Elisa: Theory and Practice (Methods in Molecular Biology) (НЕ полн.)

&IAPTER

1

Basic Immunology 1. The Immune

Response

A knowledge of immunology is essential in developing ELISAs. Information about the specific system and similar systems already studied from the biochemical and immunological aspects is important to allow development of assays and to determine the significance of results. It is not the intent of this chapter to provide an in-depth understanding of immunology; however, it is necessary for the ELISA operator to have a basic understanding of: 1, The immunology of infectious diseases. 2. The propertiesof certain componentsof the immune system. 3. Aspects of serology. Mammals possess a system of surveillance called the immune system that protects them from disease-causing (pathogenic) microorganisms, such as viruses, bacteria, and parasites. The immune system specifically recognizes and eliminates pathogens. The protection afforded by the immune system of the mammal is divided into two functional divisions, namely, the innate immune system and the adaptive immune system, both of which respond specifically to these foreign substances. Innate immunity acts as the first line of defense against infectious agents, and most potential pathogens are checked before they can establish infection. If these defenses are overcome, then the adaptive immune system is activated. The adaptive system produces a specific reaction against each infectious agent, and also remembers that particular agent and can prevent it causing disease in the future. Most of the applications of ELISA involve studies on the adaptive immune system, so this will be featured in more detail. Excellent books on basic immunology have been written, in particular (I), which has extensive references for specific areas outlined below.

1

2

Basic Immunology 1.1. Innate

Immunity

The factors involved in innate immunity include biochemical and physical barriers, e.g., the skin, acting as an impenetrable barrier to infectious agents, and the presence of lysozyme in tears, which destroys bacteria, such as S. aureus. The key difference in this system from that of the adaptive immunity is that resistance is not improved by repeated infection and that the system is aspecific in nature. Thus, if organisms do penetrate an epithelial surface, they encounter phagocytic cells of the reticuloendothelial system (RE), where the cells are of many types derived from bone marrow cells. Their function is to engulf, internalize, and destroy infectious agents. For this purpose, they are placed strategically where they might encounter particles, e.g., the Kupffer cells of the liver line the sinusoids along which blood flows. The blood phagocytes include neutrophil polymorph and the blood monocytes, both of which can migrate into tissues as a result of an invasive stimulus. Other cells, such as natural killer (NK) cells, are leukocytes capable of recognizing cell surface changes on virus-infected cells. Such cells bind and kill cells under the influence of substancescalled interferons, which are produced by the virus-infected cells or sometimes by lymphocytes. Other factors involved in innate immunity involve certain serum proteins. These are referred to as acute-phaseproteins. The concentration of such proteins rises dramatically on and is maintained throughout infection. The various proteins have defined properties and produce protective effects through complex interactions with other serum components, such as complement followed by lysis of disease agents. Visible signs of an early immune response are observed in inflammation, which is the body’s reaction to an injury, such as invasion with an infectious agent. Three major events occur, namely: 1, An increasedblood supply to the infected area, 2. An increasem the permeability of the capillarrescausedby retraction of endothelial cells, allowing larger moleculesto crossthe endothelium,e.g., soluble mediators. 3. Migration of leukocytes( neutrophils,polymorphs,and macrophages)from capillaries to surroundmgtissues, These features, whereby phagocytes are attracted to sites of injury, are important in immunity and initiate all levels of potential innate immune mechanisms.

Antigens

3

1.2. Adaptive Immunity Innate immunity relies on stimulation of factors through aspecific recognition of infectious agents. Problems arise when the recognition process is not activated, e.g., when phagocytes are unable to recognize the infectious agent either because they lack a suitable receptor for the agent or because the agent does not activate soluble factors. What is needed in such a situation is a specific molecule that can attach at one end to the infectious agent and at the other end to the phagocytic cell. Such molecules, called antibodies, are produced in the mammalian systems. Antibodies are produced by B lymphocytes of the adaptive immune system, which act as flexible adaptors between infectious agents and phagocytes. Any particular antibody molecule can bind only to one type of infectious agent, and the other end of the molecule binds to the phagocyte by way of a receptor, the Fc receptor. A stylized structure of an antibody is shown in Fig. 1A. Figure 1B shows that IgG is a rather bulky structure when examined at the molecular level. Antibodies are effectively bifunctional molecules. One part, which is extremely variable between different antibodies, binds to all the various infectious agents, whereas the second, constant portion binds to receptors of cells and also activates complement. 2. Antigens The mammalian immune system has the ability and capacity to recognize surface features (topography) of foreign macromolecules or microorganisms that are not normal constituents of that mammal (e.g., pathogenic microorganisms). This recognition of surface features is normally specific, and the components of the mammal that carry out this specific recognition of surface features of macromolecules or microorganisms are protein molecules called antibodies. Foreign substanceshave specific surface features, called antigens, that antibodies recognize. The portion of the antigen to which the antibody binds is called the antigenic determinant or epitope. Antibody specifically binds to epitopes on the antigen by multiple noncovalent interactions similar to the interactions that confer specificity to enzyme-substrate reactions. The part of the antibody that binds to the antigenic determinant is termed the antigen-combining site or paratope (which is complimentary to the epitope). An antigen eliciting a response from the immune system is referred to as an immunogen. Microorganisms, macromolecules, such as foreign

