which perform a helper function to enable BSA-specific B cell precursors of antibody-forming cells to develop an anti-BSA response. Several points are noteworthy regarding these experiments: (1) the effect could only be observed when donors of BSA-primed cells were subjected to immunosuppression by ALS, implying that in the presence of a normal complement of BSA-specific T cells, help from hapten-specific T cells is an irrelevant issue; and (2) the effect was obtained using extremely high doses of DNP-BSA ( loo0 pg. ) for secondary challenge. Iverson (193) and Taylor and Iverson (194) have reported similar observations using skin-painted mice as donors of hapten-specific helper cells. Unfortunately, these reports fail to mention whether or not transfer of passive antihapten antibody could substitute for primed cells in mediating the hapten-specific helper effect. This point is of considerable importance in interpreting these data in light of the recent observation of Janeway and Paul ( 195) that passively transferred anti-DNP antibody prepared BALB/ C mice for an anti-idiotype antibody response to a DNP derivative of isologous yza myeloma protein (LPC-1). This enhancement by serum antibody is not, however, comparable in magnitude to the helper effect obtained in other systems with carrierspecific T cells as described in Section II1,B. Nevertheless, the very existence of carrier specificity in hapten-specific antibody responses tells us that the expression of carrier function is not a property that can be effectively ascribed to haptens. Why, then, do haptenic determinants fail to express carrier function, or do so only marginally, although they can participate in the specificity of the cells that mediate carrier function? Three general explanations may be entertained.

~ G ~ ~ . A T O INFLUENCE R Y OF ACTIVATED T CELLS

37

1. Cells mediating carrier function ( T cells), as a class, have antigensensitive receptors generally similar to those of the precursors of antibody-forming cells ( B cells); however, the receptors of T cells differ from those of B cells in their range of specificity, inasmuch as they are specialized to recognize the determinants of native and of altered proteins and polypeptides. 2. The T cells mediating carrier function generally possess receptors of low affinity, relative to the concentration conditions in their environment, and can bind and be activated most efficiently by exposure to a total antigenic determinant and only rarely by the partial determinant which haptens constitute. 3. The T cells mediating carrier function are specialized either in the possession of an entirely different class of recognition units or in the normal binding of antigen followed by an additional operation which can only be performed on substrates of given structure. At the present time there is no hard evidence to allow a definitive choice between the above alternatives. Indeed, they are not mutually exclusive possibilities.

D. SENSITIVITY AND RESISTANCEOF T AND B CELLFUNCTION TO

X-IRRADIATION AND CORTICOSTEROIDS

A working knowledge of T and B cell functions in the immune system must take into consideration a variety of parameters. Among these are considerations of functional and receptor specificity which have been discussed in previous sections. In recent years, considerable attention has been given to the development of methodology to isolate or deplete selectively one population or the other in relatively pure terms, and a variety of techniques are now described to accomplish this end. Thus, in vitro treatment of mouse lymphoid organs with anti4 serum plus complement depletes the great majority of T cells, leaving a relatively pure population of B cells (196). Other techniques have been described but it is not relevant to discuss them at length here [for review, see Miller et al. ( 7 4 ) ] . It is of considerable interest, however, that the two cell types are functionally distinguishable on the basis of their sensitivity or resistance to X-irradiation and certain drugs. It is crucial to understand these differences in B and T cells in order to develop a reasonable clue to the nature of the interaction which occurs between them. In this section we shall present evidence demonstrating that differentiated T cells are functionally resistant with respect to helper activity to X-irradiation and corticosteroids, whereas B cells are functionally sensitive to such agents.

38

DAVID H. KATZ AND UARUJ BENACERRAF

1. X-Zrradiution The general sensitivity of the immune system to X-irradiation is a phenomenon recognized for many years. Indeed, it is on this basis that experimental models for successful adoptive transfer of immunity with lymphoid cells have been developed. Thus, X-irradiation of a host animal, under appropriate conditions, abrogates the capacity of its own lymphoid cells to engage in an immune response such that the response of normally competent transferred lymphoid cells can be assessed. In recent years, it became recognized that the functional role of macrophages in immune responses was unaffected by exposure to X-irradiation ( 8 9 ) , but until very recently it was generally believed that immunocompetent lymphocytes were all radiosensitive. Our present state of knowledge, however, permits us to make a functional distinction between T and B lymphocytes on the basis of their respective resistance and sensitivity to X-irradiation. The functional resistance to X-irradiation of T cells concerned with helper activity was first clearly established by investigators working independently in two laboratories using different cell cooperation systems. Katz et al. (68) found that carrier-primed lymphoid cells which had been exposed to high doses of X-irradiation in vitro were, nevertheless, capable of performing a helper function in enhancing secondary anti-DNP antibody responses of syngeneic guinea pig recipients to which they had been transferred. Exposure of the carrier-specific BGGprimed lymphocytes to X-irradiation in vitro prior to transfer into DNP-OVA-primed recipients did not affect the capacity of such cells to perform a helper function in the secondary anti-DNP response to DNP-BGG. This was true over a wide dose range of X-irradiation ( 1OOO-5OOO R). Furthermore, a clear functional distinction between T and B cells, with respect to radioresistance, was established in these experiments. Thus, the functional capacity of B cells in the BGG-primed lymphoid cell population not exposed to in vitro X-irradiation was shown to be intact by the presence of anti-BGG antibodies in the serum of recipients of such cells prior to DNP-BGG challenge and a clear anamnestic anti-BGG antibody response following DNP-BGG challenge. In contrast, no anti-BGG antibody could be detected either before or after DNP-BGG challenge in the sera of recipients of BGG-primed cells which had been exposed to in uitro X-irradiation ( 68). Almost simultaneously, investigators in Dutton’s laboratory found that educated helper T cells in the in uitro primary responses to SRBC and to trinitrophenyl (TNP)-erythrocytes were radioresistant (90,139, 1 4 ) . Thus, the in uitro primary anti-TNP response of normal mouse spleen

REGULATORY INFLUENCE OF ACTIVATED T CELLS

39

cells to TNP-SRBC was markedly enhanced by the addition of spleen cells from carrier (SRBC)-primed mice (139, 140). Exposure of the carrier-primed spleen cells to in vitro X-irradiation (as much as 4000 R ) completely inhibited the capacity of such cells to develop an antibody response but did not prevent their ability to function as helpers for the anti-TNP response of the normal spleen cells. Moreover, the radioresistant helper cell was shown to be a T cell (139, 197). In their studies of reconstitution of the primary in vitro anti-SRBC response of spleen cells from thyniectomized mice, Munro and Hunter (88) observed that X-irradiation ( l W R ) of the allogeneic donor of normal spleen cells immediately prior to sacrifice diminished, but did not abrogate the cooperating effect of such cells. Observations of other investigators can be interpreted as consistent with the above results ( 198-200). However, early studies of thymus-marrow synergism in mice had suggested that both T and B cell components were sensitive to X-irradiation. In the studies of Claman and co-workers (11, 77) and Miller and Mitchell (9, 57) the functional capacity of thymus-derived cells to cooperate with B cells was abrogated by X-irradiation of such cells either in situ or in vitro. It would appear, therefore, that a contradiction exists between these observations and those of Katz et al. (68) and Kettman and Dutton (140) cited above. We feel, however, that a valid explanation may be offered for the conflicting observations. The experimental protocols employed by Claman (11,77) and Miller (9, 57) and their associates required a crucial period of proliferation by T cells in the presence of antigen, as shown by their own data, before such cells could effectively facilitate the antigen-induced activity of normal B cells. That this is, indeed, the case is also illustrated by the findings of Claman and Chaperon ( 77) and Shearer and Cudkowicz ( 1 6 6 ) in the double transfer thymus-marrow synergy model that a crucial 2-3 day period of interaction between SRBC and T cells in the first host was required before such T cells could cooperate with B cells in a second host. Under these circumstances, any inhibitory influence on mitotic activity during this crucial period by X-irradiation or antimitotic drugs would understandably abolish the helper function of the T cells. In contrast, under conditions where T cell proliferation is not required, because the cell population has already been expanded and differentiated, the helper function of the cells is not affected by X-irradiation. This is well exemplified in our in v h o model where the crucial proliferative response of the helper T cells had presumably occurred early following BGG priming (68). It is noteworthy here to make the following additional points: (1) the capacity of such irradiated carrier-primed

40

D A W H. KATZ AND BARU J BENACERRAF

cells to respond to antigen in vitro with DNA synthesis was 90% inhibited with relatively low doses of in vitro X-irradiation ( W R ) and totally abolished at doses above 1500 R; and (2)the facilitative capacity of irradiated carrier cells was intact after the 6-day interval between the transfer of cells and the subsequent secondary antigenic challenge (68). These points are of major importance in developing an understanding of the T cell function in cooperative interactions with B cells. It appears, therefore, that having successfully undergone whatever degree of proliferation and differentiation is occasioned by antigenic stimulation, the T cell can perform its helper function in the absence of any further proliferation. From a purely functional standpoint this would imply that ( 1 ) the mature antigen-specific T cell cooperatively interacts with B cells in a way that requires little or no further division; ( 2 ) the heavily irradiated helper T cell can survive and perform its specific function in vivo for at least as long as 6 days; or, alternatively, the helper T cell, once fully differentiated, can perform its role through interaction with other cells prior to subsequent immunization and, therefore, its presence is not required at the precise time of antigenic stimulation. There have been several previous demonstrations of functional radioresistance among T cells participating in cell-mediated immune reactions. Thus, Moller and Moller (201) observed that in vitro killing of target cells by mouse lymphoid cells was not inhibited by exposure of the latter to 1500 R in vitro. Asherson and Loewi (202) found that delayed hypersensitivity reactions in guinea pigs could be transferred with donor cells exposed in vitro to lo00 R. Feldman (203)was able passively to transfer delayed hypersensitivity in rats with donor lymphocytes which had been exposed in vitro to 6000 R X-irradiation. It seems likely, therefore, that T cell function, in cellular immunity as well as in antibody production, does not require proliferation once the initial antigen-induced process of differentiation and clonal expansion has occurred. 2. Corticosteroids It has been known for quite some time that cortisone has a general suppressive effect on immune responses in some species (204, 205), presumably reflecting the rapid atrophy of lymphoid structures which follows administration of large doses of the drug. This effect is particularly striking in the thymus where approximately 95%of the total thymic lymphocyte population is depleted by cortisone ( 2 0 6 ) .It is the 5%of thymocytes which remain after such treatment which has attracted renewed interest in the effects of cortisone on immunocompetent cells in the past few years. The cortisone-resistant thymocytes appear to be located primarily in the medulla, whereas the sensitive cells are predominantly in

REGULATORY INFLUENCE OF ACTIVATED T CELLS

41

the cortex of the thymus ( 2 0 7 ) .These cortisone-resistant thymocytes are larger and possess more histocompatibility antigens on their membranes than the sensitive thymocytes ( 2 0 8 ) . The basis for such renewed interest in this area, in light of the establishment of the two-cell concept of immunity, was an observation made by Warner ( 1 9 ) some years ago that the few remaining thymocytes of cortisone-treated chickens were capable of expressing graft-versus-host (GVH) reactivity comparable in magnitude to normal chicken thymus. Recently, Blomgren and Anderson ( 2 0 9 ) found that thymus cells obtained from cortisone-treated donor mice were of the order of ten-fold more reactive in inducing the GVH response than whole thymus from untreated donors. Cohen et al. ( 2 1 0 ) similarly observed that cells in mouse spleen, bone marrow, and thymus capable of initiating GVH reactivity were cortisone-resistant. Such observations clearly established that the T cells responsible for cell-mediated immunity were resistant to the effects of cortisone. Since administration of corticosteroids to mice at the same time as primary SRBC immunization markedly suppressed the humoral response (205), the question was raised as to whether a differential effect on T and B lymphocytes could be established. Levine and Claman (211) found that spleens from cortisone-treated donor mice were significantly decreased in capacity to transfer adoptively anti-SRBC responses to irradiated recipients; bone marrow cells from cortisone-treated donors, however, were perfectly capable of cooperating with normal thymocytes in the adoptive anti-SRBC response. Subsequent studies by Cohen and Claman ( 2 1 2 ) established that the cortisone-sensitive spleen cells were the bone marrow-derived antibody-forming cell precursors, and that the T cells responsible for helper function in the anti-SRBC system were cortisoneresistant. Similar observations were made by Anderson and Blomgren (213) in humoral responses of mice to SRBC, BSA, and the NIP hapten. Thus, adult thymectomized, irradiated, bone marrow-reconstituted mice which were unable to develop humoral immune responses to any of these antigens could be restored by injection of cortisone-resistant thymus cells from syngeneic donors. Thus, it appears that cortisone exerts a differential effect on immunocompetent lymphocytes in the following ways. (1) On the one hand, B cell precursors of antibody-forming cells which have migrated to peripheral lymphoid organs are cortisone-sensitive; those which have not yet left the bone marrow, on the other hand, appear to be cortisone-resistant. Whether this finding reflects a true difference in sensitivity based on qualitative differences in the maturation stage of the cell or merely reflects differences in effective doses of cortisone reaching the cell in bone

42

DAVID H. KATZ AND BARUJ BENACERRAF

marrow versus periphery is not known. ( 2 ) The T cells which are predominantly located in the thymic cortex and presumably not functionally mature are cortisone-sensitive; T cells of the thymic medulla and those that have migrated to peripheral lymphoid organs and are, therefore, fully differentiated to participate in humoral and cellular immune reactions are functionally resistant to the effects of cortisone. The nature of the maturational differences between sensitive and resistant T cells is not established. Inasmuch as it may be tempting to ascribe cortisone sensitivity to those cells undergoing mitotic turnover, it is noteworthy that cortisone administration during antigen-induced mitosis of T cells does not have a suppressive effect on the capacity of such cells to participate in either humoral or cell-mediated immune responses (212,214). VI. Mechanism of Regulation of B Cell Function by T Cells

When considering the cellular events critical to the expression of a regulatory function by T cells on the response to antigen of B cells, several possibilities have been entertained by various workers in the field. These will be considered, in order, in our opinion, of their increasing relevance as follows: ( 1) transfer of genetic information; (2) antigen concentration or presentation; ( 3 ) regulation of B cell response to antigen by products of activated T cells. A. TRANSFER OF GENETIC INFORMATION There is no evidence whatever that transfer of genetic information could explain the regulatory role of T cells on the response of B cells to certain antigens. Nonetheless, from a purely hypothetical standpoint, any consideration of cell interactions must include the possibility that a signal from one cell type, in the form of genetic information, may be transferred to another cell type leading to a fundamental change in the specificity and/ or function of the second cell. This possibility was entertained in early studies of Mitchell and Miller ( 5 6 ) . If this were a relevant consideration with respect to T and B cell interaction, then one would not expect to find the high level of specificity restriction existing, that, indeed, exists among B cells in the nonimniune animal (see Section V,A ) . Moreover, the possibility of genetic information transfer from T cells to B cells of functional significance has been recently definitively excluded on the basis of the experiments described below. Mitchison et al. (65, 7 0 ) showed in the adoptive transfer response to hapten-carrier conjugates in mice that, if the donors of carrier-primed cells had immunoglobulin allotype markers different from donors of hapten-primed cells, then the antihapten antibodies produced after secondary challenge with hapten carrier were of the allotype of the hapten-

REGULATORY INFLUENCE OF ACI’NATED T CELLS

43

primed cell donors. Similarly, Jacobson et al. (215) using congenic mice as donors of T and B cells in the adoptive transfer response to SRBC observed that all of the anti-SRBC PFC were of the allotype of the B cell donors. Since the allotypic markers are located on the nonvariable Fc portion of the heavy chains of immunoglobulin molecules (216), it may be considered that these observations are not completely convincing arguments against the possibility of genetic information transfer. However, studies of Katz et al. (66) have shown that the affinity of antihapten antibodies produced during enhanced primary and secondary antihapten responses in guinea pigs and rabbits is determined by the mode and time of immunization with the hapten-carrier conjugate and not by the mode and time of immunization with the free carrier (see Section 111,B). Since antibody affinity is a marker of the variable region of the immunoglobulin molecule, these findings taken together with the allotype data cited above definitively exclude the possibility of T cell regulation of B cell function by virtue of macromolecular information transfer.

B. ANTIGENPRESENTATION AND CONCENTRATION The hypothesis of “antigen focusing” has been, perhaps, the earliest and most popular explanation for the nature of T and B cell interaction in antibody responses. This concept, first proposed by Mitchison (63, 65), assumes a totally passive role for the specific helper T cell, whereby antigen is recognized and bound by such cells, transported to sites in lymphoid organs where B lymphocytes reside, and then presented to specific B cells in a concentration appropriate to stimulate them. The antigen-focusing hypothesis is immediately appealing in that it has a certain degree of elegance in its simplicity and can be used to explain many established features of T and B cell interactions. Thus, T cell specificity is required for recognition and binding of antigen by certain of its determinants, not necessarily and, indeed, probably not, the same as those determinants recognized by specific B cell receptors. The antigen would thereby be “bridged between the two cells having different determinant specificities for distinct moieties of the complex antigen. Since T cells comprise the major portion of the recirculating small lymphocyte population (217), they can be considered to be particularly well-suited for the role of binding antigen and concentrating it at sites where B cells reside. There are alternative ways in which T cells may serve in antigen concentration and presentation other than forming an antigen bridge with B cells. One such possibility is that T cells elaborate a unique class of antibody ( IgX) with specific binding capacity for certain determinants of the antigen which it then concentrates either onto the B cell itself or

44

DAVID H. KATZ AND BARU J BENACERRAF

onto macrophages ( 7 4 ) .The advantage to the B cell of an antigen concentration mechanism, either through bridging or IgX, is felt to be related to presentation of antigenic determinants in such a way as to favor multivalent binding of these determinants by the B cell receptors. The argument for the antigen concentration hypothesis rests mainly on logical speculation and circumstantial evidence. Direct evidence that T cells serve predominantly in an antigen presentation role is largely lacking. Indeed, more evidence exists in favor of a more sophisticated role for T cells in antibody responses as will be discussed in Section V1,C. There are, nevertheless, certain observations which are more easily explained by an antigen concentration scheme. Mitchison ( 7 0 ) , in the adoptive transfer of secondary antihapten responses in mice, has shown that the hapten must be present on the carrier molecule in order to elicit a cooperative effect. When NIP-OVA-primed cells are transferred together with BSA-primed cells to irradiated recipients, a cooperative response is obtained to challenge with NIP-BSA but not if an unrelated NIP carrier is administered simultaneously with free BSA. Similar observations were reported by Kettman and Dutton (140) in studies of primary in oitru antihapten antibody responses. Thus, carrier-primed spleen cells develop enhanced primary in uitro anti-TNP responses to a TNP conjugate of the same carrier used for priming, but not to a TNP conjugate of another carrier even if the original carrier used for priming is simultaneously present, in unconjugated form, in the culture. Although these observations can be interpreted as evidence for an antigen-focusing hypothesis, i.e., to explain the requirement for hapten and carrier to be together in the same molecule, a word of caution should be added concerning the experimental protocols employed. It is possible that the design of Mitchison’s (70) and of Kettman and Dutton’s (140)experiments inadvertently favored a negative result for reasons other than a requirement for hapten to be present on the carrier molecule. We refer to the possibility of antigenic competition at the level of the helper T cell developing under the conditions employed, inasmuch as we shall present compelling evidence in later ( Section VI1,B) that such competition does exist with respect to helper function in antihapten antibody responses. It may be alternatively considered, at least with respect to the in uiuo situation, that a negative result in such experiments may reflect a lack of close proximity between the responding T and B cells. Even a mechanism, such as we shall propose in the next section, whereby stimulated T cells may regulate B cell function via release of soluble mediators may require a reasonably close association between the respective cells, especially if such mediators are relatively fast-acting and short-lived. If the helper function of T cells reflected simply a passive role of

REGULATORY INFLUENCE OF ACTIVATED T CELLS

45

these cells as vehicles of antigen presentation and/ or concentration to B cells, then it should be possible to supplant the need for T cells in antibody production by the use of inert antigen-coated particles or cells. Several attempts to accomplish this have been unsuccessful in our own laboratory and in those of others. Thus, Unanue (218) has shown that thymectomized mice are markedly depressed in their capacity to develop antibody responses to KLH. Moreover, he failed to restore their responsiveness by administering the KLH bound to live macrophages. Since the macrophages localized in the vicinity of B cells, this result indicates that more than antigen presentation is required for B cell stimulation. Katz et al. (219), using the model of enhanced secondary anti-DNP responses in guinea pig recipients of syngeneic carrier-primed lymphoid cells ( 6 7 ) , were unable to observe a cooperative effect when anticarrier antibody-coated polyacrylic resin particles were substituted for carrierprimed lymphoid cells. Polyacrylic resin particles coated with highaffinity guinea pig anti-BGG immunoglobulin were injected intravenously in various quantities into groups of DNP-OVA-primed guinea pigs. Similarly, primed control groups received either syngeneic BGG-primed lymphoid cells or lymphoid cells from nonprimed donors. All groups were then challenged with a heterologous DNP-BGG conjugate. The recipients of BGG-primed lymphoid cells displayed enhanced secondary anti-DNP responses to DNP-BGG challenge, whereas the other groups made no secondary response. Had the BGG-specific lymphoid cells simply picked up the DNP-BGG via specific receptors and concentrated it at sites of B cells, we may have expected a similar passive effect of the antiBGG-coated particles. The argument from such an experiment, we realize, is far from compelling because, among other reasons, although particles such as those we used may be the same size as lymphoid cells, they are not lymphoid cells and, therefore, make any physiological interpretation fruitless. The best approach to the problem, obviously, should employ lymphoid cells as the antigen-coated particles. Recently, Miller et al. (220) tested the possibility that T cells cooperated with B cells in antibody responses by acting as passive carriers of antigen. This was done using nonreactive T or B lymphocytes artificially coated with the antigen, FyG, in an adoptive transfer system in mice (Table V ) . They found that T cells from FyG-tolerant donors which were coated with FyG failed to cooperate with B cells in irradiated recipients in response to FyG, although such cells could cooperate in response to HRBC. However, FyGcoated T cells from nontolerant donors were capable of cooperating with B cells in response to FyG. Similar observations were obtained using antigen-coated B cells (220). Thus, B cells were incubated in oitm with

TABLE V ANTI-FyG PFC RESPONSE IN SPLEENS OF T X BM CBA MICE RECEIVINQ FALG-INCUBATED TDL FROM NONTOLERANT OR FyG-TOLERANT (CBA X C57Bl)Fl MICE'.~

TDL given None 2 X 10' FALG- coated cells from FyGtolerant mice 2 X lo7 FALG- coated cells from cyclophosphamidepretreakd mice

Antigen challenge in viva

+

No. of mice in groupc

Peak anti-FyG PFC per spleend

Reduction of anti-FyG with

19SPFC

7 S PFC

CBA antiC57B1 serum

2(4-1)

-

500 pg. F r G AP 4 X 108 HRBC 2 wk postirradiation 4 X 108HRBC

16

510 (700-370)

16

140 (160-120)'

4 X 108 HRBC

16

2(4-1p

7690 14,730 (8890-6640)e(16,640-13,040)6

-

20 (19 S) 0 (7 S)

C57B1 anti-CBA serum

Peak 19 S anti-HRBC PFC

a

390 (690-230)

x

32,230 (40,640-25,560) 99 (19 S) 29,210 99 (7 S) (38,680-22,070)

We thank Dr. J. F. A. P. Miller for permission to reproduce these data from Miller et al. (127). This experiment illustrates the inability of antigen-coated tolerant thoracic duct lymphocytes (TDL) to induce a response to fowl -,-globulin (FyG) in adult, thymectomized, irradiated, bone marrow-protected (T X BM) recipient mice. The TDL from nontolerant or FyG-tolerant Fl donor mice (tolerance induced by cyclophosphamide method) were reacted in oitra with fowl antimouse lymphocyte globulin (FALG) and then transferred to T X BM CBA mice. As control, these mice also received antigen challenge with human red blood cells (HRBC) int.ravenously.Control T X BM mice received no TDL but were challenged intravenously with alum-precipitated FyG plus pertussis and HRBC. A t the times indicated after transfer and challenge the numbers of 19 and 7 S anti-FyG plaque-forming cells (PFC) and 19 S anti-HRBC PFC in spleens of recipient mice were determined. Half the mice were killed at 5 to 6 days to determine 19 S anti-FyG PFC, the other half at 7 days to assay 7 S anti-FyG PFC and 19 S anti-HRBC PFC. Geometric means (upper and lower limits of SE). e P values < 0.005.

%

u

*

E

;?

3

5

m cl

z

m

n

8%

REGULATORY INFLUENCE OF ACTIVATED T CELLS

47

an antigen-antibody complex consisting of NIP-FyG-anti-FyG, and the coated B cells were transferred with or without normal T cells into irradiated hosts. A primary anti-NIP response occurred only if both B and T cells were given. Furthermore, transfer of such coated B cells together with spleen cells from NIP-HGG-primed donors resulted in a substantial secondary anti-NIP response; this effect was abolished by treatment of the NIP-HGG-primed spleen cells with anti-8 serum and complement. These observations can only be interpreted in one way: antigen-coated T or B lymphocytes cannot substitute for specific T cells in either primary or secondary antibody responses. Since Miller et al. ( 220) determined that the artificially antigen-coated cells they employed followed normal homing patterns to the lymphoid organs, the data argue strongly against a passive antigen-focusing role for T lymphocytes in antibody responses. Finally, in this context it would be well to recall the observations of Gershon and Paul ( 1 6 5 ) concerning the relationship of T cell activity to affinity of antibodies produced. Reiterating what we said earlier (Section IV,C), if the predominant role of T cells in antibody production was a relatively passive one serving to localize and increase the effective antigen concentration in the microenvironment of B cell receptors, then one would expect that with decreased number of T cells, antigen would most likely stimulate B cells with high-affinity receptors. The resulting serum antibodies should, therefore, be of high average affinity. In contrast, Gershon and Paul (165) found just the opposite, i.e., thymusdeprived animals produced low-affinity antibodies and the addition of T cells led to production of antibodies of increasing affinity. Their results imply a more complex and sophisticated role of T cells in the regulation of B cell function in antibody production. OF B CELL FUNCTION IN ANTIBODYPRODUC~ION BY C. REGULATION MEDIATORS PRODUCED AND SECRETED BY T CELLS

The last possibility regarding T cell function in humoral immune responses, and the one we favor, is that T cells play an active regulatory role on B cell precursors of antibody-forming cells reacting with antigen. The expression of this regulatory effect would be mediated through soluble factors synthesized and secreted by T cells following antigenic stimulation. It may well be that all differentiated T cells responsible for a variety of cellular immune reactions mediate their function in large part through the release of nonspecific soluble factors. The elaboration of migration inhibitory factor (MIF) by T cells involved in delayed hypersensitivity reactions, which has been so elegantly demonstrated by David (221, 222) and Bloom and Bennett (223, 224), is exemplary of

48

DAVID H. KATZ AND BARUJ BENACERRAF

this point. We feel that elaboration of soluble mediators capable of affecting, in some way, the function of other cells of the immune system is a general property of specific T cells in contrast to the general property of B cells which is to differentiate into antibody-secreting cells. Since, as yet, no one has isolated or characterized a soluble factor produced by T cells which can be definitively shown to express all aspects of cooperative T cell function with respect to B cell responses to antigen, this discussion must rest to a certain extent on logical speculation. Nevertheless, the evidence which we interpret to argue in favor of this concept of T cell function in humoral immunity is rapidly mounting in a very compelling way. In many instances, certain experimental observations which support the concept were either puzzling to the reporting investigator( s ) or went unnoticed by them. We have taken obvious liberties, therefore, in placing our own interpretations on them. We have attempted to bring together appropriately related observations and/or experimental systems, and, therefore, the discussion has been divided in subsections covering the following: (1) the enhancement of immunocompetent cell function by GVH or comparable reactions, i.e., the “allogeneic effect”; (2) enhancement of immune responses by soluble factors released by T cells in culture; and ( 3 ) enhancement of antibody responses following nonspecific T cell stimulation.

I . Allogeneic Effect In the course of our studies of cell interactions in hapten-specific immune responses in guinea pigs, we noted that induction of a transient GVH reaction in otherwise perfectly normal and fully immunocompetent adult animals had a rather remarkable effect on several parameters of immune function (153, 225, 226). This phenomenon, which we have termed “allogeneic effect,” appears, in our opinion, to have great general relevance to the nature of the regulatory function of T cell activity on other cells, either B cells or other T cells, of the immune system. The characteristics of the allogeneic effect will be reviewed in some detail here. In the initial studies (225), it was observed that intravenous injection of lymphoid cells from inbred strain 2 guinea pig donors into DNP-OVAprimed allogeneic strain 13 recipients resulted in a significant rise in the levels of both anti-DNP and anti-OVA antibodies in the recipients’ serum (Fig. 3). This rise in antibody titers, which occurred spontaneously in the absence of any further exogeneous antigen challenge, began between the sixth and tenth day after allogeneic cell transfer, reached a peak by day 13 and then gradually diminished thereafter. An increase in total serum 7-globulins which was also observed during this time was of rela-

REGULATORY INFLUENCE OF ACI'IVATED T CELLS

200

1

'

I

I

1

'

1

'

1

'

SERUM ANTI-DNP ANTIBODY

I

'

1

I

'

I

is g 100Q

'

1

'

1

Stroin 2 Donors Stroin 13 Recipients-No Boost

P

a

1

TOTAL GAMMA GLOBULINS

d % - 140E \ p 120-

s

'

1

8)

80-

z

SERUM ANTIANTIBODY OVA

I

-

,? 60-

I= 2

a

40 -

-

20 0; '

h ' 11' L ' b

'

I d ' 16'

Ib' I;' DAYS AFTER TRANSFER

u 6

10

14

FIG. 3. Stimulation of antibody production as a result of transfer of immunocompetent strain 2 cells into primed strain 13 recipients in the absence of secondary antigenic challenge. 200 x 10" Strain 2 lymph node and spleen cells were injected intravenously into strain 13 recipients primed 3 weeks earlier with 2,4-dinitrophenyl ( DNP)-ovalbumin (OVA), Recipients were bled at various times after cell transfer without a subsequent antigenic challenge, and their sera were analyzed for anti-DNP and anti-OVA antibody and total y-globulin concentrations. The numbers in parentheses refer to the number of recipients analyzed at a given time after transfer. [These data from our laboratory appeared in J . Ex&. Med. 133, 169, 1971 ( reference 225) .]

tively small magnitude in comparison to the increase in specific antibodies. In addition to the spontaneous synthesis of both anti-DNP and antiOVA antibodies by such recipients in the absence of antigenic challenge, a striking secondary anti-DNP antibody response could be elicited by an appropriately timed, secondary challenge with DNP-BGG ( Fig. 4 ) (225). The crucial feature of the latter effect is that the requirement for carrier specificity is completely abrogated by the allogeneic effect. In a syngeneic cell transfer system, we had previously shown that DNP-OVAprimed guinea pigs developed enhanced secondary anti-DNP antibody responses to challenge with DNP-BGG provided they had received BGGprimed lymphoid cells from syngeneic donors; transfer of nonprimed

50

DAVID H. KATZ AND BARUJ BENACERRAF

STRAIN 13 RECIPIENTS -Strain

2 CFA Cells, No Boost

-

Strain 2 BGG Cells, No Boost

600 U

5

500-

6 400-

8'

88

z

2 0

300-

1

I

U

I-

I

5 200W

I

I

I I

W

c3

I

I

100 -

T

DAYS AFTER TRANSFER

FIG.4. Stimulation of anti-2,4-dinitrophenyl ( DNP) antibody production as a result of transfer of immunocompetent strain 2 cells into primed strain 13 recipients, showing the effect of secondary antigenic challenge with a DNP-heterologous conjugate 6 days after transfer. About 200 X loe strain 2 lymph node and spleen [ (CFA) or bovine y-globulin (BGG)] cells were injected intravenously into strain 13 recipients primed 3 weeks earlier with DNP-ovalbumin (OVA). Recipients were either not boosted or boosted with DNP-BGG 6 days after transfer. The numbers in parentheses refer to the number of recipients in a given group. The left panel illustrates the anti-DNP responses of recipients of CFA cells, whereas the right panel illustrates the anti-DNP responses of recipients of BGG-specific cells. [These data from our laboratory appeared in J . Exptl. Med. 133, 169, 1971 (reference 225) .I

lymphoid cells was ineffective ( 6 7 ) . In marked contrast, DNP-OVAprimed guinea pigs displayed dramatic secondary anti-DNP responses to DNP-BGG challenge administered 6 days following transfer of allogeneic lymphoid cells, irrespective of whether or not the allogeneic cell donors were primed with BGG (225). Hence, we have a situation in which the need for carrier-specific T cells has been apparently abolished. Indeed, this has been conclusively demonstrated in subsequent studies by Katz et al. (153)in which a secondary anti-DNP response has been elicited in DNP-OVA-primed guinea pig recipients of allogeneic cells following

51

REGULATORY INFLUENCE OF ACTIVATED T CELLS

challenge with a nonimmunogenic DNP copolymer. This compound, a DNP conjugate of the copolymer of D-glutamic acid and D-lysine (DNPD-GL), is normally very highly tolerogenic in guinea pigs, but in the presence of the allogeneic effect, it is capable of eliciting a secondary response comparable to that induced with DNP-BGG (Fig. 5). Since it is fair to assume that T cells specific for D-GL either do not exist or are not functionally significant, the allogeneic effect appears to have permitted the direct triggering of DNP-specific B cell precursors of antibody-forming cells by the nonimmunogenic DNP-D-GL. The mechanism of the allogeneic effect has been clearly established by our studies to be the result of a specific immunological attack of T ' 1779)

u

E

-

Strain 13 Recipients Strain 2 Donors MDNP-BGG Secondary Immunization *--b DNP-

D -GL Secondary Immunization

I

WEEKS AFTER PRIMARY IMMUNIZATION

7

1

II

DAYS AFTER SECONDARY IMMUNIZATION

FIG. 5. Stimulation of anti-2,4-dinitrophenyl ( DNP ) antibody production as a result of transfer of immunocompetent strain 2 cells into primed strain 13 recipients, showing the effect of secondary antigenic challenge with ( DNP)-bovine y-globulin (BCG) and DNP-copolymer of D-glutamic acid and n-lysine ( D-GL), About 550 X 10" strain 2 lymph node and spleen cells were injected intravenously into strain 13 recipients primed 3 weeks earlier with DNP-ovalbumin (OVA). Recipients were either not boosted or boosted with 1.0 mg. of DNP-BGG or DNP-D-GL in saline 6 days after transfer. Serum anti-DNP antibody concentrations just prior to secondary immunization and on days 7 and 11 are illustrated. The numbers in parentheses refer to the numbers of animals in the given groups. [These data from our laboratory appeared in J. Erptl. Med. 134, 201, 1971 (reference 153).]

52

DAVID H. KATZ AND BARUJ BENACERRAF

grafted cells on host cells (225, 226). This conclusion is based on the following observations. 1. Transfer of lymphoid cells from allogeneic or semiallogeneic donors under circumstances where the transferred cells are incapable of reacting against tissue antigens of the host fails to elicit the allogeneic effect. This is true despite the fact that the transferred cells, by virtue of possessing foreign histocompatibility antigen, elicit a specific rejection response on the part of the host. These points have been demonstrated by transferring into DNP-OVA-primed strain 13 recipients the following lymphoid cells (225): ( a ) L,C leukemia lymphocytes from strain 2 guinea pigs-these leukemia cells possess the strain 2 histocompatibility antigens but are immunologically incompetent; ( b ) strain 2 lymphoid cells which have been rendered immunologically incompetent by exposure to in vitro X-irradiation (3000 R ) ; and ( c ) lymphoid cells from ( 2 X 13)F, hybrid donors-these cells are immunologically competent but cannot mount a rejection response against the parental recipients. 2. Transfer of lymphoid cells from parental strain 2 donors elicits the allogeneic effect in ( 2 X 13)F, hybrid recipients which have been primed with DNP-OVA. In this situation, the parental lymphoid cells can mount a GVH reaction against the F, recipients, but the hosts are incapable of rejecting the parental cells (226). The data of these experiments are summarized in Table VI. Thus, the allogeneic effect phenomenon reflects the development in the lymphoid organs of the host of a specific GVH reaction. Moreover, the phenomenon is expressed irrespective of the concomitant development of a host rejection response. The degree to which the phenomenon influences antibody production is related to the following crucial factors : 1. The host must be primed with the antigen in question before the transfer of allogeneic cells. We have made several unsuccessful attempts to enhance primary anti-DNP responses in guinea pigs inoculated with allogeneic cells at various times prior to primary immunization (153,225, 226). This raises the possibility that a critical change related to antigenic stimulation must occur in the specific B cell population before such cells can be affected by the allogeneic effect. 2. The intensity of the allogeneic effect on antibody synthesis is related to the intensity of the GVH reaction induced. In guinea pigs, transfer of as little as 50 x lo6 allogeneic lymphoid cells enhanced the production of anti-DNP antibodies in recipients, but an almost linear increase in magnitude of the phenomenon was observed with progressively higher numbers of cells transferred (225). We also observed that donor cell inocula, consisting of both lymph node and spleen cells, were more potent in mediating the effect than an equivalent number of spleen cells

REGULATORY INFLUENCE OF ACTIVATED T CELLS

53

TABLE VI REACTION AS THE UNDERLYING MECHANISM OF ALLOGENEIC EFFECT IN GUINEA

GRAFT-VERSUS-HOST THE

No. and strain of DNP-OVA-primed recipien t s b

Type of donor lymphoid cells transferred

5 Strain 13

Nonirradiated strain 2 Nonirradiated strain 2 Irradiated (3000 R) strain 2 Strain 2 leukemia, nonirradiated (2 X 13)F1 hybrid, nonirradiated

5 (2 X 13)Fi 5 Strain 13 5 Strain 13

5 Strain 13

Increase in geometric mean anti-DNP antibody concentrations (pg/ml) from days 6 to 13 after cell transfer 728.4 461.3 -5.8

-23.1 -7.0

a This table represents a composite of dat,a from several studies of Katz et al. (2259 226). * These observations demonstrate that the allogeneic effect is mediated through the development of a graft-versus-host reaction and does not require the concomitant existence of a host rejection response. Strain 13 and (2 X 13)R hybrid guinea pigs were immunized intraperitoneally with 2,Cdinitrophenyl (DNP)-ovalbumin (OVA) in saline. Three weeks later allogeneic lymphoid cells of the types indicated were injected intravenously into individual recipients. Six days later, all recipients were boosted with DNP-bovine r-globulin (BGG). The data are expressed as the increase in geometric mean anti-DNP antibody concentrations from the day of secondary boost (day 6 after transfer) to 7 days later (day 13).

alone ( 2 2 5 ) . This superiority of lymph node cells is consistent with observations in the mouse that fewer lymph node cells than spleen cells are required to obtain a GVH reaction of a given magnitude ( 2 2 7 ) . Finally, the relationship between intensity of the GVH reaction and the magnitude of the allogeneic effect can be illustrated by varying the strength of histocompatibility differences between the strains of guinea pigs employed. Thus, when lymphoid cells from a random-bred line of guinea pigs, the NIH strain, are used to elicit the allogeneic effect in strain 2 recipients, a much higher cell number (1.0 X lo9) is required to obtain the same magnitude elicited by a moderate number (0.200 X lo9) of lymphoid cells from strain 13 donors ( 2 2 8 ) . 3. A very critical time relationship exists between the transfer of allogeneic cells and the administration of a secondary antigenic challenge with respect to the elicitation of an enhanced secondary antibody response. The optimal time interval between cell transfer and challenge is

54

DAVID H. KATZ AND BARU J BENACERRAF

6 days. Administration of antigen either at earlier or later times after transfer may result in little or no increase, or even some suppression in antibody production (225, 226). The relevance of the allogeneic effect with respect to understanding the regulatory function of T cells in antibody production derives from the observation that during the peak period of the GVH reaction the operation of normal helper T cell function is no longer required for anti-DNP secondary responses. Two possible explanations for this phenomenon may be entertained: (1) the GVH reaction may lead to a general proliferative response within the host T cell population among which are contained normally occurring carrier-specific helper cells, and active proliferation of these cells as a result of the GVH reaction could result in sufficient numbers of helpers available at the time of the secondary DNP challenge; ( 2 ) the GVH reaction may exert a facilitative effect of some sort on B cell precursors of antibody-forming cells and, perhaps, on antibody-forming cells themselves, which results in a fundamental change in such cells with respect to their ability to respond or, alternatively, on the nature of their response to antigen. It should be noted that we refer to B cell precursors of host origin, not of donor origin, since it has been established that the anti-DNP antibodies are made by cells of the host (226). It is the second of these possible explanations which is most consistent with the data and which places the greatest relevance on the allogeneic effect phenomenon insofar as normal regulatory function of T cells is concerned. The strongest evidence in support of an effect of the GVH reaction on B cells derives from the following two observations: ( 1 ) haptenspecific memory, presumably reflecting a B cell memory, is markedly enhanced when antigenic challenge is appropriately timed during later stages of the GVH reaction (226); and ( 2 ) antibody responses can be elicited with antigens for which presumably no T cells exist (or are nonfunctional ) when such antigens are administered during the GVH reaction (153). The first observation was made in the guinea pig allogeneic transfer system (226). As we mentioned earlier, secondary DNP-BGG challenge at very early or later (than 6 days) times after allogeneic cell transfer consistently resulted in little or no secondary anti-DNP response. Nevertheless, a striking change had occurred in the memory cell population; when such animals were subjected to a final antigenic challenge with the original immunizing antigen, DNP-OVA, the magnitude of anti-DNP antibody responses was always higher in recipients of allogeneic cells than respective controls (not receiving allogeneic cells ) . Hence, it appears that the allogeneic effect enhances not only antibody production

REGULATORY INFLUENCE OF ACTIVATED T CELLS

55

but also the development of memory in the host B lymphocyte population as well. The second observation concerns the induction of anti-DNP responses with DNP-D-GL (153).As described earlier, the DNP conjugate of this copolymer is nonimmunogenic and, indeed, highly tolerogenic in both strain 13 and strain 2 guinea pigs. Yet, when DNP-OVA-primed strain 13 guinea pigs were given allogeneic lymphoid cells and then challenged 6 days later with DNP-D-GLor DNP-BGG, striking secondary anti-DNP responses were obtained in both cases (Fig. 5). It is a fair assumption that, particularly insofar as strain 13 guinea pigs are concerned, few or no D-GL-specific T cells capable of performing helper function exist in these animals. Any manipulation which would permit the development of an anti-DNP antibody response to DNP-D-GL must reflect the exertion of some influence on B cells or their progeny, and this is precisely what we interpret to occur during the allogeneic effect. The capacity of B cells to be triggered by DNP-D-GL in the presence of the allogeneic effect exemplifies, therefore, the case we are making for T cell regulation of B cell function via nonspecific mediators. Since administration of DNP-D-GL to guinea pigs ( 153) or mice (228) not undergoing a GVH reaction leads to profound DNP-specific tolerance, it would appear that the T cell mediators play a very crucial role in B cell triggering. The allogeneic effect phenomenon is not restricted to guinea pigs. Hirst and Dutton (90)have reported that small numbers of allogeneic nonadherent spleen cells from normal donors enhanced the primary in uitro anti-SRBC response of spleen cells from neonatally thymectomized or normal mice. Schimpl and Wecker (96) have found that mouse spleen cells treated with anti4 serum and complement can be restored to respond in uitro to SRBC by addition of allogeneic thymocytes capable of recognizing the T-deprived spleen cells as histoincompatible. EkpahaMensah and Kennedy (229) reported enhancement of the primary in vitro anti-SRBC response of normal mouse spleen cells by the addition of irradiated allogeneic spleen cells. Recently, we have elicited an in vivo allogeneic effect in mice ( 2 3 0 ) .Thus, DNP-KLH-primed CAF, mice injected intravenously with spleen cells from normal parental A strain donors displayed markedly enhanced secondary anti-DNP antibody responses to a challenge with DNP-BGG administered 6 days after cell transfer. The phenomenon has also recently been demonstrated in mice by Ordal and Grumet (231) working with C,H( Hzklk), congenic C,H * Q( HDq/q),and F, ( Hzk/q) mice which are all genetic nonresponders to (T,G)-A--L. These nonresponders normally produce only 19 S antibodies to one or more challenges with (T,G)-A--L (163). However, non-

56

DAVID H. KATZ AND BARUJ BENACERRAF

responder F, ( HZk/q)mice displayed enhanced antibody responses, and of the 7 S class, when nonresponder parental C,H( HzkIk)lymph node and spleen cells were injected intravenously on the same day as antigen administration ( 231 ). This observation precisely confirms our findings with DNP-D-GL in guinea pigs (153) and reflects the release of T cell mediators during the GVH reaction which permit or facilitate triggering of B cells by antigen. There are other recently reported findings which we interpret as reflecting phenomena related to the allogeneic effect insofar as the concept of T cell factors regulating B cell function is concerned. The one which is most relevant in this regard is that of Hartmann (91 ) in studies of the in vitro antierythrocyte response of mouse bone marrow-derived cells. The experimental design has been described in detail in Section II1,A. In brief, he found that the capacity of B cells to respond in vitro to SRBC could be restored by the addition of educated T cells. When only one in uitro immunogen was employed, the response of B cells occurred only if the T cells had been specifically educated to that immunogen. Thus, B cells developed an in uitro antibody response to SRBC when SRBC-educated T cells were added but not when HRBC-educated T cells were used instead. However, if both SRBC and HRBC were added as in uitro immunogens to mixtures of B cells with SRBC- or HRBCeducated T cells, then PFC specific for both erythrocytes were produced. Hence, T cell stimulation by the specific antigen to which the population had been primed permits triggering of B cells to develop an antibody response to another, non-cross-reacting antigen. We consider this observation to be analogous in interpretation to the allogeneic effect phenomenon. Other studies which we interpret in a similar fashion include: (1) the observation by Sulica et aE. (232) that suspension cell cultures of spleen cells from rabbits immunized with DNP-BSA produced antiDNP antibodies when exposed in vitro to the carrier, BSA, alone; ( 2 ) the report by Rubin and Coons (233) that spleen cells from mice which had previously been immunized with an antigen, such as tetanus toxoid or ovalbumin, developed significantly enhanced primary in vitro antibody responses to SRBC if the priming antigen was also added to the culture; ( 3 ) the observation by Storb and Weiser (234) that adoptively transferred spleen cells developed higher anti-SRBC responses in irradiated allogeneic recipients than in irradiated syngeneic recipients; and (4)the studies of Moller (235) on the suppressive effect of the GVH reaction on primary anti-SRBC responses in mice, showing that, in certain H-2 histocompatibility combinations, transfer of parental spleen cells into adult F, recipients resulted in enhanced rather than suppressed antibody responses.

REGULATORY INFLUENCE OF ACXIVATED T CELLS

57

In conclusion, we propose that the normal role of T cells in antibody responses is to exert a regulatory influence on B cell precursors of antibody-forming cells with respect to their reactivity to antigen. This regulatory influence is required or not required depending on the physicochemical state and quantity of the antigenic moiety when it is capable of interacting with the specific B cell. Most of the antigens with which we are familiar require the existence of a regulatory influence for optimal, or any, B cell response to become manifest; only a few molecules with unique physicochemical structural features appear to act on B cells independently of such a regulatory influence (see Section IV,A). The regulatory influence of T cells on B cells is, we believe, mediated through soluble factors which are not yet defined. These soluble factors are not antigen-specific and most likely act rapidly and have a short half-life. Although the factors themselves may be nonspecific, the T cells which synthesize and secrete them are, indeed, antigen-specific. We propose that under normal circumstances specific antigenic stimulation of T cells must occur to signal the elaboration of mediators by such cells. The release of such factors in the microenvironment of specific B cells exerts a regulatory influence on these cells insofar as their immediate response to antigen is concerned. For the sake of simplicity of the model, we shall consider, at this point, only a positive regulatory effect of such factors on B cells, i.e., the factor facilitates or permits triggering of the B cell by antigen. The possible existence of T cell mediators with the opposite effect will be dealt with in a later section. This model is particularly appealing to us because it allows for specificity restriction among T cells and does not create an unlikely or unique functional requirement upon such cells, i.e., as discussed earlier, we know that T cells responsible for other immune reactions perform their role via released mediators. It is most logical then to assume a similar situation exists for T cells performing helper function in antibody responses. Indeed, in a recent paper Jimenez et al. (235a) suggest that carrier-specific helper T cells are identical to delayed-type hypersensitivity T cells. Furthermore, the model relieves the T cell from an unlikely passive participation serving only to concentrate and present antigen to B cells, which, in essence, represents a wasteful duplication of macrophage function. 2. Soluble Factors Released by T Cells in Culture Several investigators have reported findings suggesting the release of a soluble factor from T cells active in antibody production. Haskill et al. (200) studying in vitro anti-SRBC responses of density-gradientseparated mouse spleen cells found that cells of the light-density region

58

DAVID H. KATZ AND BARUJ BENACERRAF

of the gradient could be stimulated to respond to SRBC in culture by addition of cells from the heavier-density region or by thymus cells even when the two populations were separated by a dialysis or Nucleopore membrane. The effect, however, was quite variable. Kennedy et al. (236) reported that a supernatant obtained from peritoneal exudate lymphocytes exposed to gentle heating was active in permitting B cells to develop an antiburro erythrocyte (BRBC) response upon adoptive transfer to irradiated recipient mice. This heated cell supernatant was antigenspecific but not species-specific, i.e., the supernatant was active only if obtained from cells of donors specifically immunized with BRBC, but either mice or rats could serve as peritoneal exudate cell donors. Furthermore, the supernatant itself was not immunogenic. The authors suggest that the active factor is most likely some form of antibody, but no studies have been reported on physicochemical characterization of the supernatant. Dona et al. (237) were able to demonstrate that a cell-free medium of thymus cell cultures were effective in restoring the primary in vitro anti-SRBC response of spleen cells from neonatally thymectomized mice. Thus, thymus cells were incubated alone or in the presence of SRBC for 24 to 40 hours; the medium obtained from such cultures was separated from all cells by centrifugation and then added to spleen cells of thymectomized mice in culture with SRBC. The capacity of such cells to respond to SRBC was restored by addition of the T cell medium irrespective of whether or not T cells had been exposed to SRBC in vitro. Gorczynski et al. (238) have also successfully restored the capacity of T cell-depleted mouse spleen cells to develop primary in vitro anti-SRBC responses with a soluble factor released by cultured T cells. In their system, SRBC-educated T cells (derived from spleens of irradiated recipients of syngeneic thymus and SRBC) were cultured for 3 days with SRBC, after which they were gently heated (48°C X 30 minutes) and removed by centrifugation. This heated cell culture supernatant was active in reconstituting the response of mouse spleen cells treated with anti-8 serum and complement to approximately 50%of the normal control response. In contrast to the results of Doria et al. (237), active supernatants were obtained only from T cell cultures containing SRBC. Moreover, the active factor was not dialyzable in contrast to results of Haskill et al. (200). In studies of the enhancement of in vitro antibody responses by the allogeneic effect, investigators in at least three laboratories have preliminary evidence for a soluble factor active in these systems, Thus, Dutton et al. (239) have obtained a cell-free supernatant from 24-hour mixtures of allogeneic spleen cells which can enhance the in vitro anti-

REGULATORY INFLUENCE OF ACTIVATED T CELLS

59

SRBC response of both normal and T cell-deprived spleen cells. EkpahaMensah and Kennedy (229) reported enhancement of primary in vitro anti-SRBC responses of normal mouse spleen cells by addition of irradiated histoincompatible spleen cells. Moreover, they state that the anti-SRBC response of spleen cells can be stimulated by mixtures of allogeneic lymphoid cells on the opposite side of a 0.5-pm. Nucleopore membrane. Similar preliminary observations have been made in our laboratory in studies of the secondary in vitro anti-TNP antibody response of mouse spleen cells cultured in double-chamber vessels separated by 0.2-p. Millipore filters ( 2 2 8 ) .Thus, the anti-TNP secondary response of spleen cells on one side of the vessel can be enhanced by the appropriate mixture of allogeneic cells on the opposite side of the Millipore filter (Table VII). It must be stressed that all of the above observations should be interpreted with some degree of caution. In vitro antibody responses may be subject to wide variability determined by many unknown factors concerning culture conditions. 3. Nonspecific Stimulution of T Cell Function

In this category we shall discuss the agents that exert a stimulatory influence on T cells irrespective of the antigen specificity of their receptors. A variety of substances are known to do this, perhaps chief among them being the mitogenic agents phytohemagglutinin (PHA ), pokeweed mitogen, and concanavalin A (Con A). It is known that some plant mitogens can induce in lymphoid cells many of the morphological changes characteristic of antigenic stimulation (240, 241 ). It is generally accepted that specific receptors exist for these substances on the surface membrane of T lymphocytes. However, in contrast to antigen-specific receptors, no clonal restriction exists for these mitogens and, therefore, most if not all T cells can be stimulated by them. There are several reports in the literature demonstrating an enhancing effect of mitogens in antibody production. Tao (242) observed that lymph node fragments from rabbits primed with BSA or human chorionic gonadotropin developed anamnestic antibody responses in vitro when exposed transiently to PHA at the beginning of culture. Makela and Pasanen (243) found that NIP-CGG primed mouse spleen cells adoptively transferred to syngeneic irradiated recipients could be stimulated to some extent to produce anti-NIP antibodies by the intraperitoneal administration of pokeweed mitogen. The degree of stimulation by the mitogen, although only 1/ 1500 the magnitude of the antigen-induced secondary response, was significantly higher than nonstimulated control values. It is highly possible that in both of these models, small traces of priming antigen were carried over to either the in vitro culture or to the

6,

0

EFFECT O F ALLOGENEIC INTER-4CTIONS No antigen chamber

Spleen cellsb

Anti-TNP PFC/IOO (direct)

TNP-burro

731 970 598

3

TNP-bumo

526 684 557

5

TNP-burro C57B1

220 93

1

+

TABLE VII O N HAPTEN-SPECIFIC I M M U N E RESPONSES in vitrOD

Geom. Ave. 751+--

TNP-Burro RBCc chamber

Spleen cells

2

0

585 +--+

4

TNP-burro

1.43

6

TNP-bumo

-

Anti-TNP PFC/lW

Geom. Ave.

4

0 0 0

0

1480 1320 928

1219

2440 2110

U

sU

9 w >

2253

~

This table represents data from our laboratory (228) which illustrates the capacity of allogeneic cell interactions on one side of a cellimpermeable membrane (chamber 5) to enhance the specific antihapten response of cells on the other side of the membrane (chamber 6) most probably as a result of active soluble mediators. Cells were cultured in doublechambered vessels separated by ultrathin Millipore filters (pore size, 0.20 p ) . Opposing chambers are indicated by arrows (-). Numbers of anti-trinitrophenyl (TNP) plaque-forming cells (PFC) in individual cultures and geometric averages of respective culture groups are presented. Shtistical analysis by student’s t test: comparison of chambers 4 and 6 yielded a P value of 0.05 > P > 0.02. * Dissociated spleen cells were obtained from BALB/C mice primed 7 days earlier with 1 X lo8 heavily conjugated (TNP)-burro erythrocytes and from nonimmune C57B1/6N mice; 1 X lo7 cells/ml. were cultured in each chamber. Chamber 5 contained 0.5 x 10’ of each cell type. Chambers 1, 3, and 5 contained primed cells without in vitro antigen, whereas chambers 2,4, and 6 contained primed cells plus in vitro immunogen. In vitro immunogen was 1 X 105 heavily conjugated TNP-burro erythrocytes.

8

4 2: tr

m

$

REGULATORY INFLUENCE OF ACTIVATED T CELLS

61

irradiated recipient. Such small quantities of antigen, although insufficient on their own to induce antibody formation, can do so in the presence of a sufficient concomitant T cell stimulation which is exerted by the added nonspecific mitogen. An intriguing observation of Makela and Pasanen ( 2 4 3 ) which they could not explain was that incubation of primed donor cells at 37°C in vitro for a short time before adoptive transfer resulted in a significant synthesis of anti-NIP antibodies in recipients in the absence of any antigenic or mitogenic stimulation; this did not occur if the cells were kept at 0°C in vitro. In the context of recent observations on movement of B cell membrane receptors (244,245), it is likely that at 37°C stimulation of B cells by small traces of antigen in the suspension could have occurred in vitro prior to transfer. More recently, Rich and Pierce ( 2 4 6 ) have significantly enhanced in vitro responses to SRBC with Con A. Thus, submitogenic doses of Con A either administered in vivo to donor mice or added to spleen cell suspensions in vitro results in an increase up to tenfold in the magnitude of the in vitro anti-SRBC response. Recent observations of Katz and Unanue ( 2 4 7 ) are also relevant in this regard. When spleen cells from HRBC-primed mice are cultured in vitro in the presence of small quantities of ALS they develop antigenspecific yM and yG PFC in vitro responses in the presence of small quantities of soluble HRBC immunogen. In addition, there is also an enhancement of the secondary in vitro response of such cells to particulate HRBC antigen when cultured in presence of ALS. The action of ALS appears to be a stimulatory influence on T cells since the effect is abrogated by treatment of donor spleen cells with anti4 serum and complement. A further observation of considerable interest is that transient exposure of the HRBC-primed spleen cells to anti-Ig or anti-Fab serum prior to culture significantly increases the magnitude of both soluble antigen-induced antibody production and enhanced particulate antigen-induced secondary response resulting from ALS activity. However, exposure of the cells to only anti-Ig or anti-Fab has no effect ( 2 4 7 ) . We interpret the above data ( 2 4 7 ) as suggesting the following sequence of cellular events leading to antibody production: T and B lymphocytes can be stimulated to perform their respective functions in antibody production by an appropriate immune reaction on their surface membranes, usually related to the clonally-restricted antigen-specificity of their receptors. However, an additional immune reaction of anti-Ig or anti-Fab with Ig receptors on B cells may significantly enhance the triggering event of such cells to produce their specific antibodies, provided an essential regulatory influence exerted by mediators elaborated by T cells is operating. It would appear also that T cells can be stimulated to

62

DAVID H. KATZ AND BARU J BENACERRAF

release their mediators by the occurrence of a similar reaction on their membrane as shown by the enhancing activity of ALS on these cells. VII. Suppressive Effects of

T Cells on Antibody Synthesis

Heretofore our discussion has been concerned with the positive regulatory influence of T cell function of B cell precursors of antibodyforming cells. It is rapidly becoming clear that the regulatory function of T cells consists also of a negative aspect with regard to antibody production. This need not imply the existence of distinct T cells responsible for the respective roles of either enhancing or suppressing B cell function. It is equally possible that the same T cell may express either of these regulatory influences via the same or distinct mediators, the difference in outcome being determined by (1) the relative concentration of T cell products released in the microenvironment of B cells, ( 2 ) the range at which these mediators are active and their respective half-lives, and (3) the amount and physicochemical properties of effective antigen. In this section we shall consider the evidence suggesting the existence of a suppressive T cell influence on antibody formation under the following experimental conditions: ( 1) enhancement of immune responses by measures which diminish the number of T cells and ( 2 ) suppression of antibody responses by the administration of more than one antigen (antigenic competition).

A. ENHANCEMENT OF IMMUNE RESPONSES BY DEPLETION OF T CELLS Several studies have clearly demonstrated that in vivo administration of ALS or antilymphocyte globulin ( ALG) may cause enhanced antibody responses to subsequent immunization with antigen. Baum et al. (248) found that rats pretreated with lymphocytopenic doses of ALG developed primary antibody responses to KLH that were sixteen-fold higher in magnitude than saline or normal rabbit globulin-treated controls. In contrast, the same dose of ALG abrogated the capacity of another experimental group to develop primary anti-SRBC responses. Baker et d.(156) reported that mice treated with ALS developed primary PFC responses to Type 111 pneumococcal polysaccharide ( SIII) which were ten-fold higher in magnitude than controls not treated with ALS, although there was no significant difference in serum hemagglutinin titers between the nontreated and ALS-treated animals. Previous studies have demonstrated that the response of mice to SIIIconsists of the development of two types of y M PFC in equal numbers, i.e., direct PFC and indirect PFC which can be developed only with appropriate anti-yM sera ( 249). It is interesting that Baker et al. (156) found that the increased PFC responses in ALS-treated mice were restricted to the direct PFC.

REGULATORY INFLUENCE OF ACTIVATED T CELLS

63

The usual effect of ALS treatment in vivo is to depress the humoral response to a variety of antigens (250). Its immunosuppressive activity has been shown to be due to a preferential elimination of thymus-derived recirculating small lymphocytes ( 251 ) . The implication of the studies cited above (156, 248) is that under certain circumstances removal or diminution of T cells permits greater antibody production that occurs in their presence. This reasoning is easily extended to the concept that T cells exist which exert a suppressive regulatory influence on B cells. Alternatively, we must consider also the possibility that the enhanced antibody responses observed following ALS treatment reflects the release from destroyed cells of large amounts of nucleic acids which may then cause an adjuvant effect such as we shall discuss in Section VII1,C with respect to synthetic polynucleotides. This possibility has been cited to explain the enhanced antibody production sometimes observed following whole-body X-irradiation ( 252, 253) or the administration of cytotoxic agents ( 254-256). A final possibility to explain the enhancing effect of ALS is that this reagent may be exerting a stimulatory influence on T cell function. It is well established that in the appropriate dose range ALS may have a marked stimulatory effect on lymphoid cells. Many studies have shown that ALS stimulates blast transformation and DNA synthesis of lymphoid cells in culture (257-261 ). Moreover, ALS has been found to enhance response of lymphoid cells to other antigens in culture. Thus, Greaves et al. (261) observed that addition of submitogenic doses of ALS to cultures of PPD-sensitive human leukocytes significantly enhanced the DNA synthetic response of such cells to PPD. More recently, as reported in Section VI,C, Katz and Unanue (247) have demonstrated that the response of primed mouse spleen cells to antigen in vitro can be markedly enhanced by the exposure of such cells to small quantities of ALS and antiimmunoglobulin. It is possible that a similar mechanism operating in vivo could explain the results of Baum et aZ. (248) and Baker et al. (156). However, we are inclined to believe that this is not the explanation for their findings and favor the explanation based on elimination of T cell function. Recent observations which support this explanation will now be given. Baker et aZ. (262) have recently extended their observations of ALS enhancement of antibody responses to SrIr in mice. The experiment was designed to study the effect of various types of passively transferred syngeneic lymphoid cells on the enhanced responses to SrII displayed hy ALS-treated mice. They found that passively transferred thymocytes suppressed the anti-SIIIresponse, whereas peripheral blood lymphocytes further enhanced the response. Spleen cells had no effect. The authors

64

DAVID H. KATZ AND BARUJ BENACERRAF

interpret these data as indicating the existence of two functionally distinct cells-one in the thymus which suppressed the response and another in peripheral blood which amplifies the response. Perhaps analogous is the observation of Armstrong et al. (147) who found that the addition of thymic lymphocytes to thymus-deprived mice depressed the antibody responses of such mice to the thymus-independent antigen, POL. Perhaps more intriguing are the observations of Okumura and Tada (263) in studies on the formation of homocytotropic antibody (HTA) in rats immunized with DNP-Ascaris (As). These investigators found that thymectomy or splenectomy in adults 3-10 days before primary immunization with DNP-As resulted in greatly enhanced and prolonged HTA titers in such rats compared to controls, A similar result was obtained with ALS administration ( 2 6 4 ) .These manipulations had no effect on the yM and yG anti-DNP antibody responses, however. The data have added relevance inasmuch as neonatally thymectomized rats are markedly reduced in their capacity to develop anti-DNP HTA responses to DNP-As ( 2 6 3 ) .In the same system, it has also been observed that totalbody X-irradiation (265) and certain immunosuppressive drugs (266) cause enhanced HTA production, although the interpretation of these data is complicated by additional considerations concerning the effects of such treatments. The enhancing effects of thymectomy provide the clearest support of all for the possible existence of a suppressive regulatory T cell function. This is further strengthened by published observations of Okumura and Tada (267) that passively transferred thymocytes from hyperimmune donor rats terminate preexisting anti-DNP HTA production in recipients. Finally, Kerbel and Eidinger (268) recently reported results of a very detailed study on the effects of adult thymectomy and/or ALS administration on primary antibody responses of mice to a variety of antigens. On the one hand, the antibody responses to erythrocyte antigens of mice treated with ALS alone or by a combination of adult thymectomy plus ALS were markedly suppressed; the responses to KLH and PVP, on the other hand, were considerably enhanced by treatment with ALS alone or combined with adult thymectorny. These authors monitored the serum antibody levels over 25 days, in contrast to Baum et at. (248), and observed that after 15 days the treated mice manifested a suppression in their anti-KLH titers as compared to controls. Moreover, the enhancement of antibody production by ALS alone or in combination with thymectomy was restricted to anti-KLH antibodies of the yM class ( 2 6 8 ) . Indeed, as the antibody response shifted from the yM to the yG phase, an actual suppression could be attributed to the treatment. This brings up a very important point to consider inasmuch as all of the examples of

REGULATORY INFLUENCE OF ACTIVATED T CELLS

65

ALS enhancement described above, with the exception of HTA formation in rats, have the common feature of being related to synthesis of antibodies predominantly of the yM class. It may well be that the suppressive regulatory T cell function is related to promotion of early yG antibody synthesis which, in turn, exerts a feedback type suppression on yM antibody synthesis. OF ANTIBODYRESPONSES BY THE ADMINISTRATION OF B. SUPPRESSION MORETHANONEANTIGEN ( ANTIGENICCOMPETITION )

The phenomenon of antigenic competition, first recognized by Michaelis (269) three-quarters of a century ago, occurs when an animal is immunized with two antigens simultaneously or in close sequence resulting in a depressed antibody response to one or both antigens [for review, see Adler (270)l. The underlying mechanism has been the subject of much study but is still not known. However, recent investigations have provided evidence which favors the existence of soluble factors acting in this phenomenon. It is relevant, therefore, to consider certain aspects of the phenomenon in this review, since we believe that such soluble inhibitory factors responsible for antigenic competition are most likely elaborated by T cells and are related, if not identical, to the T cell mediators active in other regulatory influences on antibody production. Evidence for the existence of a soluble factor active in antigenic competition was first advanced by the studies of Radovich and Talmage (271). They demonstrated that injection of mice with HRBC several days before primary immunization with SRBC resulted in a suppressed anti-SRBC response; no suppression occurred when the two antigens were administered simultaneously. Furthermore, they demonstrated that competition between antigens given to irradiated recipients of spleen cells increased proportionately with the number of spleen cells transferred. Moller and Sjoberg (272) confirmed and extended these observations. Employing an adoptive transfer system in mice, they found that spleen cells from HRBC-primed donors developed adoptive primary anti-SRBC responses in irradiated recipients which were comparable in magnitude to the responses obtained with normal donor cells, indicating that the number of antigen-sensitive cells to SRBC was not appreciably affected by the HRBC priming. They also demonstrated that adoptively transferred normal spleen cells into irradiated recipients which had previously been primed with HRBC resulted in marked suppression of the response to SRBC given at the same time as cell transfer (272). Analogous results were obtained when the transferred spleen cells were derived from donors which had been preimmunized with SRBC. These findings indicate that differentiation and proliferation of precursors into antibody-

66

DAVID H. KATZ AND BARUJ BENACERRAF

forming cells have been impaired, The same observation has been reported by Waterston (273). Brody and Siskind (274) studied the phenomenon in a haptenic system in rabbits. They found that competition between two haptens was equally pronounced whether haptens were attached to the same or different carriers-this argues against competition for a common antigensensitive cell. Moreover, the affinity of the antibody produced in antigenic competition is approximately equal to the antibody produced by animals immunized with only one antigen, suggesting that neither tolerance nor antibody suppression operated in the phenomenon. They also observed that antigenic competition only occurred when both antigens are injected so as to drain into the same regional lymph node (274). The latter point is of some controversy since it superficially conflicts with results of other investigators (275-277). However, differences in route and mode of immunization may explain the discrepancies in some instances (275), whereas the method of determining the immune response may explain others (276,277). Hence, Fauci and Johnson (276, 277) studied antigen competition by determining the PFC response in regional lymph nodes of rabbits. They found that local lymph node PFC responses to p-arsanilic acid ( An)-KLH injected into front and hind footpads on one side were significantly suppressed by simultaneous injection of TNP-KLH into the contralateral footpads. They interpreted these observations as indicating the presence of a circulating inhibitory factor. It is not possible, however, to determine whether or not their data reflect a migration of the specific PFC out of the regional lymph nodes. They further observed a diminished anti-TNP PFC response in local lymph nodes of rabbits that had been preimmunized with soluble KLH prior to footpad immunization with TNP-KLH (277). This finding may be analogous to the “preemption” phenomenon of O’Toole and Davies (278) whereby mice immunized with SRBC intraperitoneally manifest anti-SRBC PFC responses which were depressed in the local lymph nodes draining the site of a subsequent subcutaneous injection of SRBC. In our own laboratory, we have recently observed marked inhibition of secondary anti-DNP responses in guinea pigs which have been simultaneously immunized with two unrelated carrier molecules (279). Thus, DNP-OVA-primed guinea pigs supplementally immunized with BGG in adjuvant develop markedly enhanced secondary anti-DNP responses to DNP-BGG secondary immunization. However, this enhanced secondary response can be either diminished or abolished by the following manipulations of this system: ( 1 ) simultaneously administering KLH with BGG in the supplemental immunization diminishes the anti-DNP response by 91%; ( 2) administering soluble BGG intraperitoneally 1-2 days before secondary DNP-BGG challenge completely abolished the response re-

REGULATORY INFLUENCE OF ACTIVATED T CELLS

67

gardless of whether BGG alone or BGG together with KLH were used in the supplemental immunization; ( 3 ) administering soluble KLH 1-2 days before secondary challenge with DNP-BGG inhibits the response by 98%,provided KLH was administered simultaneously with BGG in the supplemental immunization. Perhaps most relevant was the finding that a nonimmunogenic amino acid copolymer, D-GL, failed to cause any inhibition of the anti-DNP response as was observed with the immunogens KLH and BGG ( 2 7 9 ) . We interpret these observations to indicate the existence of soluble inhibitory mediators in the phenomenon and, furthermore, that such mediators are most likely elaborated by carrier-specific T cells. The fact that the greatest degree of suppression occurred when the specific carrier (BGG) was used may argue against an explanation based on a general shift in frequency of specific helper T cells such as recently proposed by Kerbel and Eidinger ( 2 8 0 ) . The best evidence for a critical role of T cell-released inhibitory substance derives from the recent studies of Gershon and Kondo (281, 282). These authors found that the phenomenon of antigenic competition was thymus-dependent. Thus, antigenic competition elicited by sequential immunizations with SRBC and HRBC did not occur in thymus-deprived mice. Moreover, passively transferred normal thymocytes restored the competition phenomenon and the magnitude of inhibition was related to the numbers of thymocytes transferred (281).They also observed that antigenic competition developed irrespective of circulating antibodies specific for the competing antigen ( 2 8 2 ) .They interpreted their findings as strongly suggestive of an inhibitory factor released by T cells which seems valid in our opinion. There is even some very recent evidence that the inhibitory effect can be overcome to some extent by administration of polynucleotides ( 2 8 3 ) . VIII. Functions of T and B lymphocytes in Various Immunological Phenomena

A. IMMUNOLOGICAL TOLERANCE The establishment of T and B cell interactions in humoral immune responses demanded a reappraisal of the conceptual and experimental approach to the study of immunological tolerance. Since this review cannot consider at length the various parameters and details of tolerance induction, the reader is referred to a recent review on the subject by Dresser and Mitchison ( 2 8 4 ) .

1. Target Cells for Tolerance Induction The most immediate question concerns which cell, T or B lymphocyte, is the target for tolerance induction. It was clear very early that T cells could manifest a specific tolerant state and, thereby, a specificity restric-

68

DAVID H. KATZ AND BARU J BENACERRAF

tion in this population is shown. Thus, Isakovic et al. (285) demonstrated that thymus grafts from BGG-tolerant rats transferred the specific tolerance to thymectomized irradiated recipients. In mice rendered tolerant by repeated injections of SRBC, Gershon et al. (286) found a reduced T cell mitotic response to SRBC. Taylor (61, 287) demonstrated that T cells of mice made tolerant to BSA were unable to cooperate with normal B cells in an adoptive transfer response in irradiated recipients. Miller and Mitchell (288) induced tolerance to SRBC in mice with cyclophosphamide and observed that recirculating T cells in the thoracic duct lymph (but not in thymus) were specifically tolerant. Paul et al. (67) demonstrated that carrier-specific T cells could be rendered specifically tolerant thereby abolishing their helper function in secondary antihapten antibody responses. Whereas tolerance in T lymphocytes could be readily demonstrated, it was much more difficult to establish the existence of tolerance in B lymphocytes. In the experiments above of Taylor (61, 287) and Miller and Mitchell (288), B cells from tolerant animals were able to cooperate with normal T cells in adoptive transfer responses. Playfair (289) used higher doses of SRBC plus cyclophosamide to induce tolerance in mice and found that B cells were transiently tolerant. Gershon and Kondo (290) observed that B cells could be rendered tolerant by repeated injections of SRBC. Chiller et al. (62) found that tolerance existed in both B and T cells of mice rendered tolerant by a single injection of deaggregated HGG. More recently, techniques have been devised to induce a high level of hapten-specific tolerance which probably reflects a selective state of B cell tolerance (153-155, 291). The very elegant experiments of Chiller et al. (292) have elucidated the critical kinetic differences in tolerance induction of T and B lymphocytes and have placed a much clearer perspective on this matter. Their experimental design was based on the adoptive transfer response to HGG in irradiated syngeneic recipients with T and B lymphocytes from normal donors and/or donors made tolerant to HGG by a single injection of deaggregated HGG. They found a marked difference between T and B lymphocytes with respect to both kinetics of tolerance induction and the dose of tolerogen required. As shown in Fig. 6, T cells were rendered tolerant very early ( 2 days) after administration of the tolerogen and remained tolerant for a very long period (77 days), In contrast, B cells did not exhibit the tolerant state until later (11 days) after tolerance induction and recovered much earlier (49 days) than T cells. The crucial point is that the immune state of the whole animal reflected the immune state of the respective T cells. In other words, although at a given time the B cells were normally immunocompetent, the existence of tolerance in

REGULATORY INFLUENCE OF ACTIVATED T CELLS

-

69

100

n n Y

.-n c Y

-

0 Y n

c

= E

30

, i A-A

1

14

21

DH66 Injecttd Donor

28

35

42

49

Days Following Tolerogon

FIG.6. Kinetics of tolerance induction and spontaneous loss of unresponsiveness in thymus and bone marrow cells. Mice were rendered tolerant with deaggregated human y-globulin (DHGG) administered as a single intravenous dose of 2.5 mg. At various times thereafter, groups were sacrificed and suspensions of their thymus or bone marrow cells were injected intravenously along with normal bone marrow cells or thymus cells, respectively, into lethally irradiated recipients. At the same time and again 10 days later these recipients were challenged with 0.4 mg. of aggregated HGG, and 5 days after the second injection their spleens were assayed for plaque-forming cells to HGG. The results are presented as percent unresponsiveness in the respective cell types. [Taken from Chiller et al., Science 171, 813, 1971 (reference 292); we thank Drs. Chiller, Habicht, and Weigle for permission to reproduce this figure which they made available to us.]

the T cell population reffected a tolerant state of the whole animal. Moreover, tolerance induction was found to occur at much lower doses of tolerogen in the T cell population than in the B cell population (Fig. 7 ) . These observations (292) clearly show that both B and T cells can be rendered immunologically tolerant and also explain the apparent discrepancies in earlier studies.

2. Fate of Tolerant Cells What happens to a tolerant cell? Does it exist in a functionally unresponsive or unrecognizable state or is it eliminated from the system? Here again, the available evidence may appear conflicting in that some investigators have demonstrated specific antigen-binding cells still present in relatively normal amounts in tolerant animals (7, 293-295), whereas others have observed diminished numbers of such cells (153, 170, 296). Perhaps the most logical explanation for these differences can be based on the different target cells involved in the various systems studied and the degree of specific tolerance existing at the time cells are

70

DAVID H. KA'IZ AND BARUJ BENACERRAF

0

1

WT

NBM

I

TT

NBM

NT TIM

FIG.7. The effect of varying the dose of tolerogen on the induction of unresponsiveness in thymus and bone marrow cells. Each point represents the arithmetic mean of the individual response [plaque-forming cells ( P F C ) ] obtained in 6 mice. NT, NBM-thymus and bone marrow, respectively, obtained from normal donors; TT, TBM-thymus and bone marrow, respectively, obtained from tolerogen-injected 0.5 mg. ( B ) , or 2.5 mg. ( A )of deaggregated donors which received 0.1 mg. ,).( Both thymus and bone marhuman gamma globulin 11 days prior to sacrifice. (0) row cells obtained from normal donors. [Taken from Chiller et al., Science 171, 813, 1971 (293); we thank Drs. Chiller, Habicht, and Weigle for permission to reproduce this figure which they made available to us.]

examined. Hence, where tolerance exists predominantly among T cells, it would be expected to find relatively normal numbers of specific antigen-binding cells (presumably representing B cell precursors of antibody-forming cells), Another important factor is the existence of antigenbinding cells in a tolerant animal in which some degree of B cell tolerance exists; in such a circumstance the antigen-binding cells may be predominantly of low-affinity receptor type, whereas the high-affinity cells may be significantly diminished. Previous studies have demonstrated, indeed, that induction of tolerance results in preferential diminution of high-affinity antibodies (297, 298). Ideally, studies of this type should be employed in a model where tolerance is more or less restricted to the B cell population. Since the threshold of tolerance induction ir, T cells is indeed lower than it is in B cells (292), at least insofar as protein antigens are concerned, it is difficult to obtain a selective B cell tolerance in vivo once the T cells have already been tolerized. Recently, however, several investigators have reported the successful induction of hapten-specific tolerance which may reflect such a state of restricted B cell tolerance (153-155, 291, 299). We define hapten-specific tolerance as the specific suppression of antihapten antibody responses to conjugates of a given hapten resulting from

REGULATORY INFLUENCE OF ACTIVATED T CELLS

71

the prior admininstration of the haptenic determinant alone or of the hapten conjugated to an unrelated carrier. The induction of tolerance to hapten-carrier conjugates has been described by several investigators ( 117, 300, 301 ) where, in general, impaired antihapten antibody responses were restricted to the hapten-carrier conjugates used for tolerance induction. The tolerant animals displayed either no suppression or only transient decrease of antihapten responses when challenged with the hapten coupled to a carrier unrelated to that used in the paralysisinducing regimen. Such experiments (117, 300, 301 ), therefore, cannot be considered examples of hapten-specific tolerance since unresponsiveness in these circumstances may reflect tolerance in either or both T and B lymphocytes as has been shown recently (302). As we define it, haptenspecific tolerance should reflect a selective B cell tolerance. Induction of hapten-specific tolerance in vivo has been accomplished by administering hapten conjugates of essentially nonimmunogenic carrier molecules, at least as far as the T cells are concerned. Thus, Bore1 (291) induced DNP-specific tolerance in mice by administering DNP coupled to mouse serum proteins. Later experiments demonstrated that DNP-mouse 7-globulin was the tolerogenic moiety (154), confirming earlier studies of Havas (155). Borek and Battisto (299) induced tolerance to the p-azobenzenearsonate hapten in guinea pigs by administering the hapten coupled to autologous red cells. Also in guinea pigs, Katz et al. (153) found that administration of DNP coupled to nonimmunogenic amino acid copolymers resulted in a profound state of DNP-specific tolerance. This was achieved with DNP-D-GL in both strain 2 and strain 13 guinea pigs and with DNP-L-GL in the PLL nonresponder strain 13 animals. Thus, administration of DNP-D-GL in soluble form intraperitoneally to nonimmune guinea pigs resulted in a significantly depressed anti-DNP antibody response to a subsequent challenge to DNP-OVA. Moreover, administration of DNP-D-GL to guinea pigs that had been previously primed with DNP-OVA and were actively producing antiDNP antibodies resulted in abrogation of antibody synthesis and their capacity to respond to a secondary challenge with DNP-OVA (Fig. 8 ) . In both situations, the tolerance was hapten-specific as evidenced by normal anti-OVA antibody responses. Recently, several investigators have induced hapten-specific tolerance in vitro. Feldmann (152) induced DNP-specific tolerance in mouse spleen cells exposed in vitro to high concentrations of DNP-POL. Moreover, induction of tolerance was related to the degree of DNP molar substitution ( a high degree being required for tolerance induction). Rittenberg and Bullock (303) induced TNP-specific tolerance in primed mouse spleen cells with high doses of TNP-KLH in uitro. Naor and

72

DAVID H. KA'IZ AND BARUJ BENACERRAF

J C ~ U MnNTI-DNP

-

e-+

5a

ANTIBODY

No Intervening Immunization DNP-D-GL DNP-L-GL

0 1 2 3 4 WEEKS AFTER PRIMARY IMMUNIZATION

0

2

4

!I,

SERUM ANTI-OVA ANTIBODY

300

6

DAYS AFTER SECCNDARY IMMUNIZATION WITH DNP-OVA

FIG.8. 2,4-Dinitrophenyl-specifictolerance induced in DNP-ovalbumin ( OVA ) primed guinea pigs as a result of an intervening treatment with DNP-copolymer of D-glutamic acid and D-lysine (D-GL) or DNP-copolymer of L-glutamic acid and L-lysine ( L-GL). Strain 13 guinea pigs received a primary immunization with 3.0 mg. of DNP-OVA administered intraperitoneally in saline at week 0. Two weeks later, an intervening treatment with 3.0 mg. of either DNP-D-GL or DNP-L-GL in saline or saline alone was carried out. Four weeks after primary immunization, the animals were challenged with 1.0 mg. of DNP-OVA in saline. Serum anti-DNP and anti-OVA antibody concentrations just prior to challenge and on day 7 are illustrated. The numbers in parentheses refer to the number of animals in the given groups. [These data from our laboratory appeared in J . Erptl. Med. 134, 201, 1971 (reference 153).]

Mishell (304) specifically inhibited in vitro antihapten responses of mouse spleen cells to TNP-SRBC or penicillin-SRBC by the addition of the respective hapten coupled to isologous (mouse) erythrocytes in relative excess. The mechanism of DNP-specific tolerance in guinea pigs in our hands is a central one as evidenced (153) by (1) a significant diminution in DNP-specific PFC in the spleen and ( 2 ) a marked depression in the number of DNP-specific antigen-binding cells in peripheral blood, lymph nodes, and spleen. More recent extension of these studies of antigenbinding cells (296) have revealed that in DNP-tolerant animals there is a reduction predominantly of high-affinity B cell precursors and antibody-forming cells. Hence, although low-affinity cells are diminished earIy in the tolerant state, they reappear during recovery from tolerance much sooner than high-affinity cells (296). We should make perfectly clear that the observed diminution in numbers of DNP-specific antigenbinding cells does not necessarily mean that such B cell precursors are eliminated from the system. They may, indeed, be lost as a result of

REGULATORY INFLUENCE OF ACTIVATED T CELLS

73

tolerance induction or, alternatively, the failure to detect them in the antigen-binding assay may reflect a change in the capacity of these specific precursors to express their surface receptors, i.e., the receptor molecules may be pinocytized or no longer synthesized, or both. Hence, it is not possible at this time to know whether tolerance in specific B cells means that such cells are lost or whether they are still present but no longer express detectable surface receptor molecule and are, therefore, nonfunctional immunologically. The above discussion dealt with B cells. What about T cells in the tolerant state? The situation is, indeed, more complicated here because of the great difficulty in directly detecting antigen-specific T cells as compared to B cells. These technical problems render the determination of the existence of specifically tclerant T cells and their fate in vivo difficult to resolve at the present time. However, some recent observations provide indirect evidence that specifically tolerant T cells may exist (74, 290, 305, 306). 3. Mechanism

The studies cited above and others reported in recent years have increased our understanding of the kinetics and the cells involved in tolerance induction. Insofar as the precise cellular events resulting in specific unresponsiveness are concerned, however, we still have no definitive answers. Nonetheless, certain observations permit a few general predictions to be made. First, it is now readily apparent that the interaction of antigenic determinants and receptors of B cell precursors of antibody-forming cells may lead to very different outcomes determined by the differences in form and concentration in which the antigen is present or, particularly, by the state of the cell at the time of interaction. Hence, as shown recently by Klinman (307),spleen fragments from mice primed with DNPKLH could be stimulated to produce secondary anti-DNP responses in vitro with DNP-KLH over a wide range of molar DNP concentrations. When the spleen fragments were exposed to DNP conjugates of nonhomologous or nonimmunogenic carriers, he found that weak stimulation occurred only at very low molar DNP concentrations, whereas at higher concentrations such complexes had an inhibitory effect on anti-DNP antibody synthesis. This suggests that in the presence of a T cell regulatory influence both low and high determinant densities at the surface receptors provide a stimulatory signal to B cells, whereas in the absence of T cell regulation, low determinant density only provides a weak stimulatory signal and high density provides a tolerogenic signal. The same interpretation may be given to observations of Diener and Arm-

74

DAVID H. KATZ AND BARUJ BENACERRAF

strong (308) and Feldmann (152) who found that POL and DNP-POL, respectively, induced antibody production by mouse spleen cells exposed in uitro to low concentrations of antigen, whereas high concentrations induced specific tolerance. It is noteworthy that POL is a thymus-independent antigen (147).However, even in the presence of T cell activity, decreased B cell responsiveness may result from exposure to very high concentrations of antigen. The observation of Bullock and Rittenberg (309) that, with increasing time after in vivo priming with TNP-KLH, successively lower doses of TNP-KLH stimulated secondary anti-TNP responses in vitro whereas higher doses (which were stimulatory to cells obtained at earlier times after priming) suppressed the anti-TNP response can be interpreted in the same manner. The operational significance of determinant-B cell receptor interaction in the presence or absence of T cell regulation may be exemplified by some recent observations of our own (153).As we have mentioned before, administration of DNP-D-GL, a nonimmunogen, to normal or DNP-OVA-primed guinea pigs results in profound DNP-specific tolerance manifested by depressed antibody production and diminished numbers of detectable DNP-specific antigen-binding cells and anti-DNP antibody-producing cells. We have interpreted the ease of tolerance induction with these and other nonimmunogenic or weakly immunogenic substances (154,155, 291), for which few or no specific helper T cells exist, to result from direct interaction in the appropriate dose range with specific precursors of antibody-forming cells without intervening helper cells. This concept may explain the puzzling phenomenon described by some investigators studying the “termination” of tolerance to protein antigens by administration of modified or cross-reactive proteins (310-312). In those experiments, the simultaneous administration of even very small quantities of the tolerated protein prevented termination of the unresponsive state. It has been postulated that the mechanism in the termination of tolerance involves the presentation of the determinant of the tolerated protein to precursors of antibody-forming cells through the action of helper T cells specific for the cross-reactive or altered protein (292). In this set of circumstances, injection of the tolerated protein simultaneously with the cross-reactive protein should lead to a potent suppressive effect on the precursors of antibody-forming cells by direct interaction with their receptors in the absence of helper cells specific for the tolerated substance. If this reasoning is correct and if a suitable substitute for specific helper T cell function exists, then it should be possible to convert a normally tolerogenic signal into an immunogenic one. We have tested this hypothesis by employing the potent nonspecific stimulatory influence

REGULATORY INFLUENCE OF ACTIVATED T CELLS

75

of the allogeneic effect ( 153). Thus, DNP-OVA-primed guinea pigs were injected intravenously with allogeneic lymphoid cells and, 6 days later, challenged with either the immunogenic DNP-BGG conjugate or the normally tolerogenic conjugate DNP-D-GL. A comparable secondary anti-DNP response was obtained with both conjugates with respect to serum anti-DNP antibodies as well as anti-DNP PFC in the spleens of such animals (Fig. 5 ) . As discussed in Section V I , C , the elicitation of a GVH reaction renders the primed recipient, therefore, refractory to tolerance induction by DNP-D-GL and permits this molecule to behave as an immunogen capable of stimulating a strong anti-DNP secondary response. The observation of McCullagh (313) that adult rats tolerant to SRBC, nevertheless, are stimulated to form an anti-SRBC antibody response by administration of allogeneic immunocompetent cells and SRBC is analogous to the allogeneic effect described above. A second general prediction concerning tolerance induction can be made in light of the very elegant recent studies of Feldmann and Diener ( 314-316) concerning antibody-mediated suppression of in vitro immune responses to polymeric and monomeric flagellin. These investigators have provided impressive evidence that, in the absence of T cell function, mixtures of antigen and specific antibody in a very critical ratio lead to a profound state of specific tolerance in spleen cells exposed to the mixture in vitro. The authors interpret their data as suggesting that a lattice of antigen-antibody complexes on top of the specific antigen recognition receptor of immunocompetent cells provides the tolerogenic signal. Further experimentation is necessary to determine whether or not this interpretation is correct. Nevertheless, their observations clearly suggest a critical relationship between the extent and valency of determinant binding and the possible interpretation of the event by the specific cell, i.e., as a tolerogenic or as an immunogenic signal. If, as we postulate, it is at this level that the regulatory role of T cell mediators is effective, the prediction may be made that the tolerogenic conditions used by Feldmann and Diener should result in an antibody response in the presence of activated T cells.