Basic Immunology A

DISUPHIDE

NH+

BON6 LIGHT

HEAVY

B

CHAIN

CHAIN

Heayy chain

Fig. 1. (A) Representationof basic structureof an IgG molecule. (B) Model of IgG molecule basedon X-ray crystallographicanalysis. proteins, nucleic acids, carbohydrates, polysaccharides, and so forth, are usually effective immunogens. Molecules with mol wt below 5000 usually are not effective immunogens. However, many of these small nonimmunogenic molecules, when covalently attached to a large molecule, can stimulate an immune response. These molecules, which are nonimmunogenic, are termed haptens, and the large molecules to which they can be covalently attached, generally proteins, are termed carriers. Once the hapten is attached covalently to the carrier protein and introduced into an organism, a specific antibody response to the hapten and the carrier (if the latter is recognized) occurs. This antibody response can be specific to the hapten. Thus, nonimmunogenic molecules (haptens) can be recognized by an organism when covalently attachedto a carrier protein.

Antigens

5

2.1. Antigen Presentation, Processing, and Recognition Lymphocytes account for between 20 and 80% nucleated cells in the blood, and over 99% nucleated cells in lymphatic fluid. Lymphocytes contact and respond to antigens in specialized lymphoid organs. Included in these specialized organs are the spleen, thymus, lymphatic tree and the lymph nodes positioned along them, bone marrow, and Peyer’s patches (appendix, adenoids, tonsils [Bursa of Fabricius]). The lymphoid system has three principal functions, namely: 1. Concentration of antigens from all parts of the body into lymphoid organs. 2. Circulation of lymphoid cells through these organs to ensure antigen exposure to antigen-specific lymphocytes in a short period of time. 3. Transmission and dissemination of the products of the immune response, e.g., antigen-specific effector B + T-cells, humoral antibodies, throughout the body.

Antigens are collected and processed by different lymphoid organs depending on their route of entry into the body, i.e., respiratory system, gastrointestinal tract, skin, vector transmission, venereal, and so on. Antigen processing in all these lymphoid organs involves macrophages,a short period of time after infection (injection). The antigen(s) becomes incorporated in special vesicles (phagolysosomes) within the macrophage. The macrophage cell surface either retains or receives a small amount of immunogenic material for presentation to antigen-specific lymphocytes. Binding of this macrophage surface antigen to a B-cell or T-cell (presentation of antigen) induces a general activation of the cell. This process, known as blast-transformation, causes the B-cell (or T-cell) with the appropriate receptor specificities to recognize the antigen-presenting macrophage to enlarge. Such activated cells initiate DNA synthesis, divide, and give rise to effector cells and memory cells of the B-cell (or T-cell) lineage. Most memory B-cells re-enter the general circulation, whereas most effector B-cells are retained in the lymph node. An individually activated B-cell proliferates and differentiates to form plasma cells that begin to produce identical antibodies with a single antigen specificity (at the rate of 3000-30,000 molecules/cell/s). An organism’s total response even to a simple antigen is almost always heterogeneous with respect to antibody specificity because most antigens have multiple epitopes, which can trigger the activation of different