B. IMMUNOLOGICAL MEMORY

The existence of specificity among both B and T cell populations implies that both cell types are capable of expressing immunological memory. This is, indeed, the case, as will be shown in this section. The existence of memory in the T cell population is implicit in observations made in the double transfer system response to SRBC in which T cells were specifically activated in the primary host (11,56, 74, 166) and in in vitro systems where activated T cells were employed (91, 92,

76

DAVID H. KATZ AND BARUJ BENACERRAF

94). Attempts to demonstrate B cell memory in the SRBC system, however, gave some conflicting results. The results of some investigators clearly indicate that T cells carry memory, whereas B cell memory was not so evident (166, 317, 318). Thus, Shearer and Cudkowicz ( l 6 6 ) , using the double transfer system in mice, found that activation of T cells in the first host by SRBC was necessary to obtain synergy with B cells in the second host, but the quality of the antibody response was the same whether or not the B cells had been exposed to SRBC in the primary host. Takahashi et al. (318) studied the adoptive secondary response to SRBC in mice and found that treatment of immune spleen cells with anti-8 serum and complement abolished the adoptive second response and that addition of normal thymus or spleen cells failed to restore the response. Addition of immune spleen cells to anti-0-treated spleen cells restored the adoptive second response and this was interpreted as evidence for memory being the responsibility of the @-bearingT cell population. However, by using as T and B cell donors congenic mice differing genetically only at the loci coding for immunoglobulin allotype, Jacobson et al. (215) showed that all of the yG anti-SRBC PFC produced in the adoptive transfer response to a second antigenic challenge were of the B cell allotype, thus indicating memory in the B cell population. Kennedy et al. (236) found that a synergistic effect between adoptively transaddition of normal thymus or spleen cells failed to restore the response. to burro erythrocytes was optimal when both cell populations came from donors primed to burro red blood cells. This also supports the existence of specific memory in the B cell population. Recently, Jehn and Karlin (319) have shown that both thymocytes and bone marrow cells could adoptively transfer memory in the response to SRBC. Whereas the demonstration of B cell memory is not without complicating problems in studies of the anti-SRBC response, the use of different cooperating systems has led to a clearer definition of immunological memory in both T and B cells. In the models of cooperative cell interactions in the immune response to hapten-carrier conjugates, the existence of T and B cell memory is well established. This is so because the carrier-specific helper cell has been demonstrated to be a T cell (126, 1 4 3 ) , and it is quite clear that carrier priming leads to enhanced hapten-specific antibody responses in viuo (63, 64, 6 6 ) , and in uitro ( 137-140). Moreover, the development of heightened antihapten antibody production on the part of the B cell population, as a result of prior immunization in such systems, indicates the existence of immunological memory in the B cell. This conclusion is further strengthened by the following observations. 1. In the hapten-carrier cooperating systems involving free carrier

REGULATORY INFLUENCE OF ACTIVATED T CELLS

77

immunization, the class of antihapten antibody produced is determined by the conditions employed for sensitizing the animal to the hapten (not to the secondaiy carrier). This has been observed in the enhanced secondary anti-SULF response in rabbits (65,134) and in both enhanced primary and secondary anti-DNP responses in guinea pigs (66). These findings demonstrate that the class of primary or secondary antihapten antibodies is predetermined by the hapten-primed memory cells of the B type and not by the carrier-specific helpers of the T population. 2. In the adoptive transfer system, Mitchison et al. (65,70)have used mice differing in immunoglobulin allotype as donors of carrier-primed ( BSA) or hapten-primed (NIP-OVA) spleen cells. Following transfer of the donor cell mixture into irradiated recipients the antihapten antibody produced upon challenge with NIP-BSA bore the allotype of the haptenprimed donor spleen cells. This observation clearly illustrates specific immunological memory among the hapten-specific B cell population. 3. In the hapten-carrier cooperating system in which free carrier preimmunization or supplemental immunization leads to enhanced primary or secondary antihapten responses, respectively, Katz et al. (66) have shown that the affinity of antihapten antibodies is characteristic of the mode and time of immunization with the hapten-carrier conjugate and not the mode and time of immunization with free carrier. This finding can only be explained by the existence of specific immunological memory in the hapten-primed B cell population. Recently, Miller and Sprent ( 78 ) have provided conclusive evidence for the existence of immunological memory in both B and T cells in a cooperating system not involving hapten-carrier antibody responses, By making use of their observation that thoracic duct lymphocytes (TDL) consist of both T and B cells ( 2 1 7 ) , they reconstituted neonatally thymectomized CBA mice with thymus cells from (CBA X C57B1)F1 donors and analyzed the capacity of TDL obtained from such reconstituted mice to transfer humoral responses adoptively to FyG in irradiated recipients. Such TDL obtained from reconstituted mice primed with FyG produced excellent adoptive memory responses to FyG. The PFC in the irradiated recipient spleen were shown by anti-H-2 serum to be derived from the thymectoinized CBA host, not from the F, cells. By treating the TDL with anti-C57B1 serum and complement, they were able to remove selectively the F, T cell component of the TDL. This permitted precise analysis of memory in each cell population and enabled them to show specific memory in both T and B cells. Thus, T-depleted TDL which failed to transfer the response adoptively to FyG were optimally restored by addition of TDL from F y G-primed mice demonstrating memory in the T cell population. The response of irradiated recipients

78

DAVID H. KATZ AND BARUJ BENACERRAF

of primed T cells to FyG could only be enhanced by the addition of FyG-primed B cells and not with unprimed B cells, thus, demonstrating, the existence of memory in B cells. The question of whether memory is an expression of a qualitative or quantitative change in specific T and B cells is less clearly defined. Moreover, there is no reason to discount the possibility that memory in the two cell populations reflect distinctly different cellular events. It is clearly acceptable to consider that immunization bears a definite selective pressure on specific precursors of antibody-forming cells ( B cells) leading to production of antibodies of progressively higher affinity depending on time and antigen dose (298). This selective pressure is evident in the studies described above dealing with class and affinity of antihapten antibodies (65,66,134).These must be considered as qualitative changes in the B cell population. We do not know whether such B cell memory reflects an important quantitative change as well. The observation of Miller and Sprent (78) that unprimed B cells, even in very large numbers, could not substitute for primed B cells in enhancing the adoptive transfer response of primed T cells to FyG suggests that B cell memory reflects more complex events than mere quantitative considerations would explain. The T cell memory is more readily explicable in quantitative terms. The enhancement of primary or secondary antihapten antibody responses by free carrier immunization or by the adoptive transfer of carrierprimed cells can easily be thought of in terms of merely increasing the number of carrier-specific helper cells as a result of mitotic events in the T cell population occasioned by antigenic stimulation. Also relevant in this respect is the observation of Miller and Sprent (78) that very large numbers of unprimed T cells could substitute for FyG-primed T cells in enhancing the adoptive transfer response of primed B cells. Nevertheless, it is important to bear in mind that we do not have, at present, the methodology, such as we have for B cells, to delineate qualitative changes in the T cell population which may reflect selective pressures by antigen. Studies of the adoptive transfer of antihapten antibody responses in mice which have demonstrated that helper T cell activity is greatest at earlier rather than later times after immunization (65, 70, 124) may argue against the existence of antigen-related selective pressure comparable to that which exists for B cell precursors (298). The argument is strengthened by the fact that the specificity characteristics of T cells mediating delayed hypersensitivity reactions do not change significantly during the course of immunization (8, 103). The lack of selective pressure by antigen on T cells, if true, does not imply a lack of heterogeneity among such cells with respect to their receptor specificities. Indeed, that

REGULATORY INFLUENCE OF ACTIVATED T CELLS

79

such heterogeneity in affinity of antigen-binding receptors among T lymphocytes exists is well illustrated by the studies of Paul et al. ( 1 0 3 ) . They showed that lymphocytes obtained from DNP-GPA-primed guinea pigs manifest a DNA synthetic response in uitm to the immunizing antigen over a wide dose range response, in contrast to a narrow dose range response to a nonspecific mitogen, clearly reflecting heterogeneity among the specific T cell population. It is clear, therefore, that both B and T cells are capable of expressing specific immunological memory. What memory precisely reflects is uncertain. It is likely that development of memory in B cells requires distinct qualitative changes in this population either with or without quantitative changes. Development of memory in T cells involves an increase in the number of antigen-specific reactive cells, but the possibility of a concomitant qualitative change of significance in these cells is not definitively ruled out. C. IMMUNOLOGICAL ADJWANTS Studies on the immunoenhancing effects of synthetic polynucleotides has become an increasingly active area of investigation in recent years. It is not within the scope of this review to consider this subject in any detail, and the reader is referred to Braun et al. (320) and Johnson et al. (321) for recent reviews. It is relevant to note that double-stranded polynucleotides such as poly A poly U, when administered with antigen, can exert a marked stimulatory effect on antibody formation and cellmediated immune responses in normal animals. Moreover, administration of such agents to genetic low responder, thymectomized or aged animals restores their deficient antibody responses to normal levels (320, 321 ). These agents appear, therefore, to act in such a way as to stimulate or replace the normal regulatory influence of T cells on antibody production. The literature is somewhat confusing at present as to which alternative is correct. Thus, Cone and Johnson (322) have recently presented evidence which they interpret as demonstrating that poly A-poly U stimulates T cell activity, whereas Campbell and Kind (323) interpret results obtained in a somewhat different system as indicating that poly Aapoly U acts directly on B cells (thereby replacing T cell function). Perhaps the clearest studies in this regard are those of Diamantstein et al. ( 3 2 4 ) who found that adult, thymectomized, irradiated, bone marrow-reconstituted mice could be restored to develop both yM and yG anti-SRBC responses by the administration of polyanionic substances such as dextran sulfate and polyacrylic acid. These observations, if -true, indicate that the polyanions are capable of replacing T cell function in antibody responses.

80

DAVID H. KATZ AND BARU J BENACERRAF

What is, perhaps, most intriguing to consider in light of the biological relevance of these findings is the possibility that their effect is mediated through a cyclic 3’,5’-adenosinemonophosphate ( cyclic AMP) mechanism. It has been shown that cyclic AMP can enhance both in vivo and in vitro anti-SRBC responses in mice (325). More recently, Braun and his associates ( 326-329) have obtained results indicating that enhancement of immune responses by synthetic polynucleotides is related to the capacity of such substances to enhance formation of cyclic AMP in immunocompetent lymphocytes. It is attractive to consider that some of the mediators, which we propose are elaborated by T cells, may have similar physicochemical properties to the agents discussed above and that they exert their regulatory influence on B cell function through a cyclic AMP mechanism. Such agents, indeed, fill certain criteria that we have suggested earlier, namely they are rapidly acting and short-lived. It is likely that this will be definitively elucidated in the next few years. Before concluding this section, we should briefly consider the action of other immunological adjuvants in this same context. For many years it has been an unanswered puzzle to immunologists how classic adjuvants exert their enhancing properties on the immune system. It appears that adjuvants exert their effect to some extent on macrophage function. This follows from the observation of Unanue et al. (330) that nonparticulate or particulate adjuvants taken up by macrophages in vitro and injected into syngeneic mice increased the antibody response of the recipients to Maia squiwdo hemocyanin; adjuvants taken up by lymphoid cells, in contrast, had no similar enhancing effect. The first direct evidence that T cell function is required for potentiation of antibody responses by some adjuvants was obtained by Unanue (218). He found that beryllium, a very potent adjuvant in normal animals, had no enhancing effect on the antibody responses of thymectomized mice immunized with KLH. Recently, Allison and Davies (331) also examined the effect of thymus deprivation on the capacity of several adjuvants to enhance antibody responses of mice to BSA. Their findings confilmed those of Unanue ( 2 1 8 ) in demonstrating that the potentiation of antibody formation by adjuvants, such as Escherichia coli lipopolysaccharide ( endotoxin ) and Bordetellu pertussis vaccine require the presence of T cells. Thus, thymectomized, irradiated, bone marrow-reconstituted mice failed to respond to antigen plus adjuvant unless additional reconstitution with a thymus graft is made. It appears, therefore, that regardless of the effects of adjuvants on macrophage function, stimulation of T cells is an important aspect of the mechanism of action of at least some adjuvants. This does not exclude the possibility that other adjuvants may have a

REGULATORY INFLUENCE OF ACTIVATED T CELLS

81

direct effect on B cell function, i.e., if one considers the polyanionic substances in the category of adjuvants ( 3 2 4 ) .

D. CELL-MEDIATED IMMUNITY Recognition of T and B lymphocyte interactions in humoral immune

responses prompted studies to determine whether analogous interactions existed in the ontogeny of cell-mediated immunity. Globerson and Auerbach (332) found that in vitro GVH reactivity occurred only if both thymocytes and bone marrow cells were present. Barchilon and Gershon (333) and Hilgard (334) reported observations of synergy between thymus and bone marrow cells in the production of GVH splenomegaly in X-irradiated F, recipients of parental donor cells. Inasmuch as splenomegaly in GVH reactions is predominantly a host cell proliferative response (335) and the bone marrow is the source of this host cell proliferation ( 3 3 4 ) ,these findings are hardly analogous to T and B interaction in induction of humoral responses, Very recently, Eidinger and Ackerman (336) have reported evidence for a synergism between T and B cells in the development of delayed hypersensitivity responses in mice. Considerable interest presently exists in the question of whether or not cooperative interactions between T cells are essential for the development of cell-mediated immune reactions. Several lines of investigation have provided indirect evidence for T-T cell interaction in cellmediated immunity. Thus, Cantor and Asofsky (337) reported synergy between mouse lymphoid cell populations from different sources in the production of GVH splenomegaly in F, hosts. Recent studies suggest that, at least in their system, two classes of T cells may be involved: one class consisting of precursors af the effector cells and the other serving to amplify the activity of the former ( 3 3 8 ) .Studies in our own laboratory (339,340) have shown that potent nonspecific stimulation of the immune system by induction of transient GVH reactions affords significant protection to inbred guinea pigs against lethal inocula of lymphatic leukemia cells. Since tumor immunity is generally a cell-mediated phenomenon, we have interpreted these findings as indicating that the allogeneic effect can enhance cellular as well as humoral immunological mechanisms (339, 340). Indeed, preliminary studies suggest the heightened development of delayed hypersensitivity in guinea pigs as a result of the allogeneic effect ( 3 4 1 ) .This is consistent with observations of McBride et a2. (342) made in chickens several years ago. The latter studies (342) demonstrated that induction of a GVH reaction in chick embryos resulted in accelerated maturation of immunocompetence in such chick embryos as manifested by their precocious capacity to develop specific cellular immune responsiveness.

82

DAVID H. KATZ AND BARUJ BENACERRAF

The above observations indicate that the activity of some T cells may enhance the development of cell-mediated immune reactions by effector T cells. Although evidence is accumulating for the existence of functional heterogeneity among T cell population (343, 344), it is premature to draw any conclusions, at this time, as to whether such cells are truly distinct in the functional sense or rather that they reflect different stages of maturation within the same cell line or class. IX. Biological and Pathophysiological Significance of the Regulatory Influence of T Cells on Antibody Production

As shown in the preceding sections, the division of the immune system into two classes of lymphocytes, T and B cells, differentiated to mediate, respectively, the two fundamental expressions of immunity, cellular or humoral immune responses, although basically correct, was recognized to account only in part for the complex events of antibody responses, It has become abundantly clear that the specific response of B cells to antigen is affected to a considerable extent by the concomitant activity of differentiated T cells. The cooperative interaction between specific T and B cells and antigen which has been discussed at length in this review may be most appropriately interpreted as a regulatory function of activated T cells on the conditions and manner in which B cells respond to the antigen bound to their immunoglobulin receptors. This interpretation takes into account that specific B cells may be stimulated to differentiate and synthesize IgM antibodies, under certain limits, by appropriate polymeric antigens in a narrow concentration range. It stresses also, however, that the contribution of stimulated T cells is required for the following essential activities of B cells: ( 1) the antibody response by B cells at a high, otherwise tolerogenic, antigen (or epitope) concentration (which results in a marked increase, over several logs, in the antigen concentration range immunogenic for B cells); ( 2 ) the enhancement of B cell antibody responses of all classes, but particularly in the IgG classes and the switch from IgM to IgG; ( 3 ) the emergence of B cell memory populations in all classes; (4)the effective selective pressure by antigen on B cell proliferation and differentiation resulting in increase of antibody affinity with time of immunization; ( 5 ) the suppressive effect of antibody synthesis under certain conditions heretofore observed in antigen competition phenomena. The considerable importance of the regulatory function of T cell activity for the antibody response of B cells to antigen is dramatized by the explosion of excellent work in immunology which the recognition of the various aspects of this phenomenon has stimulated in a short space of time as reviewed here. Although the mechanism or mechanisms of the

REGULATORY INFLUENCE OF AmIVATED T CELLS

83

various effects of T cell activity on B cell function listed above are far from completely understood, it is our opinion that most of these phenomena have been conclusively shown to result from the active participation of antigen-stimulated T cells. Furthermore, there is excellent evidence that this activity is mediated by distinct factors capable of affecting B cell function, secreted by T cells for this purpose. Rather than relating again the data in support of these interpretations, we propose to discuss briefly in this last section the biological significance of these regulatory mechanisms in the evolution of antibody responses and in the pathogenesis of selected pathological conditions. Since the regulatory activity of T cells is both specific and antigen-dependent, the dual specificity mechanism evolved in T and B cells insures several levels of control of effective specific humoral responses. The T cell level of control is all the more relevant, since the range of specificity which can activate helper T cells is in some way considerably more restricted than the range of antibody specificities which can be synthesized by B cells. There is ample evidence that this is precisely the case, as (1) based on studies of specific immune responses controlled by histocompatibility-linked Ir genes in both guinea pigs and mice (162, 351) and ( 2 ) based also on the evidence, discussed at length in Section V,A, that haptenic determinants, for which there exist many specific B cells, are not able to activate helper T cells unless they are part of a molecule recognized by T cells as immunogenic on a genetic basis. The evolutionary advantage of a T cell level of specificity control for the stimulation of antibody synthesis can be viewed as providing a safety mechanism against the possibility that antibody synthesis will be triggered indiscriminately by molecules with little biological relevance but which are, nevertheless, capable of binding to B cell receptors. This is a possibility which must be considered very seriously as the immunoglobulin receptors of B cells encompass indeed an extremely wide range of specificity, not matched by the range of antigens capable of stimulating T cells. In addition, as specific tolerance is achieved more easily with lower antigen concentrations and for longer periods of time in the T cell than in the B cell population, as shown in Section VIII,A, the specificity control at the T cell level must be of the greatest importance in the prevention of autoimmune responses. Self-tolerance may, therefore, reflect a rigid state of tolerance to self-antigens in the T cell population. The B cells expressing receptor specificity for self-antigens may be continuously generated from the stem cell pool, but in the absence of T cell activity, such B cells are subjected to a constant tolerizing signal following direct interaction with the self-antigenic determinants readily available to them.

84

DAVID H. KATZ AND BARUJ BENACERRAF

This mechanism insures self-tolerance at the level of both potentially responsive cells-a situation analogous to one in which two different keys must be used to open the same lock. How then is self-tolerance lost in certain pathological states? An analogy is perhaps found in the experiments where the allogeneic effect permitted B cells to develop antibody responses to a normally tolerogenic substance (153) or terminated a preexisting state of tolerance to SRBC ( Section VII1,A) ( 3 1 3 ) .The implication of these findings is that in the presence of sufficiently strong T cell activity (regardless of the specificity of the T cells involved), B cells can be triggered to respond to determinants that otherwise would have turned them off. It is conceivable that similarly potent T cell stimulation can follow certain infectious or otherwise toxic insults of exogenous origin which may permit a B cell response to self-antigens, leading to transient or prolonged autoantibody formation which may or may not have pathogenetic consequences. Aside from the possible relevance to the maintenance of self-tolerance, a two-cell mechanism of antibody production may offer definite advantages in terms of host defense mechanisms, If, as we have proposed, T cells exert a regulatory influence on the rate of B cell differentiation in the presence of antigen-induced selective pressures, the advantages to the host which has been invaded, for example, with bacteria or viruses are obvious. The kinetics of optimal antibody production are increased, the response longer sustained, and the development of specific memory is most likely greatly enhanced by the presence of T cell activity. Simultaneously, of course, T cell memory has developed among those cells destined to perform their function in cell-mediated immune reactions. In this regard, there is already evidence to suggest that the existence of the two-cell mechanism can be appropriately manipulated to permit enhanced antibody production to certain bacterial antigens characteristically weak in immunogenicity. Thus, we have recently elicited considerably enhanced antibody responses in rabbits to SIIIby administering it covalently bound to BGG ( 3 5 2 ) . If augmented antibody responses can be elicited in this manner and can be shown to be saciently protective against subsequent infection by virulent organisms, then such an approach to relatively safe human immunization against a variety of bacterial polysaccharides would be feasible. Finally, the possibility that some T cell function may regulate the activity of differentiated effector T cells in cellular immune reactions may have considerable relevance and therapeutic potential in problems of cancer immunity. This is particularly true inasmuch as the major obstacles to be dealt with in this area concern the following two points: ( 1) most tumor-specific transplantation antigens (TSTA) appear to be

REGULATORY INFLUENCE OF ACTIVATED T CELLS

85

relatively weak, thereby eliciting weak or ineffective cellular immunity; and ( 2 ) antibodies produced against TSTA are in many, if not all, instances protective for the tumor cell against potential cytotoxic T cells (353). It is highly probable that both such obstacles can be overcome under appropriate conditions whereby selective heightened T cell activation may occur. In this instance, the provocateur of T cell activation need not reflect specificity related to TSTA provided the nonspecific stimulation is of sufficient magnitude. This is, perhaps, exemplified by the capacity of the allogeneic effect to confer significant protection in guinea pigs with highly fatal lymphatic leukemia transplants (339, 340) (Section VII1,D).

ACKNOWLEDGMENTS We are deeply indebted to Dr. William E. Paul for his active contribution to all of our studies on hapten-carrier cooperation phenomena. We thank Drs. J. F. A. P. Miller, Martin C. Raff, and William 0. Weigle and their colleagues who generously permitted us to reproduce some of their published data, and those investigators who provided us with unpublished observations. The excellent assistance of Miss Karen Ellis in the preparation of the manuscript is greatly appreciated.

REFERENCES 1. Burnet, F. M. (1957). Aust. J. Sci. 20, 67. 2. Burnet, F. M. (1959). ‘The Clonal Selection Theory of Acquired Immunity.” Cambridge Univ. Press, London and New York. 3. Burnet, F. M. (1967). Cold Spring Harbor Symp. Quunt. Biol. 32, 1. 4. Makela, 0. (1970). Transplant. Reu. 5, 3. 5. Greaves, M. F. (1970). Transplant. Reu. 5, 45. 6. Wipell, H. (1970). Transplant. Reu. 5, 76. 7. Ada, G. L. (1970). Transplant. Rev. 5, 105. 8. Paul, W. E. (1970). Transplant. Reu. 5, 130. 9. Miller, J. F. A. P., and Mitchell, G. F. (1969). Transplant. Rev. 1, 3. 10. Davies, A. J. S. (1969). Transplant. Rev. 1, 43. 11. Claman, H. N., and Chaperon, E. A. (1969). Transplant. Rev. 1, 92. 12. Landsteiner, K. (1962). ‘The Specificity of Serological Reactions,” rev. ed. Dover, New York. 13. Barnes, D. W. H., and Loutit, J. F. (1967). Lancet 2, 1138. 14. Moore, M. A. S., and Owen, J. J. T. (1967). Lancet 2!, 658. 15. Wolf, N. S., and Trentin, J. J. (1968). J . Exp. Med. 127, 205. 16. Warner, N. L., Szenberg, A., and Burnet, F. M. (1962). Aust. J . E x p . Biol. Med. Sci. 40, 373. 17. Warner, N. L., and Szenberg, A. ( 1962). Nature (London) 196, 784. 18. Warner, N. L., and Szenberg, A. (1964). Annu. Rev. Microbiol. 18, 253. 19. Warner, N. L. (1964). Aust. J . Exp. Biol. Med. Sci. 42, 401. 20. Warner, N. L., and Szenberg, A. (1964). I n “The Thymus in Immunobiology” (R. A. Good and A. E. Gabrielson, eds.), p. 395. Harper (Hoeber), New York.

86

DAVID H. KATZ AND BARUJ BENACERRAF

21. Cooper, M. D., Peterson, R. D. A,, South, M. A., and Good, R. A. (1966). J . Exp. Med. 123, 75. 22. Warner, N. L. (1965). Azrst. J. Exp. Biol. Med. Sci. 43, 439. 23. Aspinall, R. L., Meyer, R. K., Graetzer, M. A., and Wolfe, H. R. (1963). J. Immunol. 90, 872. 24. Miller, J. F. A. P., and Osoba, 1). (1967). Physiol. Reu. 47, 437. 25. Warner, N. L., Ovary, Z., and Kantor, F. S. (1971). Int. Arch. Allergy Appl. Immunol. 40, 719. 26. Miller, J. F. A. P. (1961). Lancet 2, 748. 27. Jankovi6, B. D., Waksman, B. H., and Arnason, B. G. (1962). J. Exp. Med. 116, 159. 28. Amason, B. G., Jankovi6, B. D., Waksman, B. H., and Wennersten, C. ( 1962). J. Exp. Med. 116, 177. 29. Good, R. A., Dalmasso, A. P., Martinez, C., Archer, 0. K., Pierce, J. C., and Papermaster, B. W. (1962). J. Exp. Med. 116, 773. 30. DiGeorge, A. M. ( 1965). J. Pediat. 67, 907. 31. Lischner, H. W., Punnett, H. H., and DiGeorge, A. M. (1967). Nature (London) 214, 580. 32. DiGeorge, A. M. (1968). In “Birth Defects, Immunologic Deficiency Diseases in Man,” p. 116. National Foundation-March of Dimes, New York. 33. August, C. S., Rosen, F. S., Filler, R. M., Janeway, C. A., Markowski, B., and Ray, H. E. M. (1968). Lancet 2, 1210. 34. Lischner, H. W., and DiGeorge, A. M. (1969). Lancet 2, 1044. 35. Miller, J. F. A. P. (1962). Proc. Roy. SOC., Ser. B 156, 415. 36. Papermaster, B. W., Dalmasso, A. P., Martinez, C., and Good, R. A. (1962). Proc. SOC. Exp. Biol. Med. 111, 41. 37. Martinez, C., Kersey, J., Papermaster, B. W., and Good, R. A. (1962). Proc. SOC. Exp. Biol. Med. 109, 193. 38. Dalmasso, A. P., Martinez, C., Sjodin, K., and Good, R. A. (1963). J. E x p . Med. 118, 1089. 39. Taylor, R. B. (1963). Nature (London) 199, 873. 40. Leuchars, E., Cross, A. M., and Dukor, P. (1965). Transplantation 3, 28. 41. Miller, J. F. A. P., Leuchars, E., Cross, A. M., and Dukor, P. (1964). Ann. N . Y. Acod. Sci. 120, 205. 42. Osoba, D., and Miller, J. F. A. P. (1963). Nature (London) 199, 653. 43. Miller, J. F. A. P. (1963). Lancet 1, 43. 44. Leuchars, E., Cross, A. M., Davies, A. J. S., and Wallis, V. (1964). Nature (London) 203, 1189. 45. Davies, A. J. S., Leuchars, E., Wallis, V., and Koller, P. C. (1966). Transplantation 4, 438. 46. Koller, P. C., Davies, A. J. S., Leuchars, E., and Wallis, V. (1966). In “Germinal Centers in Immune Responses” (H. Cottier et al., eds.), p. 161. Springer-Verlag, Berlin and New York. 47. Miller, J. F. A. P., DeBurgh, P. M.,Dukor, P., Grant, G., Allman, V., and House, W. (1966). Clin. Exp. Immunol. 1, 61. 48. Kennedy, J. C., Siminovitch, L., Till, J. E., and McCulloch, E. A. (1965). Proc. SOC. E x p . Biol. Med. 120, 868. 49. Celada, F. (1967). J. E x p . Med. 125, 199. 50. Gregory, C. J., and Lajtha, L. G. (1968). Nature (London) 218, 1079. 51. Bussard, A. E., and Lurie, M. (1967). 1. Exp. Med. 125, 873.

REGULATORY INFLUENCE OF ACTIVATED T CELLS

87

52. Claman, H. N., Chaperon, E. A., and Triplett, R. F. (1966). PTOC. SOC. Exp. Biol. Med. 122, 1167. 53. Playfair, J. H., Papermaster, B. W., and Cole, L. J. (1965). Science 149, 998. 54. Davies, A. J. S., Leuchars, E., Wallis, V., Marchant, R., and Elliott, E. V. ( 1967). Transplantation 5, 222. 55. Miller, J. F. A. P., and Mitchell, G. F. (1967). Nature (London) 216, 659. 56. Mitchell, G . F., and Miller, J. F. A. P. (1968). Proc. Nut. Acud. Sci. U. S. 59, 296. 57. Miller, J. F. A. P., and Mitchell, G. F. (1968). J . E x p . Med. 128, 801. 58. Mitchell, G. F., and Miller, J. F. A. P. (1968). J. Exp. Med. 128, 821. 59. Nossal, G. J. V., Cunningham, A., Mitchell, G. F., and Miller, J. F. A. P. (1968). J. Exp. Med. 128, 839. 60. Gowans, J. L., and McGregor, D. D. (1965). Progr. Alkrgy 9, 1. 61. Taylor, R. B. ( 1969). Transplant. Reu. 1, 114. 62. Chiller, J. M., Habicht, G. S., and Weigle, W. 0. (1970). Proc. Nut. Acud. Sci. U. S. 65, 551. 63. Mitchison, N. A. (1969). In “Immunological Tolerance” (M. Landy and W. Braun, eds.), p. 149. Academic Press, New York. 64. Rajewsky, K., Schirrmacher, V., Nase, S., and Jerne, N. K. (1969). J. Exp. Med. 129, 1131. 65. Mitchison, N. A., Rajewsky, K., and Taylor, R. B. (1971). In “Developmental Aspects of Antibody Formation and Structure” (J. Sterzl and I. m a , eds.), Vol. 2, p. 547. Academic Press, New York. 66. Katz, D. H., Paul, W. E., Goidl, E. A., and Benacerraf, B. (1970). J . Exp. Med. 132, 261. 67. Paul, W. E., Katz, D. H., Goidl, E. A., and Benacerraf, B. (1970). J. Exp. Med. 132, 2.83. 68. Katz, D. H., Paul, W. E., Goidl, E. A., and Benacerraf, B. (1970). Science 170, 462. 69. Hamaoka, T., Takatsu, K., and Kitagawa, M. (1971). Immunology 21, 259. 70. Mitchison, N. A. (1971). Eur. 1. Immunol. 1, 18. 71. Boak, J. L., Mitchison, N. A., and Pattisson, P. H. (1971). Eur. J. Immunol. 1, 63. 72. Britton, S., Mitchison, N. A,, and Rajewsky, K. ( 1971). Eur. J. Immunol. 1, 65. 73. Mitchison, N. A. (1971). Eur. J. lmmunol. 1, 68. 74. Miller, J. F. A. P., Basten, A., Sprent, J., and Cheers, C. (1971). Cell. Immunol. 2, 469. 75. Claman, H. N., Chaperon, E. A., and Selner, J. C. (1968). Proc. SOC. Exp. Biol. Med. 127, 462. 76. Claman, H. N., Chaperon, E. A., and Hayes, L. L. (1969). Transplantation 7,87. 77. Claman, H. N., Chaperon, E. A., and Triplett, R. F. (1966). J . Zmmunol. 97, 828. 78. Miller, J. F. A. P., and Sprent, J. (1971). J. Exp. Med. 134, 66. 79. Shearer, G. M., Cudkowicz, C., Connell, M. St. J., and Priore, R. L. (1968). J. Exp. Med. 128, 437. 80. Shearer, G. M., Cudkowicz, G., and Priore, R. L. (1969). J. Exp. Med. 129, 185. 81. Shearer, G. M., and Cudkowicz, G. (1969). J. Exp. Med. 129, 935. 82. Shearer, G . M., Cudkowicz, G., and Priore, R. L. (1969). J. Exp. Med. 130, 467.

88

DAVID H. KATZ AND BARU J BENACERRAF

Mishell, R. I., and Dutton, R. W. (1967). J. Exp. Med. 126, 423. Dutton, R. W. (1967). Aduan. Immunol. 6, 254. Marbrook, J. (1967). Lancet 2, 1279. Mosier, D. E., and Coppleson, L. W. (1968). Proc. Nut. Acad. Sci. U. S . 61, 542. 87. Mosier, D. E., Fitch, F. W., Rowley, D. A., and Davies, A. J. S. (1970). Nature (London) 225, 276. 88. Munro, A., and Hunter, P. ( 1970). Nature (London) 225, 277. 89. Roseman, J. ( 1969). Science 165, 1125. 90. Hirst, J. A,, and Dutton, R. W. (1970). Cell. Immunol. 1, 190. 91. Hartmann, K.-U. (1970). J. E x p . Med. 132, 1267. 92. Chan, E. L., Mishell, R. I., and Mitchell, G. F. (1970). Science 170, 1215. 93. Schimpl, A,, and Wecker, E. (1970). Nature (London) 226, 1258. 94. Hartmann, K.-U. (1971). 1. Exp. Med. 133, 1325. 95. Doria, G., Martinozzi, M., Agarossi, G., and DiPietro, S. (1970). Experientia 26, 410. 96. Schimpl, A., and Wecker, E. (1971). Eur. J. lmmunol. 1, 304. 97. Salvin, S. B., and Smith, R. F. (1960). J. E x p . Med. 111, 465. 98. Benacerraf, B., and Cell, P. G . H. (1959). Immunology 2, 219. 99. Gell, P. G. H., and Benacerraf, B. (1961). J. E x p . Med. 113, 571. 100. Benacerraf, B., and Levine, B. B. (1962). J. Exp. Med. 115, 1023. 101. Silverstein, A. M., and Gell, P. G. H. ( 1962). J. E x p . Med. 115, 1053. 102. Dutton, R. W., and Bulnian, H. N. (1964). lmrnunology 7, 54. 103. Paul, W. E., Siskind, G. W., and Benacerraf, B. (1968). J. Exp. Med. 127, 25. 104. Ovary, Z., and Benacerraf, B. (1963). Proc. SOC. E x p . Biol. Med. 14, 72. 105. Paul, W. E., Siskind, G. W., Benacerraf, B., and Ovary, Z. (1967). J. Immunol. 99, 760. 106. Mitchison, N. A. (1967). Cold Spring Harbor Symp. Quant. Biol. 32, 431. 107. Eisen, H. N., Carsten, M. E., and Belman, S. (1954). J. Immunol. 73, 296. 108. Levine, B. B. (1963). J. E x p . Med. 117, 161. 109. Levine, B. B. (1965). J. E x p . Med. 121, 873. 110. Borek, F., and Silverstein, A. M. ( 1965). Nature (London) 205, 299. 111. Paul, W. E., Siskind, G . W., and Benacerraf, B. (1966). J. Exp. Med. 123, 689. 112. Parker, C. W., Gott, S. M., and Johnson, M. C. (1966). Biochemistry 5, 2314. 113. Little, J. R., and Counts, R. B. (1969). Biochemistry 8, 2729. 114. Benacerraf, B., Ojeda, A., and Maurer, P. H. (1963). 1. Erp. Med. 118, 945. 115. Levine, B. B., Ojeda, A., and Benacerraf, B. ( 1963). Nature (London) 200, 544. 116. Green, I., Paul, W. E., and Benacerraf, B. (1966). J. Exp. Med. 123, 859. 117. Cinader, B., and Pearce, J. H. ( 1958). Brit. J. E x p . Pathol. 39, 8. 118. Rajewsky, K., Rottlander, E., Peltre, G., and Miiller, B. (1967). J. Exp. Med. 126, 581. 119. Schierman, L. W., and McBride, R. A. (1967). Science 156, 658. 120. Rajewsky, K., and Rottlander, E. (1967). Cold Spring Harbor Symp. Quant. Biol. 32, 547. 121. Green, I,, Paul, W. E., and Benacerraf, B. (1968). J. E x p . Med. 127, 43. 122. Paul, W. E., Thorbecke, G. J., Siskind, G. W., and Benacerraf, B. (1969). Immunology 17, 85. 123. Benacerraf, B., Green, I., and Paul, W. E. (1967). Cold Spring Harbor Symp. Quunt. Biol. 32, 569. 124. Kontiainen, S . (1971). Eur. J. Immunol. 1, 276. 83. 84. 85. 86.

REGULATORY INFLUENCE OF ACTWATED T CELLS

89

125. Roelants, G. E., and Askonas, B. A. (1971). Eur. J. Immunol. 1, 151. 126. Raff, M. C. (1970). Nature (London) 226, 1257. 127. Miller, J. F. A. P., Sprent, J., Basten, A., Warner, N. L., Breitner, J. C. S., Rowland, G., Hamilton, J., Silver, H., and Martin, W. J. (1971). J . Exp. Med. 134, 1266. 128. Salvin, S. B., and Smith, R. F. (1960). Proc. SOC. E x p . Biol. Med. 104, 584. 129. Tada, T., and Okumura, K. (1971). J . Immunol. 107, 1137. 130. Ashley, H., and Ovary, Z. (1965). Proc. SOC. Exp. Biol. Med. 119, 311. 131. Coe, J. E., and Salvin, S. B. (1964). J. lmmunol. 93, 495. 132. Rittenberg, M. B., and Amkraut, A. (1966). J . Immunol. 97, 42. 133. Levine, B. B. (1967). J. Immunol. 99, 1173. 134. Schirrmacher, V., and Rajewsky, K. (1970). J. E x p . Med. 132, 1019. 135. Rittenberg, M. B., and Pratt, K. L. (1969). Proc. SOC. Exp. Biol. Med. 132, 575. 136. Kettman, J., and Dutton, R. W. (1970). J. lmmunol. 104, 1558. 137. Katz, D. H., Bluestein, H. G., Rouques, R. J., and Pierce, C. W. (1971). J. Immunol. 106, 271. 138. Trowbridge, I. S., Lennox, E. S., and Porter, R. R. (1970). Nature (London) 228, 1087. 139. Dutton, R. W., Campbell, P., Chan, E., Hirst, J., Hoffmann, M., Kettman, J., Lesley, J., McCarthy, M., Mishell, R. I., Raidt, D. J., and Vann, D. ( 1971). In “Cellular Interactions in the Immune Response” ( S . Cohen, G. Cudkowicz, and R. T. McCluskey, eds.), p. 31. Karger, Basel, 140. Kettman, J., and Dutton, R. W. ( 1971). Proc. Nut. Acad. Sci. U. S . 6 4 699. 141. Segal, S., Globerson, A., and Feldman, M. (1971). Cell Immunol. 2, 2.05. 142. Kunin, S., Shearer, G. M., Segal, S., Globerson, A,, and Feldman, M. (1971). Cell. lmmunol. 2, 229. 143. Cheers, C., Breitner, J. C. S., Little, M., and Miller, J. F. A. P. (1971). Nature New Biol. 232, 248. 144. Humphrey, J. H., Parrott, D. M. V., and East, J. (1964). Immunology 7, 419. 145. Davies, A. J. S., Carter, R. L., Leuchars, E., Wallis, V., and Dietrich, F. M. (1970). Immunology 19, 945. 146. Howard, J. G., Christie, G. &., Courtenay, B. M., Leuchars, E., and Davies, A. J. S. (1971). Cell. Immunol. 2, 614. 147. Armstrong, W. D., Diener, E., and Shellam, G. R. (1969). J . Exp. Med. 129, 393. 148. Moller, G., and Michael, G. (1972). To be published. 149. Anderson, B., and Blomgren, H. (1971). Cell. Immunol. 2, 411. 150. Basch, R. S. (1966). Int. Arch. Allergy Appl. Immunol. 30, 105. 151. Feldmann, M., and Basten, A. (1971). J. E x p . Med. 134, 103. 152. Feldmann, M. ( 1971). Nature New Biol. 231, 21. 153. Katz, D. H., Davie, J. M., Paul, W. E., and Benacerraf, B. ( 1971). J. E x p . Med. 134, 201. 154. Golan, D. T., and Borel, Y. (1971). J. Exp. Med. 134, 1046. 155. Havas, H. F. ( 1969). Immunology 17, 819. 156. Baker, P. J., Barth, R. F., Stashak, P. W., and Amsbaugh, D. F. (1970). J. Immunol. 104, 1313. 157. Britton, S., and Moller, G. (1968). J. Immunol. 100, 1326. 158. Taylor, R. B., and Wortis, H. H. (1968). Nature (London) 220, 927. 159. Aird, J. (1971). Immunology 20, 617.

90

DAVID H. KATZ AND BARUJ BENACERRAF

Mdcela, 0. (1970). Transplant. Rev. 5, 3. Hamaoka, T . ( 1971 ), Private communication. McDevitt, H. O., and Benacerraf, B. (1969). Aduan. Immunol. 11, 31. Grumet, F. C. (1970). Fed. Proc., Fed. Amer. SOC. E x p . Biol. 30, 469. Mitchell, G. F., Mishell, R. I., and Herzenberg, L. A. (1971). In “Progress in Immunology” (B. Amos, ed.), p. 323. Academic Press, New York. 165. Gershon, R. K., and Paul, W. E. (1971). J . Immunol. 106, 872. 166. Shearer, G. M., and Cudkowicz, G. (1969). J. Exp. Med. 130, 1243. 167. Naor, D., and Sulitzeanu, D. ( 1967). Nature (London) 214, 687. 168. Ada, C. L., and Byrt, P. (1969). Nature (London) 222, 1291. 169. Wigzell, H., and Makela, 0. (1970). 1. Exp. Med. 132, 110. 170. Humphrey, J. H., and Keller, H. U. (1971). In “Developmental Aspects of Antibody Formation and Structure” (J. Sterzl and H. Riha, eds.), Vol. 2, p. 485. Academic Press, New York. 171. Raff, M. C., Steinberg, M., and Taylor, R. B. (1970). Nature (London) 225, 553. 172. Pernis, B., Forni, L., ancl Amante, L. (1970). J . E x p . Med. 132, 1001. 173. Rabellino, E., Colon, S., Grey, H. M., and Unanue, E. R. ( 1971). J. Exp. Med. 133, 156. 174. Unanue, E. R., Grey, H. M., Rabellino, E., Campbell, P., and Schmidtke, J. (1971). J . E x p . Med. 133, 1188. 175. Davie, J. M., Paul, W. E., Mage, R. G., and Goldman, M. B. ( 1971). Proc. Nut. Acad. Sci. U. S . 68, 430. 176. Coombs, R. R. A., Gumer, B. W., Janeway, C. A., Jr., Wilson, A. B., Cell, P. G. H., and Kelus, A. S. (1970). Immunology 18, 417. 177. Byrt, P., and Ada, G. L. (1969). Immunology 17, 503. 178. Dwyer, J. M., and Mackay, I. R. (1970). Lancet 1, 164. 179. Warner, N. L., Byrt, P., and Ada, G. L. (1970). Nature (London) 226, 942. 180. Unanue, E. R. (1971). J. Immunol. 107, 1168. 181. Davie, J. M., and Paul, W. E. (1971). J. E x p . Med. 134, 495. 182. Davie, J. M., Rosenthal, A. S., and Paul, W. E. (1971). J. E x p . Med. 134, 517. 183. Wigzell, H., and Anderson, B. (1969). J. E x p . Med. 129, 23. 184. Truffa-Bachi, P., and Wofsy, L. (1970). Proc. N u t . Acad. Sci. U . S . 66, 685. 185. Unanue, E. R. (1971). J. Immunol. 107, 1663. 186. Basten, A., Miller, J. F. A. P., Warner, N. L., and Pye, J. (1971). Nature New Biol. 231, 104. 187. Greaves, M. F., Torrigiani, G., and Roitt, I. M. (1969). Nature (London) 222, 885. 187a. Stupp, Y.,Paul, W. E., and Benacerraf, B. (1971). Immunology 21, 595. 188. Mason, S., and Warner, N. L. (1970). J. Immunol. 104, 762. 189. Reithmiiller, G., Rieber, E.-P., and Seeger, I. (1971). Nature New Biol. 230, 248. 190. Greaves, M. F., Mdler, E., ancl Moller, G. (1970). Cell Immunol. 1, 386. 191. Lesley, J. F., Kettman, J. R., and Dutton, R. W. (1971). J. E x p . Med. 134, 618. 192. Mitchison, N. A. (1970). Fed. Proc., Fed. Amer. SOC. E x p . Biol. 29, 222. 193. Iverson, G. M. (1970). Nature (London) 827, 273. 194. Taylor, R. B., and Iverson, G. M. (1971). Proc. Roy. Soc., Ser. B 176, 393. 195. Janeway, C. A,, Jr., and Paul, W. E. ( 1972). Fed. Proc., Fed. A ~ TSOC. . Exp. Biol. 31, 746. 196. Raff, M. C., and Wortis, H. H. (1970). Immunology 18, 931. 197. Vann, D. C., and Kettman, J. (1972). J. Immrtnol. 108, 73. 160. 161. 162. 163. 164.

REGULATORY INFLUENCE OF ACTIVATED T CELLS

91

Osoba, D. (1970). J. E x p . Med. 132, 368. Goldie, J. €I., and Osoba, D. (1970). Proc. SOC. E x p . Biol. Med. 133, 1.265. Haskill, J. S . , Byrt, P., and Marbrook, J. (1970). J. E x p . Med. 131, 57. Moller, C., and Miiller, E. ( 1965). Nature (London) 208, 260. Asherson, C. L., and Loewi, G. (1967). Immunology 13, 509. Feldman, J. D. (1968). J. Immunol. 101, 563. Kass, E . H., and Finland, M. (1953). Annu. Rev. Microbiol. 7, 361. Berglund, K. (1956). Actu Puthol. Microbiol. Scund. 38, 311. Dougherty, T. F. (1952). Phys. Rev. 32, 379. Ishidate, M., and Metcalf, 1). (1963). Aust. J . E x p . Biol. Med. Sci. 41, 637. Cerottini, JX.,and Brunner, K. T. (1967). Immunology 13, 395. Blomgren, H . , and Andersson, B. (1969). E x p . Cell Res. 57, 185. Cohen, J. J., Fischbach, M., and Claman, H. N. (1970). J. Immunol. 105, 1146. Levine, M. A., and Claman, H. N. (1970). Science 167, 1515. Cohen, J . J., and Claman, H. N. (1971). J. E x p . Med. 133, 1026. Andersson, B., and Blomgren, H. (1970). Cell. Zmmunol. 1, 362. Cohen, J. J . , and Claman, H. N. (1971). Nature (London) 229, 274. Jacobson, E. B., L'Age-Stehr, J., and Herzenberg, L. A. (1970). J. E x p . Med. 131, 1109. 216. Potter, M . , and Lieberman, R. ( 1967). Advan. Immunol. 7 , 91. 217. Miller, J . F. A. P., and Sprent, J. ( 1971). Nature New Bid. 230, 267. 218. Unanue, E. R. (1970). J. Immunol. 105, 1339. 219. Katz, D. H . , Paul, W. E., and Benacerraf, B. (1970). Unpublished observations. 220. Miller, J. F. A. P., Sprent, J., Basten, A., Warner, N. L., Breitner, J. C. S., Rowland, G., Hamilton, J., Silver, H., and Martin, W. J. ( 1971). J. Exp. Med. 134, 1266. 221. David, J. R. ( 1965). J . E x p . Med. 122, 1125. 222. David, J . R . (1966). Proc. Nut. Acad. Sci. U. S . 56, 72. 223. Bloom, B. R., and Bennett, B. (1966). Science 153, 80. 224. Bennett, B., and Bloom, B. R. (1968). Proc. Nut. A c d . Sci. U.S. 59, 756. 225. Katz, D. H . , Goidl, E. A., Paul, W. E., and Benacerraf, B. (1971). J. E x p . Med. 133, 169. 226. Katz, D. H., Paul, W. E., and Benacerraf, B. (1971). J. Imrnunol. 107, 1319. 227. Cantor, H . , Mandel, M. A., and Asofsky, R. (1970). J. Imrnunol. 104, 409. 228. Katz, D. H . ( 1971). Unpublished observations. 229. Ekpaha-Mensah, A,, and Kennedy, J. C. (1971). Nature New Biol. 233, 174. 230. Osborne, D. P. Jr., and Katz, D. H. ( 1972). J. Exp. Med. (in press). 231. Ordal, J. C., and Grumet, F. C. ( 1972). Fed. Proc., Fed. Amer. SOC. E x p . Biol. 31, 777. 232. Sulica, A., Tarrab, R., and Sela, M. (1971). Science 171, 1165. 233. Rubin, A. S., and Coons, A. H. (1971). Proc. Nat. Acad. Sci. U . S . 68, 1665. 234. Storh, U., and Weiser, R. S. (1971). J . Imrnnnol. 105, 957. 235. Moller, G. ( 1971). Immunology 20, 597. 235u. Jimenez, L., Bloom B. R., Blume, M. R., and Oettgen, H. F. (1971). J. E x p . Med. 133, 740. 236. Kennedy, J. C., Treadwell, P. E., and Lennox, E. S. (1970). J. E x p . Med. 132, 353. 237. Doria, G.,Agarossi, G., and DiPietro, S. (1972). J. Immunol. 108, 268. 238. Gorczynski, R. M., Miller, R. G., and Phillips, R. A. (1972). J. Immunol. 108, 547. 239. Dutton, R. W., Falkoff, R., Hirst, J. A., Hoffmann, M., Kappler, J. W., Kettman, 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215.

92

DAVID H. KATZ AND BARUJ BENACERRAF

J. R., Lesley, J. F., and Vann, D. (1971). In “Progress in Immunology” (B. Amos, ed.), p. 355. Academic Press, New York. 240. Mitchison, N. A. (1957). j . Cell. Comp. Physiol. 50, Suppl. 1, 247. 241. Chessin, L. N., Biirjeson, J,, Welsh, P. D., Douglas, S. D., and Cooper, H. L. (1966). J. E x p . Med. 124, 873. 242. Tao, T. W. (1964). Science 146, 247. 243. Mikela, O., and Pasanen, V. J. (1969). Int. Arch. Allergy Appl. Immunol. 35, 468. 244. Taylor, R. B., Duffus, W. P. H., Raff, M. C., and de Petris, S. (1971). Nature New Biol. 233, 225. 245. Perkins, W. D., Karnovsky, M. J., and Unanue, E. R. (1972). J. Exp. Med. 135, 267. 246. Rich, R., and Pierce. C. W. (1972). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 31, 662. 247. Katz, D. H., and Unanue, E. R. (1972). J. lmmunol. 109, in press. 248. Baum, J., Lieberman, G., and Frenkel, E. P. (1969). J. Immunol. 102, 187. 249. Baker, P. J., and Stashak, P. W. (1969). J . Immunol. 103, 1342. 250. James, K. (1969). Progr. Surg. 7, 140. 251. Martin, W. J., and Miller, J. F. A. P. ( 1968). J. Exp. Med. 128, 855. 252. Dixon, F. J., and McConahey, P. J. ( 1963). J. E?p. Med. 117, 833. 253. Taliaferro, W. G. ( 1954). J. Infec. Dis. 95, 117. 254. Chanmoupan, D., and Schwartz, R. S. (1966). J. E x p . Med. 124, 363. 255. Merritt, K., and Johnson, A. G. (1963). J . Immunol. 91, 266. 256. Swanson, M. A., and Schwartz, R. S. (1967). N. Engl. J. Med. 277, 163. 257. Griisbeck, R., Nordman, C., and de la Chapelle, A. (1963). Lancet 2, 385. 258. Holt, L. J., Ling, N. R., and Stanworth, D. R. (1966). lmmunochemistq 3, 359. 259. Woodruff, M. F. A., Reid, B. L., and James, K. (1967). Nature (London) 216, 758. 260. Foerster, J., Lamelin, J. P., Green, I., and Benacerraf, B. (1969). J. Exp. Med. 129, 295. 261. Greaves, M. F., Roitt, I. M., Zamir, R.. and Carnaghen, R. B. A. (1967). Lancet 2, 1317. 262. Baker, P. J., Stashak, P. W., Amsbaugh, D. F.. Prescott, B., and Barth, R. F. (1970). J. Immunol. 105, 1581. 263. Okumura, K., and Tada, T. (1971). J. Immunol. 106, 1019. 264. Ochiai, T., and Tada, T. (1972). To be published. 265. Tada, T., Taniguchi, M., and Okumura, K. ( 1971). J. Immunol. 106, 1012. 266. Taniguchi, M., and Tada, T. ( 1971). J. Immunol. 107, 579. 267. Okumura, K., and Tada, T. ( 1971). J. Immunol. 107, 1682. 268. Kerbel, R. S . , and Eidinger, D. (1971). J. Immunol. 108, 917. 269. Michaelis, L. ( 1904). Deut. Med. Wochenschr. 30, 1240. 270. Adler, F. L. (1964). Progr. Allergy 8, 41. 271. Radovich, J,, and Talmage, D. W. (1967). Science 158, 512. 272. Moller, G., and Sjoberg, 0. (1970). Cell. Immunol. 1, 110. 273. Waterston, R. H. (1970). Science 170, 1108. 274. Brody, N. I., and Siskind, G. W. (1969). J. Erp. Med. 130, 821. 275. Eidinger, D., Khan, S. A., and Millar, K. G. (1968). J. E x p . Med. 128, 1183. 276. Fauci, A. S . , and Johnson, J. S. ( 1971). J. Immunol. 107, 1052. 277. Fauci, A. S., and Johnson, J. S. ( 1971). j . Immunol. 107, 1057. 278. O’Toole, C. M., and Davies, A. J. S. (1971). Nature (London) 230, 187.

REGULATORY INFLUENCE OF ACTIVATED T CELLS

93

Katz, D. H., Paul, W. E., and Benacerraf, B. (1972). J. lmmunol. 109, in press. Kerbel, R. S., and Eidinger, D. (1971). J. Exp. Med. 133, 1043. Gerson, R. K., and Kondo, K. (1971). J. lmmunol. 106, 1524. Gershon, R. K., and Kondo, K. ( 1971). J. lmmunol. 106, 1531. Matsumoto, T., Braun, W., and Moller, G. (1972). J. lmmunol. (in press). Dresser, D. W., and Mitchison, N. A. (1968). Aduan. Immunol. 8, 129. Isakovic, K., Smith, S. B., and Waksman, B. H. (1965). I . Exp. Med. 122, 1103. Gershon, R. K., Wallis, V., Davies, A. J. S., and Leuchars, E. (1968). Nature (London) 218, 381. 287. Taylor, R. B. (1968). Nuture (London) 220, 611. 288. Miller, J. F. A. P., and Mitchell, G. F. (1970). J. Exp. Med. 131, 675. 289. Playfair, J. H. L. ( 1971). Nature (London) 222, 882. 290. Gershon, R. K., and Kondo, K. ( 1970). Immunology 18, 723. 291. Borel, Y. (1971). Nature (London) 230, 180. 292. Chiller, J. M., Habicht, G. S., and Weigle, W. 0. (1971). Science 171, 813. 293. Howard, J. G., Elson, J., Christie, G. H., and Kinsby, R. G. (1969). Clin. Exp. lmmunol. 4, 41. 294. Naor, D., and Sulitzeanu, D. (1969). lnt. Arch. Allergy Appl. lmmunol. 36, 112. 295. Sjoberg, 0. (1971). J. Exp. Med. 133, 1015. 296. Davie, J. M., Katz, D. H., Paul, W. E., and Benacerraf, B. (1972). J. Exp. Med. (in press). 297. Theis, G. A., and Siskind, G. W. (1968). J. lmmunol. 100, 138. 298. Siskind, G. W., and Benacerraf, B. (1969). Aduun. lmmunol. 10, 1. 299. Borek, F., and Battisto, J. R. (1971). J. lmmunol. 106, 677. 300. Weigle, W. 0. (1965). J. lmmunol. 94, 177. 301. Brownstone, A., Mitchison, N. A., and Pitt-Rivers, R. ( 1966). Immunology 10, 481. 302. Segal, S., Globerson, A., and Feldman, M. ( 1971). Cell. Immunol. 2, 222. 303. Rittenberg, M. B., and Bullock, W. W. ( 1972). lmmunochemistry (in press). 304. Naor, D., and Mishell, R. I. (1972). J. lmmunol. 108, 246. 305. Gershon, R. K., and Kondo, K. ( 1971). Fed. Proc., Fed. Amer. SOC. E x p . Biol. 30, 650. 306. Borel, Y. (1971). Private communication. 307. Klinman, N. R. (1971). J. Exp. Med. 133, 963. 308. Diener, E., and Armstrong, W. D. (1969). J. Exp. Med. 129, 591. 309. Bullock, W. W., and Rittenberg, M. B. ( 1970). J. Exp. Med. 132, 926. 310. Porter, R. J. (1962). Fed. Proc., Fed. Amer. SOC. E x p . Biol. 21, 28. 311. Weigle, W. 0. ( 1964). Immunology 7, 239. 312. Humphrey, J. H. (1964). lmmunology 7, 462. 313. McCullagh, P. J. (1970). J. Exp. Med. 132, 916. 314. Feldmann, M., and Diener, E. (1970). J. Exp. Med. 131, 247. 315. Diener, E., and Feldmann, M. (1970). 1. Exp. Med. 132, 31. 316. Feldmann, M., and Diener, E. (1971). lmmunology 21, 387. 317. Cunningham, A. J. (1969). lmmunology 17, 933. 318. Takahashi, T., Carswell, E. A., and Thorbecke, G. J. (1970). J. Exp. Med. 132, 1181. 319. Jehn, U. W., and Karlin, L. (1971). J . lmmunol. 106, 946. 320. Braun, W., Ishizuka, M., Yajima, Y., Webb, D., and Winchurch, R. (1971). In 279. 280. 281. 282. 283. 284. 285. 286.

94

321.

322. 323. 324. 325. 326. 327. 328. 329. 330. 331. 332. 333. 334. 335. 336. 337. 338. 339. 340. 341. 342. 343. 344. 345. 346. 347. 348. 349. 350. 351. 352. 353.

DAVID H. KATZ AND BARUJ BENACERRAF

“Biological Effects of Polynucleotides” (R. Beers and W. Braun, eds.), p. 26. Springer Publ., New York. Johnson, A. G., Cone, R. E., Friedman, H. M., Han, I. H., Johnson, H. G., Schmidtke, J. R., and Stout, R. D. (1971). In “Biological Effects of Polynucleotides” (R. Beers and W. Braun, eds.), p. 52. Springer Publ., New York. Cone, R. F., and Johnson, A. G. (1971). J. E x p. Med. 133, 665. Campbell, P. A., and Kind, P. (1971). J. Immunol. 107, 1419. Diamantstein, T., Wagner, B., L’Age-Stehr, J., Beyse, I., Odenwald, M. V., and Schultz, G. (1971). EUT. J. Immunol. 1, 302. Ishizuka, M., Gafni, M., and Braun, W. (1970). Proc. Soc. Exp. Biol. Med. 134, 963. Braun, W., and Ishizuka, M. (1971). Proc. Nut. Acad. Sci. U. S . 68, 1114. Winchurch, R., Ishizuka, M., Webb, D., and Braun, W. (1971). J. Immunol. 106, 1399. Braun, W., and Ishizuka, M. (1971). J . Immunol. 107, 1036. Ishizuka, M., Braun, W., and Matsumoto, T. (1971). 1. Immunol. 107, 1027. Unanue, E. R., Askonas, B. A., and Allison, A. C. (1969). J. Immunol. 103, 71. Allison, A. C., and Davies, A. J. S. ( 1971). Nature (London) 233, 330. Globerson, A., and Auerbach, R. (1967). J. E r p . Med. 126, 223. Barchilon, J., and Gershon, R. K. (1970). Nature (London) 227, 71. Hilgard, H. R. (1970). J. E r p . Med. 132, 317. Simonsen, M. (1962). Progr. Allergy 6, 349. Eidinger, D., and Ackerman, A. (1971). 1. E r p . Med. 133, 5 . Cantor, H., and Asofsky, R. (1970). J. E x p . Med. 131, 235. Asofsky, R., Cantor, H., and Tigelaar, R. E. (1971). In “Progress in Immunology” (B. Amos, ed.), p. 369. Academic Press, New York. Katz, D. H., Ellman, L., Paul, W. E., Green, I., and Benacerraf, B. (1972). Cancer Res. 32, 133. Ellman, L., Katz, D. H., Green, I., Paul, W. E., and Benacerraf, B. (1972). Cancer Res. 32, 141. Paul, W. E. (1971). Private communication. McBride, R. A,, Coppleson, L. W., Nisbet, N. W., Simonsen, M., SkowronCendrzak, A,, and Wigzell, H. L. R. ( 1966). Immunology 10, 63. Stobo, J. D. (1972). Transplant. Reu. 11, in press. Raff, M. C., and Cantor, H. (1971). In “Progress in Immunology” ( B . Amos, ed.), p. 83. Academic Press, New York. Raff, M. C., Nase, M., and Mitchison, N. A. (1971). Nature (London) 229, 50. Reif, A. E., and Allen, J. M. V. ( 1964). J. E x p . Med. 120, 413. Raff, M. C., and Wortis, H. H. (1970). Immunology 18, 931. Nussenzweig, V., Bianco, C., Dukor, P., and Eden, A. (1972). In “Progress in Immunology” (B. Amos, ed.), p. 73. Academic Press, New York. Basten, A., Sprent, J., Mandel, T., Miller, J. F. A. P., and Pye, J. (1972). J. E x p . Med. 135, 610. Engers, H. D., and Unanne, E. R. (1971). Private communication. Benacerraf, B., and McDevitt, H. 0. (1972). Science 175, 273. Paul, W. E., Katz, D. H., and Benacerraf, B. ( 1971). J. Immunol. 107, 685. Hellstrom, K. E., Hellstrom, I., Sjogren, H. O., and Warner, G. A. (1971). In “Progress in Immunology” (B. Amos, ed.), p. 940. Academic Press, New York.

The Regulatory Role of Macrophages in Antigenic Stimulation E.

R. UNANUE

Department of Pathology, Harvard Medical School, Boston, Massachusetts

I. Introduction . . . . . . . . . . . 11. Association of Macrophage-Bound Antigen with Immunogenicity A. I n Vivo Experiments . . . . . . . . . B. I n Vitro Experiments . . . . . . . . . 111. Handling of Antigen by Macrophages . . . . . . A. Uptake of Antigen . . . . . . . . . B. The Immunogenic Moiety . . . . . . . . 1V. Macrophage-Lymphocyte Contact . . . . . . . V. Macrophages and Adjuvants . . . . . . . . VI. Summary . . . . . . . . . . . . References . . . . . . . . . . . .

,

. . . . . . . . . .

. . . . . .

. . . . .

95 97 97 120 128 128 137 147 149 151 157

I. Introduction

An optimal immune response to most antigens usually needs interaction or collaboration among three distinct cell types as is evidenced by the results of extensive research in cellular immunology during the past five years. These three cells are the B and T types of lymphocytes (i.e., B lymphocytes are those derived from the bone marrow in mammals or the bursa of Fabricius in chickens; T lymphocytes are those derived from the thymus) and the macrophages. The B and T lymphocytes have specific recognition units on their surfaces and express humoral or cellmediated immunity, respectively. Macrophages do not elaborate antibody; they do interact with a multiplicity of antigens, catabolize, and destroy antigen and also play a helper role in the inductive process, in part, by presentation of some of the antigen. This kind of collaboration between macrophages ( or their products ) and lymphocytes was actually the first immunological cell-to-cell interaction recognized, preceding by a few years the discovery of the interaction between T and B cells. The present review considers the role played by macrophages in handling antigen and in regulating the response of B and T lymphocytes. The history of the macrophage is complex. The immunological importance given to this cell has varied throughout the years. In the 1930s when our understanding of lymphocytes and antibody formation was poor, it was thought that macrophages themselves were responsible for both uptake of antigen and synthesis of antibody (Sabin, 1923). As a 95

96

E. R. UNANUE

consequence, a great part of research in immunology employed techniques which evaluated uptake of antigen or physiology of the macrophage system. Following recognition that lymphocytes and plasma cells were the cells involved in antibody responses, the interest in macrophages dwindled for a time only to be revived by Fishman and Adler’s provocative observations in the early 1960s (Fishman, 1961; Fishman and Adler, 1963a,b). These authors demonstrated that under certain conditions antibody responses by lymphocytes did not occur upon exposure to antigen alone but rather upon interaction with extracts of macrophages which had previously phagocytized the antigen. These early experiments were interpreted as denoting that a special processing of antigen by macrophages was an essential step in immune recognition. The macrophage again became the center of attention but only for a short time. Later, as more experimental data became available, it was clear to many that this early interpretation might not be entirely correct. Subsequent experiments have shown that lymphocytes do interact with antigens bound to macrophages but that this interaction on occasion may not be absolutely necessary and, moreover, may not involve processing of antigen molecules. During recent years the interactions between T and B lymphocytes and antigen have concerned immunologists. We now better understand how lymphocytes interact with antigens and the nature and the state of immunogenic molecules. Although all the facts about macrophages and lymphocytes are not available and will not be for some time, our knowledge of the physiology of these cells is, nevertheless, growing, and the role of the macrophage in the inductive state of immunity may now be placed in a better perspective. Apart from its role in the induction of immunity, the macrophage has held the interest of those studying delayed hypersensitivities ( or cell-mediated immunities) ( Benacerraf and Green, 1989). We now recognize the macrophage as one of the cells involved in the nonspecific components of these reactions, playing an essential role in resistance to some bacterial and viral infections (Mackaness and Blanden, 1967). This review will focus on the role of the macrophage in removing antigen from extracellular fluids, degrading the larger part of this antigen while presenting a small part of it in persisting immunogenic form. This handling of antigen is done without contributing to the specificity of the immune response which is determined by the antigenreactive T and B lymphocytes. During the process of uptake of antigen, macrophages appear to retain a few molecules of antigen, undegraded or with few chemical changes. Macrophage-associated antigen becomes an effective immunogenic stimulus mainly in conditions that require that two types of lymphocytes meet with antigen molecules. These lympho-

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

97

cytes are specifically antigen committed and few in number. (Perhaps with some antigens, the molecules need to be partially degraded in order to “allow” them to trigger lymphocytes.) The positive immunogenic role of macrophages is related to their capacity to ( I ) remove extracellular antigen which might be capable of interacting with and eliminating isolated T or B lymphocytes and ( 2 ) retain antigen in lymphoid tissues and promote its necessary meeting with both T and B cells. The importance of the regulatory role of macrophages depends upon the need for concentration of antigen molecules, a need which is conditioned to number of T and B cells, requirements for T cells, physiochemistry of the antigen, etc. Other interpretations of the role of macrophages have been published (Fishman and Adler, 1970; R. s. Schwartz et d,1970; Perkins, 1970; Cohn, 1968). II. Association of Macrophage-Bound Antigen with lmmunogenicity

The role of macrophages in the induction of an immune response has been evaluated in uivo and in uitro in a wide variety of experiments. Among the in uiuo procedures are included those experiments that involve either modifying the antigen so as to make it more-or-less phagocytizable by the macrophage system or, alternatively, modifying or altering the macrophage or reticuloendothelial system (RES) in order to reduce the amount of antigen taken up. Also analyzed in this section are those experiments that involve isolation and transfer of macrophages with bound antigen as well as attempts to sensitize lymphocytes in uitro upon contact with macrophages. Tissue culture experimentation involves mainly the study of the proliferative or antibody response of lymphocytes to antigen in the presence or absence of macrophages.

A. In Vivo EXPERIMENTS 1. Fate of Antigen in Viuo For several years immunologists have attempted to follow antigen in experimental animals. One approach used was to label the antigen with isotopes, such as lZ5I,3gS, or 32P, and, subsequently to study the tissue extracts for the presence or absence of labeled material and/or immunogenic moieties. The latter was investigated by immunizing nonimmune or boosting immune animals with the tissue extracts. These early approaches, of the 1950s and early 196Os, have been reviewed (Humphrey, 1960; Campbell and Garvey, 1961, 1963). The main information from these extensive studies is that small immunogenic moieties could be extracted from tissues many weeks and months after the initial injection of antigen. Campbell and Garvey identified small fragments of antigen

98

E. R. UNANUE

which they believed were associated with ribonucleic acid ( RNA) . Yet, information on the localization of the antigen and its role in uivo was not elucidated by these studies. A better means to follow antigen in vivo utilized fluorescent antibody ( Kaplan et al., 1950) or radioautography. The fluorescent method, although giving precise histological localization, was not sensitive enough for the detection of only a few molecules of antigen. The method of choice then has been to use proteins tagged with radioiodine at a high specific activity. This method has been used extensively and successfully by Nossal, Ada, and collaborators at both the light and electron microscopic levels. The work of the Australian group has been recently summarized in a book to which the reader is referred for a detailed analysis of their results (Nossal and Ada, 1971). The usual method is to inject antigen labeled with carrier-free lZ5Iat high specific activity (Greenwood et al., 1963; McConahey and Dixon, 1986) and to trace it in the lymph node or spleen, hours or days after local or intravenous injection, respectively. The tissues can be processed for radioautography and/or immunofluorescence ( in order to investigate for antibody) at the same time. The latter procedure enables us to locate the labeled antigen and indicates the presence of immunoglobulin ( I g ) either extracellularly or associated with cells ( McDevitt et al., 1966). Fates of a synthetic polypeptide labeled externally with lZ5Iand internally with tritium were compared and gave similar results, indicating the validity of the external label (Humphrey et al., 1967). It is clear that radioautography, although showing precise histological localization, does not yield quantitative information. Some quantitative aspects of radioautography are discussed in a paper by Ada et al. (1966). Still, identification of antigen in a tissue does not necessarily imply that it participates in immune induction. With antigen tracing studies, two main areas or sites in lymph nodes where antigen becomes trapped have been identified: one is associated with typical macrophages situated in medullary areas or perifollicular zones of lymph nodes, or perifollicular or marginal zones of spleen; the second is associated with the lymphoid follicles. The characteristics of antigen trapping in these two areas appear to differ (Ada et al., 1964; Nossal et at., 1964; Nossal, 1967; Humphrey and Frank, 1967; Humphrey, 1969; Nossal and Ada, 1971; Hanna and Hunter, 1971). At the first site, antigens enter into a node via afferent lymphatics and are readily taken up by the macrophages of the medulla and perifollicular areas. The experimental antigens most frequently used include albumins, hemocyanins, Salmonella flagellin, Ig, ferritin, and synthetic polypeptides. Ultrastructurally, flageIlin-lY was readily found in association with macrophages and usually in intracellular vesicles or in

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

99

phagolysosomes ( Nossal et al., 1968a). The macrophages responsible for this uptake showed morphological characteristics in general similar to macrophages isolated from bone marrow or the peritoneal cavity. Of importance was the observation that injection of antigen into tolerant, unresponsive recipients led to its uptake by these macrophages to an extent comparable to that exhibited by macrophages from normal animals ( Humphrey and Frank, 1967; Humphrey, 1969). Injection of autologous proteins, such as Ig, also led to their trapping by extrafollicular macrophages, although no attempts were made to exclude aggregates from the preparation of autologous Ig. The amount of antigen associated with these macrophages rapidly declined and became difficult to demonstrate after a few days (Nossal, 1967). Most likely, a great part of the antigen had been lost as a result of catabolism by the macrophage. To this effect, synthetic polypeptides made up of D-amino acids, which are degraded poorly by the enzymes of macrophages, persisted in the sinus and perifollicular macrophages for prolonged periods of time and probably throughout the life of the experimental animal (Medlin et al., 1970; Janeway and Humphrey, 1969). Uptake of antigen by macrophages was not reduced by X-irradiation (G. M. Williams, 1966b; Jaroslow and Nossal, 1966), but was greatly enhanced if the antigen was bound to antibody as an immune complex (Nossal et al., 1965). The second site of antigen uptake is the lymphoid follicles. White (1963), by using fluorescent antibody, was the first to call attention to the presence of Ig in a weblike pattern among cells of lymphoid follicles ( 1963). Later, Nossal, Ada, and collaborators reported on the appearance and persistence of flagellin-lZ5Iin these structures and on the marked enhancement of the trapping mechanism in immune animals (Nossal et al., 1965; Jaroslow and Nossal, 1966; Lang and Ada, 1967). Humphrey and Frank's work (1967) on follicular localization of albumins or hemocyanins in normal and tolerant rabbits has clarified our understanding of the mechanisms of localization of antigen. They found, on the one hand, that adult rabbits made tolerant by injections of antigen at birth were unable to localize antigen in the lymphoid foIIicles (although, as mentioned before, the medullary macrophages were perfectly capable of antigen uptake) ; normal rabbits, on the other hand, localized the antigen in the follicle but not until several days after its injection, at a time when antibody synthesis had started. Further studies in which radioautography was combined with fluorescent antibody methods indicated that both antigen and Ig were present on the same sites in the follicles (Balfour and Humphrey, 1966). It should be noted that early studies showed localization of flagellin-lz~Iin the follicles of tolerant rats (Mitchell and Nossal, 1966), but this trapping was most likely asso-

100

E. R. UNANUE

ciated with the presence of natural antibodies in the serum (G. M. Williams, 1966a). Hence, antigen must be bound with antibody in order to be localized in the lymphoid follicles. As would be expected, Ig-lz5I (either autologous or foreign) injected subcutaneously could be found localized in the follicles of draining nodes (Hanna et al., 1968; Herd and Ada, 1969a,b). Herd and Ada indicated that this localization appeared to be associated with the presence of the Fc portion of the Ig molecules ( 1969a,b) . The third component of complement ( C ) , C3, was also identified in the follicles together with antigen and Ig; furthermore, there was good evidence that C could fix in uitro to these areas ( Gajl-Peczalska et al., 1969). The localization of C in the follicles adds further proof that antigen is found at this site as part of an immune complex. The fate of antigen in the medullary and perifollicular macrophages and in follicles is strikingly different. By ultrastructural examination it was possible to detect most of the antigen in the follicles in between closely packed cells on what appeared to be membranes of specialized dendritic-type cells (Mitchell and Abbott, 1965; Szakal and Hanna, 1967; Nossal et al., 1968b; Cottier and Sordat, 1971). Another characteristic of antigen localized in the follicle was its persistence for prolonged periods of time, whereas most of the antigen associated with the extrafollicular macrophages was eliminated. Finally, in contrast to perifollicular and medullary macrophages, the antigen-trapping mechanism appeared to be sensitive to X-irradiation (G. M. Williams, 1966b; Hanna and Hunter, 1971; Nettesheim and Hammons, 1971). Table I is a summary of the antigen-trapping systems. Two points are not yet clarified with respect to the follicular trapping system. They are (1) the kinds of cells that trap antigen and ( 2 ) the manner in which antigen or trapping cells reach the follicles. Lymphoid follicles appear to have a complicated web of interdigitating cell processes, from lymphocytes in various stages of differentiation and nonlymphocytic cells characterized by large pale nuclei. The nonlymphoid cell has received several descriptive names : dendritic macrophage, reticulum cell, and dendritic cell. The term macrophage has been abandoned since small amount of antigen is inside the vesicles of this cell in contrast to typical macrophages. The evidence that this cell is a reticulum cell is not clear and, hence, the descriptive, noncommital term of dendritic cell appears most appropriate. Based on observations of fluorescent antibody, White postulated that the antibody present in follicles was localized on these special “dendritic”-type cells. As previously mentioned, by combined radioautography and immunofluorescence antigen and antibody were identified at the same loci. Ultrastructural studies showed that the antigen (or the antibody) was localized in between

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

101

TABLE I ANTIGEN-TRAPPING SYSTEMSIN LYMPHNODESAND SPLEEN Extrafollicular Associated with typical macrophages Many antigens are taken up without apparent need of antibody; antibody increases trapping Tolerance does not affect binding of protein antigen Not affected by X-irradiation Most of the antigen is rapidly degraded-persistence of antigen is short May participate in early immune induction Trapping does not necessarily correlate with immunogenicity

Follicular Associated with dendritic cells and perhaps with some type of lymphocytes Antigen must be complexed to antibody (and C?) in order to bind to the follicles (foreign Ig can bind directly) Tolerant animals do not trap antigen unless natural antibody is present in serum Affected by X-irradiation Antigen persists for long periods May participate in immune memory Trapping follows early immune induction and does not insure immunogenicity

tightly packed cells. In many instances the antigen was clearly localized on the membrane of the dendritic cell in close contact to the surrounding lymphocytes. The problem in this study is the proper identification of the site of localization of the immune complex. As Nossal et al. carefully stated in their ultrastructural study in 1968(b), “it could not be claimed that every thin process seen in the follicles came from the reticular cell, nor was every antigen depot necessarily associated with a reticular cell process. . . . In fact in many cases where a complexity of processes was encountered, it was not possible to determine from which type of cell a particular process was derived.” We now know that follicles contain B-type lymphocytes which have a membrane receptor for C3 (Bianco et al., 1970) and can bind antigen-antibody complexes, Therefore, an important new problem is to determine the extent to which these lymphocytes contribute to the binding of antigen-antibody complexes in the follicles. Notwithstanding this reservation and assuming that, as expected, the dendritic cell bears most of the antigen-antibody complexes in the follicles, we must still question the nature of this cell. Is this a new type of cell differing completely from the monocyte-macrophage line while also having a membrane receptor for Ig? Or, is it possible that this cell may represent a stage of maturation or functional differentiation of monocyte-macrophages? I raise this latter possibility because of indications that monocytes and macrophages do exhibit some degree of functional changes which might make them express one function or

102

E. R. UNANUE

another to a variable extent during their differentiative process. The physiological behavior of the brain macrophage may be pertinent to keep in mind-in inflammation the microglia appears as a round cell, whereas under normal conditions it is a small cell with interdigitating processes. Similar morphological changes are seen in macrophages of the omentum (Fischer et al., 1970). The manner in which antigen (with antibody) becomes trapped in the follicles is not determined. The two obvious considerations are that ( a ) the antigen-antibody complexes enter as free complexes to become trapped by the cells or ( b ) the complexes enter already bound to the membranes of cells. White and collaborators (1967) have done a very detailed analysis of this process in the chicken spleen and are of the impression that the dendritic cells outside the arteriolar sheath trap the antigen, move along the sheath, and trap the lymphocyte to form the reactive follicle. In the mammalian spleen there is also some suggestion that antigen is taken up by the cells of the marginal zone and is carried into the follicles (Nossal et al., 1966). The association between uptake of antigen by extrafollicular macrophages or follicular cells and the immune process is not known. It is understandable that this problem must depend for its solution upon a series of indirect pieces of evidence. For example, we know that extrafollicular macrophages handle antigen in a way very similar to that of peritoneal macrophages (Ada and Lang, 1966; J. M. Williams and Ada, 1967); we also know that some antigen bound to transplanted peritoneal macrophages is immunogenic. ( Some of these macrophages even localize to perifollicular areas after injection. ) The logical assumption, then, is that antigen bound to extrafollicular macrophages of the spleen and lymph node should be available for immune recognition at least for some time. Unfortunately little is know about the turnover and fate of extrafollicular macrophages. Insofar as the follicular antigen-trapping system is concerned, we know that it is not necessary for immune induction. Induction of antibody precedes localization of antigen in the follicle (Humphrey and Frank, 1967). Indeed, the first histological evidence of antibody formation usually involves the perifollicular areas ( Oort and Turk, 1965; Leduc et aZ., 1968). It is also apparent that follicular localization of antigen is not a necessary condition for induction of antibody (Humphrey, 1969). It has been reasonably speculated by many that the follicular trapping of antigen may be a mechanism for maintenance of immunological memory. In order to fully understand the meaning of this process of follicular localization of antigen, information must be sought on dynamics of cells in the follicles, effects of concentration of antibody, reaction of lymphocytes to antigen-antibody-C complexes, origin of dendritic cells, etc.

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

103

2. Persistence of Immunogenicity Potentially immunogenic molecules can persist in tissues for prolonged periods of time. Whether some or all of these molecules are associated with macrophages and the extent to which they contribute to the maintenance of a state of immunity (i.e., of memory) have not been ascertained. The experimental approaches that have been used to study persistence of immunogens are diverse. Humphrey in 1960 reviewed an early technical procedure employed during the 1950s. It consisted of making a tissue extract days or weeks after injection of antigen into one experimental animal and testing for immunogenicity by injection into a second animal. Indeed, immunogenic material from pneumococcal polysaccharide, heniocyanin, Ig, or albumins was identified in tissues several days after its injection. A different approach consisted of transferring immune cells into animals that previously had received antigen and which were immunoincompetent ( by either tolerance induction or X-irradiation) . Murine lymphocytes primed to human albumin made antibody when transferred into tolerant mice that had received the antigen 10 weeks previously (Mitchison, 1965). In this system, but with X-irradiated hosts as recipients, immunogens from sheep red blood cells (SRBC) and Escherichia coli lipopolysaccharide were found to persist in mice for as long as 14 and 45 days, respectively (Britton et d.,1968). However, the immunogenicity of human albumin was short-lived, about 17%hours (Britton and Celada, 1968). These experiments, therefore, indicated that potential immunogens persisted although their quantity varied from antigen to antigen. Is the persisting immunogen necessary for the production of immune memory? There have not been many experiments designed to answer this important question. The evidence suggests that the initial interaction with antigen leads to a population of long-lived, memory lymphocytes that do not necessarily require continuous antigenic stimulation ( Celada, 1967). Nevertheless, the retained immunogen may add to an optimal state of low active immunity or of immune memory. Celada (1967) transferred lymphocytes from immune mice into X-irradiated recipients and challenged these with antigen at periods of up to 6 months after transfer. Because his dose-response system was well-calibrated and sensitive, he estimated that the capacity of the transferred cells to mount a secondary response decayed in two phases: during the first month with a half-life of 15 days, and thereafter of 100 days. The reader should recall that the decline of immunogenicity of the same antigen in viuo was 17%hours. Hence, the memory response to albumin did not appear to be dependent upon the presence of retained antigen.

104

E. R. UNANUE

The well-known fact that antibody can block the immune response at the level of the antigenic stimulus has been used to estimate whether persisting immunogen may be active in uiuo. Britton and Moller (1968) observed an interesting cyclical fluctuation in the number of antibodysynthesizing cells to E. coli lipopolysaccharide in mice for as long as 50 days. The phenomenon was explained by feedback suppression of the response by antibody, implying that antigen persisted throughout the observed period of time and became periodically available for immune recognition. Graf and Uhr (1969) injected rabbits with bovine albumin. Two to 3 weeks later the rabbits were bled; the antialbumin antibody was removed from the plasma by a special method, and the absorbed plasma was reinfused into the rabbits. This resulted in exclusive removal of the specific antibody. Subsequently, the level of antialbumin antibody rose in the circulation, The interpretation was that the antibody in the serum was, in fact, playing a homeostatic, inhibitory role, and once removed, immunogenic material again became available for immune stimulation. The induction of optimal immune memory to hemocyanin required the persistence of antigen for at least 3 to 4 weeks as also judged by effects of passive antibody (Cerottini and Trnka, 1970). The site where immunogenic antigen persists has not been determined. Perhaps this immunogen is on the dendritic cells of the follicles or on typical extrafollicular macrophages or even is associated with other cells. Immunogens could be shown to persist in macrophages for at least 3 weeks by the use of the macrophage transfer system (Unanue and Askonas, 1968b). Cells containing antigen could be isolated from spleens of mice as long as 7 months after antigen injection (Mitchison, 1969a). In summary, there is strong evidence that a small amount of immunogen can persist for prolonged periods of time in tissues. Its availability to immunocompetent cells may depend upon the level of extracellular antibody and, perhaps, anatomical factors. The exact locus of the antigen is not known, although there is a possibility that it could be associated with RES cells. The persisting immunogen, although not entirely necessary, may contribute to the maintenance of memory; it persists in extremely small amounts, probably of the order of 0.001%or less of the injected dose ( McConahey et aZ., 1968).

3. Immune Response to Antigens Differing in Their Degree of Phagocytosis There is an association between the degree of uptake of antigen by macrophages of the lymphoid organs and immunogenicity. Antigens, such as red cells and bacteria, trigger strong immune responses with

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

105

ease, while soluble proteins or polysaccharides trigger poor immune states. Experiments that show the clearest relationship between uptake of antigen by macrophages and immunogenicity are those that employ as antigens proteins differing in their state of polymerization (Dresser, 1962; Claman, 1963; Frei et al., 1965; Biro and Garcia, 1965; Spiegelberg and Weigle, 1967; Golub and Weigle, 1969). The two proteins most commonly employed are albumins and Ig. Solutions of albumins or Ig are made up of monomers and a variable quantity of polymers of differing sizes. When injected into an animal, the monomers are poorly taken up by the RES and circulate for some time; the polymers are immediately taken up and disappear from the circulation. Monomers can be obtained free of polymers either by ultracentrifugation (Dresser, 1962; Claman, 1963) or by biological filtration (Frei et al., 1965; Golub and Weigle, 1969). Biological filtration consists of injecting the solution containing a mixture of monomers and polymers into an experimental animal which is exsanguinated some hours later. The serum contains the monomers inasmuch as the polymers have been removed from the circulation by the RES. For experimental purposes, polymers are readily obtained by aggregation of the protein by either physical or chemical means. There is general agreement that removal of the polymers from a solution of foreign albumin or Ig by ultracentrifugation (Dresser, 1962; Claman, 1963) or biological filtration (Frei et al., 1965; Golub and Weigle, 1969) results in a significant loss of the immunogenic potency of these proteins. Furthermore, one or several injections of albumins or Ig in their monomeric form may lead to a state of tolerance (Dresser, 1962; Frei et al., 1965; Biro and Garcia, 1965; Dresser and Mitchison, 1968; Golub and Weigle, 1969). In contrast, the polymeric form of albumin or Ig is highly immunogenic. Some commercial preparations of albumins are contaminated with endotoxin (Dvorak and Bast, 1970) and are unsuitable for evaluating immunogenicity, inasmuch as endotoxins are powerful immunological adjuvants. Yet, the higher immunogenicity of polymers cannot be accounted for by contamination with endotoxin, since polymers free of endotoxin are themselves highly immunogenic (Fauci et al., 1971; Schmidtke and Unanue, 1971a). These observations led to the “direct access” theory of tolerance, i.e., the interaction of lymphocytes “directly” with monomers produces immunological paralysis, whereas the interaction after macrophage uptake produces immunity ( Dresser and Mitchison, 1968). Many events apart from uptake by the RES could explain the immunogenicity of polymers and monomers. It is possible that aggregated moieties trapped in lymphoid tissues not necessarily by the RES may be the best form of antigen and at the best site to trigger a lymphocyte to proliferate and differentiate. Perhaps the mere

106

E. R. UNANUE

fact that these aggregates are readily removed from the serum and extracellular tissues favors the immune response, since specific lymphocytes are then free to interact with antigen now concentrated in critical lymphoid organs. Nevertheless, there is direct evidence that part of the above effect can be explained by the uptake of polymeric molecules by the macrophages. This evidence derives from experiments in which macrophages containing antigens are isolated from experimental animals and transplanted into a syngeneic recipient which is then tested for its capacity to achieve an immune state. The recipient makes a strong response to some macrophage-associated antigens. The results of the experiments involving antigen bound to live macrophages are detailed in a later section. In contrast to the above situation, there are other experiments that show an immunologically detrimental effect of macrophages. In these experiments a highly phagocytizable antigen, such as a red cell or an aggregated protein, is injected into the peritoneal cavity of normal mice or mice having a rich macrophage exudate. Such mice exhibit smaller immune responses than counterparts injected intravenously (i.v. ) (where the antigen has not had contact with macrophages before its arrival in the spleen) (Perkins and Makinodan, 1964; Franzl and Morrello, 1966; Perkins, 1970). Moreover, if before intraperitoneal ( i.p. ) injection of antigen, a particle capable of blocking the RES is injected i.p., the immune response is restored. In these experiments there is a direct relationship between the amount of radiolabeled red cell or aggregated Ig trapped in the spleen after injection i.v. or i.p. and the degree of immunogenicity-those mice injected with antigen i.v. or i.p. into a cavity with blockade of the RES had about ten to twenty-fold more antigen in their spleens than mice injected i.p. without RES blockade. These experiments clearly indicate that ( 1 ) depriving a lymphoid organ of antigen results in poor immunogenicity, and ( 2 ) local peritoneal macrophages can degrade antigens and reduce the amount of effective immunogenic stimulus to reach lymphoid sites. The catabolic function (or negative immunogenic function) of macrophages is readily demonstrable under these conditions. These experiments do not rule out that in the spleen at least part of the response could be regulated or associated with the antigen taken up by the local RES. Still, since local lymphoid macrophages are in a better anatomical relationship with lymphocytes than are peritoneal macrophages, the former would be more suitable for helping in immune induction (aside from catabolizing antigen), In contrast, local peritoneal macrophages have little chance for interaction with lymphoid cells, since there are relatively few lymphocytes in a normal peritoneal cavity. An alternative and not mu-

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

107

tually exclusive explanation is that direct interaction with some antigens and splenic lymphocytes results in immune induction. The above experiments gave different results with antigens that were poorly taken up by macrophages. Injection of poorly phagocytizable antigens into the peritoneal cavity of mice with a rich macrophage exudate results in immune responses comparable or higher than those obtained in mice injected i.v. (Unanue, 1968a). In this case, free antigen flowing directly into the spleen is taken up very poorly, and the presumed arrival into the spleen of peripheral macrophages with antigens, although small in number, results in enhanced immunity. It is reasonable to conclude that many antigens taken up by macrophages at nonlymphoid sites, such as the peritoneal cavity or liver, may be lost from interacting with lymphocytes because the macrophages destroy the antigen, and the macrophage-associated antigen that is not destroyed has little opportunity at these sites to interact with immunocompetent elements. Hence, the helper role of macrophages (or positive immunogenic function) should be sought mainly among macrophages in intimate association with lymphoid elements such as in spleen or local lymph nodes.

4. RES Blockade There have been numerous attempts to study immune responses to antigen administered at the time of an RES blockade (Gay and Clark, 1924; Cannon et al., 1929; Stern et al., 1955; Cruchaud, 1968, Sabet and Friedman, 1969 ) . Reticuloendothelial blockade consists of saturation of the RES as a result of overloading with a given colloid and is measured by estimating the rate of elimination of a test colloidal suspension from the circulation. The extent and duration of blocking depend upon the dose of colloid administered. Colloidal carbon which has been used frequently produces RES blockade for up to 3 days. Other colloids, such as saccharated iron oxide depress for only a few hours (Benacerraf et al., 1956) and are followed by a hyperphagocytic phase. Blockade of the RES is never complete. Also large doses of some of the materials injected may be “toxic” for experimental animals. For these reasons, it is obvious that many studies of the immune response during RES blockade are of dubious value and are difficult to interpret conclusively. In some instances, RES blockade has even resulted in increased antibody formation (Lewis, 1954; C. R. Jenkin et al., 1965; Fisher, 1966). Recently, Friedman and collaborators have carefully examined the immune response to SRBC in mice that have received injections of colloidal carbon (Sabet and Friedman, 1969; Sabet et al., 1969) or ethyl stereate ( Melnick and Friedman, 1969). Their experiments show quite clearly that a significant depression in the number of antibody-forming cells

108

E. R. UNANUE

can be consistently obtained if the blocking agents are given usually a day or two before injection of antigen. Injection of carageenan, a high molecular weight polygalactose specifically cytotoxic for macrophages, can result in depression of delayed hypersensitivity in sensitized guinea pigs (H. J. Schwartz and Leskowitz, 1969). More important in the context of our review is that injection of antigen and carageenan into unsensitized guinea pigs led to depressed cell-mediated reactions (H. J. Schwartz, 1971).

5. Use of Antimacrophage Serum Attempts to use antibodies against a cell line in order to evaluate its role in a given phenomenon date as far back as 1899 when Metchnikoff proposed to use antimacrophage serum (AMS). However, it was not until the past two decades that the use of specific antibodies to cells became more popular, probably as a result of Humphrey’s successful experiments (1955a,b) in which he showed that Arthus reactions in guinea pigs depleted of circulating leukocytes by antipolymorphonuclear leukocyte sera were markedly reduced. In this way it was established that the infiltration by polymorphonuclear leukocytes was essential in the production of the inflammatory reaction. Heterologous antibodies to polymorphonuclear leukocytes ( Humphrey, 1955a,b; Cochrane et al., 1965), mast cells (Valentine et al., 1967), and lymphocytes (Lance, 1970) have been used quite extensively. Heterologous antibodies specific for macrophages have been obtained now and used to study different aspects of the function of these cells. Macrophages have specific, surface antigenic determinants identifiable by their reaction with heterologous antibodies. Thus, for example, antibodies made in rabbits against mouse macrophages react only with macrophages and spare lymphocytes and hematopoietic cells ( Cayeux et al., 1966; Unanue, 1968b; Despont and Cruchaud, 1969). In order to obtain antibodies to macrophages, experimental animals have been immunized with purified or semipurified suspensions of macrophages (usually peritoneal or alveolar) either lightly, i.e., with relatively few cells and for a few times, or strongly (Boros and Warren, 1971), i.e., with large number of cells, several times, and usually associated with strong adjuvants. In the former cases, selected antisera can be shown to react with macrophages exclusively; in the latter cases, all antisera have to be absorbed first with lymphocytes in order to make them specific for macrophages. The specificity of AMS has been tested by agglutination, cytotoxic, or immunofluorescent tests ( Fig. 1).By each of these methods, appropriate antibodies can be shown to bind to macrophage membranes and to kill the cell in the presence of C. Lymphocytes are

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

109

FIG.1. Photomicrograph of peritoneal exudate cells after reaction with antimacrophage serum (AMS). Guinea pig cells were incubated first with rabbit antiguinea-pig macrophage serum and second with fluorescein-conjugated sheep antirabbit I&. The macrophage reaction is strongly positive, whereas that of the lymphocytes (arrows) is negative. The photomicrograph has been overexposed in order to bring out the faint negative lymphocyte. (From J. D. Feldman and Unanue, 1971, p. 289.)

not killed by specific AMS indicating that either macrophages contain specific surface determinants or, alternatively, that the lymphocytes have the antigens but in amounts and/or spatial distributions below the level of detection or C fixation. Cohn and Parks (1967) reported the presence of natural macroglobulin antibodies in bovine sera directed against mouse macrophages and erythrocytes. The effects of antimacrophage antibodies were studied both in vitro and in uiuo. Macrophages treated i n uitro with the antibodies (in the absence of C ) adhered poorly to glass (Unanue, 196813; Hirsch et al., 1969) and showed reduced uptake of some particulate (Unanue, 1968b; Hirsch et al., 1969; Jennings and Hughes, 1969; Argyris and Plotkin, 1969) and soluble antigens (Unanue, 196813). This blocking of the uptake probably takes place by direct or indirect interference by AMS with the site on the cell surface to which the antigens bind. Macrophages cultured for several days in the presence of bovine sera containing AMS activity had increased pinocytic activity, increased levels of acid hydrolases, and larger numbers of lysosomes (Cohn and Parks, 1967). Experimental animals were injected with AMS i.v., i.p., or subcutaneously. Injection of AMS into the peritoneal cavity produced rapid agglutination and/or death of the local macrophage population (Unanue,

110

E. R. UNANUE

1968b, Boros and Warren, 1971). This effect, as expected, was transitory, and by a day or two after injection of AMS, the peritoneal cavity contained newly formed macrophages. Macrophages with nuclear aberrations were frequently seen within hours after AMS injection (Unanue, 1968b); peritoneal lymphocytes were not affected by injection of AMS (Unanue, 1968b). The i.v. injection of specific AMS led to a transitory drop in monocyte counts (Boros and Warren, 1971) but not in the number of circulating lymphocytes (Hirsch et al., 1969; Boros and Warren, 1971). The phagocytic index was depressed after several injections of AMS (Hirsch et al., 1969). There were few histological studies of spleen or lymph nodes following AMS injection-some congestion of splenic red pulp and death of local macrophages were reported (Unanue, 1968b; Muller-Hermelink and Muller-Ruchholtz, 1971) . Antimacrophage serum has been used to study three types of phenomena: ( 1 ) immune responses to a variety of antigens; ( 2 ) delayed hypersensitivity responses in sensitized animals; and (3) infectious processes. Although only the first interests us in this review, a brief analysis of the other two will be made. The use of AMS in in vitro immune responses will be covered in a later section. The use of AMS for evaluating the role of macrophages in in vivo immune reactions has not been very rewarding, In only some instances did AMS, given at a time close to the administration of antigen, produce some depression of immunity ( Argyris and Plotkin, 1969; Isa, 1971); however, the specificity of an AMS was not altogether clear. Moreover, it has been reported that injection of some batches of normal foreign sera might depress immune responses ( Despont and Cruchaud, 1969). The lack of consistent immunosuppressive effects by AMS (Unanue, 1968b; Despont and Cruchaud, 1969; Panijel and Cayeux, 1968) is difficult to interpret, since no clear answer can be given on whether or not macrophages play a role in immune induction. It may well be that AMS does not reach tissue macrophages at adequate concentrations or during a sufficient period of time to block macrophage function. Along these lines, it is known that antilymphocyte serum is an effective immunosuppressant mainly because it binds and opsonizes circulating lymphocytes (Lance, 1970). Immune reactions not dependent on circulating T cells are not affected by antilymphocyte sera. Antimacrophage serum has been shown to be more effective in reducing delayed hypersensitivity elicited in sensitized animals. Thus, Heise and Weiser (1970) noted a small depression of skin tuberculin reactions in guinea pigs treated 2 days earlier with AMS. J. D. Feldman and Unanue (1971) resorted to an adoptive transfer system: irradiated guinea pigs that received immune lymphocytes plus bone marrow cells (as a

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

111

source of macrophages ) showed depressed skin reactions if treated with AMS. Boros and Warren (1971) showed reduced formation of immune lung granulomas in mice treated with AMS. Most likely the effectiveness of AMS is due to its capacity to reduce levels of circulating monocytes, the cell line responsible for the nonspecific exudate of cell-mediated responses. Still, the results published so far are not impressive, and, perhaps, the use of stronger AMS or of continuous injection of the antisera might produce clearer answers. One point of importance in explaining the action of AMS in vivo is whether AMS can bind to macrophage precursors. J. D. Feldman and Unanue (1971) noted that AMS was not binding to cells in bone marrow, the tissue containing a rich pool of precursor cells. More recently, Virolainen and collaborators (1972) have shown directly that macrophage precursors did not react with AMS as did the mature monocytes and macrophages. These results indicate that the AMS-reacting antigen( s ) in the macrophage develops during the process of differentiation and suggest that continuous production of this cell line can occur despite the presence of extracellular AMS. Administration of AMS potentiates some viral infections ( Hirsch et al., 1969). In these infections the degree of viremia depends upon the capacity of the macrophage to take up viruses from the circulation. For example, AMS treatment hastened mortality of mice with vesicular stomatitis virus (Hirsch et al., 1969), since depression of the macrophage system affects the degree of viremia and the consequent amount of virus seeded into the brain. Similar results were obtained with mice infected with yellow fever virus (Panijel and Cayeux, 1968). In contrast, treatment with AMS of mice infected with encephalomyocarditis virus lowered their mortality-apparently in this case the macrophage permitted the survival and multiplication of the virus (Panijel and Cayeux, 1968) . 6. Transfer of Macrophages

A system employed to study the immunogenicity of macrophagebound antigens consists of isolating live macrophages containing the antigen and transplanting them to syngeneic hosts which are then observed for their immune responses (Cohn, 1962; Gallily and Feldman, 1967; Argyris, 1967; Argyris and Askonas, 1968; Unanue and Askonas, 1968b; Unanue, 1969; Mitchison, 196913; Spitznagel and Allison, 1970b; Cruchaud et al., 1970). Hence, the immune response to an antigen initially carried exclusively by macrophages in the absence of free antigen is studied. The macrophages in these experiments were obtained mainly from

112

E. R. UNANUE

the peritoneal cavities of mice or rats after injection of an irritating or inflammatory substance, such as glycogen, proteose peptone, or thioglycollate broth. These peritoneal macrophages are well-differentiated cells that have originated from a rapidly proliferating pool of cells in the bone marrow (Volkman, 1966; Volkman and Gowans, 1965a,b). In this experimental system there have been few comparisons of macrophages from different organs or of macrophages showing different degrees of maturation. However, when the immunogenicity of bacteriophage T4 bound to Kuppfer cells was compared to those bound to peritoneal macrophages, the former group showed reduced immunogenicity ( Inchley, 1989). A critical analysis of the variables that are introduced in such a system in order to purify the Kuppfer cell was not done. Spitznagel and Allison compared the handling of antigen by normal macrophages and macrophages obtained after thioglycollate injection. The latter took up more antigen, although weight by weight the antigen in both cells as assayed in the transfer system had comparable immunogenicity ( 1970b). Other unpublished experiments (mentioned by Howard, 1970) claim that bovine albumin is of comparable immunogenicity when bound to isolated Kuppfer cells, alveolar macrophages, or peritoneal macrophages (cells were isolated and then transplanted into mice to assay for immunogenicity). The technique of the macrophage transfer system is shown in Fig. 2. The macrophages are exposed to the antigen in uiuo, i.e., after an i.p. injection, or in oitro; the macrophages are washed well after the uptake of antigen in order to eliminate extracellular antigen, and injected i.p. into histocompatible recipients. The recipient mice are subsequently studied either for their primary or secondary antibody responses and/or for a delayed hypersensitivity response. The antigens used in the system have included soluble proteins such as albumins, foreign Ig, hemocyanin, lysozyme or ovalbumins, or particulate antigens comprising Escherichia coli, red cells, or Shigella. In many instances the antigens are tagged with an isotope as in the cases of proteins with radioiodine, red cells with W r or 1251, or bacteria with szP. This allows an estimate of the amount of antigen associated with the live macrophage before transfer. The fates of the injected macrophages as well as the site where these cells meet with immunocompetent cells are not completely established. The problem with following macrophages after their injection is finding a suitable label which is sensitive, nontoxic, and not reutilized if released by the cell. Roser and collaborators (Roser, 1965, 1968; Russell and Roser, 1966), by using lg8Auas a marker, obtained evidence that a small number of injected macrophages end up in lymph nodes, liver, and spleen. In the spleen, macrophages were localized to the marginal zone

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

PRIMARY

Respa

or

A A

113

SECONDARY

kw

9 Serum Ab

Serum Ab

FIG. 2. Illustration of the technique of the macrophage transfer system, The antigen (in triangles) is injected i.p. and the peritoneal macrophages are isolated 30 minutes to 1 hour later. Alternatively, the macrophages can be isolated and incubated in uitro with the antigen. After uptake of antigen the cells are washed several times and then injected (i.p., i.v., or subcutaneously) into a normal recipient. If the antigen is tagged with an isotope, the radioactivity in the cells is measured before injection. The recipients are studied for their primary or secondary response to antigen.

and in lymph nodes to the medulla and perifollicular areas. Although all indications are that lg8Auis a good tag for following migration of macrophages, some released material may be reutilized. Gillette and Lance ( 1971) followed macrophages labeled with either 1251-5’-iodo-2’deoxyuridine ( UDR) or W r . (Both labeling methods gave comparable results.) After i.v. or i.p* injection the label was found in liver and spleen, and, as regards the migration pattern of peritoneal macrophages after stimulation by different inflammatory substances, it was found that more of the “stimulated tagged cells than controls left the peritoneal cavity and appeared in the spleen. Several characteristics of the immune response to antigens transferred in live macrophages have been studied. It is clear that immunocompetent lymphocytes must be present in order to interact with antigen and initiate the immune process. Transfer of macrophages into X-irradiated recipients ( Argyris, 1967; Unanue and Askonas, 1968b; Cruchaud and Unanue, 1971a,b) or into recipients immunologically tolerant ( Mitchison, 196913) to the antigen does not lead to immune responses. This type of experiment clearly emphasizes the well-known fact that macrophages are not involved in antibody synthesis. Antigen on dead macrophages (Mitchison, 1969b) or those from allogeneic (Mitchison, 1969b) or xenogenic (Unanue and Askonas, 1968b) donors (which are rejected by the host)

114

E. R. UNANUE

IMMUNOGENICITY OF

Antigen

TABLE I1 ANTIGENSI N THE M . ~ C R O P H . 4 G I ~ - T l ~ A N s F ISYSTIs:Ma ~~R

Donor of macrophages Recipients

Albumin (bovine)

Mice

Mice

Albumin (human)

Mice

Mice

Escherichia coli

Rabbits

Mice

?-Globulin (bovine)

Iiabbits

Rabbits

7-Globulin (human )

Mice

Mice

Hemocyanin (from keyhole limpets, Megathuru crenulala )

Mice

Mice

Main findings (Refs.)* Immunogenic (considerably more than free antigen when priming) (1) Macrophages catabolized most of the antigen taken up (1) Viable macrophages were required (1) Macrophages from tolerant mice were effective (1) Irradiation of donors of macrophages decreased immunogenicity of antigen in macrophages (1) Transfer into allogeneic recipients decreased immunogenicity (1) Immunogenic (results comparable to those with bovine albnmin) (1) Injection of free albumin decreased response to macrophage-bound albumin ( 2 ) Most of the antigen was catabolized except for a small fraction-some of the retained antigen was on surface of the membrane and accounted for a great part of the immunogenic moieties (i.e., trypsin-inhibited immunogenicity) (3) Immunogenic (4) Immiinogenicity persisted (4) Immunogenic (5) Irradiated recipients did not respond to transfer unless reconstituted with lymph node rells (5) Macrophages from irradiated (750 It) donors had decreased immunogenicity of antigen in macrophages ( 5 ) Immunogenic (ti) Inhibition of RNA synthesis had no effect on the immunogenicity of antigen in macrophages (6) Immunogenic (less than free antigen) (7) Most of the antigen was catabolized except for a small fraction (7) Immunogenic molecules were retained on plasma membrane (8, 9 ) Immuiiogeiiicity was inhibited by incubating macrophages in vrlro with antibody prior to transfer (9)

TABLE I1 (Continued) Antigen

Hemocyanin (from Maia squinado )

Donor of macrophages Recipients

Mice

Mice

Main findings (Itefs.)b Immunogenicity was increased in X-irradiated macrophages which retained more antigen on the membrane (10) Activated macrophages of cell-mediated immunities have more retained antigen on their membranes (11) Immunogenic (more than free antigen) (12) Antigen in live macrophages was more immunogenic than in dead macrophages (12) Antigen in macrophages placed in diffusion chambers into peritoneal cavity was less immunogenic than when injected i.p. (12) Immunogenicity persisted for long times and was associated with a retained fraction of antigen (13), some of which was on the plasma membrane (81

Lysozyme

Mice

Mice

Ovalbumin

Mice

Mice

Red blood cells (sheep)

Mice

Mice

Shigella

Mice

Mice

4

Immunogenicity was inhibited by incubating macrophages in uitro with antibody prior to transfer (14) Immunogenic (comparable to free antigen) (1) Immunogenic (more than free antigen) (1) Immunogenic (comparable to free antigen) (15) Some red cell antigen may be released from macrophages (16) Immunogenic and capable of eliciting antibody in 550-R irradiated mice (17) Antigen in macrophages from irradiated mice was not immunogenic (17)

This table was published in part in Seminars in Hematology 7,225 (1970). I n most

of these experiments, a transfer system similar to that described in Fig. 2 was used.

* Kev to references:

1. Mitchison, 1969b 2. Spitznagel and Allison, 1970 3. Schmidtke and Unanue, 1971 4. Cohn, 1962 5. Pribnow and Silverman, 1967 6. Cruchaud el al., 1970 7. Unanue, 1969 8. Unanue et al., 1969 9. Unanue and Cerottini, 1970

10. Schmidtke and Dixon, 1972 11. Lane and Unanue, 1972 12. Unanue and Askonas, 1968a 13. Unanue and Askonas, 1968b 14. Unanue, 1972 15. Argyris, 1967 16. Cruchaud and Unanue, 1971a,b 17. Gallily and Feldman, 1967

116

E. R. UNANUE

induces smaller responses than antigen bound to living syngeneic cells. Finally, antigen bound to other cells, such as fibroblasts, induces poor immune responses ( Mitchison, 1969b). Intimate association between macrophages and lymphocytes is required, since macrophages isolated in impermeable diffusion chambers induce diminished immune responses ( Unanue and Askonas, 1968b; Cruchaud and Unanue, 1971a). Antigen bound to live macrophages is immunogenic and can induce primary immune responses, prime for a secondary challenge (i.e., induce a memory state), and induce a state of delayed hypersensitivity. Most of these studies are summarized in Table 11. In all examples, the extent of the immune response depends upon the amount of antigen transferred in the cells ( Mitchison, 1969b; Spitznagel and Allison, 1970b; Unanue and Askonas, 1968b). For example, small amounts of hemocyanin bound to macrophages primed mice for a small y M immune response, whereas larger amounts primed for strong yM (19s) and yG ( 7 s ) responses (Unanue and Askonas, 1968b). A partial inhibition of the immune response to macrophage-bound antigen occurs if normal macrophages are injected simultaneously. The reasons for this phenomenon are not known (Unanue and Askonas, 1968b). Delayed hypersensitivity has been induced in rats (Unanue and Feldman, 1971) and guinea pigs ( Seeger and Oppenheim, 1970a). Such animals exhibited strong cutaneous delayed reactions if previously injected with live macrophages containing antigen. Immune responses to a given amount of antigen injected either bound to live macrophages or in free form have been compared. Generally, those antigens that were taken up poorly by tissue macrophages, if transferred exclusively bound to macrophages, induced much higher immune responses than a comparable amount of free antigen (Mitchison, 196913; Unanue and Askonas, 196813). For example, albumin bound to live macrophages was about 1000-fold more immunogenic than free albumin ( Mitchison, 1969b; Spitznagel and Allison, 1970b). ( Such experiments support the contention that at least part of the strong immune responses to aggregated antigens may be explained by their uptake by tissue macrophages.) A different situation was observed when one compared the immune response of free or macrophage-bound antigen but using antigens which in vivo were taken up to a great extent by macrophages (Unanue, 1969). In this situation, the free antigen induced a higher immune response than the macrophage-bound antigen. For example, foreign red cells (Cruchaud and Unanue, 1971a) or keyhole limpet hemocyanin (Unanue, 1969) produced much greater responses (two to ten fold) when injected free rather than when bound to live macrophages. These results appear similar to those discussed in Section II,A,3. These large particulate antigens are readily trapped by lymph node and

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

117

spleen macrophages and, hence, are in suitable sites for triggering immune states. Thus, transfer of these latter antigens already bound to macrophages does not necessarily enhance the concentration of antigen in a lymphoid tissue; however, such enhancement does occur with antigens that are poorly taken up by tissue macrophages. In the latter case, transfer of antigen that is exclusively bound to live macrophages leads to a stronger immune state as a result of more antigen being concentrated in the lymphoid tissue (see Fig. 3 ) . Responses to macrophage-bound antigen in immune and nonimmune mice can differ. Mitchison (1969b) observed that the relative immunogenicity of albumin administered either macrophage-bound or free was much higher in nonimmune mice than in primed mice. With other antigens, such as hemocyanin, the immune response of primed or nonimmune mice was comparable (Unanue and Askonas, 1968b). The first results can be interpreted as denoting that immune mice with a large number of immunocompetent lymphocytes are less dependent on an antigenconcentrating cell; or, alternatively, that small amounts of antibody in the primed mice serve to concentrate the antigen. The macrophage transfer system has been used to evaluate the role of macrophages in tolerance, in T and B lymphocyte collaboration, in some genetic disorders, and in immunoincompetence produced by X-irradiation. It is now very clear that immunological unresponsiveness is associated with a lack or an incapacity of T and/or B lymphocytes to respond to antigen (Chiller et al., 1971). Macrophages from tolerant

Degree of Phagocytoris In Viva

Enhanceinon! by MacrophaEr-Bound - - - -AntiLen -- -

J

I KLH SRBC

AgpHSA

MSH

HSA

FIG.3. Hypothetical graph illustrating the relationship between uptake of antigen by macrophages and the apparent enhancement (or decrease) of immunogenicity of antigen when injected macrophage-bound or free. Those antigens that are well taken up by macrophages show little improvement or even a decrease when administered macrophage-bound-the reverse is true for those antigens that are taken up poorly by macrophages where the response to macrophage-bound antigeii is much higher than to a comparable amount of free antigen. (KLH) keyhole limpet hemocyanin; ( SRBC) sheep red blood cells; ( Agg-HSA) heat-aggregated human serum albumin; (MSH) hemocyanin from Maia squinado; (HSA) human serum albumin.

118

E. R. UNANUE

animals handle immunogens in the same way as do macrophages from normal animals. In viuo it was shown by radioautography that macrophages from tolerant animals took up as much antigen as macrophages from normal responsive animals ( Humphrey and Frank, 1967). The association of antigen with cells in the follicles of normal and tolerant animals was analyzed in a previous section. By using the transfer system, Mitchison ( 1969b) showed directly that immunogenicity of antigens bound to macrophages of tolerant mice was not impaired and, thus, reemphasized that the state of tolerance was not associated with macrophage dysfunction. Similar results were obtained by Harris ( 1967) by using macrophages from tolerant rabbits and assaying for immunogenicity in vitro. It is also quite clear that the association of most catabolizable antigens with macrophages is a pathway for immunogenicity that does not lead to tolerance induction. In no instance has transfer of antigen bound to macrophages produced an unresponsive state (Table 11). Furthermore, there appears to be competition between the tolerogenic forms of some antigens and macrophage-associated, immunogenic antigen for the immunocompetent cells. This very important observation was made by Spitznagel and Allison ( 1970b), who injected bovine albumin bound to live macrophages together with free albumin into syngeneic mice. The injection of albumin in live macrophages induced strong responses which were markedly reduced by concomitant injection of free antigen. Macrophage-associated antigen can apparently interact with both T and B lymphocytes. This has been inferred from three sets of experimental observations. First, as mentioned above, recipient animals injected with antigen bound to live macrophages developed satisfactory cell-mediated immunity, presumably a manifestation of activity of T lymphocytes (Unanue and Feldman, 1971; Seeger and Oppenheim, 1970a). Along these same lines, if sensitized animals were injected with macrophages containing antigen, they readily developed strong delayed reactions ( superimposed on a nonspecific inflammatory reaction ) (Unanue and Feldman, 1971). A second line of evidence for the interaction of both T and B lymphocytes with macrophage-associated antigens derived from studies in thymectomized mice (Unanue, 1970) which did not respond to antigen bound to live macrophages. Also, the tissues of thymectomized mice were not impaired in their capacity to take up antigens ( Unanue, 1970). These observations indicated that the response to antigen in macrophages required the presence of both types of lymphocytes and that the immunoincompetence of thymectomized mice was not associated with an impaired mechanism for concentrating antigen. Macrophages from thymectomized mice were normal in uitro (Mosier et al., 1970). The third piece of evidence stems from the obser-

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

119

vation of Hamaoka et al., who showed that hapten-primed cells and carrier-primed cells responded to hapten carriers bound to spleen macrophages ( Hamaoka et al., 1971). Presumably the carrier-primed lymphocytes are thymic derived, and the hapten-primed cells correspond to the B type. Similar results have been obtained by Katz and Unanue (1971), using peritoneal macrophages. The macrophage transfer system has been used to evaluate the participation of macrophages in the genetically controlled immune response to hemocyanin among various strains of mice (Cerottini and Unanue, 1971). Macrophages from high and low responder strains handled antigen in the same way. Moreover, low responders were not restored to a high response by transferring macrophage-bound antigens from high responder strains. Clearly, the macrophage was not responsible for the differing responses among species in agreement with well-known observations in other systems ( McDevitt and Benacerraf, 1969). The macrophage transfer system was used to study the effects of X-irradiation on macrophages. Gallily and Feldman (1967) made the interesting observation that X-irradiated mice could mount an immune response to Shigella only if the organism was carried by macrophages and not free; such a response to Shigella bound to macrophages did not take place if the macrophages came from X-irradiated donors. Their conclusion was that part of the immunological defect of the X-irradiated mice could be attributed to defective antigen handling by macrophages. This study is difficult to interpret since there is at present no information on how Shigellu immunogens are handled by macrophages. Pribnow and Silverman ( 1967) had some indications that macrophages from X-irradiated rabbits after uptake of bovine y-globulin did not induce immune responses. Mitchison (1969b) also observed the effects of X-irradiation on the transfer of human albumin bound to murine macrophages. He subsequently showed with Kolsch that there were some differences in the intracellular distribution of antigens between macrophages from heavily X-irradiated mice and nonirradiated mice ( Kolsch and Mitchison, 1968)) although it was not clear whether these differences in handling correlated strictly with the impaired capacity to induce human albumin immunogenicity. Such effects of X-irradiation have not been observed with other antigens. Responses to hemocyanins (Unanue and Askonas, 1968b; Unanue and Cerottini, 1971) and SRBC bound to macrophages are not impaired by X-irradiation. Recently, Schmidtke and Dixon (1972) studied the handling of hemocyanin by X-irradiated macrophages and found that more antigen was retained on surface membranes of X-irradiated cells associated with increased immunogenicity. There was no effect by X-irradiation on the function of macrophages in in vitro (Roseman et aZ., 1969). In summary, there is no

120

E. R. UNANUE

clear explanation for the impaired responses observed to only some antigens associated with macrophages from X-irradiated animals. Detailed studies on antigen handling by these cells should eventually explain the discrepancies.

7. Attempts to Sensitize Lymphocytes in Vitro Ford, Gowans, and McCullagh attempted to sensitize lymphocytes by incubation on macrophage monolayers before adoptive transfer ( Ford et al., 1966; McCullagh, 1970). Rat thoracic duct lymphocytes were isolated and incubated with either solubilized SRBC or with monolayers of macrophages that had taken up SRBC. After 3 hours the lymphocytes were transplanted into syngeneic recipients. (The lymphocytes were separated from macrophages using a technique whereby iron particles were added to the cells which were then placed in a magnetic fieldmacrophages that took up the iron adhered, whereas the lymphocytes did not.) It was found that lymphocytes first incubated on macrophages made an anti-SRBC response, but those incubated in solubilized SRBC did not. This interesting experimental model, if developed further, could be used to study interactions of macrophage-bound antigen with lymphocytes. B. I n Vitro EXPERIMENTS In 1967, Dutton reviewed the early literature concerning in vitro systems and included the work associated with macrophage extracts (Dutton, 1967, p. 295). The two in vitro systems most frequently used in recent years include the antibody-forming cell [plaque-forming cells ( PFC)] response to foreign red cells (notably SRBC) or the proliferating cell response to soluble antigens or allogeneic cells. In this review we attempt to summarize the published results so far as they concern the role of the macrophage.

1. Response in the MishelGDutton

M

Marbrook System

An antibody-forming cell response can be readily obtained as outlined by Mishell and Dutton (1966) and Marbrook (1967). In both methods a suspension of spleen cells is used, most often from mice. In the Mishell-Dutton system the cells (usually 5-15 X loe in 1 nil of media) are cultured in dishes that are rocked for several days; the medium which contains selected fetal calf sera is supplemented daily with culture nutrients. In the Marbrook method the cells ( 4107-1.6 x lo7) are placed in a dialysis bag that is immersed in a flask with tissue culture medium; this method is stationary and does not require daily supplementation of the medium. In both methods the immune response

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

121

is measured by numbers of PFC present 4 or 5 days after the culture is initiated. There is general agreement that the primary anti-RBC response requires the presence of at least three types of interacting cells. Two of the cells are lymphoid and correspond to the T and B lymphocytes identified in vivo; the remaining cell has all the characteristics of a macrophage and can best be identified by its property of adherence to glass surfaces. The evidence that this third type of cell (i,e., adherent cell) is a macrophage is based on the following: ( 1 ) the cell adheres to glass or plastic surfaces; ( 2 ) it has phagocytic properties; ( 3 ) it has classic morphological criteria of macrophages; ( 4) it is radioresistant; (5) it does not synthesize antibodies; (6) it is affected by treatments that block the RES in uivo; and, finally ( 7 ) it is affected by specific AMS (Shortman and Palmer, 1971). In general, other in vitro methods employed for separating cells identify the activity of the adherent cells with macrophages. In conclusion, it is rather unlikely that apart from macrophages there is another cell having similar properties of adherence, sedimentation, and reactivity with AMS. Therefore, I shall use the term adherent cell synonymously with macrophage. The first demonstration that an adherent cell population was required in this system came from the studies of Mosier (1967). He cultured murine spleen cells on plastic dishes for 30 minutes and separated those that attached to the dish from those that did not attach. Neither cell type (i.e., nonadherent comprising about 90%of the original population and adherent comprising about 10%) by itself would respond in vitro to SRBC. However, when both were mixed together the anti-SRBC response was readily reconstituted. In other experiments the adherent cells were shown to interact with SRBC. The SRBC were added to the adherent cells for 0.5 hour after which most of the SRBC were removed. The adherent cells, presumably containing 5%phagocytized SRBC, were capable of interacting with the nonadherent lymphocytes. However, no exact estimate of the ratio of phagocytized to nonphagocytized SRBC was made, and it is most likely that some nonphagocytized SRBC must have remained along with the adherent cells. Subsequently, these early experiments were confirmed and expanded (Mosier, 1969; Pierce, 1969a,b; Pierce and Benacerraf, 1969; Talmage et d.,1969; Dutton et al., 1970; Hartman et al., 1970; Haskill et al., 1970; Shortman et al., 1970; Shortman and Palmer, 1971). Thus, it was later shown conclusively that the nonadherent cells comprised the population of progenitors of PFC (Hartman et d.,1970; Shortman and Palmer, 1971). This was best done by mixing adherent and nonadherent spleen cells, each from a different mouse strain. The source of resulting PFC was

122

E. R. UNANUE

easily determined by killing one or the other cell with appropriate antibodies to strain histocompatibility antigens prior to testing for PFC. Indeed, all PFC originated from the mouse strain that donated the nonadherent cells. Adherent cells obtained 1 4 8 hours after in vivo X-irradiation of mice (with 250 to lOOOR) were capable of performing in culture as effectively as adherent cells from nonirradiated mice (Roseman, 1969; Roseman et al., 1969; Dutton et al., 1970; Shortman et al., 1970). However, 3 days after X-irradiation the mouse spleens were depleted of adherent cells (Gorcyznski et al., 1971). As expected, the nonadherent population from X-irradiated mice were unable to respond. The degree to which the adherent cells influence the response to SRBC may depend upon the number of progenitors of PFC and/or the cell density in culture. Thus, spleen cells from immune mice appeared, in the hands of some investigators, to be less sensitive to depletion of the adherent cell population than were cells from nonimmune mice (Pierce, 1969a). The influence of cell density in rabbits was observed by Theis and Thorbecke (1970); nonadherent cells were more responsive when cultured in stationary tubes where the total surface area was small than in dishes where the total area was large. Although most investigators find that the macrophage is required for the in vitro response to SRBC, there is no general agreement on the mode of action or function of this cell. Is the macrophage favoring a response by handling the antigen in an appropriate way? Is the macrophage simply favoring cell-to-cell interactions regardless of its handling of antigen? Or, is the macrophage providing a source of nutrients or special factors required for the in vitro growth of lymphocytes? These three possible mechanisms are not mutually exclusive, and evidence in favor of one or two has been provided. Unfortunately, in vitro responses are not uniform, and results vary among different investigators. Each investigator has found subtle and sometimes ill-defined requirements for support of antibody formation in culture. We should be aware that in vitro systems introduce experimental variables which may apply only to the in vitro method and not to events occurring in uivo. The fate of the SRBC (and its antigen) among the adherent cell population has not been fully studied. It is quite important to determine the form in which the SRBC immunogen is presented to the lymphocyte. Does such presentation occur before or after endocytosis of SRBC? Is modification or change of the SRBC by intra- or extracellular macrophage products a necessary step? Mosier’s original method gave one the impression that most SRBC were phagocytized by the macrophages. Hence, immune recognition must have followed the in-

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

123

teriorization process. However, recently Leserman and Roseman ( 1971) have indicated that ammonium chloride treatment of adherent cells exposed to SRBC abolishes their capacity to support in vitro responses of lymphocytes. The ammonium chloride presumably is lysing extracellular SRBC. The treatment was not affecting macrophage function, since addition of SRBC restored their response. This last piece of evidence raises the possibility that the in vitro response may be directed to SRBC antigens not engulfed by macrophages but remaining extracellularly. Shortman and Palmer ( 1971) have recently presented some evidence for a soluble SRBC antigen resulting from interaction of macrophage culture fluid and SRBC. They incubated SRBC in the fluid from 24-hour culture of macrophages; after several hours the SRBC were centrifuged and discarded. The resulting supernatant stimulated nonadherent cells for an anti-SRBC response. (Sheep red blood cells by themselves were not stimulatory. ) The conclusion from this experiment is that macrophages may generate some factors which “release” or “liberate” antigen from SRBC, and these are the antigenic moieties capable of triggering the response of the lymphocyte. Shortman and Palmer have cautioned against overinterpreting their data in that their supernatant-antigen fluid was about twenty-five-fold less immunogenic than SRBC associated with the adherent cell layer. In favor of the macrophage serving as a cell that facilitates cellular interactions are the results of Mosier (1969) and Pierce and Benacerraf (1969). They found that after 24 or 48 hours of culture in the MishellDutton system, clusters of cells appear, some of which contain PFC. A macrophage was required for the cluster in that nonadherent cells formed few or small aggregates of cells (Pierce and Benacerraf, 1969). All the PFC in a cluster ( a mean in some cultures of about 6 per cluster) responded to a given foreign red cell, either sheep or burro, if both were added to the culture (Mosier, 1969). That these clusters may have a functional significance was gathered from results in which the cultures were interrupted at various times after their initiation and continued by keeping stationary instead of by rocking, i.e., the culture was initiated and interrupted at 6, 24, or 48 hours; the cells were then dispersed and recultured either by rocking or remaining stationary (Pierce and Benacerraf, 1969). Those cultures that were rocked developed clusters again, whereas those left stationary did not. The cultures left stationary after an interruption at 6 hours were found to have few PFC at day 4 or 5 as compared to similar cultures rocked after interruption at 6 hours. Interruption at 24 hours produced a lesser effect in that the responses of stationary cultures were one-third or one-half that of rocking cultures. At 48 hours there was no effect. On the basis

124

E. R. UN’ANUE

of these studies, Pierce and Benacerraf concluded that after a period of about 1 day the lymphocytes were already “activated upon interaction with macrophage (and its antigen) and, hence, did not require close cell contacts and aggregates. It should be noted that not all culture systems require movement. The Marbrook system for studying SRBC response is a stationary one, although also sensitive to the depletion of adherent cells. Evidence for macrophages supplying factors that improve the growth of lymphocytes comes from the experimental results of Dutton and collaborators (Dutton et al., 1970; Hoffman and Dutton, 1971). They have found that culture fluids of adherent spleen cells or, better, of peritoneal macrophages will allow nonadherent spleen cells to respond to antigen, The chemical property of the factor in question has not been solved (it is stable to heating at 57°C for 30 minutes but sensitive to freezing and thawing). Hoffman and Dutton postulated that the factor could be an enzyme necessary for SRBC processing, an antibody-like molecule or a growth-conditioned media-type factor. The observation that SRBC treated in vitro with the culture fluid can stimulate nonadherent cells for antibody production favors an “antibody-like molecule” theory. Recent findings of Shortman and Palmer (1971) are important in establishing a coherent picture among results in the literature. They have shown quite clearly that a given culture of nonadherent lymphocytes that does not respond to SRBC will respond to a different antigen -the polymerized form of flagellin from Salmonella adelaide (Table I11 shows one of their results). In some experiments both antigens were added to the same culture, yet the cells only responded to flagellin. This experiment tends to minimize the role of macrophages in vitro as cells simply providing a conditioned kind of media and implies a more active role of these cells in the process of antigen handling and presentation. It also emphasizes that the macrophage pathway may vary in importance depending upon the antigen in question. Along these lines, M. Feldman and Palmer (1971) claimed that macrophages were not required for the in oitro immune response to a soluble extract of SRBC. Their results imply that SRBC must be degraded to a small size in order to trigger lymphocytes for antibody synthesis. This important study needs confirmation. Gorczynski et al. (1971) have correlated an adherent cell in oitm with spleen macrophages operational in uiuo. They found that spleen cell suspensions devoid of adherent cells 3 days after X-irradiation of mice did not support in vitro responses by lymphocytes; at the same time suspensions of lymphocytes ( depleted of macrophages ) transplanted

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

125

TABLE I11 EFFECTOF ANTIMACROPHAGE ANTISERUMON IMMUNE RESPONSES IN TISSUE CULTURW~ Antibody-forming cells/culture Tissue cultures Unfractionated spleen cells Untreated Normal serum Antimacrophage serum Adherence column filtrate Untreated Normal serum Antimacrophage serum

+ +

+ +

SRBC antigen

POL antigen

1660 (1020-2280) 893 (313-1500) 36 (0-150)

1090 (568-1350) 696 (278-1020) 446 (242-846)

92 (30-205) 6 (0-13) 0 (0-0)

1330 (468-2240) 405 (164-648) 298 (177-363)

a From Shortman and Palmer (1971), p. 405. I thank the authors for allowing to reproduce the table. * This table shows the effect of macrophage depletion on the response to sheep red blood cells (SRBC) or polymerized flagellin (POL). Purified lymphocytes (by column filtration or as a result of antimacrophage serum treatment) respond to POL but not to SRBC. One milliliter of undiluted antiserum was added to each 8 ml. of tissue culture medium. The cells (12 X 106 viable nucleated cells in 1 ml. culture) were preincubated with this mixture for 90 minutes at 37"C, prior to initiation of the response by antigen addition. Results are the mean of three separate experiments, each involving three to four cultures per point. The range of individual experiments is given in parenthesis. Background values of untreated unfractionated spleen cells in the absence of added antigen averaged 65 antibody-forming cells per culture for POL, 99 per culture for SRBC.

into mice irradiated 3 days previously did not form antibody in vivo. A similar correlation was found when mice were injected with large doses of horse red cells; such mice did not make antibodies when immunized to SRBC in vivo (unless replenished with macrophages) nor did the spleens contain active adherent cells for in vitro responses (to SRBC). Furthermore, cells isolated from normal spleens were capable of restoring both in vivo and in vitro responses in both sets of experiments. These cells, presumably macrophages, had identical physical properties when separated in density gradients or velocity sedimentation. This last experiment strongly suggests that the role of macrophages in vitro may be akin to their role in vivo. 2. Proliferative Response of Immune Cells Lymphoid cells from immunized experimental animals or humans proliferate in culture when exposed to antigens for which they are specific. This specific proliferative response could be determined by either pulsing the cells with a radiolabeled deoxyribonucleic acid (DNA)

126

E. €3. UNANUE

precursor ( such as tritiated thymidine) hours before harvesting or estimating the number of large blasts in cell smears. In humans the lymphoid cells were obtained from peripheral blood and the antigen most frequently used was purified protein derivative ( PPD) . In experimental animals (such as rabbits, guinea pigs, or mice) cells were obtained from the spleen, and the antigens included a variety of soluble proteins such as albumins, Ig, tetanus, and diphtheria toxoid. Cells from nonimmune individuals responded poorly, if at all, in this in vitro system. The culture consisted of cells placed in stationary tubes with antigen and was examined 3 to 5 days later. Several points seem to have been reasonably well established. For example, the presence of adherent cells or purified macrophages appears to be important in order for a purified population of lymphocytes to respond to antigen (Hersch and Harris, 1968; Gordon, 1968; Oppenheim et al., 1966; Lamvik, 1969; Schechter and McFarland, 1970). This effect has been tested most often by separating the macrophages through their ability to adhere to glass. Customarily, human lymphocytes have been separated by passing them through a column of glass beads with appropriate media. These cultures of purified lymphocytes which responded poorly to antigen were restored by addition of purified macrophages. Other cell types, such as fibroblasts (Hersch and Harris, 1968) or neutrophiles (Cline and Sweet, 1968), could not be substituted for macrophages. The extent of macrophage contribution to responses depended in part upon total cell density (Oppenheim et al., 1968). In ordinary cultures purified lymphocytes responded poorly but, if the cells were concentrated 3 to 4 times, there was a partial restoration of the proliferative response. This effect could also be seen to a lesser extent on cultures of lymphocytes stimulated by phytohemagglutinin ( PHA), where poor stimulation was noted at high dilutions of the mitogen unless the cell concentration was increased. This would imply that macrophages functioned as a means for concentrating lymphocytes with antigen or lymphocytes with lymphocytes. To this effect, many reports have emphasized and illustrated the presence of groups of lymphocytes, some of which are blasts around macrophages (Cline and Sweet, 1968; Schechter and McFarland, 1970). Depletion of macrophages from a suspension of spleen cells by AMS reduced the proliferative response to antigen but not to PHA (Jehn et al., 1970). These last results agree with the previous series of experimental results. Part of the effect of macrophages in these culture systems which utilize soluble antigens could be explained by the uptake of antigen. The evidence was reasonably convincing that the soluble antigen taken up by

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

127

the macrophages was capable of stimulating the DNA synthetic response. This was done by obtaining macrophages (often with a great degree of purity) from spleen (Harris, 1965), peripheral blood ( Oppenheim et uZ., 1966; Hersch and Harris, 1968; Schechter and McFarland, 1970; Cline and Sweet, 1968), peritoneal cavity (Seeger and Oppenheim, 1970b), or lung alveolar spaces (Seeger and Oppenheim, 1970b) and exposing them to antigen for various periods of time, The cells were washed well, mixed with the lymphocyte suspension, and then cultured for several days. [In one experiment the cells were separated from the glass and mixed with lymphocytes. This procedure insured that antigen adhering to glass was not operative in the system (Schechter and McFarland, 1970).] The lymphocytes, which in the usual culture conditions responded poorly to soluble antigen, now proliferated upon interaction with the macrophage-bound antigen. The response required close cell contact and was not associated with a soluble antigen product released from the macrophage. Thus, separation of the macrophages from the lymphocytes by a membrane (of 0.45 p pore size) impaired the response (Hersch and Harris, 1968; Cline and Sweet, 1968). Also, culture fluid that harbored macrophages (with antigen) was not stimulatory (Cline and Sweet, 1968). Harris claimed that the antigen responsible for this response was on the membrane of the macrophage (1965). Macrophages containing antigen, if treated with the specific antibody, were inhibited in triggering the response of the lymphocyte. Live macrophages were also important here in that after antigen uptake, if killed by exposure to heat, they would not trigger lymphocyte proliferation (Cline and Sweet, 1968). Macrophages from immune individuals were as effective as those from normal unimmunized subjects (Cline and Sweet, 1968; Seeger and Oppenheim, 1970b). Unfortunately, no complete quantitative studies comparing the relative effectiveness of macrophage-bound antigen and soluble antigen are available. Harris had some indication of the uptake of radiolabeled antigen by the macrophage and claimed that macrophagebound antigen must be more immunogenic than free ( 1965); others have made similar claims (Cline and Sweet, 1968). Two further points are of importance and these concern the ratio of macrophage to lymphocyte in culture and the effects of X-irradiation. The ratio of macrophages to lymphocytes was critical in that high ratios depressed the response (as also observed in the previous section) (Harris, 1965; Parkhouse and Dutton, 1966). The best ratio at which to obtain a response was about 1 monocyte/1050 lymphocytes. In all reports, macrophages exposed to antigen and irradiated were not impaired and responded as well as nonirradiated macrophages. Whether the macrophage also contributed in producing a conditioned

128

E. R. UNANUE

media is under consideration. Bach and collaborators (1970) have shown that fluid from macrophage cultures can be substituted for these cells and allow lymphocytes to respond to allogeneic cells or soluble antigens. Hence the problem of distinguishing among several effects is still present in this system, and again technical differences probably account for nonuniformity of results. 111. Handling of Antigen by Macrophages

The handling of antigen by macrophages will be reviewed here with emphasis on identifying the immunogenic moiety of the macrophage bound antigens. Cohn (1968) described the physiology of macrophages in a previous volume of this series and presented details on the process of endocytosis. Also, excellent reviews on phagocytosis ( Rabinovitch, 1968) and on lysosomes (Weissman and Dukor, 1970) have appeared. Uptake of soluble and particulate antigens by macrophages consists of a two-stage process. The first is attachment of antigen to the surface of the cell, and the second is endocytosis of the bound antigen. Subsequently, some of the pinocytic or phagocytic vesicles with the endocytized antigen fuse with lysosomes; the lysosomal enzymes then degrade most of the antigen. We consider two main points: (1) the capacity of a macrophage to take up antigen and to discriminate between foreign and nonforeign elements and (2) the possible locus of the immunogenic moiety among the total antigen taken up by this cell.

A. UPTAKE OF ANTIGEN In trying to understand the extent to which the macrophage plays a role in the immunological process, it is necessary to evaluate the capacity of these cells to discriminate among a variety of antigenic or nonantigenic materials. To do so one needs information on types and amounts of antigen taken up, on whether the macrophage has the capacity to discriminate among antigens or between foreign and nonforeign substances, on whether there are different populations of macrophages differing in their capacity to take up antigen, and finally on the biochemical nature of macrophage-antigen interaction. Both in vivo and in vitro procedures have been employed to study uptake of antigen by macrophages. In vivo procedures include (1) the study of blood clearance of suitably labeled antigens and ( 2 ) radioautographic studies of the fate of antigen. (The techniques and main results of this last procedure were described in Section II,A,6.) The in vitro methods consist of studying the uptake by macrophages of a variety of suitably tagged antigens under a series of tissue culture conditions. It is obvious that the in vitro methods are the methods of choice because

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

129

they give direct quantitative results amenable to experimental variation without interference with i n vivo variables. Macrophages are cells that are relatively easy to culture for hours or days without fastidious cultural requirements (Jacoby, 1965). Knowledge of the mechanisms of antigen uptake and of the physiology of macrophages has been obtained by studying the rate of elimination of suitably tagged antigens from the circulation. This method gives an indication of the activity of the RES mainly of liver and spleen. Colloidal suspensions, such as carbon, saccharated iron oxide ( Benacerraf et al., 1954), chromic phosphate (Dobson and Jones, 1951; Dobson, 1957), or gold (Zilversmit et al., 1952), as well as particulate antigens such as heat-aggregated proteins (Halpern et al., 1957) have been employed. The ideal material must not be toxic, should be of a homogeneous size, and should be easy to trace and measure in the blood. The kinetics of clearance from the circulation of colloidal suspensions were studied extensively by Benacerraf, Biozzi, and collaborators ( reviewed in Biozzi et al., 1957; Benacerraf et al., 1957a; Benacerraf, 1964). They found that the liver and spleen took up most of the injected material (8040%) of which two-thirds was taken by the liver. Indeed there was a direct relationship between size of these organs and the rate of blood clearance. The distribution of material between liver and spleen varied depending upon the amount injected. The liver RES was probably the most active because of its large number of RES cells and great blood flow (Benacerraf et al., 1955a,b, 1957b; Biozzi et al., 1958). In one passage through the liver about 80 to 90%of material could be cleared [in liver cirrhosis the amount cleared was greatly decreased (Halpern et al., 1959)l. At suitable concentrations the particles or particulate antigens were eliminated exponentially from the circulation. A suitable mathematical formula has been applied to estimate the phagocytic activity (i.e., phagocytic index). As the concentration of particles increased, the index decreased-an indication of saturation of the capacity of the RES to take up particles. The clearance of bacteria (Benacerraf et d.,1959; Wardlaw and Howard, 1959; Biozzi et al., 19sO), viruses ( Mims, 1959), and foreign red cells was also studied. Among other tests, the rate of blood clearance indicated that uptake of particles was ( 1 ) a radioresistant process, ( 2 ) increased by estrogen treatment, ( 3 ) decreased by cortisone treatment, (4)increased after active infection with tubercule bacillus (Biozzi et al., 1954) or after injection with mycobacterial products or gram-negative endotoxin. All these effects can be easily explained. Macrophages are radioresistant and are easily activated during delayed hypersensitivity reactions. Macrophages take up a variety of colloidal substances, proteins,

130

E. R. UNANUE

microorganisms, and red cells; although many of these need to be opsonized by antibody and C first. The molecular basis of the uptake is poorly understood. The uptake of an antigen, either particulate or soluble, by macrophages depends upon two main considerations: (1) whether the antigen can interact with the macrophage membrane “directly,” i.e., without serum antibody or C, and ( 2 ) whether the antigen is complexed to antibody and/or C (or other serum proteins). Not all antigens interact well with the macrophage surface. Some bacteria, many of which are virulent, do not bind with macrophage and, hence, are not phagocytized. These include pneumococci (Wood, 1951), some strains of Salmonella (C. Jenkin and Benacerraf, 1960), and staphylococci ( MacKaness, 1960; see, also, Allen and Cook, 1970; Brumfitt et al., 1965). These bacteria usually have a thick capsule of carbohydrate. It would appear that interactions between capsular polysaccharides and macrophages are weak or poor resulting in poor attachment. However, Wood‘s classic experiment on phagocytosis of pneumococci indicated that, under suitable conditions, macrophages were able to take up some encapsulated organisms in the absence of antibody (Wood et al., 1946), particularly in the presence of inert material. The macrophages were thought to “corner” the organisms against inert substances and in this way take up some organisms. Bacteria that were poorly taken up by macrophages were easily phagocytized if complexed to specific antibody and/or to serum C. This phenomenon is described in detail below. Macrophages can take up a variety of foreign erythrocytes without the aid of antibody. Moreover, Perkins and Leonard have indicated that there was preferential uptake of some species of erythrocytes ( 1963) . For example, mouse macrophages preferred chicken red cells over sheep or rabbit red cells. In their experiments, in which uptake of nonopsonized erythrocytes was small, about 15 to 20%of macrophages would take up both types of red cells. Macrophages did not take up autologous red cells unless these became effete (Vaughan and Boyden, 1964; Lee and Cooper, 1966). Trypsin treatment of the macrophages removed the receptor for effete red cells (Vaughan, 1965); those chemically modified by glutaraldehyde readily bound to macrophages ( Rabinovitch, 1967a,b) , Most protein antigens that have been examined can be taken up by macrophages directly. The degree of uptake in part depends upon the state of aggregation of the protein. Evidence that some protein antigens can interact “directly” with macrophages at least without the intervention of antibody is based on both in vivo and in vitro observations. Some antigens injected into animals could be taken up by macrophages in the absence of serum or cell-bound antibodies. Injection of hemocyanin, aggregated albumins, or Ig into an unimmunized animal resulted in their

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

131

rapid uptake by the perifollicular macrophages of lymph nodes and spleen (see Section 11,AJ). Such large or highly aggregated proteins, if they reached the circulation, were cleared very rapidly by macrophages of the liver and spleen. However, in these in vim experiments the possible presence of small, undetectable amounts of natural serum antibodies could not be completely ruled out. The experiments of Humphrey and Frank (1967) described earlier, using hemocyanin or albumin in rabbits, indicated strongly that natural or acquired antibodies were not playing a role in the uptake of antigen by lymph node macrophages. (The rabbits were immunologically tolerant to the antigens and the presence of antibodies was ruled out by the absence of follicular trapping of the antigen. ) By tissue culture experimentation this problem has been evaluated directly. In some of our own experiments macrophages exposed in vitro to protein antigens, in the absence of serum, took up the antigen efficiently (Unanue et aE., 1972; Schmidtke and Unanue, 1971a). It seemed unlikely that small amounts of membrane-bound antibody were involved. This possibility was then eliminated by incubating macrophages with anti-Ig antibody before contact with the antigen (Unanue et al., 1972; Schmidtke and Unanue, 1971a). Murine macrophages were washed well, isolated, and planted in culture dishes; different antibodies to mouse Fab and class determinants were added to the dish; at a later stage radioactive heniocyanin or aggregated albumin was added. Treated and untreated macrophages took up equal amounts of the radiolabeled antigen. If antibody was on the membrane of the macrophage, then addition of antibody to Ig should hinder its combination with antigen. Similar antibodies, for example, could block the uptake of the same antigens by B-type lymphocytes which are known to have membrane-bound Ig (Unanue et al., 1972; Warner and Byrt, 1970) (Figs. 4 and 5 are representative experiments). However, if such macrophages were treated with AMS, which presumably covers part of the surface, then uptake of the antigen was greatly impaired. In accordance with these results, treatment of macrophages with trypsin (which removes any surface-bound Ig) did not impair their capacity to take up labeled antigens. Hence, a variety of antigens could, indeed, be taken up in the absence of extracellular or cellular-bound specific antibody. Aggregation or polymerization of many protein antigens favors interaction with the macrophage surface. This has been studied best with albumins or Ig. For example, the rate of elimination from the circulation of these radiolabeled proteins in their polymeric or monomeric forms is strikingly different. Polymers are cleared in minutes by the RES of liver and spleen, whereas monomers circulate with a half-life similar to that of comparable autologous proteins (i.e., 2-3 days).

132

E. R. UNANUE

16 Inhibition of fixation of KLH 10 20 30 40 50 60 70 80 80 100

Anti Mouse V2A Globulin

cI

Anti Mours Globulin

Anti Mours Globulin labs F nucrophagss)

FIG.4. Results of an experiment in which macrophages were incubated with a series of antisera prior to exposure to radioiodinated keyhole limpet hemocyanin (KLH). The uptake of KLH was estimated by counting the cells. Note that incubation of the macrophages with antimouse Ig (i.e,, an antimouse IgGs. or a polyvalent antimouse Ig) had little effect on the binding of KLH (the small effect produced by antimouse globulin was abolished by absorption of the antisera cells-see last item in figure ). As expected, antimacrophage sera ( AMS ) blocked the effect-the absorption of AMS by macrophages reduced its binding to macrophages and its capacity to block KLH uptake. Antilymphocyte sera which cross-react with macrophages also block the uptake of KLH. Compare these results with the effects of anti Ig in uptake of K L H by lymphocytes in Fig. 5. (From Unanue et al., 1972).

The handling of polymerized and monomer albumins was also studied in tissue culture experiments (Schmidtke and Unanue, 1971a). It was found that monomer human albumin was taken up to a small extent by murine macrophages and that it readily dissociated from the cell surface. In contrast, polymer human albumin was taken up to a much larger extent and, after uptake, a great part of the molecule was retained on the membrane ( and subsequently endocytized ) . Apparently macrophages take up autologous albumin (Schmidtke and Unanue, 1971a) and Ig (Herd and Ada, 1969a) to the same extent as their foreign counterparts, the main consideration being, as with the foreign material, the degree of polymerization. All these results bring out two important considerations: first they suggest that any polymeric antigen with multiple points of attachment to a membrane forms, as expected, a stronger interaction with the cell than that of a monomeric antigen with few points of contact; second, they indicate the poor capacity of the macrophage to discriminate foreign vs. nonforeign substances. There is little information on the nature and specificity of the inter-

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

X Reduction of Lymphocytes Binding

10

I

20

30

40

SO

60

70

80

133

KLH

90

100

Anti I g A

FIG. 5. Results of an experiment in which lymphocytes were incubated with keyhole limpet hemocyanin (KLH)-””T, and the number binding KLH was determined by radioautography. Prior incubation of lymphocytes with anti-Ig antibodies reduced the number of cells binding KLH-an indication that class-specific Ig was mediating the uptake. Compare with results of experiments of Fig. 4. (From Unanue et al., 1972).

action between different antigens and the macrophage membrane, We do know that protein antigens bind to areas of the membrane surface distinct from those to which Ig and C bind. However, the interaction or possible competition of different surface areas of the cell with the various proteins has not been studied critically as yet. Are there populations of macrophages differing in their capacity to discriminate among various antigens? All indications are that such populations do not exist and that any macrophage under optimal conditions can take up any antigen. There can be, however, a difference in the degree of uptake of antigen between monocytes and macrophages. About two-thirds or more of the peritoneal macrophages take up protein antigens. A single macrophage can take up two different protein antigens (Rhodes et d.,1969). Perkins and Leonard (1963), however, had some

134

E. R. UNANUE

indication that selected macrophages preferred to take up foreign red cells, but whether this could be explained by the state of cell maturation was not ascertained. Macrophages from tolerant animals take up the tolerogen as efficiently as macrophages from normal, nontolerant animals ( Mitchison, 1969b; Humphrey and Frank, 1967). 1. Immunoglobulin Receptor on Macrophage Antigen complexed to antibody binds avidly to macrophages as a result of the interaction of the antibody molecule with a specific surface receptor on the macrophage membrane. This antibody has been termed cytophilic ( Boyden, 1963). It is well known that aggregated antigens, red cells, or bacteria in the presence of antibody were immediately cleared from the circulation by the RES of liver and spleen. In 19631964, Boyden developed an in vitro assay to study binding of antigenantibody complexes (1963, 1964). He employed red blood cells which in the presence of antibody adhered around the macrophage-the antigen (red cell)-antibody complex was easily visualized as a rosette around a central macrophage. Bacteria coated with antibody were also used in this type of test ( Uhr, 1965; Auzins and Rowley, 1963). The in vitro assay has two varieties: in one, a monolayer of macrophages is exposed to antisera, usually at room temperature, then washed and exposed to the antigen; a second consists of adding antigen-antibody complexes directly to the macrophage monolayer (Jonas et al., 1965; Berken and Benacerraf, 1966). (The complexes could be formed by immunological reaction of antigen with specific antibody or by chemically coupling the antigen to any Ig.) As one would expect, the latter assay was more sensitive. Ultrastructural studies on rosettes of red cells on macrophages showed adherence and many interdigitations of the two-cell surfaces ( Abramson et al., 1970a). It is interesting to note that the red cells after adhering to macrophages became spherical ( LoBuglio et al., 1967) for no apparent reason. It should be made clear that the attachment of opsonized particles (i.e., particles coated with antibody) to the membrane without their interiorization is a result of working at room temperature. Usually increasing the temperature to 37°C results in rapid interiorization of the antigen ( Berken and Benacerraf, 1966). Only some classes of Ig can interact with the macrophage surface. This can be studied with the rosette test by either isolating specific antired-cell antibodies of a particular Ig class or chemically coupling a purified Ig class, usually myeloma proteins, to red cells. A different approach is to perform inhibition tests with a purified class of Ig by reacting macrophages with soluble Ig before the reaction with opsonized red cells

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

135

occurs, In guinea pigs, where it was first shown, y z antibodies to red cells formed rosettes on the macrophages, but yl antibodies of a comparable antibody titer did not (Berken and Benacerraf, 1966). In humans, IgG antibodies are cytophilic (Huber and Fudenberg, 1968; Abramson et al., 1970b; Inchley et al., 1970). Among the IgG subclasses, only IgG, and IgG, bind to macrophages (Huber and Fudenberg, 1968; Abramson et al., 1970b). In mice, IgG and IgM are cytophilic (Lay and Nussenzweig, 1969). Cytophilic Ig’s are specific in that their attachment is blocked only by homologous Ig (Inchley et al., 1970; Huber and Fudenberg, 1968; Huber et al., 1969; Abramson et al., 1970b). Note that IgM is not cytophilic in all species: mouse IgM binds to macrophages, but human IgM does not. As expected, cytophilic Ig binds to the macrophage through the Fc part of the molecule. Red cells coated with Fab antibodies do not form rosettes on macrophages (Berken and Benacerraf, 1966); purified Fc but not Fab inhibits the rosette reaction (Inchley et al., 1970; Abramson et al., 1970b). A word of caution must be expressed concerning cytophilic tests in which macrophages and antibodies derive from different species. Indeed, not all Ig reacts equally with macrophages of every species. For example, Kossard and Nelson (1968a) found that guinea pig antibodies were taken up to a limited extent, if at all, by mouse and rabbit macrophages, although guinea pig macrophages did react well with mouse and rabbit antibodies. The nature of the cytophilic receptor on macrophages is not known. Treatment of the macrophages with trypsin does not affect the binding of red cells coated with IgG antibodies (Kossard and Nelson, 1968b; Howard and Benacerraf, 1966) but does affect the binding of red cells coated with IgM antibodies (Lay and Nussenzweig, 1969). Free sulfhydryl (SH) groups form part of the receptor site, since treatment of macrophages with reagents that react with free SH groups (such as iodoacetamide ) stops the attachment of opsonized red cells. The number of Ig receptors has not been definitely established. Claims have been made that an adult macrophage may have about 2 X lo6 sites. These claims are based on the amounts of radiolabeled Ig that binds to saturation on a macrophage. These figures may be an overestimate in that the Fab portion of the Ig niay interact with the macrophage on a site different from that where the Fc region acts. The mechanisms of binding of Ig in free form or as an immune complex to the macrophage membrane have been the subject of some controversy. It was speculated that when antibody bound to antigen it suffered some conformational change in its Fc region and that this change was responsible for attachment to the cell. The observed facts, however, do not support this argument. Soluble Ig devoid of aggregates binds to

136

E. R. UNANUE

the macrophage. This can be tested by showing the binding of radioiodinated Ig to macrophages and, best of all, by showing that Ig can block to some extent the binding of antigen-antibody complexes to the macrophage. It is apparent, however, that the interaction of monomeric Ig with the macrophage is weak and dissociates readily. However, antibody complexed to antigen or aggregated strongly interacts with the macrophage. Phillips-Quagliata and coworkers ( 1969, 1971) studied the interaction of macrophages with monovalent, bivalent, and polyvalent haptens bound to antibodies. Polyvalent and bivalent haptens coniplexed a t equivalence with antibody were avidly taken up by macrophages. However, the same haptens in antigen excess were unable to enhance this uptake. Moreover, uptake by macrophages of soluble antibody or antibody bound to a monovalent hapten was comparable. The authors as well as others hypothesized (and we agree) that the increased interaction of antibody with the macrophage, in the presence of antigen, is not the result of an allosteric change in the Ig molecule but probably results from additional binding sites from each Ig molecule forming the complex.

2. Complement Receptor on Macrophages Macrophages have a receptor on their membranes for C (Lay and Nussenzweig, 1968; Huber et al., 1968; Henson, 1969). Lay and Nussenzweig (1968) found that red cells coated with rabbit yM antibodies did not bind to mouse macrophages unless serum was added. This binding was dependent on divalent cations and could be reversed by addition of ethylenediaminetetraacetate ( EDTA) . The receptor on the macrophage was sensitive to treatment with trypsin. Huber et al. (1968) showed that the binding of antigen-antibody-C complexes to human macrophages was due to the presence of bound C3. Thus red cells that had reacted with antibody and C1, C14, or C142 did not attach to macrophages; however, red cells that had reacted with antibody and C1423 attached to the macrophages. Soluble Ig or serum did not interfere with the binding of antigen-antibody-Cl423 complexes to macrophages. As would be expected from these results, serum from animals deficient in C5 or C6 could be used as a source of C in this test (Lay and Nussenzweig, 1968). The role of macrophage-bound C3 may be of considerable importance in promoting phagocytosis. For example, complexes of hunian IgG with antigen could be easily blocked from attaching to macrophages by soluble Ig in vitro. Still, soluble Ig or serum did not block the binding of antigen-antibody-C complexes to macrophages. Spiegelberg et al. (1963) have shown the importance of C in the clearance of bacteria and red cells from the circulation. Decomplemented mice were no longer able to eliminate antibody-coated Escherichia coli and rat erythrocytes from

ROLE OF MACROPHAGES I" ANTIGENIC STIMULATION

137

the circulation rapidly, yet they quickly eliminated E . coli previously opsonized with antibody and fresh mouse serum (as a C source). The IgG and C3 receptors on the macrophages enjoy cooperative interaction (Huber et al., 1968). About 100 molecules of C3 lead to ingestion of IgG antibody-antigen complexes; however, tenfold more C3 molecules are needed to favor ingestion of complexes made by IgM (which by itself has no attachment site for human monocytes). The C3 receptor on macrophages is probably similar to the immune adherence receptor described previously ( Nelson, 1963). Granulocytes also possess a receptor for C3 (Shin et al., 1969; Gigli and Nelson, 1968). There is suggestive evidence that C5 may also be involved in the promotion of phagocytosis, at least by granulocytes. Phagocytosis of Candida albicans (Morelli and Rosenberg, 1971) and Pneumococcus (Shin et al., 1969) was somewhat depressed in serum from congenitally C5-deficient mice. M. E. Miller and Nillson (1970) have described a hereditary dysfunction of C5 which results in deficient phagocytosis of yeast by granulocytes. In summary, although we do not understand the basic molecular interactions involved in the uptake of a variety of antigenic and nonantigenic particles, nevertheless certain conclusions can be reached insofar as the process of antigen recognition by macrophages is concerned. Although it has become quite apparent that only few lymphocytes can interact with a single antigen (Naor and Sulitzeanu, 1967; Ada and Byrt, 1969; Humphrey and Keller, 1970), presumably these being a selected clonal population, specific immune recognition by macrophages is nonexistent (unless the antigen is itself coated with antibody and/or antibody and c).Lymphocytes bind to antigen as a result of antibody receptors synthesized by the cell; macrophages bind antigen directly or because the antigen is coated by antibody (made by a lymphocyte). Thus, macrophages take up a wide variety of substances regardless of whether these are immunogenic or noninimunogenic. The responsibility to recognize foreignness in an immune response lies in the lymphocyte population-if antigens do not reach lymphocytes which are specific for them, then there will be no recognition (Ada and Byrt, 1969; Humphrey and Keller, 1970) and no immune process.

B. THE IMMUNOGENIC MOIETY Most of the antigen taken up by macrophages is degraded and does not participate in immune induction. Of the total antigen taken up by the cell, only a small fraction escapes complete digestion and becomes available for lymphocyte recognition. Various experimental approaches have been used for identifying immunogenic antigen and its locus on the

138

E. R . UNANUE

macrophage. In some experiments the antigens have been followed by using live macrophages in a transfer system. In other experiments, cells have been disrupted and the fractions tested for immunogenicity. Obviously this approach offers no certainty that any extracted immunogenic material is functional in vivo. A combination of tissue culture methods plus the macrophage transfer system, although showing that a great part of the antigen is lost as a result of catabolism, has indicated two possible pathways by which antigen may be presented to lymphocytes for immune recognition: one is by retention of nonpinocytized antigen on the surface membrane, and the second is by release of intracellular antigen, some of which may reattach to the surface of macrophages. Most antigens that bind to the plasma membrane are interiorized in vesicles. Attachment, however, is not necessarily followed by endocytosis. Chemically treated red cells bind to the macrophage but are not endocytized unless specific antibody is added ( Rabinovitch, 1967a,b). A similar example was seen with M y c o p h m u pulmonis in tissue culture; the organisms attached to the macrophages but were not phagocytized until antibody was added to the culture (Jones and Hirsch, 1971). Presumably there was need for linkage of several close attachment sites for interiorization of the membrane. The antibody may be linking two red cells (or organisms) or linking one of them to the membrane site where the antibody Fc piece is attached. Endocytosis is a temperature- and energy-dependent process ( Cohn, 1968), and the metabolic requirements vary depending upon the size of the interiorized vesicle (Cohn, 1968). After endocytosis, the endocytic vesicle fuses with primary lysosomes. The antigen is then exposed to the hydrolytic enzymes of the lysosomes and degraded (Cohn, 1968; Rabinovitch, 1968; Weissman and Dukor, 1970). Most of the endocytized antigen is extensively catabolized. For example, in tissue culture, macrophages digested most pinocytized lZ5Ihuman albumin down to the level of amino acids (Ehrenreich and Cohn, 1967). Similar results were found with hemoglobin (Ehrenreich and Cohn, 1968). Also, many types of bacteria are well known to be extensively degraded by macrophages (Gill and Cole, 1965). However, synthetic polypeptides formed by D-amino acids or some carbohydrates, such as pneuniococcal polysaccharides, were taken up by macrophages but were not digested. Presumably the macrophages did not have the enzymes necessary to catabolize these antigens. The effects of lysosomal enzymes on different materials will not be reviewed since Weissman and Dukor have recently examined this topic ( 1970). Cohn in 1962 attempted to correlate immunogenicity with catabolism of antigen. He studied the degradation of :j2PP-labeledEscherichin coli by rabbit macrophages and polymorphonuclear leukocytes and found

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

139

that the rates of ingestion, killing, and degradation were comparable (Cohn, 1963a,b). However, the immunogenicities of E . coli in live peritoneal macrophages and polymorphonuclear leukocytes ( as assayed in a transfer system) were quite different. Although the immune response to E. coli ingested by macrophages persisted and was relatively unaltered for up to 2 hours, it was quickly destroyed by ingestion by polymorphonuclear leukocytes (Cohn, 1963b). The conclusion was that in some way immunogens from E. coli had resisted digestion by peritoneal macrophages and were available for immune recognition. [ It had previously been shown that injection of Salmonella typhi into serous cavities of rabbits with a macrophage-rich exudate or a polymorphonuclear leukocyte-rich exudate led to immune responses only in the former case ( Walsh and Smith, 1951).]

1. Protein Antigens on the Plasma Membrane Some molecules of antigen could be retained on the surface of the macrophage for long periods of time. In one of our experiments, mouse macrophages were exposed to 1Y61-labeledhemocyanin and, after uptake, were placed in suspension culture (Unanue and Askonas, 1968a). At various time intervals the amount of radioactivity associated with macrophages and in the culture fluid was determined; the live cells were then transferred to syngeneic hosts in order to study their immunogenicity. It was found that most of the protein radioactivity was readily lost from the macrophage and appeared in the fluid as protein-free iodine; however, after a few hours, a small residual amount of radioactivity persisted in the cells for up to 2 days of culture. Interestingly enough, the immune response to a constant number of macrophages in a sensitive system was comparable whether macrophages were transferred early or late after ingestion of the protein. (The immune response was comparable if macrophages were transferred early with 100%of the antigen initially taken up or late with about 10%of the antigen remaining on the cell.) Thus, the immune response must be directed to residual antigen (represented by, at the most, 10%of the total) and not to the 90%catabolized. Some of the persisting protein antigen has been identified on the plasma membrane. Indications are that this plasma membrane-bound antigen mostly comprises molecules of sizes comparable to native antigen that has eluded endocytosis. The evidence is based on the following findings : 1. By electron microscopy using 1’51-labeled antigen, it was possible to detect a few molecules of antigen on the surface of the macrophage, although most of the antigen was associated with lysosomes (Unanue et al., 1969b) (Fig. 6 ) . ( N o evidence was obtained for two populations

140

E. R. UNANUE

FIG.6. An electron photomicrograph of a macrophage several hours after uptake of Muiu squinudo hemocyanin (MSH)-’”I. Grains are on the membrane of the cell. (From Unanue et al., 196913.)

of functional macrophages; most macrophages had antigen on the plasma membrane as well as in intracellular vesicles. ) 2. Macrophages that had membrane-bound antigen were capable of reacting with specific antibodies or antibody fragments (Unanue et al., 1969b; Unanue and Cerottini, 1970). 3. Treatment with trypsin or EDTA of macrophages that had taken up antigen removed a small amount of antigen from the surface of the cell ( Unanue and Cerottini, 1970). 4. The antigen removed from the membrane was of a size comparable to the native molecule. 5. Removal of surface antigen on macrophages by treatment with trypsin or by covering with antibody abrogated most of the immune response to it (Unanue and Cerottini, 1970); see Table IV. Other investigators have confirmed these observations. With EDTA, Askonas and Jaroskova (1970) removed Maia squinudo hemocyanin that was retained on macrophages and thus abrogated a great part of their capacity to induce a response in the macrophage transfer system. Recently, Garcia found that macrophages exposed to diptheria toxoid retained a few molecules on their membranes; such macrophages when

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

141

TABLE IV EFFECTOF TREATING MACROPHAGES WITH ANTIBODY OR TRYPSIN".)

Trypsin Antibody

HSA ABG33

KLH ABC-33 (control/experimental)

MSH HA titer

18.6/0.0" Not done

9.4/2.4 11.3/1.8

Not done 80/neg.

From Unanue et al. (1972). Treatment of macrophages in uitro with trypsin or antibody markedly reduced the immune response to macrophage bound antigen. The immune response to human serum albumin (HSA) w&sdetermined 2 weeks after transfer of macrophages containing 4 pg. of HSA (titers are expressed as micrograms of HSA N bound per milliliter of serum X 1W). The immune response to keyhole limpet hemocyanin (KLH) represents the secondary antibody titers. I n these experiments the mice were primed with 4 pg. of KLH. The immune response to Maia squinudo hemocyanin (MSH) represents secondary hemagglutination titers. Mice were given 1 pg. of MSH in 4.5 X 106 cells, either treated or untreated with antibody, and challenged 4 weeks later with 1 pg. of free MSH i.p. c Titers after transfer of control; untreated cells/titers after transfer of treated cells. a

6

added to lymph node cell fragments in uitro carried antigens and initiated a strong immune response (Garcia et aZ., 1972). Removal of the membrane-bound antigen completely abrogated the immune response. Harris claimed the presence of membrane antigen in macrophages during in vitro DNA synthesis by lymphocytes (1965). Schmidtke and Dixon (1972) correlated the antigen retained on the membranes of X-irradiated macrophages with the immune response to macrophage-bound antigen. The amount of antigen remaining on the plasma membrane in tissue culture experiments varied depending upon the type of antigen and state of activation of the macrophage. For example, 1-2% of hemocyanin molecules were retained for prolonged periods. In the case of polymeric albumin, although the amount was about the same, the retained molecules dissociated faster (half-life of about 1 day), Monomer albumin was readily lost from the membrane (Schmidtke and Unanue, 1971a). It is clear that different antigens have to be analyzed critically before generalizations are made. Concerning the state of activation of macrophages, the amount of the hemocyanin on the membranes of macrophages activated by Listeria monocytogenes infection was about 5 to 10 times more than in macrophages isolated from the unstimulated peritoneal cavity-as would be expected, the activated macrophages induced, when transferred, a somewhat higher immune response (Lane and Unanue, 1972). The origin and persistence of the few molecules of protein antigens on the membrane of macrophages, a cell that displays strong endocytotic

142

E. R. UNANUE

activity, needs explaining. The evidence appears reasonably clear that most of the molecules derived from the first contact with extracellular antigen ( Unanue and Cerottini, 1970; Schniidtke and Unanue, 1971a). This conclusion was reached after removing the surface-bound antigen by trypsin and then observing that new antigen did not appear on the membrane. No relationship was apparent between the antigen in intracellular vesicles and the antigen on the membrane (Unanue and Cerottini, 1970; Schmidtke and Unanue, 1971a). One explanation for the persistence of these few molecules may be that they were held on surface areas having little endocytotic activity. This would imply that the plasma membrane of macrophages is not homogeneous but is made up of different functional areas insofar as endocytosis is concerned. In fact, stimulating pinocytosis on macrophages that have retained antigen on their membranes did not result in the interiorization of the retained antigen [i.e., macrophages were exposed to antigen and cultured for a day at which time a small amount of antigen was retained on the surface whereas the bulk had been pinocytized and catabolized; the culture conditions were then changed to increase endocytosis (Unanue et ul., 1972)l. The recent experiments of Tsan and Berlin (1971) tend to support this explanation of different functional areas on the macrophage membrane. They found that the sites for amino acid transport 011 the macrophage remained the same before and after phagocytosis, i.e., despite interiorization of membrane as a result of endocytosis the transport sites were preserved. They envisioned the membrane as “mosaic in character with geographically separate transport and phagocytic sites.” 2. Antigen in lntrucellular Vesicles Information on the immunogenicity of complex particulate antigens, such as red blood cells or microorganisms, when taken up by macrophages is confused by the chemical complexity of the particulate antigen and the poorly understood nature of inimunogens. Apart from this, particulate antigens are not always easy to separate from macrophages, and there is always the possibility that tests of macrophage-associated antigens are contaminated with free nonphagocytized antigens. Some experiments showing immunogenicity of particulate antigens bound to live macrophages appear in Table I. In most of these experiments the antigen presumably had been interiorized by the macrophage, although there was no conclusive proof of this. It is possible, although unlikely, that these antigens remained on the surface membrane without being endocytized for some time (evidence for intact SRBC in or around macrophages has been brought forth in Mishell-Dutton in vitro responses).

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

143

Cruchaud and I attempted to follow some antigens of foreign red cells to obtain some idea of the possible fate of the endocytized immunogen (Cruchaud and Unanue, 1971a). We used SRBC the stroma of which were labeled with Iz5Ior to which specific 1 g G - Y was bound. The macrophages were exposed to SRBC for a period of time after which the nonphagocytized SRBC were eliminated. The cells were then placed in culture for up to 1-3 days. It was found that, after the period of culture, partially degraded material appeared in the culture fluid. In the case of 12JI-labeledantibody the released material was a fragment of the Ig molecule. We concluded that small amounts of the endocytized antigens had escaped complete degradation, were released, and, hence, had become available for immune recognition. Evidence also indicated that the antigens from particulate substances liberated by the macrophage could bind de novo to their surface membrane for presentation to lymphocytes. Indeed, treatment of macrophages that had taken up SRBC with trypsin or antibody curbed their immunogenicity (as assayed in the macrophage transfer system), The exact pathway that the material followed in the cell and the conditions determining its release have not been established. Indeed, there is not much information available on the final intracellular fates of different phagocytic vesicles. Is it possible that some phagocytic vesicles are channeled into a different pathway, away from typical primary lysosomes, or that some vesicles fuse with lysosomes that have a small content of hydrolases? Can the digestive process in a phagolysosome be uneven with a resulting pool in the vesicle of undigested or partially digested material? Many other experiments have demonstrated that immunogenic material could be extracted from intracellular vesicles for some time after uptake. Uhr and Weissman isolated a “large granule” fraction from liver macrophages that had taken up bacteriophages and which was immunogenic as well as containing active phage (Uhr and Weissman, 1965, 1968). Franzl ( 1962) isolated “lysosomal” fractions, which were also immunogenic, from spleens of mice injected 1 and 3 days previously with SRBC. Kolsch and Mitchison (1968) identified antigen in two cell compartments: one was present in a light sedimenting fraction with a rapid turnover to which 90%of the antigen was channeled; the other was present in a dense sedimenting fraction of slow turnover and associated with the remaining 10%of the antigen. This last dense, sedimenting compartment was heterogeneous and most likely contained cell membrane, nuclei, and debris. These same investigators also found that antigen bound to live macrophages was more immunogenic than the antigen extracted from the cytoplasm. The SRBC antigens extracted from macrophages also induced strong delayed reactions (Pearson and Raffel, 1971).

144

E. R. UNANUE

Antigen extracted from macrophages has been found complexed with RNA (see below). The release process of intracellular antigen is undetermined. It appears clear that soluble antigens do not cross the membrane of the secondary lysosome under normal circumstances. The extent to which the lysosomal membrane is permeable to small products of catabolism was investigated by Ehrenreich and Cohn ( 1969). They exposed macrophages to di- or tripeptides made of D-amino acids which were not degradable by macrophage enzymes. The macrophages took up the peptides and stored them in secondary lysosomes for some periods of time-the distended secondary lysosomes containing the peptides were easily visualized as translucent vesicles by phase microscopy. In contrast, small peptides of L-amino acids were degraded, easily crossed the membrane, and left the cell. The conclusion was that the lysosomal membrane was only permeable to small peptides up to the level of 2 or 3 amino acids. Antigen could be released from a cell as a result of death, by channels connecting the phagocytic vesicle with the surface of the cell or by exocytosis. Death of cells is known to lead to the release of cytoplasmic material and could account for release of intracellular antigens. Yet in the experiments of Cruchaud and Unanue (1971a) the estimated amount of cell death could not account for the release of antigen, although direct measurement of other cell cytoplasmic components in the culture fluids was not done. It should be pointed out that this release of antigen from macrophage may be akin to the release of lysosomal enzymes that has been observed to take place after phagocytosis of large particles by both macrophages ( Weissman et al., 1971b) and polymorphonuclear leukocytes (Henson, 1971; Weissman et al., 1971b). Weissman et al. ( 1971b) found that macrophages which had taken up zymosan particles released lysosomal P-glucuronidase into the culture fluid but did not release lactic dehydrogenase, a cytoplasmic enzyme. To this effect both Henson (1971) and Karnovsky and Karnovsky ( 1971) have ultrastructural evidence in polymorphonuclear leukocytes for openings or channels from the phagocytic vesicles to the surface through which the enzymes could be released. The final possible pathway of release of intracellular antigen is by reversed pinocytosis or exocytosis. Exocytosis of antigen is a possibility that has been raised to explain the continuous presence of small amounts of undigested antigens in body fluids (Janeway and Humphrey, 1969). Such antigens are taken up and stored by the RES. There is at present no morphological documentation of exocytosis by macrophages.

3. RNA-Antie ?n Complexes In early 1960, Fishman (1961) and Fishman and Adler (1963b) reported that lymphocytes responded ii Tmunologically to extracts of peri-

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

145

toneal exudate cells from rats injected a few minutes previously with T2 phage. The experimental system consisted of culturing lymph node cells or, alternatively, of placing lymph node cells in diffusion chambers and implanting them into X-irradiated rats (reviewed in Fishman and Adler, 1963a). The filtrate from peritoneal cells exposed to T2 phage induced neutralizing antibodies by 8 days of culture; T2 phage by itself or normal extract of macrophage (i.e., not exposed to T2 phage) did not induce antibody. The antibody response was specific in that the cultures did not contain neutralizing antibodies to T1 or T5 phage. The same authors demonstrated that the extract was sensitive to ribonuclease ( RNase ) treatment; indeed, extracting RNA with cold phenol yielded active material which sedimented as a light molecular weight RNA. It was these kinds of experimental results that led many to conclude that for an antigen to be immunogenic it had to be “processed by macrophages. The system itself has been criticized by those familiar with tissue culture assays because some of the early results were not reproducible, and antibody titers were low. It was subsequently found that antigen-or still better, a fragment of it-was associated with the RNA extract. Askonas and Rhodes ( 1965a,b) isolated RNA (by cold phenol extraction) from murine macrophages that had taken up lZ5I-labeledhemocyanin. A small amount of the iodinated protein (about 0.01%of the total taken by the macrophage) was found associated with RNA. The material, when injected into mice, was discovered to be highly immunogenic. Friedman et al. (1965) incubated peritoneal exudate cells, in witro, with T2 phage, isolated RNA, and found, by using complement fixation assays, antigens of head, tail, and internal protein of the phage in the extract. Subsequent studies have confirmed these two early findings; indeed, after uptake of many antigens by macrophages, a complex of RNA with antigen can be isolated from these cells (Gottlieb and Doty, 1967; Roelants and Goodman, 1968; Herscowitz and Stelos, 1970a,b; Bishop and Gottlieb, 1971). The biological significance of this finding is in doubt, however, and strong evidence has been presented that the phenomenon may even be a laboratory artifact ( Roelants and Goodman, 1969; Goodman, 1972). Some of the characteristics of the RNA-antigen complex are known. The complex isolated from macrophages recently exposed to the antigen (Fishman and Adler, 1963a) did not involve newly synthesized RNA (Gottlieb and Doty, 1967). In one such experiment, macrophages were pulsed with tritiated uridine at the same time that T2 phage was added to the cells. The RNA was extracted and analyzed in a gradient. The fraction containing the immunogenic moiety was not associated with the radioactive label. The size of the RNA-antigen complex appeared to be heterogeneous probably because of the different sizes of the antigen

146

E. R. UNANUE

associated with the RNA (Saha et al., 1964; Roelants and Goodman, 1968; Roelants et al., 1971). Gottlieb, however, found the RNA banding in cesium sulfate gradients at an invariable density of 1.588 (Gottlieb, 1969; Gottlieb and Straus, 1969); this band disappeared after treatment with proteolytic enzymes ( Gottlieb and Doty, 1967). Roelants and Goodman (1969) estimated that their RNA-antigen complex had an S value of about 4. Many antigens taken up by macrophages could complex with RNAthe common denominator of these antigens was the presence of charged groups in the molecule (Roelants and Goodman, 1969). Uncharged antigens, although immunogenic in vivo, were not found to be associated with RNA from macrophages. Mixing RNA with antigen in vitro did not result in complex formation ( Roelants and Goodman, 1968). Roelants and Goodman (1969) studied the binding of pure RNA to negatively charged antigen and found that the presence of Mg salts complemented binding; they also found that the RNA-antigen complex dissociated after exhaustive dialysis against buffers containing chelating agents. Their explanation was that antigen chelated to Mg would bind and aggregate a transfer-type RNA present as a precursor in an unspun or expanded form. Thus, the complex may be entirely a laboratory aitifact resulting from cellular disruption which allows RNA to mix with antigenic fragments. Indeed, the lysosomal membrane was highly impermeable to small antigen fragments of up to a size of 2 or 3 amino acids (Ehrenreich and Cohn, 1969). It cannot be ruled out, however, that local combination of RNA with antigen fragments can occur as a result of cell death in tissues. Garvey et al. (1967) have reported on the presence of RNA complexed to antigen in urine. Is the RNA-antigen complex specific to macrophages? Not all investigators agree here. Gottlieb contended that the complex was, indeed, peculiar to macrophages since he was unable to find it among other types of cells. However, the recent experiments of Roelants et al. ( 1971) clearly indicated that in vitro complexing of antigen to RNA of other cells like HeLa or even bacteria (Escherichia coli) could occur. Moreover, in their hands, these complexes formed with RNA of other nonmacrophagic cells were highly immunogenic. Whether their antigenRNA complexes are identical to those reported by Gottlieb has not been ascertained. Finally, it is also clear that there was no relationship between complexing of antigen to RNA and its immunogenic capacity when injected in vivo (Roelants and Goodman, 1969). For example, a poly-y-D-glutamyl polypeptide was found to complex with 4s RNA of macrophages. Such a polypeptide, however, was nonimmunogenic when administered in vivo (unless complexed with methylated albumin ) (Roelants and Goodman, 1968).

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

147

In summary, although the complexing of antigen with RNA may result in a strong immunogenic complex in uitm, we still have no indication that such a complex operates in uiuo. Roelants and Goodman’s experiments indicate that when a cell containing antigen in the vesicles is disrupted, it is practically unavoidable for the RNA and antigen to mix and not form a complex. Hence, the existence of complexes preceding cell disruption is questionable. As mentioned previously, however, such complexes could conceivably form in uivo in areas where cell death readily occurs after phagocytosis. The reasons for the high immunogenicity of the RNA-antigen complex are not known, but complexing of antigen to many chemicals may enhance iriimunogenicity by means not yet understood. It should be remembered that synthetic polynucleotide complexes, such as polyadenilic and polyuridilic acids, are potent adjuvants to antigens administered simultaneously ( Braun et al., 1968; Schmidtke and Johnson, 1971). Some observations on RNA from macrophages are not explicable on the basis of an RNA-antigen complex. In these RNA extracted from peritoneal exudate cells exposed to phage from rabbits of a given allotype was mixed with normal lymphocytes of rabbits of a different allotype. Part of the antibody molecules (the IgM) carried the allotypic specificities of the donor (Adler et al., 1968). Similar experiments with extracts from lymphocytes (Bell and Dray, 1969, 1970) suggest a messenger-type function in the RNA extract, the significance of which is not known. These results have been difficult to reproduce (Bluestein et al., 1970) and need to be confirmed by other investigators. Furthermore, this phenomenon may merely represent a laboratory exercise with no relevance to in viuo experiments. Recently, Goodman (1972) has discussed these experiments at length. IV. Macrophage-Lymphocyte Contact

If one of the functions of macrophages is to present some molecules of antigen to lymphocytes for immune recognition, it is reasonable to expect that there should be a close anatomical association of both cells at some time. Clusters of lymphocytes and plasma cells around macrophages are frequently observed in the perifollicular areas of lymph nodes (Thiery, 1962; H. R. P. Miller and Avrameas, 1971). Needless to say, the association of lymphocytes around dendritic cells of the follicles is a very close one. There is one ultrastructural study (Schoenberg et al., 1964) on cell fusion between macrophages and lymphocytes of the spleen, but this phenomenon, if present, appears to be the exception rather than the rule. In the Mishell-Dutton tissue culture system, lymphocytes readily cluster around macrophages; the importance of this clustering for immune induction has been discussed previously ( Mosier,

148

E. R. UNANUE

1967; Pierce and Benacerraf, 1969). Electron microscopy of these clusters reveals close approximation of lymphocytes with macrophages but no cytoplasmic fusion ( McIntyre et al., 1971 ) . In tissue cultures employing diffusion chambers, it was possible to identify cytologically the lymphocytes clustering around a central macrophage containing the antigen (Sulitzeanu et al., 1971). Lymphocytes have also been shown to cluster around blastlike large cells of unknown origin (Sulitzeanu et at., 1971). McFarland and colleagues have studied the movement of lymphocytes around macrophages (or other cells) and have identified a special area in one pole of the lymphocytes-the uropod-which may be important in cell-to-cell contact (Fig. 7 ) ( McFarland and Heilman, 1965; McFarland et al., 1966; McFarland, 1969). Recently it has been shown that B lymphocytes tend to attach preferentially to macrophages ( Schmidtke and Unanue, 1971b). This adherence resulted from interaction of Ig receptors on the B lymphocytes with the Fc cytophilic receptor on macrophages, since it was blocked by aggregated Ig. Moreover, B lymphocytes did not attach preferentially to fibroblasts, cells lacking receptors for Ig. There is a report of adherence of guinea pig thymocytes to macrophages-adherence was increased by addition of guinea pig serum and abolished by trypsinization of macrophages. Heterologous thymocytes did not adhere to the guinea pig macrophages ( Siegel, 1970).

FIG.7. Photomicrograph of a mixed leukocyte culture. Lymphocytes are in close approximation to a maLrophage. The typical appearance of two lymphocytes forming the uropod is clearly seen. Original photographs were taken with Nomarski interference microscopy at an approximate magnification of 600. I thank Dr. W. McFarland (Veterans Administration Hospital, Washington, D. C. ) for allowing us to publish this photograph.

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

149

V. Macrophages and Adjuvants

In recent years it was speculated that macrophages might be involved in the heightened immune response produced by adjuvants, although nowhere has direct, clear evidence been introduced. 1. Many adjuvants such as the water-in-oil emulsions or alum emulsions, devised originally by Freund, produced a marked inflammatory reaction at the tissue of inoculation or at distinct sites characterized by a heavy macrophage infiltrate (White et al., 1955; French et al., 1970; Steinberg et al., 1970). The macrophages of this inflammatory reaction formed epithelioid and giant cells. Plasma cells secreting specific antibody were abundant in the granulomatous tissue. White and collaborators, when studying the immunological response in chickens to human albumin mixed with water-in-oil emulsions (French et al., 1970; Steinberg et al., 1970) observed that there was a small early peak of antibody synthesis (which was also present in animals given albumin without adjuvants) followed by a very large peak about 40 days after injection. The onset of the last peak coincided with the development of a large granulomatous reaction at the site of the adjuvant-antigen depot. The granuloma was characterized by abundant plasma cells secreting specific antibody in contrast to the spleen which showed a marked paucity of plasma cells. Interestingly, thymectomy but not bursectomy of the chicken resulted in marked reduction of the granuloma. Bursectomy resulted in a reduced peak antibody response and in fewer plasma cells in the granuloma; the hyperactive macrophages were still abundant in the granuloma of these chickens. The conclusion, which seems reasonably clear, was that the late granuloma reaction (with the hyperactivated macrophages ) was associated with the presence of an intact thymus-perhaps abundant macrophages at the granuloma contributed in some way to the heightened immune reaction. In fact, Humphrey and Turk, in 1963, reported and Leskowitz (1970) confirmed that injection of diphtheria toxoid at the skin site of a delayed hypersensitivity reaction to tuberculin (which is full of macrophages) resulted in 2 to 4 times more antitoxoid antibody production. 2. A series of adjuvants have marked effects on lysosomal membranes. These adjuvants include silica (Pernis and Paronetto, 1962), vitamin A or retinol ( Dresser, 1968; Spitznagel and Allison, 1970a), endotoxin ( Spitznagel and Allison, 1970a), beryllium ( Unanue et al., 1969a)) and Bordetella pertussis vaccine (Unanue et al., 1969a). Culture with these adjuvants led to marked effects on macrophages (Kessel et al., 1963; Kessel and Braun, 1965; Allison et al., 1966; Wiener and Levanon, 1968; Heilman and Robert, 1967; Heilman, 19%) (Fig. 8) characterized

150

E. R. UNANUE

FIG.8. Changes in appearance of macrophages and number of lysosomes after exposure to immunological adjuvants. Lysosomes were visualized by exposing the cells to the fluorescent dye, Euchrysine. The treated macrophages to the left show large Euchrysine-positive vacuoles; the macrophages on the right are controls. Results similar to these can be obtained by incubating macrophages with endotoxin, beryllium, Bordetelkz pertussis, etc. [see J. Immunol. 103, 7 (1969). This technique of staining live cells with Euchrysine is described by Allison and Young (1964).]

by formation of large lysosomes, increased hydrolytic enzymes, and, at large doses, death. Macrophages containing antigen and treated with adjuvants were studied in the macrophage transfer system. It was found that mice made a stronger immune response to hemocyanin bound to beryllium- or B. pertussis-treated macrophages than to normal niacrophages. This heightened immune response was unrelated to an effect on the handling of antigen, since transfer of macrophages containing antigen together with macrophages treated with the adjuvant resulted in an increased response. In essence, part of the adjuvant effect could be transferred by macrophages exposed to these adjuvants. However, the nature of this effect was not determined. Adjuvants increased the formation of lysophosphatides and free fatty acids in macrophages, and some have speculated that these surface active substances may mediate the adjuvant effect (Munder et al., 1970).