6

Basic Immunology

B-cells. Consequently, the serum of the host reflects the heterogeneous collection of immunoglobulin molecules previously secreted. Although individual B-cells are committed to produce ,one or at most two antibody isotypes, the B-cell response to any antigen can produce antibody molecules of all five classes of immunoglobulins. Since antibody molecules of different classes differ in their heavy-chain-constant regions (Fc), they may exhibit identical antigen-binding specificities and, hence, identical variable regions. Antibodies of the five classes mediate different physiological effector functions, and are present in serum at different concentrations and for different half-life periods. The different classes are also produced in different relative amounts in primary and secondary immune responses. 2.1.1. IgM This is the first antibody produced in responseto an irnmunogen and is particularly effective against invading microorganisms. It is a pentamer in serum, and although the affinity of each active site on the pentamer (10 in number) for an epitope may be low, the overall avidity of the pentamer for a complex antigen is high becauseof the repeatingnature of epitopes on many cell membrane antigens.Becauseit is presentas a pentamer in serum, IgM is about 1000 times more effective (on a molar basis) at agglutinating cells by crosslinking them, than a monomeric antibody against the same epitope. IgM coated onto the antigen of target cells stimulates target cell ingestion by macrophagesand target cell destruction (lysis) by compliment fixation. 2.1.2. IgG This monomeric antibody is normally produced later in the immune response than IgM. This is the most prevalent antibody in blood and tissue spaces, and is capable of fixing complement. It also activates macrophage ingestion of opsonized (coated) antigen particles. IgG is the only class of antibody that can cross the placenta to provide passive immunity to the developing fetus. Normally, the affinity of IgG antibodies toward a specific antigen increases with time after immunization, a process known as affinity maturation. 2.1.3. IgA This is also produced later in the immune response than IgM. It can exist as a monomer, dimer, or trimer of the basic Y-shaped structural unit. IgA antibodies are important at numerous epithelial surfaces and

Adaptive

Immunity

and Clonal Selection

7

act as a potential protective barrier at several points of entry, e.g., gastrointestinal tract, respiratory tract, genitourinary tract, eyes, and so forth. Some epithelial cells produce a polypeptide, called the secretory component, which complexes to the Fc region of IgA and mediates their transport across the epithelial cell surface to the lumen. IgA B-cell precursors are especially frequent in lymphoid organs draining the gastrointestinal tract and in the mammary glands. IgA is the major immunoglobulin in colostrum and milk, and is also present in sweat, tears, and saliva. 2.1.4. IgE This monomeric antibody is heat-labile. It is present in blood in very low concentrations. IgE antibodies are produced in response to infection mainly the helminth parasites and in allergic atopic conditions. IgE antibodies can bind via their Fc regions to mast cells or blood basophils. Further interaction of this bound IgE with a cognate (known and recognized by IgE) antigen can trigger cell degranulation and the liberation of vasoactive compounds, such as histamine and heparin. 2.1.5. IgD This monomeric antibody is present only in minute concentrations in blood. Its functions are unknown. 3. Adaptive Immunity and Clonal Selection The immune system as a whole can specifically recognize many thousands of antigens. The specificity of the adaptive immune response is based on the specificity of the antibodies and lymphocytes, and since it has been shown that each lymphocyte is only capable of recognizing one particular antigen, this means that the lymphocytes recognizing any particular antigen are a very small proportion of the total. Thus, we have to explain how an adequate response to an infectious agent is mounted. The answer is clonal selection, whereby antigen binds to a small number of cells that can recognize it and induces them to proliferate. Thus, the antigen selects the specific clones of antigen-binding cells. This is illustrated in Fig. 2. This process occurs in both B-lymphocytes, where they mature into antibody-producing cells, and T-lymphocytes, which areinvolved in the recognition and destruction of infected cells. A basic requirement for the production of an antibody response is that the immunogen possessessurface features that are recognized as foreign in the animal into which it is introduced or in which it occurs.

Basic Immunology

8

Antigen selection

Production of antibody 2 Fig. 2. The antibody-producing cells (B-cells) are programmed to make a single antibody only. The antibody is placed on an Fc receptor on the cell’s surface. Each B-cell has a different receptor, and antigen binds to those cells with the appropriate receptor. The cells become stimulated to multiply and mature onto antibody-producing cells and memory cells, which can live longer. All the cells have the same antigen-binding capacity. 4. Antibodies Antibodies are fundamental reagents in ELISA, and the determination of their presence and/or concentration in the blood is vital in understanding disease processes and in diagnosis of disease. A knowledge of the properties of antibodies is fundamental to the development of specific assays. An understanding of the variation in antibody composition of different mammals is also important.

9

Antibodies ANTIGEN

BINDING (PARATOPE)

SITE

CARBOHYDRATE

CARBOHYDRATE-

FC

I@

s s c#& HEAVY

WSULPHJDE

BOND

s s s s

CHAIN

I

I COO-

coo-

Fig. 3. Structural elements of an IgG molecule. 4.1. Antibody Structure and Function Antibodies form a group of glycoproteins present in the serum and tissue fluids of all mammals. The group is also termed immunoglobulins, indicating their role in adaptive immunity. All antibodies are immunoglobulins, but not all immunoglobulins are antibodies, i.e., not all the immunoglobulin produced by a mammal has antibody activity. Five distinct classes of immunoglobulin molecule have been recognized in most higher mammals. These are immunoglobulin (Ig) G (IgG), IgA, IgM, IgD, and IgE. These classes differ from each other in size, charge, amino acid composition, and carbohydrate content. There are also significant differences (heterogeneity) within each class. The basic polypeptide structure of the immunoglobulin molecule is shown in Fig. 3.