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

151

It has been shown that beryllium (Schmidtke and Unanue, in preparation), or vitamin A (Dresser et al., 1970), which were so avidly taken up by tissue macrophages, could produce marked effects on the distribution pattern of lymphoid cells in spleen and lymph nodes. These effects were seen early (usually within the first 24 hours) and were characterized by transitory increase of lymphocytes in the paracortex. Injection of adjuvant with Wr-labeled lymphocytes led to their increased entrapment in the spleen. 3. There was a marked adjuvant effect when antigen was administered together with lactic dehydrogenase virus (Notkins et al., 1966; Mergenhagen et al., 1967) or Listeria monocytogenes (Lane and Unanue, 1972). Lactic dehydrogenase virus is readily taken up by macrophages (DuBuy and Johnson, 1969), whereas listeriosis is known to produce marked macrophage activation. The mechanisms of adjuvant action and the role of the macrophage were not established in any of these three situations. The evidence for some role of macrophages-perhaps by antigen presentation, by affecting in some way neighboring cells through release of inflammatory factors, or by affecting the traffic of cells in a lymphoid organ-is circumstantial but suggestive enough to warrant close attention and further experimentation. VI. Summary

The immune response to many antigens involves the participation of two lymphocytes ( T and B ) that are specifically committed to react with different but closely spaced antigenic determinants (J. A. F. P. Miller et al., 1971). These lymphocytes appear to have Ig receptors on their membranes and can bind antigen. Participating in this interaction between antigen and T and B lymphocytes is the macrophage, a cell line endowed with the properties of taking up, degrading, and retaining a wide range of antigens. There is agreement that the specificity of the immune response lies in the T and B lymphocytes-the macrophage handles a variety of materials either immunogenic or not without much immunological discrimination. The macrophage does not instruct the lymphocyte to react wth antigen, but rather it is the committed lymphocyte that selects the molecule of antigen with which it can react. The experimental results that show immunogenic material associated with macrophages appear reasonably clear, although the molecular basis for macrophage-antigen-lymphocyte interaction and for lymphocyte triggering needs to be determined. The participation of macrophages in the context of the lymph node and spleen histology is not resolved. Further work needs to be done in order to clarify the origin and func-

152

E. R. UNANUE

tion of the elusive dendritic cells of the lymphoid follicles. The pathways for presentation of immunogenic moieties by macrophages, although outlined, need further exploration with several types of antigen. Finally, improvement of and experience with tissue culture methodology should make possible clarification of many of the problems associated with cellular interactions involving macrophages and lymphocytes. The effects of macrophages in immune induction can be discussed along three separate lines: ( 1 ) the capacity of the macrophage to remove and eliminate extracellular antigen; ( 2 ) their property of holding a few immunogenic moieties; and ( 3 ) their ability to degrade antigen. 1. The macrophage of both lymphoid and nonlymphoid tissues participates in the removal and degradation of soluble or particulate antigen molecules from the circulation and extracellular fluids, a process which is critical for immune induction. The presence of disseminated antigen molecules might eventually lead to their interaction with isolated T or B lymphocytes. Such an isolated interaction (without other cells participating) might result in an ineffective process. On the one hand, a T lymphocyte that meets antigen outside the lymphoid tissues, even if triggered to proliferate, does not have a responding B lymphocyte nearby and, therefore, part of its biological function is lost; on the other hand, a B lymphocyte that meets antigen will not be triggered unless a T cell is in close proximity (see Katz and Benacerraf, 1972). Critical experiments showing the deleterious effects of a large pool of unconcentrated antigen are that ( u ) the injection of molecules of monomer albumin or Ig leads to tolerance, whereas injection of polymeric molecules that leave the circulation produces immunity ( Section II,A,3) ; ( b ) injection of free antigen together with macrophage-bound antigen reduces the response to the latter-in essence, a competition for the lymphocyte between the extracellular diffusable antigen and the concentrated moieties exists (Section II,A,5); and, very important, ( c ) , in cases where the RES cannot degrade antigen, these molecules persist throughout the life of the animal and lead to life-long tolerance (here the RES may even contribute to tolerance by continuous release of the undegraded antigen ) (Howard and Siskind, 1969; Janeway and Humphrey, 1969; Medlin et d.,1970). In these experiments we also have to consider the possibility that “monomeric” molecules by themselves may not trigger the T and B lymphocyte and the “polymeric” forms, whether macrophagebound or free, may be needed to induce the cells to proliferate-analysis of the mechanism of cell triggering will presumably solve this problem. 2. We conclude from the preceding discussion that the diffusion of isolated molecules of antigen throughout tissues may be detrimental and that for antigen to initiate an immune process effectively it should

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

153

be concentrated in critical sites. There are indications for a homeostatic process that concentrates antigen and lymphocytes in the lymphoid tissues. Antigen is concentrated in part or totally by macrophage uptake ( Section 11,AJ ) ; after antigen injection, circulating lymphocytes pool into a lymph node or the spleen by mechanisms yet unknown (Zatz and Lance, 1971). Hence, the proposed helper role of the macrophage is to favor the meeting of antigen with the few committed T and B lymphocytes at one point. This statement does not rule out that free molecules of some antigens in a lymphoid tissue, if available, may lead to effective induction of immunity. Clearly, antigen bound to macrophages in the absence of free extracellular antigen may be a very effective pathway for immunogenicity, as was discussed in Section II,A,5 where it was mentioned that both T and B cells may need to interact with some macrophage-bound antigen. The length of time that antigen bound to macrophages helps in the inductive process is not clearly determined (see Section II,A,2). 3. The final point concerns the need for processing or for changing the structure of the antigen in order that it can be recognized by the lymphocytes. This point can be analyzed with respect to different kinds of antigen. Is there a need for degradation or change of polypeptides or proteins? It is apparent that such antigens are recognized before degradation, since most of the antibodies that they elicit are directed against steric or conformational determinants (Sela et al., 1967; Sela, 1969). Moreover, the need for two close determinants ( a “hapten” and a “carrier”) for immune induction precludes any breakdown or digestion of the antigen before recognition. The need to maintain native configuration does not rule out participation of macrophages in induction of antibody. We now know that the immune response to proteins bound to macrophages is directed against molecules, although few in number, which have not been digested (Unanue and Askonas, 1968a; Unanue and Cerottini, 1970). To this end the response to hapten carriers can take place after uptake of antigen by macrophages (Hamaoka et al., 1971)-another indication that the immunogenic protein moieties in macrophages are represented by nondegradable antigen. It is also clear that native proteins can bind to lymphocytes (Naor and Sulitzeanu, 1967; Ada and Byrt, 1969; Humphrey and Keller, 1970) and, in fact, can trigger the cells under appropriate circumstances (Britton et al., 1971). Hence, the role of the macrophages related to protein antigens may be explained mainly by their capacity to concentrate the antigen and allow for cell interactions as discussed above. Insofar as large particulate antigens are concerned, the information available is neither extensive nor conclusive. Lymphocytes can react

TmLX V ROLEOF MACROPHAGE IN IMMUNE RESPONSE TO FOUR KINDSOF ANTIGEN

Removal of antigen from extracellular fluids

Concentrationand presentation of antigen

Change of antigen

1. Soluble antigens: mainly proteins (extent of uptake may depend upon degree of polymerization)

Very important in nonimmune individuals (1,2). I n immune individuals antibody may aid in eliminating antigen.

Importantdepending upon number of specific T and B lymphocytes (1). No information available on the ext,ent to which antigen is presented by macrophages as compared to other pathways.

Macrophage does not alter the molecule but plays a passive role by presentation of “native” molecule (3, 4). More information needed in molecular mechanisms of lymphocyte triggering.

2 . Particulate antigens: bacteria and red cells (highly phagocytized by macrophages)

Important depending upon Important (6)-depends site of entry of antigen and number of T and B upon the amount of antilymphocytes. gen administerednonlymphoid macrophages may reduce the amount of antigen available to lymphoid organs and, therefore, reduce immune response (5).

Antigen

upon

Some experiments indicate that partial degradation may be necessary (7,8).

M

Tj

zr z

ti

3. T independent antigens, such as pneumococcal polysaccharide or polymerized flagellin (antigens with repeating antigen determinants)

Important-in the case of SIII (or other noncataboliaable antigens) the lack of removal may lead to persistence of tolerance (9, 10)

May depend upon number of B lymphocytes-no information available.

4. Tissue or cell antigens, such as tumors or grafts (no

Not necessary

Not necessary

Not necessary (7)

3 0

May not be necessary

diffusable antigen)

Key t o references: 6. Gorczynski et d.,1971 1. Mitchison, 1969b 7. Shortman and Palmer, 1971 2. Spitznagel and Allison, 1970b 8. M. Feldman and Palmer, 1971 3. Unanue and Cerottini, 1970 9. Howard and Siskind, 1969 4. Unanue et d., 1969 10. Medlin et al., 1970 5 . Perkins, 1970 The references refer to key papers that have dealt with the particular problem-the conclusions in this Table are mine and do not necessarily reflect that of the authors of the reference. a

L

$

z

b3

i 0

n

v)

Q

*

z

3

s

156

E. R. UNANUE

with red cells or Shigella, for example, but we do not know if that interaction can trigger the lymphoid cell. Some experiments suggest that breakdown by macrophages of the SRBC (Shortman and Palmer, 1971; M. Feldman and Palmer, 1971) or Shigella (Gallily and Feldman, 1967) may be needed for immune induction. These interesting observations need confirmation and should be extended to other large particulate antigens. Apart from this point, there is indirect information that there may be in viuo immune recognition of “hidden” bacterial antigenic determinants which most likely have been exposed as a result of partial degradation and release by macrophages (Schwab and Brown, 1968; Schwab and Ohanian, 1967). In this case the macrophage may contribute to the heterogeneity of antibody by releasing these hidden antigens; along these lines the RNA-antigen complex, whether formed before or after cell death, may add to the pool of immunogenic moieties. The extent to which the three macrophage processes-antigen removal, antigen concentration and presentation, and antigen changeinfluence the immune response depends upon (1) amount of antigen, ( 2 ) physicochemistry of the antigen, ( 3 ) number of reactive T and B lymphocytes, ( 4 ) T dependency of the antigen, ( 5 ) presence or absence of adjuvant-type effects, ( 6 ) the anatomical site of antigen arrival, and ( 7 ) the molecular mechanisms of lymphocyte triggering. Some of these variables have been considered. It is apparent that the degree to which the immune response needs macrophage participation may be very different in the case of a soluble antigen and in the case of an antigen forming part of a cell membrane, such as a tumor cell. In the latter case there is no need for antigen removal nor for concentration (i.e., the antigens are already concentrated on a membrane and there is no extracellular diffusable antigen ) ; also, the lymphocyte apparently can recognize the concentrated tissue antigen without any need of macrophages. In contrast, in the case of soluble antigen, macrophage participation is clearly present but is influenced b y the number of reactive lymphocytes, amount of antibody, or by concomitant injection of adjuvants. In judging such a complicated process as immune induction, we must take into consideration each variable. Table V is an attempt to summarize these points with few different antigenic stimuli and with the incomplete series of observations now at hand. Perhaps in the future when more data will be available these points can be better documented.

ACKNOWLEDGMENTS The work reported here was done in collaboration with many colleagues t o whom I am deeply indebted. I wish to thank Brigitte A. Askonas. John H. Humphrey, Frank J. Dixon, and Baruj Benacerraf in whose laboratories I have had the opportunity to work and discuss many problems in the field of cellular immunology and

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

157

to express my appreciation to those investigators who sent me unpublished observations or who authorized me to reproduce some of their material. I thank Miss Karen Ellis for excellent secretarial assistance.

REFERENCES Abramson, N., LoBuglio, A. F., Jandl, J. H., and Cotran, R. S. (197Oa). J. Ersp. Med. 132, 1191. Abramson, N., Gelfand, E. W., Jandl, J. H., and Rosen, F. S. (1970b). J. Exp. Med. 132, 1207. Ada, G. L., and Byrt, P. (1969). Nature (London) 222, 1291. Ada, G. L., and Lang, P. G. (1966). Immunology 10, 431. Ada, G. L., Nossal, G. J. V., and Austin, C. M. (1964). Aust. J. E x p . Biol. Med. Sci. 42, 331. Ada, G. L., Humphrey, J. H., Askonas, B. A., McDevitt, H. O., and Nossal, G. J. V. (1966). E x p . Cell Res. 41, 557. Adler, F. L., Fishman, M., and Dray, S. (1968). J. Immunol. 97, 554. Allen, J. M., and Cook, G. M. W. (1970). E x p . Cell Res. 59, 105. Allison, A. C., and Young, D. D. (1964). Life Sci. 3, 1407. Allison, A. C., Harington, J. S., and Birbeck, M. (1966). J. Exp. Med. 124, 141. Argyris, B. F. (1967). J. Immunol. 99, 744. Argyris, B. F., and Askonas, B. A. (1968). Immunology 14, 379. Argyris, B. F., and Plotkin, D. H. (1969). J. Immunol. 103, 372. Askonas, B. A., and Jaroskova, L. (1970). In “Developmental Aspects of Antibody Formation and Structure” ( J . Sterzl and I. Riha, eds.), 2nd ed., Vol. 1. Academic Press, New York. Askonas, B. A., and Rhodes, J. M. (1965a). Proc. S y m p . Mol. Cell. Basis Antibody Formation, 1964, p. 503. Askonas, B. A., and Rhodes, J. M. (1965b). Nature (London) 205, 470. Auzins, I., and Rowley, D. (1963). Aust. J. Exp. Biol. Med. Sci. 41, 539. Bach, F. H., Alter, B. J., Solliday, S., Zoschke, D. C., and Janis, M. (1970). Cell. Immunol. 1, 219. Balfour, B. M., and Humphrey, J. H. (1966). In “Germinal Centres in Immune Responses” (H. Cottier et al., eds.), pp. 80-85. Springer-Verlag, Berlin and New York. Bell, C., and Dray, S. ( 1969). J. Immunol. 103, 1196. Bell, C., and Dray, S. (1970). J. Immunol. 105, 541. Benacerraf, B. (1964). In “The Liver” (C. Rouiller, ed.), Vol. 2, p. 37. Academic Press, New York. Benacerraf, B., and Green, I. (1969). Annu. Reo. Med. 20, 141. Benacerraf, B., Stiffel, C., Biozzi, G., and Halpern, B. M. (1954). C. R. SOC. Biol. 148, 486. Benacerraf, B., Biozzi, G., Cuendet, A., and Halpern, B. M. (1955a). J . Physiol. (London) 128, 1. Benacerraf, B., Halpem, B. N., Stiffel, C., Cruchaud, S., and Biozzi, G. (195%). Ann. Inst. Pasteur, Paris &9, 601. Benacerraf, B., Stiffel, C., and Biozzi, G. (1956). C. R. SOC. Biol. 150, 1161. Benacerraf, B., Biozzi, G., Halpem, B. N., and Stiffel, C. (1957a). In “Physiopathology of the Reticuloendothelial System” (B. N. Halpern, B. Benacerraf, and J. F. Delafresnaye, eds. ), p. 52. Thomas, Springfield, Illinois.

158

E. R. UNANUE

Benacerraf, B., Biozzi, G., Halpern, B. N., Stiffel, C., and Mouton, D. (1957b). Brit. J . Exp. Pathol. 38, 35. Benacerraf, B., Sebestyen, M. h4., and Schlossman, S. (1959). J. Exp. Med. 110, 27. Berken, A., and Benacerraf, B. (1966). J. E x p . Med. 123, 119. Bianco, C., Patrick, R., and Nussenzweig, V. (1970). 1. Exp. Med. 132, 702. Biozzi, G., Benacerraf, B., Grumbach, F., Halpern, B. N., Levaditi, J., and Rist, N. ( 1954). Ann. Inst. Pasteur, Paris 87, 291. Biozzi, G., Halpem, B. N., Benacerraf, B., and Stiffel, C. (1957). In “Physiopathology of the Reticuloendothelial System” (B. N. Halpern, B. Benacerraf, and J. F. Delafresnaye, eds.), p. 205. Thomas, Springfield, Illinois. Biozzi, G., Benacerraf, B., Halpern, B. N., Stiffel, C., and Hillemand, B. (1958). J. Lab. Clin. Med. 51, 230. Biozzi, G., Howard, J. G., Halpern, B. N., Stiffel, C., and Mouton, D. (1960). Immunology 3, 74. Biro, C. E., and Garcia, G. (1965). Immunology 8, 411. Bishop, D. C., and Gottlieb, A. A. (1971). 3. lmmunol. 107, 269, Bluestein, H. G., Green, I., and Benacerraf, B. (1970). Proc. SOC. Exp. Biol. Med. 135, 146. Boros, D. L., and Warren, K. S. ( 1971). J. Immunol. 107, 534. Boyden, S. V. (1963). In “Cell-Bound Antibodies” (B. Amos and H. Koprowski, eds. ), p. 7. Wistar Inst. Press, Philadelphia, Pennsylvania. Boyden, S. V. (1964). Immunology 7, 474. Braun, W., Nakano, M., Jaraskova, L., Yajima, Y., and Jimenez, L. (1968). In “Nucleic Acids in Immunology” (0. J. Plescia and W. Braun, eds.), p. 347. Springer Publ., New York. Britton, S., and Celada, F. (1968). Immunology 14, 503. Britton, S., and Moller, G. (1968). J. Immunol. 100, 1320. Britton, S., Wepsic, T., and Moller, G. (1968). Immunology 14, 491. Britton, S., Mitchison, N. A,, and Rajewsky, K. (1971). Eur. J. Immunol. 1, 65. Brumfitt, W., Glynn, A. A., and Percival, A. (1965). Brit. J. Exp. Pathol. 46, 213. Campbell, D. H., and Garvey, J. S. (1961). Lob. Inuest. 10, 1126. Campbell, D. H., and Garvey, J. S. (1963). Aduan. Immunol. 3, 261-313. Cannon, P. R., Baer, R. B., Sullivan, F. L., and Webster, J. R. (1929). J. lmmunol. 17, 441. Cayeux, P., Panijel, J., Cluzan, R., and Levillain, R. ( 1966). Nature (London) 212, 688. Celada, F. (1967). J. E r p . Med. 125, 199. Cerottini, J.-C., and Tmka, Z. (1970). Int. Arch. Allergy Appl. rmmunol. 38, 37. Cerottini, J.-C., and Unanue, E. R. (1971). J. Immunol. 106, 732. Chiller, J. M., Habicht, G. S., and Weigle, W. 0.(1971). Science 171, 813. Claman, H. N. (1963). J . Immunol. 91, 833. Cline, M. J., and Sweet, V. C. (1968). 1. E x p . Med. 128, 1309. Cochrane, C. G., Unanue, E. R., and Dixon, F. J. (1965). J. E x p . Med. 122, 99. Cohn, Z. A. ( 1962). Nature (London) 196, 1066. Cohn, Z. A. (1963a). J. E r p . Med. 117, 27. Cohn, Z. A. (1963b). J. E r p . Med. 117, 43. Cohn, Z. A. (1968). Aduan. Immunol. 9, 163-214. C o b , Z. A., and Parks, E. (1967). J . E r p . Med. 125, 1091. Cottier, H., and Sordat, B. (1971). Aduan. E x p . Med. Biol. 12, 203. Cruchaud, A. ( 1968). Lab. Znoest. IS, 15.

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

159

Cruchaud, A,, and Unanue, E. R. (1971a). J. Immunol. 107, 1329. Cruchaud, A., and Unanue, E. R. (1971b). In “Morphological and Functional Aspects of Imniunity” (K. Lindahl-Kiessling, G. A h , and M. G. Hanna, Jr., eds.), pp. 187-195. Plenum, New York. Cruchaud, A., Despont, J. P., Girard, J. P., and Mach, B. (1970). J. Immunol. 104, 1256. Despont, J. P., and Cruchaud, A. (1969). Nature (London) 223, 838. Dobson, E. L. ( 1957). In “Physiopathology of the Reticuloendothelial System” (B. N. Halpern, B. Benacerraf, and J. F. Delafresnaye, eds.), p. 80. Thomas, Springfield, Illinois. Dobson, E. L., and Jones, H. B. ( 1951). Acta Med. Scand. 145, Suppl. 273, 1. Dresser, D. W. (1962). Immunology 5, 378. Dresser, D. W. (1968). Nature (London) 217, 527. Dresser, D. W., and Mitchison, N. A. (1968). Aduan. Immunol. 8, 129-181. Dresser, D. W., Taub, R. N., and Krantz, A. R. (1970). Immunology 18, 663. DuBuy, 13. G., and Johnson, M. L. (1969). Proc. SOC. Exp. Biol. Med. 128, 1210. Dutton, R. W. (1967). Aduan. Immunol. 6, 253-336. Dutton, R. W., McCarthy, M. M., Mishell, R. I., and Raidt, D. J. (1970). Cell. Immunol. 1, 196. Dvorak, H. F., and Bast, R. C., Jr. (1970). Immunochemistry 7, 118. Ehrenreich, B. A., and Cohn, Z. A. (1967). J. E x p . Med. 126, 941. Ehrenreich, B. A., and Cohn, Z. A. (1968). J. Cell Biol. 38, 244. Ehrenreich, B. A., and Cohn, Z. A. (1969). J. Exp. Med. 129, 227. Fauci, A. S., Benacerraf, B., and Wolff, S. M. (1971). PTOC. SOC. E x p . Biol. Med. 136, 264. Feldman, J. D., and Unanue, E. R. (1971). Cell. Immunol. 2, 275. Feldman, M., and Palmer, J. (1971). Immunology 21, 685. Fischer, H., Ax, W., Freund-Molbert, E., Holub, M., Krusman, W. F., and Matthes, M. L. (1970). In “Mononuclear Phagocytes” ( R . Van Furth, ed.), p. 528. Davis, Philadelphia, Pennsylvania. Fisher, S. ( 1966). Immunology 11, 127. Fishman, M. (1961). 1. Exp. Med. 114, 837. Fishman, M., and Adler, F. L. (1963a). Proc. Int. Symp. Immunopathol., 3rd, 1962, p. 79. Fishman, M., and Adler, F. L. (196313). 1. Exp. Med. 117, 595. Fishman, M., and Adler, F. L. (1970). In “Mononuclear Phagocytes” ( R . Van Furth, ed. ), p. 581. Davis, Philadelphia, Pennsylvania. Ford, W. L., Gowans, J. L., and McCullagh, P. J. (1966). Thymus: Exp. Clin. Stud., Ciba Found. Symp., 1965, p. 58. Franzl, R. E. (1962). Nature (London) 195, 457. Franzl, R. E., and Morrello, J. A. (1966). J. ReticuEoendothel. SOC. 3, 351. Frei, P. C., Benacerraf, B., and Thorbecke, G . J. (1965). Proc. Nut. Acad. Sci. U.5. 53, 20. French, V. I., Stark, J. M., and White, R. G . (1970). Immunology 18, 645. Friedman, H. P., Stavitsky, A. B., and Solomon, J. M. (1965). Science 149, 1106. Gajl-Peczalska, K. J., Fish, A. J., Meuwissen, H. J., Frommel, D., and Good, R. A. (1969). J. E r p . Med. 130, 1367. Gallily, R., and Feldman, M. (1967). Immunology 12, 197. Garcia, D. A., Little, J. B., and Ourth, D. D. (1972). J. Immunol. (Submitted). Garvey, J. S., Campbell, D. H., and Das, M. L. (1967). J. E x p . Med. 125, 111.

160

E. R. UNANUE

Gay, F. P., and Clark, A. R. (1924). J. Amer. Med. Ass. 83, 1296. Gigli, I., and Nelson, R. A., Jr. (1968). E x p . Cell Res. 51, 45. Gill, F. A., and Cole, R. M. (1965). J. Immunol. 94, 898. Gillette, R. W., and Lance, E. M. ( 1971). J. Reticuloendothel. SOC. 10, 223. Golub, E. S., and Weigle, W. 0. (1969). J. Immunol. 102, 389. Goodman, J. W. ( 1972). In “Contemporary Topics in Immunochemistry” (F. P. Inman, ed.). Plenum, New York (in press). Gorczynski, R. M., Miller, R. G., and Phillips, R. A. (1971). J. E x p . Med. 134, 1201. Gordon, J. (1968). Proc. SOC.E x p . Biol. Med. 127, 30. Gottlieb, A. A. (1969). Biochemistly 8, 2111. Gottlieb, A. A,, and Doty, P. (1967). Proc. Nut. Acod. Sci. U . S . 57, 1849. Gottlieb, A. A., and Straus, D. S. (1969). J. Biol. Chem. 244, 3324. Graf, M. W., and Uhr, J. W. (1969). J. E x p . Med. 130, 1175. Greenwood, F. C., Hunter, W. M., and Clober, J. S. (1963). Biochem. J. 89, 114. Halpern, B. N., Biozzi, G., Benacerraf, B., and Stiffel, C. (1957). Amer. J . Physiol. 189, 520. Halpern, B. N., Biozzi, G., Pequignot, G., Delaloye, B., Stiffel, C., and Mouton, D. (1959). Pathol. Biol. 7, 1637. Hamaoka, T., Takatsu, K., and Kitagawa, M. (1971). Immunology 21, 259. Hanna, M. G., Jr., and Hunter, R. L. (1971). In “Morphological and Functional Aspects of Immunity” (K. Lindahl-Kiessling, G. Alm, and M. Hanna, eds. ), p. 257. Plenum, New York. Hanna, M. G., Jr., Francis, M. W., and Peters, L. C. (1968). Immunology 15, 75. Harris, G. (1965). Immunology 9, 529. Harris, G. (1967). Immunology 12, 159. Hartmann, K.-U., Dutton, R. W., McCarthy, M. M., and Mishell, R. I. (1970). Cell. Immunol. 1, 183. Haskill, J. S., Byrt, P. N., and Marbrook, J. (1970). J. Exp. Med. 131, 57. Heilman, D. H. (1965). Int. Arch. Allergy Appl. Immunol. 26, 63. Heilman, D. H., and Robert, C. B. (1967). J. Bacteriol. 93, 15. Heise, E. R., and Weiser, R. S. (1970). J. Immunol. 104, 704. Henson, P. M. (1969). Immunology 16, 107. Henson, P. M. (1971). J. E x p . Med. 134, No. 3, Part 2, 1145. Herd, Z. L., and Ada, G. L. (1969a). Aust. J . E x p . Biol. Med. Sci. 47, 63. Herd, Z. L., and Ada, G. L. (1969b). Aust. J. E x p . Biol. Med. Sci. 47, 73. Herscowitz, H. B., and Stelos, P. (1970a). J . Immunol. 105, 771. Herscowitz, H. B., and Stelos, P. (1970b). J. Immunol. 105, 779. Hersh, E. M., and Harris, J. E. (1968). J. Irnmunol. 100, 1184. Hirsch, M. S., Gary, G. W., and Murphy, F. A. (1969). J. Immunol. 102, 656. Hoffmann, M., and Dutton, R. W. (1971). Science 172, 1047. Howard, J. G. (1970). In “Mononuclear Phagocytes” ( R . Van Furth, ed.), p. 179. Davis, Philadelphia, Pennsylvania. Howard, J. G., and Benacerraf, B. (1966). Brit. J. E x p . Pathol. 47, 193. Howard, J. G., and Siskind, G. W. (1969). Clin. E x p . Immunol. 4, 29. Huber, H., and Fudenberg, H. H. (1968). Int. Arch. Allergy Appl. Immunol. 34, 18. Huber, H., Polley, M. J., Linscott, W. D., Fudenberg, H. H., and Miiller-Eberhard, H. J. ( 1968). Science 182, 1281. Huber, H., Douglas, S. D., and Fudenberg, H. H. (1969). Immunology 17, 7. Humphrey, J. H. (1955a). Brit. J. E x p . Pathol. 36, 268. Humphrey, J. H. (1955b). Brit. J. E x p . Pothol. 36, 283. Humphrey, J. H. (1960). In “Mechanisms of Antibody Formation” ( M . Holub and

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

161

L. Jaroskovi, eds.), p. 44. Academic Press, New York. Humphrey, J. H. (1969). Antibiot. Chemother. (Basel) 15, 7. Humphrey, J. H., and Frank, M. M. (1967). Immunology 13, 87. Humphrey, J. H., and Keller, H. U. (1970). In “Developmental Aspects of Antibody Formation and Structure” (J. Sterzl and I. m a , eds. ), 2nd ed.. Vol. 1, p. 181. Academic Press, New York. Humphrey, J. H., and Turk, J. L. (1963). Immunology 6, 119. Humphrey, J. H., Askonas, B. A,, Auzins, I., Schechter, I., and Sela, M. 1967 1. Immunology 13, 71. Inchley, C. J. (1969). Clin. E x p . Immunol. 5, 173. Inchley, C. J., Grey, H. M., and Uhr, J. W. ( 1970). J. Immunol. 1.05, 362. Isa, A. M. (1971). J. Immunol. 107, 595. Jacoby, F. (1965). In “Cells and Tissues in Culture” (E. N. Willmer, ed.), Vol. 2, p. 1. Academic Press, New York. Janeway, C. A., Jr., and Humphrey, J. H. (1969). Isr. J. Med. Sci. 5, 185. Jaroslow, B. M., and Nossal, G. J. V. (1966). Aust. J. Exp. Biol. Med. Sci. 44, 609. Jehn, U. W., Musher, D. M., and Weinstein, L. (1970). Proc. SOC. Exp. Biol. Med. 134, 241. Jenkin, C., and Benacerraf, B. (1960). J. Exp. Med. 112, 403. Jenkin, C. R., Auzins, I., and Reade, P. P. (1965). Aust. J. Exp. Biol. Med. Sci. 43, 607. Jennings, J. F., and Hughes, L. A. (1969). Nature (London) 221, 79. Jonas, W. E., Gurner, B. W., Nelson, D. S., and Coombs, R. R. A. (1965). Int. Arch. Allergy A p p l . Immunol. 28, 86. Jones, T. C., and Hirsch, J. G. (1971). J. Exp. Med. 133, 231. Kaplan, M. H., Coons, A. H., and Deane, H. W. (1950). J. Exp. Med. 91, 15. Karnovsky, M. L., and Karnovsky, M. J. (1971). Unpublished data. Katz, D. H., and Benacerraf, B. (1972). Adoan. Immunol. 15, 1 (this volume). Katz, D. H., and Unanue, E. R. ( 1971). Unpublished experiments. Kessel, R. W. I., and Braun, W. (1965). Aust. J. E xp. Biol. Med. Sci. 43, 511. Kessel, R. W. I., Monaco, L., and Marchiso, M. A. (1963). Brit. J. Exp. Pathol. 44, 351. Kolsch, E., and Mitchison, N. A. (1968). J. E x p . Med. 128, 1059. Kossard, S., and Nelson, D. S. (1968a). Aust. J. E xp. Biol. Med. Sci. 46, 51. Kossard, S., and Nelson, D. S. (1968b). Aust. J. E xp. Biol. Med. S c i . 46, 63. Lamvik, J. 0. (1969). S c a d J. Haemutol. 6, 149. Lance, E. M. (1970). Clin. Exp. Immunol. 6, 789. Lane, F. C., and Unanue, E. R. (1972). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 31, 610. Lang, P. G., and Ada, G. L. (1967). Immunology 13, 523. Lay, W. H., and Nussenzweig, V. (1968). J. E x p . Med. 128, 991. Lay, W. H., and Nussenzweig, V. (1969). J. Immunol. 102, 1172. Leduc, E. H., Arrameas, S., and Bouteille, M. (1968). J . Exp. Med. 127, 109. Lee, A., and Cooper, G . N. (1966). Aust. 1. Exp. Biol. Med. Sci. 44, 527. Leserman, L. D., and Roseman, J. M. ( 1971). Fed. Proc., Fed. Amer. SOC. E r p . Biol. 30, 589. Leskowitz, S. (1970). In “Developmental Aspects of Antibody Formation and Structure” (J. Sterzl and I. Riha, eds.), 2nd ed., Vol. 1, p. 165. Academic Press, New York. Levis, W. R., and Robbins, J. H. (1969). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 28, 566.

162

E. R. UNANUE

Lewis, L. A. (1954). Amer. J. Physiol. 179, 285. LoBuglio, A. F., Cotran, R. S., and Jandl, J. H. (1967). Science 158, 1582. McConahey, P. J., and Dixon, F. J. (1966). Int. Arch. Allergy Appl. Immunol. 29, 185, McConahey, P. J., Cerottini, J.-C., and Dixon, F. J. (1968). J. Exp. Med. 127, 1003. McCullagh, P. J. (1970). Aust. J. Exp. Biol. Med. Sci. 48, 551. McDevitt, H. O., and Benacerraf, B. (1969). Aduan. Immunol. 2, 31-74. McDevitt, H. O., Askonas, B. A., Humphrey, J. H., Schechter, I., and Sela, M. (1966). Immunology 11, 337. McFarland, W. (1969). Science 163, 818. McFarland, W., and Heilman, D. H. (1965). Nature (London) 205, 887. McFarland, W., Heilman, D. H., and Moorhead, J. F. (1966). J. E x p . Med. 124, 851. McIntyre, J. A,, Niblack, G. D., Prater, T. F. K., and LaVia, M. F. (1971). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 30, 649. Mackaness, G. B. (1960). J. Erp. Med. 112, 35. Mackaness, G. B., and Blanden, R. V. (1967). Progr. Allergy 11, 89. Marbrook, J. (1967). Loncet 2, 1279. Medlin, J., Humphrey, J. H., and Sela, M. (1970). Folia Biol. (Praque) 16, 156. Melnick, H. D., and Friedman, H. (1969). Proc. SOC. Exp. Biol. Med. 133, 423. Mergenhagen, S. E., Notkins, A. L., and Dougherty, S. F. (1967). J. Immunol. 99, 576. Metchnikoff, E. (1899). Ann. Inst. Pasteur, Paris 13, 47. Miller, H. R. P., and Avrameas, S . (1971). Nature (London) 229, 184. Miller, J. F. A. P., Basten, A., Sprent, J., and Cheers, C. (1971). Cell. Immunol. 2, 469. Miller, M. E., and Nillson, U. R. (1970). N . Engl. J. Med. 282, 354. Mims, C. A. (1959). Brit. J. Exp. Puthol. 40, 533. Mishell, R. I., and Dutton, R. W. (1966). Science 153, 1004. Mitchell, J., and Abbot, A. (1965). Nature (London) 208, 500. Mitchell, J., and Nossal, G. J. V. (1966). A&. 1. Exp. Biol. Med. Sci. 44, 211. Mitchison, N. A. (1965). Immunology 9, 129. Mitchison, N. A. (1969a). Isr. J. Med. Sci. 5, 230. Mitchison, N. A. (1969b). Immunology 16, 1. Morelli, R., and Rosenberg, L. T. (1971). J. Immunol. 107, 476. Mosier, D. E. (1967). Science 151, 1573. Mosier, D. E. (1969). J. Exp. Med. 129, 351. Mosier, D. E., RowIey, D. A,, and Davies, A. J. S . (1970). Nature (London) 255, 276. Muller-Hermelink, H. K., and Muller-Ruchholtz, W. ( 1971). In “Morphological and Functional Aspects of Immunity” (K. Lindahl-Kiessling, G . Alm, and M. C . Hanna, Jr., eds.), p. 451. Plenum, New York, Munder, P. G., Modolell, M., Ferber, E., and Fischer, H. (1970). In “Mononuclear Phagocytes” ( R . Van Furth, ed.), p. 445. Davis, Philadelphia, Pennsylvania. Naor, D., and Sulitzeanu, D. (1967). Nature (London) 214, 687. Nelson, D. S. (1963). Aduan. Immunol. 3, 131. Nettesheim, P., and Hammons, A. S. (1971). 1. Irnmunol. 107, 518. Nossal, G. J. V. (1967). Gamma Globulins, Nobel Symp., 3rd, 1967. Nossal, G. J, V., and Ada, G. L. (1971). “Antigens, Lymphoid Cells, and the Immune Response.” Academic Press, New York. Nossal, G. J. V., Ada, G. L., and Austin, C. M. (1964). Aust. J. Exp. Biol. Med. Sci. 42, 311.

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

163

Nossal, G. J. V., Ada, G. L., Austin, C. M., and Pye, J. (1965). Immunology 9, 349. Nossal, G. J. V., Austin, C. M., Pye, J., and Mitchell, J. (1966). Int. Arch. Allergy Appl. Immunol. 29, 368. Nossal, G. J. V., Abbot, A., and Mitchell, J. (1968a). J. Exp. Med. 127, 263. Nossal, G. J. V., Abbot, A., Mitchell, J., and Lummus, Z. (196813).J. Exp. Med. 127, 277. Notkins, A. L., Mergenhagen, S. E., Rizzo, A. A., Sheele, C., and Waldmann, T. A. ( 1966). 1. Exp. Med. 123, 347. Oort, J., and Turk, J. L. (1965). Brit. J. Exp. Pathol. 46, 147. Oppenheim, J. J., Hersh, E. M., and Block, J. B. (1966). In “Proceedings of the Symposium: The Biological Effects of Phytohemagglutinin” ( M. W. Elves, ed.). England. Oppenheim, J. J,, Leventhal, B. G., and Hersh, E. M. (1968). J. Immunol. 101, 262. Panijel, J., and Cayeux, P. (1968). Immunology 14, 769. Parkhouse, R. M. E., and Dutton, R. W. (1966). J. Immunol. 97, 663. Pearson, M. N., and Raffel, S. (1971). J. E x p . Med. 133, 494. Perkins, E. H. (1970). In “Mononuclear Phagocytes” (R. Van Furth, ed.), p. 35. Davis, Philadelphia, Pennsylvania. Perkins, E. H., and Leonard, M. R. ( 1963). J. Immunol. 90, 228. Perkins, E, H., and Makinodan, T. (1964). J. Immunol. 92, 192. Pernis, B., and Paronetto, F. (1962). Proc. SOC. E x p . Biol. Med. 110, 390. Phillips-Quagliata, J., Levine, B. B., and Uhr, J. W. (1969). Nature (London) 222, 1290. Phillips-Quagliata, J. M., Levine, B. B., Quagliata, F., and Uhr, J. W. (1971). J. Exp. Med. 133, 589. Pierce, C. W. (1969a). J. Exp. Med. 130, 345. Pierce, C. W. (1969b). J. Exp. Med. 130, 365. Pierce, C. W., and Benacerraf, B. (1969). Science 166, 1002. Pribnow, J. F., and Silverman, M. S. (1967). J. Immunol. 98, 225. Rabinovitch, M. (1967a). J. Immunol. 99, 232. Rabinovitch, M. (196713). J. Immunol. 99, 1115. Rabinovitch, M. (1968). Semin. Hemutol. 5, 134. Rhodes, J. M., Lind, I., Birch-Andersen, A., and Ram, H. (1969). Immunology 17, 445. Roelants, G. E., and Goodman, J. W. (1968). Biochemistry 7, 1432. Roelants, G. E., and Goodman, J. W. (1969). J. Exp. Med. 130, 557. Roelants, G. E., Goodman, J. W., and McDevitt, H. 0. ( 1971). 3. Immunol. 106, 1222. Roseman, J. M. (1969). Science 185, 1125. Roseman, J. M., Fitch, F. W., and Rowley, D. A. (1969). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 28, 814. Roser, B. (1965). Aust. 3. Biol. Med. Sci. 43, 553. Roser, B. (1968). J. Reticuloendothel. SOC. 5, 455. Russell, P., and Roser, B. (1966). Aust. J . Biol. Med. Sci. 44, 629. Sabet, T. W., and Friedman, H. (1969). Proc. SOC. Exp. Biol. Med. 131, 1317. Sabet, T., Newlin, C., and Friedman, H. (1969). Immunology 16, 433. Sabin, F. R. (1923). Bull. Johns Hopkins Hosp. 34, 277. Saha, A., Garvey, J. S., and Campbell, D. H. (1964). Arch. Biochem. Biophys. 105, 179. Schechter, G. P., and McFarland, W. (1970). J. Immunol. 105, 661.

164

E. R. UNANUE

Schmidtke, J. R., and Dixon, F. J. (1972). J. Immunol. (in press). Schmidtke, J. R., and Johnson, A. G. ( 1971). J. Immunol. 106, 1191. Schmidtke, J. R., and Unanue, E. R. (1971a). J. Immunol. 107, 331. Schmidtke, J. R., and Unanue, E. R. (1971b). Nature (London) 223, 84. Schoenberg, M. D., Mumaw, W. R., Moore, R. D., and Weisberger, A. S. (1964). Science 143, 964. Schwab, J. H., and Brown, R. R. (1968). J. Immunol. 101, 930. Schwab, J. H., and Ohanian, S. H. (1967). J. Bacten’ol. 94, 1346. Schwartz, H. J. ( 1971). Cited in Progr. Immunol. (1971) p. 1382. Schwartz, H. J., and Leskowitz, S. (1969). J. Immunol. 103, 87. Schwartz, R. S., Ryder, R. J. W., and Gottlieb, A. A. (1970). Progr. Allergy 14, 81. Seeger, R. C., and Oppenheim, J. J. (1970a). Fed. PTOC.,Fed. Amer. SOC. E x p . Biol. 30, 693. Seeger, R. C., and Oppenheim, J. J. (1970b). J. Exp. Med. 132, 44. Sela, M. (1969). Science 166, 1365. Sela, M., Schechter, B., Schechter, I., and Borek, F. (1967). Cold Spring Harbor Symp. Quant. Biol. 32, 537. Shin, H. S., Smith, M. R., and Wood, W. B. (1969). J. E x p . Med. 130, 1229. Shortman, K., and Palmer, J. ( 1971). Cell. Immunol. 2, 399. Shortman, K., Diener, E., Russell, P., and Armstrong, W. D. (1970). J. E x p . Med. 131, 461. Siege], I. (1970). J. Immunol. 105, 879. Spiegelherg, H. L., and Weigle, W. 0. (1967). J. Immunol. 98, 1020. Spiegelberg, H. L., Miescher, P. A., and Benacerraf, B. (1963). J. Immunol. 90, 751. Spitznagel, J. K., and Allison, A. C. (1970a). J. Immunol. 104, 119. Spitznagel, J. K., and Allison, A. C. (1970b). J. Immunol. 104, 128. Steinberg, S. U., Munro, J. A,, Fleming, W. A., French, V. I., Stark, J. M., and White, R. G. (1970). Immunology 18, 635. Stern, K., Spencer, K., and Farquar, M. (1955). Proc. SOC. Exp. Biol. Med. 89, 126. Sulitzeanu, D., Kleinman, R., Benezra, D., and Gery, I. (1971). Nature (London) 229, 254. Szakal, A. K., and Hanna, M. G., Jr. (1967). Exp. Mol. Pathol. 8, 75. Talmage, D. W., Radovich, J., and Hemmingsen, H. (1969). J. Allergy 43, 323. Theis, G. A,, and Thorbecke, G. J. (1970). J. Exp. Med. 131, 970. Thiery, J. P. ( 1962). J. Microsc. (Paris) 1, 275. Tsan, M., and Berlin, R. D. (1971). J. Erp. Med. 134, 1016. Uhr, J. W. (1965). Proc. Nut. Acad. Sci. U.S. 54, 1599. Uhr, J. W., and Weissman, G. (1965). J. Immunol. 94, 544. Uhr, J. W., and Weissman, G. ( 1968). J. Reticuloendothel. SOC. 5, 243. Unanue, E. R. (1968a). Unpublished results. Unanue, E. R. (1968b). Nature (London) 218, 36. Unanue, E. R. (1969). J. Immunol. 102, 893. Unanue, E. R. (1970). J. Immunol. 105, 1339. Unanue, E. R. ( 1972). J. Immunol. (in press ). Unanue, E. R., and Askonas, B. A. (1968a). 1. Erp. Med. 127, 915. Unanue, E. R.,and Askonas, B. A. (1968b). Immunology 15, 287. Unanue, E. R., and Cerottini, J.-C. (1970). J. Exp. Med. 131, 711. Unanue, E. R., and Cerottini, J.-C. (1971). I n “Developmental Aspects of Antibody Formation and Structure” (J. Sterzl and I. Riha, eds.), 2nd ed., Vol. 2, p. 521. Academic Press, New York.

ROLE OF MACROPHAGES IN ANTIGENIC STIMULATION

165

Unanue, E. R., and Feldman, J. D. ( 1971). Cell. Immunol. 2, 269. Unanue, E. R., Askonas, B. A., and Allison, A. C. (1969a). J. Immunol. 103, 71. Unanue, E. R., Cerottini, J.-C., and Bedford, M. (196913). Nature (London) 222, 1193. Unanue, E. R., Schmidtke, J., Cruchaud, A., and Grey, H. M. (1972). Proc. Int. Symp. Immunopathol., 6th, 1971 (in press). Valentine, M. D., Bloch, K. J., and Austen, K. F. (1967). J. Immunol. 99, 98. Vaughan, R. B. (1965). Immunology 8, 245. Vaughan, R. B., and Boyden, S. V. (1964). Immunology 7, 118. Virolainen, M., Lahti, A., and Hayry, P. (1972). Aduan. Exp. B i d . Med. (in press). Volkman, A. (1966). J. Exp. Med. 124, 241. Volkman, A,, and Gowans, J. L. (1965a). Brit. J. Exp. Pathol. 46, 50. Volkman, A., and Gowans, J. L. (196513). Brit. 1. Exp. Pathol. 46, 62. Walsh, T. E., and Smith, C. A. ( 1951). J. Immunol. 66, 303. Wardlaw, A. C., and Howard, J. G. (1959). Brit. J. Exp. Pathol. 40, 113. Warner, N. L., and Byrt, P. ( 1970). Nature (London) 226, 942. Weissman, G., and Dukor, P. (1970). Aduan. Immunol. 12, 283-331. Weissman, G., Dukor, P., and Zurier, R. B. (1971a). Nature (London) 231, 131. Weissman, G., Zurier, R. B., Spieler, P. J., and Goldstein, I. M. (1971b). J. Exp. Med. 134, No. 3, Part 2, 1495. White, R. G. (1963). Ciba Found. Study Group [Pap.] 16, 6. White, R. G., Coons, A. H., and Connolly, J. M. (1955). J. Exp. Med. 102, 73. White, R. G., French, I., and Stark, J. M. (1967). In “Germinal Centers in Immune Responses” (H. Cottier et al., eds.), p. 131. Springer-Verlag, Berlin and New York. Wiener, E., and Levanon, D. (1968). Lab. Znoest. 19, 584. Williams, G. M. (1966a). Immunology 11, 467. Williams, G. M. (1966b). Immunology 11, 475. Williams, G. M., and Ada, G. L. (1967). Immunology 13, 249. Wood, W. B., Jr. (1951). Harvey Lect. 47, 72. Wood, W. B., Jr., Smith, M. R., and Watson, B. (1946). J . Exp. Med. 84, 387. Zatz, M. M., and Lance, E. M. (1971). J . E x p . Med. 134, 224. Zilversmit, D. B., Boyd, G. A., and Brucer, M. (1952). J. Lab. Clin. Med. 40, 255.

Immunological Enhancement: A Study of Blocking Antibodies'

.

JOSEPH D FELDMAN Deparfmenf of Experimenfal Pafhology. Scripps Clinic and Research Foundation. La Jolla. California

I . History . . . . . . . . . I1. Definitions . . . . . . . . I11. Components . . . . . . . . A . Antibodies . . . . . . . . B. Antigens . . . . . . . . C . Host . . . . . . . . . IV . Immunological Enhancement . . . . A . Experimental Tumors . . . . . B. Autochthonous Tumors . . . . . C. Normal Tissue Grafts . . . . . D. Delayed Hypersensitivity and Autoimmunity . . . . . . V . Fetus as Homograft VI . Tolerance . . . . . . . . VII. Theories . . . . . . . . . A . Afferent Blockade . . . . . . B. Efferent Blockade . . . . . . C. Central Blockade . . . . . . D . Miscellany . . . . . . . E . Personal Preference . . . . . . VIII . Prospects . . . . . . . . . References . . . . . . . .

. .

. . . .

. . . . . . . . . . .

. . . .

. . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

.

. . . . . . . . . . .

. . . . . . . . . .

. . . . .

. . . . .

. . . .

. . . .

. . . .

.

.

. . . .

.

.

167 169 170 170 172 174 177 177 182 183 186 188 190 191 192 195 197 201 202 204 205

I. History

Immunological enhancement ( IE ) was for many years a laboratory curiosity. usually fatal. wherein transplanted tumors showed an unanticipated growth in rodents . In 1907. Flexner and Jobling (1907a.b ) described the operation of a tumor-promoting factor . When nonviable Jensen sarcoma was inoculated into rats some days before a challenge with viable tumor. the challenge inoculum. instead of regressing. grew progressively and often overwhelmed the recipient . Haaland and his colleagues confirmed this work using a different tumor in mice (Bashford et al., 1908; Haaland. 1910). Chambers and Scott (1924. 1926) tried to 'This is Publication No . 561 from the Department of Experimental Pathology. Scripps Clinic and Research Foundation. La Jolla. California. The work was supported by U . S . Public Health Service Grant AI-07007 and Atomic Energy Com: mission Contract AT( 04-3) .779 . 167

168

JOSEPH D. FELDMAN

isolate the tumor-promoting agent and were unsuccessful. During the 20 years following the original description by Flexner and Jobling, a small number of conflicting reports appeared on the subject of tumor susceptibility. Woglom ( 1929), in a critical evaluation of experimental tumorigenesis, concluded that hypersusceptibility to tumor transplants (i.e., IE) was a poorly delineated subject and evidence for its existence was unconvincing. Just 3 years later, Casey initiated a series of studies and established that this type of tumor promotion was an immunological phenomenon and a specific event, i.e., a neoplasm could not be actively enhanced by inoculation of tumors differing in source and type from the test neoplasm (for complete list of references, see Casey, 1932, 1933, 1941). Casey also introduced the word “enhancing” and later attributed the progressive growth to an XYZ factor (Casey et al., 1949). In 1946, Snell and his associates first related the phenomenon of IE to histocompatibility antigens ( H antigens), and this was confirmed by Casey et al. (1949). Within the ensuing decade, IE was shown to operate in extending the survival of normal tissue grafts (Billingham and Parkes, 1955; Billingham and Sparrow, 1955; Billingham et al., 1956). In this samc period, Kaliss explored the phenomena associated with IE and offered the first evidence that IE was mediated by antibodies (Kaliss and Molomut, 1952; Kaliss and Kandutsch, 1956). It became obvious that enhancement of a malignant neoplasm to the detriment of its host was similar to the enhancement of survival of normal tissue grafts and in each instance depended upon an immune response of the host. In the past 15 years the number of investigators and studies concerned with IE has proliferated logarithmically, and the concept of the process has been expanded to include not only progressive growth of transplanted tumor and prolonged survival of normal tissue grafts but also somc aspects of suppressed cellular immunity. The latter subsumes inhibition of delayed hypersensitivity ( D H ) (Flax and Waksman, 1962; Crowle and Hu, 1965, 1968, 1969; Axelrad, 1968; Axelrad and Rowley, 1968); inhibition of autoimmune diseases (Voisin and Kinsky, 1962; Chutnh and Rychlikovl, 1964; Voisin, 1965; Paterson, 1966; Chutnh and Pokornh, 1967; Voisin et al., 1968b; Kaliss, 1969; Buckley et al., 1971a); and the progression of autochthonous tumors to autonomy in both animals and people (Attia and Weiss, 1966; Gershon et al., 1968; K. E. Hellstroni and Hellstriim, 1970). The phrase “immunological enhancement” was primarily descriptive of certain facets of tumor growth in a grafted host. It does not encompass the wider spectrum of immunological processes recently attributed to it nor does it reflect the underlying immunological mechanisms. For these reasons I shall use immunological blockade ( I B ) almost syn-

IMMUNOLOGICAL ENHANCEMENT

189

onymously with immunological enhancement, despite the fact that the latter is already embedded in the language of common usage. In this review the components and putative mechanisms of IB will be considered and IB will be depicted as a nicely balanced interaction among antibody, antigen, and responding host. II. Definitions

Immunological blockade may be defined as the inhibition or delay by antibodies of an immune response to specific antigens. The key word here is antibodies, since they are the eventual mediators and modifiers of this complex process, and their detection is the sine 9ua non of IB. The definition separates IB from immunological unresponsiveness or tolerance. The latter is characterized by a suppression or abolition of antibody production. The former requires the demonstration and operation of antibodies. The distinction needs emphasis because a number of experiments concerned with the study of IB have been complicated by design and manipulations that might have induced unresponsiveness (Allen et al., 1952; Werder and Hardin, 1954; Billingham and Sparrow, 1955; Billingham et al., 1959; Stark and Dwyer, 1959; Hildemann and Walford, 1960; Cock, 1962b; Halasz, 1963; Medawar, 1963; Halasz et al., 1964; Snyder et al., 1964; Halasz and Orloff, 1965; Heslop, 1966; Nelken et al., 1966; Ruszkiewicz, 1967; R. E. Wilson et al., 1969, 1971; Wheeler et al., 1970; C. E. Zimmerman, 1971). In the opposite direction, some experiments designed to induce tolerance may have elicited enhancing antibodies that helped to prolong survival of grafts, particularly when histocompatibility differences were minor (Lapp and Bliss, 1966, 1970; Wachtel and Silvers, 1971). A distinction is also made here between enhancing antibodies and antilymphocyte antibodies, derived from allogeneic or xenogeneic sources. Antilymphocyte antibodies are capable of extending survival of allo- and xenografts (Mayer et al., 1965; Guttmann et al., 1967; Cerilli and Treat, 1969; Grob and Inderbitzen, 1969; Mandel and DeCosse, 1969; Ono et al., 1969; Taub, 1969; Hall-Allen et al., 1970; Jeekel and Westbrock, 1971; C. E. Zinimernian, 1971; Witz et d.,1971). They presumably achieve this effect in vivo because of their action on irnmunocompetent cells, particularly cells derived from the thymus (Brent et al., 1967; Denman and Frenkel, 1968; Martin and Millcr, 1968; Bachvaroff et al., 1969; Barth et al., 1969; Barth, 1969; Hardy et al., 1969; Schlesinger and Yron, 1969; Sloboda and Landes, 1970; Oliver and Feldman, 1971); but antilymphocyte antibodies may also bind to tissues other than lymphoid ( Witz et al., 1968, 1969~1,1971; Lindquist et aZ., 1969; Kamoun and Hamburger, 1970; Orr et al., 1970; Oliver and Feldman, 1971; C. B. Wilson, et al., 1971)

170

JOSEPH D. FELDMAN

and to antigens shared by lymphocytes and other cell types (Walford et al., 1965; Barth et al., 1969; M. Cohen and Nelken, 1970; Milgrom, 1970). It has not been determined whether these antigens are H antigens or other types of tissue antigens. Antibodies with enhancing activity, created by grafting allogeneic tissues into suitable recipients, certainly bind to lymphocytes of donor source (Walford et al., 1%). Their primary action, however, is not directed against lymphocytes nor lymphocyte antigens, but against H antigens. For these reasons the subject of antilymphocyte sera and their relation to IE will be excluded from this review. Antibodies that mediate IB may be produced by active immunization of the recipient prior to grafting, and they are generated by the injection of antigens that would usually induce a state of immunity and rejection. The antigens may be presented to the host by infusions of viable cells, by inoculation of dead tissue or tissue incorporated into adjuvants, by injection of tissue extracts or of a single class of relatively homogeneous proteins, such as serum albumins or globulins. Immunological blockade may also be accomplished by passive immunization, i.e., by the infusion of antisera or certain immunoglobulin fractions derived from these antisera into virgin recipients that will be or have been grafted. A rigorous proof of the operation of IB in an actively immunized host requires the demonstration that passive transfer of antibody will prolong survival of a homograft, promote tumor growth, and suppress or abolish cellular immunity. I l l . Components

A. ANTIBODIES Antibodies that mediate IB belong to the 7 S yG class. Only a small number of studies have examined this matter rigorously (Bubenik et al., 1965; Irvin et al., 1967; Brunner et aZ., 1968; Takasugi and Hildemann, 1969a,b,c; B. Zimmerman and Feldman, 1969a, 1970), and the evidence at this time seems to exclude other immunoglobulin classes. Of the subclasses of yG, the most exacting work has implicated yG, in the mouse (Irvin et al., 1967; Tokuda and McEntee, 1967; Batchelor, 1968; Cruse, 1969a; Takasugi and Hildeniann, 1969a,b; Gillespie et al., 1971) and its homolog in the guinea pig (Nelson, 1962; Broder and Whitehouse, 1968) and the rat (B. Zimmerman and Feldman, 1969a, 1970). The yG2 is a complement-fixing immunoglobulin and is the slowest moving, most cathodal of the immunoglobulins in electrophoretic fields. A few investigators have stated that 7 S yG, or a fast moving 7 S yG was responsible for IB (Voisin et al., 1966, 1969; Chard, 1968; Crowle and Hu, 1969), but

IMMUNOLOGICAL ENHANCEMENT

171

their methods for separating fast and slowly moving yG were not sensitive enough to achieve a clean division. Infrequently, yM has been suggested as a blocking agent (Bubenik and Koldovsky, 1965; Bubenik et al., 1965; Halek et al., 1968) or its role could not be assessed (Irvin et al., 1967; Brunner et al., 1968). Voisin has suggested that yA may also mediate enhancement ( Voisin et al., 1969). The evidence, however, is overwhelmingly in favor of yG. Since antibodies with blocking capacity have been considered to be mediating agents in IE, it was reasonable to test the ability of light chains to operate in IB. Both Fab’ and F ( ~ i b ’ fragments )~ of yG antibodies have been successfully used to promote growth of tumors in mice (Chard et al., 1967; Batchelor, 1968; Broder and Whitehouse, 1968; Chard, 1968; Whitehouse and Broder, 1968). These fragments of immunoglobulin molecules contain the antibody-combining site specific for its particular antigen. Presumably, Fab’ and F( ab’), derived from immunoglobulins other than yG should be capable of blocking antigenic sites, but to date the polypeptide light chains of yM, yA, and other immunoglobulins have not been tested for their enhancement capacity. yG Immunoglobulins with enhancing activity have also been associated with hemagglutinating and cytotoxic activities. Indeed, the association has been so constant that in the recent past many investigators were satisfied to demonstrate hemagglutinating and cytotoxic qualities in the circulation of enhanced hosts and to conclude from this the existence of enhancing antibodies. The association of these measurable activities, however, is only one way; that is to say, immunoglobulins with enhancing activity display hemagglutinating and cytotoxic qualities, but the reverse is not true, i.e., there are classes of immunoglobulins capable of hemagglutinating and of killing target cells that lack any enhancing qualities. For example, yM antibodies are generally potent cytotoxic agents, more so than yG antibodies, probably because yM has ten specific binding sites for antigen and five Fc chains, any two of which will bind complement ( C ) ( S. Cohen and Milstein, 1967; S. Cohen, 1968). There is little firm evidence that yM antibodies are capable of IB. In future work it will certainly help to measure the potency of blocking antibodies by means other than by hemagglutination and cytotoxicity. In fact, it is not known if blocking as well as hemagglutinating and cytotoxic activities all reside in one moleculc or reflect separate minority populations within a larger group of antibody molecules. Blocking antibodies exhibit a paradoxical effect which is undoubtedly related to their ability to fix and activate C. Both in wivo and in vitro smaller doses of antibody promote tumor growth or protect tumor cells from injury, whereas larger amounts of the same antibody either inhibit

172

JOSEPH D. FELDMAN

tumor progression or destroy target tumor cells (Gorer and Kaliss, 1959; Batchelor and Silverman, 1962; Boyse et al., 1962; Bubenik et al., 1965; Crowle and Hu, 1965; Hutchin et al., 1967; Buckley et al., 1971a). In either case, whether blockade or destruction is achieved, the amount of antibody is minute, probably in the range of nanograms. The opposing effects depend upon antigenic density over the surface of cells, the ratio of antibody molecules to antigenic sites, and the ability of antibody to activate c. Passive transfer of lymphoid cells sensitized to tumor or H- antigens may effect enhancement of tumor and skin grafts (Mitchison and Dube, 1955; Cruse et al., 1965; Hutchin et al., 1967; Miiller, 1967; Lapp and Mijller, 1969; Irvin and Eustace, 1970, 1971; Irvin, 1972). The subject of cellular immunity and host participation in IB will be discussed below, but it is emphasized here that a population of sensitized cells contains both cells that produce conventional circulating antibody (G. Moller, 1964b; Irvin and Eustace, 1970, 1971; Irvin, 1972) and cells that react directly and specifically with antigen. In view of the fact that minute quantities of antibody may mediate passive IB, it is extremely difficult to evaluate and quantitate the contribution made to IB by the fraction of cells that produce conventional antibody and by the fraction of specifically reactive sensitized cells, when a population of lymphoid elements from an immune donor is passively transferred to a naive recipient. Recently, Cerottini et al. (1970a,b, 1971) have elaborated a method to separate alloantibody-producing lymphocytes from cytotoxic lymphocytes. It should be possible, now, to assess the role of alloantibodies and that of effector lymphocytes in IE.

B. ANTIGENS When promotion of tumor growth under the conditions described above was discovered to belong to the province of immunology, a number of experiments were devised to determine the antigens that were involved, and they revealed that H antigens, particularly those of the H-2 locus, were the ones that elicited enhancing antisera (Snell et al., 1946; Casey et al., 1949; Snell, 1954, 1957; Brent and Medawar, 1962; E. MGller, 1963; G. Mijller, 1964a, 1966; Corson, 1970; C. E. Z’immerman, 1971) Histocompatibility antigens induce transplantation immunity and distinguish the individual within an outbred species, as among humans, or the individuals of an inbred strain from those of another strain within the species, as among mice (Snell and Bunker, 1965; Snell, 1957; Snell and Stimpfling, 1966). Considerable effort has been and is being exerted to isolate and analyze chemically H-2 antigens (Reisfeld and Kahan, 1970). Complete chemical identification has not yet been achieved. It

.

IMMUNOLOGICAL ENHANCEMENT

173

is the fond hope of many investigators to accomplish this, since H antigens lie at the root of graft rejection or acceptance. It is not known if all or only some of the H antigens will produce sera with blocking activity, but minor H antigens, i.e., non-H-2 antigens, may elicit IB as well as H-2 antigens do (White and Hildemann, 1968; Ceppellini et al., 1969; Thoenes et al., 1970; Hildemann, 1971; Mullen and Hildemann, 1971), and in one instance a single H-antigen difference between donor and recipient permitted the expression of IB (Winn, 1970). Histocompatibility antigens are located on or in cell surface membranes (Boyse et al., 1968a; Aoki et al., 1969, 1970; Kourilsky et al., 1971) and, perhaps, constitute part of some intracellular membranes ( Kandutsch, 1960). Antigens other than H antigens are also capable of inducing IB. Tumor-specific antigens are one class of these (Prehn, 1959) and may serve to effect IB (Gorer, 1942; Mitchison and Dube, 1955; H. N. Green and Wilson, 1956; Gorer, 1961; G. Moller, 1963d; Berne, 1965; Bubenik et al., 1965; E. Moller, 196513; Ruhkiewicz, 1967; Ferrer, 1968a; Kaliss and Suter, 1968; Cruse, 1969a,b; Bloom and Hildemann, 1970; K. E. HelIstr6m, and Hellstrom, 1970; Jehn et al., 1970); these antigens may be induced by viral infection or by chemical carcinogens, or they may be autochthonous, i.e., we are ignorant of their source. As with H antigens, tumor-specific antigens are located on cell surface membranes, and there is suggestive evidence that they may be linked to or closely associated with H antigens. Organ-specific antigens may also bring about IB, although the solubilized antigens that were used to influence graft survival were not rigorously characterized as to composition or even specificity (Jankovid and Flax, 1963; Chutnl and Rychlikovi, 1964; Berne, 1965; Paterson, 1966; Chutni and Pokorni, 1967; Nelken and Cohen, 1968; R. E. Wilson et al., 1969, 1971; M. Cohen and Nelken, 1970; Hall-Allen et al., 1970; Richardson and Paterson, 1970; Buckley et al., 1971a). If one accepts that inhibition of the lesions of cellular immunity, exemplified by DH, is encompassed within the expanded concept of IB, then a wide variety of immunogens, with no relation to histocompatibility, tumor, or tissue antigens, may be used to produce enhancing antibodies (Flax and Waksman, 1962; Asherson and Stone, 1965; Crowle and Hu, 1965, 1968, 1969; Axelrad, 1968; Axelrad and Rowley, 1968; Kaliss, 1969, 1970). The way in which antigens are presented to the living host may determine whether IB will develop or not. Viable tissue is presumed to be more effective than nonviable; the route of injection (intradermal, intraperitoneal, subcutaneous, intravenous ) of antigens may be influential; and administration of antigen with or without adjuvant may affect

174

JOSEPH D. FELDMAN

the outcome of IB ( h e l l , 1957; Gorer, 1958, 1961; Kaliss, 1958, 1962, 1968, 1969; G. Moller and Moller, 1962, 1966; Batchelor, 1963; E. Moller et al., 1968). In addition to these variables, the immunogenicity of different tissues and organs varies considerably. For example, skin is more immunogenic than kidney or ovarian grafts (Billingham and Parkes, 1955; 0. E. A. Linder, 1961; Moseley et al., 1966; Feldman et al., 1968; Parkes, 1968; White and Hildemann, 1969; Mullen and Hildemann, 1971). Finally, the type of graft that is applied to a host will affect the development, rate of appearance, and duration of IB, i.e., organs grafted to recipients by vascular anastomosis have a better chance of creating the conditions for IB and for extended survival than do tissues such as skin that heal-in by capillary anastomosis and are more likely to enjoy prolonged survival as a result of blocking antibodies. A shibboleth that has been handed down from article to subsequent article is that all the antigens of grafted target tissues must be covered by antibodies in order to effect IB. Dissenting reports indicate that this dictum is not true (Ceppellini et al., 1969; B. Zimmeiman and Feldman, 1969b; Brunner, 1970). Brunner and his co-workers, in particular, have clearly demonstrated that only some antigens need be blocked to effect blockade (Brunner, 1970; Mauel et al., 1970). This minor aspect of the number of antigens involved in IB is emphasized here because, as will be discussed later, antigenic density and spatial arrangement of antigen on cell surfaces play a major role in the phenomena of effective blockade. C. HOST The host provides an integral ingredient to the process of IB, the nature of which is obscure. To extend the viability of skin homografts in adults that differ at the major histocompatibility locus (H-2 in mice, AgB in rats, HL-A in man) for more than a few days beyond a normal survival period of 7 to 10 days has been difficult and often unattainable. In neonatal rats, however, survival of skin homografts may continue for weeks in a fraction of grafted recipients; the younger the recipient neonates the more prolonged the viability (Heslop, 1969, 1971; B. Zimmerman and Feldman, 1969a). Also, in adults that have been irradiated with sublethal doses of X-rays, have been thymectomized, or have been manipulated with immunosuppressive reagents, skin graft survival may be prolonged days and weeks beyond the usual time of rejection (Winn, 1970; Forbes et al., 1971; Jones and Feldman, 1971a; Jeekel et d., 1972). These observations imply that the immunological apparatus of the recipient must be poorly developed, deficient, or interfered with, in order that homografts, with the aid of blocking antibodies, may survive for lengthy periods.

IMMUNOLOGICAL ENHANCEMENT

175

Peter and Feldman (1972) measured cell mediated immunity in grafted rats that were infused with blocking alloantibodies or normal rat serum, and in neonatal rats, cyclophosphamide treated rats or neonatally thymectomized adult rats bearing skin allografts. In the latter animals the calculated number of cytotoxic lymphocytes, as determined by 51Cr release from target cells and by use of the one-hit theory, was reduced to 50% or less of the number measured in intact hosts. In rats infused with blocking antibodies there was also a diminution of cytotoxic lymphocytes. However, peak activity of effector lymphocytes reached control levels 4 days later than the peak activity found in control rats, i.e., blocking antibodies deferred the full development of cell mediated immunity by some days. Immunological deficiency may be effected in ways other than those noted above. Virus infection, for example, may alter humoral and cellassociated immune responses in recipients of homografts so that enhancement becomes manifest (Howard et d., 1969; Blair et al., 1971; Weiner et al., 1971). Precisely how viral infection of a host disturbs the immune response is still the subject of much study (Ceglowski and Friedman, 1968). Inhibition may result from the cytopathic effects of virus proliferation with simple reduction of the number of immunocompetent units or from more subtle effects on intracellular synthetic processes. Moreover, virus as antigen may act as a competitive antigen and thereby interfere with an immune response to grafts. Antigenic competition is also a subject sub iudice, but there is one report that indicates that multiple antigens injected into a recipient in certain doses and in exacting temporal sequences can alter the normal function of lymphoid tissues (Eidinger et aZ., 1968). Under these circumstances IB may be detected. The genetic composition of the host may also influence the development of blocking antibodies. The immune response is linked in I some fashion to the histocompatibility genes, i.e., whether an animal responds to an antigen or not depends upon his genic composition (McDevitt and Tyan, 1968; Cudkowicz and Bennett, 1969; Gasser, 1969; Ellman et al., 1970; Green et al., 1970; Rathbun and Hildemann, 1970; Vaz and Levine, 1970; Bailey and Hoste, 1971; Gasser and Silvers, 1971). Some claim that the genetic determination of an immune response resides in the process of synthesizing immunoglobulins ( Biozzi et al., 1970); others propose that genes may control antigenic recognition ( McDevitt and Benacerraf, 1969). Beyond the relation of gene structure to response and nonresponse is the connection between H-2 locus and the classes of antibodies produced in response to antigenic stimulation, i.e., mice or guinea pigs of one strain may synthesize antibodies of one class,

176

JOSEPH D. FELDMAN

for example, yGz, and not another, for example, yGl, whereas mice of other strains form all classes of antibodies (Coe, 1966; Reif and Allen, 1966; McDevitt and Chinitz, 1969; Ellman et al., 1970). It is obvious, then, that genetic structure is associated with IB through its control and determination of immunological reactions ( LengerovP and Matousek, 1968). Within this area of genetics and possibly related to the development of blocking antibodies is the recent hypothesis put forward by Ford and Simonsen (1971) on the factor of immunization. They postulate that the number of cells reactive to AgB-determined cellular antigens is high, while the number of cells reactive to non-AgB-determined cellular antigens is low. The latter group can be augmented 10- to 50-fold by immunization with homologous cells and tissues; the former group is already maximally represented and cannot be increased very much. They reported that alloantibody did not alter graft-versus-host reactions between AgB-different hosts, presumably because a large number of cells were already reactive to homologous tissue antigens. This hypothesis might explain why it is difficult to enhance grafts exchanged between parental lines in which the genetic difference is maximal and the number of responding units is maximal; why repeated doses of blocking antibodies are of little avail; why recipients already immunized are not readily “enhanced by blocking antibodies; and why animals that differ at non-H-2 or non-AgB loci are more easily enhanced. What is important is that genetic composition determines response, type of response, and number of responding units. The host is the source of effector cells that react with foreign antigens of grafts or with soluble antigens that are capable of inducing DH. These aggressor cells, most likely thymus derived lymphocytes ( Cerottini et al., 1970a, 1971; Henney, 1970, 1971; Henney and Nordin, 1971; Canty and Wunderlich, 1970), are highly specific in their killer activities, do not generate lymphokines, and are unaffected by antisera directed to immunoglobulin molecules. Other lymphocytes, presumably derived from bone marrow, may also exhibit cytotoxic activities in the presence of antibody specific for membrane associated antigens ( MacLennan and Loewi, 1968; MacLennan and Harding, 1970; Harding et al., 1971; Perlmann and Holm, 1969; Perlmann and Perlmann, 1970; Perlmann et al., 1970, 1972). The cells need not be derived from immune animals, nor is C needed for lysis of target cells. The antibody is IgG, may be xenogeneic or allogeneic in origin, and is effective with nonimmune lymphocytes at a concentration considerably below the concentration needed for the antibody, with C, to lyse targets in the absence of lymphocytes. There may also be a third category of lymphocyte that can be activated

IMMUNOLOGICAL ENHANCEMENT

177

to aggressive action by phytohemagglutinin, by mitogens and by antigens (Perlmann and Holm, 1969; Berke et al., 1969; Lonai and Feldman, 1970; Sabbadini, 1970; Chapius and Brunner, 1971; Ginsburg et al., 1971; Sin et al., 1971). This kind of lymphocyte, when activated, attacks target cells nonspecifically and seems not to generate mediators. However, Ruddle and Waksman (1968) and Elkins (1966, 1970) have described aggressor lymphocytes that are of this category and produce toxic substances. Macrophages accumulate in relatively large numbers at sites of cell mediated injury, e.g., in graft rejection, DH, tumor cell destruction, etc. The phagocytic activities of these cells, i.e., their aggressive action, appears to be enhanced by sensitized lymphocytes specifically interacting with antigen (Mackaness, 1969). Keller and Jones (1971) reported that IgG2 from rats infected with parasites inhibited the aggressive activities of peritoneal macrophages against tumor cells, i.e., the tumor was enhanced by inhibition of the phagocytic function of macrophages. Besides the contributions of the host described above, there are other more nebulous factors that disturb the internal milieu of an organism, for example, illness, starvation, stress, and these may also affect the acceptance or destruction of a homograft and the development of cellular immunity. IV. Immunological Enhancement

A. EXPERIMENTAL TUMORS The phenomena associated with IE were first detected by observing the behavior of transplanted tumors in pretreated recipients. The usual procedure was to inoculate a naive recipient with nonviable neoplastic tissue or small amounts of viable tumor insufficient to kill the grafted host. After several weeks to months, viable tissue from the same neoplasm was implanted into the pretreated recipient. The second inoculum would then display an increased rate of growth, as measured by the time of its gross appearance or by its size, or an augmented capacity to kill the host, as measured by mortality. When untreated controls were inoculated with similar amounts of viable neoplasm, the recipients would eventually destroy the tumor and survive. After Casey and his colleagues established the immunological specificity of this seemingly simple event (Casey, 1941), Snell et aZ. ( 1946, 1948), developed laboratory models that reproduced the effect consistently and demonstrated the necessary relationship between tumor acceptance or rejection and the genetic makeup of the host (Snell, 1956,

178

JOSEPH D. FELDMAN

1957). In order for IE to operate there must be a histocompatibility difference between the neoplasm and its host. After IE could be consistently reproduced in the laboratory, investigators explored a variety of biological factors that might influence the phenomenon. Sarcomas and carcinomas were thought to be more readily enhanced than leukemias (Gorcr, 1956; Snell, 1957; Gorer and Kaliss, 1959). Several years later, when suitable proportions of tumor cells and antibodies were prepared in mixtures, leukemias were also enhanced (Boyse et al., 1962; Chouroulinkov et al., 1962; Phillips and Stetson, 1962). The interpretation of this result was based on the number of antigenic receptors on the neoplastic cells available to antibodies in enhancing sera. An exacting balance was needed between numbers of antigens and numbers of antibodies of the right immunoglobulin class (G. Moller and Moller, 1962, 1966; G. Moller, 1963d; Uhr and MSller, 1968). Sex was considered to influence IE, which was more easily achieved in males than in females ( Kaliss, 1957; Gorer and Kaliss, 1959). Female mice respond more vigorously to immunogenic stimuli and, presumably, their augmented response was inimical to survival of transplanted tumors. However, an actual documentation of the qualities and quantities of the different antibody classes elicited by neoplasms in males and females has not been provided to explain the above observation. Before 1929 the spleen was considered to play a significant role in tumorigenesis ( see Wogloni, 1929, for references). Its contribution to IE of neoplasms was assessed by E. Moller (1965b) by Ferrer (1968a,b), and by I. Hellstrom, et al. (1970b). Growth of challenge leukemia several weeks after pretreatment of mice with the same tumor was diminished in splenectomized animals as compared with intact mice. However, other types of malignancies were unaffected by splenectomy. Sera from splenectomized mice contained lower titers of antibody (E. Moller, 1965b; I. Hellstrijm et al., 1970b) and were less efficacious in passive transfer of IE ( Ferrer, 1968b). Again, quantitative studies of antibody classes in these experiments were not carried out. Several investigators attempted to discriminate by morphological examination between the reaction engendered by enhanced and nonenhanced neoplasms. Histological scrutiny of the rejection process disclosed no obvious difference in cellular response ( Gorer, 1958; Batchelor and Silverman, 1962; Chantler, 1967), at least during the first 7 days after inoculation. Nor were there any differences in spleen size and weight in mice or rats bearing enhanced and nonenhanced tumors (Gorer, 1942; H. N. Green and Wilson, 1956). Cruse (1969a) suggested that C might affect IE. He reported that if inactive C8 was fixed to tumor cells, the inactive C component blocked C lysis and, thcrefore, permitted enhancement of

IMMUNOLOGICAL ENHANCEMENT

179

inoculated tumor. The role of C in graft rejection or acceptance is unclear, since one may find evidence that C has little or nothing to do with the “take” of a graft or its rejection and also that a deficiency of C may delay destruction of a graft (Guiney et al., 1964; Volk et al., 1964; Caren and Rosenberg, 1965; Crisler and Frank, 1965; Coppola and Villegas, 1967; Gewurz et al., 1967; K. Rother et al., 1967; U . Rother et al., 1967; Phillips et al., 1968; Villegas and Coppola, 1968). It is unlikely that inactivated C8 and its fixation to tumor cells can explain all the phenomena associated with IB. Also, several investigators showed that it was easier to enhance tumor growth and spread than to prolong survival of normal tissue grafts ( Berne, 1965; Rukzkiewicz, 1967). Suggested reasons for this are offered below. Most of the observations briefly presented above are peripheral to the core of the problem. The major agents that affect IE are antibodies and sensitized cells. A good deal of effort was expended to understand why “early” sera, i.e., sera removed from animals within a few days after tumor inoculation, did not possess enhancing qualities and, indeed, were toxic to tumor cells, whereas “late” sera, removed 15-100 days after tumor transplantation and even after tumor disappearance, did exhibit enhancing activity (Kaliss, 1956, 1958, 1969; Chantler and Batchelor, 1964; Cruse et al., 1965; Bloom and Hildemann, 1970). When methods were elaborated to separate clearly and quantitatively different immunoglobulin classes and when it was learned that yG antibodies generally appear late after immunization, whereas yM antibodies are readily detected within hours to days after primary antigenic stimulation, the enigma about early and late enhancing antibodies was solved. Since most, if not all, blocking antibodies belong to the yG, class, one would expect to find blocking activity in late antisera. Less easy to explain was the paradox of antibody quantities. Relatively large amounts of antisera, between 0.5 to 5 p1, were frequently toxic to tumor cells in vitro (Boyse et al., 1962; G. Moller, 1963a; Hutchin et al., 1967) and increased the resistance of the host in vivo to challenge neoplasms (Gorer and Kaliss, 1959; Boyse et al., 1962; Hutchin et al., 1967; Amos et al., 1968; Haughton and Nash, 1969). By contrast, minute amounts of antibody, 0.1-0.5 pl, displayed the opposite effect, i.e., they were enhancing. A reasonable explanation of this will be offered in a discussion of the interplay among number and distribution of cellular antigens, non-complement-fixing antibodies, and cytotoxic lymphocytes resulting in IE (Section VII). The other major element that affects IE of tumors is the activity of effector lymphocytes. Certain populations of lymphocytes, thought to be derived froin thymus cells (Mason and Warner, 1970; Canty and

180

JOSEPH D. FELBMAN

Wunderlich, 1970; Cerottini et al., 1971); are capable of being sensitized by antigens, particularly by cellular antigens, e.g., H-2, tumor-specific, and tissue-specific. Such sensitized lymphocytes can destroy target cells bearing the specific antigens, in the absence of C and of detectable conventional-type antibodies. Blocking antibodies interfere with the cytotoxic action of effector lymphocytes. In the context of tumor enhancement, effector lymphocytes can be inhibited by blocking antibodies from killing target neoplastic cells in uitro (G. Moller, 196313; K. E. Hellstrom and Hellstrom, 1970). In uiz)o, blocking antibodies may act synergistically with effector lymphocytes and augment a host’s resistance to a growing neoplasm, or they may act antagonistically and enhance the progression of tumor (G. Moller, 1963b; Batchelor and Howard, 1965; Cruse et al., 1965; Amos et al., 1970; Forbes et al., 1971). However, once lymphocytes have been sensitized by immunization of the host, blocking antibodies may not inhibit their cytotoxic activity (Klein and Sjogren, 1960; Brondz, 1965). The situation is further confused by the behavior of early and late lymphocytes removed from immunized hosts. Early and late here refer to the time that has elapsed between the inoculation of a tumor and removal of lymph node cells, and this varies from days to weeks according to the tumor model used. Generally, lymphocytes tested within days after tumor inoculation destroy the target neoplasm and inhibit its progression in viuo (Bubenik et al., 1965; Cruse et al., 1965; Hutchin et al., 1967; Bloom and Hildemann, 1970; Irvin and Eustace, 1970, 1971). By contrast, late lymphocytes help to promote tumor growth or are inactive. Muller’s observations (1967) have been the opposite. He reported that peritoneal exudate cells removed 10 days after immunization of mice enhanced challenge tumors, and those removed 17 days after immunization retarded challenge tumors. The dynamic balance among changing surface antigens, varying proportions of yG2 and other immunoglobulin antibodies, and effector lymphocytes determines whether or not a neoplasm will be destroyed or enhanced. Kaliss has emphasized that both resistance and enhancement may exist in animals prepared for active enhancement of a neoplasm and that enhancement cannot occur in the absence of an immune response (Kaliss, 1958, 1962, 1966, 1969; Kaliss and Bryant, 1958; Kaliss and Rubinstein, 1968). Others have substantiated this fact ( M. Feldman and Globerson, 1980; Hgskovh and Svoboda, 1962; Chantler and Batchelor, 1964; Chantler, 1967). In tumor systems, resistance to neoplastic growth may be detected prior to the appearance of enhancement (Mitchison and Dube, 1955; HiIskovA and Svoboda, 1962; Chantler and Batchelor, 1964; Kaliss and Rubinstein, 1968).

IMMUNOLOGICAL ENHANCEMENT

181

Several hypotheses have been invoked to explain this paradox, and they advocate either a physiological, antigenic, or metabolic alteration of the graft. Such alterations imply a change in the character of the primary neoplasm to the extent that the host does not respond in the same fashion that it would to an unaltered tumor. A physiological difference between enhanced and nonenhanced tumors has not been measured despite frequent proposals (Kaliss, 1958, 1962; Kaliss and Bryant, 1958; Gorer and Kaliss, 1959). Antigenic alteration has been suggested to occur in skins of tumor-bearing mice (Kaliss and Suter, 1968) and in tumors that have been transplanted to H-2 incompatible recipients ( E . Moller, 1964, 1965b). However, the majority of investigators have shown a constant and reliable behavior of serially inoculated tumors in several generations of recipients (Gorer, 1961; Batchelor, 1963; Hgskovti et al., 1965; E. Moller et al., 1968; Uhr and Moller, 1968; Kaliss, 1969). A putative metabolic change in enhanced tumor grafts was suggested by Hutchin et al. (1967). These investigators did not show any relationship between the metabolic disturbances and IE. Nor did they rule out the possibility that the enzymatic alterations within tumor cells reacting with antibody might also have occurred from injury with nonimmunological agents. In brief, there is no convincing evidence that a tumor is altered by serial passage or by enhancing antibodies. Gorer (1958) suggested that an enhanced neoplasm is stimulated by antibody. He based this on histological observations that revealed mitotic activity and explosive growth in the enhanced tumor after 7 days’ residence in the recipient. There are no data, however, to corroborate the fact that antibody stimulates tumor growth. The increased mitotic activity and cellular proliferation noted by Gorer might well have been a response of neoplastic cells to necrosis that occurred in the tumor. The paradoxes of the behavior of early and late effector lymphocytes and of the simultaneous existence of both enhancement and resistance in the same host will be considered further in Section VII. There are both advantages and disadvantages in the use of tumors for studying IE. The advantages are twofold: (1) the sensitivity of the system is high and the operation of IE may be easily and convincingly established; ( 2 ) it is important to have model systems in laboratory animals to clarify some facets of autochthonous tumor progression in humans. The disadvantages are also twofold. ( I ) There is no satisfactory measurement of the effect an expanding neoplasm has upon its host and particularly upon the immunological apparatus and cellular immunity. There are many reports indicating that progressing tumors may repress or alter an immune response (Prehn, 1963; Alexander, 1970).

182

JOSEPH D. FELDMAN

( 2 ) Nor is there any satisfactory way to quantitate the amount of antigen that is needed to effect IB when the antigen is replicating and mobile. There are, then, two uncontrolled variables that complicate analysis of IE of tumors. It is helpful that the phenomena associated with blocking antibodies occur in other non-neoplastic systems, e.g., in the survival of normal tissue grafts or in suppression of DH. B. AUTOCHTHONOUS TUMORS In the circulation of people with carcinoma of the large bowel, with neuroblastoma, and with other malignancies, the Hellstrijms and their co-workers have detected an immunoglobulin with the characteristics of antibody. It is present in the blood when the neoplasm is progressing and cannot be detected after removal of the malignancy or late in the disease (I. Hellstrom et al., 1968c, 1970a, 1971a,b; K. E. Hellstrom et al., 1969). I. Hellstriim and Sjogren (1965; I. Hellstrom 1967) developed a method to detect the antibody. The tumor, or portions of it, is extirpated and explanted. After establishment of growth in petri dishes, a small number of cells in gel is plated on slides. Layered above them are lymphocytes alone, serum alone, and serum plus lymphocytes from the patient in whom the tumor originated. After 7 to 10 days of cultivation, the numbers of tumor cell colonies, in the absence of the patient’s lymphocytes and serum, are counted and compared with the number that developed in the presence of the patient’s lymphocytes, which arc presumably sensitized to specific tumor antigens, are toxic to the tumor cells, and inhibit the development of colonies (I. Hellstrom et al., 1968a,b,c). In the presence of the patient’s serum the inhibitory action of the sensitized lymphocytes is abrogated or diminished, and this difference reflects the capacity of antibody to inhibit cytotoxicity of lymphocytes. In brief, antibodies enhance in vitro growth of autochthonous neoplasms by antagonizing the destructive action of lymphocytes. Tumors from one source, e.g., gastrointestinal tract, may be enhanced by sera from other patients with the same kind of neoplasm. Similar results have been obtained in dealing with spontaneous animal tumors or tumors provoked by oncogenic viruses (I. Hellstrom et al., 1968c, 1969, 1970b; K. E. Hellstrom and Hellstrom, 1969; Heppner, 1969; I. Hellstriim and Hellstrom, 1970) or by carcinogenic chemicals (Kaliss, 1962; G . Moller, 1964a; Bloch, 1965; Bubenik et al., 1965; Bubenik and Koldovsky, 1965; Batchelor, 1968). Current theory states that these experimental tumors have at least one antigen not present in the host. The tumor-specific antigen may elicit an antibody response, which, in turn, promotes the progressive growth of the malignancy. After

IMMUNOLOGICAL ENHANCEMENT

183

the neoplasm has attained a critical mass, the blocking antibodies, no longer needed for promotion of tumor growth, disappear from the circulation. An increasing number of reports have offered evidence that in patients with autochthonous tumors a detectable cellular immunity may become manifest in some of the afflicted (Hughes and Lytton, 1964; Fridman and Kourilsky, 1969; K. E. Hellstrom and HellstrGm, 1969; Savel, 1969; Stewart, 1969; Fass et al., 1970; Jehn et al., 1970). Cellular immunity has been measured by the development of DH skin reactions (Hughes and Lytton, 1964; Fass et al., 1970) or by increased incorporation of 3H-thymidine into autologous lymphocytes that have been stimulated in vitro with extracts of tumor tissue (Fridman and Kourilsky, 1969; Savel, 1969; Jchn et al., 1970). If the toxic activity of autologous lymphocytes in patients with neoplasia can be abrogated by antibodies and if this abrogation truly reflects a diminution of cellular immunity, then the operation of blocking antibodies may be a general phenomenon that occurs in a wide spectrum of immunological injury in which cellular immunity is a feature. (For further discussion of this mechanism, see Section VII.) C. NORMALTISSUEGRAFTS Until recently IB of normal tissue grafts has commanded relatively little attention in comparison with the intensive and extensive efforts invested in the study of tumor enhancement. This was probably because successful prolongation of normal tissue graft survival has been less than dramatic and often dubious. Survival of skin homografts on adult recipients that differ at the H-2 locus in mice or its homologous locus in other species may be extended, either by active or passive manipulations, for only 2 to 3 days beyond their anticipated survival, provided no immunologically depressing treatments are applied. Within the past few years, however, there have been several reports of long-term survival and function of renal grafts in rats (Stuart et d., 1968b; French and Batchelor, 1969; White and Hildemann, 1969; Lucas et aZ., 1970; Biesecker et al., 1971; Mullen and Hildemann, 1971) and of endocrine organs (0. E. A. Linder, 1961; Cock, 1962a,b; Hgskovh and Svoboda, 1962; 0. Linder, 1962; G. Moller, 1966; Parkes, 1968). These unexpected successes have renewed and stimulated interest in enhancement of normal tissue grafts. Prolonged survival of normal tissue grafts has been restricted to skin, kidney, heart, ovary, testis, and lymphoid cells. Perusal of the literature disclosed that skin grafts appeared to be the most difficult to enhance. In neonates, survival of about 10% of test skin grafts can be prolonged for many days and even weeks ( Heslop,

184

JOSEPH D. FELDMAN

1969; B. Zimmerman and Feldman, 1969a,b, 1970). In adults in which the immunological apparatus has been tampered with prolonged survival for significant periods of time has been achieved in a high proportion of manipulated recipients (Snyder et al., 1964; Winn, 1970; Chutni, 1971; Forbes et al., 1971; Jones et al., 1972; Jeekel et aZ., 1972). Heslop (1971) has reported extended survival of skin grafts in 2%of unmanipulated rats. There has been no adequate explanation as to why skin grafts should be more difficult to enhance than organ grafts. Some have proposed that skin is more immunogenic than other tissues, that it possesses adjuvant qualities, or that healing-in by capillary anastomosis makes it more susceptible to the destructive processes of rejection. When the difference between donor skin and allogeneic recipient is less than the major histocompatibility difference ( non-H-2, non-AgB, non-HL-A, etc.), survival time of the graft is generally longer than the usual 8 to 10 days of survival obtained when histoincompatibility is at the major locus. Under these circumstances, significant extended viability may be more readily accomplished ( Billingham and Parkes, 1955; Hildemann and Walford, 1960; 0. E. A. Linder, 1961; 0. Linder, 1962; HaSek et al., 1968; I. R. Cohen et al., 1971b; Heslop, 1971). The most striking successes for long-term survival of homografts have been achieved with transplantation of kidneys, ovaries, and testes, and most recently with heart. With renal homografts in rats there have been several reports of exceptionally long, perhaps indefinite, survivals (Stuart et al., 1968a,b; White and Hildemann, 1968, 1969; French and Batchelor, 1969; White et al., 1969; R. E. Wilson et al., 1969, 1971; Lucas et al., 1970; Ockner et al., 1970a,b; Thoenes et al., 1970; Marquet et al., 1971; Quadracci et al., 1971; Mullen and Hildemann, 1971). It should be noted that in these experimental renal grafts with extended survival periods, the histocompatibility difference has usually been nonAgB or less than a full AgB difference, i.e., transplantation of F, donor kidney to a parental recipient. Furthermore, most of these studies have been complicated by prior manipulations of the recipient so that IB was not always clearly operating. Stuart et al. (1968a,b), Lucas et al. ( 1970), and Marquet et aZ. (1971) injected immunocompetent lymphoid cells (spleen, lymph node, or peripheral blood) into prospective recipients before grafting. The effect of this type of preparatory infusion might well have been one of IB. It might also have represented antigenic overload with consequent induction of partial tolerance or have reflected an interference with parental immunological response as a result of host-versus-graft activity, i.e., the recipient Lewis rat lymphoid tissues might have been preoccupied with F, immunocompetent cells that filtered into them following parenteral injection. Moreover, infusion

IMMUNOLOGICAL ENHANCEMXNT

185

of donor bone marrow cells into prospective recipients of grafts apparently prepared the host for IB which was shown to be operative by passive transfer of alloantibodies ( Ockner et al., 1970b). Wilson and his colleagues have used antigenic material, presumably H-2 antigens, to induce subsequent extended survival of kidney grafts in homologous canine recipients (R. E. Wilson et al., 1969, 1971). Again, operation of IB in these long-term survivals of kidney grafts was not demonstrated. When non-AgB differences exist between donor and recipient, manifestations of IB are readily detectable, and survivals of renal grafts are practically indefinite (White and Hildemann, 1968, 1969; White et al., 1969; Thoenes et al., 1970; Mullen and Hildemann, 1971). In several of these studies, the difference between skin and renal grafts has been clearly delineated. Recipients were grafted with both kidney and skin, and in all instances the skins were rejected before the kidneys were destroyed (White and Hildemann, 1968; White et al., 1969; Lucas et al., 1970; Ockner et al., 1970a; Marquet et al., 1971). These results point up the more stringent requirements for IE of skin as compared with other organs. They also indicate that some organs, such as heart, may be easier to enhance than kidney (Jenkins and Woodruff, 1971; Marquet et al., 1971). Grafting of endocrine tissue has had a long history of partial success, particularly when viability in homologous recipients has been measured by detection of endocrine function or of hormones (Harris and Eakin, 1949; Krohn, 1959). There is no adequate explanation why endocrine tissues should enjoy such privileged viability in a foreign host. The few reports of IB of ovary and testis are, therefore, not surprising (Billingham and Parkes, 1955; Cock, 1962a,b; 0. Linder, 1962; G. Mdler, 1964a; Parkes, 1968). But it is difficult to ascertain in these reports whether active IB was responsible for extended survival or whether some intrinsic quality of the tissue or tissuehost relationship accounted for prolonged viability. Direct demonstration of antibodies in the prolongation of survival of these endocrine organs was not attempted, Furthermore, in all successful long-term survivals, the genetic disparity between donor and recipient has been small or unknown. Voisin and his colleagues have been most ardent proponents of IB of lymphoid cells when these were transferred to suitable recipients. Detection of the phenomenon, called “enhancement facilitation,” is based on the diminution or abolition of runt disease in newborn hosts (Voisin and Kinsky, 1962) or wasting disease in F, adults (Voisin, 1965; Voisin et al., 196%). The procedure is carried out by injecting immunocompetent spleen cells into newborn recipients or into F, adults that have been simultaneously given a small amount of blocking serum, Under

186

JOSEPH D. FELDMAN

these circumstances the recipients are protected from wasting. Similar success in protection of recipients could be achieved by pretreating prospective donors of spleen cells with antigenic preparations of recipients’ tissues ( Batchelor and Howard, 1965; Schlesinger and Gotein, 1965; Voisin et al., 196813; Wright, 1968) or with antibody directed against recipient antigens ( Safford and Tokuda, 1970). Comparable graft versus host reactions are being observed in humans who have been grafted with bone marrow from allogeneic sources (Buckley et al., 1971b; Jose et al., 1971). In both of these reports blocking antibodies, presumed to be directed against the recipients’ antigens, were claimed to be beneficial in reducing the injury and symptoms that accompany the attack of grafted immunocompetent cells against host tissues. The speculation has been offered that antigen-antibody complexes, derived from the grafted host, might bind to the foreign infused cells and inhibit their cytotoxic activities (see also Klein, 1971). It is a distortion of language, however, to consider this laboratory model as a type of IE. What is “enhanced in this system is the recipient that must be regarded as a graft, and it is protected from the destructive action of infused allogeneic lymphoid cells. Alternatively, protection of the recipient might be considered a “central blockade,” in which the infused immunocompetent cells are prevented by antibodies from exerting their full cytotoxic action upon the host. These few reports of enhancement of lymphoid tissue might better be regarded as the in vivo counterpart of cellular inhibition by antibody in &To-a phenomenon that is being increasingly explored both in the context of IB and in other immunological situations. D. DELAYED HYPERSENSITIVITY AND AUTOIMMUNITY The concept of IB has been expanded to include certain aspects of DH (Crowle and Hu, 1965, 1968, 1969; Axelrad, 1968; Axelrad and Rowley, 1968; Kaliss, 1969) and of some autoimmune diseases (Jankovi6 and Flax, 1963; Chutnh and RychlikovB, 1964; Paterson, 1966, Chutnh, 1968; Richardson and Paterson, 1970; Tung et al., 1970, 1971). In our view, however, prolonged survival of a normal tissue graft or enhancement of tumor growth are not precisely comparable with the development or lack of development of inflammatory lesions characteristic of DH and autoimmune pathology. It is a kind of elision to state that IB is due to a depression or inhibition of cellular immunity, that DH and some autoimmune diseases are due to cellular immunity, and therefore, suppression or abrogation of DH and lesions of autoimmunity are manifestations of IB. The mechanics of the development of inflammatory lesions in DH

IMMUNOLOGICAL ENHANCEMENT

187

and autoimmune diseases are reasonably understood (J. D. Feldman and Najarian, 1963). A suitable recipient is immunized by inoculation of antigen or tissue incorporated into adjuvant, usually complete Freunds adjuvant ( CFA), i.e., a water and oil emulsion containing tubercle bacilli. After an incubation period of some days a minor population of lymphocytes has become sensitized. When antigen is deposited again in the animal, as in a skin test, the specifically reactive lymphocytes interact with antigen and generate a spectrum of lymphokines or chemical mediators (at least in the test tube), The lymphokines have several qualities: one inhibits the migration of macrophages; another is toxic for lymphocytes; a third is chemotactic for polymorphonuclear leukocytes (PMN), etc. Hematogenously borne monocytes and macrophages, issuing from bone marrow, accumulate in the site of antigen-lymphocyte interaction. In this manner a zone of inflammation is slowly evolved, with edema, increased vascular permeability, and infiltration of inflammatory cells. The development of lesions in autoimmune diseases presumably evolves in the same way, only the antigen may be tissuederived or tissue-fixed. The details of this process have been spelled out in order to present critically the concept of IB in DH and autoimmune diseases, i.e., the inhibition or abrogation of lesions induced by antigen. Axelrad (1968), Axelrad and Rowley ( 1968), and Crowle and Hu (1968, 1969) have suggested that DH may be inhibited or completely suppressed by pretreating the animal with antigen or antibody before challenge with antigen in adjuvant. The design of their experiments is as follows. Antigen (sheep red blood cells or serum proteins) is injected intravenously into rats or mice several days or several weeks before immunization by inoculation of antigen in CFA. The recipients are then prepared by intradermal administration of antigen in adjuvant and some days later they are skin-tested. The skin test lesions that appear are smaller than those that develop in animals not pretreated, or the lesions may be completely abolished. Other reported results belong to this category (Flax and Waksman, 1962; Asherson and Stone, 1965). A variant of this design is to inject minute quantities of antiserum into the prospective test animals instead of antigen (Crowle and Hu, 1965, 1968; Axelrad, 1968; Axelrad and Rowley, 1968). Microliter quantities of antiserum are equally effective in suppressing or abolishing DH lesions in suitably prepared animals. If the suppression of DH lesions and prolongation of graft survival by antibody are equivalent expressions of inhibited cellular immunity, then in both cases IB is probably operating. Unfortunately, the proponents of IB in DH have not presented quantitative and temporal information about the antigens or antibodies used.