10

Basic Immunology

The basic structure of all immunoglobulin molecules is a unit of two identical light (L) polypeptide chains and two identical heavy (H) polypeptide chains linked together by disulfide bonds. The class and subclass of an immunoglobulin molecule are determined by its heavy-chain type. Thus, in the human, there are four IgG subclasses, IgGl, IgG2, IgG3, and IgG4, which have heavy-chains called 1,2,3, and 4. The differences between the various subclasses within an individual immunoglobulin class are less than the differences between the different classes.Thus, IgGl is more closely related to IgG2, and so on, than to IgA, IgM, IgD, or IgE. The most common class of immunoglobulin is IgG. IgG molecules are made up of two identical light chains of mol wt 23,000 Daltons and two identical heavy chains of mol wt 53,000 Daltons. Each light chain is linked to a heavy-chain by noncovalent association, and also by one covalent disulfide bridge. For IgG, each light-heavychain pair is linked to the other by disulfide bridges between the heavy chains. This molecule is representedschematically in the form of a Y, with the amino (N-) termini of the chains at the top of the Y and the carboxyl (C) termini of the two heavy chains at the bottom of the Y-shape. A dimer of these light-heavy-chain pairs is the basic subunit of the other immunoglobulin isotypes. The structures of these other classes and subclasses differ in the positions and number of disulfide bridges between the heavy chains, and in the number of light-heavy-chain pairs in the molecule. IgG, IgE, and IgD arecomposed of one light-heavy-chain pair. IgA may have one, two, or three light-heavy-chain pairs. IgM (serum) has five light-heavy-chain pairs, whereas membrane-bound IgM has one light-heavy-chain pair. In the polymeric forms of IgA and IgM, the lightheavy-chain pairs are held together by disulfide bridges through a polypeptide known as the J chain. In both heavy and light chains, at the N-terminal portion, the sequences vary greatly from polypeptide to polypeptide. In contrast, in the C-terminal portion of both heavy and light chains, the sequences are identical. Hence, these two segments of the molecule are designated variable and constant regions. For the light chain, the variable region (V) is approx 110 amino acid residues in length, and the constant region (C) of the light chain is similarly about 110 amino acids in length. The variable region of the heavy chain (Vu) is also about 110 amino acid residues in length, but the constant region of the heavy-chain (C,) is about 330 amino acid residues in length.

Antibodies

11

The N-terminal portions of both heavy- and light-chain pairs comprise the antigen-combining (binding) sites in an immunoglobulin molecule. The heterogeneityin the amino acid sequencespresentwithin the variable regions of both heavy and light chains accounts for the great diversity of antigenspecificities among antibody molecules. In contrast, the constant regions of the heavy chain make up the part of the molecule that carries out the effector functions, which are common to all antibodies of a given class. From Fig. 3, it can be seen that there must be two identical antigenbinding sites (more in the case of serum IgM and secretory IgA). Hence, the basic Y-shaped immunoglobulin molecule is bivalent. This bivalency permits antibodies to crosslink antigens with two or more of the same epitope. Antigenic determinants that are separated by a distance can be bound by an antibody molecule. The antigen-combining site (active site) is a crevice between the variable regions of the light- and heavy-chain pair. The size and shape of this crevice can vary because of differences in the relationship of VL and VH regions, as well as differences in the amino acid sequencevariation, Thus, the specificity of antibody will result from the molecular complementarity between determinant groups (epitopes) on the antigen molecule and amino acid residues present in the active site. From this we can see that an antibody molecule has a unique threedimensional structure. However, a single antibody molecule has the ability to combine with a range (spectrum) of different antigens. This phenomenon is known as multispecificity. Thus, the antibody can combine with the inducing antigenic determinant or a separate determinant with similar structures (crossreacting antigen). Stable antigen-antibody complexes can result when there is a sufficient number of short-range interactions between both, regardless of the total fit. This is a problem for the immunoassayist, and care must be taken to ensure that the operator is assaying for the correct or desired antigen; therefore, careful planning of negative and positive controls is essential. Figure 4 demonstrates the digestion of IgG using papain or pepsin proteolytic enzymes. Mild proteolysis of native immunoglobulin at the hinge regions of the heavy-chain by papain will cleave IgG into three fragments. Two of these fragments are identical and are called “fragment antigen-binding” or “Fab.” Each Fab consists of the variable and constant regions of the light chain and the variable and part of the constant (ChI domain) regions of the heavy-chain. Therefore, each Fab carries