188

JOSEPH D. FELDMAN

For example, it is known that only a minute fraction of injected antigen is actually immunogenic, and, furthermore, this immunogenic fraction persists for some time in the recipient ( McConahey et al., 1968). Therefore, the suppression of DH lesions may be due to pre-emption of the antigen-reactive cells by the antigen that was given intravenously before challenge with antigen in adjuvant; or to the development of partial unresponsiveness. However, if antibody or antiserum is used to abrogate DH lesions, this process may be due to the complexing of antibody with the antigen in adjuvant and the degradation of the complexes before immunization can occur (afferent blockade) or to the inhibition of antibody formation by antibody (central blockade). It should be noted that some mononuclear cell lesions have been induced by antibody, and these duplicate histologically the lesions of DH that are presumed to be mediated by sensitized cells (Oldstone and Dixon, 1970). Since none of the above alternatives has been examined in the studies cited (Crowle and Hu, 1965, 1968; Axelrad, 1968; Axelrad and RowIey, 1969), it is difficult to ascertain whether or not IB operates in DH. Laboratory models of some autoimmune diseases may be initiated in the same fashion as is DH, and their pathology is also similar. Experimental allergic encephalitis ( Paterson, 1966; Richardson and Paterson, 1970), allergic orchitis (Chutnh and RychlikovL, 1964; Tung et al., 1970, 1971), and experimental thyroiditis (Jankovib and Flax, 1963; Weigle, 1965; Weigle and Nakamura, 1969) belong in this category. In all three of these models the course of the disease and its pathology may be modified by parenteral administration of antiserum, i.e., there is inhibition or abolition of tissue injury and amelioration of clinical disability. Whether antiserum in these instances promotes a state of IB is still moot and for the same reasons presented above under DH. Jerusalem et al. (1971) described the enhancement of malaria in mice by both active and passive immunization. Enhancement of the disease was measured by parasitemia and by mortality. In their model, small doses of parasite antigens provoked an enhancing phenomenon, and passive transfer of antibodies to malarial parasites was also effective. V. Fetus as Homografi

The conceptus is a foreign tissue implanted in the uterine mucosa of a pregnant female. It enjoys a protected position for many days beyond the usual survival times of grafted normal tissues. What protects the conceptus from expulsion by its maternal host? In an immunological context, the fetus might be aided by blocking maternal antibodies and thus maintain its privileged position for the necessary gestation

IMMUNOLOGICAL ENHANCEMENT

189

periods (Billingham, 1964). The following questions arise: ( 1 ) Can the mother be immunized by the conceptus? ( 2 ) Does she produce antibodies to H antigens? ( 3 ) Do these antibodies traverse the placenta and fix in the fetus? Lastly, ( 4 ) does IE play any role of biological significance in this situation? Immunization of the pregnant female by concepti has been known for many years, and passage of both red and white blood cells from fetus to mother and mother to fetus has been demonstrated in a variety of ways. In the circulation of the pregnant female antibodies to paternal H antigens have been identified (Hertzenberg and Gonzales, 1962; Payne, 1962; Kaliss et al., 1963; Mishell et al., 1963; Kaliss and Dagg, 1964; Kaliss, 1968). Antifetal antibodies traverse the placenta (Lanman et al., 1964; Kaliss, 1968), but fixation in the fetus with radiolabeled antibodies was not demonstrated (Witz et al., 19691,). The existence of maternal antifetal antibodies and their passage across the placenta do not prove that IE operates in this situation. One would need to demonstrate that the fetus is protected or that the cellular immunity engendered in the mother is inhibited. Breyere and his colleagues presented good evidence that female mice of one inbred strain could be induced by matings to males of another strain to tolerate, that is, to enhance, tumors that were genetically related to the males (Breyere and Barrett, 1960, 1961; Breyere and Burkoe, 1963; Breyere, 1967). Kaliss and Dagg (1964) confirmed these observations, at least insofar as their mated females were capable of enhancing the growth of a sarcoma. Others have tested the system with skin grafts with conflicting results. In humans, mice, and rabbits, mothers were grafted with skin of their own neonates and accepted them for periods longer than the usual survival times of unrelated neonatal skin grafts (Bardawill et al., 1962; LengerovL and VojtiSkovh, 1963; Najarian and Dixon, 1963). Nonetheless, Vener et al. ( 1961) were unable to observe prolonged survival of male skin grafts in multiparous female recipients, and Billingham et al. (1965) could not demonstrate enhancement of skin homografts on infant rats. None of the above experiments with tumor and skin grafts is directly applicable to the question originally posed: Is the fetus protected by blocking antibodies? More recently, blocking antibodies have been detected in the serum of female mice (K. E. Hellstrom et al., 1969) and female rats (Biesecker et al., 1971) impregnated by males of a different strain. They have also been found in sera of pregnant women (Buckley et al., 1971a). These blocking antibodies abrogated the cytotoxic action of maternal lymphocytes against embryonic fibroblasts. The simple detection of such antibodies in this experimental model does not determine their biological significance, particularly with regard to IE during pregnancy.

190

JOSEPH D. FELDMAN

VI. Tolerance

Tolerance is defined differently by those who graft viable cells and tissues and are concerned chiefly with cellular immunity from those who study antibody response to antigen and are concerned chiefly with synthesis of conventional-type antibody. On the one hand the former accept the existence of tolerance if allogeneic or xenogeneic grafts are maintained in their recipients for periods longer than mean or median survival times (partial tolerance) or indefinitely (complete tolerance). In cellular terms this would imply a diminution or absence of specifically sensitized effector cells to destroy the graft. On the other hand, the latter consider tolerance to be diminution (partial) or absence (complete) of measurable conventional antibody. Antibody may be quantitated by direct binding or precipitating reactions or by enumeration of antibody-forming cells. There are, then, two systems of immunocompetence, and the relationship of one to the other is undergoing intensive study. With the development of the colony inhibition method (I. Hellstrom and Sjogren, 1965; I. Hellstriim, 1967) the Hellstriims introduced a means of measuring cellular immunity by quantitating cytotoxicity of lymphocytes in uitro. Others have also elaborated quantitative methods to measure effector cell cytotoxicity (Brunner et al., 1967; Perlmann and Holm, 1969; Ford et al., 1970; Mauel et al., 1970; Henney, 1971; Nordin et al., 1971a,b), The methods, whether involving inhibition of colony formation, 51Cr release, or increase in lymph node weight, can also be used to estimate the effects of antisera and antibodies on effector cell cytotoxicity. For example, if an aliquot of sensitized lymphocytes will decrease colony formation by 50%,the cytotoxicity can be inhibited by incubation of the cytotoxic lymphocytes with antisera from varying sources, but especially from grafted recipients. It is presumed that the antisera in some unknown fashion (see Section VI1,C) inhibit the full expression of lymphocyte cytotoxicity and permit the growth of colonies to levels associated with control lymphocytes. These prefatory comments are inserted here because of recent reports about the mediation of tolerance by putative antibodies. First, the type of tolerance referred to is that defined as acceptance of graft. I. Hellstrom and Hellstrom (1971), by injection of F, spleen cells, induced tolerance in neonatal mice the existence of which was demonstrated by acceptance of skin grafts. I. Hellstrom et al. (1970a) also created chimeric dogs by infusing allogeneic bone marrow into lethally irradiated recipients. The dogs were tolerant to the allogeneic cells for months to years. They were then able to show that serum from tolerant hosts abrogated the cytotoxicity of the tolerant host’s lymphocytes. The

IMMUNOLOGICAL ENHANCEMENT

191

conclusion was that tolerance, under these circumstances, might be the result of putative alloantibodies preventing sensitized lymphocytes from destroying a target graft. In similar fashion, BildsZe et al. ( 1971), French et al. (1971), and Stuart et al. (1971), demonstrated that lymphoid cells from rats bearing long-surviving renal grafts were capable of mounting a graft-versus-host reaction, i.e., the cells exhibited an immunological response but were presumably inhibited from destroying grafts in t h o . Others have reported the presence of alloantibodies in the circulation of rats carrying long-surviving renal grafts (Ockner et al., 1970b; Thoenes et al., 1970; Marquet et al., 1971). Voisin et al. (196813) also induced tolerance in mice by injection of allogeneic cells and defined tolerance as acceptance of skin grafts. They demonstrated in these tolerant mice the following: ( a ) immunoglobulin-producing cells; ( b ) anti-H-2 antibodies against the tolerated donor tissue; ( c ) antibodies that fixed to cells of both donor and recipient; and ( d ) antibodies with enhancing capacity after passive transfer to normal mice. In the most extreme model of “tolerance,” Wegmann et al. (1971) demonstrated specific cytotoxicity of lymphocytes derived from allophenic mice ( a combination of C57 and C3H strains) directed against fibroblasts of either C57 or C3H origin. The cytotoxicity could be abolished by incubating the allophenic lymphocytes and target fibroblasts in ullophenic serum. The allophenic serum was presumed to contain antibody or antigen-antibody complexes that interfered with the killing action of the allophenic lymphocytes. This extended discussion of tolerance and its relation to blocking antibodies is offered to emphasize that tolerance of a graft is quite different from immunological unresponsiveness associated with absence of antibody formation. Furthermore, the original definition of graft tolerance requires modification in the light of current information (Bilds@ et al., 1971; I. Hellstrom and Hellstrom, 1971). Moreover, the evidence so far available is not conclusive that blocking antibodies mediate tolerance. It is unreasonable to think that autologous blocking antibodies appearing in X host would specifically bind to and affect the lymphocytes of X host. There must be other factors involved. Also, the putative blocking antibodies have not been characterized and have not been definitively shown to be antibodies to H antigens; and the operation of blocking antibodies in vivo has not been analyzed. VII. Theories

The immune response is a dynamic process and has been conceptually divided into three compartments in order better to examine and explain certain phenomena. The afferent arc is that part of the response that

192

JOSEPH D. FELDMAN

begins with the presentation of immunogenic material to the host, either of endogenous or exogenous origin, and includes the concepts of recognition, preparation, and meeting of antigen with immunocompetent cells. The efferent arc encompasses the events that occur in the already immunized host and describes the phenomena associated with antigenantibody reactions, antigen-cell interactions, activation of C, generation of phlogogenic substances, and evolution of injury. The central arc consists of a cellular charge or molecular alterations of cells to a state capable of responding specifically to antigen, i.e., the changeover of a naive cell to one endowed with intracellular machinery and which can synthesize antibody or to an effector element that can react directly and specifically with antigen. Theories of IB rest on this kind of artificial concept to explain the plethora of details that have been accumulated from heterogeneous in vivo and in vitro situations.

A. AFFERENTBLOCKADE The idea expressed in this theory is that IB occurs because antibodies with enhancing activity bind to target cellular antigens and block them from reaching immunocompetent cells of the host, either cells that course through the graft or those that reside in lymphoid tissues. Gorer (1942) first implied that this might occur when he reported the enhancement of leukemia in A strain mice that were inoculated with tumor cells incubated in C57 immune serum. At the same time, the spleens of inoculated A strain mice did not enlarge, whereas the spleens of A strain mice did when these recipients received leukemic cells incubated in nonimmune serum. This gross observation was interpreted as a block of recognition. Later, Snell (1956) and Billingham et al. (1956) elaborated a similar point of view based, respectively, on enhancement of tumor that was inoculated into recipient mice with lymph node cells and on slightly prolonged survival of skin homografts. Both suggested a “walling off of antigen. Proponents of afferent blockade have exploited other in vitro and in vivo models (Gorer, 1942; G. Moller, 1963a,d, 1964a; Crowle and Hu, 1969; I. R. Cohen et al., 1971b; McKenzie et al., 1971; Jones et al., 1972). The concept of afferent blockade rests on two findings. ( 1 ) In uitro, target tissues, either tumor cells or normal tissue grafts, when immersed in blocking antibodies, will bind antibodies as shown by fluorescence examination of target cells or by isotope concentration in target tissues (Snell, 1957; Kaliss, 1958; E. Moller and Miiller, 1962; G. Moller, 1963a,d; E. Moller, 1965a; Hutchin et al., 1967; Ceppelini et al., 1969; B. Zimmerman and Feldman, 1970; Jones and Feldman, 1971b). When grafted into allogeneic recipients, tumors are enhanced or skin graft survival prolonged. (2) In uiuo, specific uptake of labeled antibodies of en-

IMMUNOLOGICAL ENHANCEMENT

193

hancing antiserum may be detected in the grafts (Morris and Lucas, 1971; Jones et al., 1972). In direct measurements of cell mediated immunity, expressed as the number of cytotoxic lymphocytes, Peter and Feldman ( 1972) showed that blocking antibodies deferred the development of peak cellular immunity for 3 to 4 days, presumably because alloantigens were prevented from reaching immunocompetent cells. However, none of these experiments has demonstrated directly that access of antigen to immunological centers and immunocompetent cells has been impeded, reduced, or abolished. Nor does the binding of blocking antibodies to target sites distinguish between afferent and efferent blockade. Indeed, a major argument against the concept of afferent blockade is that immunity may be detected at the same time and in the same recipient bearing an enhanced graft (Kaliss and Bryant, 1958; Casey et al., 1959; G. Moller, 1964a; Halasz and Orloff, 1965; Hutchin et al., 1967; Kodama et al., 1967; Kaliss, 1969, 1970; B. Zimmerman and Feldman, 1970; Jones et al., 1972). For example, a Lewis rat is grafted with ACI skin and given an injection of enhancing antibodies. Twenty to 40 days later, when the first graft is healed-in, a second ACI skin is applied. The second graft is rejected in 10 days, but the primary skin graft survives (Jones and Feldman, 1971a). Similarly, when tumors are serially inoculated into mice the secondary tumor is destroyed in accelerated fashion, whereas the primary tumor enlarges and often overwhelms the host ( Kaliss, 1958, 1969; Kaliss and Bryant, 1958). Furthermore, Kaliss emphasized that a tumor inoculum can be enhanced by injection of blocking antiserum 7-10 days after inoculation and after hemagglutinating antibodics have appeared in the circulation (Kaliss, 1958). He stressed the point that IE becomes manifest only in a host that is making an immunological response. Others have examined the histological response and numbers of phagocytic cells in and around tumor grafts and have concluded that no measurable difference is discernible between the response to tumor inocula without blocking antisera and the response to similar inocula in the presence of blocking antisera (Batchelor and Silverman, 1962; Chantler, 1967), i.e., such inocula, enhanced or not, elicit the same type of cellular reaction and, presumably, similar immunological responses. Evidence of immunity is also indicated by the appearance of hemagglutinating and cytotoxic antibodies during the period of extended survival of enhanced grafts, but, as noted before, no direct relationship has been established between these in vitro measurements and in vivo enhancement. To explain the altered relationship between graft and host, one theory proposes either a physiological, antigenic, or metabolic alteration of the neopIastic graft. As noted earlier, there is no evidence of change in en-

194

JOSEPH D. FELDMAN

hanced tumors. Nor is there evidence that normal tissue grafts are altered after long-term residence in allogeneic hosts. If skin homografts remain viable for extended periods of time, when they are regrafted to a naive recipient of the same strain, they are rejected in the usual firstset fashion; and when they are regrafted back to the strain of the skin donor, they are accepted indefinitely ( HhskovL et al., 1965; Barker and Billingham, 1971; Jones et al., 1972). This conclusion does not mean that normal tissue grafts may not be altered by unusual laboratory manipulations (Klaue and Jolley, 1971) nor that they may not show quantitative variations in antigenic composition as tissue matures from neonatal to adult life (E. Moller, 1963; G. Moller, 1963d; Heslop, 1969). Nor does it mean that grafts may not acquire or lose antigens as a result of ingress and egress of passenger leukocytes (see below). The theory of antigenic alteration cannot be dismissed without an additional comment. Several reports indisputably establish that the antigenic composition of cell surfaces may fluctuate (E. Moller, 1963; G. Moller, 1963d; Reif and Allen, 1964; Boyse et al., 1968b,c; Old et al., 1968; Kourilsky et al., 1971; Lance et al., 1971). Particularly is this true for mouse lymphocytes, in which cell membrane antigens, such as TL, Ly, and even H-2, may increase or decrease under certain environmental conditions or may be modulated by antibody. Hellman and Duke (1967) reported an alteration of antigenicity in skin when two allogeneic skins were incubated in the same medium. More pertinent to the subject of enhancement, Morris and Lucas ( 1971) claim modulation of antigenic sites in renal homografts induced by blocking antibody. They based their conclusion on the diminished binding capacity of long-term renal grafts for fixation of specific enhancing antibody. What is important to stress here is that modulation of the antigenic character of cells appears possible, and if this is so, there is no conceptual reason that grafts might not also be altered in appropriate environments ( Boyse, 1971). There is one further comment to be made concerning afferent blockade. Macrophages have been considered as one of several cellular elements involved in immunological responses. Speculations are appearing that afferent blockade may be related to nonrecognition of graft antigens by macrophages ( M . Feldman and Globerson, 1960; Haughton and Nash, 1969; Amos et al., 1970). Pearsall and Weiser (1968a,b) examined this problem directly. They depleted allograft recipients of macrophages by administering silica ( an agent that destroys macrophages) . In mice so manipulated, skin grafts enjoyed prolonged survival ( Pearsall and Weiser, 1968a). In a second study they showed that macrophages, derived from donors immunized by grafts, were capable of causing accelerated rejection of specific tests grafts in mice that were infused with

IMMUNOLOGICAL ENHANCEMENT

195

immune macrophages (Pearsall and Weiser, 1968b). The effect blocking antibodies might exert on macrophages and, thus, on IB has not been ascertained. In summary, the data neither prove nor refute the operation of afferent blockade in IB nor delineate the extent of influence this kind of inhibition might contribute to the expression of IB. Until the antigens involved in IB are isolated and characterized, it will not be possible to determine quantitatively if antibody alters the immunogenicity of grafts and at what time in the complicated sequence of events during graft rejection this might occur. B. EFFERENT BLOCKADE This theory states that IB occurs after immunity has been established because effector elements, either sensitized cells, antibodies, or both, are prevented from reaching and reacting with the antigens responsible for IB. The block is presumed to be in one of two places, at the periphery of the graft where blocking antibodies “cover” the antigens responsible for initiation of IB and prevent effector elements from “seeing” their targets, or centrally in the lymphoid organs where blocking antibodies depress or inhibit cellular immunity, i.e., they bind to effector cells and prevent them from interacting with target antigens. Some might consider the latter action one of central blockade. A number of experiments have been designed to examine this concept, and the results in general have indicated in an indirect way that efferent blockade may operate in IB (G. Moller, 1963a,b,c, 1964a,b; E. Moller, 1965a). The prototype experiment is constructed as follows. Recipient animals are immunized with tissues from donors, particularly to involve draining or regional recipient lymph nodes. Tumor cells from donors are incubated with blocking antibodies or normal serum, and then washed to remove excess medium. The treated tumors are then incubated with sensitized lymph node cells of the recipient, and the extent of destruction ( “Cr release, dye exclusion, or other measurement ) is compared between experimental ( tumor plus blocking serum) and control targets (tumor plus control serum). Or the treated tumors may be injected into sensitized recipients, and the extent of enhancement and/ or rejection is compared between experimental and control groups. The consensus of evidence suggests that blocking antibodies somehow protect the tumor, i.e., they prevent or diminish target destruction by effector cells. The interpretation of the results favors efferent blockade at the periphery. In one study carried out i n vivo, however, the investigators measured an actual decrease of cytotoxicity in the draining nodes of recipients injected with tumor and enhancing antisera, i.e., a presumed loss of effector

196

JOSEPH D. FELDMAN

cells from the regional lymph nodes ( Snell et al., 1960). This would represent central efferent blockade. In two other reports, the authors were unable to demonstrate an inhibition of sensitized cytotoxic lymphocytes by the action of blocking antisera ( Klein and Sjogren, 1960; Brondz, 1965). Another experimental variant to demonstrate efferent blockade involves the use of varying amounts of blocking antibodies and constant amounts of antigen (tumor inocula) . Relatively large quantities of cytotoxic antisera (0.5-1.0 ml.) in the presence of C destroy target cells, but minute quantities of the same antibodies (0.1-.5 pl.) enhance the growth of tumor cells or protect them in uitro from destruction (Gorer and Kaliss, 1959; Batchelor and Silverman, 1962; Hutchin et al., 1967; Haughton and Nash, 1969; Amos et al., 1970; Linscott, 1970). By rough calculations one may determine the number of immunoglobulin molecules and antibody molecules in minute volumes of blocking antisera. Unfortunately, an actual computation of antigenic sites is not possible, since there are no quantitative measurements of H antigens on living cells, nor is there any accurate information on the surface distribution of antigens, which may determine whether antibody may be toxic or protective to target cells. Haughton and Nash (1969) computed the number of blocking antibodies in an experiment and concluded that the number of molecules was insufficient to cover all antigenic sites. They proposed, instead of efferent blockade, an interference with macrophage function. The latter is pure speculation, and the computation of molecules needed cannot yet be ascertained. All the arguments, proposals, and speculations offered to explain afferent blockade apply equally well to the theory of efferent blockade. The single difference between the two concepts is one of time. In the former, antibodies block antigenic sites ab initio; in the latter, after immunity has developed, antibodies block antigenic sites and “hide” the graft. Also, just as quantitative data about tissue antigens are lacking to support the theory of afferent blockade, no reliable information exists about the number of blocking antibodies available for blockade, their avidity, distribution, and half-life. Morris and Lucas ( 1971) have reported a rapid disappearance of blocking antibodies from the circulation of target rats during the first 4 hours after passive administration of labeled molecules. Jones and Feldman have obtained similar results (1971a). Many of the reported observations of IB might be reasonably explained by efferent blockade (French et al., 1971), but solid quantitative data are needed to prove beyond a reasonable doubt that such a mechanism indeed operates in vivo and in uitro. For example, that IB may occur after tumor cells and blocking antibodies are incubated to-

IMMUNOLOGICAL ENHANCEMENT

197

gether may be due not to a covering of antigenic sites but to dissociation of antibodies which may, then, affect sensitized lymphoid cells (Brody and Beizer, 1963; Amos et al., 1970; B. Zimmerman and Feldman, 1970). Or it is conceivable that blocking antibodies might destroy so-called passenger leukocytes within the graft. Passenger leukocytes are generally mobile and may contribute to the total antigenic stimulus of a graft (Steinmuller, 1967, 1969; Billingham, 1971). If they are destroyed by yG antibodies and C, the total quantity of immunogenicity of the graft is reduced, the number of leukocytes draining to regional lymph nodes may be reduced or eliminated, and sensitization of these draining lymphoid centers may be delayed. One might even imagine that passenger leukocytes trapped in the host’s draining lymph nodes can initiate a graft-versus-host reaction and transiently impair the host’s immunological response. Other arguments can be invoked to explain the detrimental action of blocking antibodies on cytotoxic effector lymphocytes, and these must assume that the sensitized effector lymphocytes hold bound antigen on their surfaces. The blocking antibodies would complex with the ccll-bound antigen. Consequently, there may be a reduction in number due to destruction by C or due to exit from the local nodes, or antigen-antibody complexes may bind to effector lymphocytes and inhibit their toxic activity against target cells ( Klein, 1971). A major argument against efferent blockade as the mechanism of IE is that blocking antisera given repeatedly to recipients with enhanced neoplastic or normal tissue grafts do not prolong survival of the grafts indefinitely ( Winn, 1970; McKenzie et al., 1971; Jones et al., 1972). If blocking antibodies operate by covering antigenic sites of target grafts or by inhibiting effector lymphocytes, then multiple doses of these antibodies, given over extended intervals of time, should be capable of covering antigenic sites that have been uncovered by dissociation of antibodies or that have been recently synthesized; or they should be capable of reacting with effector lymphocytes that are being continuously formed and released. To date, neither multiple doses of blocking antibodies nor polyvalent sera administered repeatedly have extended the period of enhancement. C. CENTRAL BLOCKADE

The theory states that blocking antibodies regulate or inhibit either the production of antibodies that contribute to graft rejection or the number, mobility, and efficacy of sensitized lymphocytes that may react specifically to target antigens and effect the destruction of homografts (Snell et al., 1960; Gorer, 1961; HAskovA et al., 1965; Brunner et al., 1968; Uhr and Moller, 1968; Lapp and Mdler, 1969; Ryder et al., 1969;

198

JOSEPH D. FELDMAN

Takasugi and Hildemann, 1969a,b; Brunner, 1970; Bloom and Hildemann, 1970; K. E. Hellstrom and Hellstrom, 1970; Kaliss, 1970; Safford and Tokuda, 1970; Chutnh, 1971; Forbes et al., 1971; McKenzie et al., 1971; Tada and Okumura, 1971; Klein, 1971). The block is presumed to be at the level of the effector cell, most likely within the confines of the lymphoid centers, e.g., lymph nodes and spleen. There is a good deal of overlap between the concepts of central blockade and afferent and efferent blockades, and sharply delineated boundaries do not exist to separate one concept from the other. If the central arc of the immune response is that part of an integrated dynamic process concerned with the conversion of a neutral host and cell into one immunologically prepared to meet with antigen, then central blockade encompasses those phenomena that interfere with this conversion. Stated in less abstract terms, antigenreactive elements ( lymphocytes and/or macrophages in any combination) may be prevented from being activated to react specifically (afferent central blockade) ; may be inhibited intracellularly from acquiring the cellular machinery to become specifically reactive (true central blockade) ; may be prevented from releasing and delivering effector elements to target sites (efferent central blockade) ; or may be reduced in numbers. All of these are examples of central blockade and, as above, there is little firm evidence that this theory actually describes the operation of IB. The most cogent arguments for true central blockade have been offered by Takasugi and Hildeniann ( 1969a,b). These investigators have shown that yG,, in mice, is an effective immunoglobulin that mediates IB of tumor inocula. Neither yM nor yG, was efficacious, even though the former is cytotoxic, and both classes of immunoglobulins are composed of similar light chains which, with the heavy chain, form specific antibody sites. Next, they reported that the passive transfer of yG, into mice inoculated with tumor inhibited the lymphocytosis in the circulation that accompanies the immune response to the tumor. Neither yM or yG, abolished the lymphocytosis in inoculated recipients. Together, these observations suggested that yG, inhibited those cells that synthesized yG2 and, thus, permitted the enhancement of tumor. There are, however, no direct data in Takasugi and Hildemann’s work that yG2 actually inhibited the formation of yG2, since no quantitative determinations of this immunoglobulin were made, nor that the lymphocytosis, abolished by YG,, was in fact the result of an increase in the circulation of lymphocytes that produce yG,. Production and release of yM antibody may also be inhibited or, perhaps, delayed by the putative suppressive action of blocking yG antibodies on synthesis of yM ( McKenzie et al., 1971; Peter and Feldman, 1972. If yM antibody is toxic to allogeneic cells, its sup-

IMMUNOLOGICAL ENHANCEMENT

199

pression would contribute to enhancement of survival. The proposal of Takasugi and Hildemann is, therefore, attractive and is supported to some degree by the evidence of feedback regulation of antibody production by antibody, purported to operate in the immunological centers of immunized animals (Henry and Jerne, 1968; Uhr and Moller, 1968; Cerottini et al., 1969; Lang et al., 1969; Pierce, 1969; Ryder et al., 1969; Tada and Okumura, 1971). Discussions of central blockade may also be found in other reviews (Uhr and Moller, 1968; Snell, 1970). Four recent reports offer preliminary data that may illuminate the process of central blockade. All suggest that blocking antibodies complexed with antigens interfere in vltro with effector cytotoxic action of sensitized lymphocytes. Brunner ( 1970) incubated H-2 antigens with alloantibody and decreased cytotoxicity of sensitized lymphocytes against target cells. Sjogren et al. (1971) prepared from tumor tissue a small molecular weight substance, 10,000-100,000, which, when added to putative antibodies, blocked the cytotoxic effect of mouse lymphocytes sensitized to tunior-specific antigens. Brawn (1971), in the same laboratory, reported similar results except that the mouse lymphocytes were sensitized to H-2 antigens. The last two mentioned reports are too incomplete to identify the low molecular weight substance. Klein (1971) established an in vitro model to examine the effects of blocking antibodies on effector cells directed against mouse tumor. He reported that blocking antibodies with antigen, i.e., tumor, reduced the toxic activity of effector cells but blocking antibodies alone did not. Increasing numbers of investigators suggest that IB occurs because of central blockade, and they define central blockade as inhibition of cellular immunity (Mitchison and Dube, 1955; Schlesinger and Gotein, 1965; Brunner et al., 1967; Axelrad, 1968; Axelrad and Rowley, 1968; Stuart et al., 196813; Ono et aZ., 1969; Alexander, 1970; Bloom and Hildemann, 1970; Corson, 1970; I. Hellstrijm et al., 1970b; Richardson and Paterson, 1970; Blair et al., 1971; Buckley et al., 1971a; ChutnL, 1971; Forbes et aZ., 1971; Gillespie et al., 1971). Cellular immunity is a vague phrase difficult to define and measure. It does not help to invoke depression of cellular immunity as an explanation of IB, since the phrase encompasses a spectrum of activities and interactions that have only recently been scrutinized and are poorly understood. For example, the role of macrophages, of long-lived and short-lived lymphocytes, of blast transformation, of cell cooperation, of release of chemical mediators, and of vascular permeability, all of these roles rightfully belong to biological activities effected by immunized cells. Therefore, depression of cellular immunity, a putative ingredient of IB, means a depression of what? Macrophage processing of antigen? Inhibition of blast transformation?

200

JOSEPH D. FELDMAN

Elimination of antigen-reactive lymphocytes? Destruction of sensitized immunocompetent cells? Related to central blockade are several other activities which, apparently, operate at a cellular level. Miller et al. (1971) reported the inhibition by alloantibodies of one-way mixed lymphocyte cultures prepared from combinations of dog lymphocytes. Four to 14 days after placement of renal allografts into recipient dogs, the reactivity of recipients’ lymphocytes against donor cells was reduced or eliminated if the mixed lymphocyte cultures were carried out in recipients’ plasma or in IgG fractions of these plasmas. Similar results were observed by Gordon et al. (1971), following placement of heart allografts in recipient rats. Hornung et ul. (1971) blocked mixed lymphocyte cultures of human cells with F( ab’), fragments of 7G obtained from human anti-HL-A antisera. Bernstein and Wright (1971) reported that alloantiserum made by repeated injections of allogeneic lymphoid cells into donor mice, abrogated the inhibition of migration that occurred when immunized donor peritoneal cells in capillary tubes were exposed to recipient peritoneal cells in chambers, These cited papers indicate that alloantibodies with blocking activity affect lymphocyte functions that are associated with lymphokines. Antigenic competition has been invoked as an explanation of the mechanism of IB (Flax and Waksman, 1962; Liacopoulos et al., 1962; Eidinger et al., 1968; Howard et al., 1969). It implies that the immune response to an antigen is inhibited by the simultaneous presence of a second antigen that competes for immunological recognition and cellular response. With respect to IB, only a few laboratory models might be accommodated by this theory. For example, the injection of certain viruses into mice has been associated with prolonged survival of normal tissue grafts (Howard et ul., 1969), and, similarly, the inoculation of heavy doses of CFA just before or simultaneously with the placement of a graft has been accompanied by extended viability of the graft (Flax and Waksman, 1962; RuSkiewicz, 1967; ChutnL, 1968; see, also, Kaliss, 1969). In both examples one might postulate that the virus or adjuvant pre-empted some of the immunological apparatus and, thus, denied full immunological response to the graft. Nevertheless, some viruses suppress the manifestation of cellular immunity ( Ceglowski and Friedman, 1968; Blair et al., 1971; Weiner et al., 1971) by directly affecting immunocompetent cells, and adjuvants and Bacillus CalmetteGukrin (BCG) usually intensify immune responses to antigen and accelerate graft rejection (Balner et al., 1962). Furthermore, it would be difficult to imagine how survival of a skin or renal graft might be prolonged by these agents after immunity has developed. The cellular basis

LMMUNOLOGICAL ENHANCEMENT

201

for comprehension of antigenic competition is too meager at this time to help elucidate the development and operation of IB, and, to date, passive blockade has not been accomplished with any of the above models. Indeed, it is difficult to conceive how antibody might mediate IB of grafts after preparing a recipient with virus or adjuvants, unless one invokes the idea of antigenic ubiquity, i.e., there are shared antigens in the virus or adjuvants that elicit antibodies that cross-react with graft antigens. Central blockade may also be effected by passenger leukocytes. It has been suggested that white blood cells trapped in a graft enjoy mobility and may exit from the graft to reach regional lymphoid organs where they may mount an attack on host lymphocytes or be attacked by host cells (Lapp and Moller, 1969; Ockner et al., 1970a; Billingham, 1971; Guttmann and Lindquist, 1971) . The host’s immunocompetent tissues are thereby preoccupied and unable to respond fully to graft antigens other than those carried by leukocytes. Finally, some obscure central mechanism appears to be operating in a recipient bearing an enhanced graft. In adult animals it is extremely difficult and often impossible to prolong for more than a few days beyond the usual period of rejection, survival of normal tissue grafts that differ from the recipient at the H-2 or AgB locus. However, in adults that have been manipulated by thymectomy, by immunosuppressants, or by X-ray treatment, significant prolongation of graft survival may be achieved (Winn, 1970; Forbes et al., 1971; Jeekel et al., 1972; Jones et al., 1972), and neonatal recipients may retain viable skin homografts for weeks after treatment with enhancing antibodies ( Heslop, 1969, 1971; B. Zimmerman and Feldman, 1969a,b, 1970). The implication of these experiments is that the immunological apparatus of a graft recipient must be either incompletely developed, reduced, or dampened by experimental manipulations to allow a graft to take and survive for more than 10 days,

D. MISCELLANY Autoenhancement is a term recently assimilated into the subject of IB and it offers no novel ideas about the mechanism of blocking antibodies (Cock, 1962b; Voisin, 1965; E. Moller et al., 1968; I. Hellstriim et al., 1969; Bloom and Hildemann, 1970; Lucas et al., 1970). Simply stated, when blocking sera are administered to a recipient of an allogeneic graft, they prolong its survival, during which extended period of viability the graft has an opportunity to immunize actively the recipient. Once the recipient responds, he produces additional blocking antibodies which, in turn, maintain the survival of the graft. In some experiments, long-term survival of skin homografts has been attributed to tolerance

202

JOSEPH D. FELDMAN

or to large size of graft, but the observations might just as well be interpreted as a form of enhancement (Silvers and Billingham, 1966; Silvers, 1968; Lapp6 et al., 1969; Wachtel and Silvers, 1971). The idea of autoenhancement is particularly applicable to the progressive growth of autochthonous tumors, a situation in which there is a continuous increase of antigen and a continuing antibody response driven by antigen. Amos et al. (1970) reported the enhancement of a mouse tumor after combining blocking antisera with sensitized lymph node cells and then mixing the treated lymphocytes with tumor in several temporal combinations. They suggested that the enhancement might be explicable if one assumed generation from the sensitized lymphocytes of an immunosuppressant mediator. Since the putative mediator was not isolated, characterized, or defined in any way, the idea need not compel our serious attention at this time. However, the recent preliminary reports of Brunner (1970), Brawn ( 1971), and Sjogren et al. (1971) indicate that the immunosuppressant might be antigen that has complexed with antibody and inhibits cytotoxicity of effector cells.

E. PERSONAL PREFERENCE The theory we would like to offer is based upon certain features of antigenic density on cell surfaces, peculiarities of yG immunoglobulins, and some unknown contributions by the enhanced host. Nothing in this offering is claimed to be novel or complete, but the theory does explain many, if not all, of the phenomena of IB. It assumes that IB can occur only in a host that is making an immune response (Kaliss, 1958, 1969). Slowly moving yG is a most suitable agent to mediate IB. It is capable of binding avidly to antigen and, therefore, is not readily dissociated from the cell surface; it has a half-life of 5 to 15 days in different species and is, thus, available for sufficient time to allow a graft to take (S. Cohen and Milstein, 1967). It may inhibit the production of yM, which is usually the first and most cytotoxic antibody elicited by antigen (Uhr and Moller, 1968). It will not cause passive cutaneous anaphylaxis, i.e., it is not bound, as is yG,, to tissues and mast cells because of its Fc chain and, therefore, is not diverted from antigen. Most importantly, it does not activate C unless two or more heavy chains, in relatively close proximity to each other, are available on the surface of a membrane to bind C l q (Borsos and Rapp, 1965; Boyse et d.,1968c; S. Cohen, 1968; Linscott, 1970). [An alternative pathway of C activation has recently been documented and bypasses the usual antigen-antibody-Clq aggregate that activatcs the C system ( Gotze and Muller-Eberhard, 1971) .] If these are, indeed, the requirements of an immunoglobulin to participate in IB, then F( ab’):! light chains of any immunoglobulin should

IMMUNOLOGICAL ENHANCEMENT

203

be effective in binding to antigen without activating C. Chard (1968) has reported just this, i.e., the light chains of a fast moving yG did, indeed, facilitate IB. One might predict that the light chains of yM or yA, if they could be isolated and bound to cell surface antigens without activating C, would also enable IB to occur. Of equal importance are the density and spatial distribution of cell surface antigens. Several investigators (E. Mijller and Mdler, 1962; G. Moller and Moller, 1962, 1966; Winn, 1962; E. Moller, 1963; G. Moller, 1963d; E. Mdler et al., 1968; Leikola and Pasanen, 1970; Linscott, 1970) have designed suitable experiments that cogently indicate the importance of antigenic density and spatial arrangement on the cell surface membrane. Linscott, in particular, has presented a strong case for the importance of antigenic density and lack of C fixation in IB. The spatial arrangement and sparsity of different antigens on lymphoid cell surfaces have been clearly illustrated by Aoki et al. (1969,1970) and by Kourilsky et al. (1971). Parenthetically, it has been calculated that transplantation antigens constitute a small fraction of all the protein in cell surface membranes (Reisfeld and Kahan, 1970; Kourilsky et al., 1971), and this sparsity of antigenic content would reinforce the concept of how IB operates. It is worth observing that IB is more difficult to achieve when the genetic disparity between donor and recipient exists at the strong H-2 locus (in the mouse, for example), whereas when minor antigenic differences exist, such as H-1 or H-3, or semiallogeneic differences, as between F, and parent, IB is more easily achieved. Is this because there is less antigen and it is sparsely distributed on cell surfaces? The host is the third, but by no means the least, part of the equation that contributes to IB. Here we have little firm data to propose except to reiterate that IB is readily discernible in hosts whose immunological machinery is either incomplete or damaged. The host is the source of cellular immunity and, at present, methods are being devised to measure this kind of immunity (Cerottini et al., 1970a,b; Henney, 1971; Nordin et al., 1971a,b). Small numbers of sensitized lymphocytes, apparently derived from the thymus, appear to be the effector elements in cellular immunity (Perlmann and Holm, 1969; Jimenez et al., 1971; Cerottini et al., 1971; Henney, 1972). We have calculated the number of effector lymphocytes, using a modified Brunner assay, in spleens and lymph nodes of neonatal and adult rats bearing enhanced skin grafts (Peter and Feldman, 1971). In neonates the number is considerably smaller than in adults and is further reduced by blocking yG antibodies. Similarly, thymectomized adults have fewer effector elements than intact adult rats, and blocking antibodies decrease the number still further. One might imagine, then, that IB is effected by a constellation of

204

JOSEPH D. FELDMAN

events operating in a temporally ordered fashion. Once yG antibodies are available, either by active or passive immunization, they may bind to antigen to effect both afferent and efferent blockades. Simultaneously, antibody complexed to antigen may effect central blockade, either by occupying functional antigenic receptor sites on sensitized cells, by diminishing the number of effector cells, or by feedback inhibition of both effector cell proliferation and production of cytotoxic antibodies. Whether IB or rejection will take place will depend upon the interplay and temporal sequence of these processes, and any temporary interruption of the rejection process will allow the expression of extended survival. In the living animal, immunization by grafts seems to be deferred by passive infusion of blocking antibodies (Jones et al., 1972; Peter and Feldman, 1972). A full immune response may develop after the blocking antibodies are destroyed or after new H antigens are synthesized. Once immunity has been established in the recipient, blocking antibodies are ineffective, unless the immune response is deficient as a result of immunological depression and a reduction of effector lymphocytes. VIII. Prospects

Immunological blockade has an unanticipated potential to influence tumorigenesis, prolong survival of normal tissue grafts, and ameliorate the pathology associated with DH and autoimmune diseases. If the progression of cancer is truly associated with the presence of blocking antibodies, then regulation by biological means of spread and growth of neoplasms becomes possible. This might entail exaggerating the immune response to foreign antigens of tumors or inhibiting the blocking action of certain kinds of antibodies. The achievement of indefinite survival of normal tissue homografts remains a goal of immunologists, geneticists, and surgeons, but, most importantly, of those people in need of new organs for diseased ones. If alloantisera in minute quantities can aid this biological problem, then intensified research at all levels is warranted. Furthermore, the pathology associated with the activity of some immune cells has not been fully appreciated nor its extent properly delineated. An augmented investigation of viruses, resistant bacteria, and pollutants in our environment, all of which may contribute to diseases of cellular immunity, may help to uncover the extent of discomfort and disability that issue from consequences of this type of immunological activity-IB may be one of several possible solutions. It is presumptuous to prophesy the future, but it is reasonable to hope that tumor-specific, organ-specific, and H antigens can be isolated and purified; that antisera of the appropriate immunoglobulin classes can be manufactured specifically for such antigens; and that the person

IMMUNOLOGICAL ENHANCEMENT

205

or animal can be mildly manipulated to suppress slightly his immune responses. All three of these ingredients of IB are feasible projects for successful solution. And the baleful effects of cellular immunity may be mitigated or even abolished.

REFERENCES Alexander, P. ( 1970). In “The Biology and Surgery of Tissue Transplantation” (J. M. Anderson, ed.), p. 151. Davis, Philadelphia, Pennsylvania. Allen, H. L., Williams, R. D., Lovingood, C. G., and Ellison, E. H. (1952). Ann. Surg. 135, 239. Amos, D. B., Prioleau, W. H., Jr., and Hutchin, P. (1968). J. Surg. Res. 8, 122. Amos, D. B., Cohen, I., and Klein, W. J., Jr. (1970). Transplant. Proc. 2, 68. Aoki, T., Hhmerling, U., de Harven, E., Boyse, E. A., and Old, L. J. (1969). J . Exp. Med. 130, 979. Aoki, T., Boyse, E. A., Old, L. J., de Harven, E., Hammerling, U., and Wood, H. A. (1970). Proc. Nut. Acad. Sci. U. S . 65, 569. Asherson, G. L., and Stone, S. H. (1965). Immunology 9, 205. Attia, M. A. M., and Weiss, D. W. (1966). Cancer Res. 26, 1787. Axelrad, M. A. (1968). Immunology 15, 159. Axelrad, M. A,, and Rowley, D. A. (1968). Science 160, 1465. Bachvaroff, R., Galdiero, F., and Grabar, P. ( 1969). 1. lmmunol. 103, 953. Bailey, D. W., and Hoste, J. (1971). Transplantation 11, 404. Balner, H., Old, L. J., and Clarke, D. A. (1962). Proc. SOC. E x p . B w l . Med. 109, 58. Bardawill, W. A., Mitchell, G. W., Jr., McKeogh, R. P., and Marchant, D. J. (1962). Amer. 1. Obstet. Gynecol. 84, 1283. Barker, C. G., and Billingham, R. E. (1971 ). J. Exp. Med. 133, 620. Barth, R. F. (1969). J. Immunol. 103, 648. Barth, R. F., Hunter, R. L., Southworth, S., and Rabson, A. S. (1969). J. Immunol. 102, 932. Bashford, E. F., Murray, J. A., and Haaland, M. (1908). S c i . Rep. Imp. Cancer Res. Fund, London, 3 4 p. 376. Batchelor, J. R. (1963). Guy’s Hosp. Rep. 112, 345. Batchelor, J. R. (1968). Cancer Res. 28, 1410. Batchelor, J. R., and Howard, J. G. (1965). Transplantation 3, 161. Batchelor, J. R., and Silverman, M. S. (1962). Tramplent., Ciba Found. Symp., 1961, p. 216. Berke, G., Yagil, G., Ginsburg, H., and Feldman, M. (1969). Immunobgy 17, 723. Beme, B. H. (1965). Proc. SOC. E x p . Biol. Med. 118, 228. Bemstein, I., and Wright, P. W. ( 1971). Nature ( N e w Biol.) 231, 25. Biesecker, J., Fitch, F. W., Stuart, F. P., Hellstrom, K. E., and Hellstriim, I. (1971). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 30, 358 (abstr.). BildsZe, P., Ford, W. L., Pettirossi, O., and Simonsen, M. (1971). Transplantation 12, 189. Billingham, R. E. ( 1964). N . EngE. J. Med. 270, 667. Billingham, R. E. ( 1971). Cell. Immunol. 2, 1. Billingham, R. E., and Parkes, A. S. (1955). Proc. Roy. Soc., Ser. B 143, 550. Billingham, R. E., and Sparrow, E. M. (1955). J. Embryol. Exp. Morphol. 3, 265. Billingham, R. E., Brent, L., and Medawar, P. B. (1956). Transplant. B d . 3, 84.

206

JOSEPH D. FELDMAN

Billingham, R. E., Brent, L., Brown, J. E., and Medawar, P. B. (1959). Transplant. Bull. 6, 410. Billingham, R. E., Palm, J., and Silvers, W. K. (1965). Science 147, 514. Biozzi, G., Asofsky, R., and Lieberman, R. (1970). J. E x p . Med. 132, 752. Blair, P. B., Kripke, M. L., Lapp&, M. A., Bonhag, R. S., and Young, L. (1971). I. lmmunol. 106, 364. Bloch, K. J. (1965). Fed. Proc., Fed. Amer. SOC. E r p . Biol. 24, 1030. Bloom, E. T., and Hildemann, W. H. (1970). Transplantation 10, 321. Borsos, T., and Rapp, F. (1965). Science 150, 505. Boyse, E. A. (1971). I n “Immune Surveillance” (R.T. Smith and M. Landy, eds.), p. 7. Academic Press, New York. Boyse, E. A., Old, L. J., and Stockert, E. (1962). Nature (London) 194, 1142. Boyse, E. A., Old, L. J., and Stockert, E. (1968a). Proc. Nut. Acad. Sci. U . S. 60, 886. Boyse, E. A., Miyozawa, J., Aoki, T., and Old, L. J. (196813). Proc. Roy. SOC., Ser. B 170, 175. Boyse, E. A., Stockert, E., and Old, L. J. ( 1 9 6 8 ~ )J. . Exp. Med. 128, 85. Brawn, R. J. ( 1971). Proc. Nat. Acad. Sci. U . S . 68, 1634. Brent, L., and Medawar, P. B. ( 1962). Proc. Roy. SOC., Ser. B 155, 392. Brent, L., Courtenay, T., and Gowland, G. (1967). Nature (London) 215, 1461. Breyere, E. J. (1967). Transplantation 5, 1504. Breyere, E. J., and Barrett, M. K. (1960). J. Nut. Cancm lnst. 2.4, 699. Breyere, E. J., and Barrett, M. K. (1961). J. Nut. Cancer lnst. 27, 409. Breyere, E. J., and Burkoe, S. 0. (1963). J. Nut. Cancer lnst. 31, 179. Broder, S., and Whitehouse, F., Jr. (1968). Science 162, 1494. Brody, J. I., and Beizer, L. H. (1963). Blood 22, 139. Brondz, B. D. ( 1965). Immunology 14, 356. Brunner, K. T. (1970). In “The Biology and Surgery of Tissue Transplantation” (J. M. Anderson, ed.), p. 163. Davis, Philadelphia, Pennsylvania. Brunner, K. T., Mauel, J,, and Schindler, R. ( 1967). Nature (London) 213, 1246. Brunner, K. T., Mauel, J., Cerottini, J.-C., and Chapuis, B. (1968). lmmunology 14, 181. Bubenik, J., and Koldovsk?, P. (1965). Folia Biol. (Prague) 11, 258. Bubenik, J., Ivinyi, J., and Koldovskf, P. ( 1965). Folia Biol. (Prague) 11, 426. Buckley, R. H., Amos, D. B., Kremer, W. B., and Stickel, D. L. (1971a). N . Engl. J. Med. 285, 1035. Buckley, R. H., Schiff, R. I., and Amos, D. B. (1971b). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 30, 358 (abstr. ). Canty, T. G., and Wunderlich, J, R. (1970). J. Nut. Cancer Inst. 45, 761. Caren, L. D., and Rosenberg, L. T. (1965). Immunology 9, 359. Casey, A. E. (1932). Proc. SOC.E x p . Biol. Med. 29, 816. Casey, A. E. (1933). Proc. SOC. Exp. Biol. Med. 31, 663. Casey, A. E. (1941). Cancer Res. 1, 134. Casey, A. E., Ross, G. L., and Langston, R. R. (1949). Proc. SOC. E x p . Biol. Men. 72, 83. Casey, A. E., Casey, J. G., Hathaway, C. O., and Dowling, E. A. (1959). Proc. SOC. Exp. Biol. Med. 190, 762. Ceglowski, W. S., and Friedman, H. (1968). J. Nut. Cancer lnst. 40, 983. Ceppellini, R., Bigliani, S., Curtoni, E. S., and Leigheb, G. (1969). Transplant. Proc. 1, 390.

IMMUNOLOGICAL ENHANCEMENT

207

Cerilli, G. J., and Treat, R. C. (1969). Transplantation 8, 774. Cerottini, J.-C., McConahey, P. J., and Dixon, F. J. (1969). J. Immunol. 102, 1008. Cerottini, J.-C., Nordin, A. A,, and Brunner, K. T. (1970a). Nature (London) 227, 72. Cerottini, J.-C., Nordin, A. A., and Brunner, K. T. (197Ob). Nature (London) 228, 1308. Cerottini, J.-C., Nordin, A. A., and Brunner, K. T. (1971). J. Exp. Med. 134, 553. Chambers, H., and Scott, G. M. (1924). Brit. J. Exp. Pathol. 5, 1. Chambers, H., and Scott, G. M. (1926). BTit. J. Exp. Pathol. 7, 33. Chantler, S. M. ( 1967). Transplantation 5, 1400. Chantler, S. M., and Batchelor, J. R. (1964). Transplantation 2, 75. Chapuis, B., and Brunner, K. T. (1971). Int. Arch. Allergy 40, 321. Chard, T. (1968). Immunology 14, 583. Chard, T., French, M. E., and Batchelor, J. R. (1967). Transplantation 5, 1266. Chouroulinkov, I., Boyse, E. A., and Old, L. J. (1962). PTOC. SOC. Exp. Biol. Med. 111, 263. Chutni, J. ( 1968). Nature (London) 217, 175. Chutnl, J. (1971). Transplantation 12, 28. Chutnl, J., and Pokornk, Z. ( 1967). Folia Biol. (Prague) 13, 68. Chutni, J., and Rychlikovl, M. (1964). Folia Biol. (Prague) 10, 177. Cock, A. G. ( 1962a). Transplant. Bull. 29, 95. Cock, A. G. (1962b). Immunology 5, 169. Coe, J. E. ( 1966). J. Immunol. 96, 744. Cohen, I. R., Globerson, A,, and Feldman, M. (1971a). J . E x p . Med. 133, 821. Cohen, I. R., Brautbar, H., and Nelken, D. (1971b). Transplantation 11, 135. Cohen, M., and Nelken, D. (1970). Clin. Exp. lmmunol. 6, 137. Cohen, S. (1968). J. Immunol. 100, 407. Cohen, S., and Milstein, C. (1967). Advan. Immunol. 7, 1. Coppola, E. D., and Villegas, G. R. (1967). Proc. SOC. Exp. Biol. Med. 125, 1071. Corson, J. M. (1970). Transplantation 10, 491. Crisler, C., and Frank, M. M. (1965). Fed. PTOC.,Fed. Amer. SOC. Exp. Biol. 24, 447. Crowle, A. J., and Hu, C. C. (1965). J. Immunol. 94, 555. Crowle, A. J., and Hu, C. C. (1968). Proc. SOC. E x p . Biol. Med. 127, 190. Crowle, A. J., and Hu, C. C. (1969). J. Allergy 43, 209. Cruse, J. M. (1969a). 66th Annu. Meet., Sci. PTOC. Amer. Assn. Puthol. Bacteriol., p. 9a. Cruse, J. M. (196913). 66th Annu. Meet., Sci. PTOC.AWT. Assn. Pathol. Bacteriol., p. 83a. Cruse, J. M., Germany, W. W., and Dulaney, A. D. (1965). Lab. Invest. 14, 1554. Cudkowicz, G., and Bennett, M. (1969). PTOC.Amer. Ass. Cancer Res. 10, 16. Denman, A. M., and Frenkel, E. P. (1968). Immunology 14, 1. Eidinger, D., Khan, S . A,, and Millar, K. G. (1968). J. Exp. Med. 128, 1183. Elkins, W. L. (1966). J . Exp. Med. 123, 103. Elkins, W. L. (1970). Transplantation 9, 273. Ellman, L., Green, I., Martin, W. J., and Benacerraf, B. (1970). PTOC.Nut. Acad. Sci. U. S . 66, 322. Fass, L., Ziegler, J. L., Herberman, R. B., and Kiryabwire, J. W. M. (1970). Lancet 1, 116. Feldman, J. D., and Najarian, J. S. (1963). J. Immunol. 91, 306.

208

JOSEPH D. FELDMAN

Feldman, J. D., Pick, E., Lee, S., Silvers, W. K., and Wilson, D. B. (1968). Amer. J . Pathol. 52, 687. Feldman, M., and Globerson, A. (1960). J. Nut. Cancer Znst. 25, 631. Ferrer, J. F. (1968a). Transplantation 6, 160. Ferrer, J. F. (1968b). Transplantation 6, 167. Flax, M. H., and Waksman, B. H. (1962). J. Immunol. 89, 496. Flexner, S., and Jobling, J. W. (1907a). Proc. SOC. Erp. Biol. Med. 4, 156. Flexner, S., and Jobling, J. W. (1907b). Proc. SOC. Exp. Biol. Med. S9 16. Forbes, J. T., Lewis, C. K., Cruse, J. M., and Sprunt, D. H. (1971). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 30, 358 (abstr.). Ford, W. L., and Simonsen, M. (1971). J . Exp. Med. 133, 938. Ford, W. L., Burr, W., and Simonsen, M. (1970). TranspZantution 10, 258. French, M. E., and Batchelor, J. R. (1969). Lancet 2, 1103. French, M. E., Batchelor, J. R., and Watts, H. C. (1971). Transplantation 12, 45. Fridman, E. H., and Kourilsky, F. M. (1969). Nature (London) 224, 277. Gasser, D. L. (1969). J . Zmmunol. 103, 66. Gasser, D.L., and Silvers, W. K. (1971). J . Zmmunol. 106, 875. Gershon, R. K., Carter, R. L., and Kondo, K. (1968). Science 159, 646. Gewurz, H., Clark, D. S., Cooper, M. D., Varco, R. L., and Good, R. A. ( 1967). Transplantation 5, 1296. Cillespie, C.Y., Cruse, J. M., and Crawford, L. V. (1971). Fed. Proc., Fed. Amer. SOC. E x p . Biol. 3.0, 650 (abstr.). Ginsburg, H., Hollander, N., and Feldman, M. (1971). J . Exp. Med. 134, 1062. Gordon, R. O., Stinson, E. B., Souther, S. C.,and Oppenheim, J. J. ( 1971). Transplantation 12, 484. Gorer, P. A. (1942). J . Pathol. Bacteriol. 54, 51. Gorer, P. A. (1956). Aduan. Cancer Res. 4, 149. Gorer, P. A. (1958). Ann. N . Y. Acad. Sci. 73, 707. Gorer, P. A. ( 1961). Aduan. Zmmunol. 1, 345. Corer, P. A., and Kaliss, N. ( 1959). Cancer Res. 19, 824. Cijtze, O., and Miiller-Eberhard, H. J. (1971). J. Exp. Med. 134, 905. Green, H. N., and Wilson, R. (1956). Nature (London) 178, 851. Green, I., Inman, J. K., and Benacerraf, B. (1970). Proc. Nut. Acad. Sci. U. S. 66, 1267. Grob, P. J., and Inderbitzin, T. M. (1969). Int. Arch. Allergy Appl. Zmmunol. 35, 141. Guiney, E. J,, Austen, K. F., and Russell, P. S. (1964). Proc. SOC. E x p . Biol. Med. 115, 1113. Guttmann, R. D., and Lindquist, R. R. (1971). Transplant. Proc. 3, 717. Guttmann, R. D., Carpenter, C. B., Lindquist, R. R., and Merrill, J. P. (1967). J . Elcp. Med. 126, 1099. Haaland, M. (1910). Proc. Roy. SOC., Ser. B 82, 293. Halasz, N. A. (1983). J . Swg. Res. 10, 503. Halasz, N. A., and Orloff, M. J. (1965). J . Immunol. 94, 253. Halasz, N. A., Orloff, M. J., and Hirose, F. (1964). Transplantation 2, 453. Hall-Allen, R. T. J,, Scharli, A. F., Maggs, P. R., Busch, G. J., Simonian, S. J., and Wilson, R. E. (1970). Transplantation 10, 472. Harding, B., Puditin, D. J., Gotch, F., and MacLennan, I. C. M. ( 1971). Nature ( N e w Biol.) 232, 80. Hardy, M. A., Quint, J., and State, D. (1969). Transplantation 7, 223.

IMMUNOLOGICAL ENHANCEMENT

209

Harris, E. M., and Eakin, R. M. (1949). J. Exp. Zool. 112, 131. Hdek, M., Skamene, E., Karakoz, I., Chutni, J., Nouza, K., Bubenik, J., SovovB, V., NBmec, M., and Jonik, J. ( 1968). Folia B i d . ( PragZre) 14, 411. Habkovl, V., and Svoboda, J. (1962). Zn “Mechanisms of Immunological Tolerance” (M. HaHek, A. Langerovi, and M. Vojtiskovi, eds.), p. 431. Publ. House Czech. Acad. Sci., Prague. Habkovi, V., Chutnl, J., and Biege, G. (1965). Folia Biol. (Prague) 11, 446. Haughton, G., and Nash, D. R. (1969). Transplant. PTOC.1, 616. Hellman, K., and Duke, D. (1967). Transplantation 5, 184. Hellstrijm, I. (1967). Int. J. Cancer 2, 65. Hellstrom, I., and Hellstriim, K. E. (1970). Znt. J. Cancer 5, 195. Hellstrom, I., and Hellstrom, K. E. (1971). Nature (London) 230, 49. Hellstrijm, I., and Sjogren, H. 0. (1965). E x p . Cell Res. 40, 212. Hellstrom, I., Hellstrbm, K. E., Pierce, G. E., and Yang, J. P. S. (1968a). Nature (London) 220, 1352. Hellstriim, I., Hellst~m,K. E., Pierce, G. E., and Bill, A. H. (1968b). PTOC. Nut. Acad. Sci. U . S. 60, 1231. Hellstrom, I., Hellstrdm, K. E., and Pierce, G. E. ( 1 9 6 8 ~ ) .Znt. 1. Cancer 3, 467. Hellstrom, I., Hellstrijm, K. E., Evans, C. A., Heppner, C. H., Pierce, G. E., and Yang, J. P. S. (1969). PTOC. Nut. Acud. Sci. U. S. 62, 362. Hellstriim, I., Hellstrom, K. E., Storb, R., and Thomas, E. D. (1970a). Proc. Nut. Acad. Sci. U . S . 66, 65. Hellstrijm, I., Hellstriim, K. E., and Sjogren, H. 0. (1970b). CeU. Zmmunol. 1, 18. Hellstrom, I., Hellstriim, K. E., and Warner, G. A. (1971a). Znt. 1. Cancer 7, 1. Hellstrom, I., Sjogren, H. O., Warner, G., and Hellstrom, K. E. (1971b). Znt. 1. Cancer 7 , 226. Hellstrom, K. E., and Hellstrom, I. (1969). Aduan. Cancer Res. 12, 167. Hellstriim, K. E., and Hellstriim, I. (1970). Annu. Rev. Microbiol. 2.4, 373. Hellstrom, K. E., Hellstriim, I., and Brawn, J. (1969). Nature (London) 224, 914. Henney, C. S. (1970). J. Zmmunol. 105, 919. Henney, C. S. (1971). J . Immunol. 107, 1558. Henney, C. S., and Nordin, A. A. ( 1971). 1. Immunol. 106, 20. Henry, C., and Jerne, N. K. (1968). J. Exp. Med. 128, 133. Heppner, G. H. ( 1969). Znt. J . Cancer 4, 608. Herzenberg, L. A., and Gonzales, B. (1962). Proc. Nut. Acud. Sci. U . S. 48, 570. Heslop, B. F. (1966). Transplantation 4, 32. Heslop, B. F. (1969). Transplant. Proc. 1, 560. Heslop, B. F. ( 1971). Transplantation 11, 497. Hildemann, W. H. (1971 ). TTUnsplUnt. Proc. 3, 76. Hildemann, W. H., and Walford, R. L. (1960). Ann. N. Y. Acud. Sci. 87, 56. Hornung, M. O., Layne, E., and McDonald, J. C. ( 1971). J. Immunol. 107, 979. Howard, R. J., Mergenhagen, S. E., Notkins, A. L., and Dougherty, S. F. (1969). Transplant. Proc. 1, 586. Hughes, L. E., and Lytton, B. (1964). Brit. Med. J. 1, 209. Hutchin, P., Amos, D. B., and Prioleau, W. H., Jr. (1967). Transphntation 5, 68. Irvin, G . L., I11 (1972). J. Zmmunol. 108, 207. Irvin, G . L., 111, and Eustace, J. C. (1970). Transplantation 10, 555. Irvin, G . L., 111, and Eustace, J. C. (1971). J. Zmmunol. 106, 956. Irvin, G. L., 111, Eustace, J. C., and Fahey, J. L. (1967). J. Immunol. 99, 1085. Jankovib, B. D., and Flax, M. H. (1963). J. Zmmunol. 90, 178.

210

JOSEPH D. FELDMAN

Jeekel, J. J., and Westbrock, D. L. (1971). Transplant. PTOC.3, 714. Jeekel, J. J., McKenzie, I. F. C., and Winn, H. J. (1972). 1. Immunol. 108, 1017. Jehn, U. W., Nathanson, L., Schwartz, R. S., and Skinner, M. (1970). N. Engl. 3. Med. 283, 329. Jenkins, A. M., and Woodruff, M. F. A. (1971). Transplantation 12, 57. Jerusalem, C., Weiss, M. L., and Poels, L. ( 1971). J. Immunol. 107, 260. Jimenez, L., Bloom, B. R., Blume, M. R., and Oettgen, H. F. (1971). J. Exp. Med. 133, 740. Jones, J. M., and Feldman, J. D. (1971a). Unpublished results. Jones, J. M., and Feldman, J. D. (1971b). Transplantation 12, 114. Jones, J. M., Peter, H.-H., and Feldman, J. D. (1972). J . Immunol. 108, 301. Jose, D. G., Kersey, J. H., Choi, Y. S., Biggar, W. D., Gatti, R. A,, and Good, R. A. (1971). Lancet 2,841. Kaliss, N. (1956). Proc. Nut. Acad. Sci. U. S . 42, 269. Kaliss, N. (1957). In “La biologie des homogreffes,” p. 225. CNRS, Paris. Kaliss, N. (1958). Cancer Res. 18, 992. Kaliss, N. (1962). Ann. N. Y. Acad. Sci. 101, 64. Kaliss, N. (1966). Ann. N. Y. Acad. Sci. 129, 155. Kaliss, N. (1968). Proc. SOC. Exp. Biol. Med. 129, 83. Kaliss, N. (1969). Int. Reo. Exp. Puthol. 8, 241. Kaliss, N. (1970). Transplant. Proc. 2, 59. Kaliss, N., and Bryant, B. F. (1958). J . Nut. Cancer Inst. 20, 691. Kaliss, N., and Dagg, M. K. (1964). TTansplantation 2, 416. Kaliss, N., and Kandutsch, A. A. (1956). Proc. SOC. E x p . Biol. Med. 91, 118. Kaliss, N., and Molomut, N. (1952). Cancer Res. 12, 110. Kaliss, N., and Rubinstein, P. ( 1968). In “Advance in Transplantation” (J. Dausset, J. Hamburger, and G. Math&,eds.), p. 531. Munksgaard, Copenhagen. Kaliss, N., and Suter, R. B. ( 1968). Transplantation 6, 844. Kaliss, N., Dagg, M. K., and Stinipfling, J. H. (1963). Transplantation 1, 535. Kamoun, P. P., and Hamburger, J. (1970). Transplantation 10, 53. Kandutsch, A. A. (1960). Cancer Res. 20, 264. Keller, R., and Jones, V. (1971). Lancet 2, 847. Klaue, P., and Jolley, W. B. (1971). Transplantation 11, 30. Klein, E., and Sjogren, H. 0. (1960). Transplant. Bull. 26, 442. Klein, W. J., Jr. ( 1971). 3. Exp. Med. 134, 1238. Kodama, H., Sachin, I. N., and Shabart, E. J. (1967). Transplantation 5, 1450. Kourilsky, F. M., Silvestre, D., Levy, J. P., Dausset, J., Nicolai, M. G., and Senik, A. ( 1971) . J. Immunol. 106, 454. Krohn, P. L. (1959). In “Transplantation of Tissues” ( L . Peer, ed.), Vol. 2, p. 401. Williams & Wilkins, Baltimore, Maryland. Lance, E. M., Cooper, S., and Boyse, E. A. (1971). Cell. Immunol. 1, 536. Lang, W., Nase, S., and Rajewsky, K. (1969). Nature (London) 223, 949. Lanman, J. T., Herod, L., and Fikrig, S. (1964). J. E x p . Med. 119, 781. Lapp, W. S., and Bliss, J. Q. (1966). Transplantation 4, 754. Lapp, W. S., and Bliss, J. Q. (1970). Transplantation 10, 389. Lapp, W. S., and Moller, G. (1969). Immzcnology 17, 339. Lapp& M. A., Graff, R. G., and Snell, G. D. (1969). Transplantation 7 , 372. Leikola, J., and Pasanen, V. J. (1970). Int. Arch. Allergy Appl. Immunol. 39, 352. LengerovL, A., and Matousek, V. (1968). Folia Biol. (Prague) 14, 1.

IMMUNOLOGICAL ENHANCEMENT

211

LengerovA, A., and VojtiBkovi, M. ( 1963). Folia Biol. (Prague) 9, 72. Liacopoulos, M., Neven, T., Biozzi, G., and Halpern, B. ( 1962). C . R. Acad. Sci. 254, 3765. Linder, 0. E. A. (1962). Immunology 5, 195. Linder, 0. E. A. (1961). J . Nut. Cancer Inst. 27, 351. Lindquist, R. R., Guttmann, R. D., Carpenter, C. B., and Merrill, J. P. (1969). Transplantation 8, 545. Linscott, W. D. (1970). Nature (London) 228, 824. Lonai, P., and Feldman, M. (1970). Transplantation 10, 372. Lucas, Z. J., Markley, J., and Travis, M. (1970). Fed. PTOC.,Fed. Amer. SOC. E x p . Biol. 29, 2041. McConahey, P. J., Cerottini, J.-C., and Dixon, F. J. (1968). J . E x p . Med. 127, 1003. McDevitt, H. O., and Benacerraf, B. (1969). Adoan. Immunol. 11, 31. McDevitt, H. O., and Chinitz, A. (1969). Science 163, 1207. McDevitt, H. O., and Tyan, M. L. (1968). J . E r p . Med. 128, 1. Mackaness, G. B. (1969). J . Exp. Med. 129, 973. McKenzie, I. F. C., Koene, I., and Winn, H. J. ( 1971). Transplant. PTOC.3, 711. MacLennau, I. C. M., and Harding, B. (1970). Immunology 18, 397 and 405. MacLennan, I. C. M., and Loewi, G. ( 1968). Nature (London) 219, 1069. Mandel, M. A., and DeCosse, J. J. (1969). J. Immunol. 103, 1288. Marquet, R. H., Heystek, G. A., and Tinbergen, W. J. (1971). Transplant. PTOC. 3, 708. Martin, W. J., and Miller, J. F. A. P. (1968). J. Exp. Med. 128, 855. Mason, S., and Warner, N. L. (1970). J . Immunol. 104, 762. Mauel, J., Rudolf, H., Chapuis, B., and Brunner, K. T. (1970). Immunology 18, 517. Mayer, D. J., Kronman, B., and Dumont, A. E. ( 1965). SUTg. Forum 16, 243. Medawar, P. B. (1963). Transplantation 1, 21. Milgrom, F. (1970). Fed. PTOC.,Fed. Amer. SOC. E x p . B i d . 29, 114. Miller, J., Hattler, B., Davis, M., and Johnson, M. C. (1971). Transplantation 12, 65. Mishell, R. I., Herzenberg, L. A., and Herzenberg, L. A. (1963). J . Immunol. 90, 628. Mitchison, N. A,, and Dube, 0. L. (1955). J. Exp. Med. 102, 179. Moller, E. (1963). Transplantation I, 165. Miiller, E. (1964). J . Nut. Cancer Inst. 33, 979. Moller, E. (1965a). J . E x p . Med. 122, 11. Moller, E. (1965b). J. Nut. Cancer Inst. 35, 1053. Moller, E., and Moller, G. ( 1962). J . Exp. Med. 115, 527. Moller, E., Britton, S., and Moller, G. (1968). In “Regulation of the Antibody Response” ( B . Cinader, ed.), p. 141. Thomas, Springfield, Illinois. Moller, G. (1963a). J. Nut. Cancer Inst. 30, 1153. Moller, G. (1963b). J. Nut. Cancer Inst. 30, 1177. Moller, G. ( 1 9 6 3 ~ )J.. Nut. Cancer Inst. 30, 1205. Moller, G. (1963d). J . Immunol. 90, 271. Moller, G. (1964a). Transplantation 2, 281. Moller, G. ( 1964b). Transplantation 2, 405. Moller, G. (1966). J . Immunol. 96, 430. Moller, G., and Moller, E. ( 1962). Ann. N . Y. Acad. Sci. 99, 504. Moller, G., and Moller, E. ( 1966). In “AntilJodies to Biologically Active Molecules” ( B. Cinader, ed. ), p. 349. Pergamon, Oxford.

212

JOSEPH D. FELDMAN

Morris, R., and Lucas, Z. J. ( 1971). Transplant. Proc. 3, 697. Moseley, R. V., Sheil, A. G. R., Mitchell, R. M., and Murray, J. E. (1966). Transplantation 4, 678. Miiller, M. (1967). Cancer Res. 27, 2272. Mullen, Y., and Hildemann, W. H. (1971). Transplant. Proc. 3, 669. Najarian, J. S., and Dixon, F. J. (1963). Proc. SOC. Exp. Biol. Med. 112, 136. Nelken, D., and Cohen, M. (1968). Nature (London) 218, 693. Nelken, D., Gabriel, M., and Boss, J. H. (1966). Lab. Inuest. 15, 1268. Nelson, D. S. (1962). Brit. J. E x p . Pathol. 43, 1. Nordin, A. A., Cerottini, J.-C., and Brunner, K. T. (1971a). Eur. 1. Immunol. 1, 55. Nordin, A. A,, Cerottini, J.-C., and Brunner, K. T. (1971b). Int. Arch. Allergy Appl. Immunol. 40, 321. Ockner, S. A., Guttmann, R. D., and Lindquist, R. R. (1970a). Transplantation 9, 30. Ockner, S. A., Guttmann, R. D., and Lindquist, R. R. (1970b). Transplantation 9, 39. Old, L.J., Stockert, E., Boyse, E. A., and Kim, J. H. (1968). J. Exp. Med. 127, 523. Oldstone, M. B. A,, and Dixon, F. J. (1970). J. Exp. Med. 131, 1. Oliver, C. B., and Feldman, J. D. (1971). Transplantation 11, 412. Ono, K., DeWitt, C. W., Wallace, J. H., and Lindsey, E. S. (1969). Transplantation 7, 122. Orr, W. McN., Birtch, A. G., Diethelm, A. G., Dubernard, J. M., Duquella, J., and Glassock, R. J. (1970). Clin. Exp. Immunol. 6, 305. Parka, A. S. (1968). Transplant. Bull. 5, 45. Paterson, P. Y. (1966). Aduan. Immunol. 5, 131. Payne, R. (1962). Blood 19, 411. Pearsall, N. N., and Weiser, R. S. (1968a). J . ReticuZoendotheZ. SOC. 5, 107. Pearsall, N. N., and Weiser, R. S. (196813).J . Reticuloendothel. SOC. 5, 121. Perlmann, P., and Holm, G. ( 1969). Aduan. Zmmunol. 11, 117. Perlmann, P., and Perlmann, H. (1970). Cell. Immunol. 1, 300. Perlmann, P., Perlmann, H., Wasserman, J., and Packalbn, T. (1970). Int. Arch. AUergV 3 4 204. Perlmann, P., Perlmann, H., and Biberfeld, P. (1972). J . Immunol. 108, 558. Peter, H.-H., and Feldman, J. D. (1972). J . Exp. Med. 135, in press. Phillips, M. E., and Stetson, C. A. (1962). Proc. SOC. Exp. Biol. Med. 111, 265. Phillips, M. E., Rother, U., and Rother, K. (1968). J. Immunol. 100, 93. Pierce, C. W. (1969). J . Exp. Med. 130, 365. Prehn, R. T. ( 1959). In “Biological Problems of Grafting” (F. Albert and P. B. Medawar, eds.), p. 163. Blackwell, Oxford. Prehn, R. T. (1963). J . Nut. Cancer Inst. 31, 791. Quadracci, L. J., Hellstrijm, I., Striker, G. E., Marchioro, T. L., and Hellstrom, K. E. ( 1971). CeU. Immunol. 1, 561. Rathbun, W. E., and Hildemann, W. H. (1970). J. Immunol. 105, 98. Reif, A. E., and Allen, J. M. (1964). J. Exp. Med. 120, 413. Reif, A. E., and Allen, J. M. (1966). Nature (London) 209, 521. Reisfeld, R. A., and Kahan, B. D. (1970). Aduan. Immunol. 12, 117. Richardson, W. P., and Paterson, P. Y. (1970). Clin. Exp. Immunol. 6, 337. Rother, K., Rother, U., and Ballantyne, D. L. (1967). Proc. SOC. E x p . Biol. Med. 114, 439. Rother, U., Ballantyne, D. L., Jr., Cohen, C., and Rother, K. (1967). J. E x p . Med. 1% 565.

IMMUNOLOGICAL ENHANCEMENT

213

Ruddle, N. H., and Waksman, B. H. (1968). 1. Exp. Med. 128, 1237 and 1255. Ruizkiewicz, M. (1967). In “Specific Tumor Antigens” (R. J. C. Harris, ed.),p. 107. Munksgaard, Copenhagen. Ryder, R. J. W., Kilham, L. K., and Schwartz, R. S. (1969). TmnspZunt. Proc. 1, 524. Sabbadini, E. (1970). J. Reticuloendothel. SOC. 7, 551. Safford, J. W., Jr., and Tokuda, S. (1970). Proc. SOC. Exp. Biol. Med. 133, 651. Savel, H. (1969). Cancer 24, 56. Schlesinger, M., and Gotein, R. (1965). 1. Immunol. 95, 197. Schlesinger, M., and Yron, I. (1969). Science 164, 1412. Silvers, W. K. (1968). J. Exp. Med. 128, 69. Silvers, W. K., and Billingham, R. E. (1966). J. E x p . 2002.161, 413. Sin, Y. M., Sabbadini, E., and Sehon, A. H. (1971). Cell. Immunol. 2, 239. Sjogren, H. O., Hellstrom, I., Bansal, S. C., and Hellstrijm, K. E. (1971). Proc. Nat. Acad. Sci. U. S . 6 4 1372. Sloboda, A. E., and Landes, J. (1970). J. Immunol. 104, 185. Snell, G. D. (1954). J. Nut. Cancer Inst. 15, 665. Snell, G. D. (1956). Transplant. Bull. 3, 83. Snell, G. D. (1957). Annu. Reu. Microbiol. 11, 439. Snell, G. D. (1970). Surg., Gynecol. Obstet. 130, 1109. Snell, G. D., and Bunker, H. P. (1965). Transplantation 3, 235. Snell, G. D., and Stimpfling, J. H. (1966). In “Biology of the Laboratory Mouse” (E. L. Green, ed.), p. 457. McGraw-Hill, New York. Snell, G. D., Cloudman, A. M., Failor, F., and Douglas, P. (1946). 1. Nut. Cancer Inst. 6, 303. Snell, G. D., Cloudman, A. M., and Woodworth, E. F. (1948). J. Nut. Cancer Inst. 6, 429. Snell, G. D., Winn, H. J., Stimpfling, J. H., and Parker, S. J. (1960). J . Exp. Med. 112, 293. Snyder, G. B., Kinsky, R., and Voisin, G. A. (1964). P h t . Reconstr. Surg. 33, 110. Stark, R. B., and Dwyer, E. (1959). Surgey 46, 277. Steinmuller, D. (1967). Science 158, 127. Steinmuller, D. (1969). Transplant. Proc. 1, 593. Stewart, T. H. M. (1969). Cancer 23, 1368. Stuart, F. P., Saitoh, T., Fitch, F. W., and Spargo, B. H. (1968a). Surgery 64, 17. Stuart, F. P., Saitoh, T., and Fitch, F. W. (1968b). Science 160, 1463. Stuart, F. P., Fitch, F. W., Rowley, D. A., Biesecker, J. L., Hellstrom, K. E., and Hellstrom, I. (1971). Transplantation 12, 331. Tada, T., and Okumura, K. (1971). I. Immunol. 106, 1012. Takasugi, M., and Hildemann, W. H. (1969a). 1. Nut. Cancer. Inst. 43, 843. Takasugi, M., and Hildemann, W. H. (1969b). J. Nut. Cancer Inst. 43, 857. Takasugi, M., and Hildemann, W. H. ( 1 9 6 9 ~ )Transplant. . Proc. 1, 530. Taub, R. N. (1969). Transplant. Proc. 1, 445. Thoenes, G. H., White, E., and Hildemann, W. H. (1970). J. Immunol. 104, 1447. Tokuda, S., and McEntee, P. F. (1967). Transphntation 5, 606. Tung, K. S. K., Unanue, E. R., and Dixon, F. J. (1970). Amer. J. Pathol. 60, 313. Tung, K. S. K., Unanue, E. R., and Dixon, F. J. ( 1971). J. Immunol. 106, 1453. Uhr, J. W., and Moller, G. (1968). Aduan. Immunol. 8, 81. Vaz, N. M., and Levine, B. B. (1970). Science 168, 852. Vener, J., Martinez, C., and Good, R. A. ( 1961). Proc. Soc. Exp. Biol. Med. 106, 480.

214

JOSEPH D. FELDMAN

Villegas, G. R., and Coppola, E. D. (1968). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 27, 1825. Voisin, G . A. (1965). In “Immunopathology” (P. Grabar and P. Miescher, eds.), p. 165. Grune & Stratton, New York. Voisin, G. A., and Kinsky, R. (1962). Transplant., Ciba Found. Symp., 1961 p. 287. Voisin, G. A., Kinsky, R. G., and Jansen, F. K. (1966). Nature (London) 210, 138. Voisin, G. A,, Kinsky, R. G., and Maillard, J. (1968a). Ann. Inst. Pusteur, Paris 115, 855. Voisin, G. A., Kinsky, R., and Maillard, J. (1968b). Transplantation 6, 187. Voisin, G. A., Kinsky, R., Jansen, F., and Bernard, C. (1969). Transplantation 8, 618. Volk, H., Mauersberger, D., Rother, K., and Rother, U. (1964). Ann. N . Y. Acud. Sci. 120, 26. Wachtel, S . S., and Silvers, W. K. (1971). J. Exp. Med. 133, 921. Walford, R. L., Gallagher, R., and Troup, G. M. (1965). Transplantation 3, 387. Wegmann, T. G., Hellstrom, I., and Hellstrom, K. E. (1971). Proc. Nut. Acad. Sci. 6 4 1644. Weigle, W. 0.(1965). J. Exp. Med. 122, 1049. Weigle, W. O., and Nakamura, R. M. (1969). Clin. Exp. Immunol. 4, 643. Weiner, L. P., Cole, G . A., and Nathanson, N. (1971). J . Immunol. 106, 427. Werder, A. A,, and Hardin, C. A. (1954). Surgery 36, 371. Wheeler, H. B., DeFronzo, A., and Corson, J. M. (1970). Transplantation 9, 78. White, E., and Hildemann, W. H. (1968). Science 162, 1293. White, E., and Hildemann, W. H. (1969). TranspIant. Proc. 1, 395. White, E., Hildemann, W. H., and Mullen, Y. (1969). Transplantation 8, 602. Whitehouse, F., Jr., and Broder, S. (1968). Proc. SOC. Exp. Biol. Med. 127, 1064. Wilson, C. B., Dixon, F. J., Fortner, J. G., and Cerilli, G. J. (1971). J . Clin. Inuest. 50, 1525. Wilson, R. E., Rippin, A., Dagher, R. K., Kinneart, P., and Busch, G . J. (1969). Transplantation 7, 360. Wilson, R. E., Maggs, P. R., Wanwyck, R., Shaiponich, T., Hall-Allen, R. T. J., Luke, P., and Simonsen, S. J. (1971). Transplant. Proc. 3, 705. Winn, H. J. ( 1982). Ann. N . Y. Acad. Sci. 101, 23. Winn, H. J. (1970). Transplant. Proc. 2, 83. Witz, I. P., Yagi, Y.,and Pressman, D. (1968). J. Immunol. 101, 217. Witz, I. P., Yagi, Y.,and Pressman, D. ( 1969a). Proc. SOC. Exp. Biol. Med. 130, 928. Witz, I. P., Kaliss, N., and Pressman, D. (196913). J. Immunol. 102, 283. Witz, I. P., Kaliss, N., and Samuel, T. (1971). Cell. Immunol. 2, 362. Woglom, W. H. (1929). Cancer Rev. 4, 129. Wright, P. W. ( 1968). In “Advance in Transplantation” (J. Dausset, J. Hamburger, and G. Math&,eds.), p. 41. Munksgaard, Copenhagen. Zimmerman, B., and Feldman, J. D. (1969a). J. Immunol. 102, 507. Zimmerman, B., and Feldman, J. D. (1969b). J. Immunol. 103, 383. Zimmerman, B., and Feldman, J. D. (1970). J . Immunol. 104, 626. Zimmerman, C. E. (1971). TranspZant. Proc. 3, 701.

Genetics and Immunology of Sex-Linked Antigens DAVID 1. GASSER AND WlLLYS K. SILVERS lmmunobiology Research Unit, Departments of Medical Genetics and Pothology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania

. . . . . . . . . . . . .

I. Introduction 11. The Y Antigen in Mice . . . . . . . . . . . . A. Genetic Determination of Y Antigen B. Tissue Localization . . . . . . . . . . . C . Genetic Basis of Response to Y Antigen D. Induction of Tolerance to Y Antigen . . . . E. Graft-versus-Host Reactions and Y Antigen . . . F. Effect of Y Antigen on Placental Growth and Sex Ratio 111. The Xg Blood Group Locus in Man . . . . . . . . IV. Other Sex-Limited and Sex-Linked Antigens V. Summary . . . . . . . . . . . References . . . . . . . . . . .

. . .

. . .

. . .

. . .

. . .

. . .

.

.

.

.

.

.

. . . . . . . . .

215 216 216 220 222 227 233 234 236 238 242 243

I. Introduction

For many years the Y chromosome was thought to be largely, if not entirely, genetically inert. Muller in 1914 believed that the lack of any known mutations on this chromosome indicated that either ( a ) the genes on the Y chromosome did not mutate, ( b ) these mutations were all recessive to dominant alleles on the X chromosome, or ( c ) Y-linked genes were degenerate or entirely absent. The fact that in Drosophilu, as well as in man and other mammals, individuals without a Y chromosome (XO) were viable, whereas those without an X were not, seemed to support the idea of an inactive Y chromosome. The first demonstration of a Y-linked gene was reported by Stem (1927, 1929), who showed that in Drosophilu a wild-type allele of bobbed, a gene located on the X chromosome, was represented on the Y. Subsequently, other examples of the genetic activity of the Y chromosome of this species were demonstrated (Schultz, 1956; Brosseau, 1960; Hess and Meyer, 1963), and it became quite clear that in spite of the fact that the Y chromosome of Drosophila consists entirely of heterochromatin, it is not genetically inert. Nevertheless, as recently as 1957, Stern concluded that not a single function could convincingly be attributed to the Y chromosome of man. Within a few years this situation too changed as it was shown that maleness in man is determined by the Y chromosome (Jacobs and Strong, 215

216

DAVID L. GASSER AND WILLYS K. SILVERS

1959; Ford et al., 1959). Convincing evidence has also been obtained that hypertrichosis of the pinnae of the ears is Y-linked (Dronamraju, 1964, 1965). The purpose of this article is to review the literature pertaining to an antigen associated with the Y chromosome of the mouse and probably with that of other species as well. Discussions of X-linked antigens are also included because of the possible homology between the X and Y chromosomes ( Ferguson-Smith, 1965, 1966; Ohno, 1969) and because of the fact that crossing over does occur between them (Stern, 1929; Kaufmann, 1933). Traditionally the term “sex-linked has been synonymous with “X-linked,” but since a growing number of possible Y-linked traits are being discovered, the term “sex linkage” could imply either X linkage or Y linkage. It is important to stress the fact that the occurrence of a trait in only one sex does not necessarily indicate sex linkage-it only implies sex limitation. Thus, although the Y antigen of mice is clearly a sex-limited trait, its sex-linked status remains somewhat controversial ( see Section 11). Other traits are known, e.g., the sex-limited protein (Slp) antigen, which are sex-limited but definitely not sex-linked (see Section IV). The subject of this review article had its historical beginnings in an untitled communication by Eichwald and Silmser ( 1955). These authors observed that among two inbred strains of mice, most of the male-tofemale skin grafts were rejected, whereas transplants made in other sex combinations nearly always succeeded. Although no explanation was initially offered for these results, this phenomenon soon became a popular model system for studying various aspects of transplantation biology. The interesting genetic and immunological concepts which have emerged from these studies form the main theme of this review. II. The Y Antigen in Mice

A. GENETICDETERMINATION OF Y ANTIGEN Soon after Eichwald and Silmser reported their observation that male skin isografts are often rejected by females, Hauschka (1955) offered two possible explanations for this incompatibility: male skin requires some androgenic hormone for indefinite maintenance or an antigen1 is determined by a gene on the nonpairing short segment of the Y chromosome. Hauschka suggested that if the latter possibility were correct, In mice this antigen has been referred to as the Y antigen, the male antigen, the Y factor, or H-Y.

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

217

crossing over between the X and Y chromosomes would explain why some females do not reject isologous male skin grafts (see Section I1,C). It was soon shown that male skin rejection was an immunological phenomenon. This was demonstrated by the fact that (1) females that rejected male skin isografts rejected subsequent grafts in accelerated fashion, i.e., “second-set” reactions were evident, and ( 2 ) females could be rendered tolerant of male skin isografts by exposing them to male cells at birth or shortly thereafter (Eichwald et al., 1957; Billingham and Silvers, 1958; Mariani et al., 1958). Furthermore, the rejection process, when studied histologically, was indistinguishable from known homograft reactions ( Eichwald and Lustgraaf, 1961). Once the immunological nature of male skin rejection was established, it was tacitly assumed by many that since the Y antigen was associated with the presence of a Y chromosome, it must be determined by a Y-linked histocompatibility locus (or loci). However, Fox ( 1956a,b, 1959; Fox and Yoon, 1958) pointed out that this antigen could result from the occurrence of a single X chromosome rather than from the presence of a Y chromosome. Indeed, this is the case in Drosophilu where the occurrence of two male-specific antigens, as well as two-female specific antigens, is determined by the number of X chromosomes and not by the presence or absence of a Y chromosome (see Section IV). It was, therefore, evident that only by examining XO and XXY mice could the sex chromosome responsible for determining H-Y be ascertained. Fortunately, such mice subsequently became available. In 1959, Welshons and Russell demonstrated that mice with an XO chromosomal constitution are females. This indicated that, in contrast to the situation in Drosophila where the balance between the number of X chromosomes and autosomes determines sex (Bridges, 1922), in mice, as in man (Jacobs and Strong, 1959; Ford et al., 1959)) maleness is determined by the Y chromosome. Further evidence for this was obtained when Cattanach (1961) reported that two exceptional XXY mice in his colony were males. Since none of the available XO and XXY mice were isogeneic, the role of the Y chromosome in determining the male-specific antigen could not be tested directly by challenging females with male skin grafts. Instead, the presence or absence of the Y factor was demonstrated indirectly by the capacity of tissues bearing it to induce immunity (Celada and Welshons, 1963). Normal females were inoculated with spleen cells of XO or XXY origin and subsequently challenged with isologous male spleen cells from donors presensitized to rat erythrocytes. The demonstrably impaired production of antierythrocyte antibody by some of these transferred cells-a reflection of their accelerated destruction-was taken

218

DAVID L. GASSER AND WILLYS K. SILVERS

as evidence of presensitization of the hosts to the Y antigen. When XO female spleen cells were used for pretreatment, the final antibody titers of the subsequently transferred male cells were comparable to those obtained when normal females were used, indicating that XO females behaved antigenically like normal XX females. However, when females were treated with cells from XXY males, the antierythrocyte antibody titers produced by the subsequently inoculated isologous male spleen cells were significantly reduced, resembling those obtained when the females had been presensitized with normal male cells. Evidence that the expression of the male antigen is correlated with the presence of the Y chromosome was also provided by the observation that testicular teratomas that retain this chromosome express the male antigens, whereas those testicular teratomas that have lost the Y chromosome do not (Bunker, 1966). These results made it apparent that the murine male antigen is associated with the presence of a Y chromosome and not with a single X chromosome as is the case in Drosophila. The possibility remained, however, that the male antigen was autosomally determined but was expressed only in a male hormonal milieu, a milieu which in mammals occurs only in the presence of a Y chromosome. If this is the case, one might expect that females exposed very early in life to male hormone would express the Y antigen. Conversely, males castrated at birth should not express this antigen. To obtain evidence for this thesis, Englestein (1967) attempted to induce the formation of the Y antigen in infant C57BL/6 females by testosterone injections, but no antigenic change in the skin of treated females was observed. He also attempted to suppress the antigen in males by treating them with estradiol and did report a prolongation of survival when skin from such treated animals was grafted to normal females, However, for some unknown reason, skin from control males injected with oil lasted even longer on female recipients. To determine the consequence of raising female skin in a “complete” male environment, Englestein grafted skin from neonatal C57BL/6 females onto adult C57BL/6 males and, after various periods of time, retransplanted them to C57BL/6 females that had been sensitized to the Y factor. Because 8 of 10 such female grafts, which had been maintained from 40 to 43 days on their intermediate male hosts, were rejected when returned to sensitized females, it was concluded that they had, indeed, been induced to express the male antigen. The validity of this conclusion is, however, dubious inasmuch as very similarly designed studies have yielded just the opposite results (Silvers et al., 1968; PolAEkovA, 1969). Thus Silvers and his associates found that neonatal C57 female skin grafts which had been in residence on adult C57 male hosts for upward

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

219

of 100 days were always accepted by secondary females whether they had or had not been sensitized to the Y antigen. However, such grafts, especially when retransplanted to sensitized animals, did display areas of necrosis of variable extent within 2 weeks after grafting, which led to a certain degree of contracture and transient cosmetic imperfection. In no case, however, did these initial lesions progress: affected grafts, although diminished in size, made a full recovery, regenerating hair crops of normal density. Subsequent experiments demonstrated these “healing-in’’crises to be the result of nonspecific ischemic necrosis caused by the excessive thickness of the grafts and by “contamination” from immigrant male cells, reactions against which were damaging to the entire transplant ( Lambert and Frank, 1970). The importance of such “passenger cells” was first demonstrated by Steinmuller (1967) and has been reviewed recently by Billingham ( 1971). In the light of the findings just described, it seems likely that the synthesis of the Y antigen is determined by a Y-linked histocompatibility locus ( or loci), but hormonal factors may influence its immunogenicity. Evidence for this interpretation is provided by the observation that 50% of skin grafts derived from 65- to 80-day-old C57 male animals, castrated within 12 hours of birth, were permanently accepted by normal adult females ( VojtiJkovi and PoliEkovi, 1966; PoliCkovi and Vojti8kovi, 1968). Similarly, it has been shown that long maintenance of C57 neonatal male skin grafts on isologous adult females reduces their immunogenicity, as reflected by the fact that such grafts survive for prolonged periods, occasionally persisting indefinitely, when removed and transplanted to normal C57 female recipients ( Silvers, 1968). Evidence could not be obtained for the complete absence of the male antigen from castrated males, for in no case could the rejection of male skin by neonatally castrated males be demonstrated ( VojtiHkovi and PoliEkovi, 1966). That Y antigen appears prenatally ( PoliEkovA, 1970) is obviously relevant to this observation. It has also been demonstrated that long-term maintenance of adult C57 male skin in a female endocrinological environment afforded by immunologically tolerant females does not weaken its antigenicity ( Silvers et al., 1968). This finding is not surprising inasmuch as orchiectomy carried out 12 to 24 hours after birth is significantly less effective in reducing the antigenicity of male skin than when it is performed sooner after birth ( PoliCkovi and VojtiJkovP, 1968), and castration of adult male mice has no effect on the antigenicity of grafts taken from them (Silvers et al., 1968). In conclusion, the most convincing lines of evidence that the male antigen is determined by the Y chromosome rather than the male hor-

220

DAVID L. GASSER AND WILLYS K. SILVERS

monal environment are (1) the failure of an alteration in the hormonal milieu to induce expression of the antigen in females or to prevent its expression in males entirely, and ( 2 ) the occurrence of the male antigen correlates with the presence of the Y chromosome. B. TISSUE LOCALIZATION One of the more interesting characteristics of the Y antigen is its nearly ubiquitous distribution in tissues of male mice. Although it was initially discovered in skin, its occurrence in many other tissues was soon demonstrated. Eichwald et al. (1958) observed that small pieces of lung tissue isografts placed in bilateral subpannicular pockets were rejected if male tissue was transplanted to female recipients, but control grafts involving all other sex combinations were not rejected. If female recipients were hyperimmunized against male salivary gland, rejection of this tissue was readily demonstrated also, with no rejections observed among the control grafts. Billingham and Silvers (1960) obtained permanent tolerance of male skin isografts in C57BL/6 females by injecting the recipients at birth with various concentrations of male peripheral blood leukocytes. Evidence for the occurrence of the Y antigen on male red blood cells was reported by Furasawa et al. (1963), who developed a number of isoantisera which could discriminate between male and female erythrocytes in hemagglutination tests, and by Hildemann and Pinkerton (1966) who used a plaque assay. Cell suspensions prepared from the spleen, liver, and kidney of males have all been shown to possess enough male-specific antigen to induce immunity (Sachs and Heller, 1958; Elkins, 1964) or tolerance (Billingham and Silvers, 1960; Martinez et al., 1963b; Kelly et d.,1964) to subsequent male skin isografts. Judd and Trentin (1971) transplanted a large series of hearts from fetal and neonatal C57 donors to the pinnae of adult C57 mouse ears. Approximately one-half of isografted hearts from unsexed donors were rejected by female but not by male recipients, suggesting that the Y antigen is also present on cardiac tissue. Cardiac transplants were never rejected by CBA females, an observation consistent with the inability of CBA females to reject male skin isografts (see Section I1,C). Gittes and Russell (1961) demonstrated the occurrence of the Y antigen on parathyroid, thyroid, and adrenal cortical tissue. Histological evidence of rejection was reported in all three cases, and an interesting demonstration of functional failure of the parathyroid grafts was obtained by the performance of serial serum calcium assays. When male parathyroid glands were transplanted to female recipients which also received male skin grafts, serum calcium levels dropped sharply at approximately the same time that the skin grafts were being rejected

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

221

(Fig. 1 ) . Pituitary glands of mice have also been shown to elicit malespecific rejection, as evidenced by both functional and histological criteria (Hoshino and Moore, 1968). PolhEkovh (1970) has shown that the Y antigen can be detected on cells from 11-day embryos but not 8-day embryos, indicating that H-Y is similar to H-2 and H-3 in that it appears prenatally (Billingham et al., 1956; J. Klein, 1965). One of the most important breakthroughs in the study of the H-Y antigen was the recent development by Goldberg et al. (1971) of cytotoxic antisera specific for the male antigen, as the availability of such antisera made it possible to demonstrate that the Y antigen occurs on mouse spermatozoa (see also Katsh et al., 1964). Anti-H-Y serum was obtained from C57BL/6 females 5-14 days after they had received a fourth or fifth skin graft from C57BL/6 males. Equal volumes of ( a ) diluted anti-H-Y serum, ( b ) sperm cell suspension, and ( c ) absorbed rabbit serum (as a source of complement) were incubated for 45 minutes at 37°C. Sperm counts were made in a hemocytometer after the addition of trypan blue. The results of this test with five strains of mice indicated that there were strain differences in the susceptibility of sperm to antiH-Y serum. This could be accounted for by strain differences in either the amount or specificity of H-Y antigen present. When anti-H-Y serum RAFTS ED

GLAh GRAFl

SKIN REJECTION

mmmrm

-

2

0

2

4

6

8

10

WEEKS

FIG. 1. Serum calcium changes after concomitant grafts of strain A male skin and parathyroid glands in a previously parathyroidectomized strain A female. The “chronic” skin rejection coincides with the cessation of parathyroid function. (From Gittes and Russell, 1981.)

222

DAVID L. GASSER AND WILLYS K. SILVERS

was absorbed with spleen, lymph node, and thymus cells from male or female C57BL/6 donors, there was only minimal loss of cytotoxic activity when female cells were used, but essentially complete loss when male cells were employed-a convincing demonstration that the cytotoxic reaction was specific for the H-Y antigen. The percentage of sperm specifically lysed by anti-H-Y often exceeded 50, which suggests that at least some of the X-bearing sperm possess the Y antigen. This is important as it suggests that the Y antigenic constitution of mouse sperm is determined by its precursor cells rather than the genetic material contained in the sperm, If borne out this would preclude the use of an immunological method of either distinguishing or separating X- or Y-bearing spermatozoa. TO Y ANTIGEN C. GENETIC BASISOF RESPONSE

Soon after the initial report on male skin rejection by females, it was observed that this phenomenon is not demonstrable in some strains (Prehn and Main, 1956; Short and Sobey, 1957) and that in others only a variable proportion of females reject male skin isografts (Eichwald et al., 1957; Bernstein et al., 1958). Moreover, there are interstrain differences in the rapidity with which these grafts are rejected. Michie and McLaren (1958) pointed out that these differences could arise from three possible causes: ( 1 ) strain differences in the Y antigen, ( 2 ) strain differences in “the donor’s capacity to give effect to the male-specific antigen,” and ( 3 ) strain differences in the female’s capacity to respond to the male antigen. It was suggested that the third situation would arise if ( a ) the females of some strains possessed an autosomally controlled replica of the Y antigen or if ( b ) strain differences existed in the female’s “hospitality to skin grafts in general, or male skin grafts in particular.” It was later shown by Billingham and Silvers (1960) that there are no strain differences in the Y antigen which would account for the observed variability of male skin rejection. When neonatal C57 females were injected with bone marrow cells from males of strains A, C3H, CBA, or AU, they were all found to be subsequently tolerant of C57 inale skin, Because of the very high degree of specificity of immunological tolerance (Billingham et al., 1952, 1956; HaLek and HaHkov6, 1958; Cinader and Dubert, 1956), these results were taken as formal proof of the identity of the male antigen among the five strains. Since C3H, CBA, and AU females are almost completely nonreactive to isologous male skin, it became clear that allelic differences in the Y antigen could not explain the strain variations that were observed. Additional support for this contention stemmed from the observation that skin grafts were accepted when exchanged between reciprocal F, hybrid males, i.e., males that were identical in every respect except for the origin of the

GENETICS' AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

223

X and Y chromosomes (Bernstein et al., 1958; Zaalberg, 1959; Silvers and Billingham, 1967). The second explanation proposed by Michie and McLaren proved to be somewhat prescient, but it still does not account for interstrain variation in male skin rejection. It was shown by Silvers and Billingham (1967) that strains differ significantly in the potencies of their malespecific antigen( s ) , but that the differences are not in the direction expected if nonrejection by some females is accounted for by this fact. When (CBA x C57)F, females were grafted with C57 male skin, i.e., grafts that are uniformly destroyed by C57 females, only half (14/28) were rejected; but when these hybrid females were grafted with CBA male skin, which usually survives indefinitely on CBA females, 29/30 grafts were sIoughed in Iess than 50 days. These data are shown in Table I and clearly indicate that C57 females can evoke a stronger immune response against male skin isografts than can CBA females although the Y antigen is more potent in the latter strain2 This report by Silvers and Billingham is at variance with an earlier communication by Bailey (1965) who failed to find evidence for any influence of genetic background on the expression of the Y antigen. However, this discrepancy could arise from the fact that the strains Bailey used, BALB/c and C57BL/6, do not differ in the genes that influence expression of the Y antigen, or because the donor mice he employed were all derived from intercrosses between these same two strains. It thus appears that the best explanation for the interstrain variations observed in male skin rejection rests with the third explanation proposed by Michie and McLaren; namely, that the genotype of the female can greatly influence her capacity to manifest an immune response to the antigen. Nevertheless, the basis for this inff uence remains controversial. Michie and McLaren's hypothesis ( 3 a ) was similar to that suggested by Hauschka (1955) and was defended by Hauschka and Holdridge ( 1962), who argued that a piece of Y-chromosomal material was translocated to an autosome or the X chromosome in those females that did not reject male skin. A similar hypothesis was later revived by Bailey and Hoste (1971). If some females are nonreactive to the male antigen because they possess translocated genetic information for this antigen, it follows that such females shouId express H-Y on their tissues at a level sufficient to induce self-tolerance. We know from the work of Triplett (1962) that 'Recent experiments by Wachtel et al. (1973) suggest that H-2' antigens enhance the immunogenicity of the Y antigen, which would account for the greater potency of the male antigen in CBA as compared to C57 male skin grafts.

GENETIC BACKQROUND

TABLE I AND

EXPRESSIVITY OF THE Y ANTIQEN~

Distribution of graft survival times (days) Donor

Recipient (No.)

15-24

25-50

51-75

76-100

>lo0 ____

~

c57 8 CBA 8’ c57 8’ CBA 8’ FI (C57Y) 8 Fi (CBAY) 8’ Fi (CBAY) 8’ Fi (C57Y) 3

11 2 18 12

11

9 11 5 5

13 14 1

1 1 2

-

1 1

1 15 14

*

25.6 4.0 24.2 +_ 4.2 24.0 +_ 5.5 23.0 f 3.3

~~

Data from Silvers and Billingham (1967). Copyright 1967 by the American Association for the Advancement of Science. Median survival time. c Standard deviation.

W

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

225

organisms are intolerant of unexpressed antigens for which they have genetic information. Billingham and Silvers (1960) injected C57 neonatal females with bone marrow cells from AU females which were shown to be nonreactive to AU male skin grafts. That none of these neonatally inoculated females proved to be tolerant of male skin isografts argues against the possibility that nonreactive AU females possess the male antigen. Furthermore, if females of a nonreactive strain (e.g., CBA) possess a translocation for the Y antigen, then this gene should be passed on to their progeny, and they should be incapable of rejecting male skin. However, this appears not to be the case, as E. Klein and Linder (19Sl) reported that, although 35/35 CBA females were nonreactive to CBA male skin grafts, all 20 (CBA X C57)F, females rejected such grafts. This finding that responsiveness to male skin is a dominant trait is difficult to reconcile with the translocation hypothesis. Evidence against the possibility that females that accept male skin isografts harbor a replica of the Y factor has also been presented by Zaalberg ( 1959). This investigator found that C57BL females inoculated with CBA female spleen cells, i.e., from females that accept male skin isografts, manifested a normal or “first-set” reaction when subsequently challenged with C57BL male skin grafts. It has also been shown by Hannover Larsen (1971) that if a sufficiently sensitive isograft experiment is designed, the rejection of C3H male cells by C3H females can be demonstrated. In this case, neonatal C3H males and females were made tolerant of lymphocytic choriomeningitis (LCM) virus by intraperitoneal inoculation of 1000 LD,, of the virus, which results in a persistent viremia. Tolerance can be abolished with subsequent disappearance of viremia if spleen and lymph node cells sensitized against LCM virus are injected into these mice. However, if the host rejects the injected cells, viremia will again appear. On the one hand, Hannover Larsen found that when he injected male C3H cells sensitized against LCM virus into tolerant C3H males, the viremia disappeared and did not return. On the other hand, when sensitized C3H male cells were injected into C3H females tolerant of LCM virus, the viremia temporarily disappeared, but again returned. The C3H females had apparently rejected the inoculated C3H male spleen and lymph node cells, in spite of the fact that such females do not reject isologous male skin. This observation does not appear to be compatible with the notion that females nonreactive to male skin possess the Y antigen and are naturally tolerant of it. Similar results, utilizing essentially the same principle, have also been reported by Feldman ( 1958).3 That the Y antigen is indeed foreign to “nonresponder” females is supported by our recent observation that 13/17 C3H females challenged with two successive C3H male skin grafts, eight weeks apart, rapidly rejected both grafts.

226

DAVID L. GASSER AND WILLYS K. SILVERS

It was observed by E. Klein and Linder (1961) that when 110 (CBA X C57)F, X CBA backcross females were grafted with CBA male skin, 87%of the grafts were rejected. They tentatively attributed this to the segregation of two autosomal dominant genes, each by itself capable of determining reactivity against the Y antigen. Recently, Gasser and Silvers (1971a) and, independently, Bailey and Hoste (1971) reported that the female’s ability to reject male skin is largely determined by the H-2 locus or a gene closely linked to it. As shown in Table 11, the congenic strain C3H.SW which has the genetic background of C3H but possesses the H-2b allele is a moderately good reactor to isologous male skin. However, BlO.BR, which has the genetic background of C57BL/10 but possesses the H-2’‘ allele, is a very poor reactor. It has been shown by McDevitt and Tyan (1968) that the ability of inbred mice to respond to a synthetic antigen called (T,G)-A-L is determined by a single gene, Ir-1, which is genetically linked to the H-2 locus and, in fact, maps within the H-2 region ( McDevitt et al., 1969). Reaction to H-Y is, therefore, similar to the responses to (T,G)-A-L, ( H,G) -A--L, and ( Phe,G ) -A--L ( McDevitt and Chinitz, 1969), to TNPMSA (Rathbun and Hildemann, 1970), and to low doses of BPO-BSA (Vaz and Levine, 1970), in that the immune reaction to each of these antigens is largely determined by the H-2 region. There are quite clearly other genetic factors in addition to H-2 which influence male skin rejection (Gasser and Silvers, 1971a,b), a situation similar in some ways to the murine response to the Ea-1 alloantigens ( Gasser and Shreffler, 1972). It would be of great interest to know what specific site in the H - 2 region is associated with male skin rejection. The Ir-1 locus referred to above is believed to map just to the right of the serum substance ( S s ) locus and just to the left of the “K” region in the right-hand part of the H-2 locus ( McDevitt et al., 1969). Male skin rejection was studied in a number of mice possessing recombinant H-2 alleles, with conflicting reTABLE I1 RELATIONSHIP BETWEEN H-2 GENOTYPE AND SURVIVAL SKINISOGRAFTS ON FEMALES

Strain C57BL/10 C3H.SW C3H B1O.BR 0

H-2 type

NO.of recipients

H-2b

20 17 15 19

H-Rb

H-2k

H-Rk

TIMICS OF

MALE

Cumulative yo rejected at (days): 25

50

75

100

70 0 0 0

100 59 0

100 77 0

100 88 0 53

11

37

Median survival time rt 95% confidence limits (days).

MST

95% CLa

19 lt 3 . 1 6 40 f 11.88 -

-

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

227

sults (Bailey, 1971). Mice possessing the H-2h allele, which is a recombinant between the K end of H-2" and the D end of H-2b, displayed male skin rejection comparable to that of H-2a mice. Animals with the H-2' allele, which is a recombinant between the D end of H-2" and the K end of H-Zb,displayed male skin rejection comparable to that of H-2b mice. This suggested that the immune response gene for H-Y is to the right of the crossover point, i.e., either within the X region or distally linked to it (Bailey, 1971; Stimpfling and Reichert, 1971). Females of the HTG strain, however, gave an H-P-like response, which is in conflict with the other results (Bailey, 1971). This strain possesses H-2", which is a recombinant between the D region of H-2b and the K region of H-2a. Since the H-a"-derived regions of H-2' and H-2s do not overlap, the gene that endows strain HTG with a good response to male skin is obviously different from the one conferring the same ability on the strains possessing H-2" and H-2'. Two possible explanations could account for these conflicting results: ( a ) there is more than one site within H-2 which influences male skin rejection or ( b ) the H-2 region is not the only locus determining H-Y rejection. It was shown by Gasser and Silvers (1971b) that numerous cases exist in which animals of the same H-2 type display significantly different responses to H-Y. Localizing the non H-2 genes which are responsible for these differences as well as specifically mapping the site or sites within H-2 that influence male skin rejection are problems which remain for future investigations.

D. INDUCTION OF TOLERANCE TO Y ANTIGEN The Y antigen, especially as expressed in C57BL mice, a strain in which nearly all females reject male skin isografts, has been widely utilized in studies concerned with the induction of tolerance in neonatal (Billingham and Silvers, 1958, 1960; Mariani et al., 1958, 1960b) and adult (Mariani et al., 1960b; Lustgraaf et al., 1960; Kelman et al., 1961; Martinez et al., 1963a,b; Kelly et al., 1964; Billingham d. al., 1965) animals. In fact, the observation that unresponsiveness to the Y antigen was readily induced in female mice subsequent to the neonatal period contributed greatly to reevaluating some of the original concepts of tolerance induction, namely that tolerance of living cells could be induced only during a finite period of life and that there was an age, or age range, termed the neutral or "null period," during which exposure to an antigen conferred neither tolerance nor immunity (Billingham, 1958). Thus Billingham and Silvers (1960) found that more than 90% of C57BL/6 females inoculated when they were 12 days old with 5 to 10 million adult male spleen cells, and 60%inoculated as late as 17 days after birth with 12 to 20 million cells, permanently accepted subsequent male skin isografts. Moreover, after the intravenous inoculation of about

228

DAVID L. GASSER AND WILLYS K. SILVERS

40 million isologous male spleen cells, adult strain A and C57BL/1 females were found to be unresponsive to male skin isografts (Mariani et al., 1960a,b). Alternative methods for inducing a high degree of tolerance of the Y antigen in adult C57BL/1 females include inoculating them intravenously with at least 0.25 cc. of whole blood from isologous male donors (Kelman et al., 1961) or maintaining females in parabiosis with males for a period as short as 4 days (Mariani et al., 1960a,b). Indeed, it was found that even adult C57BL/ 1 females, which had been rendered tolerant of the Y antigen by intravenous inoculation of male spleen cells, could transfer this unresponsive state to previously nonconditioned females when placed in parabiosis with them (Mariani et d., 1960a). Results with C57BL/ 6 animals indicate that more heroic procedures are required to induce tolerance of the Y antigen in adult animals of this subline ( Billingham et al., 1965). Here at least 500 million isologous male spleen cells, given intravenously or intraperitoneally, are necessary to render a high proportion of adult females tolerant of male skin isografts. The intravenous route is the more effective, and repeated exposures to the antigenic stimulus are more effective than single ones. Moreover, about 3 weeks of parabiotic union with isologous males (Lustgraaf et al., 1960; Billingham et al., 1965) or with isologous females that are themselves tolerant of the Y factor as a consequence of neonatal inoculation with isologous male cells ( Billingham et al., 1985), is necessary to induce high degrees of unresponsiveness in hitherto normal females. In contrast to the demanding nature of the treatments needed to induce tolerance in adult C57BL/6 females, as few as 300,000 isologous male cells render all neonatally inoculated females highly tolerant of male skin isografts and as few as 10,000 cells induce unresponsiveness in 25% of such recipients ( Billingham and Silvers, 1960). Inasmuch as it is relatively easy to induce tolerance of the Y antigen in adult females by using living cells, it seemed likely that this might also be achieved with antigenic extracts. This proved to be the case as most young adult C57BL strain females (sublines 1, 4, and 6 ) that were inoculated repeatedly with disrupted, freeze-thawed suspensions of isologous male splenic tissue, accepted male skin isografts (Martinez et al., 196313). Furthermore, similar results were subsequently obtained in C57BL/1 mice by using disrupted cellular material prepared from male liver and kidney ( Kelly et al., 1964). The greatest tolerance-providing effect was obtained when the disrupted cell preparations were given by the intravenous route, both before and after transplantation of skin. The dose of material employed was important inasmuch as a single intravenous injection of one spleen equivalent of disrupted cells did not induce tolerance of male skin, whereas four injections of the same amount of mate-

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

229

rial given before grafting together with a series of injections after grafting yielded a high incidence of unresponsiveness (Martinez et al., 1963a). In contradistinction to the tolerant state produced by inoculation of viable, isologous, male spleen cells in adult females or that produced by parabiotic union, tolerance of the Y antigen produced by repeated inoculations of disrupted spleen cells preparations cannot be transferred to isogeneic neonatal mice with spleen cells of the tolerant animal (Martinez et aZ.,1Wb). The phenomenon known as “split to1erance”l (Lustgraaf et al., 1960) has also been well exemplified in studies concerned with the Y antigen and, indeed, as already stated in Section II,C, provided evidence for the identity of this antigen in different strains of mice. This stemmed from the observation that, although neonatal C57 females inoculated with allogeneic male cells were found not to be tolerant of skin grafts from the strain that supplied the inoculum, they were tolerant of C57 male skin grafts (Billingham and Silvers, 1960; Lustgraaf et al., 1960). Split tolerance can occasionally also be induced by injecting females at birth with subcellular antigenic materials obtained from allogeneic male spleen cells previously disrupted by several cycles of freezing and thawing. Thus, Martinez and Smith (1964) found that at least a few C57BL/1 females injected from birth until 7 weeks of age, with a series of intraperitoneal injections of disrupted spleen cells from strain A males, were tolerant of male skin isografts but not of A strain skin. This finding that tolerance of skin grafts could be induced to some antigens but not to others carried on the same cellular vehicle (see, also, Billingham and Brent, 1959) was initially believed to indicate that the persistence of tolerance of the Y antigen does not require persistence of the antigen ( Billingham and Silvers, 1960). However, this contention has had to be reevaluated in the light of more recent experiments (Brent and Courtenay, 1963; Billingham et al., 1965). Thus, on the basis of a sensitive test for chimerism, a test which is capable of revealing the presence of as few as 12,500Y antigen-containing cells in a standard population of 20 million spleen cells, it was found that 1 out of 4 C57BL/6 females rendered tolerant of the Y factor by neonatal inoculation with CBA male marrow cells, was chimeric with respect to the Y ‘This term was coined to describe the situation which occurs when cells induce tolerance of only some of the transplantation antigens with which they confront the host. For example, early experiments by Billingham and Brent (1959) demonstrated that some strain A mice injected at birth with spleen cells from either ( A U x CBA) or (C57BL/6 X CBA)F1 hybrids were tolerant of CBA skin grafts but not of skin from AU, C57BL/6, or F1 hybrid animals.

230

DAVID L. GASSER AND WILLYS K. SILVERS

factor. Furthermore, the Y antigen was expressed by at least 1%of the spleen cells of C57BL/6 females made tolerant of male skin isografts as a result of neonatal exposure to 5 million (CBA X C57) F, male spleen cells-a dosage completely ineffective in conferring tolerance of CBA skin grafts upon C57 hosts (Billingham et d.,1965). This raises the possibility that, although detectable levels of Y antigen are not apparent in some females made tolerant of male skin isografts, this factor is, nevertheless, harbored by them. Perhaps the Y antigen, like pneumococcal polysaccharide (Felton and Ottinger, 1942; Felton, 1949; Felton et al., 1955a,b), is only very slowly metabolized, or the male skin graft itself provides an adequate antigenic stimulus for maintaining unresponsiveness. Another possibility is that the persisting foreign cells responsible for the chimerism are located so that they are protected from specific sensitivity on the part of the host. In addition to these more-or-less orthodox methods of inducing unresponsiveness to male skin isografts in mice, other means of rendering females unresponsive to the Y antigen have also been reported. Martinez et al. ( 1961) found that tolerance of male skin isografts in adult C57BL/ 1 mice could be induced by grafting the females with a male skin graft involving approximately 60%of the total body surface. However, it was not possible to transfer unresponsiveness from these animals to naive females. The discovery of Breyere and Barrett (1960) that multiparous female mice of the BALB/cAn strain which had borne litters of strain DBA/2 males became tolerant of a transplanted DBA/2 sarcoma, stimulated a number of investigators to challenge C57BL females with male skin isografts after they had given birth to a variable number of interstrain litters (Prehn, 1960; Vener et al., 1961; Lengerovh and Vojtiikovh, 1962, 1963; Billingham et al., 1965; Silvers, 1968). In all of these studies except one (Vener et al., 1961), tolerance of such grafts was observed. Moreover, the degree of unresponsiveness was directly related to the number of litters borne by the female recipient. In two reports (Lengerovh and Vojtiikovh, 1962, 1963) there was some evidence that antigenic material-probably cells-introduced into the female genital tract during insemination made some contribution to the unresponsiveness. Although the basis for pregnancy-induced tolerance remains to be determined, it is probably a consequence of repeated transient exposures of the mother to very low dosages of transplantation isoantigens from their fetuses (Billingham, 1964). Cell chimerism has not been demonstrated in such multiparous animals ( Billingham et al., 1965). Attempts to produce comparable degrees of tolerance by successive male skin grafts rather than by number of pregnancies have been

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

231

successful in one case (Prehn, 1961) but not in another (Billingham et

al., 1965).

Finally, unresponsiveness to the Y antigen can be induced in C57BL/6 mice by exposing adult females to neonatal male skin isografts, about 75%of which are permanently accepted (Table 111) (Billingham et al., 1965; Silvers, 1968; Wachtel and Silvers, 1971). When adult C57BL/6 females were challenged concomitantly with neonatal male skin grafts, on one side of their thorax, and with adult male skin grafts, on the other, about 25%accepted both grafts ( Silvers, 1968). Moreover, this percentage is doubled if 50 days are allowed to elapse between exposure to the neonatal and adult male skin grafts, respectively. Occasionally, even presensitized females can be rendered tolerant after exposure to neonatal male skin ( Silvers, 1968). Although the manner in which skin grafts from newborn males induce tolerance remains to be elucidated, it is apparent that, unlike the anomalous survival of neonatal C3H skin grafts on CBA mice (Wachtel and Silvers, 1971) , the tolerogenic influence of passenger leukocytes emigrating from the vasculature of the graft cannot account for its survival ( Wachtel and Silvers, 1972a). Evidence for this thesis stems from the following observations: (a) when neonatal male skin isografts of 50 days standing were removed and 50 days elapsed before their hosts were rechallenged with adult male skin isografts, not only were these adult grafts almost invariably rejected, but they were usually rejected in an accelerated fashion (Silvers, 1968); ( b ) neonatal male skin grafts persisted on C57BL/6 females after heavy irradiation of the donor or after passage on intermediate adult male hosts-moreover, after these treatments the grafts did not lose their ability to protect subsequent adult male skin isografts (Wachtel and Silvers, 1972a); ( c ) chimerism was not detectable in female C57 mice made tolerant by exposure to infant grafts (Billingham et al., 1965). Attempts have been made to abolish tolerance of male skin isografts in females made tolerant by most of the procedures described above (Billingham at al., 1965). It was found that, whereas tolerance of male skin isografts induced by inoculation of infant females with isologous cells or after parabiosis of adult females with isologous male partners, could be abolished by transfer of either normal or immune isologous female cells, tolerance induced by inoculation of male cells of homologous origin (from C3H, A, CBA, or AU strains) could not be abolished by these procedures. Indeed, even the transfer of three donor equivalents of immune cells, i.e., the spleens and axillary and brachial lymph nodes of three sensitized females, was ineffective in abolishing tolerance in females treated at birth with homologous male cells. Similarly, it has not

w to

TABLE I11 SURVIVAL TIMESOF ADULTAND NEONATAL MALESKIN ISOGRAFTS TRANSPLANTED TO ADULTFEMALE Hosmo

Donor group

Distribution of graft survival times (clays)

No. of recipients

214

15-24

25-50

51-75

76-100

30 46

0 0

16 2

13 4

1 4

2

(A) Adult

(B) Newborn

Data from Silvers (1968).

* Median survival time. e.Standarddeviation.

0

> 100 0

34 (74%)

MSP

S.D.C

(days)

(days)

24.5

f_

-

2.6

1.35 -

8

L!cc cn

p

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

233

been possible to immunize adoptively females rendered tolerant as a consequence of multiple pregnancies, and only rarely have neonatal male skin isografts of long standing been affected when their female hosts were challenged with large numbers of specifically sensitized female cells. These results are consistent with the thesis that when normal or immune isologous lymphoid cells are introduced into females tolerant of male skin isografts, their ability to destroy these grafts is a function of the chimeric status of the host. On the one hand, in those instances in which the females were made tolerant by exposure to isologous male cells and, therefore, are almost certainly highly chimeric, tolerance is easily abolished; on the other hand, when females are rendered unresponsive to the Y antigen by homologous male cells, by multiparity, or with neonatal male skin grafts, they possess, at the most, only a very low level of cellular chimerism and, accordingly, are resistant to adoptive immunization. Further evidence that this is so stems from the observation that if tolerant females in which chimerism could not be detected were inoculated intraperitoneally with isologous male spleen cells and, 2 weeks later, received one donor equivalent of isologous female lymphoid cells, their male skin isografts were rejected within 12 to 37 days (Billingham et al., 1965).

E. GRAFT-VERSUS-HOST REACTIONSAND Y ANTIGEN That no apparent graft-versus-host reactions are observed when newborn C57 males are inoculated with isologous lymphoid cells from either normal or sensitized females (Billingham and Silvers, 1960) is not too surprising since, as already noted, the Y antigen is not as immunogenic in neonatal mice as it is in adult animals (VojtiHkovh and Poll6kovh, 1966). To investigate this situation further neonatal males were inoculated with both adult male spleen cells and sensitized female cells on the premise that the adult male cells might provoke a more vigorous female antimale response which would adversely affect the infant hosts. This, however, is not the case as newborn males inoculated intraperitoneally and intravenously with 30 million lymphoid cells from highly sensitized females and similar numbers of adult male spleen cells, did not display any signs of runting ( Wachtel and Silvers, 197213). Attempts to induce graft-versus-host reactions in adult males have been more successful. Billingham and his associates (1965) reported that parabiotic union of specifically sensitized C57 females with isologous males resulted in parabiosis intoxication in 50%of the latter, and secondary disease has been reported to occur in C57 males protected with isologous female marrow cells after an otherwise lethal dosage of X-irradiation

234

DAVID L. GASSER AND WILLYS K. SILVERS

( Lengerovti and Chutnh, 1959). However, DeVries and Vos ( 1958) failed to obtain any evidence of homologous disease in similarly treated C57 animals.

F. EFFECT OF Y ANTIGEN ON PLACENTAL GROWTH AND SEXRATIO With the exception of highly inbred strains, the mammalian fetus must differ antigenically from the mother so that the placenta is, in effect, a homograft. It was first recognized by Billington (1964) that antigenic dissimilarity between mother and fetus results in increased placental size (see, however, Finkel and Lilly, 1971). James (1965) confirmed Billington’s findings and showed that if C57BL females were immunized against A2G antigens before mating with A2G males, an even greater increase in placental size occurred. If C57BL females were made tolerant of A2G male antigens before mating with A2G males, the placental size was no greater than in control females. An interesting application of this principle to man was reported b y Warburton and Naylor (1971) who demonstrated that there is a significant increase in placental weight between a mother’s first and second pregnancies, but that this increase does not occur if the second father is different from the first. Toivanen and Hirvonen ( 1970) demonstrated that fetal-maternal ABO incompatibility is associated with increased placental weight in the case of male but not female fetuses of the first pregnancy. This suggested that a cumulative effect of ABO incompatibility and Y chromosomedependent incompatibility is needed for an increase in placental weight to be manifested. There seems to be little doubt that fetal-maternal incompatibility increases the size of the placenta and that the Y antigen can contribute to this effect. Ounsted and Ounsted (1970) have presented evidence that Y incompatibility enhances fetal growth rate as well, The effect of male antigenic incompatibility on the sex ratio has been a subject of considerable interest. Renkonen et al. (1962) examined a large amount of family data and concluded that as the number of pregnancies increased the sex ratio (male births/100 births) decreased. They suggested that pregnancies that produce male fetuses immunize a small portion of mothers against male antigens which are then harmful to subsequent male fetuses. The methods which were used to reach this conclusion were criticized by Edwards (1963), and the immunological implications were tested by McLaren (1962, 1965) who found that deliberately immunizing female mice against the Y antigen did not alter the sex ratio of the progeny. Subsequent investigations suggested that rather than decrease the sex

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

235

ratio, H-Y incompatibility may actually increase the proportion of male offspring. It was shown by Hull (1964) that mice which were heterozygous at the agouti locus seemed to be favored over homozygotes. Clarke and Kirby (1966) suggested that this effect could be attributed to histocompatibility genes (e.g., H - 3 ) which were closely linked to agouti. When Hull (1969) reexamined the problem using a congenic pair of mice which differed at the H-3 locus, he found that homozygous females mated to heterozygous males produced significantly less than 50%homozygous offspring. In one experiment this deficiency of H-3 homozygotes occurred only among male offspring, which suggested that the Y antigen contributed to the advantageous maternal-fetal incompatibility. If male antigenic incompatibility affects the sex ratio, one might expect the effect to be more pronounced in the offspring of human consanguineous marriages than in those from random outbred matings in which more histocompatibility determinants are segregating. Kirby et al. ( 1967) reported that isolated first-cousin marriages produced 554%males among their progeny, which differed significantly from the national average. In those families in which there had already been consanguinity within three generations before the first-cousin marriage, the sex ratio was 0.61. Slatis et al. (1958) found no disturbed sex ratio among the children of consanguineous marriages in the Chicago area, but Bonn6 (1963) reported a sex ratio of 0.62 among the children of consanguineous marriages in the Samaritan communities of Jordan and Israel. Allan (1959) showed that AB mothers have offspring with a significantly higher sex ratio (0.557) than those of other blood groups combined (0.519). Since AB mothers cannot react against A or B antigens of the zygote, it would appear that a Y chromosome-dependent antigen plays a more important role in these cases than in mothers of other blood types. Evidence for the possibility that enhancing antibodies may play a role in protecting the fetus from immunological injury was obtained in an in vitro experiment by Hellstrom et al. (1969). They reported that ( a ) lymph node cells from BALB/c mice pregnant by C3H males, or from BALB/ c mice immunized with C3H embryonic tissues, reduced colony formation of plated C3H embryonic cells, and ( b ) sera from BALB/c mice pregnant with C3H (but not BALB/c) abrogate the inhibitory effect of specifically immune lymph node cells. This intriguing hypothesis was explored further in an in vivo experiment by Lapp6 and Schalk (1971). Since Hellstrom’s serum-blocking factors appear to be largely dependent on the presence of a spleen (Hellstrom et al., 1969), it was predicted that maternal splenectomy would expose male but not

236

DAVID L. GASSER AND WILLYS K. SILVERS

female fetuses to the cytopathic effects of sensitized lymphocytes and, consequently, a lower sex ratio would result. Female mice were splenectomized before immunization against the Y antigen and then mated to isogeneic males. The C57BL females that received this treatment gave birth to progeny with a sex ratio of 0.594, which was significantly greater than that from the control group of sham-splenectomized, Y-immunized mice (0.490) as well as the untreated controls (0.516). Lappi. and Schalk's interpretation, therefore, differed from that of Hellstram et aE. ( 1969), in that the maternal cellular immune response was perceived as a positive selective force in fetal development. The increased proportion of males in the splenectomized, immunized group was attributed to preferential implantation of male zygotes. It was proposed that males would have a selective advantage afforded by the maternal immune response against the Y antigen but that this advantage was neutralized by spleendependent blocking antibodies. Ill. The Xg Blood Group locus i n Man

In 1962, Mann et al. reported a new sex-linked bIood group called Xg. The antibody defining this antigen was present in the serum of a Caucasian patient who had been transfused many times because of severe nosebleeds. The antigen recognized by this antibody was called Xg", the two alternative phenotypes Xg( a + ) and Xg( a - ), the gene responsible for the antigen X$, and the gene for its so far silent allele, Xg. The evidence that this gene is on the X chromosome is overwhelming. ( a ) In positive X positive matings, all daughters were positive for the antigen, whereas both negatives and positives were found among the sons; ( b ) in matings of positive males X negative females, all sons were negative and all daughters were positive; ( c ) matings of negative males by positive females produced all possible phenotypes; ( d) the frequency of the Xg( a ) phenotype was much greater among females than males; ( e ) females with Turner's syndrome (only one X chromosome) have the male distribution of Xg groups, and males with Klinefelter's syndrome (an extra X chromosome) have Xg frequencies much closer to the female than the male distribution (Race and Sanger, 1988). The frequencies of the Xg" and X g alleles have been determined in a number of populations. Although the genes are polymorphic in all populations reported so far, there are some interesting differences. The frequencies in Britain, for example, differ significantly from those on the mainland of northern Europe (Race and Sanger, 1968). The erythrocytes of a number of mammalian species other than man have also been tested for the occurrence of X$, and of these, only gibbons (Hylobutes 2ar l a r ) have been reported to include positive

+

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

237

individuals (Gavin et al., 1964). Moreover, in this species, as in man, Xg appears to be X-linked. In 1964, Sanger et al. reported that the pedigrees of two families were not consistent with X linkage of the Xg groups. In both of these families the mothers were Xg( a - ) and the fathers were Xg(a+ ). Five sons in these families were exceptional in being Xg( a ). Since a male gets his sole X chromosome from his mother, all the sons in these families should have been Xg( a- ). There are several types of cytogenetic abnormalities which could account for these exceptions, but no abnormal karyotypes were observed in these two families. One possible explanation was that a small portion of the X chromosome containing the Xg locus had become translocated to an autosome or the Y chromosome, which would make it possible for the sons to inherit the X$ allele from their fathers. The possibility that X-Y interchange could occur is strengthened by evidence for homology between X and Y ( Ferguson-Smith, 1965) and is especially attractive since it would account for the occurrence of XX males ( Ferguson-Smith, 1966). In 1971 a very interesting and exceptional family was reported by Buckton et al. A woman and her husband were phenotypically Xg( a - ), but two of their children were Xg( a + ). In this case the peripheral blood leukocytes of the mother presented a mosaic karyotype 45,X/ 46,XX, the two cell lines being in the approximate proportion of 1 to 2. In her bone marrow about 90%of the cells were of the type 45,X, which suggested that her red cells were derived from a predominantly 45,X line. Presumably she was an X $ / X g heterozygote, passed the X $ allele on to two of her children, and was Xg( a- ) herself because the X chromosome possessing the X g allele was lost from the cells of her erythroid series. Whether this explanation can account for the exceptional families reported previously remains to be determined. Since the Xg locus is on the X chromosome, it was of considerable interest to determine whether the “single active X” hypothesis (Lyon, 1962, 1968) applied to Xg. Gorman et al. (1963) examined the red cells of female patients heterozygous for both Xg”/Xg and X-linked G6PD deficiency. The authors attempted to fractionate these red cells with anti-Xga serum and then to determine whether the Xg( a- ) population differed from the Xg( a + ) population with respect to G-6PD levels. Despite the fact that their technique proved to be very successful when artificial mosaics were used, they could find no evidence that an Xg( a - ) component existed among the red cells of heterozygous patients, and the Lyon hypothesis seemed not to apply to Xg. Evidence to the contrary was reported by Lee et al. (1968). They studied a family in which three female members were heterozygous for

+

238

DAVID L. GASSER AND WILLYS K. SILVERS

hereditary, X-linked, sideroachrestic anemia and, therefore, possessed two morphologically distinct erythrocyte populations. These erythrocyte populations could be separated according to density by centrifugation in layered gum acacia solutions. In two of the female patients the normal erythrocytes were found to be Xg(a-) and the microcytic erythrocytes were Xg( a + ) . These results suggested that X inactivation applies to both the red cell disorder and to the Xg” antigen. Weatherall et al. (1970) attempted to repeat this experiment on the red cells of another Xg”/Xg female patient who also had a dimorphic erythrocyte population because of a sideroblastic anemia. When they separated the cells into two groups by centrifugation, both populations were positive for Xg”. Fialkow et al. (1970) studied 11 X g / X g heterozygous patients who had chronic myelocytic leukemia. In these individuals most of the peripheral blood erythrocytes are derived from a single clone. If X inactivation applies to Xg, then some of these patients should be phenotypically Xg( a- ), since there is a 50% chance that the Xg-carrying X chromosome will be inactivated. Since all 11 were typed as Xg( a + ), there did not appear to be inactivation of the Xg“ alleles in any of the patients. If random, fixed Xg locus inactivation occurs, the probability that all 11 heterozygotes would be Xg( a + ) is one in 2048, which suggests that X inactivation in man is incomplete. This does not seem at all surprising, since females with only one X chromosome, or with extra X chromosomes, and males with an extra X chromosome are not normal. There is some evidence suggesting that in the “inactive” X chromosome, inactivation may emanate from a certain point and spread along gradients, perhaps leaving distal regions unaffected ( Russell, 1964). There is no clear evidence of close linkage between Xg and G-GPD, hemophilia A, hemophilia B, deutan, protan, X-borne dystrophy, or testicular feminization (Race and Sanger, 1968). This lack of close linkage with other loci on the same chromosome may indicate that Xg is far enough from the “center of inactivation” that it is not affected by “Lyonization.” IV. Other Sex-Limited and Sex-Linked Antigens

Evidence for the occurrence of a male-specific transplantation antigen in rats has been reported (Billingham and Silvers, 1959; Billingham et al., 1962; Zeiss et al., 1962; Heslop, 1969a,b). Billingham et al. (1962) found that about 60% of inbred BN females rejected BN male skin grafts, but that Lewis females did not reject Lewis male grafts. However, 40% of (Lewis X BN)F, hybrid females rejected Lewis male skin, suggesting that Lewis males possess the Y antigen, but that Lewis females are

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

239

incapable of responding to it. Zeiss et al. (1962) reported that male skin isografts were rejected by females of two locally developed strains of inbred rats, hooded and black, but not by females of a third strain, albino. Heslop (1969a) reported male skin rejections by females of strains AS, AS2, BS, and HS. She also observed that partial tolerance of adult male skin isografts could be induced in females by simultaneous grafting of neonatal male skin (Heslop, 1969b), which is similar to results obtained in mice (see Section 11,D). Attempts have been made to determine whether the Y factors of the mouse and rat are antigenically similar (Billingham et al., 1962). In some experiments, newborn and very young C57BL/6 female mice were repeatedly injected with fairly large dosages of BN male rat spleen or marrow cells, and subsequently challenged with C57BL/6 male isografts to determine whether a state of tolerance of the Y factor had been induced. In other experiments, adult C57 female mice were injected with suspensions of spleen cells from male BN rat donors in the hope of demonstrating sensitization in respect to subsequent isografts of male skin. Although no convincing evidence was obtained of any cross-reactivity between the antigens determined by the Y chromosomes of these two species, such evidence has been obtained by Yang and Silvers (1971). Thus, by utilizing males of a different rat strain (Fischer), it has been found that lymph node cells from these animals can sensitize C57 females with respect to subsequent male skin isografts. The basis for this discrepancy in the results obtained with BN male donors, on the one hand, and Fischer males, on the other, is currently under investigation. Evidence for a male-specific antigen has also been reported in rabbits ( Chai, 1968) and platyfish ( Miller, 1962; Kallman, 1970). Bailey (1963a) reported evidence for the occurrence of a histocompatibility locus on the X-chromosome of mice. Tail skin grafts were exchanged between male F, hybrids produced by reciprocal matings between the strains C57BL/6 and BALB/c. Many rejections were observed in the transplants exchanged between reciprocal hybrids, but very few rejections occurred when exchanges were made between F1 males of the same parental combinations. When sensitized F, males were given second reciprocal hybrid male grafts, their survival time was considerably shortened. Moreover, when F, males were challenged with parental strain grafts, the paternal but not maternal strain skin was sloughed. These results suggest that the observed histoincompatibility was associated with the X chromosome. Data confirming these findings were reported by Rosenau and Horwitz (1968). Bailey (1964) subsequently reported that when (C57BL/6 X

240

DAVID L. GASSER AND WILLYS K. SILVERS

BALB) Fl male hosts were challenged with paternal strain skin grafts, these grafts were rejected more slowly than (BALB X C57BL/6)F1 male skin grafts. Bailey attributed this difference either to donor-host differences in Y-determined antigens or to nonhistocompatibility genes which modify the rejection process. With regard to the former, there is one report which suggests that the male antigen may occur in different forms in different mouse strains (Hildemann and Cooper, 1967). Evidence for X-linked histoincompatibility between the inbred strains A/ J and C57BL/6 was reported by Hildemann and Cooper ( 1967) who used skin graft exchanges and by Hildemann and Pinkerton (1966) who used an agar plaque technique. The effect of the Lyon hypothesis (Lyon, 1961) on the expression of a histocompatibility locus on the X chromosome must also be considered, for if Lyonization occurs and the action of H-X is localized, this should be reflected by a variable or mosaic rejection pattern of F, female grafts, heterozygous at H-X, by F, males. This follows from the fact that skin grafts from such hybrid females should consist of two populations of cells-one which expresses the H-X antigen of the paternal X and the other of the maternal X. That such a homograft response does, in fact, occur has been demonstrated by Bailey (196313). He challenged (BALB/c X C57BL)Fl males with tail skin grafts from hybrid females and found that, whereas some of these grafts were almost completely accepted, others were totally destroyed. Moreover, the rejection pattern of some grafts was visibly mosaic which would be expected from an early random determination of X-chromosome inactivation. This observation presents another excellent example of the specificity of the homograft reaction at the cellular level (Billingham and Silvers, 1970; Mintz and Silvers, 1970). In 1966, Berg and Bearn reported the detection of an inherited serum antigen in man which they designated Xml.The antigen was identified by a rabbit antiserum made specific by absorption and was shown to reside in the a2 macroglobulin fraction of the serum. Evidence for X linkage of this trait was obtained from family studies as well as from the distribution of phenotypes among four populations of unrelated individuals. Passmore and Shreffler (1970) reported the discovery of a serum antigen in mice which was normally detected only in males of the proper genotype. The gene determining this trait, Slp, maps within the H-2 region and is genetically inseparable from the Ss locus, which determines the quantity of an unidentified serum substance (Shreffler and Owen, 1963). Evidence was presented for the occurrence of the SZp determinant on the same molecule as the Ss protein. Males of a

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

241

strain which normally express the S l p trait failed to do so when castrated at 3%weeks of age, whereas females of the same strain ovariectomized at 3%weeks and treated with testosterone propionate expressed the Slp antigen by 10 weeks of age (Passmore and Shreffler, 1971). Inasmuch as the female is the heterogametic sex (ZW) in birds whereas the male is homogametic ( Z Z ) , it is of special interest to examine the sex-linked transplantation antigens in avian species. As early as 1932, Kozelka reported that significantly more female-to-male comb, wattle, and spur grafts were rejected than grafts exchanged in other sex combinations. Gilmour (1967) confirmed the existence of a heterogametic antigen in females which was similar in some respects to the Y antigen of mammals. In two highly inbred Reaseheath lines of chickens ( IA and WA), female-to-male integumental grafts were reacted against much more strongly than transplants made in the other three sex combinations. Most first-set grafts induced tolerance to subsequent female grafts, but in one case a state of immunity to other female grafts was demonstrated after the initial transplant. Bacon and Craig ( 1966) also reported evidence for a heterogametic transplantation antigen, and in a subsequent paper (1969) described variable responses of roosters to the female antigen. Two inbred lines of white Leghorn chickens, C and RPL-6, and a brown Leghorn line, R, were studied. There were no female-to-male rejections observed among chickens of line R, but in most cases fcmale-to-male grafts were rejected in the other two lines. The F, hybrid males produced by reciprocal matings between lines R and RPL-6 in most cases rejected integumental grafts from females of either parental strain or the F, generation. This suggests that ( a ) although female-to-male grafts in line R were not rejected, the females in this line, nonetheless, must possess the heterogametic antigen, and ( b ) the genetic factors associated with the ability to reject female grafts were dominant. Thus this situation bears a striking similarity to the rejection of male skin isografts by females in mice and rats. Bacon reported (1970a) that when line C male embryos were injected at 14 days of incubation with white blood cells from line C females, permanent tolerance to female wattle grafts was induced. Evidence for a histocompatibility locus associated with the Z chromosome has been reported by Bacon (1970b). Scheinberg and Reckel (1961) found extracts of seeds of Lathyrus cicera or Pisum ar0en.w capable of agglutinating the red cells of female chickens which have reached sexual maturity. Agglutinability fell after ovariectomy and was restored by injections of diethylstilbesterol. Al-

242

DAVID L. GASSER AND WZLLYS K. SILVERS

though males do not normally express the agglutinogen (the “Hi” antigen), they could be induced to do so by repeated injections of diethylstilbesterol ( Scheinberg and Reckel, 1962). Evidence has been presented suggesting that the ability to produce the Hi substance is determined by a single autosomal gene (Scheinberg and Reckel, 1962). A similar antigen involving a hormone-influenced serum protein (Hip) in chickens was described by David (1971). This protein occurs in some laying hens and a naturally occurring anti-Hip antibody occurs in some hens lacking Hip, but neither the antigen nor the antibody were found in sera of roosters. Injections of progesterone have been shown to induce Hip, but not anti-Hip, in some roosters. Although Hip is similar in some respects to the Hi agglutinogen, it was shown by David that they are not the same antigen. There seems to be a genetic component involved in the determination of Hip, but the mode of inheritance has not been worked out as yet. A number of sex-specific antigens in Drosophila have been described which uniquely illustrate phenomena not yet observed in other organisms. Fox (1959) has described two male-specific antigens ( d-1 and 6”-2) and two female-specific antigens ( ? -1 and ? -2) which can be detected by rabbit antisera using the Ouchterlony agar diffusion technique. Fox has demonstrated that the presence or absence of these antigens is determined by the number of X chromosomes. Flies with two X chromosomes with a partial or complete Y chromosome express the female antigens but not the male antigens, whereas flies with a single X chromosome with no Y chromosomes, one Y chromosome, or more than one Y chromosome all express the male antigens and not the female antigens. Since sex determination in Drosophila is determined by the number of X chromosomes rather than by presence or absence of a Y chromosome ( Bridges, 1922), it seems possible that these four antigens could be products of sex differentiation. V. Summary

1. The Y antigen of inbred mice provides ready made what appears to be the simplest conceivable transplantation immunity system, in which individuals differ according to whether they have, or do not have, a single antigen of almost ubiquitous tissue distribution. 2. Although the question remains controversial, most of the evidence supports the hypothesis that the Y antigen is determined by a gene on the Y chromosome. 3. Nonreactivity of females of some inbred strains to isogenic male skin grafts is largely but not entirely determined by the H-2 locus or a gene closely linked to it.

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

243

4. The penetrance and expressivity of the gene determining the Y antigen are strongly influenced by the genetic background. 5. There is considerable evidence that Y incompatibility exerts a stimulatory effect on placental growth and some evidence that the sex ratio is also affected by such incompatibility. 6. Studies on inducing unresponsiveness to the Y antigen have been particularly informative because of the comparative ease, and variety of ways, with which tolerance of male skin isografts can be induced in neonatal and adult females. 7. There is evidence for an X-linked histocompatibility locus (H-X) in mice which conforms to the Lyon hypothesis. 8. The Xg locus in man does not appear to conform to the Lyon hypothesis. 9. A system which is apparently homologous to the murine H-Y occurs in rats, and a heterogametic antigen exists in female chickens which is similar in some respects to the murine H-Y system. 10. Several sex-limited antigens in D~osophilaare determined by the number of X chromosomes rather than the presence or absence of a Y chromosome, which corresponds to the mechanism of sex determination in this genus. 11. In addition to cases of sex-linked antigens, there are a number of well-documented examples of sex limited autosomally determined antigenic factors in various species. ACKNOWLEDGMENTS The investigations carried out in the authors’ laboratories were supported by research grants from the National Institutes of Health. One of us (DLG) is a Postdoctoral Fellow of the U. S. Public Health Service. We are grateful to Drs. Stephen S. Wachtel and William J. Mellman for critically reading the manuscript.

REFERENCES Allan, T. M. (1959). Brit. Med. J . 1, 553. Bacon, L. D. (1970a). Transplantation 10, 124. Bacon, L. D. (1970b). Transplantation 10, 126. Bacon, L. D., and Craig, J. V. (1966). Poultry Sci. 45, 1066. Bacon, L. D., and Craig, J. V. (1969). Transplantation 7, 387. Bailey, D. W. (1963a). Transplantation 1, 70. Bailey, D. W. (196313). Science 141, 631. Bailey, D. W. (1964). Transplantation 2, 203. Bailey, D. W. (1965). Transphtation 3, 531. Bailey, D. W. (1971). Transplantation 11, 426. Bailey, D. W., and Hoste, J. ( 1971). Transplantation 11, 404. Berg, K., and Bearn, A. G. (1966). 1. Exp. Med. 123, 379.

244

DAVID L. GASSER A N D WILLYS K. SILVERS

Bernstein, S. E., Silvers, A. A., and Silvers, W. K. (1958). J. Nut. Cancer Inst. 20, 577. Billingham, R. E. ( 1958). In “The Chemical Basis of Development” (W. D. McElroy and B. Glass, eds.), pp. 575-592. Johns Hopkins Press, Baltimore, Maryland. Billingham, R. E. (1964). N . Engl. J. Med. 270, 667 and 720. Billingham, R. E. ( 1971). Cell. Immunol. 2, 1. Billingham, R. E., and Brent, L. (1959). Phil. Tram. Roy. SOC. London, Ser. B 242, 439. Billingham, R. E., and Silvers, W. K. (1958). Science 128, 780. Billingham, R. E., and Silvers, W. K. ( 1959). Transplant. Bull. 6, 399. Billingham, R. E., and Silvers, W. K. (1960). J. Immunol. 85, 14. Billingham, R. E., and Silvers, W. K. (1970). J . Exp. Med. 131, 101. Billingham, R. E., Lampkin, G. H., Medawar, P. B., and Williams, H. L. (1952). Heredity 6, 201. Billingham, R. E., Brent, L.,and Medawar, P. B. (1956). PhiZ. Trans. Roy. SOC. London, Ser. B 239, 357. Billingham, R. E., Hodge, B. A,, and Silvers, W. K. (1962). Proc. Nut. Acad. Sci. U.S. 4 4 138. Billingham, R. E., Silvers, W. K., and Wilson, D. B. (1965). Proc. Roy. SOC., Ser. B 163, 61. Billington, W. D. (1964). Nature (London) 202, 317. Bonn&,B. (1963). Hum. Biol. 35, 61. Brent, L., and Courtenay, T. H. (1963). In “Mechanisms of Immunological Tolerance” (M. HaHek, A. Lengerovi, and M. VojtmovP, eds.), pp. 113-121. Academic Press, New York. Breyere, E. J., and Barrett, M. K. (1960). J. Nut. Cancer Inst. 25, 1405. Bridges, C. B. ( 1922). Amer. Nutur. 56, 51. Brosseau, G. E. (1960). Genetics 45, 257. Buckton, K. E., Cunningham, C., Newton, M. S., O’Riordan, M. L., and Sanger, R. ( 1971). Lancet 1, 371. Bunker, M. C. (1966). Can. 3. Genet. Cytol. 8, 312. Cattanach, B. M. (1961). Genet. Res. 2, 156. Celada, F., and Welshons, W. J. (1963). Genetics 48, 139. Chai, C. K. (1968). Transplantation 6, 689. Cinader, B., and Dubert, J. M. (1956). PTOC.Roy. SOC., Ser. B 146, 18. Clarke, B., and Kirby, D. R. S. (1966). Nature (London) 211, 999. David, C. S. ( 1971). J. Immunol. 106, 1104. DeVries, M. J., and Vos, 0. (1958). J. Nut. Cancer Inst. 21, 1117. Dronamraju, K. R. (1964). Nature (London) 201, 424. Dronamraju, K. R. (1965). Advan. Genet. 13, 227. Edwards, A. W. F. (1963). Nature (London) IQS, 1106. Eichwald, E. J., and Lustgraaf, E. C. (1961). J. Nut. Cancer Inst. 26, 1395. Eichwald, E. J., and Silmser, C. R. (1955). Transphnt. Bull. 2, 148 Eichwald, E. J., Silmser, C. R., and Wheeler, N. (1957). Ann. N . Y. Acad. Sci. 64, 737. Eichwald, E. J., Silmser, C. R., and Weissman, I. (1958). J. Nut. Cancer Inst. 20,

563.

Elkins, W. L. (1964). J. Immunol. 92, 275. Engelstein, J. M. (1967). Proc. SOC. Exp. B i d . Med. 126, 907. Feldman, M. (1958). Transplant. E d . 5, 15. Felton, L. D. (1949). J . Immunol. 81, 107.

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

245

Felton, L. D., and Ottinger, B. (1942). J. Bacterial. 43, 94. Felton, L. D., Kauffman, G., Prescott, B., and Ottinger, B. (1955a). J. Immunol. 74, 17. Felton, L. D., Prescott, B., Kauffman, G., and Ottinger, B. (195513). J . Immunol. 74, 205. Ferguson-Smith, M. A. (1965). J. Med. Genet. 2, 142. Ferguson-Smith, M. A. ( 1966). Lancet 2, 475. Fialkow, P. J,, Lisker, R., Giblett, E. R., and Zavala, C. (1970). Nature (London) 226, 367. Finkel, S . I., and Lilly, F. (1971). Nature (London) 234, 102. Ford, C. E., Jones, K. W., Miller, 0. J., Mittwoch, U., Penrose, L. S., Rider, M., and Shapiro, A. ( 1959). Lancet 1, 709. Fox, A. S. (1956a). Transplant. Bull. 3, 131. Fox, A. S. (195613). Physiol. 2001.29, 288. Fox, A. S. (1959). J. Nut. Cancer Inst. 23, 1297. Fox, A. S., and Yoon, S. B. ( 1958). Transplant. Bull. 5, 52. Furusawa, M., Kotani, M., and Takeuchi, H. (1963). Nature (London) 200, 182. Gasser, D. L., and Shreffler, D. C. (1972). Nature ( N e w Biol.) 235, 155. Gasser, D. L., and Silvers, W. K. (1971a). J. Immunol. 106, 875. Gasser, D. L., and Silvers, W. K. (1971b). Transplantation 12, 412. Gavin, J., Noades, J., Tippett, P., Sanger, R., and Race, R. R. (1964). Nature (London) 2&& 1322. Gilmour, D. G. (1967). Transphntation 5, 699. Gittes, R. F., and Russell, P. S. (1961). J. Nut. Cancer Inst. 26, 283. Goldberg, E. H., Boyse, E. A., Bennett, D., Scheid, M., and Carswell, E. A. (1971). Nature (London) 232, 478. Gorman, J. G., Dire, J., Treacy, A. M., and Cahan, A. (1963). J. Lab. Clin. Med. 61, 642. Hannover Larsen, J. (1971). J. Immunol. 107, 185. Halek, M., and HaHkovL, V. (1958). Cesk. Biol. 7, 282. Hauschka, T. S. (1955). Transplant. Bull. 2, 154. Hauschka, T. S., and Holdridge, B. A. (1962). Ann. N . Y . had. Sci. 101, 12. Hellstrom, K. E., Hellstkim, I., and Brawn, J. (1969). Nature (London) 224, 914. Heslop, B. F. (1969a). Proc. Uniu. Otago Med. Sch. 47, 53. Heslop, B. F. (196913). PTOC. Uniu. Otago Med. Sch. 47, 55. Hess, O., and Meyer, G. F. (1963). J. Cell Biol. 16, 527. Hildemann, W. H., and Cooper, E. L. (1967). Transplantation 5, 707. Hildemann, W. H., and Pinkerton, W. (1966). J. E x p . Med. 124, 885. Hoshino, K., and Moore, J. E. (19s8). Int. J. Cancer 3, 374. Hull, P. (1964). Genetics 50, 563. Hull, P. (1969). Heredity 24, 203. Jacobs, P. A., and Strong, A. (1959). Nature (London) 183, 302. James, D. A. (1965). Nature (London) 205, 613. Judd, K. P., and Trentin, J. J. (1971). Transplantation 11, 298. Kallman, K. D. (1970). Transplant. Proc. 2, 263. Katsh, G. F., Talmage, D. W., and Katsh, S. (1964). Science 143, 41. Kaufmann, B. P. (1933). PTOC.Nut. Acad. Sci. U . S. 19, 830. Kelly, W. D., Smith, J. M., Martinez, C., and Good, R. A. (1964). Proc. SOC. Exp. Biol. Med. 115, 8. Kelman, H., Martinez, C., and Good, R. A. (1961). Proc. SOC. Exp. Biol. Med. 106, 875.

246

DAVID L. GASSER AND WILLYS K. SILVERS

Kirby, D. R. S., McWhirter, K. G., Teitelbaum, M. S., and Darlington, C. D. (1967). Lancet 2, 139. Klein, E., and Linder, 0. ( 1961). Transphnt. Bull. 27, 457. Klein, J. (1965). Folia Biol. (Prague) 11, 169. Kozelka, A. W. (1932). 1. Exp. Zoo.?. 81, 431. Lambert, P. B., and Frank, H. A. (1970). J . Exp. Med. 132, 868. LappB, M., and Schalk, J. (1971). Transplantation 11, 491. Lee, G. R., MacDiarmid, W. D., Cartwright, G. E., and Wintrobe, M. M. (1968). Blood 32, 59. LengerovP, A,, and ChutnP, J. ( 1959). Folia Biol. (Prague) 5, 24. LengerovP, A., and VojtiSkovL, M. (1962). Folia Biol. (Prague) 8, 21. LengerovP, A., and VojtiBkovL, M. (1963). Folia Biol. (Prague) 9, 72. Lustgraaf, E. C., Fuson, R. B., and Eichwald, E. J. (1960). Transplant. Bull. 26, 145. Lyon, M. F. (1961).Nature (London) 190, 372. Lyon, M. F. (1962). Amer. J. Hum. Genet. 14, 135. Lyon, M. F. (1968). Annu. Reo. Genet. 2, 31. McDevitt, H. O., and Chinitz, A. (1969). Science 163, 1207. McDevitt, H. O., and Tyan, M. L. (1968). 1. Exp. Med. 128, 1. McDevitt, H. O., Shreffler, D. C., and Stimpfling, J. H. (1969). J . Clin. Inoest. 4 4 57a. McLaren, A. (1962). Nature (London) 195, 1323. McLaren, A. ( 1965). Transplantation 3, 259. Mann, J. D., Cahan, A,, Gelb, A. G., Fisher, N., Hamper, J., Tippett, P., Sanger, R., and Race, R. R. ( 1962). Lancet 1, 8. Mariani, T., Martinez, C., Smith, J. M., and Good, R. A. (1958). Proc. SOC. Exp. Biol. Med. 99, 287. Mariani, T., Martinez, C., and Good, R. A. (1960a). Int. Arch. Allergy Appl. Immunol. 18, 216. Mariani, T., Martinez, C., Smith, J. M., and Good, R. A. (196Ob). Ann. N. Y. Acad. Sci. 8'7, 93. Martinez, C., and Smith, J. M. (1964). Nature (London) 202, 508. Martinez, C., Shapiro, F., and Good, R. A. ( 1961). Proc. SOC. Erp. Biol. Med. 108, 476. Martinez, C., Blaese, M., and Good, R. A. (1963a). Nature (London) 198, 1319. Martinez, C., Smith, J. M., Blaese, M., and Good, R. A. (1963b). J . E r p . Med. 118, 743. Michie, D., and McLaren, A. (1958). Transplant. Bull. 5, 17. Miller, L. (1962). Transplant. Bull. 30, 147. Mintz, B., and Silvers, W. K. (1970). Transplantation 9, 497. Muller, H. J. (1914). J. E x p . ZOO^. 17, 325. Ohno, S. (1969). Annu. Rev. Genet. 3, 495. Ounsted, C., and Ounsted, M. (1970). Lancet 2, 857. Passmore, H. C., and Shreffler, D. C. (1970). Biochem. Genet. 4, 351. Passmore, H. C., and Shreffler, D. C. (1971). Bbchern. Genet. 5, 201. Pol&15kovP, M. (1969). Fol6 Biol. (Prague) 15, 181. PolAZikovP, M. (1970). Folia Biol. (Prague) 16, 12. PolA&kovP, M., and VojtEkovB, M. (1968). Folia Biol. (Prague) 14, 93. Prehn, R. T. (1960).J . Nut. Cancer Inst. 25, 883. Prehn, R. T. (1961). Transplant. Bull. 20, 487.

GENETICS AND IMMUNOLOGY OF SEX-LINKED ANTIGENS

247

Prehn, R. T., and Main, J. M. (1956). J. Nut. Cancer Inst. 17, 35. Race, R. R., and Sanger, R. (1968). “Blood Groups in Man.” Davis, Philadelphia, Pennsylvania. Rathbun, W. E., and Hildemann, W. H. (1970). J . Immunol. 105, 98. Renkonen, K. O., Makela, O., and Lehtovaara, R. (1962). Nature (London) 194, 308. Rosenau, W., and Horwitz, C. (1968). Lab. Invest. 18, 298. Russell, L. B. (1964). Trans. N . Y. Acad. Sci. [2] 26, 726. Sachs, L., and Heller, E. (1958). J. Nut. Cancer lmt. 20, 555. Sanger, R., Race, R. R., Tippett, P., Gavin, J., Hardisty, R. M., and Dubowitz, V. (1964). Lancet 1, 955. Scheinberg, S. L., and Reckel, R. P. (1961). Nature (London) 189, 295. Scheinberg, S. L., and Reckel, R. P. (1962). Ann. N. Y. Acud. Sci. 97, 194. Schultz, J. (1956). Cold Spring Harbor Symp. Quant. Biol. 21, 307. Short, B. F., and Sobey, W. R. (1957). Transplant. Bull. 4, 110. Shreffler, D. C., and Owen, R. D. (1963). Genetics 4 4 9. Silvers, W. K. (1968). J . E x p . Med. 128, 69. Silvers, W. K., and Billingham, R. E. (1967). Science 158, 118. Silvers, W. K., Billingham, R. E., and Sanford, B. H. (1968). Nature (London) 220, 401. Slatis, H. M., Reis, R. H., and Hoene, R. E. (1958). Amer. J. Hum. Genet. 10, 446. Steinmuller, D. ( 1967). Science 158, 129. Stem, C. (1927). Z. Indukt. Abstamm.- Vererbungsl. 44, 187. Stern, C. (1929). Z. Indukt. Abstamm.- Vererbungsl. 51, 253. Stern, C. (1957). Amer. 1. Hum. Genet. 9, 147. Stimpfling, J. H., and Reichert, A. E. (1971). Transplantation 12, 527. Toivanen, P., and Hirvonen, T. ( 1970). Clin. Exp. Imrnunol. 7, 533. Triplett, E. L. ( 1962). J. Immunol. 89, 505. Vaz, N. M., and Levine, B. B. (1970). Science 168, 852. Vener, J., Martinez, C., and Good, R. A. ( 1961). Proc. SOC. Erp. Biol. Med. 108, 480. VojtBkovA, M., and PoliEkovA, M. (1966). Folia Biol. (Prague) 12, 137. Wachtel, S. S., and Silvers, W. K. (1971). Adoan. Biol. Skin 11, 223. Wachtel, S. S., and Silvers, W. K. (1972a). J. E x p . Med. 135, 388. Wachtel, S. S., and Silvers, W. K. (1972b). In preparation. Wachtel, S. S., Gasser, D. L., and Silvers, W. K. (1973). Transplant. Proc. In press. Warburton, D., and Naylor, A. F. ( 1971). Amer. J . Hum. Genet. 23, 41. Weatherall, D. J., Pembrey, M. E., Hall, E. J., Sanger, R., Tippett, P., and Gavin, J. ( 1970). Lancet 1, 744. Welshons, W. J., and Russell, L. B. (1959). Proc. Nut. Acad. Sci. U.S . 45, 560. Yang, S. L., and Silvers, W. K. ( 1971). Unpublished experiments. Zaalberg, 0. B. (1959). Transplant. Bull. 6, 433. Zeiss, I. M., Nisbet, N. W., and Heslop, B. F. (1962). Transplant. Bull. 30, 161.

Current Concepts of Amyloidl EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN Rheumatic Diseases Study Group, Deparlment of Medicine, New York University Medical Center, New York, New York

I. Introduction . . . . . . . . . . . . . . . 11. General Approaches to the Study of Amyloid 111. Morphological Studies in Sectioned Tissues . . . . IV. Morphological, Biochemical, Physicochemical, and Antigenic Properties of Amyloid Fibrils . . . . . . . A. Dye Binding . . . . . . . . . . . B. Morphological Studies . . . . . . . . C. Physicochemical and Biochemical Studies of Amyloid Fibrils and Their Subunits . . . . . . . . . D. Immunological Studies . . . . . . . . E. Comparative Studies of Different Amyloid Preparations . F. Discussion and Conclusions . . . . . . . . . . . V. Speculations on the Pathogenesis of Amyloid References . . . . . . . . . . . .

. . . . . .

249 253 255

. . . . . .

264 264 266

.

. .

. .

.

. . .

. . .

270 277 285 289 294 299

I. Introduction

Amyloid appears in the light microscope as an amorphous, eosinophilic, hyaline, extracellular substance which is readily recognized by its unique green birefringence after staining with Congo red when viewed in the polarizing microscope and by its metachromatic properties with dyes such as Crystal Violet, Toluidine Blue, or Methyl Violet (Bennhold, 1922; Cohen, 1966b, 1967). When present, it is usually widely distributed in many organs of the body and frequently replaces the cellular elements present in these tissues. Although most commonly it accompanies a large number of apparently unrelated disorders, it is occasionally seen as a primary illness unassociated with any underlying pathological conditions. Although organs with the gross appearance of amyloid infiltrates have been recognized for over 300 years, the first clear-cut descriptions of the “lardaceous liver” and “sago spleen,” now recognized as being characteristic of this disorder, can be attributed to von Rokitansky (1846) and Virchow ( 1854), respectively. The substance which infiltrates these tissues and which is presumably responsible for this appearance was mis‘This work is supported by U. S. Public Health Service Grants #AM 02594, #AM 01431, and #AM 012274; the Health Research Council of the City of New York and U. S. Public Health Service Research Career Development Award #5-K03 A1 09572. 249

250

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

named “amyloid by Virchow because it resembled starch in its staining properties with iodine and sulfuric acid. The name has remained, despite the fact that as early as 1859, Friedreich and Kekulb discovered the now generally accepted protein nature of amyloid. Progress in elucidating the pathogenesis of these disorders, the nature of the deposits, and their precise origin has been slow. Difficulties in defining the etiology of the disease arise mostly because in man, and many species of animals, amyloid is associated with a variety of diseases that appear to have few if any features in common. Confusion has been compounded by the plethora of agents and procedures, again with few obvious common properties, which can induce amyloid in experimental animals. As a result, it has proved difficult to identify any single mechanism responsible for this disorder. As mentioned below, it seems possible that many different factors may ultimately interact or initiate a limited number of stimuli that can induce the synthesis and deposition of amyloid. (For a review, see Mandema et al., 1968.) Nevertheless, during the past few years a number of significant observations have been made which have given us a clearer insight into the nature of this material and promise to provide us with a more unified concept of the pathogenesis of amyloid. Impetus for a new look at amyloid was provided by the demonstration in several laboratories of distinct and characteristic fibrils in what had been presumed previously to be a homogeneous structureless material (Cohen and Calkins, 1959; Spiro, 1959). The availability of purified fibrils free of all other soluble and insoluble tissue constituents then permitted direct chemical, physical, and immunological analyses of the major constituent of amyloid. As a result of these approaches we have gained knowledge of the ultrastructure of amyloid (Shirahama and Cohen, 19f37b; Pras et al., 1968) and have obtained antisera which appear to react with amyloid specifically. This may prove to be of great value diagnostically (Franklin and Pras, 1969; Glenner et al., 1969; Cathcart et al., 1970a). Most importantly, it has been clearly established that amyloid is often closely related structurally to immunoglobulins, in particular to the variable region of the light polypeptide chain (Glenner et al., 1970b, 1971b; Harada et al., 1971). If the fragments of immunoglobulins can, indeed, be shown to be present in all types of amyloid, it seems likely that the excessive production of such fragments may be the final common effector mechanism resulting in the deposition of amyloid. In spite of the promise that these studies will in time permit the development of an accurate and meaningful system of classification based primarily on the biochemical properties of different types of amyloid, analogous to that now known for the closely related immunoglobulins,

CURRENT CONCEPTS OF AMYLOJD

251

these studies have not progressed sufficiently to permit the delineation of such a system at this time. As a consequence, we shall introduce this review with a composite classification which incorporates features of several of those currently employed. It seems quite likely that all of these may become obsolete and will be replaced eventually by a rational system based on pathogenetic mechanisms and chemical properties of different types of amyloid. Even in our present state of knowledge, it is apparent that none of the currently employed systems of classification permits a clear-cut separation of amyloid into different classes since there exists a great deal of overlap in the tissue distribution and staining characteristics of the several major types of amyloid. Consequently, although they are widely used and are of some clinical value, it is commonly felt that each one of these classifications is somewhat arbitrary and indistinct. The simplest and most widely used classification of amyloidosis is based on the four major categories introduced by Reimann (Reimann et al., 1935). ( 1 ) Primary amyloid occurs with no known antecedent or coexistent disease and usually involves mesodermal tissues, such as smooth or skeletal muscle, or the cardiovascular system. Generally there is variability in staining of the deposits with Congo red, iodine, and metachromatic dyes. ( 2) Secondary amyloid is usually associated with chronic diseases such as infections, neoplasms, neurological disorders, or connective tissue diseases especially rheumatoid arthritis. It generally involves spleen, liver, kidney, intestines, and adrenals and tends to have reproducible and characteristic staining properties with the abovementioned dyes. ( 3 ) Amyloid associated with myeloma tends to resemble the primary type but is invariably associated with a neoplasm involving plasma cells or lymphocytes, such as multiple myeloma, macroglobulinemia, or heavy chain disease. (4)Tumor-forming amyloid is characterized by small masses of amyloid in the skin, eye, bladder, urethra, respiratory tract, or other organs and is generally unassociated with any underlying disease. In addition to these four major classes of amyloid, a large number of familial types of amyloidosis, each involving a different organ system and characterized by a different form of inheritance, have been reported from many parts of the world ( Andrade et al., 1970). The most interesting, and probably also the most common one is the type of amyloid associated with familial Mediterranean fever (FMF), most often seen in Sephardic Jews, but also occasionally in other ethnic groups (Sohar et al., 1967). Although many of the clinical and pathological features of these familial types of amyloidosis might be included in the primary and secondary types, it appears better at this time to keep them separate since

252

EDWARD C. FRANKLIN AND W R O T H E A ZUCKER-FRANKLIN

the basic mechanism responsible for the deposition of amyloid may not be the same as in the sporadic cases. Nevertheless, it seems appropriate to mention at this point that electron microscopically amyloid from all of these types is indistinguishable and that chemical studies of some of the proteins isolated from patients with the primary, secondary, myelomaassociated amyloid, and at least one familial type, that associated with FMF, have revealed similarities (Pras et al., 1969) and have recently been shown to consist of a fraction of immunoglobulin light chains (Glenner et al., 1970b, 1971a,b,c; Harada et al., 1971). Thus, on the basis of these studies it seems possible that the amyloid produced in all of these disorders may be chemically similar and that the characteristic tissue distribution may be related to other, as yet poorly defined, factors. However, one should bear in mind the recent observations of Benditt and Eriksen (1971), Franklin et al. (1972), and Pras et al. ( 1971) which have raised the possibility that some amyloids, especially those obtained from patients with the so-called secondary type of amyloid and the type associated with FMF, may be chemically different and consist of a major protein that can be isolated in dilute acid and which, on the basis of its amino acid analysis and size, does not appear to be related to any known immunoglobulin fraction. Rather detailed reviews of two large and well-studied series of patients with amyloidosis have been published recently (Brandt et al., 1968; Barth et al., 1969b). These two reports provide ample clinical descriptions of the various types of amyloidosis and also emphasize some similarities in the clinical and pathological features of the three major types of amyloidosis. This subject will not be discussed in detail in this review. A different type of classification, initially based on the positive birefringence of amyloid and its conversion to negative by phenol or glycerol, has divided amyloid into the perireticular and the pericollagenous types (Missmahl and Hartwig, 1953; Missmahl, 1957, 1968; Heller et al., 1964). In the perireticular form, there appears to be a generalized vascular disease resulting from the deposition of amyloid starting at the basement membrane of the vessels and spreading outward. This type of amyloid is seen in the secondary type, certain familial forms including FMF, and a number of instances of primary or idiopathic amyloidosis with the so-called “typical distribution.” The pericollagenous type is said to start in the connective tissue, beginning in the adventitia of the vessel from where it appears to spread inward, and is most commonly seen in the classic primary type, in amyloidosis associated with multiple myeloma, in the tumor-forming types of amyloid, and in certain other familial types of amyloidosis. Although this classification has been used

CURRENT CONCEPTS OF AMYLOID

253

in a number of recent studies, it appears to us to add little in terms of a greater understanding of the disease and is mentioned here only for the sake of completeness. Before proceeding to a detailed discussion of certain aspects of amyloidosis, a few words seem appropriate concerning the aims of this review. Rather than to provide an exhaustively documented, detailed, complete, historical coverage of the field, we shall discuss in a selective fashion current concepts and in many instances controversial and as yet unsettled questions related to the structure, chemistry, immunology, cellular origin, metabolism, and possibly pathogenesis of amyloid, citing only certain references that appear to us to be pertinent to these areas. The interested reader who wishes a more detailed documentation of the field might do well to read the reviews by Cohen (196613, 1967) and Mandema et al. (1968). I I . General Approaches to the Study of Amyloid

Two major approaches have been employed in recent years to characterize amyloid more definitively. One of these, the morphological approach, has used the electron microscope ( E M ) to study amyloid. Initially EM studies were limited to examination of infiltrated tissues from man and experimental animals. With the identification of characteristic fibrils as the major constituent of amyloid and the development of methods for their isolation, similar studies have now also been carried out on the isolated amyloid fibrils. As a result of studies of the partially purified fibrils, two major structural components have been identified: ( 1 ) the characteristic fibrillar component makes up more than 955%of amyloid and appears to be unique to this material (Shirahama et al., 196713; Pras et al., 1968; Glenner et al., 1969; Harada et al., 1971); (2) another structure, known as the “rod,” the “doughnut,” or the “P” component has not been identified in tissues but has been seen in partially purified preparations (Bladen et al., 1966; Cathcart et al., 1 9 6 7 ~ ) .It appears to represent a normal serum a,-globulin and makes up a much smaller fraction of the amyloid complex. It is largely removed during most fractionation procedures and its precise role in the pathogenesis of amyloid remains to be elucidated. The other approach to study the nature of amyloid has attempted to characterize it chemically and immunologically. Many investigators have stained amyloid tissues with fluorescent antisera to a variety of serum proteins in the hope of detecting components that may be important constituents of amyloid deposits. In this manner, a large number of different serum proteins have been identified in amyloid deposits. Of particular interest, in the light of the recent chemical studies and be-

254

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

cause their direct bearing on immunological factors thought to be of importance in the pathogenesis of the disorder, have been the immunoglobulins and certain complement components, demonstrated in amyloid by Mellors and Ortega (1956), Cathcart and Cohen ( 1966), Cathcart et al. (1967a), Lachmann et al. (1962), Milgrom et al. (1966), Muckle (1968a,b), Vazquez and Dixon (1956), Vogt and Kochem (1960), Williams et d . ( 1960), and others. Although these observations have at times been conflicting in nature and difficult to interpret they have greatly influenced our concepts of etiology and pathogenesis. A more direct approach, which during the past year has provided clear-cut evidence that some amyloid is composed of immunoglobulin light-chain fragments, has been to isolate the major constituents of amyloid and to subject them to chemical and immunological analysis. Such studies have clearly demonstrated that the major fibrillar component isolated from a number of different types of amyloid is composed of a subunit ranging in molecular weight from 5000 to 30,000 daltons which includes at least the first part of the variable region of the light chain of ,-globulin. The possibility that it encompasses additional parts of the 7-globulin molecule has not been ruled out and in certain instances appears likely (Glenner et al., 1970b, 1971a,b,c; Harada et al., 1971). The likelihood that not all amyloid is related to light chains was raised by the unusual protein devoid of cysteine described by Glenner et al. (1970b). This was also suggested by the recent finding of large amounts of a homogeneous protein which does not appear to be related to immunoglobulins in dilute acid extracts from amyloid tissues (Benditt and Eriksen, 1971) and pure amyloid fibrils (Pras and Reshef, 1972; Franklin et al., 1972). In contrast, the second constituent, known as the “doughnut” (Glenner et al., 1970a,b), is a serum a,-globulin that assumes a peculiar morphological appearance in amyloid (Bladen et al., 1966; Cathcart et al., 1967a,c). This protein is not unique to amyloid and its role in the formation and pathogenesis of amyloid deposits remains as yet unexplained. None of the other serum constituents that have been identified in amyloid tissue sections can be detected after further purification, and their specific role in the deposition of amyloid deposits, therefore, remains doubtful. Since the most meaningful studies in recent years have been carried out with isolated or partially purified fibrils, it seems appropriate to list a few of the methods most commonly employed in their purification prior to providing the detailed results from these studies. One of these, the method of Cohen and Calkins (1964) subjects the amyloid-laden tissue to homogenization and maceration in physiologic saline. The homogenate is then centrifuged at 12,OOOg for 30 minutes

CURRENT CONCEPTS OF AMYLOID

255

and the top layer, rich in amyloid fibrils, is subsequently subjected to repeated sucrose gradient centrifugation ( Cohen, 1966a) to obtain a relatively pure fibril preparation. The other, the method of Pras et al. (1968), is based on the observation that amyloid fibrils are insoluble in physiologic saline but that they are readily extracted in distilled water. Consequently, the tissue is repeatedly homogenized in solutions of physiologic saline and Centrifuged at about 9OOOg for 30 minutes. The contaminating proteins are discarded in the supernatant, and the pellets are rehomogenized and recentrifuged until the O.D. of the supernatant at 280 nm. is 0. The pellet is then repeatedly rehomogenized in distilled water and the resultant homogenate is centrifuged at 80,000 g for 1hour. After three or four such extractions, the supernatant becomes opalescent, contains Congo red-binding material, and can be shown to contain from 20 to 95%of the amyloid fibrils present in the starting tissue. Further purification can be achieved by column chromatography on Sepharose or Sephadex in solvents containing guanidine with a reducing agent in order to dissociate noncovalent and disulfide bonds ( Glenner et al., 1969). This step provides not only some additional purification but also dissociates the polymers into their basic subunits which can then be subjected to further chemical studies. For electron microscopy, the purified fibrils can be pelleted, fixed, dehydrated, and embedded by standard methods (Luft, 196l), following which they are sectioned and positively stained with lead (Millonig, 1961) and uranium salts (Watson, 1958). Alternatively, the purified fibrils can be placed directly on Formvar-covered EM grids and then negatively “stained with heavy metals ( Brenner and Horne, 1959). Ill. Morphological Studies in Sectioned Tissues

The fibrillar nature of amyloid tissue deposits was reported in 1959 (Cohen and Calkins, 1959; Spiro, 1959) and has since been illustrated repeatedly (Hjort and Christensen, 1961; Gueft and Ghidoni, 1963; Thiery and Caroli, 1961; Boere et al., 1965; Sorenson and Bari, 1968; to name a few). A typical example of an amyloid infiltrated lymph node is seen in Fig. 1. It is not surprising that the fibrils were not recognized earlier since even on electron microscopy at low magnification the material appears amorphous and mixed with cellular debris; only at higher resolutions are the fibrils clearly delineated (Figs. 2 to 4). In some tissues, particularly in those where parenchymal cells are held together in a loose fashion, such as the spleen, lymph nodes, and bone marrow, the fibrils are oriented at random and can be seen to crisscross in all directions. In such sections, individual fibrils measure about 120~4.in width and usually do not bend or branch. In other areas, amyloid fibrils may

256

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

CURRENT CONCEPTS OF AMYLOID

257

be seen in more organized arrays (Fig. 3 ) at right angles to or in parallel with adjacent cell membranes or bundles of collagen fibrils. To date no significance has been attributed to any particular organization the fibrils may display; rather it is tacitly assumed that parallel bundles of fibrils reflect the stresses and strains of the tissues in which they are located or, as the case may be, the deformability of the surrounding cells. In view of the close relation of some amyloid fibrils to immunoglobulin light chains, it becomes of utmost importance to delineate the mechanisms involved in the deposition of amyloid and to define the cells responsible for its synthesis. In the absence of detailed in oitro biosynthetic studies, some information bearing on these questions can be obtained from E M observations. As regards the deposition of the type of amyloid related to immunoglobulins, at least two possibilities, which are not mutually exclusive, must be considered. On the one hand, light chains may be synthesized and secreted by plasma cells and circulate in the blood. They may, by as yet unknown mechanisms, undergo the changes requisite for assuming the characteristic morphological appearance of amyloid and then interact with, as yet undefined, tissue constituents or receptors to form amyloid deposits. Evidence possibly in favor of such a mechanism was provided by the studies of Osserman (Osserman et al., 1964) which demonstrated the binding of fluoresceinated Bence-Jones proteins to tissue sections prepared from muscle, kidney, liver, and intestine. It is of interest that Bence-Jones proteins both from patients with and without amyloid bound to tissues, with possibly somewhat more intense staining by proteins from patients with amyloidosis. However, there was no significant degree of tissue specificity which paralleled the distribution of amyloid infiltrates in the cases from which the Bence-Jones proteins were obtained. Alternatively, amyloid may be deposited in close proximity to cells that synthesize a substance which assumes the typical fibrillar appearance in their immediate vicinity without first having to circulate in the bloodstream. Such a view is favored by the not infrequent existence of masses of amyloid in close proximity to plasma cells (see Fig. 1).Although the precise mechanism for the synthesis and the sites of assembly of the fibrils from the soluble precursor subunits remains to be estabFIG. 1. Thin section of a lymph node infiltrated with amyloid obtained from a patient with multiple myeloma. The amyloid deposit (Am) is surrounded by plasma cells ( P ), monocytoid cells ( M ), lymphocytes ( L ), and reticuloendothelial ( RE ) cells. Magnification: X 4000. ( From Zucker-Franklin and Franklin, 1970.) FIG. 2. Detail of a lymphocytoid plasma cell surrounded by amyloid fibrils (AM). Arrow points to intracytoplasmic fibrils. Nucleus ( N); rough endoplasmic reticulum ( E R ) . Magnification: X 23,000. (From Zucker-Franklin and Franklin, 1970.)

258

EDWARD C. FRANKLIN AND DOROTHFA ZUCKER-FRANKLIN

259

CURRENT CONCEPTS OF AMYLOID

lished, it seems possible that the fibrils are formed intracellularly and released by cellular dissolution or, alternatively, that a soluble precursor is released which polymerizes to form the fibrils either on the outer cell membrane or in the surrounding tissue fluid. Whereas EM studies have demonstrated fibrils intracellularly on occasion ( see below), examination of fixed tissues cannot provide kinetic evidence allowing a choice between the different pathways of fibril formation. The problem is compounded by the fact that the deposits are often found in such close association with surrounding cells that it is difficult to determine whether the fibrils are inside or outside the cell (Figs. 4 and 5 ) . This is further aggravated by the state of the cells concerned which are often damaged or necrotic (Fig. 5A and B). Even when the cells are well preserved, their plasmalemmas have often disappeared in the area adjacent to the amyloid deposit. When no fibrils are seen in the cytoplasm of such cells, there seems to be little question that the fibrils have indented the cell from the outside (Fig. 4). At a recent symposium on amyloidosis, such illustrations prompted Bywaters to draw an analogy between the appearance of these cells and San Sebastian transfixed with arrows: “He looks a bit sick too, but nobody has suggested that he was secreting the arrows” (Bywaters, 1968). Although San Sebastian adorns the cover of the Proceedings of this meeting, the final word on the subject has not been spoken. Ranlov ( 1968) clearly demonstrated intracellular fibrils which appear to form an integral part of the cytoplasm in the spleen of experimental mice, and in our own studies on human specimens, this impression has been amply corroborated ( Zucker-Franklin and Franklin, 1970). In organs heavily infiltrated with amyloid, cells are often seen to contain fibrils which are morphologically identical to those in the extracellular space (Figs. 2, 6, and 7 ) . In many instances, the cell appears otherwise normal and displays an intact surface membrane, at no time suggesting that penetration into the cell by the fibrils has taken place. Other cells show degenerative changes with displacement of organelles by the cytoplasmic fibrils. It should be emphasized that intracellular fibrils are only seen in a small percentage of specimens and then mostly in heavily infiltrated organs such as lymph nodes, spleen, and bone marrow. Ranlov (1968) and Ranlov and Wanstrup (1968) interpreted the intracellular amyloid precipitation as a “fatigue phenomenon” on the part of the cell, ~~

~~

FIG. 3. Amyloid fibrils seen in bundles parallel to cell membrane. Paired filaments may be distinguished within each fibril. Magnification: ~64,000. FIG. 4. Detail of a reticuloendothelial cell surrounded by amyloid (Am)fibrils. The arrows point to areas where the fibrils appear to invaginate the cell. Intracytoplasmic fibrils can also be seen. Magnification: X29,OOO. (From Zucker-Franklin and Franklin, 1970.)

260

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

CURRENT CONCEPTS OF AMYLOID

261

and postulated that it only takes place in cases where amyloid deposition is unusually fast. Indeed, it is easy to accept the hypothesis that intracellular polymerization or precipitation of the protein could occur if the excretory mechanism of the cells were impaired. Although the intracytoplasmic fibrils illustrated in Figs. 6 and 7 are morphologically similar to those in the extracellular space and may, on the basis of immunofluorescence studies, represent amyloid, one should not ignore the fact that structurally identical fibers are occasionally encountered in epithelial cells of normal subjects. Such fibrils are usually referred to as tono fibrils (Fawcett, 1966) or stress fibrils (Buckley and Porter, 1967). The biochemical nature or function of such fibrils in normal cells is entirely conjectural. However, their occurrence raises the possibility that small amounts of amyloid are synthesized by such cells under normal conditions. If this were true, extracellular amyloid would reflect overproduction of a normal cell constituent rather than an abnormal protein. Apart from the debate as to whether or not the intracellular fibrils represent amyloid is the controversy over the specific type of cells responsible for its synthesis. In early studies, Cohen et al. (1965) performed tissue culture studies on amyloid-infiltrated rabbit spleens and, with the help of isotope-labeled amino acids, showed that the label was incorporated in the reticuloendothelial ( R E ) cells before it appeared on the extracellular fibrils. Other investigators studying liver and splenic tissues (Heefner and Sorenson, 1962; Gueft and Ghidoni, 1963; Teilum, 1968; Laufer and Tal, 1967), also incriminated RE cells, and Battaglia ( 1961, 1962) even suggested that during the induction of experimental amyloidosis in mice, amyloid could be seen in Kupffer cells. He postulated that following rupture of these cells there was spillage of the fibrils into the space of Disse. In studies of bone marrow, Ben-Ishay and Zlotnick (1968) also found intracellular fibrils in RE cells. In the central nervous system, Terry (1963, 1964) described typical fibrils within socalled senile plaques. The amyloid nature was confirmed by staining with Congo red and birefringence on polarization microscopy. In this location the fibrils related closely to the surrounding glial cells-the nervous system’s counterpart of RE cells. However, other cells not belonging to the RE system (RES) in a strict sense, have also been considered to produce FIG. 5. ( A ) Degenerated lymphoid cell entirely surrounded by amyloid (Am) obtained from bone marrow specimen of a patient with IgA myeloma. The plasma membrane of this cell does not seem to constitute a barrier between the cytoplasm and the extracellular space (arrow). Nucleus ( N u ) . ( B ) Area in inset of Fig. 5A shown at a higher magnification. At this site the plasma membrane has disappeared; intra- and extracellular fibrils are seen in continuity. Magnification: x 120,000.

262

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

CURRENT CONCEPTS OF AMYLOID

263

amyloid. In serial studies on renal amyloidosis, experimentally induced with casein in rabbits, Cohen and Calkins (1960) found the earliest deposits between endothelial cell cytoplasm and basement membrane. When deposits became more abundant, the epithelial side of the basement membrane also became involved. The function of the basement membrane as a barrier or accomplice in the development of the renal lesions has never been clearly defined. From such studies, one would have to conclude that the endothelial cell is also capable of synthesizing amyloid. However, Shibolet et al. ( 1967), who followed the development of amyloid in the kidneys of Leishmania-infected Syrian hamsters, concluded that fibroblasts infiltrating the interstitial space between capillary and tubular membranes were involved in the deposition of amyloid. Still a different opinion was expressed by Sorenson and Shimamura (1964) who induced renal amyloidosis in mice and found the first accumulations of fibrils in the mesangial area of the glomerulus. In the human only a limited number of serial studies on the deposition of amyloid in the kidney are available. However, particular attention has been drawn to the mesangial cell matrix and basement membrane as the earliest region involved in human renal amyloidosis (Suzuki et aZ., 1963). Furthermore, in light of recent chemical studies of amyloid and in view of the concepts that immunological mechanisms are important in the pathogenesis of amyloid, the most interesting cell that has been considered to play a role in the production of amyloid is the plasma cell. On the basis of histochemical techniques, Teilum (1956, 1964a,b) contends that it is this cell which is primarily responsible for the synthesis of this protein. His view has been shared by Caesar (1960). In our own studies ( Zucker-Franklin and Franklin, 1970), intracellular fibrils were seen in RE, “monocytoid” cells, and also in well-delineated plasma cells (Figs. 6 and 7 ) surrounding or immediately adjacent to extensive extracellular deposits. If it is difficult to consider the possibility that all the above-named cells are capable of elaborating amyloid or a soluble precursor substance, it may be helpful to recall that all these cells are of mesenchymal origin and that they may, therefore, be potentially capable

FIG. 6. Detail of a plasma cell showing intracytoplasmic fibrils obtained from a lymph node of a patient with multiple myeloma. Rough endoplasmic reticulum (ER) as well as free and clustered ribosomes are abundant. Mitochondria ( M ) . Magnification: X 37,000. (From Zucker-Franklin and Franklin, 1970.) FIG.7. Detail of a plasma cell with intracytoplasmic fibrils. Most of the fibrils are located around the Golgi zone ( G ) , but some bundles run more peripheral. Mitochondria ( M ) ; rough endoplasmic reticulum ( ER ) ; nucleus ( N ) ; extracellular amyloid (Am). Magnification: x20,OOO.

264

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

of synthesizing the same protein, as a result of dedifferentiation or in response to a particular stimulus. IV. Morphological, Biochemical, Physicochemical, and Antigenic Properties of Amyloid Fibrils

Many controversial data have accumulated during the last 125 years concerning the nature of amyloid. This confusion persisted because until recently most of the analyses have been carried out on tissue extracts and homogenates which were contaminated to various extents by nonamyloid components. Only with the availability of the major fibrillar components of amyloid in pure form has it been possible to obtain reasonably firm data on the properties of amyloid. Even here, however, some confusion arose initially because the P component, which has a characteristic rodlike or doughnutlike appearance in the EM (see below) was initially thought by some to represent the fibrillar component (Glenner and Bladen, 1966). In this section we shall, therefore, concentrate primarily on the recent studies of the purified fibrils and summarize only briefly the limited biochemical studies on the P component. The interested reader can find complete historical reviews of the studies on the chemistry of amyloid elsewhere (Cohen, 1966b, 1967; Ashkenazi et al., 1968; Mandema et al., 1968).

A. DYEBINDING Probably the single most useful property of the amyloid fibrils, which has permitted their extraction from tissues in high yield and great purity, is their unusual behavior in the presence of salt (Pras et al., 1968). Unlike most other proteins the major fibrillar protein component of amyloid is completely insoluble and presumably highly polymerized even at low concentrations of salt (0.025 M NaCl) , thus permitting the removal of most contaminants by repeated extraction in 0.15 M NaCI. In the absence of salt, and following homogenization, amyloid changes its physical properties sufficiently so that it can be removed from the remaining insoluble substance by ultracentrifugation. The distilled water extract of amyloid is generally rather opalescent, quite unstable, and tends to precipitate on prolonged standing. Because of this and the presence of typical fibrils in the EM, it seems likely that amyloid in distilled water represents a suspension or colloidal solution rather than a true solution. Although the distilled water extract has a number of properties characteristic of amyloid, we do not know if it contains the intact fibril or a natural subunit which can polymerize to yield the fibril seen in the EM. The unusual solubility properties of amyloid are shown in Fig. 8.

CURRENT CONCEPTS OF AMYLOID

0005 0010 a015 0.0200025

265

,,

0.075"0.15 SALT CONCENTRATION , rnole/liter

FIG. 8. Relation between the amount of amyloid left in solution and the concentration of added salt, NaCl or CaCL The amount of amyloid left in solution was determined in two ways: ( a ) by the percent of the initial absorbance at 280 mp. left in the supernatant solution after adding salt and removing precipitated amyloid by centrifugation, and ( b ) by measuring the percent of Congo red in saline not precipitated when added to the same supernatant solution. The solid Congo red line was obtained with the supernatant solutions after addition of NaCI. (From Pras et d., 1968.)

The binding of Congo red by amyloid and the associated green birefringence have long been recognized and used to detect amyloid in tissue (Bennhold, 1922; Gafni et al., 1968; Benditt et al., 1970). More recently, the ability of amyloid to combine with Congo red quantitatively and presumably stoicheometrically has proved of great value in following the purification of the amyloid material (Pras et al., 1968) and in studying the reconstitution of fibrils from soluble precursors. Another characteristic property of amyloid is metachromasia with a number of dyes, such as Crystal Violet or Toluidine Blue. Thus the addition of amyloid to these dyes causes a shift in the absorption peak from about 600 to 550 mp. which can be perceived grossly and measured accurately with a spectrophotometer (Pras and Schubert, 1968). It is of interest that the acid mucopolysaccharides responsible for the metachromasia are probably not an integral part of the fibril protein and that they can be extracted from it (Pras et al., 1971). The solubility and dyebinding properties appear to be characteristic of all amyloid fibrils studied to date; however, the quantitative aspects of the Congo red binding may vary for different specimens suggesting chemical differences between different amyloid preparations. The availability of purified fibrils has permitted a series of ultrastructural, biochemical, physicochemical, and immunological studies designed to characterize these substances and has, in recent years, provided significant insight into the possible nature of amyloid.

266

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

B. MORPHOLOGICAL STUDIES When the isolated fibrils are embedded, sectioned, and examined in the EM, they appear identical to those in freshly fixed amyloid-laden tissues obtained at autopsy or on biopsy (Fig. 9A and B). Their width is about 100 to 150A and their length cannot be determined since they crisscross into and out of the plane of section. A large percentage of the fibrils consists of two longitudinal subunits or filaments equal in width and separated by a space of +25 A. The two subunits remain parallel throughout their course. Scattered among the fibrils are irregularly shaped dots which are assumed to represent tangential and cross sections of the fibrils. Frequently, the cross sections show a radiolucent core which is more apparent in some preparations than in others (Fig. 9B ) . When the “soluble” fibrils are salt-precipitated rather than precipitated by ultracentrifugation, they form thicker bundles which run parallel for short distances after which they twist, cross, or divert. Single subunits are more difficult to find in such preparations. However, the basic structure does not change suggesting that salt merely promotes

FIG.9. ( A and B ) Isolated fibrils that have been embedded and sectioned. Note cross sections of fibrils in Fig. 9B show hollow core (arrows) ; also fibrils appear different in quality and size. Both taken at same magnification: x102,OOO.

CURRENT CONCEPTS OF AMYLOID

267

aggregation andlor polymerization of the fibrils and that the material obtained in distilled water represents a finer dispersion of the basic subunits, When “soluble amyloid is negatively stained, i.e., when the fibrils are directly placed and stained on an EM grid, without prior fixation or embedding, greater resolution can be obtained. In such preparations the most common configuration seen is the paired filament (Fig. 10). Individual filaments measure 50-75A. in diameter and range from several millimicrons in length to the smallest resolvable fragments. The filaments have a beaded or helical substructure (Shirahama and Cohen, 1965). These may lend themselves to further structural analysis as smaller longitudinal subunits or protofibrils, measuring only 20-25 A. in width, can occasionally be seen alongside filaments in our preparations (Pras et al., 1968). Shirahama and Cohen (1967b) have attempted to define further subunits of the protofibril, but since these studies are still controversial, they will not be described in detail here. Thus, presently we use the following terminology: the amyloid fibril seen in thin tissue sections consists of a number of filaments, aggregated side by side, and often forms thick bundles. The filaments measure 50-75 A. in width and, in tissue sections as well as negatively stained preparations, are often found in pairs. The protofibril, which has only been seen in negatively stained material so far, measures 25-35A. in width. Its precise relationship to other protofibrils within the amyloid filament has not been established beyond reasonable doubt (Boere et al., 1965; Emesen et al., 1966). Morphological studies on the P component of amyloid (for a discussion, see Section IV,D) have been limited to negatively stained material (Fig. 11). No structure resembling either the doughnut or rod configuration of this component has ever been described in tissue sections derived from patients or experimental animals. This could be attributable to the fact that the P component is estimated to constitute only 5%of the total amyloid mass and could conceivably have been overlooked. However, in a review of our own material comprising several hundred micrographs, no structure reserribling the P component could be found. It is conceivable that the P component is extracted during the embedding procedure-a possibility made more likely by the claim that this small fraction of amyloid represents an @-globulin,a soluble serum protein. If this is true, it ought to be possible to perform EM studies on the component isolated from normal blood. The appearance of the negatively stained P component extracted from amyloid-laden tissues was first described by Bladen et (11. (1966; see, also, Fig. 11) and Benditt and Eriksen (1966) who equated this rod structure with the main

268

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

CURRENT CONCEPTS OF AMYLOID

269

FIG. 11. Negatively stained periodic rods representing the P component of amyloid. Arrow points to “doughnuts,” the unit structural components of the rod. (From Glenner and Bladen, 1966; electron micrograph obtained through courtesy of Dr. H. A. Bladen.)

amyloid fibril. Subsequently the structural appearance of the negatively stained P component was confirmed by Cathcart et al. (1967a,c) and compared and contrasted with the true amyloid fibril (Shirahama and Cohen, 1967b). Since the P component, at least as defined by biochemical and immunological techniques, is found in the serum of normal subjects, it may not play a significant role in the pathogenesis of amyloidosis but may merely represent another serum component, nonspecifically adsorbed to this peculiar extracellular protein. Another way of studying the physical structure of the amyloid fibril is by X-ray diffraction, and these studies on various types of human amyloid as well as on mouse and duck amyloid have been carried out by three different groups with virtually identical results (Eanes and Glenner, 1968; Bonard et al., 1969; Shmueli et al., 1969). All X-ray diffraction patterns have shown a sharp meridional arc at about 4.7 A. and a more diffuse equatorial arc at 9.8A. These characteristic reflections have been interpreted as indicative of a cross-p or a pleated-sheet configuration by two groups (Eanes and Glenner, 1968; Bonard et al., 1969) and as an ordered polymer of globular protein subunits by another (Shmueli et aZ., 1969). All three consider that this rather unusual appearance is characteristic of amyloid in tissue and not the result of denaturation or preparative artifacts. FIG. 10. Isolated purified amyloid negatively stained with phosphotungstic acid. Arrows point to single filaments, Note that most filaments are paired and end together (half-circles). Some thicker bundles ( B ) in which filaments are twisting around each other can also be seen. Magnification: X96,OOO.

270

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

C. PHYSICOCHEMICAL AND BIOCHEMICAL STUDIESOF AMYLOID F m w AND THEIRSU~UNITS Previous studies measuring the binding of certain dyes to amyloid (Carnes and Forbes, 1956; Goldberg and Deane, 1960) suggested that the isolectric point of amyloid was between 4.5 and 5. Free electrophoresis of a single sample of the water-soluble protein in unbuffered Tris at pH 10 showed the material to be negatively charged, but it was not possible to obtain an exact electrophoretic mobility in the absence of salt (Pras et al., 1968). Ultracentrifugation studies of intact fibrils have also been difficult to interpret since they had to be carried out in distilled water in the absence of any ions. Most preparations, when examined shortly after extraction, were homogeneous and had a major component with a sedimentation coefficient of about 45 S. However, if extraction proved more difficult or after prolonged standing, larger polymers with sedimentation coefficients of 7 5 s and higher appeared. Within a short time, they begin to sediment rapidly even at low centrifugal forces, and after a while these larger aggregates have generally shown a tendency to precipitate from solution spontaneously. These precipitates could generally by redispersed to the 45 S component by vigorous homogenization. Unfortunately, it has not yet been possible to correlate the ultrastructure of the fibrils or their subunits with the sedimentation rates obtained in the ultracentrifuge since EM studies are always carried out on material which has been precipitated either by salt or dehydration. As mentioned below, a few of the preparations contained a more slowly sedimenting component with sedimentation coefficients ranging from 8 to 15s the nature of which remains as yet unknown (Pras et al., 1968, 1969). From the start, it seemed likely that the 45s component extracted in distilled water was composed of smaller subunits and, consequently, a number of attempts have been made to define chemically homogeneous subunits and to isolate the constituent polypeptide chains. With some of the initial preparations studied by us, urea, guanidine, and reducing agents aIone and in combination brought the material out of solution and could not be used as such for chemical studies. Because of these difficulties we employed 0.1 M NaOH to produce a smaller subunit which was soluble in 0.15 M NaCl and which we referred to as degraded amyloid (DAM) (Pras et al., 1969). This material has a sedimentation coefficient of 1 to 2 s and an estimated molecular weight of about 30,000 to 40,000 daltons based on peptide maps and the content of arginine and lysine. This subunit, produced by as yet undefined mechanisms, still contained some small fragments of fibrils on E M (Fig, 12). Although this material did not lend itself to detailed chemical analysis,

CURRENT CONCEPTS OF AMYLOID

271

it proved to be very useful in immunological and some superficial chemical studies to be described below. Use of alkali to degrade amyloid is not new. Prior studies by Hass (Hass and Schulz, 1940) and by Cohen’s group ( Newcombe and Cohen, 1964) had already investigated the effect of alkali; Hass had probably extracted amyloid from tissues as DAM using sodium hydroxide, whereas Newcombe (Newcombe and Cohen, 1964) and Shirahama (Shirahama and Cohen, 1967a) probably stopped short of the pH required to dissociate the molecule into a subunit when they studied the effect of pH on the solubility and extractability of amyloid. Several other attempts at dissociation have been reported recently. Miller et al. ( 1968) employed cyanogen bromide designed to cleave at methionine residues, urea, a reducing agent, and acid. All of these procedures yielded smaller fragments. Degradation was almost complete with CNBr, whereas urea, alkali, and acid yielded 50% or less of a smaller subunit. Although all the agents appeared to produce subunits with similar behavior on Sephadex filtration and polyacrylamide electrophoresis, these subunits were not sufficiently well characterized to permit any estimates as to their size or nature. An unusual feature, which has severely hampered the biochemical analysis of amyloid and which may be of some biological significance in terms of long-term persistence of amyloid in tissues, is the resistance of the native undenatured fibrils to many proteolytic enzymes used in the biochemical analysis of other proteins (Sorenson and Binington, 1964; Emeson et al., 1966; Ruinen et al., 1967; Pras et al., 1969; Kim et al., 1969). Incubation of native undenatured amyloid with papain, pepsin, and trypsin failed to produce identifiable subunits and left much of the residual fibrillar protein in tact, Only pronase seemed to be effective in degrading the fibrils more or less completely. Following denaturation, the fibrils become more susceptible to several of these proteolytic agents. Probably the single most important observation dealing with the nature of amyloid came through the efforts of Glenner and his collaborators (1970b, 1971a,b,c; Harada et al., 1971). These investigators treated several purified preparations of amyloid fibrils with 6 M guanidine and a reducing agent-a procedure that destroyed the fibrillar appearance of amyloid, the characteristic appearance on X-ray diffraction, and eliminated the capacity to bind Congo red. Following filtration on Sephadex and Sepharose, they isolated a smaller subunit from these fibrils ranging in molecular weight from 5000 to 31,000 daltons which, in each instance, appeared relatively homogeneous on SDSZdisc electro-

* Sodium

dodecyl sulfate.

272

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

CURRENT CONCEPTS OF AMYLOID

273

phoresis. Chemical studies of several of these preparations by Glenner et al. (1970b, 1971a,b) showed these small subunits to resemble immunoglobulin chains in having either pyroglutamic acid (PCA), Asp, or Glu as the N-terminal residue, a finding reported also for the intact fibril by Skinner and Cohen (1971). In a series of short notes which promise to be followed by additional full length reports in the near future, Glenner et al. (1970b, 1971a,b) subjected two of these fragments having unblocked N termini to analysis in the automatic sequencer and demonstrated in each of them a striking homology to the first 30 residues of the variable region of the k light chain. In one instance, the major subunit appeared similar to the whole light chain present in the patient’s urine by peptide mapping (Glenner, 1972). This finding has led to the exciting conclusion that the major protein component of amyloid is a portion of an immunoglobulin polypeptide chain. Recently, as might be expected, these workers have also demonstrated structural identity of the major amyloid protein and the light chain derived from the same subject (Glenner et al., 1972). At the present time the precise structure of the amyloid proteins remains unknown since the molecular weight of a light chain is approximately 22,500 daltons, whereas the major amyloid protein ranges from 5000 to 30,000 daltons. It seems possible that the amyloid protein may represent a part of a light chain, primarily the variable region, either with an internal deletion or with degradation and consequent loss of varying amounts of the constant and possibly the variable region. The ability to make amyloid fibrils, consisting primarily of the variable region from selected Bence-Jones proteins, with a variety of proteolytic enzymes strongly favors the second alternative. It is worth noting that structural studies of this type have so far been limited to k-like proteins since these have unblocked N termini, but that h-type proteins with blocked N termini are far more commonly found in amyloid. To date the published information is not sufficient to decide whether amyloid consists of a series of identical subunits or if the molecule is composed of more than one type of polypeptide chain, nor is there any estimate of the

FIG. 12. Fragments seen in denatured amyloid following treatment of isolated amyloid fibrils with 0.1 M NaOH for 1 hour. Fragments range from 200 to 1500A. in length. In specimens digested for 3 hours, there is an increase in smaller-sized fragments. Magnification: X 75,000. ( From Franklin and Zucker-Franklin, 1972.) FIG. 13. Precipitate obtained by pepsin digestion of a A-type Bence Jones protein, negatively stained with 1%phosphotungstic acid at pH 5.4. The fibrils in this case are indistinguishable from those of negatively stained isolated amyloid. Magnification: x75,OOO.

274

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

number of subunits of the major 4 5 s component. The chemical homogeneity of amyloid and the ability to make amyloid fibrils form purified Bence Jones proteins (see below) favor the presence of only a single type of subunit. In a parallel, but so far less precise approach to the study of amyloid subunits, we have noted the presence of about 6 to 8 characteristic common peptides in the peptide maps of similarly prepared amyloid subunits (Levin et al., 1972a). Most of these are present also in k and A Bence Jones proteins and appear to be derived from the variable region of the molecule. The constant occurrence of these peptides in all preparations studied and also in “synthetic amyloid prepared by proteolysis of Bence Jones proteins (see below) suggests that certain structural features are required for all amyloid proteins but that similar sequences may be found also in proteins from patients without amyloidosis. The conclusions obtained from the studies of amyloid fibrils regarding the role of light chains in amyloid have derived significant support from a parallel line of investigations which has demonstrated that treatment of many Bence Jones proteins, regardless of their origin, with pepsin or trypsin can give rise to fibrils which show green birefringence when stained with Congo red and which have the characteristic appearance of amyloid in the EM and on X-ray diffraction (Glenner et al., 1971c; Linke et al., 1972). Detailed studies of one such prepar.‘1t’ion revealed complete homology of the 4600 daltons molecular weight fragment with a Bence Jones protein from residue 3 to 22. It is of interest that, in this protein, pepsin appears to have removed the first two residues thus making the protein susceptible to analysis in the sequencer. Peptide maps revealed, as expected, absence of the constant region peptides ( Glenner et al., 1 9 7 1 ~ )Studies . in our laboratory (Linke et al., 1972) of 1 8 X and 13k Bence Jones proteins have yielded a fragment from four of these proteins having a molecular weight of about 8 to 12,000 daltons which no longer has any of the antigenic properties of the native Bence Jones proteins, yet which on acidification can give rise to fibrils resembling amyloid on EM (Fig. 13). Preliminary chemical studies suggest that this fragment lacks most of the common region peptides and yet contains the same peptides that are characteristic of the major component of amyloid isolated from tissues. Several clues that the situation may be somewhat more complex and that exceptions to these findings may occur have recently been noted by three groups of investigators. Glenner has described a single subunit having a molecular weight of 5000 daltons which lacked cysteine residues (Glenner et al., 1970a). Further analysis of this protein revealed complete lack of homology to any known immunoglobulin polypeptide

275

CURRENT CONCEPTS OF AMYLOID

chain (Ein et al., 1972)-a finding which clearly indicates that other proteins may also be involved in the formation of amyloid. In a series of studies dealing with the major constituents of amyloid, Benditt et aZ. (1962, 1968, 1970; Benditt and Eriksen, 1964, 1966) have presented evidence for the existence of several low molecular weight subunits extractable from amyloid tissues with acid and urea. In a recent study, Benditt and Eriksen (1971) have partly characterized the A protein obtained from amyloid-containing tissues of patients with secondary amyloidosis and from a patient with FMF. The A protein extracted from amyloid tissues has the same electrophoretic mobility as the major protein found in purified amyloid fibrils. It has a molecular weight of about 7000, and amino acid analyses of three of these were quite similar. The absence of cysteine and threonine and the low values of proline have led these observers to conclude that the A protein is not related to the variable portion of an ordinary immunoglobulin. This initial conclusion was strengthened by a subsequent report from the same group which presented the amino acid sequence of the first 24 residues of the A protein isolated from the liver of a patient with secondary amyloidosis accompanying tuberculosis and clearly demonstrated the lack of homology of this fragment to any known immunoglobulin (Benditt et al., 1971). In an independent study, Pras and Reshef (1971) have succeeded in extracting a homogeneous protein with a molecular weight of about 8000 with 0.02 M HC1 from purified amyloid fibrils isolated from patients with FMF, rheumatoid arthritis, bronchiectasis, Hodgkin’s disease, and tuberculosis. This acid soluble component constituted up to 608 of the weight of the purified fibrils and could be reconstituted into a fibrillar shape. Amino acid analyses of three of these preparations showed them to lack cysteine. They resembled the unusual protein of Glenner and were virtually identical to the A protein of Benditt and Eriksen. Preliminary sequence studies (Franklin et al., 1972) show no homology to any immunoglobulins. The complete amino acid sequence of one of these proteins from a patient with FMF (Levin et al., 197213) is as follows: 10

5

1

15

Arg-Ser-Phe-Phe-Ser-Phe-Leu-Gly-Glu-Ala-Phe-Asp-Gly-Ala-Arg-~p-Met-Trp-Arg 20

25

30

35

Ala-Tyr-Ser-Asp-Met-Arg-Glu-Ala-Asn-Tyr-Ile-Gly-Ser-Asp-Lys-Tyr-Phe-H~-Ala 45

40

*

50

55

Arg-Gly-Asn-Tyr-Asp-Ala-Ala-Lys-Arg-Gly-Pro-Gly-Gly-Ala-Arg-Ala-Ala-Glu-Val 60

65

70

76

Ile-Ser-Asn-Ala-Arg-Glu-Asri-Ile-Gln-Arg-Leu-Thr-Gly-Arg/Gly-Ala-Glu-Asp-Ser

* Uncertain.

276

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

Partial amino acid sequences for four other proteins studied in our laboratory and one reported by Benditt et al. (1971) were identical in the regions examined (1152 residues). To date, we have failed to extract this component from the amyloid of two patients with macroglobulinemia and multiple myeloma who had immunoglobulin-like subunits. The ultrastructure of the acid soluble fraction (Fig. 13A) bears no resemblance to the structure of the native amyloid fibrils. It is vermiforni in appearance, varies in thickness between 350-450 A depending on the method used for negative staining, and does not seem to have any resolvable substructure.

FIG. 13A. Electron micrograph of acid soluble fraction (ASF) prepared from isolated amyloid fibrils, negatively stained with uranyl acetate. Magnification: X 96,000.

CURRENT CONCEPTS OF AMYLOID

277

D. IMMUNOLOGICAL STUDIES Attempts to characterize the major constituents of amyloid immunologically have proceeded in two major directions. On the one hand, there has been a search for a variety of serum proteins within amyloidcontaining tissues and, later, when they became available, in preparations of purified fibrils. On the other hand, numerous attempts have been made to induce antibodies specifically reactive with amyloid by immunization with extracts of amyloid tissues. More recently purified amyloid fibrils, degraded amyloid fibrils, and the protein subunit have been used for this purpose. When used as antigens these preparations seem to have given rise to two distinct types of antibodies. One of these, that produced by immunization with the amorphous protein subunit, appears to react with idiotypic antigens unique for each amyloid and with k or h Bence-Jones proteins. The other, produced to NaOHor guanidine-degraded partially fibrillar amyloid, reacts only with amyloid (and with no other serum or tissue protein) and may detect some conformational antigenic site common to all amyloid fibrils (Franklin and Zucker-Franklin, 1972). The first type of antibody, that directed against the light-chain determinants, has proved of great value in further elucidating the nature of amyloid, whereas the antibodies to partially fibrillar amyloid have proved to be useful as diagnostic reagents for detecting amyloid in tissues ( Zucker-Franklin and Franklin, 1970). Over the years immunological mechanisms have often been implicated in the pathogenesis of amyloid. These ideas are based in part on the frequent association of the disease with multiple myeloma (Magnus Levi, 1956; Osserman et al., 1964; Pick and Osserman, 1968) and, in part, on a number of experimental as well as naturally occurring instances of amyloidosis associated with the administration of large amounts of foreign protein which, in turn, result in a state of hyperimmunization ( Teilum, 1968; Muckle, 1968a,b,c). Because of these theoretical considerations and the ready availability of antisera to y-globulins, the major effort has usually been directed toward the demonstration of immunoglobulins and complement components in amyloid-containing tissues and in association with amyloid fibrils. In spite of much contradictory data and difficulties in interpretation of some experimental observations, most of these studies suggested that y-globulin is usually present in amyloid deposits, possibly in amounts greater than in control tissues (Vazquez and Dixon, 1956; Mellors and Ortega, 1956; Schultz et al., 1966, 1968) and that complement components are also frequently found in this location. In the light of current knowledge dealing with the relation of amy-

278

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

loid to y-globulins and the antigenic properties of amyloid, many of these studies are difficult to evaluate critically. On the one hand, as mentioned above, the basic aniyloid protein frequently appears to be similar to a light-chain fragment. On the other hand, in spite of these structural features, few if any antisera to y-globulin or light chains react with purified amyloid fibrils (Franklin and Pras, 1969) or the basic protein subunit (Isersky et al., 1972; Franklin and Levin, 1972). In fact, synthetic aniyloid fibrils prepared from Bence-Jones proteins often fail to react with antisera to Bence-Jones proteins, even when these are prepared to the same protein used to make amyloid (Linke and Franklin, 1972). Thus, because of these unusual properties, it is not surprising that much contradictory immunological data has accumulated. Probably some of the differences are related in part to the antisera employed, but others probably reflect the interaction of substances other than amyloid with antisera to various y-globulin fragments. y-Globulin was first detected with fluorescent antisera to amyloid by Mellors and Ortega (1956) and by Vazquez and Dixon (1956). The latter observers attempted to quantitate the amount of y-globulin by measuring the uptake of radioactively labeled antisera to 7-globulin and estimated that % 12- X as much antiserum was absorbed by amyloid tissues as by normal control tissue. Vogt and Kochem (1960) and 2 years later, Lachmann et al. (1962) demonstrated the presence of complement (PIC in particular) in amyloid deposits from 3 patients. Then, Schultz et a2. (1964, 1966) using mixed agglutination and antiglobulin consumption tests detected y-globulin in some but not all amyloidcontaining tissues and in partially purified amyloid fibrils and concluded that there were more heavy than light chains in amyloid tissue. In spite of these positive findings, however, some questions were raised by several observers about the amounts of y-globulin and its specificity in amyloid deposits. In 1958, Calkins et al. using the antiglobulin consumption test, concluded that amyloid from several individuals contained less than 1% y-globulin and that, consequently, amyloid cannot represent simply an antigen-antibody complex. Similarly, Benditt et al. (1962) failed to demonstrate a reaction between a urea extract of human amyloid and antisera to y-globulin; this treatment failed to alter the reactivity of y-globulin with antisera. Additional data pointing away from a specific association of y-globulin and amyloid were subsequently obtained by studies with isolated amyloid fibrils. Paul and Cohen ( 1963) , using ferritin-conjugated antisera to 7-globulin to detect y-globulin in purified preparations of amyloid fibrils, found no adherence of ferritin to the amyloid fibrils and concluded that y-globulin was not present as an integral part of the fibrils. These studies were

CURRENT CONCEPTS OF AMYLOID

279

extended by Cathcart and Cohen (1966) who demonstrated that purified amyloid fibrils, solubilized at p H 9.5, failed to react with antisera to yG, M, and A and also with antisera against k and h chains and Fc fragments. Consistent with this finding was the observation that antisera against some of these amyloid fibril preparations failed to react with YG-globulin and k and h chains and that absorption of antisera to y-globulins with amyloid fibril preparations failed to remove significant antibody activity. This series of studies, therefore, concluded that little, if any, y-globulin was associated with the major constituent of amyloid, the amyloid fibril. Similar conclusions have been arrived at by Cagli et al. (1962) who failed to demonstrate antibodies to 7-globulin in a rabbit immunized with an extract from amyloid and by Muckle (1964) who did not detect antibodies to y-globulin in antisera to extracts from 1 of 2 subjects. In a subsequent study, Muckle (1968a,b,c) interpreted the finding of immunoglobulins and other serum proteins in amyloid as the result of nonspecific binding. However, due to their varying nature and amount, he concluded that they are not essential to the fundamental nature of amyloid. More recently, using highly purified amyloid and smaller fibrillar subunits, we have been unable to demonstrate a reaction with antisera to each of the four major classes of immunoglobulins and their light chains ( Franklin, 1970). A clearer insight into some of these apparently contradictory observations has resulted from the chemical studies cited above and from immunological studies with antisera prepared against either the amorphous major protein subunit or against partially degraded amyloid which still retains some of the characteristic structural features of amyloid. Two types of results, depending on the antigens used for immunization have been obtained, Immunological studies with antisera produced to the basic protein subunit known to be related to the variable region of k or h BenceJones proteins have, as expected, demonstrated a strong, presumably idiotypic reaction with the protein used for immunization and varying degrees of cross-reaction with other amyloid proteins and Bence-Jones protein belonging to the same light-chain class ( k or h ) and variable region subclass. Since the antigens used for immunization were pure and in many instances chemically well-defined, these studies offer strong support for the concept that some amyloid is closely related to immunoglobulin light chains (Glenner et al., 1971a,b,c). It seems likely that these antisera will prove of great value in studying the primary structural relationships between different amyloid fractions. In contrast, and in direct support of prior studies indicating that amyloid failed to react with antisera to immunoglobulins, is the observa-

280

EDWARD C. FRANKLIN AND W R O T H E A ZUCKER-FRANKLIN

tion that the basic subunit, and even the synthetic amyloid fibrils prepared from Bence Jones proteins by proteolysis, generally fail to react with most antisera to light chains and 7-globulins. This finding is consistent with the known poor immunogenicity of the variable region of light chains and suggests that the light-chain determinants associated with amyloid may be hidden in the molecule and not readily reactive with antisera. On the one hand, it seems likely that many of the reported reactions of amyloid tissues with antisera to y-globulin may have reflected the presence of y-globulin contaminants. On the other hand, the possibility that the antisera reacted with immunoglobulin precursors of amyloid deserves serious consideration and may explain the frequently noted positive results. The other type of antibody, which is apparently directed to some structural feature common to all amyloid and, possibly, recognizes some conformational sites characteristic of the fibril, is of interest since it fails to react with any antigenic determinant on light chains or other serum proteins. It has been long recognized that amyloid fibrils are poorly immunogenic (Ram et al., 1968; Franklin and Pras, 1969; Glenner et al., 1969). In our own experience, for example, none of five rabbits immunized with purified fibrils produced antibodies. The reason for the poor immunogenicity of the intact fibrils may be their large size (Ram et al., 1968) and their marked resistance to phagocytosis (ZuckerFranklin, 1970) and proteolysis ( Sorenson and Binington, 1964; Kim et al., 1969; Pras et al., 1969). As a result, it may be difficult for the fibrils to be taken up by cells involved in processing antigens and to be degraded to a form that can induce an active immune response manifested by the formation of antibodies. Recently, this difficulty has been overcome by using various partially or completely degraded forms of amyloid for immunization since these may be more readily processed by cells involved in the immune response. Initially, we (Franklin and Pras, 1969) used alkali-degraded amyloid, a substance which still retains small fragments of fibrils to induce antibodies specifically reactive with amyloid. By using this substance as the immunizing agent, it was possible to induce antibodies that reacted with amyloid and failed to react with normal tissues or serum proteins. Although most of these antisera reacted only with DAM and amyloid and not with normal serum or tissue extracts by precipitin analyses and even by complement fixation, they were always absorbed with a marked excess of NHS and tissue extracts so as to avoid any possible confusion with cxtraneous antigens, Since this initial report, three other groups have succeeded in producing similar antisera using various partially degraded types of amyloid. Glenner et al. (1969) used urea- or guanidine-treated amyloid

CURRENT CONCEPTS OF AMYLOID

281

fibrils as the immunogens. Shapira et al. (1970) immunized with reduced and alkylated samples, whereas Cathcart et al. (1970a) used both NaOH and 5 M guanidine to produce potent immunogens. Treatment with sodium hydroxide, guanidine, and urea is relatively mild and alters the molecule sufficiently to make it immunogenic while retaining at least some of the fibrillar structure in the EM (Glenner et al., 1969). In support of the contention that these antisera pick up the tertiary structure of the fibril is our recent finding that reduction and alkylation in 6 A4 guanidine abolishes the reactivity of DAM with these antisera, whereas similar treatment of Bence Jones proteins does not abolish their reaction with antisera to Bence Jones proteins (Franklin and ZuckerFranklin, 1972) . An interesting finding, which is not entirely consistent with this interpretation, is the failure of synthetic amyloid fibrils prepared from Bence Jones proteins to react with this antibody. It seems possible, however, that these synthetic fibrils may not have exactly the same conformation as naturally occurring amyloid fibrils.

FIG. 14. ( A ) Ouchterlony double-diffusion study of antiserum to denatured amyloid (DAM) (center well) reacting with ( a ) DAM; ( b ) normal serum, ( c ) spleen extract; ( d ) yG-globulin; ( e ) Fab fragment; and ( f ) Fc fragment. ( B ) Immunoelectrophoresis of DAM (top) and normal serum (bottom). Antiserum to DAM is in the trough. (From Franklin and Pras, 1969.)

282

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

Antisera reactive specifically with amyloid can be tested in three ways. ( 1 ) By precipitation techniques either in agar or in the fluid phase using DAM as antigen-Fig. 14A and B shows an Ouchterlony plate and immunoelectrophoretic pattern. A single broad arc moving to the anode is clearly visible. Figure 15A shows a series of quantitative precipitin curves with one antiserum and a number of DAM preparations. Intact amyloid could not be studied in the same fashion because it is insoluble in physiologic saline and too large to diffuse through the agar. ( 2 ) By complement fixation-since the antigen need not be soluble to be tested by complement fixation, DAM and amyloid proved suitable for such analyses. Figure 15B shows the reaction with a number of DAM preparations and the native amyloid fibril corresponding to the immunogen. ( 3 ) By absorption studies-amyloid fibrils proved ideal for 140

-

E100

s

g 60

s

20 40

80

120

160

200

40

80

120 160 c/g ANTIGEN

I 200

FIG. 15. ( a ) Precipitin curve with an antiserum to denatured amyloid ( D A M ) reacting with eight DAM preparations. (0-0) DAM used for immunization. ( b ) Complement fixation with anti-DAM and five DAMS including the homologous and the homologous amyloid ( 0-0). Significant cross-reaction is DAM (0-0) apparent in all. (From Franklin and Pras, 1969.)

283

CURRENT CONCEPTS OF AMYLOID

such studies since, being insoluble in salt, they had all the properties of an insoluble immunoabsorbent. Thus addition of amyloid to antisera followed by an assay of the residual antibody left after absorption provided an excellent estimate of the cross-reactivity. Table I shows the results of such an absorption study demonstrating the generally marked cross-reaction of antisera prepared against DAM with amyloid. Nevertheless, in most instances it appeared that additional determinants were exposed in DAM which were not readily detectable in amyloid. Detailed studies using these antisera with a number of different amyloid preparations will be cited later. Suffice it to say here that the antisera to DAM generally reacted best with the antigen used for immunization but that they cross-reacted with most other DAMS (the degree of cross-reaction varied from significant to minimal with different antisera) and, in general, they reacted well also with the native amyloid fibrils. Since it usually required larger amounts of amyloid fibrils to reach equivalence and since frequently it was not possible to absorb the antisera fully with intact amyloid fibrils, it seems likely that some of the conformational determinants are hidden in the molecule or that additional determinants are exposed during the denaturation procedure necessary to convert the fibrils into a good immunogen. In the one instance where the fibrils could absorb all of the antibody, 5 times more amyloid was required to absorb the antiserum fully. Because many of these antisera reacted well with all amyloids tested, they were considered useful as general reagents reactive with amyloid in tissue. A pool of several antisera was conjugated with fluorescein isothiocyanate and shown to react specifically with amyloid-containing tissues ( Zucker-Franklin and Franklin, 1970). In fact, preliminary CROSS-RE.4CTION

TABLE I BETWEEN

AMYLOIDAND DENATURED AMYLOIDQ

~

Amyloid

A B C 1)

Residual ppt. after absorption with AM (%)b 25 0 30

80

% fixedc 25 100 100 AC

C fixation, ratio AM/DAM at peak 10 5 6

Adapted from Table I of Franklin and Pras, 1969. (ppt.) left (% of total) after absorption of antisera prepared against degraded amyloid (DAM) with amyloid (AM). CHWfixed at peak by AM c yo Fixed = x 100 CHso fixed at peak by DAM a

* Precipitate

284

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

studies suggest that these antisera may be more sensitive than Congo red and that they may detect amyloid present in cells. In view of the poor immunogenicity of amyloid fibrils and the frequent use of impure preparations for immunization, many attempts to induce antibodies unique to the fibrils have resulted only in antibodies to other protein contaminants. Among these are certain serum constituents, such as y-globulins, fibrinogen (Horowitz et al., 1965), lipoproteins, and certain p- and al-,and a,-globulins (Cathcart et al., 1967a; Kim et al., 1967; Milgrom et al., 1966; Muckle, 1964) some of which appear to be selectively concentrated in amyloid. As mentioned earlier, of particular interest and worthy of some comment appears to be an al-globulin, termed the P component, which was most carefully characterized by Glenner ) ~ Ram et al. ( 1968). This and Bladen ( 1966), Cathcart et al. ( 1 9 6 7 ~and P component may well be identical to an a,-globulin detected by Dixon (1965) and to one of the two components noted by Milgrom et al. (1966). Although it is carried along with the fibrils in certain preparative procedures, the P component is not an integral part of the major fibrillar constituent of amyloid. This material, which has been shown to be a normal serum constituent by Cathcart et al. (1967c), appears to be identical to the rodlike structure consisting of a stacked array of pentagonal doughnut-like discs and can be seen either lengthwise or in cross section in the EM (Fig. 11) (Bladen et al., 1966). This P component has been found, when looked for, in all partially purified amyloid fibril preparations, but can be removed by further purification and is absent from similar extracts prepared from normal tissues. However, in all instances, it made up only a small fraction of the amyloid preparation solubilized by treatment with glycine ( Glenner et al., 1968). Although the precise chemical nature of the P component remains to be elucidated, preliminary studies by density gradient ultracentrifugation have yielded a sedimentation coefficient of 10.2 S and an approximate molecular weight of 200,000 (Cathcart and Cohen, 1968; Cathcart et al., 1 9 6 7 ~).According to Glenner and Bladen (1968), the Type I1 fibers, which correspond to the rods, are larger and have a molecular weight of 800,000 and an elution volume on Sephadex G-200 consistent with this estimate. On acrylamide gel electrophoresis, the P component migrates as a single band (Cathcart et al., 1967c), but it can be dissociated into at least five bands by guanidine (Glenner and Bladen, 1966). The concentration of the P component in serum does not seem to be increased in patients with amyloidosis and other pathological states ( Cathcart et al., 1967b), although, according to Muckle ( 1969), the possibility exists that levels may be increased in the actively progressive form of the disease,

CURRENT CONCEPTS OF AMYLOID

285

It is of interest that when amyloid tissues were stained with antisera to P component and fibrils, fluorescence was located on the same region for both, thus indicating the close association of these two constituents of amyloid (Cathcart et al., 1970a).

E. COMPARATIVE STUDIESOF DIFFERENT AMYLOID PREPARATIONS In view of the different clinical and pathological types of amyloid, and some of the differences already alluded to, it was of interest to compare different amyloids by immunological, physical, and chemical methods in the hope of correlating some of these properties with the clinical classification. While differences have definitely been documented, it is unfortunate that most of these studies are as yet too limited to permit any definite conclusions as to chemical differences between amyloid belonging to the currently recognized clinical classes. However, preliminary chemical studies done in several laboratories suggest strongly that the major constituents of amyloid derived from patients with secondary amyloidosis and amyloidosis associated with FMF may not be identical to those derived from primary or myeloma-associated amyloid (Benditt et al., 1971; Benditt and Eriksen, 1971; Pras and Reshef, 1971; Franklin et al., 1972). Comparative studies have been attempted on* the ultrastructural, biochemical, and immunological level. In the EM, with some of the exceptions noted above, most of the amyloid preparations appear identical. Similarly, X-ray diffraction studies of a limited number of samples have failed to detect differences. On the other hand, biochemical and immunological analyses have yielded clear-cut evidence that, on this level, variations among different amyloid preparations may be recognized. Table I1 lists the sedimentation coefficients and Congo red binding capacities of nine amyloid preparations examined by us in detail. Although initially we felt that some of the quantitative differences in Congo red binding reflected the presence of impurities, it is now our opinion that these differences may in reality reflect differences among various amyloids. Most of the variations in sedimentation coefficients probably represent the existence of different sized polymers, which could usually be dissociated to a 45s subunit by repeated homogenization. However, two of the preparations had 8 and 13 S peaks which may represent a different type of fibril. In one instance, the EM appearance seemed to correlate with the physical properties (see Fig. 9). The results of carbohydrate analyses are even more difficult to interpret in view of lack of data concerning the degree of purity of many

286

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

TABLE I1 PHYSICAL AND DYE-BINDING PROPERTIES OF AMYLOIDS ISOLATED FROM NINESUBJECTS' Congo red bound (mg)

Sedimentation coeff, (major component)c

No.

TYP&

protein (mg)

AM

DAM

1 2 3 5 6 8 9 10 15

P S P MM MM MM S S S

0.32 0.26 0.21 0.34 0.22 0.33 0.11 0.13 0.24

45 40 7.6 74d 7.9 41 150d 8 . 6 ; 95d 44

1.3-2.8 1.7 2.1 1.1 2.5 2.1 2.4 2.7 2.4

Adapted from Pras et al., 1969. P-primary ; S--secondary; MM-myeloma. AM-amyloid; DAM-degraded amyloid. P o l y m e r 4 0 S component.

of the preparations studied. Although the nature of the carbohydrate group remains uncertain, it seems likely that the carbohydrate moiety may be trapped by the protein rather than being an integral part of it (Pras et al., 1971); nevertheless it appears to be responsible for the binding of the various metachromatic dyes (Pras and Schubert, 1968). Recent evidence suggests that heparitin sulfate rather than chondroitin sulfate is responsible for these properties (Bitter and Muir, 1966; Mowry and Scott, 1967; Dalferes, 1968; Muir and Cohen, 1968). The technique of peptide mapping too has demonstrated similarities and differences among different preparations. Figure 16 shows maps of several amyloid subunit preparations. In general, all of these were strikingly similar, and contained certain common peptides. Nevertheless they differed from one another in certain peptides, thus indicating clear-cut chemical differences among them ( Levin et al., 1972a). Similar results have recently also been reported by Glenner's group (Glenner et al., 1970a,b). They demonstrated, in addition, the virtual identity of amyloid extracted from two organs of the same individual and the amyloid and immunoglobulin light chain prepared from the same patient. More definitive evidence not only of microheterogeneity but also for the possible existence of at least two chemically distinct types of amyloid have been obtained from end-group analyses, amino acid analy-

CURRENT CONCEPTS OF AMYLOID

287

FIG. 16. Peptide maps of ( A ) amyloid No. 1 and ( B ) its denatured amyloid derivative, ( C ) amyloid No. 6, and ( D ) low molecular weight subunit obtained from a secondary amyloid. (From Pras et al., 1969.)

ses, and partial amino acid sequence studies of a number of amyloid fibril subunits. On one hand, many of the amyloid subunits resembled light chains, in particular the variable region. Thus NH, terminal analyses of eight subunit preparations by Glenner et al. (197Oa) and of twelve amyloid fibrils by Skinner and Cohen (1971) showed the majority to be unreactive, presumably due to the presence of PCA, whereas a small number had glutamic or aspartic acid as the NH, terminal residues. Amino acid analyses also showed striking differences among different subunit preparations (Table 111) (Glenner et al., 1970a). The reason for the existence of gross similarities as well as some structural differences among certain amyloid preparations has been provided by the demonstration by Glenner and his collaborators that the major protein constituent of many amyloid fibrils consists of a fragment of light chains of immunoglobulins, in particular the variable region. Detailed chemical studies of many homogeneous light chains have clearly demon-

288

EDWARD C. FRANKLW AND DOROTHEA ZUCI(ER-FRANKLIN

TABLE I11 AMINOACID ANALYSIS, N-TERMINAL GROUP,AND MOLECULAR PROTEINS‘ WEIQHTOF MAJORAMYLOID Amyloid protein: Amino acid*

IIIa

IIIb

IV

VI

VIIa

VIIb

IX

XIV

Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half-Cystine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Lysine Histidine Arginine Tryptophan

5.51 5.85 8.18 7.77 5.70 6.09 6.71 1.71 5.44 0.23 2.18 6.12 2.31 1.60 3.93 0.85 1.81 1.00

5.98 5.24 7.87 6.85 6.11 6.33 6.55 1.29 4.86 0.33 2.79 6.27 2.09 1.81 3.39 0.60 2.05 1.00

6.71 0 3.07 2.18 0 4.47 7.76 0 0 1.89 0.82 1.09 3.58 4.96 1.09 0.89 4.80 1.00

4.30 5.30 9.51 7.60 4.85 8.36 6.46 1.80 4.50 0 3.16 6.26 3.40 1.85 4.65 1.30 3.20 1.00

9.20 6.82 9.23 9.99 4.98 6.17 5.62 1.81 5.28 0.75 4.06 7.14 3.50 4.09 4.95 1.15 3.50 1.00

8.73 7.10 8.72 9.35 6.48 7.47 6.12 1.33 5.50 0.63 5.29 6.65 4.08 5.36 4.97 1.03 3.65 1.00

7.10 6.18 7.98 7.81 5.60 7.02 4.34 1.50 3.22 0.80 3.47 4.34 3.76 3.30 3.18 0 2.38 1.00

6.91 6.15 10.95 9.02 7.32 11.13 8.26 1.64 6.20 0 4.68 7.79 3.68 3.22 2.29 1.23 1.11 1.00

N-terminusc: U u u u Asx Asx Asx u Molecular weight: 27,600 13,700 5000 15,400 31,200 18,300 7500 14,600 a

From Glenner et al., 1970a.

c

U-unreactive to fluorodinitrobenzene.

* Expressed m micromoles per micromole of tryptophan. strated that all light chains have certain class specific characteristics in the “common” region, but that they differ in amino acid sequence in the variable region. However, even in the variable region there exist common subclass properties ( Milstein and Pink, 1970). While all the immunoglobulin-related amyloid subunits seem to bear a striking chemical resemblance to each other, another unrelated substance has recently been isolated from several amyloid preparations from patients with secondary amyloidosis and amyloidosis with FMF (Benditt and Eriksen, 1971; Benditt et al., 1971; Pras et al., 1971). Amino acid analyses (Table IV) and partial amino acid sequence studies of these proteins indicate that they are identical to each other and unrelated to any known immunoglobulin (Benditt et al., 1971; Franklin et al., 1972). While this material represents the major corn-

289

CURRENT CONCEPTS OF A M Y M I D

TABLE IV AMINOACIDCOMPOSITION OF ACID-SOLUBLE FRACTION FROM Two HUMAN AMYLOID FIBRILS(RESULTEXPRESSED AS No. RESIDUES FOUND)"

Amino acid

FMF

Hodgkins

Lysine Histidine Arginine Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half-Cystine Valine Methionineb Isoleucine Leucine Tyrosine Phenylalanine Tryptophanb

2.0 1.1 7.0 9.1 0.9 4.6 6.3 1 .o 7.6 11.6

2.3 1.8 6.6 7.3 0.2 4.0 5.4 0.8 6.6 10.0

0.9 (2) 2.0 2.2 3.7 4.7

1.1 (2) 2.0 1.2 3.4 5.0

-

+

-

= Mean of three analyses after 21 hr. hydrolysis. Identified from amino acid sequence analysis (p. 275).

b

ponent of these fibrils, it seems likely on the basis of preliminary observations that they may also contain small amounts of light chain related material (Levin et al., 1972). Similarly, the possibility that this protein may exist in trace amounts in primary or myeloma-associated amyloid cannot be excluded at this time.

F. DISCUSSION AND CONCLUSIONS On the basis of the morphological, chemical, and immunological studies, the following tentative conclusions can be reached. It would appear that amyloid in man and all other species studied consists primarily of a fibril with a characteristic ultrastructural appearance. A second component, the precise significance of which remains to be elucidated, and which makes up a small percent of the total, is a serum a,-globulin which assumes a rodlike or a doughnutlike appearance. A number of other proteins have been identified in amyloid deposits but they are not present in purified fibrils and their role in the pathogenesis of amyloid remains dubious. As a result of the studies reported during the past year, there can be little doubt that in many subjects the fibrils consist of subunits rang-

290

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

CURRENT CONCEPTS OF AMYLOU)

291

ing in size from 5000 to 30,000 daltons which seem to represent primarily the variable portion of immunoglobulin light chains. Although their precise structure remains to be elucidated, preliminary sequence studies of several proteins have indicated that, with one possible exception, these proteins are not complete light chains and that they may either have internal deletions of varying lengths or that they may have lost varying amounts of the constant region. Sufficient numbers of these immunoglobulin-related proteins have not been investigated in proper detail to accurately define the number of classes of amyloid, to determine if k and proteins have certain common features necessary for fibril formation, or if other immunoglobulin chains may play a role. In view of some of these structural features and circumstantial evidence that under some conditions plasma cells may be involved in the synthesis of amyloid, it is tempting to speculate that the protein could be a primitive variant of y-globulin, the synthesis of which is induced under certain conditions (see below). A particularly promising lead, since it seems to suggest chemical differences among various types of clinically defined amyloidosis, is the description of an apparently non-immunoglobulin-related substance from patients with secondary and FMF-related amyloidosis. This observation suggests the existence of several different, and possibly unrelated, proteins all of which, under certain conditions, can give rise to the material we define as amyloid and raises the hope that in time it will be possible to develop a rational classification of amyloidosis based on the chemical properties of the amyloid fibrils. The precise origin, mode of synthesis, and assembly of the amyloid related to immunoglobulins remains to be defined. As mentioned before, several mechanisms may be responsible for the synthesis and assembly of such amyloid fibrils. On the one hand, the deposition of large masses of amyloid in the vicinity of plasma cell tumors, coupled with the finding in certain instances of intracellular fibrils, suggest the possibility that amyloid may be locally produced and that fully formed fibrils can be assembled from the precursor in the plasma cell or, alternatively, immediately after secretion as is the case in some instances of macroFIG. 17. Peripheral blood cells incubated with amyloid. No phagocytosis was evident after 30 minutes. Cell in upper left is a monocyte (Mono); cell in lower right a neutrophile. Amyloid ( A m ) . Magnification: X7000. (From Zucker-Franklin, 1970. ) FIG. 18. Peripheral blood neutrophile incubated with antiserum-treated amyloid. Note abundance of phagocytic vacuoles containing amyloid ( Am). The cell is almost completely degranulated. Ultrastructure of amyloid within various phagosomes seems to have undergone alteration. Compare fine structure of inclusions L, L, and 13. Nucleus ( N ) . Magnification: X 14,000. (From Zucker-Franklin, 1970.)

292

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

globulinemia (Buxbaum et al., 1971). The unusual insolubility of the amyloid fibrils may explain the deposition of the fibrils in direct proximity to the cells synthesizing them. On the other hand, the demonstration of light-chain fragments of immunoglobulins as the major component of some types of amyloid raises the possibility that amyloid forms as the result of the interaction of circulating Bence-Jones proteins, immunoglobulins, or possibly antigenantibody complexes with certain tissue receptors. Preceding or following this, there may he some proteolytic degradation of these molecules which causes them to assume the characteristic fibrillar appearance, In support of this view is the demonstration by Osserman et al. (1964) that fluoresceinated Bence-Jones proteins, especially those derived from patients with amyloidosis, can bind to tissue sections, and the recent demonstration by Glenner et aZ. ( 1 9 7 1 ~ )and by Linke in our laboratory

FIG. 19. Monocyte obtained from specimen incubated with antiserum-treated amyloid fibrils. The phagocytic vacuoles containing fibrils (Am) vary in size and are irregular in circumference, presumably because of coalescence with other phagosomes. The arrow points to site where such coalescence may have taken place. Lysosome ( L ) . Magnification: X 13,000. ( From Zucker-Franklin, 1970.)

CURRENT CONCEPTS OF AMYLOID

293

FIG.20. Reticuloendothelial cell from a bone marrow specimen of a patient with multiple myeloma. The fibrils occupy membrane-bound spaces reminiscent of coalesced phagocytic vacuoles. Arrow points to site where amyloid (Am) appears to be engulfed by the cell. Location of amyloid in this cell is in marked contrast to intracytoplasmic fibrils illustrated in Figs. 6 and 7. Magnification: X6000.

294

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

that amyloid can be formed in the test tube from Bence-Jones proteins. It is of some interest that in amyloid deposits as well as in the production of synthetic amyloid from Bence-Jones proteins there is an inordinately high frequency of proteins related to h light chains-a finding suggesting that this type of protein has some structural feature that makes it more susceptible to the deposition as amyloid. The recent chemical and immunological studies and others likely to derive from them in the near future, as well as a more precise characterization of the amyloid proteins not related to immunoglobulins, may well bring us closer to an answer to some of these questions. Another problem which so far has defied a clear-cut explanation is the long persistence of amyloid, often even after removal of the inciting stimulus, A partial explanation for this may be provided by the marked resistance of native amyloid fibrils to proteolysis and their failure to be ingested by peripheral blood phagocytes under conditions where these cells are able to engulf bacteria, foreign particles, and denatured serum proteins ( Fig. 17) ( Zucker-Franklin, 1970). Only fibrils treated with specific antiserum to amyloid were avidly phagocytized by these cells (Figs. 18 and 19). However, once the antibody-treated fibrils were seen within phagocytic vacuoles, they appeared to undergo ultrastnictural changes suggesting that Iysosomal enzymes may be capable of degrading amyloid intracellularly (Fig. 18). It is quite likely that in vivo, under some conditions, when the stimulus for the formation of amyloid has ceased, the fibrils can also be removed slowly by phagocytosis and intracellular degradation. Fifty years ago, Waldenstrom ( 1928 ) observed by means of serial splenic biopsies that excision of tuberculous fistulas or the amputation of limbs afflicted with osteomyelitis was accompanied by a reduction of splenic amyloidosis. More recently, isolated reports concerned with the resorption of amyloid in experimental animals (Richter, 1954) and in a few patients have also appeared (Lowenstein and Gallo, 1970). The RE cell depicted in Fig. 20 may bear witness to the phagocytosis of amyloid which may occur in the living organism. V. Speculations on the Pathogenesis of Amyloid

To the best of our knowledge, there is no other disease known to man which occurs naturally with as many associated disorders as amyloid. In addition, there is no other disorder which can be induced as readily in a variety of species with as many different agents and procedures as amyloid. Consequently, over the years there has accumulated a great mass of often contradictory information, all of which osten-

CURRENT CONCEPTS OF AMYLOID

295

sibly bears on the pathogenesis of the disease. It is obviously beyond the scope of this review to summarize all of this material in detail. We shall, therefore, discuss some of our own thoughts on this interesting question and cite only those experimental studies that have direct bearing on them. The theories published by some investigators, e.g., Teilum ( 1964a,b, 1968), Muckle ( 1968a,b), Janigan and Druet (1966, 1968a,b), and Janigan (1969) are primarily based on experimental results and, therefore, have the greatest appeal to us. For a more widely rounded discussion of the problem of pathogenesis, the reader is referred to the book “Amyloidosis” ( Mandema et al., 1968). In recent years, discussions related to pathogenesis of amyloid have dealt primarily with immunological mechanisms and have implicated the RES as the responsible organ system. Chemical and immunological studies demonstrating the relation of amyloid to immunoglobulin light chains provide direct proof for concepts based on the association of amyloid in man with a variety of chronic infectious diseases and plasma cell dyscrasias and support the idea that a hyperactive stimulated RES, frequently accompanied by excessive y-globulin production, may play a role in the pathogenesis (Teilum, 1968). However, in contrast to these conditions where the deposition of amyloid is associated with a hyperactive immunological system, amyloid also occurs with a fairly high frequency in certain of the immune deficiency diseases such as agammaglobulinemia-to date, 9 cases of amyloid complicating agammaglobulinemia have been reported (Teilum, 1968; Mawas et al., 1969). Amyloid also complicates disorders such as Hodgkin’s disease and other lymphomas where impairment in delayed hypersensitivity occurs. In animals, too, a similar association of amyloid with states of prolonged antigenic stimulation as well as immunological incompetence has been noted. It is a frequently cited fact that horses given prolonged antigenic stimulation for the production of commercial antisera often die with amyloidosis. These animals have very high levels of 7-globulin and high titers of antibodies to the immunogens used for immunization. Similarly, in the various experimental models, stimulation of the RES accompanied by active antibody synthesis is a common feature. Thus, casein and other substances (Druet and Janigan, 1966a,b) are given in large amounts, and both Freund’s adjuvant and endotoxin ( Barth et al., 1969a), now widely used to induce amyloid, have a profound stimulating effect on the RES. Nevertheless, the studies by Pierpaoli and Clerici (1964; Clerici et al., 1965) clearly demonstrated that antibody synthesis is not necessary and that animals made tolerant to casein are able to produce amyloid as readily as control animals. Contrary results were subsequently reported by Letterer and Kretschmer (1966).

296

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

Although most regimens employed to induce amyloid stimulate the

RES and present it with a profound antigenic challenge, according to

Druet and Janigan (1966a), these may, at the same time, cause thymic atrophy and peripheral lymphopenia. In addition, there appear to be a number of factors that accelerate amyloid formation seemingly by virtue of the fact that they inhibit the immune system. Thus steroids, immunosuppressive agents, and X-irradiation ( Hultgren et al., 1967; Teilum, 1968; Bradbury et al., 1964) have all been found to accelerate significantly the deposition of amyloid, as have thymectomy and bursectomy (Druet and Janigan, 1966b; Clerici et al., 1969)-a11 these procedures have in common the ability to decrease one or another phase of the immune response. Nevertheless, certain exceptions to this general rule have recently been reported since, in the hands of Ranlov (1968), both antilymphocyte serum and nitrogen mustard appear to inhibit cellular immunity and casein-induced amyloid. Thus we are faced by an apparent paradox. On the one hand, amyIoid occurs in a number of diseases which superficially have a hyperactive immune system and which are accompanied by excess y-globulin synthesis. On the other hand, amyloid often is found in association with diseases accompanied by various immune deficiency states and can be accelerated in experimental animals by a variety of procedures and agents which are generally used to inhibit the immune response. Furthermore, a recent report suggests that if lymphocyte transformation is used as a measure of delayed hypersensitivity, then patients with amyloidosis have a diminished response to herpes simplex virus but not to two other antigens nor to phytohemagglutinin (Lehner et al., 1970). We have recently confirmed the normal response to phytohemagglutinin (Friedman et d.,1970), but we have not investigated other antigens. The results presented above suggest the possibility that in all instances amyloid occurs in a situation where the antigenic stimulus is excessive for the available immune system or where, because of neoplastic change, it has escaped from normal control mechanisms. This may cause the immune system to respond with an exaggerated or an alternative response, possibly one existing during a stage of evolution before the development of classic antibodies and which may again be elicited under unusual stress. There is abundant evidence from phylogenetic studies that light chains represent the most primitive precursor of immunoglobulins. Impairment of the immune system in immune deficiency diseases and under the influence of immunosuppressives, X-rays, or steroids is obvious. Similarly, this impairment has been recognized in diseases affecting lymphopoietic tissues, such as Hodgkin’s disease and lym-

CURRENT CONCEPTS OF AMYLOID

297

phomas. Even in neoplastic disorders affecting plasma cells, such as myeloma and macroglobulinemia, despite the abundance of plasma cells and the excessive synthesis of y-globulins, it has been recognized for a long time that these cells are largely committed and that responsiveness to antigenic challenge is diminished or even absent. Thus, the presence of a plasma cell tumor or excess y-globulin may be accompanied by impaired immediate and delayed hypersensitivity. An analogous situation may well exist in chronic infections, rheumatoid arthritis, hyperimmunized horses, and casein-treated animals, all of which produce excessive amounts of antibodies. In all of these instances, it seems possible that the RES is committed to the synthesis of a specific type of antibody and, thus is immunologically deficient when a new and possibly weaker stimulus appears. In the case of casein-stimulated rabbits, Rodey and Good (1969) have recently demonstrated that mouse spleen cells from animals, pretreated with azocasein or Freund’s adjuvant, have a diminished response to phytohemagglutinin, possibly by prior activation and consequent “removal” of thymus-dependent cells. The immune response to other antigens in chronic infections has not been studied in detail; however, defects in certain aspects of the immune response have been observed in rheumatoid arthritis and systemic lupus. Particularly suggestive are the recent tissue culture studies of Herman et a?. (1971) which indicate that synovial cells removed from patients with rheumatoid arthritis who had been previously stimulated with tetanus toxoid produced very large amounts of y-globulin but that tetanus antitoxin constituted a much smaller fraction than lymphoid tissues from normals or synovia from other diseases. This finding clearly suggests that these tissues, although rich in plasma cells and actively synthesizing antibodies, are markedly defective in their response to another new antigenic stimulus. It would be of interest to examine these tissues for the presence of amyloid or if possible for their ability to synthesize amyloid in uitro. Consistent with such a view are the observations of Teilum (1954, 1964a,b, 1968) who, on the basis of morphological studies of tissues during the evolution of amyloid, has proposed a two-stage mechanism. Initially, there is a pronounced stimulation of pyroninophilic antibodyproducing plasma cells and RE cells which occurs prior to the elaboration of amyloid. As the antibody-producing system is exhausted, there is a decrease in the pyroninophilic cells and a concomitant diminution in the production of y-globulin. This is accompanied by a proliferation of periodic acid-Schiff (PAS )-positive RE cells which appear to be involved in the synthesis of amyloid. To date there are no studies of the immune response in many of the experimental models used to induce amyloidosis,

298

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

but the phenomenon of antigenic competition makes it appear possible that the response to a new antigenic challenge might be decreased. A similar situation might be found in amyloidosis commonly associated with aging (Schwartz, 1968; Wright et al., 1969). It has long been recognized that small amounts of amyloid can be found at postmortem in many elderly individuals. This material has all the morphological attributes of amyloid and is estimated to be present in at least 50%of individuals over the age of 70. Although the immune response in man has not been carefully correlated with age, there has been the general impression that it diminishes and that this diminution may be one of the possible reasons for the increased incidence of neoplasia with age (Calkins, 1968). Thus it appears likely that amyloid in the aged accompanies a less active immune system than was available in youth. What then can we postulate as a possible pathogenetic mechanism? In its simplest form we might say that amyloid is likely to occur in situations where the size of the antigenic stimulus is excessive compared to the capacity of the immune system or where the immune system has escaped from normal control mechanisms. Thus, if the immune system is normal it can tolerate an enormous antigenic load (e.g., casein) before amyloid is produced. If it is defective, as in agammaglobulinernia, a smaller stimulus is effective. Whether only a specific type of antigen is capable of initiating this process or whether any antigen can do so, whether the nature of the stimulus predetermines the type of amyloid, and whether only living infectious agents might be capable of inciting it remains to be established. Since the material deposited is obviously heterogeneous, it is difficult to propose a unified concept. The possibility that amyloid is a consequence of imbalanced immunoglobulin chain synthesis or, possibly, even represents a vestigial response of some component of the immune system, which is latently present and called upon only when it is overwhelmed or dedifferentiated, is suggested by the light-chain nature of the basic protein subunit and by the presence of amyloid fibrils in plasma cells and lymphocytes under certain circumstances. Obviously, pursuit of this line of investigation and further chemical characterization of the other types of amyloid will prove most fruitful in unraveling this difficult problem. Although not directly related to the above concepts, one other aspect of experimental amyloid research deserves mention for the sake of completeness and because of its potential interest. This deals with a model in which amyloid enhancing factor (AEF) can be recovered from tissues, subcellular fractions, and, possibly, serum of animals that are developing amyloidosis. As assayed in the recipients, AEF markedly

CURRENT CONCEPTS OF AMYLOID

299

shortens the induction time for amyloid incited with casein or casein derivates. Although AEF has not yet been precisely characterized, it has been reported by one group to occur in serum and to be related to the 7-globulin fraction ( Janigan and Druet, 1968a,b). Others have fractionated cells and recovered it in the fraction containing cell nuclei after fractionation, probably deoxyribonucleic acid protein ( Ranlov, 1968) and also in the cytoplasm (Janigan and Druet, 1968a,b). Several reports suggest that AEF can cross species lines since such fractions from human amyloid can enhance the induction of the disease in mice. Whether the material transferred is an amyloid precursor, antigen per se, or antigen coupled to some cell constituents so as to make it more potent remains to be determined. A recent report by Rimon et al. (1972) has characterized this factor as a low molecular glycoprotein, unrelated to the amyloid fibrils. At any rate, it appears that AEF acts on the RE cells of the host and causes them to produce amyloid at a more rapid rate than normally. Although the model has not provided any definite answers, it appears at this time worthy of further study,

REFERENCES Andrade, C., Araki, S., Block, W. D., Cohen, A. S., Jackson, C. E., Kuroiwa, Y., McKusick, V. A., Nissim, J., Sohar, E., and Van Allen, M. W. (1970). Arthritis Rheum. 13, 902. Ashkenazi, Y., Gafni, J., Sohar, E., Heller, H., Shmueli, U., and Hersko, H. (1968). I n “Amyloidosis” ( E . Mandema et ul., eds.), pp. 263-270. Excerpta Med. Found., Amsterdam. Barth, W. F., Willerson, J. T., Asofsky, R., Sheagren, J. N., and Wolff, S. M. (1969a). Arthritis Rheum. 12, 615. Barth, W. F., Willerson, J. T., Waldmann, T. A., and Decker, J. L. (1969b). AWW. J . Med. 47, 259. Battaglia, S. (1961). Klin. Wochenschr. 39, 795. Battaglia, S. (1962). Beitr. Pathol. Anut. Allg. Pathol. 126, 300. Benditt, E. P. (1968). In “Amyloidosis” (E. Mandema et al., eds.), p. 137. Excerpta Med. Found., Amsterdam. Benditt, E. P., and Eriksen, N. (1964). Arch. PathoE. 78, 325. Benditt, E. P., and Eriksen, N. (1966). PTOC. Nut. Acad. Sci. U. S. 55, 308. Benditt, E. P., and Eriksen, N. (1971). Amer. J. Pathol. 65, 231. Benditt, E. P., Lagunoff, D., Eriksen, E., and Iseri, 0. A. (1962). Arch. Pathol. 74, 323. Benditt, E. P., Eriksen, N., and Berglund, C. (1968). In “Amyloidosis” (E. Mandema et ul., eds.), pp. 206-216. Excerpta Med. Found., Amsterdam. Benditt, E. P., Eriksen, N., and Berglund, C. (1970). PTOC.Nut. Acud. Sci. U. S . 66, 1044. Benditt, E. P., Eriksen, N., Hermodson, M. A., and Ericsson, L. H. (1971). FEBS Lett. 19, 169.

300

EDWARD C. FRANKLIN AND WROTHEA ZUCKER-FRANKLIN

Ben-Ishay, Z., and Zlotnick, A. (1968). Isr. J . Med. Sci. 4, 987. Bennhold, H. (1922). M u e n c h . Med. Wochenschr. 69, 1537. Bitter, T., and Muir, H. (1966). J. Clin. Inuest. 45, 963. Bladen, H. A., Nylen, M. U., and Glenner, G. G. (1966). J. Ultrastruct. Res. 14, 449. Boere, H., Ruinen, L., and Scholten, J. H. (1965). J . Lab. Clin. Med. 66, 943. Bonard, L., Cohen, A. S., and Skinner, M. (1969). Proc. SOC. E x p . Biol. Med. 131, 1373. Bradbury, S., Phil, D., and Micklem, H. S. (1964). Amer. J . Pathol. 46, 263. Brandt, K., Cathcart, E. S., and Cohen, A. S. (1968). Amer. J . Med. 44, 955. Brenner, S., and Home, R. W. (1959). Biochim. Biophys. Acta 34, 103. Buckley, J. K., and Porter, K. R. ( 1967). P r o t o p l u m 64, 349. Buxbaum, J., Zolla, S., Scharff, M. D., and Franklin, E. C. (1971). J. Exp. Med. 133, 1118. Bywaters, E. (1968). In “Amyloidosis” ( E . Mandema et al., eds.), p. 80. Excerpta Med. Found., Amsterdam. Caesar, R. (1960). Z . Zellforsch. Mikrosk. Anat. 52, 653. Cagli, V., Carbonara, A,, and Mancini, G. (1962). Boll. SOC. ltal. Biol. Sper. 38, 353. Calkins, E. (1968). In “Amyloidosis” (E. Mandema et al., eds.), pp. 87-99. Excerpta Med. Found., Amsterdam. Calkins, E., Cohen, A. S., and Gitlin, D. (1958). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 17, 431. Carnes, W. H., and Forbes, B. R. (1956). Lab. Invest. 5, 21. Cathcart, E. S., and Cohen, A. S. (1966). J. Immunol. 96, 239. Cathcart, E. S., and Cohen, A. S. (1968). In “Amyloidosis” ( E . Mandema et al., eds. ), pp. 130-138. Excerpta Med. Found., Amsterdam. Cathcart, E. S., Wollheim, F. A., and Cohen, A. S. (1967a). J . lmmunol. 99, 376. Cathcart, E. S., Wollheim, F. A., and Cohen, A. S. (1967b). Proc. SOC. E x p . B i d . Med. 125, 1123. Cathcart, E. S., Shirahama, T., and Cohen, A. S. ( 1 9 6 7 ~ )Biochim. . Biophys. Acta 147, 392. Cathcart, E. S., Skinner, M., Cohen, A. S., Lawless, 0. J., and Benson, M. D. ( 1970a). Nature (London) 228, 1090. Clerici, E., Pierpaoli, W., and Romussi, M. (1965). Pathol. Microbiol. 128, 806. Clerici, E., Mocarelli, P., Smeraldi, E., and Villa, M. L. (1969). Clin. Exp. Immunol. 5, 163. Cohen, A. S. (1966a). Lab. Znuest. 15, 66. Cohen, A. S. (1966b). Int. Reu. Exp. Pathol. 4, 159. Cohen, A. S. (1967). N . Engl. J . Med. 2’77, 527, 574, and 628. Cohen, A. S., and Calkins, E. (1959). Nature (London) 183, 1202. Cohen, A. S., and Calkins, E. (1960). J. Erp. Med. 112, 479. Cohen, A. S., and Paul, W. E. ( 1963). Nature (London) 197, 193. Cohen, A. S., and Calkins, E. (1964). J . Cell Biol. 21, 481. Cohen, A. S., Gross, E., and Shirahama, T. (1965). Amer. J . Pathol. 47, 1079. Dahlin, D. C. (1950). Med. Clin. N . Amer. 34, 1107. Dalferes, E. R. (1968). Proc. SOC. E x p . Biol. Med. 127, 925. Dixon, F. J. (1965). In “Immunopathology” (P. Grabar and P. Miescher, eds.), p. 309. Grune & Stratton, New York. Druet, R. L., and Janigan, D. T. (196Sa). Amer. J . Pathol. 49, 911.

CURRENT CONCEPTS OF AMYLOJD

301

Druet, R. L., and Janigan, D. T. (1966b). Amer. J. Pathol. 49, 1103. Eanes, E. D., and Glenner, G. G. (1968). J . Histochem. Cytochem. 16, 673. Ein, D., Kimura, S., and Glenner, G. G. (1972). Biochem. Biophys. Res. Commun. 46, 498. Eliakim, M., Rachmilewitz, M., Rosenmann, E., and Niv, A. (1970). Isr. J. Med. Sci. 6, 228. Emeson, E. E., Kikkawa, Y., and Gueft, B. (1966). J . Cell Biol. 38, 570. Fawcett, D. W, (1966). In “The Cell,” p. 247. Saunders, Philadelphia, Pennsylvania. Franklin, E. C. (1970). Unpublished data. Franklin, E. C., and Levin, M. (1972). Unpublished observation. Franklin, E. C., and Pras, M. (1969). J. Exp. Med. 130, 797. Franklin, E. C., and Zucker-Franklin, D. (1972). Proc. SOC. Ezp. Biol. Med. 140, 565. Franklin, E. C., Frangione, B., Levin, M., and Pras, M. (1972a). Unpublished data. Franklin, E. C., Pras, M., Levin, M., and Frangione, B. (1972b). FEBS Lett. 22, 121. Friedman, E., Brown, S. M., and Franklin, E. C. (1970). Unpublished data. Friedreich, N., and KekulB, A. (1859). Arch. Pathol. Anat. Physiol. Klin. Med. 16, 50. Gafni, J,, and Sohar, E. (1960). Amer. J . Med. Sci. 102, 332. Gafni, J., Hersko, C., Sohar, E., Heller, H., and Ashkenazi, Y. (1968). In “Amyloidosis” ( E . Mandema et al., eds.), pp. 271-279. Excerpta Med. Found., Amsterdam. Glenner, G. G. ( 1972). Personal communication. Glenner, G. G., and Bladen, H. A. (1966). Science 154, 271. Glenner, G. G., and Bladen, H. A. (1968). In “Amyloidosis” ( E . Mandema et al., eds.), pp. 137 and 216. Excerpta Med. Found., Amsterdam. Glenner, G. G., Keiser, H. R., Bladen, H. A., Cuatrecasas, P., Eanes, E. D., Ram, J. S., Kanfer, J. N., and DeLellis, R. A. (1968). J . Histochem. Cytochem. 16, 633. Glenner, G. G., Cuatrecasas, P., Isersky, C., Bladen, H. A., and Eanes, E. D. (1969). 1. Histachem. Cytochem. 17, 769. Glenner, G. G., Harada, M., Isersky, C., Cuatrecasas, P., Page, D., and Keiser, H. (1970a). B i o c h . Biophys. Res. Commun. 41, 1013. Glenner, G. G., Harbaugh, J., Ohms, J. I., Harada, M., and Cuatrecasas, P. (1970b). Biochem. Biophys. Res. Commun. 41, 1287. Glenner, G. G., Page, D., Isersky, C., Harada, M., Cuatrecasas, P., Eanes, E. D., DeLellis, R. A., Bladen, H. A,, and Keiser H. R. (1971a). J. Histochem. Cytochem. 19, 16. Glenner, G. G., Terry, W., Harada, M., Isersky, C., and Page, D. (1971b). Science 172, 1150. Glenner, G. G., Ein, D., Eanes, E. D., Bladen, H. A., Terry, W., and Page, D. ( 1 9 7 1 ~ )Science . 174, 712. Glenner, G. G., Ein, D., and Terry, W. D. (1972). Amer. J. Med. 52, 141. Goldberg, A. F., and Deane, H. W. (1960). Blood 16, 1708. Gueft, B., and Ghidoni, J. J. (1963). Amer. J . Pathol. 43, 837. Harada, M., Isersky, C., Cuatrecasas, P., Page, D., Bladen, H. A., Eanes, E. D., Keiser, H. R., and Glenner, G. G. (1971). J . Histochem. Cytochem. 19, 1. Hass G., and Schulz, R. Z. (1940). Arch. Pathol. 30, 240. Heefner, W. A., and Sorenson, G. D. (1962). Lab. Invest. 11, 585.

302

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

Heller, H., Missmahl, H. P., Sohar, E., and Gafni, J. (1964). J. Puthol. Bacteriol. 88, 15. Herman, J. H., Bradley, J., Ziff, M., and Smiley, J. D. ( 1971). J. Clin. Inuest. 50, 266. Hjort, G. H., and Christensen, H. E. (1961). Acta Rheumutol. S c a d . 7, 65. Horowitz, R. E., Stuyvesant, V. W., Wigmore, W., and Tatter, D. (1965). Arch. Puthol. 79, 238. Hultgren, M. K., Druet, R. L., and Janigan, D. T. (1967). Amer. J . Puthol. 50, 943. Isersky, C., Ein, D., Page, D. L., Harada, M., and Glenner, G. G. (1972). J . Immunol. 10% 486. Janigan, D. T. (1969). Amer. 1. Pathol. 55, 379. Janigan, D. T., and Druet, R. L. (1966). Amer. J. Pathol. 48, 1013. Janigan, D. T., and Druet, R. L. (1968a). Amer. 1. Puthol. 52, 381. Janigan, D. T., and Druet, R. L. (196%). Zsr. J. Med. Sci. 4, 1035. Kim, I. C., Shirahama, T., and Cohen, A. S . (1967). Amer. J. Puthol. So, 869. Kim, I. C., Franzblau, C., Shirahama, T., and Cohen, A. S. (1969). Biochim. Biophys. Actu 181, 465. King, L. S. ( 1948). Amer. J . Puthol. 24, 1095. Lachmann, P. J., Miiller-Eberhard, H. J., Kunkel, H. G., and Paronetto, F. ( 1962). J. E x p . Med. 115, 63. Laufer, A., and Td, C. (1967). Brit. J . E x p . Puthol. 4 4 423. Lehner, T., Cameron, J. S., and Ware, R. G. (1970). Clin. E x p . Immunol. 6, 439. Letterer, E., and Kretschmer, R. (1966). Nature (London) 210, 390. Levin, M., Franklin, E. C., and Frangione, B. (1972a). Unpublished observation. Levin, M., Franklin, E. C., Pras, M., and Frangione, B. (1972b). J . Clin. Znuest. ( I n Press ) . Linke, R. P., and Franklin, E. C. (1972). Unpublished observations. Linke, R. P., Franklin, E. C., and Zucker-Franklin, D. (1972). Unpublished data. Lowenstein, J., and Gallo, G. (1970). N . Engl. J. Med. 282, 128. Luft, J. H. (1961). J. Biophy. Biochem. Cytol. 9, 409. Magnus Levi, A. (1956). J. Mt. Sinai Hosp., New York 19, 8. Mandema, E., Ruinen, L., Scholten, J. H., and Cohen, A. S., eds. (1968). “Amyloidosis.” Excerpta Med. Found., Amsterdam. Mawas, C., Sors, C., and Bernier, J. J. (1969). Amer. J. Med. 46, 624. Mellors, R. C., and Ortega, L. G. (1956). Amer. J. Pathol. 32, 455. Milgrom, F., Kasukawa, R., and Calkins, E. (1966). J. Immunol. 99, 245. Miller, H. I., Rotman, Y., Ben-Shaul, Y.,and Ashkenazi, Y. (1968). Isr. J . Med. Sci. 4, 982. Millonig, G. (1961). J. Biophys. Biochern. Cytol. 11, 736. Milstein, C., and Pink, R. (1970). Progr. Biophys. Mol. Blol. 21, 209. Missmahl, H. P. (1957). 2. Wiss. Mikrosk. 63, 133. Missmahl, H. P. (1968). In “Amyloidosis” (E. Mandema et al., eds.), pp. 22-33. Excerpta Med. Found., Amsterdam. Missmahl, H. P., and Hartwig, M. (1953). Virchows Arch. Puthol. Anut. Physiol. 324, 489. Mowry, R. W., and Scott, J. C. ( 1967). Histochemie 10, 8. Muckle, T. J. ( 1964). Nature (London) 203, 773.

CURRENT CONCEPTS OF AMYLOID

303

Muckle, T. J. (1968a). In “Amyloidosis” ( E . Mandema et at., eds.), pp. 335-343. Excerpta Med. Found., Amsterdam. Muckle, T. J. (196813). Isr. J . Med. Sci. 4, 1020. . “Amyloidosis” (E. Manderna et al., eds.), pp. 122-129. Muckle, T. J. ( 1 9 6 8 ~ ) In Excerpta Med. Found., Amsterdam. Muckle, T. J. (1969). Ann. Rheum. Dis. 28, 31. Muir, H., and Cohen, A. S. (1968). In “Amyloidosis” ( E . Mandema et al., eds.), pp. 280-285. Excerpta Med. Found., Amsterdam. Newcombe, D. S., and Cohen, A. S. (1964). Biochim. Biophys. Acta 104, 480. Osserman, E. F., Takatsuki, K., and Talal, N. (1964). Semin. Hemutol. 1, 3. Paul, W. E., and Cohen, A. S. (1963). Amer. J . Pathol. 43, 721. Pick, A. I., and Osserman, E. F. (1968). In “Amyloidosis” (E. Mandema et aZ., eds.), pp. 100-109. Excerpta Med. Found., Amsterdam. Pierpaoli, W., and Clerici, E. (1964). Experientia 20, 693. Pras, M., and Reshef, T. (1972). Biochim. Biophys. Acta 271, 193. Pras, M., and Schubert, M. (1968). J . Histochem. Cytochm. 17, 258. Pras, M., Schubert, M., Zucker-Franklin, D., Rirnon, A., and Franklin, E. C. (1968). J. Clin. Invest. 47, 924. Pras, M., Zucker-Franklin, D., Rimon, A., and Franklin, E. C. (1969). J . Exp. Med. 130, 777. Pras, M., Nevo, Z., Schubert, M., Rotman, J., and Matalon, R. (1971). 1. Histochem. 19, 443. Ram, J. S., DeLellis, R. A., and Glenner, G. G. (1968). Int. Arch. Allergy Appl. Immunol. 34, 269. Ranlov, P. (1968). In “Amyloidosis” (E. Mandema et al., eds.), p. 318. Excerpta Med. Found., Amsterdam. Ranlov, P., and Wanstrup, J. (1968). In “Amyloidosis” (E. Mandema et al., eds.), pp. 7 4 4 1 . Excerpta Med. Found., Amsterdam. Reimann, H. A., Koucky, R. F., and Eklund, C. M. (1935). Amer. J. Pathol. 11, 977. Richter, G. (1954). Amer. 1. Pathol. 30, 239. Rimon, A. et nl. (1972). Personal communication. Rodey, G. E., and Good, R. A. ( 1969). Proc. SOC. Exp. Biol. Med. 131, 457. Ruinen, L., Scholten, J. H., and Mandema, E. (1967). Genet. Elem.; Prop. Funct., Symp., 1966 p. 325. Schultz, R. F., Calkins, E., Milgrom, F., and Witebsky, E. (1964). Arthdtis Rheum. 7, 751. Schultz, R. F., Calkins, E., Milgrom, F., and Witebsky, E. (1966). Amer. J . Pathol. 48, 1. Schultz, R. F., Kasukawa, R., Calkins, E., and Milgrom, F. (1968). In “Amyloidosis” ( E . Mandema et d.,eds.), pp. 139-146. Excerpta Med. Found., Amsterdam. Schwartz, P. ( 1968). In “Amyloidosis” ( E . Mandema et al., eds.), pp. 400-417. Excerpta Med. Found., Amsterdam. Sellin, D., and Haferkamp, 0. (1968). 2. Gesamte Erp. Med. 147, 173. Shapira, E., Yogev, R., Pras, M., and Russell, A. (1970). Unpublished data. Shibolet, S., Merker, H. J., Sohar, E., Gafni, J., and Heller H. (1967). Brit. J . Exp. Pathol. 48, 244. Shirahania, T., and Cohen, A. S. (1965). Nature (London) 2M, 737. Shirahama, T., and Cohen, A. S. (1967a). J. Cell Biol. 35, 459. Shirahama, T., and Cohen, A. S. (1967b). J . Cell Bid. 33, 679.

304

EDWARD C. FRANKLIN AND DOROTHEA ZUCKER-FRANKLIN

Shmueli, U., Gafni, J., Sohar, E., and Ashkenazi, Y. (1969). J. Mol. Biol. 41, 309. Skinner, M., and Cohen, A. S. (1971). Biochim. Biophys. A d a 236, 183. Sohar, E., Gafni,J., Pras, M., and Heller, H. ( 1967). Amer. J. Med. 43, 227. Sorenson, G . D., and Bari, W. A. (1968). In “Amyloidosis” (E. Mandema et al., eds.), pp. 58-70. Excerpta Med. Found., Amsterdam. Sorenson, G. D., and Binington, H. B. (1964). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 23, 550. Sorenson, G. D., and Shimamura, T. (1964). Lab. Inuest. 13, 1409. Spiro, D. (1959). Amer. J . Pathol. 35, 47. Suzuki, Y.,Churg, N., Grishman, E., Mautner, W., and Dachs, S. (1963). Amer. J . Pathol. 43, 555. Symmer, W. St. C. (1956). J. C h . Pathol. 9, 187. Teilum, G. (1954). J. Lab. Clin. Med. 43, 367. Teilum, G. ( 1956). Amer. J. Puthol. 32, 945. Teilum, G. ( 1964a). J . Pathol. Bacteriol. 88, 317. Teilum, G. (1964b). Acta Pathol. Microbiol. Scand. 61, 21. Teilum, G. ( 1968). In “Amyloidosis” (E. Mandema et al., eds.), pp. 37-44. Excerpta Med. Found., Amsterdam. Terry, R. D. (1963). I. Neuropathol. Exp. Neurol. 22, 629. Terry, R. D. (1964). Amer. J . Pathol. 44, 269. Thiery, J. P., and Caroli, J. (1961). Sem. Hop. 37, 29. Vazquez, J. J., and Dixon, F. J. (1956). I. Erp. Med. 1% 727. Virchow, R. (1854). Arch. Puthol. Anat. Physiol. Klin. Med. 6, 268 and 416. Vogt, A., and Kochem, H. G. (1960). 2. Zellforsch. Mikrosk. Anat. 52, 640. von Rokitansky, C. F. ( 1846). Vienna: Braumulkr Seidel 3. Waldenstrijm, H. (1928). A d a Chir. Scand. 63, 479. Watson, M. L. (1958).J. Biophys. Biochem. Cytol. 4, 475. WilIiams, R. C., Jr., and Law, D. H. (1960). I. Lab. Clin. Med. 56, 629. Wright, J. R., Calkins, E., Breen, W. J., Stoke, G., and Schultz, R. F. (1969). Medicine (Baltimore) 48, 39. Zucker-Franklin, D. (1970). 1. Ultrastruct. Res. 32, 247. Zucker-Franklin, D., and Franklin, E. C. (1970). Amer. J . Pathol. 59, 23.

Author Index Numbers in parentheses are reference numbers and indicate that an author’s work is referred to, although his name is not cited in the text. Numbers in italics show the page on which the complete reference is listed. Ashkenazi, Y., 264, 265, 269, 271, 290, A 301,302, 304 Abbot, A., 99, 100, 101, 162, 163 Ashley, H., 18(130), 89 Abramson, N., 134, 157 Askonas, B. A., 16(125), 31, 33, 80 Ackerman, A., 81, 94 (330), 89, 94, 98, 104, 105, 111, Ada, G. L., 2(7), 31, 32( 168, 177), 69 115, 116, 117, 119, 139, 145, 113, (7), 85, 90, 98, 99, 100, 102, 132, 149, 153, 157, 161, 162, 164, 165 137, 153, 157, 160, 161, 162, 163, Asofsky, R., 53(227), 81, 91, 94, 175, 165 206, 295, 299 Adler, F. L., 65, 92, 96, 97, 144, 145, Aspinall, R. L., 3(23), 86 147, 157, 159 Attia, M. A. M., 168, 205 Agarossi, G., 13(95), 58( 237), 88, 91 Auerbach, R., 81, 94 Aird, J., 25, 89 August, C. S., 4( 33), 86 Alexander, P., 181, 199, 205 Austen, K. F., 108, 165, 179, 208 Allan, T. M., 235, 243 C. M., 98, 102, 157, 162, 163 Austin, Allen, H. L., 169, 205 Auzins, I., 98, 107, 134, 157, 161 Allen, J. M., 130, 157, 176, 194, 212 Avrameas, S., 147, 162 Allen, J. M. V., 31, 94 Allison, A. C., 80, 94, 111, 112, 115, 116, Ax, W., 102, 159 helrad, M. A., 168, 173, 186, 187, 188, 118, 149, 150, 155, 157, 164 199, 205 Allman, V., 5(47), 10(47), 86 Alter, B. J., 128, 157 B Amante, L., 31(172), 32(172), 90 Bach, F. H., 128,157 Amkraut, A., 18(132), 89 Amos, D. B., 168, 172, 173, 179, 180, Bachvaroff, R., 169, 205 186, 188, 189, 192, 193, 194, 196, Bacon, L. D., 241, 243 Baer, R. B., 107, 158 197, 199, 202, 205, 206, 209 Amsbaugh, D. F., 25(156), 62(156), 63 Bailey, D. W., 175, 205, 223, 227, 239, 240, 243 (156, 262), 89, 92 Anderson, B., 24( 149), 25( 149), 31, 32 Baker, P. J., 25( 156), 62, 63, 89, 92 Balfour, B. M., 99, 157 (183), 41, 89, 90, 91 Ballantyne, D. L., 179, 212 Andrade, C., 251, 299 Balner, H., 200, 205 Aoki, T., 173, 194, 203, 205, 206 Bansal, S. C., 199, 202, 213 Araki, S., 251, 299 Barchilon, J., 81, 94 Archer, 0. K., 4( 29), 86 Argyris, B. F., 109, 110, 111, 113, 115, Bardawill, W. A., 189, 205 Bari, W. A., 255, 304 157 Armstrong, W. D., 23, 64, 74, 89, 93, Barker, C. G., 194, 205 Barnes, D. W. H., 3(13), 85 121, 122, 164 Barrett, M. K., 189, 206, 230, 244 Arnason, B. G., 4(27, 28), 86 Barth, R. F., 25( 156), 62( 156), 63( 156, Arrameas, S., 102, 161 262), 89, 92, 169, 170, 205 Asherson, G. L., 31, 40, 91, 173, 187, Barth, W. F., 252, 295, 299 205 305

306

AUTHOR INDEX

Basch, R. S., 24( 150), 89 Bashford, E. F., 167, 205 Bast, R. C., Jr., 105, 159 Basten, A., 10(74), 16(74, 127), 24, 26 (74), 29(74), 31, 32( 186), 33, 37 (74), 44(74), 45(220), 46(127), 47(220), 73(74), 75(74), 87, 89, 90, 91, 94, 151, 162 Batchelor, J. R., 170, 171, 172, 174, 178, 179, 180, 181, 182, 183, 184, 186, 191, 193, 196, 205, 207, 208 Battaglia, S., 261, 299 Battisto, J. R., 70(299), 71, 93 Baum, J., 62, 63, 64, 92 Bearn, A. G., 240, 243 Bedford, M., 139, 140, 165 Beizer, L. H., 197, 206 Bell, C., 147, 157 Belman, S., 15( 107), 88 Benacerraf, B., 10(66, 67, 68),12(66), 13(68), 15(98, 99, 100, 103, 104, 105, 111, 114, 115, 116, 121, 122), 16, 18(66, 67), 19(66),20( 67), 25 (163), 26(66, 162), 31(68), 32 (187a), 34(67), 35(67, 187a), 38 ( a s ) , 39(68), 40(68), 43(66), 45 (67, 219), 48(153, 225, 226), 49 (225), 50(67, 153, 225), 51(153), 52(153, 226), 53(225, 226), 54 (153, 225, 226), 55( 153), 56( 153), 63(260), 66(279), 67( 279), 68( 67, 153), 69( 153, 296), 70( 153, 298), 71( 153), 72( 153, 296), 74( 153), 75( 153), 76( 66), 77( 66), 78( 66, 103, 298), 79( 103), 81(339, 340), 83(162, 351), 84(153, 352), 85 (339, 340), 87, 88, 89, 90, 91, 92, 93, 94, 96, 105, 107, 119, 121, 123, 129, 130, 134, 135, 136, 147, 148, 152, 157, 158, 159, 160, 161, 162, 163, 164, 175, 176, 183, 207, 208, 211 Benditt, E. P., 252, 254, 265, 267, 275, 276, 278, 285, 288, 299 Benezra, D., 148, 164 Ben-Ishay, Z., 261, 300 Bennett, B., 47, 91 Bennett, D., 221, 245 Bennett, M., 175, 207 Bennhold, H., 249, 265, 300

Ben-Shaul, Y.,271, 302 Benson, M. D., 250, 280, 285, 300 Berg, K., 240, 243 Berglund, C., 265, 275, 299 Berglund, K., 40(205), 41(205), 91 Berke, G., 177, 205 Berken, A., 134, 135, 158 Berlin, R. D., 142, 164 Bernard, C., 170, 171, 214 Berne, B. H., 173, 179, 205 Bernier, J. J., 295, 302 Bernstein, I., 200, 205 Bernstein, S . E., 222, 223, 244 Beyse, I., 79(324), 81(324), 94 Bianco, C., 31(348), 94 Biberfeld, P., 176, 212 Biege, G., 181, 194, 197, 209 Biesecker, J., 183, 189, 191, 205, 213 Biggar, W. D., 186, 210 Bigliani, S., 173, 174, 192, 206 Bildsgie, P., 191, 205 Bill, A. H., 182, 209 Billingham, R. E., 168, 169, 174, 184, 185, 189, 192, 194, 197, 201, 202, 205, 206, 213, 217, 219, 220, 221, 222, 223, 224, 225, 227, 228, 229, 230, 231, 233, 238, 239, 240, 244, 247 Billington, W. D., 234, 244 Binington, H. B., 271, 280, 304 Biozzi, G., 107, 129, 157, 158, 160, 175, 200, 206, 21 1 Birbeck, M., 149, 157 Birch-Anderson, A., 133, 163 Biro, C. E., 105, 158 Birtch, A. G . , 169, 212 Bishop, D. C., 145, 158 Bitter, T., 286, 300 Bladen, H. A,, 250, 252, 253, 254, 255, 264, 267, 269, 271, 273, 274, 280, 281, 284, 292, 300, 301 Blaese, M., 220, 227, 228, 229, 246 Blair, P. B., 175, 199, 200, 206 Blanden, R. V., 96, 162 Bliss, J. Q., 169, 210 Bloch, K. J., 108, 165, 182, 206 Block, J. B., 126, 127, 163 Block, W. D., 251, 299 Blomgren, H., 24( 149), 25( 149), 31, 41, 89, 91

307

AUTHOR INDEX

Bloom, B. R., 47, 57(235a), 91, 203, 210 Bloom, E. T., 173, 179, 180, 198, 199, 201, 206 Bluestein, H. G., 21( 137), 22( 137), 76 (137), 89, 147, 158 Blume, M. R., 57(235a), 91, 203, 210 Bod, J. L., 10(71), 87 Boere, H., 255, 267, 300 Biirjeson, J., 59( 241), 92 Bonard, L., 267, 300 Bonhag, R. S., 175, 199, 200, 206 Bonn&, B., 235, 244 Borek, F., 15(110), 70(299), 71, 88, 93, 153, 164 Borel, Y.,,25(154), 68( 154, 291), 70 (154, 291), 71, 73(306), 74(154, 291), 89, 93 Boros, D. L., 108, 110, 111, 158 Borsos, T., 202, 206 Boss, J. H., 169, 212 Bouteille, M., 102, 161 Boyd, G. A., 129, 165 Boyden, S. V., 130, 134, 158, 165 Boyse, E. A,, 172, 173, 178, 179, 194, 202, 203, 205, 206, 207, 210, 212, 221, 245 Bradbury, S., 296, 300 Bradley, J., 297, 302 Brandt, K., 252, 300 Braun, W., 67( 283), 79, 80(325, 326, 327, 328, 329), 93, 94, 147, 149, 158, 161 Brautbar, H., 184, 192, 207 Brawn, R. J., 182, 189, 202, 206, 209, 235, 236, 245 Breen, W. J., 298, 304 Breitner, J. C. S., 16(127), 21(143), 45 (220), 46(127), 47(220), 76(123), 89, 91 Brenner, S., 255, 300 Brent, L., 168, 169, 172, 192, 205, 206, 221, 222, 229, 244 Breyere, E. J., 189, 206, 230, 244 Bridges, C. B., 217, 242, 244 Britton, S., 10(72), 25( 157), 87, 89, 103, 153, 158, 174, 181, 201, 203, 211 Broder, S., 170, 171, 206, 214 Brody, J . I., 197, 206 Brody, N. I., 66, 92

Brondz, B. D., 180, 196, 206 Brosseau, G. E., 215, 244 Brown, J. E., 169, 206 Brown, R. R., 156, 164 Brown, S. M., 296, 301 Brownstone, A., 71(301), 93 Brucer, M., 129, 165 Brumfitt, W., 130, 158 Brunner, K. T., 41(208), 91, 170, 171, 172, 174, 176, 177, 180, 190, 197, 198, 199, 202, 203, 206, 207, 211, 212 Bryant, B. F., 180, 181, 193, 210 Bubenik, J., 170, 171, 172, 173, 182, 184, 206, 209 Buckley, J. K., 261, 300 Buckley, R. H., 168, 172, 173, 186, 189, 199, 206 Buckton, K. E., 237, 244 Bullock, W. W., 71, 74, 93 Bulman, H. N., 15(102), 88 Bunker, H. P., 172, 213 Bunker, M. C., 218, 244 Burkoe, S. O., 189, 206 Burnet, F. M., 2(1, 2, 3), 3(16), 85 Burr, W., 190, 208 Busch, G. J., 169, 173, 208 Bussard, A. E., 6, 86 Buxbaum, J., 292, 300 Byrt, P. N., 31, 32( 168, 177), 39(200), 57 (200), 58(200), 90, 91, 121, 131, 137, 153, 157, 160, 165 Bywaters, E., 259, 300

C Caesar, R., 260, 300 Cagli, V., 279, 300 Cahan, A., 236, 237, 245, 246 Calkins, E., 250, 254, 255, 263, 277, 278, 284, 298, 300, 302, 303, 304 Cameron, J. S., 296, 302 Campbell, D. H., 97, 146, 158, 159, 163 Campbell, P. A., 21(139), 31(139, 174), 32( 174), 38( 139), 39( 139), 76 (139), 79, 89, 90, 94 Cannon, P. R., 107, 158 Cantor, H., 53(227), 81, 82(344), 91, 94 Canty, T. G., 176, 179, 206

308

AUTHOR INDEX

Carbonara, A., 279, 300 Caren, L. D., 179, 206 Carnaghen, R. B. A., 63(261), 92 Carnes, W. H., 270, 300 Caroli, J., 255, 304 Carpenter, C. B., 169, 208, 211 Carsten, M. E., 15(107), 88 Carswell, E. A., 76(318), 93, 221, 245 Carter, R. L., 23( 145), 89, 168, 208 Cartwright, G. E., 237, 246 Casey, A. E., 168, 172, 177, 193, 206 Casey, J. G., 193, 206 Cathcart, E. S., 250, 252, 253, 254, 269, 279, 280, 284, 285, 300 Cattanach, B. M., 217, 244 Cayeux, P., 108, 110, 111, 158, 163 Ceglowski, W. S., 175, 200, 206 Celada, F., 6, 86, 103,. 158, 217, 244 Ceppellini, R., 173, 174, 192, 206 Cerilli, G. J., 169, 207, 214 Cerottini, J.-C., 41(208), 91, 104, 115, 119, 139, 140, 142, 153, 155, 158, 162, 164, 165, 170, 171, 172, 176, 180, 188, 190, 197, 199, 203, 206, 207, 211, 212 Chai, C. K., 239, 244 Chambers, H., 167, 207 Chan, E. L., 13(92), 14, 21(139), 29 (92), 31( 139), 38( 139), 39( 139), 76(139), 88, 89 Chanmougan, D., 63(254), 92 Chantler, S. M., 178, 179, 180, 193, 207 Chaperon, E. A., 2(11), 6(52), 7(52), 10, 11(77), 13(11, 77), 14, 29, 31, 39, 75( l l ) , 85, 87 Chapuis, B., 170, 171, 174, 177, 190, 197,206,207,211 Chard, T., 170, 171, 203, 207 Cheers, C., 10(74), 16(74), 21, 24(74), 26( 74), 29( 74), 37(74), 44( 74), 73(74), 75(74), 76(143), 87, 89, 151, 162 Chessin, L. N., 59(241 ), 92 Chiller, J. M., 10(62), 11, 68, 69(292), 70(292), 74(292), 87, 93, 117, 158 Chinitz, A,, 176, 211, 226, 246 Choi, Y.S., 186, 210 Chouroulinkov, I., 178, 207 Christensen, H. E., 255, 302

Christie, G. H., 23( 146), 69(293), 70 (293), 89, 93 Churg, N., 263, 304 ChutnC, J., 168, 171, 173, 184, 186, 188, 194, 197, 198, 199, 200, 207, 209, 234, 246 Cinader, B., 15(117), 71( 117), 88, 222, 244 Claman, H. N., 2(11), 6, 7(52), 10, 11(77), 13, 14, 29, 31, 39, 41, 42 (212, 214), 75(11), 85, 87, 91, 105, 158 Clark, A. R., 107, 160 Clark, D. S., 179, 208 Clarke, B., 235, 244 Clarke, D. A., 200, 205 Clerici, E., 295, 296, 300, 303 Cline, M. J., 126, 127, 158 Cloudman, A. M., 172, 177, 213 Cluzan, R., 108, 158 Cochrane, C. G., 108, 158 Cock, A. G., 169, 183, 185, 201, 207 Coe, J. E., 18(131), 89, 176, 207 Cohen, A. S., 249, 250, 251, 252, 253, 254, 255, 261, 263, 264, 267, 269, 271, 273, 278, 279, 280, 284, 285, 286, 287, 295, 300, 302, 303, 304 Cohen, C., 179, 212 Cohen, I., 180, 194, 196, 197, 202, 205 Cohen, I. R., 184, 192,207 Cohen, J. J., 31, 41, 42(212, 214), 91 Cohen, M., 169, 170, 173, 207, 212 Cohen, S,, 171, 202, 207 Cohn, Z. A., 97, 109, 128, 138, 139, 144, 146, 158 Cole, G. A., 175, 200, 214 Cole, L. J., 6(53), 87 Cole, R. M., 138, 160 Colon, S., 31( 173), 32( 173), 90 Cone, R. E., 79,94 Connell, M. St. J., 11( 79), 87 Connolly, J. M., 149, 165 Cook, G. M. W., 130, 157 Coombs, R. R. A,, 31, 32, 90, 134, 161 Coons, A. H., 56, 91, 98, 149, 161, 165 Cooper, E. L., 240, 245 Cooper, G. N., 130, 161 Cooper, H. L., 59(241), 92 Cooper, M. D., 3(21), 86, 179, 208

AUTHOR INDEX

309

(153), 75(153), 84(153), 89, 90, Cooper, S., 194, 210 93 Coppleson, L. W., 12, 81(342), 88, 94 Coppola, E. D., 179, 207, 214 Davies, A. J. S., 2( lo), 5, 7, 10( 10, 44, Corson, J. M., 169, 172, 199, 207, 214 45, 46, 541, 12(87), 23( 145, 146), Cotran, R. S., 134, 157, 162 66, 68(286), 80, 85, 86, 87, 88, 89, 92, 93, 94, 118, 162 Cottier, H., 100, 158 Davis, M., 200, 211 Counts, R. B., 15( 113), 88 Deane, H. W., 98, 161, 270, 301 Courtenay, B. M., 23(146), 89 DeBurgh, P. M., 5(47), 10(47), 86 Courtenay, T., 169, 206 Decker, J. L., 252, 299 Courtenay, T. H., 229, 244 DeCosse, J. J., 169, 211 Craig, J. V., 241, 243 DeFronzo, A., 169, 214 Crawford, L. V., 170, 199, 208 de Harven, E., 173,203,205 Crisler, C., 179, 207 Cross, A. M., 5(40, 41, 44), 10(44), 86 de la Chapelle, A., 63(257), 92 Crowle, A, J., 168, 170, 172, 173, 186, Delaloye, B., 129, 160 DeLellis, R. A., 252, 254, 280, 284, 303 187, 188, 192, 207 Cruchaud, A., 107, 108, 110, 111, 113, Denman, A. M., 169, 207 115, 116, 131, 132, 133, 141, 142, de Petris, S., 61(244), 92 Despont, J. P., 108, 110, 111, 115, 159 143, 144, 158, 159, 165 DeVries, M. J., 234, 244 Cruchaud, S., 129, 157 Cruse, J. M., 170, 172, 173, 174, 178, Dewitt, C. W., 169, 199, 212 Diamantstein, T., 79, 81( 324), 94 179, 180, 184, 198, 199, 207, 208 Cuatrecasas, P., 250, 252, 253, 254, 271, Diener, E., 23(147), 64(147), 74, 75, 89, 93, 121, 122, 164 286, 301 Cudkowicz, C., 11(79, SO), 12, 26( 82), Diethelm, A. C., 169, 212 27(82), 29, 39, 75(166), 76, 87, Dietrich, F. M., 23( 145), 89 DiCeorge, A. M., 4(30, 31, 32, 34), 86 90, 175, 207 DiPietro, S., 13(95), 58(237), 88, 91 Cuendet, A., 129, 157 Cunningham, A. J., 7(59), 8(59), 10 Dire, J., 237, 245 Dixon, F. J., 63(252), 92, 98, 104, 108, (59), 76(317), 87, 93 115, 119, 141, 158, 162, 164, 169, Cunningham, C., 237, 244 188, 189, 199, 207, 211, 212, 213, Curtoni, E. S . , 173, 174, 192, 206 214, 254, 277, 278, 284, 300, 304 Dobson, E. L., 129, 159 D Doria, G., 13, 58, 88, 91 Doty, P., 145, 146, 160 Dachs, S., 263, 304 Dougherty, S. F., 151, 162, 175, 200, 209 Dagg, M. K., 189,210 Dougherty, T. F., 40(206), 91 Dahlin, D. C., 300 Douglas, P., 172, 177, 213 Dalferes, E. R., 286, 300 Dalmasso, A. P., 4( 29), 5(36, 38), 86 Douglas, S. D., 59(241), 92, 135, 160 Dowling, E. A., 193, 206 Darlington, C. D., 235, 246 Dray, S., 147, 157 Das, M. L., 146, 159 Dresser, D. W., 67, 93, 105, 149, 151, Dausset, J., 173, 194, 203, 210 159 David, C . S . , 242, 244 Dronamraju, K. R., 216, 244 David, J. R., 47, 91 Davie, J. M., 25( 153), 31, 32( 175, 181, Druet, R. L., 295, 296, 299, 300, 301, 302 182), 48( 153), 50( 153), 51(153), 52( 153), 54( 153), 55( 153), 56 Dube, 0. L., 172, 173, 180, 199, 211 (153), 68(153), 69(153, 296), 70 Dubernard, J. M., 169, 212 (153), 71(153), 72(153, 296), 74 Dubert, J. M., 222, 244

310

AUTHOR INDEX

Eustace, J. C., 170, 171, 172, 180, 209 Dubowitz, V., 237, 247 Evans, C. A., 182, 201, 209 DuBuy, H. G., 151, 159 Duffus, W. P. H., 61(244), 92 Duke, D., 194, 209 F Dukor, P., 5(40, 41, 47), 10(47), 31 (348), 86, 94, 128, 138, 144, 165 Fahey, J. L., 170, 171,209 Failor, F., 172, 177, 213 Dulaney, A. D., 172, 179, 180, 207 Falkoff, R., 58( 239), 91 Dumont, A. E., 169, 211 Farquar, M., 107, 164 Duquella, J., 169, 212 Dutton, R. W., 12, 13, 15(102), 20(83, Fass, L., 183, 207 84), 21, 31, 32(191), 38(90, 139, Fauci, A. S., 66, 92, 105, 159 140), 39, 44, 55, 58, 76( 139), 88, Fawcett, D. W., 261, 301 89, 90, 91, 120, 121, 122, 124, 159, Feldman, J. D., 31, 40, 91, 109, 110, 160, 162, 163 111, 116, 118, 159, 165, 169, 170, 174, 176, 180, 184, 187, 192, 193, Dvorak, H. F., 105, 159 194, 196, 197, 198, 201, 203, 204, Dwyer, E., 169, 213 207, 208,210, 212,214 Dwyer, J. M . , 31, 32( 178), 90 Feldman, M., 21( 141, 142), 71(302), 89, 93, 111, 115, 119, 156, 159, E 176, 177, 205, 207, 208, 211, 225, 244 Eakin, R. M., 185,209 Eanes, E. D., 250, 252, 253, 254, 269, Feldmann, M., 24, 25( 152), 71, 74, 75, 89, 93 271, 301 Felton, L. D., 230, 244, 245 East, J., 23(144), 89 Ferber, E., 150, 162 Eden, A., 31(348), 94 Feryson-Smith, M. A., 216, 237, 245 Edwards, A. W. F., 234, 244 Ferrer, J. F., 170, 178, 208 Ehrenreich, B. A,, 138, 144, 146, 159 Eichwald, E. J., 216, 217, 220, 222, 227, Fialkow, P. J., 238, 245 Fikrig, S., 189, 210 228, 229, 244, 246 Eidinger, D., 64, 66 ( 275 ) , 67, 81, 92, Filler, R. M., 4( 33), 86 Finkel, S . I., 234, 245 93,94, 175,200,207 Ein, D., 252, 254, 273, 275, 278, 301, Finland, M., 40( 204), 91 Fischbach, M., 31(210), 41( 210), 91 302 Fischer, H., 102, 150, 159, 162 Eisen, H. N., 15(107), 88 Fish, A. J., 100, 159 Eklund, C. M., 251, 303 Fisher, N., 236, 246 Ekpaha-Mensah, A., 55, 59, 91 Fisher, S., 107, 159 Eliakim, M., 301 Fishman, M., 96, 97, 144, 145, 147, 157, Elkins, W. L., 177, 207, 220, 244 159 Elliott, E. V., 7(54), 10(54), 87 Fitch, F. W., 12(87), 88, 119, 122, 163, Ellison, E. H., 169, 205 183, 184, 189, 191, 199, 205, 213 Ellman, L., 81( 339, 340), 85(339, 340), Flax, M. H., 168, 173, 186, 187, 188, 94, 175, 176, 183, 207 200, 208, 209 Elson, J., 69( 293), 70( 293), 93 Fleming, W. A., 149, 164 Emeson, E. E., 301 Flexner, S., 167, 208 Engelstein, J. M., 218, 244 Foerster, J., 63( 260), 92 Engers, H. D., 31, 94 Ericsson, L. H., 275, 276, 285, 288, 299 Forbes, B. R., 270, 300 Forbes, J. T., 174, 180, 184, 198, 199, Eriksen, E., 275, 278, 299 Eriksen, N., 252, 254, 265, 267, 275, 208 Ford, C. E., 216, 245 276, 285, 288, 299

AUTHOR INDEX

Ford, W. L., 120, 159, 176, 190, 191, 205, 208 Forni, L., 31( 172), 32( 172), 90 Fortner, J. G., 169, 214 Fox, A. S., 217, 242, 245 Francis, M. W., 100, 160 Frangione, B., 252, 254, 274, 275, 285, 286, 288, 289, 301, 302 Frank, H. A., 219, 246 Frank, M. M., 98, 99, 102, 118, 131, 134, 161, 179, 207 Franklin, E. C., 250, 252, 253, 254, 255, 257, 259, 263, 264, 265, 267, 270, 271, 274, 275, 277, 278, 279, 280, 281, 283, 285, 286, 287, 288, 289, 292, 296, 300, 301, 302, 303, 304 Franzblau, C., 271, 280, 302 Franzl, R. E., 106, 143, 259 Frei, P. C., 105, 159 French, M. E., 171, 183, 184, 191, 196, 207, 208 French, V. I., 102, 149, 159, 164, 165 Frenkel, E. P., 62(248), 63(248), 64 (248), 92, 169, 207 Freund-Molbert, E., 102, 159 Fridnian, E. H., 183, 208 Friedman, E., 296, 301 Friedman, H. M., 79(321), 94 Friedman, H. P., 107, 145, 159, 162, 163, 175, 200, 206 Friedreich, N., 250, 301 Frommel, D., 100, 159 Fudenberg, H. H., 136, 137, 160 Furusawa, M., 220, 245 Fuson, R. B., 227, 228, 229, 246

311

Gary, G. W., 109, 110, 111, 160 Gasser, D. L., 175, 208, 223, 226, 227, 245, 247 Gatti, R. A,, 186, 210 Gavin, J., 237, 238, 245, 247 Gay, F. P., 107, 160 Gelb, A. G., 236, 246 Gelfand, E. W., 135, 157 Gell, P. G . H., 15(98, 99, 101), 31(176), 32(176), 88, 90 Germany, W. W., 172, 179, 180, 207 Gershon, R. K., 27, 28, 47, 67, 68, 73 (290, 305), 81, 90, 93, 94, 168, 208 Gery, I., 148, 164 Gewurz, H., 179, 208 Ghidoni, J. J., 255, 261, 301 Giblett, E. R., 238, 245 Gigli, I., 137, 160 Gill, F. A., 138, 160 Gillespie, G. Y., 170, 199, 208 Gillette, R. W., 113, 160 Gilmour, D. G., 241, 245 Ginsburg, H., 176, 177, 205, 208 Girard, J. P., 111, 115, 159 Gitlin, D., 278, 300 Gittes, R. F., 220, 221, 245 Glassock, R. J., 169, 212 Glenner, G. G., 250, 252, 253, 254, 255, 264, 267, 269, 271, 273, 274, 275, 278, 279, 280, 281, 284, 286, 287, 288, 292, 300, 301,303 Glober, J. S., 98, 160 Globerson, A,, 21( 141, 142), 71( 302), 81, 89, 93, 94, 180, 194, 207, 208 Glynn, A. A., 97,158 Gijtze, O., 202, 208 G Goidl, E. A., lO(66, 67, 68), 12(66), 13(68), 18(66, 67), 19(66), 20 Gabriel, M., 169, 212 (67), 26(66), 31(68), 34(67), 35 Gafni, J., 251, 252, 263, 264, 265, 269, ( 6 7 ) , 38(68), 39(68), 40(68), 43 299, 301,302,303,304 (66), 45(67), 48(225), 49(225), Gafni, M., 80(325), 94 50( 67, 225), 52( 225), 53( 225), 54 Gajl-Peczalska, K. J., 100, 159 (225), 68(67), 76(66), 77(66), 78 Galdiero, F., 169, 205 ( 6 6 ) , 87, 91 Gallagher, R., 170, 214 Golan, D. T., 25(154), 68(154), 70 Gallily, R., 111, 115, 119, 156, 159 (154), 74( 154), 89 Gallo, G., 294, 302 Goldberg, A. F., 270, 301 Garcia, D. A., 141, 159 Goldberg, E. H., 221, 245 Garcia, G., 105, 158 Goldie, J. H., 39( 199), 91 Garvey, J. S., 97, 146, 158, 159, 163

312

AUTHOR INDEX

Goldman, M. B., 32( 175), 90 Goldstein, I. M., 144, 165 Golub, E. S., 105, 160 Gonzales, B., 189, 209 Good, R. A., 3(21), 4(29), 5(36, 37, 38), 86, 100, 159, 179, 186, 189, 208, 210, 213, 217, 220, 227, 228, 229, 230, 245, 246, 247, 297, 303 Goodman, J. W., 145, 146, 147, 160, 163 Gorczynski, R. M., 58, 91, 122, 124, 155, 160 Gordon, J., 126, 160 Gordon, R. O., 200, 208 Gorer, P. A,, 172, 173, 174, 178, 179, 181, 192, 196,197,208 Gorman, J. G., 237, 245 Gotch, F., 176, 208 Gotein, R., 186, 199, 213 Gott, S. M., 15( 112), 88 Gottlieb, A. A., 97, 145, 146, 158, 160, 164 Gowans, J. L., 10(60), 87, 112, 120, 159, 165 Gowland, G., 169, 206 Grabar, P., 169, 205 Grasbeck, R., 63( 256), 92 Graetzer, M. A., 3( 23), 86 Graf, M. W., 104, 160 Graff, R. G., 202, 210 Grant, G., 5(47), 10(47), 86 Greaves, M. F., 2 ( 5 ) , 32( 187, 190), 63, 85, 90, 92 Green, H. N., 173, 178, 208 Green, I., 81(339, 340), 15(116, 121), 16( 123), 63( 260), 85( 339, 340), 88, 92, 94, 96, 147, 157, 158, 175, 176, 183, 207, 208 Greenwood, F. C., 98, 160 Gregory, C. J., 6, 86 Grey, H. M., 31(173, 174), 32(173, 174), 90, 131, 132, 133, 135, 141, 142, 161, 165 Grishman, E., 263, 304 Grob, P. J., 169, 208 Gross, E., 261, 300 Grumbach, F., 129, 158 Grumet, F. C., 26, 55, 56(231), 90, 91 Gueft, B., 255, 261, 267, 271, 301 Guiney, E. J., 179, 208 Gurner, B. W., 31(176), 32(176), 90, 134, 161

Guttmann, R. D., 169, 184, 185, 191, 208, 211, 212

H Haaland, M., 167, 205, 208 Habicht, G. S., 10(62), 11(62), SS(62, 292), 69( 292), 70( 292), 74( 292), 87, 93, 117,158 Hammerling, U., 173, 203, 205 Haferkamp, O., 303 Halasz, N. A., 169, 193, 208 Hall, E. J., 238, 247 Hall-Allen, R. T. J., 169, 173, 184, 185, 208, 214 Halpern, B., 200, 211 Halpern, B. N., 129, 157, 158, 160 Hamaoka, T., 10(69), 16(69), 25, 87, 90, 119, 153, 160 Hamburger, J., 169, 210 Hamilton, J., 16( 127), 45( 220), 46 (127), 47(220), 89, 91 Hammons, A. S., 100, 162 Hamper, J., 236, 246 Han, I. H., 79( 321), 94 Hanna, M. G., Jr., 98, 100, 160 Hannover Larsen, J., 225, 245 Harada, M., 250, 252, 253, 254, 271, 278, 286,301, 302 Harbaugh, J., 286, 301 Hardin, C. A,, 169, 214 Harding, B., 176, 208, 211 Hardisty, R. M., 237, 247 Hardy, M. A,, 169,208 Harington, J. S., 149, 157 Harris, E. M., 185, 209 Harris, G., 118, 127, 141, 160 Harris, J. E., 126, 127, 160 Hartmann, K.-U., 13(91, 94), 14, 29 (91, 94), 56, 75(91), 76(94), 88, 121, 160 Hartwig, M., 252, 302 Hakk, M., 171, 184, 209, 222, 245 Haskill, J. S., 39(200), 57, 58, 91, 121, 160 Hagkov6, V., 180, 181, 183, 194, 197, 209, 222, 245 H a s , G., 271, 301 Hathaway, C. O., 193, 206 Hattler, B., 200, 211 Haughton, G., 179, 194, 196, 209

313

AUTHOR INDEX

Hauschka, T. S., 216, 223, 245 Havas, H. F., 25(155), 68(155), 70 (155), 71, 74(155), 89 Hayes, L. L., 10(76), 87 Hayry, P., 111, 165 Heefner, W. A., 261, 301 Heilman, D. H., 148, 149, 160, 162 Heise, E. R., 110, 160 Heller, E., 220, 247 Heller, H., 251, 252, 263, 264, 265, 299, 301, 302, 303, 304 Hellman, K., 194, 209 Hellstrom, I., 85(353), 94, 168, 173, 178, 180, 182, 183, 184, 189, 190, 191, 193, 199, 201, 202, 205, 209, 212, 213, 214, 235, 236, 245 Hellstroni, K. E., 85( 353), 94, 168, 173 178, 180, 182, 183, 184, 189, 190, 191, 193, 199, 201, 202, 205, 209, 212, 213, 214, 235, 236, 245 Hemmingsen, H., 121, 164 Henney, C. S., 176, 190, 203, 209 Henry, C., 199, 209 Henson, P. M., 136, 144, 160 Heppner, C. H., 182, 201, 209 Herberman, R. B., 183, 207 Herd, Z. L., 100, 132, 160 Herman, J. H., 297, 302 Hermodson, M. A., 275, 276, 285, 288, 299 Herod, L., 189, 210 Herscowitz, H. B., 145, 160 Hersh, E. M., 126, 127, 160, 163 Hersko, C., 265, 301 Hersko, H., 264, 299 Herzenberg, L. A., 26( 164), 27( 164), 43(215), 76(215), 90, 91, 189, 209, 211 Heslop, B. F., 169, 174, 184, 194, 201, 209, 238, 239, 245, 247 Hess, O., 215, 245 Heystek, G. A., 184, 185, 191, 211 Hildemann, W. H., 169, 173, 174, 175, 179, 180, 183, 184, 185, 191, 198, 199, 201, 206, 209, 212, 213, 214, 220, 226, 240, 245, 247 Hilgard, H. R., 81, 94 Hillemand, B., 129, 158 Hirose, F., 169, 208 Hirsch, J. G., 138, 161 Hirsch, M. S., 109, 110, 111, 160

Hirst, J. A., 13, 21(139), 31( 139), 38 (90, 139), 39(139), 55, 58(239), 76( 139), 88, 89, 91 Hirvonen, T., 234, 247 Hjort, G. H., 255, 302 Hodge, B. A., 238, 239,244 Hoene, R.E., 235, 247 Hoffmann, M., 21(139), 31(139), 38 (139), 39( 139), 58(239), 76( 139), 89, 91, 109, 110, 111, 160 Holdridge, B. A., 223, 245 Hollander, N., 176, 208 Holm, G., 176, 177, 190, 203, 212 Holt, L. J., 63( 258), 92 Holub, M., 102, 159 Horne, R. W., 255, 300 Hornung, M. O., 200, 209 Horowitz, R. E., 284, 302 Norwitz, C., 239, 247 Hoshino, K., 221, 245 Hoste, J., 175, 205, 223, 226, 243 House, W., 5(47), 10(47), 86 Howard, J. G., 23(146), 69(293), 70 (293), 89, 93, 112, 129, 135, 152, 155, 158, 160, 165, 180, 186, 205 Howard, R. J., 175, 200, 209 Hu, C. C . , 168, 170, 172, 173, 186, 187, 188, 192, 207 Huber, H., 135, 136, 137, 160 Hughes, L. A., 109, 161 Hughes, L. E., 183, 209 Hull, P., 235, 245 Hultgren, M. K., 296, 302 Humphrey, J. H., 23, 31, 32( 170), 69 (170), 74(312), 89, 90, 93, 97, 98, 99, 102, 103, 108, 118, 131, 134, 137, 144, 149, 152, 153, 155, 157, 160, 161, 162 Hunter, P., 13, 39, 88 Hunter, R. L., 98, 100, 160, 169, 205 Hunter, W. M., 98, 160 Hutchin, P., 172, 179, 180, 188, 192, 193, 194, 205, 209

I Inchley, C. J., 112, 135, 161 Inderbitzin, T. M., 169, 208 Inman, J. K., 175, 208 Irvin, G. L., 111, 170, 171, 172, 180, 209 Isa, A. M., 110, 161

314

AUTHOR INDEX

Isakovic, K., 68, 93 Iseri, 0. A,, 275, 278, 299 Isersky, C., 250, 252, 253, 254, 271, 278, 286, 301 Ishidate, M., 31, 41(207), 91 Ishizuka, M., 79( 320), 80( 325, 326, 327, 328, 329), 93, 94 IvBnyi, J., 170, 171, 172, 173, 182, 206 Iverson, G. M., 36, 90

Jones, H. B., 129, 159 Jones, J. M., 174, 184, 192, 193, 194, 196, 197, 201, 204, 210 Jones, K. W., 216, 245 Jones, T. C., 138, 161 Jones, V., 177, 210 Jose, D. G., 186, 210 Judd, K. P., 220, 245

J

K

Jackson, C. E., 251, 299 Jacobs, P. A., 215,217,245 Jacobson, E. B., 43, 76,91 Jacoby, F., 129, 161 James, D. A., 234, 245 James, K., 63( 250, 5 9 ) , 92 Jandl, J. H., 134, 135, 157, 162 Janeway, C. A., Jr., 4(33), 31( 176), 32 (176), 36, 86, 90, 99, 144, 152, 161 Janigan, D. T., 295, 296, 299, 300, 301, 302 Janis, M., 128, 157 Jankovi6, B. D., 4(27, 28), 86, 173, 186, 188, 209 Jansen, F. K., 170, 171, 214 Jaraskova, L., 147, 158 Jaroskova, L., 140, 157 Jaroslow, B. M., 99, 161 Jeekel, J. J., 169, 174, 184, 201, 210 Jehn, U. W., 76, 93, 126, 161, 173, 183, 210 Jenkin, C. R., 107, 130, 160, 161 Jenkins, A. M., 185, 210 Jennings, J. F., 109, 161 Jerne, N. K., 10(64), 18(64), 76(64), 87, 199, 209 Jerusalem, C., 188, 210 Jimenez, L., 57, 91, 147, 158, 203, 210 Jobling, J. W., 167, 208 Johnson, A. G., 63(255), 79, 92, 94, 147, 164 Johnson, H. G., 79( 321), 94 Johnson, J. S., 66, 92 Johnson, M. C., 15(112), 88, 200, 211 Johnson, M. L., 151, 159 Jolley, W. B., 194, 197, 210 Jonik, J., 171, 184, 209 Jonas, W. E., 134, 161

Kahan, B. D., 172, 203,212 Kaliss, N., 168, 169, 172, 173, 174, 178, 179, 180, 181, 182, 186, 189, 192, 193, 196, 198, 200, 202, 208, 210, 214 Kallman, K. D., 239, 245 Kamoun, P. P., 169, 210 Kandutsch, A. A., 168, 173, 210 Kantor, F. S., 3(25), 86 Kaplan, M. H., 98, 161 Kappler, J. W., 58(239), 91 Karakoz, I., 171, 184, 209 Karlin, L., 76, 93 Karnovsky, M. J., 61( 245), 92, 144, 161 Karnovsky, M. L., 144, 161 Kass, E. H., 40( 204), 91 Kasukawa, R., 254, 277, 284, 302, 303 Katsh, G. F., 245 Katsh, S., 245 Katz, D. H., lO(66, 67, 68), 12(66), 13, 18(66, 67), 19(66), 20( 67), 21, 22, 25( 153), 26( 66), 31, 34(67), 38, 39, 40( 68), 43, 45, 48(225, 226), 49(225), 50, 51(153), 52 ( 153, 225, 226), 53, 54( 153, 225, 226), 55( 153, 228, 230), 56( 153), 59(228), 60(228), 61, 63, 66(279), 67( 279), 68( 67, 153), 69( 153, 296), 70(153), 71, 72(153, 296), 74( 153), 75( 153), 76( 66, 137), 77, 78( 66), 81(339, 340), 84( 153, 352), 85(339, 340), 87, 89, 91, 92, 93, 94, 119, 152, 161 Kauffman, G., 230, 245 Kaufmann, B. P., 216,245 Keiser, H. R., 250, 252, 253, 254, 271, 301 KekulB, A., 250, 301

315

AUTHOR INDEX

Keller, H. U., 31, 32(170), 69(170), 90, 137, 153, 161 Keller, R., 177, 210 Kelly, W. D., 22.0, 227, 228, 245 Kelman, H., 227,228, 245 Kelus, A. S., 31(176), 32(176), 90 Kennedy, J. C., 6, 10(48), 55, 58, 59, 76, 86, 91 Kerbel, R. S., 64, 67, 92, 93 Kersey, J. H., 5(37), 86, 186, 210 Kessel, R. W. I., 149, 161 Kettman, J. R., 21, 31(139, 140), 32 (191), 38(139, 140), 39, 44, 58 (239), 76(139, 140), 89, 90, 91 Khan, S. A., 66(275), 92, 175, 200, 207 Kikkawa, Y.,267, 271, 301 Kilham, L. K., 197, 199, 213 Kim, I. C., 271, 280, 284, 302 Kim, J. H., 194,212 Kimura, S., 275, 301 Kind, P., 79, 94 King, L. S., 302 Kinsby, R. G., 69(293), 70(293), 93 Kinsky, R., 168, 169, 170, 171, 184, 185, 186, 191, 213, 214 Kirby, D. R. S., 235, 244, 246 Kiryabwire, J. W. M., 183, 207 Kitagawa, M., 10(69), l6(69), 87, 119, 153, 160 Klaue, P., 194, 197, 210 Klein, E., 180, 196, 210, 225, 226, 246 Klein, J., 221, 246 Klein, W. J., Jr., 180, 186, 194, 196, 197, 198, 199, 202, 205, 210 Kleinnian, R., 148, 164 Klinman, N. R., 73, 93 Kochem, H. G., 254, 278, 304 Kodama, H., 193, 210 KoldovskL, P., 170, 171, 172, 173, 182, 206 Koller, P. C., 5( 45, 46 ), 10(45, 46), 86 Kolsch, E., 119, 143, 161 Kondo, K., 67, 68, 73 ( 290, 305 ) , 93, 168, 208 Kontiainen, S., 16( lu), 18, 78( 124), 88 Kossard, S., 135, 161 Kotani, M., 220, 245 Koucky, R. F., 251,303

Kourilsky, F. M., 173, 183, 194, 203, 208, 210 Kozelka, A. W., 241, 246 Krantz, A. R., 151, 159 Kremer, W. B., 168, 172, 173, 189, 199, 206 Kretschmer, R., 295, 302 Kripke, M. L., 175, 199, 200, 206 Krohn, P. L., 185, 210 Kronman, B., 169, 211 Krusman, W. F., 102, 159 Kunin, S., 21, 89 Kunkel, H. G., 254, 278, 302 Kuroiwa, Y., 251, 299

L Lachmann, P. J., 254, 278, 302 L’Age-Stehr, J., 43( 215), 76( 215), 79 (324), 81(324), 91, 94 Lagunoff, D., 275, 278, 299 Lahti, A,, 111, 165 Lajtha, L. G., 6, 86 Lambert, P. B., 219, 246 Lamelin, J. P., 63(260), 92 Lampkin, G. H., 222,244 Lamvik, J. O., 126, 161 Lance, E. M., 108, 110, 113, 153, 160, 161, 165, 194, 210 Landes, J., 169, 213 Landsteiner, K., 2, 14, 85 Lane, F. C., 115, 144, 151, 161 Lang, P. G., 99, 102, 157, 161 Lang, W., 199, 210 Langston, R. R., 168, 172, 206 Lanman, J. T., 189,210 Lapp, W. S., 169, 172, 197, 201, 210 Lapp&, M. A., 175, 199, 200, 202, 206, 210, 235, 246 Laufer, A., 261, 302 LaVia, M. F., 148, 162 Law, D. H., 254, 304 Lawless, 0. J., 250, 280, 285, 300 Lay, W. H., 135, 136, 161 Layne, E., 200, 209 Leduc, E. H., 102, 161 Lee, A., 130, 161 Lee, G. R., 237, 246 Lee, S., 174, 208 Lehner, T., 296, 302

316

AUTHOR INDEX

Lehtovaara, R., 234, 247 Leigheb, G., 173, 174, 192, 206 Leikola, J., 203, 210 LengerovL, A., 176, 210, 211, 230, 234, 246 Lennox, E. S., 21(138), 58(236), 76 (138, 236), 89, 91 Leonard, M. R., 130, 133, 163 Leserman, L. D., 123, 161 Leskowitz, S., 108, 149, 161, 164 Lesley, J . F., 21(139), 31(139), 32 (191), 38( 139), 39( 139), 58(239), 76(139), 89, 90, 91 Letterer, E., 295, 302 Leuchars, E., 5(40, 41, 44, 45, 46), 7 (54), lO(44, 45, 46, 54), 23( 145, 146), 68( 286), 86, 87, 89, 93 Levaditi, J., 129, 158 Levanon, D., 149, 165' Leventhal, B. G., 126, 163 Levillain, R., 108, 158 Levin, M., 252, 254, 274, 275, 278, 285, 286, 288, 289, 301, 302 Levine, B. B., 15(100, 108, 109, 115), 18(133), 88, 89, 136, 163, 175, 213, 226, 247 Levine, M. A., 31, 41, 91 Levis, W. R., 161 Levy, J . P., 173, 194, 203, 210 Lewis, G. K., 174, 180, 184, 198, 199, 208 Lewis, L. A., 107, 162 Liacopoulos, M., 200, 21 I Lieberman, G., 62( 248), 63( 248), 64 (248), 92 Lieberman, R., 43(216), 91, 175, 206 Lilly, F., 234, 245 Lind, I., 133, 163 Linder, 0. E . A,, 174, 183, 184, 185, 211, 225, 226, 246 Lindquist, R. R., 169, 184, 185, 191, 208, 211, 212 Lindsey, E. S., 169, 199, 212 Ling, N. R., 83(258), 92 Linke, R. P., 274, 278, 279, 302 Linscott, W. D., 136, 137, 160, 196, 202, 203, 21 1 Lischner, H. W., 4( 31, 34), 86 Lisker, R., 238, 245 Little, J. B., 141, 159

Little, J. R., 15(113), 88 Little, M., 21(143), 76( 143), 89 LoBuglio, A. F., 134, 157, 162 Loewi, G., 31, 40, 91, 176, 211 Lonai, P., 176, 211 Loutit, J. F., 3( 13), 85 Lovingwood, C. G., 169, 205 Lowenstein, J., 294, 302 Lucas, Z. J., 183, 184, 185, 193, 194, 196, 201, 211, 212 Luft, J . H., 255, 302 Luke, P., 169, 173, 184, 185, 214 Lummus, Z., 100, 101, 163 Lurie, M., 6, 86 Lustgraaf, E. C., 217, 227, 228, 229, 244, 246 Lyon, M. F., 237,240, 246 Lytton, B., 183, 209

M McBride, R. A., 15(119), 16, 18(119), 81, 88, 94 McCarthy, M. M., 21( 139), 31( 139), 38(139), 39( 139), 76( 139), 89, 121, 122, 124, 159, 160 McConahey, P. J., 63(252), 92, 98, 104, 162, 188, 199, 207, 211 McCullagh, P. J., 75, 84(313), 93, 12.0, 159, 162 McCulloch, E . A., 6(48), 10(48), 86 McDevitt, H. O., 26( 162), 83( 162, 351), 90, 94, 98, 119, 146, 157, 162, 163, 175, 176, 211, 226, 246 MacDiarmid, W. D., 237,246 McDonald, J. C., 200, 209 McEntee, P. F., 170, 213 McFarland, W., 126, 127, 148, 162, 163 McGregor, D. D., 10(60), 87 Mach, B., 111, 115, 1S9 McIntyre, J. A., 148, 162 Mackaness, G. B., 96, 130, 162, 177, 211 Mackay, I. R., 31, 32(178), 90 McKenzie, I. F. C., 174, 184, 192, 197, 198, 201, 210, 211 McKeogh, R. P., 189, 205 McKusick, V. A., 251, 299 McLaren, A., 222, 234, 246 MacLennan, I. C. M., 176, 208, 211 McWhirter, K. G., 235, 246

AUTHOR INDEX

Miikela, O., 2 ( 4 ) , 25, 32( 169), 59, 61, 85, 90, 92, 234, 247 Mage, R. G., 32( 175), 90 Maggs, P. R., 169, 173, 184, 185, 208, 214 Magnus Levi, A., 277, 302 Maillard, J., 168, 185, 186, 191, 214 Main, J. M., 222, 247 Makinodan, T., 106, 163 Mancini, G., 279, 300 Mandel, M. A., 53(227), 91, 169, 211 Mandel, T., 31( 349), 94 Mandema, E., 250, 253, 264, 271, 295, 302, 303 Mann, J. D., 236, 246 Marbrook, J., 12, 39( 200), 57( 200), 58 (200), 88, 91, 120, 121, 160, 162 Marchant, D. J., 189, 205 Marchant, R., 7 ( 54), 10(54), 87 Marchioro, T. L., 184, 212 Marchiso, M. A., 149, 161 Mariani, T., 217, 227, 228, 246 Markley, J., 183, 184, 185, 201, 211 Markowski, B., 4( 33), 86 Marquet, R. H., 184, 185, 191, 211 Martin, W. J., 16(127), 45(220), 46 (127), 47(220), 63(251), 89, 91, 92, 169, 175, 176, 183, 207, 211 Martinez, C., 4 (29), 5( 36, 37, 38), 86, 189, 213, 220, 227, 228, 229, 230, 245, 246, 247 Martinozzi, M., 13(95), 88 Mason, S., 32( 188), 33( 188), 90, 179, 211 Matalon, R., 252, 265, 286, 288, 303 Matousek, V., 176, 210 Matsumoto, T., 67( 283), 80(329), 93, 94 Matthes, M. L., 102, 159 Mauel, J., 170, 171, 174, 190, 197, 199, 206, 211 Mauersberger, D., 179, 214 Maurer, P. H., 15( 114), 88 Mautner, W., 263, 304 Mawas, C., 295, 302 Mayer, D. J., 169, 211 Medawar, P. B., 168, 169, 172, 192, 205, 206, 211, 221, 222, 244 Medlin, J., 99, 152, 155, 162 Mellors, R. C . , 254, 277, 278, 302

317

Melnick, H. D., 107, 162 Mergenhagen, S. E., 151, 162, 163, 175, 200, 209 Merker, H. J., 263, 303 Merrill, J. P., 169, 208, 211 Merritt, K., 63( 255), 92 Metcalf, D., 31, 41( 207), 91 Metchnikoff, E., 108, 162 Meuwissen, H. J., 100, 159 Meyer, G. F., 215, 245 Meyer, R. K., 3(23), 86 Michael, G., 24( 148), 89 Michaelis, L., 65, 92 Michie, D., 222, 246 Micklem, H. S., 296, 300 Miescher, P. A., 136, 164 Milgrom, F., 170, 211, 254, 277, 278, 284, 302, 303 Millar, K. G., 66(275), 92, 175, 200, 207 Miller, H. I., 271, 302 Miller, H. R. P., 147, 162 Miller, J., 200, 211 Miller, J. F. A. P., 2 ( 9 ) , 3(24), 4(26), 5, 7, 8(57, 58, 59), 9, 10, 11, 13, 14, 16(74, 127), 21( 143), 24(74), 26, 29, 31(186, 349), 32(186), 33 (186), 37, 39, 42, 43(217), 44( 74), 45, 46, 47, 63(251), 68, 73(74), 75( 56, 74), 76( 143), 77, 78, 85, 86, 87, 89, 90, 91, 92, 93, 94, 151, 162, 169, 211 Miller, L., 239, 246 Miller, M. E., 137, 162 Miller, 0. J., 216, 245 Miller, R. G., 58(238), 91, 122, 124, 155, 160 Millonig, G., 255, 302 Milstein, C., 171, 202, 207, 288, 302 Mims, C. A., 129, 162 Mintz, B., 240, 246 Mishell, R. I., 12, 13(92), 14(92), 20 (83), 21( 139), 26( 164), 27( 164), 29(92), 31( 139), 38( 139), 39 (139), 72, 75(92), 76(139), 88, 89, 90, 93, 120, 121, 122, 124, 159, 160, 162, 189, 211 Missmahl, H. P., 252, 302 Mitchell, G. F., 2 ( 9 ) , 3(9), 7, 8(57, 58, 59), 9, 10, 11, 13, 14, 29, 39, 42,

318

AUTHOR INDEX

Muller-Ruchholtz, W., 110,162 Mumaw, W . R., 147,164 Munder, P. G., 150, 162 Munro, A., 13,39,88 Munro, J. A., 149,164 Murphy, F. A., 109,110,111, 160 (65),31(345), 36,42, 43,44,59 Murray, J. A., 167,205 (240),67,71(301),76(63),77, 78 Murray, J . E., 174,212 ( 65, 70), 87, 88, 90, 92, 93, 94, Murray, R. M., 174,212 104, 105, 111, 113,115, 116, 117, Musher, D. M., 126,161 118, 119, 134, 143,153, 155, 158, 159, 161, 162, 172, 173, 180, 199, N 211 Najarian, J. S., 187,189,207, 212 Mittwoch, U., 216,245 Nakamura, R. M., 188,214 Miyozawa, J., 194,206 Nakano, M., 147,158 Mocarelli, P., 296,300 Naor, D., 31, 32(167), 69(294), 72,90, Modelell, M., 150,162 93, 137,153,162 Moller, E., 31,32(190),40,90, 91, 172, 173, 174,178, 181, 192, 194,195, Nase, M., 31(345),94 Nase, S., 10(64), 18(64), 76(64),87, 201,203,211 199,210 Moller, G., 24(148), 25(157), 31, 32 (190),40,65,67(283),89, 90, 91, Nash, D. R., 179,194,196,209 92, 93, 103,158, 172,173,174,178, Nathanson, L.,173,183,210 179,180, 181, 182,183, 185,192, Nathanson, N., 175,200,214 193,194, 195, 197, 199, 201,202, Naylor, A. F., 234,247 Nelken, D., 170,173,184,192,207, 212 203,210, 211 Nelson, D. S., 134, 135,137,161, 162, Molomut, N., 168, 210 170,212 Monaco, L., 149,161 Nelson, R. A., Jr., 137,160 Moore, J. E., 221,245 Nemec, M., 171,184, 209 Moore, M. A. S., 3(14),85 Nettesheim, P., 100, 162 Moore, R. D., 147,164 Neven, T., 200,211 Moorhead, J . F., 148,162 Nevo, Z., 252,265, 286,288, 303 Morelli, R., 137,162 Newconhe, D. S., 271,303 Morrello, J . A., 106,159 Newlin, C., 107,163 Morris, R., 193,194,196,212 Newton, M. S., 237,244 Moseley, R. V., 174,212 Mosier, D.E., 12, 88, 118, 121, 123, Niblack, G. D., 148,162 147,162 Nicolai, M. G., 173,194,203,210 I Mouton, D., 129,158, 160 Nillson, U.R., 137,162 Mowry, 8. W., 286,302 Nisbet, N. W., 81(342), 94, 238, 239, Muckle, T . J., 254, 277, 279,284,285, 247 295,302, 303 Nissim, J., 251,299 Miiller, B., 15(118),88 Niv, A., 301 Miiller, M., 172,180,212 Miiller-Eberhard, H . J., 136, 137, 160, Noades, J., 237,245 Nordin, A. A., 172,176, 180,190,203, 202,208, 254,278,302 207, 209, 212 Muir, H., 286,300, 303 Mullen, Y.,173,174,183,184,185,212, Nordman, C., 63(257),92 Nossal, G. J. V., 7(59), 8, 10(59), 87, 214 98,99,100, 101,102,157, 161, 162, Muller, H . J., 215,246 163 Muller-Hermelink, H . K., 110,162

68,75(56, 92),85, 87, 88, 90, 93 Mitchell, G. W., Jr., 189,205 Mitchell, J., 99,100,101,102,162, 163 Mitchell, R. M., 174,212 Mitchison, N. A., lO(63,65,70,71,72), 12,15(106), 16,18,20,26( 6 5 ) , 27

AUTHOR INDEX

Notkins, A. L., 151, 162, 163, 175, 200, 209 Nouza, K., 171, 184, 209 Nussenzweig, V., 31, 94, 135, 136, 161 Nylen, M. U., 253, 254, 267, 284, 300

0 Ochiai, T., 64( 264), 92 Ockner, S. A., 184, 185, 191, 212 Odenwald, M. V., 79(324), 81(324), 94 Oettgen, H. F., 57(235a), 91, 203, 210 Ohanian, S. H., 156, 164 Ohms, J. I., 286, 301 Ohno, S., 216, 246 Ojeda, A,, 15(114, 115), 88 Okumura, K., 18(129), 64, 89, 92, 198, 199, 213 Old, L. J,, 172, 173, 178, 179, 194, 200, 202, 203, 205, 206, 207, 212 Oldstone, M. B. A., 188, 212 Oliver, C . B., 169, 212 Ono, K., 169, 199, 212 Oort, J., 102, 163 Oppenheim, J . J., 116, 118, 126, 127, 163, 164, 200, 208 Ordal, J. C . , 55, 56(231), 91 O’Riordan, M. L., 237,244 Orloff, M. J., 169, 193, 208 Orr, W. McN., 169, 212 Ortega, L. G., 254, 277, 278, 302 Osborne, D. P., Jr., 55(230), 91 Osoba, D., 3( 24), 5, 39( 198, 199), 86, 91 Osserman, E. F., 257, 277, 292, 303 O’Toole, C. M., 66, 92 Ottinger, B., 230, 245 Ounsted, C., 234, 246 Ounsted, M., 234, 246 Ourth, D. D., 141, 159 Ovary, Z., 3(25), 15(104, 105), 16, 18 ( 130), 86, 88, 89 Owen, J . J. T., 3( 14), 85 Owen, R. D., 240,247

P PackalBn, T., 176, 212 Page, D. L., 250, 252, 253, 254, 271, 278, 286, 301, 302 Palm, J., 189, 206

319

Palmer, J., 121, 123, 124, 125, 155, 156, 159, 164 Panijel, J., 108, 110, 111, 158, 163 Papermaster, B. W., 4(29), 5(36, 37), 6(53), 86, 87 Parker, C. W., 15( 112), 88 Parker, S. J . , 196, 197, 213 Parkes, A. S., 168, 174, 183, 184, 185, 205, 212 Parkhouse, R. M. E., 127, 163 Parks, E., 109, 158 Paronetto, F., 149, 163, 254, 278, 302 Parrott, D. M. V., 23( 144), 89 Pasanen, V. J . , 59, 61, 92, 203, 210 Passmore, H. C . , 240, 241, 246 Paterson, P. Y., 173, 186, 188, 199, 212 Pattison, P. H., 10(71), 87 Paul, W. E., 2 ( 8 ) , lO(66, 67, 68), 12 (66), 13(68), 15(103, 111, 116, 121, 122), 16(123), 18(66, 67), 19(66), 20, 25( 153), 26(66), 27, 28, 31, 32(8, 175, 181, 182, 187a), 34, 35(67, 187a), 36, 38(68), 39 (68), 40( 68), 43( 66), 45( 67, 219), 47, 48( 153, 225, 226), 49 (225), 50(67, 153, 225), 51(153), 52( 153, 225, 226), 53( 225, 226), 54(153, 225, 226), 55(153), 56 (153), 66(279), 67(279), 68, 69 (153, 296), 70(153), 71(153), 72 (153, 296), 74(153), 75(153), 76 (66), 7 7 ( 6 6 ) , 78(8, 66, 103), 79, 81(339, 340, 341), 84(153, 352), 85, 87, 88, 89, 90, 91, 93, 94, 278, 300, 303 Payne, R., 189, 212 Pearce, J. H., 15(117), 71( 117), 88 Pearsall, N. N., 194, 195, 212 Pearson, M. N., 143, 163 Peltre, G., 15(118), 88 Pembrey, M. E., 238, 247 Penrose, L. S . , 216, 245 Pequignot, G., 129, 160 Percival, A., 130, 158 Perkins, E. H., 97, 106, 130, 133, 155, 163 Perkins, W. D., 61(245), 92 Perlmann, H., 176, 212 Perlmann, P., 176, 177, 190, 203, 212 Pernis, B., 31, 32(172), 90, 149, 163 Peter, H.-H., 193, 198, 203, 204, 212

320

AUTHOR INDEX

Peters, L. C., 100,160 Q Peterson, R. D. A,, 3(21),86 Quadracci, L. J., 184,212 Pettirossi, O., 191,205 Quagliata, F., 136,163 Phil, D., 296,300 Quint, J., 169,208 Phillips, M.E., 178,179,212 Phillips, R. A., 58(238),91, 122, 124,

155,160

Phillips-Quagliata, J. M.,136,163 Pick, A. I.,277,303 Pick, E., 174,208 Pierce, C. W., 21(137), 22( 137),61,76

R

Rabellino, E., 31,32(173,174), 90 Rabinovitch, M.,128,130,138, 163 Rabson, A. S., 169,205 Race, R. R., 236,237,238,245,246,247 (137),89, 92, 121, 122, 123, 148, Rachmilewitz, M.,301 163,199,212 Radovich, J., 65,92, 121,164 Pierce, G . E., 182,201,209 Raff, M.C., 16(126), 17, 18(126), 31, Pierce, J. C., 4(29),86 32(171), 37(196), 61(244), 76 Pierpaoli, W., 295,300,303 ( 126),82(344),89,90,92,94 Pink, R., 288,302 Raffel, S., 143,163 Pinkerton, W., 220,240,245 Raidt, D. J., 21(139), 31(139), 38 Pitt-Rivers, R., 71(301),93 (139),39(139),76(139), 89, 121, Playfair, J. H.L., 6(53),68,87,93 122,124.159 Plotkin, D. H., 109,110,157 Rajewsky, K., lO(64, 65, 72), 12(65), Poels, L.,188,210 15(118,120),16(65, 120), 18(64, Pokorni, Z.,168,173,207 65,la), u)(65), 26(134), 27(65, PolA6kova, M., 218,219,221,233,246,

247

Polley, M. J., 136,137,160 Porter, K.R., 261,300 Porter, R. J., 74(310),93 Porter, R. R., 21(138),76(138),89 Potter, M.,43(216),91 Pras, M.,250,251,252,253,254,255,

264, 265,267,271,275,278,280, 281,283,285,286,287,288,289, 301,302,303,304 Prater, T. F. K., 148,162 Pratt, K.L., 21,89 Prehn, R. T.,173, 181,212, 222,230, 231,246,247 Prescott, B., 63(262),92,230,245 Pressman, D., 169,189,214 Pribnow, J. F., 115, 119,163 Prioleau, W. H.,Jr., 172,179,180,188, 192, 193, 194, 205, 209 Priore, R. L., ll(79, 80), 12,26(82), 27(82), 87 Puditin, D. J., 176,208 Punnett, H.H., 4(31), 86 Pye, J., 31(186, 349), 32(186), 33 (Be),90,94,99,102,163

134), 36(65),42(65), 43(65), 76 (64),77(65, 134), 78(65, 134), 87,88, 89, 153,158, 199,210 Ram, J. S . , 280,284,303 Ranlov, P., 259,296,299,303 Rapp, F.,202,206 Ram, H.,133, 163 Rathbun, W. E.,175,212,226, 247 Ray, H.E.M., 4(33),86 Reade, P. P., 107,161 Reckel, R. P., 241,242,247 Reichert, A. E., 227,247 Reid, B. L.,63(259),92 Reif, A. E.,31,94,176,194,212 Reimann, H. A., 251,303 Reis, R. H., 235,247 Reisfeld, R. A., 172,203,212 Reithmiiller, G., 32(189),90 Renkonen, K. O., 234,247 Reshef, T.,254,275, 285,303 Rhodes, J. M.,133, 145,157, 163 Rich, R., 61,92 Richardson, W.P., 173, 186, 188, 199, 212 Richter, G., 294,303 Rider, M.,216,245

321

AUTHOR INDEX

Rieber, E.-P., 32( 189), 90 Rimon, A., 250, 252, 253, 255, 264, 265, 267, 270, 271, 280, 286, 287, 299, 303 Rist, N., 129, 158 Rittenberg, M. B., 18(132), 21, 71, 74, 89, 93 Rizzo, A. A., 151, 163 Robbins, J. H., 161 Robert, C. B., 149, 160 Rodey, G. E., 297, 303 Roelants, G . E., 16( 125), 31, 33, 89, 145, 146, 163 Roitt, I. M., 32( 187), 63(261), 90, 92 Romussi, M., 295, 300 Roseman, J. M., 13(89), 38(89), 88, 119, 122, 123, 161, 163 Rosen, F. S., 4(33), 86, 135, 157 Rosenau, W., 239, 247 Rosenberg, L. T., 137, 162, 179, 206 Rosenmann, E., 301 Rosenthal, A. S., 32(182), 90 Roser, B., 112, 163 Ross, G. L., 168, 172, 206 Rother, K., 179, 212, 214 Rother, U., 179, 212, 214 Rotman, J., 252, 265, 286, 288, 303 Rotman, Y., 271, 302 Rottlander, E., 15( 118, 120), 16( 120), 88 Rouques, R. J., 21( 137), 22( 137), 76 (137), 89 Rowland, G., 16(127), 45(220), 46 (127), 47(220), 89, 91 Rowley, D. A., 12(87), 88, 118, 119, 122, 134, 157, 162, 163, 168, 173, 186, 187, 188, 191, 199, 205, 213 Rubenstein, P., 180, 210 Rubin, A. S., 56, 91 Ruddle, N. H., 177, 213 Rudolf, H., 174, 190, 211 Ruinen, L., 255, 264, 267, 271, 295, 300, 302, 303 Russell, A., 281, 303 Russell, L. B., 217, 238, 247 Russell, P., 112, 121, 122, 163, 164 Russell, P. S., 179, 208, 220, 221, 245 Ruizkiewicz, M., 169, 173, 179, 200, 213 Rychlikova, M., 168, 173, 186, 188, 207 Ryder, R. J. W., 97, 164, 197, 199, 213

S Sabbadini, E., 177, 213 Sabet, T. W., 107, 163 Sabin, F. R., 95, 163 Sachin, I. N., 193, 210 Sachs, L., 220, 247 Safford, J. W., Jr., 186, 198, 213 Saha, A., 146, 163 Saitoh, T., 183; 184, 199, 213 Salvin, S. B., 15(97), 18(97, 128, 131), 88, 89 Sanford, B. H., 219,247 Sanger, R., 236, 237, 238, 244, 245, 246, 247 Save], H., 183, 213 Schalk, J., 235, 246 ScharE, M. D., 292,300 Scharli, A. F., 169, 173, 208 Schechter, B., 153, 164 Schechter, G. P., 126, 127, 163 Schechter, I., 98, 153, 161, 162, 264 Scheid, M., 221, 245 Scheinberg, S. L., 241, 242, 247 Schierman, L. W., 15( 119), 16, 18 (119), 88 Schiff, R. I., 186, 206 Schimpl, A., 13, 14, 55, 88 Schindler, R., 190, 199, 206 Schirrmacher, V., 10(64), 18(64, 134), 26( 134), 27( l a ) , 76( 64), 77 (134), 78( 134), 87, 89 Schlesinger, M., 169, 186, 199, 213 Schlossman, S., 129, 158 Schmidtke, J. R., 31( 174), 32( 174), 79 (321), 90, 94, 105, 115, 119, 131, 132, 133, 141, 142, 144, 148, 151, 164, 165 Schoenberg, M. D., 147, 164 Scholten, J. H., 250, 253, 255, 264, 267, 271, 295, 300, 302, 303 Schubert, M., 250, 252, 253, 255, 264, 265, 267, 270, 286, 288, 303 Schultz, G., 79( 3%), 81( 324), 94 Schultz, J., 215, 247 Schultz, R. F., 278, 298, 303, 304 Schulz, R. Z., 271, 301 Schwab, J. H., 156, 164 Schwartz, H. J., 108, 164 Schwartz, P., 298, 303

322

AUTHOR INDEX

Schwartz, R. S., 63(254, 256), 92, 97, 164, 173, 183, 197, 199, 210, 213 Scott, G. M., 167, 207 Scott, J. C., 286, 302 Sebestyen, M. M., 129, 158 Seeger, I., 32( 189), 90 Seeger, R. C., 116, 118, 127, 164 Segal, S., 21(141, 142), 71(302), 89, 93 Sehon, A. H., 177, 213 Sela, M., 56(231), 91, 98, 99, 152, 153, 155, 161, 162, 164 Sellin, D., 303 Selner, J. C., 10(75), 87 Senik, A., 173, 194, 203, 210 Shabart, E. J., 193,210 Shaiponich, T., 169, 173, 184, 185, 214 Shapira, E., 281, 303 Shapiro, A., 216, 245 Shapiro, F., 230, 246 Sheagren, J. N., 295, 299 Shearer, G. M., 11, 12, 21( 142), 26( 82), 27(82), 29, 39, 75(166), 76, 87, 89, 90 Sheele, C., 151, 163 Sheil, A. G. R., 174, 212 Shellam, G. R., 23(147), 64(147), 74 (147), 89 Shibolet, S., 263, 303 Shin, H. S., 137, 164 Shirahama, T., 250, 253, 254, 261, 263, 267, 269, 271, 280, 284, 300, 302, 303,304 Shmueli, U., 264, 269, 299, 304 Short, B. F., 222, 247 Shortman, K., 121, 122, 123, 124, 125, 155, 156, 164 Shreffler, D. C., 226, 240, 241, 245, 246, 247 Siegel, I., 148, 164 Silmser, C. R., 216, 217, 220, 222, 244 Silver, H., 16( 127), 45(220), 47(220), 46( 127), 89, 91 Silverman, M . S., 115, 119, 163, 172, 178, 193, 196, 205 Silvers, A. A., 222, 223, 244 Silvers, W. K., 169, 174, 175, 189, 202, 206, 208, 213, 214, 217, 218, 219, 220, 222, 223, 224, 225, 226, 227,

228, 229, 230, 231, 232, 233, 238, 239, 240, 244, 245, 246, 247 Silverstein, A. M., 15( 101, 110), 88 Silvestre, D., 173, 194, 203, 210 Siminovitch, L., 6(48), 10(48), 86 Sinionian, S. J., 169, 173, 208 Simonsen, M., 81(335, 342), 94, 176, 190, 191, 205, 208 Simonsen, S. J., 169, 173, 184, 185, 214 Sin, Y. M., 177,213 Siskind, G. W., 15(103, 105, 111, 122), 66, 70(297, 298), 78( 103, 298), 79 ( 103), 88, 92, 93, 152, 155, 160 Sjodin, K., 5(38), 86 Sjoberg, O., 65, 69(295), 92, 93 Sjogren, H . O., 85(353), 94, 178, 180, 182, 190, 196, 199, 202, 209, 210, 213 Skamene, E., 171, 184, 209 Skinner, M., 173, 183, 210, 250, 269, 273, 280, 285, 287, 300, 304 Skowron-Cendrzak, A., 81( 342), 94 Slatis, H. M., 235, 247 Sloboda, A. E., 169, 213 Smeraldi, E., 296, 300 Smiley, J. D., 297, 302 Smith, C. A., 139, 165 Smith, J. M., 217, 220, 227, 228, 229, 230, 245, 246 Smith, M . R., 130, 137, 164, 165 Smith, R. F., 15(97), 18(97, 128), 88, 89 Smith, S. B., 68(285), 93 Snell, G. D., 172, 177, 178, 192, 196, 197, 199, 202,210, 213 Snyder, G. B., 169, 184, 213 Sobey, W. R., 222, 247 Sohar, E., 251, 252, 263, 264, 265, 269, 299, 301, 302, 303, 304 Solliday, S., 128, 157 Solomon, J. M . , 145, 159 Sordat, B., 100, 158 Sorenson, G. D., 255, 261, 263, 271, 280, 301, 304 Sors, C., 295, 302 South, M. A., 3( 21), 86 Souther, S. G., 200, 208 Southworth, S., 169, 205 Sovovi, V., 171, 184, 209

323

AUTHOR INDEX

Spargo, B. H., 184, 213 Sparrow, E. M., 168, 169, 205 Spencer, K., 107, 164 Spiegelberg, H. L., 105, 136, 164 Spieler, P. J., 144, 165 Spiro, D., 250, 255, 304 Spitznagel, J. K., 111, 112, 115, 116, 118, 149, 155, 164 Sprent, J., 10(74), 11, 16(74, 127), 24 (74), 26( 74), 29( 74), 31(349), 37(74), 43(217), 44(74), 45(220), 46( 127), 47( 220), 73( 74), 75( 74), 77, 78, 87, 89, 91, 94, 151, 162 Sprunt, D. H., 174, 180, 184, 198, 199, 208 Stanworth, D. R., 63(258), 92 Stark, J. M., 102, 149, 159, 164, 165 Stark, R. B., 169, 213 Stashak, P. W., 25( 156), 62( 156, 249), 63(156, 262), 89, 92 State, D., 169, 208 Stavitsky, A. B., 145, 159 Steinberg, M., 31( 171), 32( 171), 90 Steinberg, S. U., 149, 164 Steinmuller, D., 197, 213, 219, 247 Stelos, P., 145, 160 Stern, C., 215, 216, 247 Stern, K., 107, 164 Stetson, C. A., 178, 212 Stewart, T. H. M., 183, 213 Stickel, D. L., 168, 172, 173, 189, 199, 206 Stiffel, C., 107, 129, 157, 158, 160 Stimpfling, J. H., 172, 189, 196, 197, 210, 213, 226, 227, 246, 247 Stinson, E. B., 200, 208 Stobo, J. D., 82(343), 94 Stockert, E., 172, 173, 178, 179, 194, 202, 206, 212 Stolte, G., 298, 304 Stone, S. H., 173, 187, 205 Storb, R., 182, 190, 209 Storb, U., 56, 91 Stout, R. D., 79(321), 94 Straus, D. S., 146, 160 Striker, G. E., 184, 212 Strong, A., 215, 217, 245 Stuart, F. P., 183, 184, 189, 191, 199, 205, 213

Stupp, Y.,32( 187a), 35( 187a), 90 Stuyvesant, V. W., 284, 302 Sulica, A., 56, 91 Sulitzeanu, D., 31, 32( 167), 69( 294), 90, 93, 137, 148, 153, 162, 164 Sullivan, F. L., 107, 158 Suter, R. B., 173, 181, 210 Suzuki, Y.,263, 304 Svoboda, J., 180, 183,209 Swanson, M. A., 63( 256), 92 Sweet, V. C . , 126, 137, 158 Symmer, W. St. C . , 304 Szenberg, A., 3(16, 17, 18, 20), 31, 41, 85

T Tada, T., 18(129), 64, 89, 92, 198, 199, 213 Takahashi, T., 76, 93 Takasugi, M., 170, 198, 213 Takatsu, K., 10(69), 16(69), 87, 119, 153, 160 Takatsuki, K., 257, 277, 292, 303 Takeuchi, H., 220, 245 Tal, C., 261, 302 Talal, N., 257, 277, 292, 303 Taliaferro, W. G., 63(253), 92 Talmage, D. W., 65, 92, 121, 164, 245 Taniguchi, M., 64( 265, 266), 92 Tao, T. W., 59, 92 Tarrab, R., 56( 232), 91 Tatter, D., 284, 302 Taub, R. N., 151, 159, 169, 213 Taylor, R. B., 5(39), lO(61, 65), 11, 12(65), 16( 65), 18(65), 20( 65), 25, 26(65), 27(65), 31(171), 32 (171), 36, 42(65), 43(65), 61 (244), 68, 77(65), 78(65), 86, 87, 89, 90, 92, 93 Teilum, G., 261, 263, 277, 295, 296, 297, 304 Teitelbaurn, M. S . , 235, 246 Terry, R. D., 261, 304 Terry, W. D., 250, 252, 254, 273, 301 Theis, G. A., 70(297), 93, 122, 164 Thiery, J. P., 147, 164, 255, 304 Thoenes, G. H., 173, 184, 185, 191, 213 Thomas, E. D., 182, 190, 209

324

AUTHOR INDEX

Thorbecke, G. J., 15(122), 76(318), 88, 93, 105, 122, 159, 164 Tigelaar, R. E., 81(338), 94 Till, J. E., 6(48), 10(48), 86 Tinbergen, W. J., 184, 185, 191, 211 Tippett, P., 236, 237, 238, 245, 246, 247 Toivanen, P., 234, 247 Tokuda, S., 170, 186, 198, 213 Torrigiani, G., 32( 187), 90 Travis, M., 183, 184, 185, 201, 211 Treacy, A. M., 237,245 Treadwell, P. E., 58(236), 76(236), 91 Treat, R. C., 169, 207 Trentin, J. J., 3( 15), 85, 220, 245 Triplett, E. L., 223, 247 Triplett, R. F., 6(52), 7(52), 10(52), 11(77), 13(77), 39, 87 Troup, G. M., 170,214 Trowbridge, I. S., 21, 76( 138), 89 Tmka, Z., 104, 158 Truffa-Bachi, P., 32( 184), 90 Tsan, M., 142, 164 Tung, K. S. K., 188, 213 Turk, J. L., 102, 149, 161, 163 Tyan, M . L., 175, 211, 226, 246

Vaz, N. M., 175, 213, 226, 247 Vener, J., 189, 213, 230, 247 Villa, M. L., 296, 300 Villegas, G. R., 179, 207, 214 Vircliow, R., 249, 304 Virolainen, M., 111, 165 Vogt, A., 254, 278, 304 Voisin, G. A., 168, 169, 170, 171, 184, 185, 186, 191, 201, 213, 214 VojtiSkovi, M., 189, 211, 219, 230, 233, 246, 247 Volk, H., 179, 214 Volkman, A,, 112, 165 von Rokitansky, C. F., 249, 304 Vos, O., 234, 244

W

Wachtel, S. S., 169, 202, 214, 223, 231, 233, 247 Wagner, B., 79(324), 81(324), 94 Waksman, B. H., 4(27, 28), 68(285), 86, 93, 168, 173, 177, 187, 200, 208, 213 Waldenstrom, H., 294, 304 Waldmann, T. A., 151, 163, 252, 299 Walford, R. L., 169, 170, 184, 209, 214 U W a k e , J. H., 169, 199, 212 Uhr, J. W., 104, 134, 135, 136, 143, Wallis, V., 5(44, 45, 46), 7(54), 10 (44, 45, 46, 54), 23(145), 68(286), 160, 161, 163, 164, 178, 181, 197, 86, 87, 89, 93 199, 202, 213 Unanue, E. R., 31, 32(173, 174, 180, Walsh, T. E., 139, 165 185), 45, 61, 63, 80, 90, 91, 92, Wanstrup, J., 259, 303 94, 104, 105, 107, 108, 109, 111, Wanwyck, R., 169, 173, 184, 185, 214 113, 115, 116, 117, 118, 119, 131, Warburton, D., 234, 247 132, 133, 139, 140, 141, 142, 143, Wardlaw, A. C., 129, 165 144, 148, 149, 151, 153, 155, 158, Ware, R. G., 296, 302 159, 161, 164, 165, 188, 213 Warner, G. A., 85( 353), 94, 182, 209 Warner, N. L., 3(16, 17, 18, 19, 20, 22, 25), l 6 ( 127), 31, 32( 186, 188), 33 V (186, 188), 41, 45(220), 46( 127), Valentine, M. D., 108, 165 47( 220), 85, 86, 89, 90, 91, 131, Van Allen, M . W., 251,299 165, 179, 211 Vann, D. C., 21(139), 37(139), 38 Warren, K. S., 108, 110, 111, 158 (139), 39(139, 197), 58(239), 76 Wasserman, J., 176, 212 (139), 89, 90, 91 Waterston, R. H., 66, 92 Varco, R. L., 179, 208 Watson, B., 130, 165 Vasquez, J. J., 254, 277, 278, 304 Watson, M. L., 255, 304 Vaughan, R. B., 130, 165

AUTHOR INDEX

325

Wilson, R. E., 169, 173, 184, 185, 208, Watts, H. G., 191, 196, 208 214 Weatherall, D. J., 238, 247 Winchurch, R., 79( 320), 80(327), 93, Webb, D., 79(320), 80(327), 93, 94 94 Webster, J . R., 107, 158 Winn, H . J., 173, 174, 184, 198, 197, Wecker, E., 13, 14,55, 88 201, 203, 210, 213, 214 Wegmann, T. G., 191, 214 Weigle, W. O., 10(62), 11(62), 68 Wintrobe, M . M., 237, 246 (62, 292), 69(292), 70(292), 71 Witebsky, E., 277, 278, 303 (300), 74(292, 311), 87, 93, 105, Witz, I. P., 169, 189, 214 Wofsy, L., 32( 184), 90 117, 158, 160, 164, 188, 214 Woglom, W. H., 168, 178, 214 Weiner, L. P., 175, 200, 214 Wolf, N. S., 3( 15), 85 Weinstein, L., 126, 161 Wolfe, H. R., 3(23), 86 Weisberger, A. S., 147, 164 Weiser, R. S., 56, 91, 110, 160, 194, Wolff, S. M., 105, 159, 295, 299 195, 212 Wollheim, F. A., 254, 269, 284, 285, 300 Weiss, D. W., 168, 205 Wood, H. A., 173, 203, 205 Weiss, M. L., 188, 210 Wood, W. B., Jr., 130, 137, 164, 165 Weissman, G., 128, 138, 143, 144, 164, Woodruff, M . F. A., 63( 259), 92, 185, 165 210 Weissman, I., 220, 244 Woodworth, E. F., 177, 213 Welsh, P. D., 59(241), 92 Wortis, H. H., 25, 31, 37( 196), 89, 90, Welshons, W. J., 217, 244, 247 94 Wennersten, C., 4( 28), 86 Wright, J. R., 298, 304 Wepsic, T., 103, 158 Wright, P. W., 186, 200, 205, 214 Werder, A. A., 169, 214 Wunderlich, J. R., 176, 180, 206 Westbrock, D. L., 169, 210 Wheeler, H. B., 169, 214 Y Wheeler, N., 217, 222, 244 White, E., 173, 174, 183, 184, 185, 191, Yagi, Y., 169, 189, 214 213, 214 Yagil, G., 177, 205 White, R. G., 99, 102, 149, 159, 164, Yajima, Y., 79(320), 93, 147, 158 165 Yang, J. P. S., 182, 201, 209 Whitehouse, F., Jr., 170, 171, 206, 214 Yang, S. L., 239, 247 Wiener, E., 149, 165 Yogev, R., 281, 303 Yoon, S. B., 217, 245 Wigmore, W., 284, 302 Wigzell, H. L. R., 2 ( 6 ) , 32(169, 183), Young, D. D., 150,157 Young, L., 175, 199,200,206 81(342), 85, 90, 94 Yron, I., 169, 213 Willerson, J . T., 252, 295, 299 Williams, G. M., 99, 100, 102, 165 Z Williams, H. L., 222, 244 Williams, R. C., Jr., 254, 304 Zaalberg, 0. B., 223, 225, 247 Williams, R. D., 169, 205 Zamir, R., 63( 261),92 Wilson, A. B., 31(176), 32(176), 90 Zatz, M . M., 153, 165 Wilson, C. B., 169, 214 Zavala, C., 238, 245 Wilson, D. B., 174, 208, 227, 228, 229, Zeiss, I. M., 238, 239, 247 230, 231, 233, 244 Ziegler, J. L., 183, 207 Ziff, M., 297, 302 Wilson, R., 173, 178, 208

326

AUTHOR INDEX

Zilversmit, D. B., 129, 165 Zimmerman, B., 170, 174, 184, 192, 193, 197, 201, 214 Zimmerman, C. E., 169, 172, 214 Zlotnick, A., 261, 300 Zolla, S., 292, 300

Zoschke, D. C., 128, 157 Zucker-Franklin, D., 250, 252, 253, 255, 259, 263, 264, 265, 267, 270, 271, 274, 277, 280, 281, 286, 287, 294, 301, 302, 303, 304 Zurier, R. B., 144, 165

Index A Adjuvants, macrophages and, 149-151 Amyloid, classification, 250-253

fibrils,

B B-cells, antigen receptors on, 32-33 cell-mediated immunity and, 81-82 immunological adjuvants and, 79-81 immunological memory and, 75-79 immunological specificity, 28-32 immunological tolerance and, 67-75 mechanism of regulation of function by T-cells, antigen presentation and concentration, 43-47 production and secretion of mediators, 47-62 transfer of genetic information, 4243 recognition of hapten and carrier determinants, 33-37 regulatory influence of activated T-cells, class of immunoglobulins synthesized, 25-26 selective pressure of antigen, 26-28 stimulation in absence of T-cell regulation, 23-25 sensitivity and resistance to corticosteroids, 37, 40-42 X-irradiation, 3 7 4 0

comparative studies, 285-289 conclusions, 289-294 dye-binding, 264-265 immunological studies, 277-285 morphological studies, 266269 physicochemical and biochemical studies, 270-276 general approaches to study, 253-255 historical background, 249-250 morphological studies in sectioned tissue, 255-264 pathogenesis, speculations on, 29P299 Antibodies, immunological enhancement and, 170172 production, significance of regulatory influence of T-cells and, 82-85 synthesis, suppressive effects of T-cells, 62-67 Antigen( s ) , immunogenic moiety, 137-147 immunological enhancement and, 1 7 2 174 macrophage-bound, immunogenicity and, 97-128 C other sex-limited and sex-linked, 238242 Carrier, presentation and concentration, regudeterminants, recognition by T and B lation of B-cell function, 43-47 cells, 33-37 protein, immune response to, 4-14 Canier effect, different determinants on receptors on B and T cells, 32-33 same antigen and, 14-22 selective pressure on B-cells, 26-28 Corticosteroids, sensitivity of T and B uptake by macrophages, 128-137 cells, 37, 40-42 Y chromosome of mouse and, 21C236 Antigenic competition, T-cells and, 65'D 67 Autoimmunity, immunological enhance- Delayed hypersensitivity, immunological enhancement and, 186-188 ment and, 186188 327

328

SUBJECT INDEX

Hapten, determinants, recognition by T and B cells, 33-37 Host, immunological enhancement and, 174-177 Human, Xg blood group locus in, 2 3 5 238

Immunological enhancement, autochthonous tumors and, 182-183 components, antibodies, 170-172 antigens, 172-174 host, 174-177 definitions, 169-170 delayed hypersensitivity and autoimmunity and, 186-188 experimental tumors and, 177-182 fetus as homograft and, 188-189 historical, 167-169 normal tissue grafts and, 183-186 theories, 191-192 afferent blockade, 192-195 central blockade, 197-201 efferent blockade, 195-197 miscellany, 201-202 personal preference, 202-204 tolerance and, 190-191 Immunological memory, functions of T and B cells in, 7%79 Immunological tolerance, functions of T and B cells in, 67-75 immunologic enhancement and, 190191 Immunology, amyloid and, 277-285

I

1

Immune response, requirement for two distinct lymphoid cell types, carrier effect and cooperative interaction, 14-22 response to foreign erythrocytes and protein antigens, 4-14 Immune system, specific cells of, 3-4 Immunity, cell-mediated, T and B cells and, 8182 Immunogenicity, association of macrophage-bound antigen, in vitro experiments, 12&128 in uiuo experiments, 97-120 Immunoglobulin, class synthesized, T-cell activity and, 25-26 Immunological adjuvants, functions of T and B cells, 79-81

Lymphocyte, macrophage contact, 147148

E Erythrocytes, foreign, immune response to, 4-14

F Fetus, as homograft, 188-189

G Genetic information, transfer, regulation of B-cell function and, 42-43 Grafts, normal tissue, immunological enhancement and, 183-186 Graft-versus-host reaction, Y antigen and, 233-234

H

M Macrophages, adjuvants and, 149-151 antigen handling by, antigen uptake, 128-137 immunogenic moiety, 137-147 lymphocyte contact, 147-148 role in antigenic stimulation, summary, 151-156 Mediators, production and secretion, regulation of B-cell function and, 47-62

T T-cells, antigen receptors on, 32-33 cell-mediated immunity and, 81-82 immunological adjuvants and, 79-81

329

SUBJECr INDEX

immunological memory and, 75-79 immunological specificity, 2 8 3 2 immunological tolerance and, 67-75 mechanism of regulation of B-cell function, antigen presentation and concentration, 4 3 4 7 production and secretion of mediators, 47-62 transfer of genetic information, 42-

43

nature of regulatory influence on Bcells, class of immunoglobulin synthesized,

25-26

selective pressure by antigen, 26-28 stimulation in absence of T-cell regulation, 23-25 recognition of hapten and carrier determinants, 33-37 sensitivity and resistance to corticosteroids, 37, 4 0 4 2 X-irradiation, 3 7 4 0 significance of regulatory influence on antibody production, 82-85 suppressive effects on antibody synthesis,

enhancement of immune response by depletion of T-cells, 62-65 antigenic competition and, 65-67 Tumors, autochthonous, immunological enhancement and, 182-183 experimental, immunological enhancement and, 177-182

X Xg blood group locus, human, 236-238 X-irradiation, sensitivity of T and B cells, 37-40

Y Y antigen,

mouse, genetic basis of response to, 222-

227

genetic determination of, 216-220 graft-versus-host reaction and, 233-

234

induction of tolerance to, 227-233 placental growth and sex ratio and,

234-236

tissue localization, 226222 Y chromosome, genes of, 215-216