Basic Immunology

12

ANTIGEN COMBINING SITE

Heavy chains

MONOVALEW

FC’ MONOVALENT

FE

Fig. 4. Enzymic cleavageof humanIgG. Pepsincleavestheheavychainto give F(Ab’), andpFc’ fragments.Furtheractionresultsin greaterfragmentationof central proteinto peptides.Papain splits the molecule in the hinge region to give two Fab fragments and the Fc fragment. Further action on the Fc can produce Fc’.

one antigen-binding site. The third fragment, consisting of the remainder of the constant regions of the heavy-chains, is readily crystallizable and is called “fragment crystallizable” or Fc. Pepsin digestion cleaves the Fc from the molecule, but leaves the disulfide bridge between the Fab regions. This molecule contains both antigen-combining sites and is bivalent. The five immunological classes (isotypes) can be distinguished structurally by differences in their heavy-chain constant regions (i.e., mainly the Fc portion). These heavy-chain classes define the corresponding immunoglobulin classes IgA, IgG, IgD, IgE, and IgM. Some classes can be divided further into subclasses.

Antibodies

13

In addition, two major typesSof light chains exist, based on the differences in the constant region Cl and are known as kappa (K) and lambda (h). Immunoglobulins from various mammals appear to conform to the above format. However, the subclass designation and variety may not be the same in all species examined, e.g., mice have IgGl, IgG2a, IgG2b, and IgG3; cows have IgGl and IgG2. 4.2. Antibody Production in Response to Antigenic Stimulus

The antibodies produced in a humoral responseto antigenic stimulus am heterogeneousin specificity andmay include all immunoglobulin classes.This heterogeneousresponseis owing to the fact that most antigenshave multiple antigenicdeterminantsthat trigger off the activationof different B-cells. Therefore, the serum of any mammal(vertebrate) contains a heterogeneousmixture of immunoglobulin molecules.The specificities of theseimmunoglobulin molecules will reflect the organism’s past antigenic exposure and history. The first antibody produced in response to a primary exposure of an immunogen is IgM. When the immunogen is persistent or the host (mammal) is re-exposed to the immunogen other classes of antibody may be produced as well as IgM. Thebbody compartment in which the immunogen is presented can determine the predominant antibody isotype produced (e.g., IgA in the gastrointestinal tract). In general, primary exposure to an immunogen stimulates the production of IgM initially, followed by the appearance of IgG, as shown in Fig. 5. If no further exposure occurs or the immunogen is removed by the mammal, a low level of IgM and IgG can be detected. If re-exposure occurs, a similar peak of IgMiantibody is produced, which declines in a similar kinetic manner to the primary IgM response,but the IgG response is not only more rapid (over time), but also reaches higher serum levels, which persist for a longer period of time. This IgG response to re-exposure is known as the “anamnestic response.” Where complex antigens occur, as in infectious diseases, the dosage (infection level), type of antigen (viral, bacterial, protozoan, and helminthic), route of infection (oral, respiratory, cutaneous), and species of mammal infected (cow, pig, camel, and human) will all affect the degree and speed by which IgG replaces IgM. These considerations are vital for the immunoassayist concerned with diagnosing infectious diseasesof mammals, and great care and planning

14

Basic Immunology

0

5

10

15

20

25

30

Days after primary

35

40

45

50

55

dose of antigen

Fig. 5. Anamnestic response following second administration of antigen. Primary response following initial antigen dose has a lag phase, where no antibody is detected (4-5 d). This is followed by a log phase, where antibody is produced. A plateau phase follows where antibody titers stabilize after which a decline in titer is observed. On secondary stimulation, there is an almost immediate rise in titer and higher levels of antibodies are achieved that are mainly IgG.

should be exercised before undertaking such immunoassays. It should also be noted at this stage that different infectious disease agents can stimulate different antibody isotypes. For example, certain viral pathogens stimulate predominantly IgM agglutinating responses, bacterial polysaccharides stimulate IgM (and IgG2 in humans) antibodies, and helminthic infections stimulate IgE antibody synthesis. In general, it can be stated that during the development of immunity to infectious disease agents, the antibodies produced become capable of recognizing antigens better, as demonstrated by improved antigen-antibody interaction, three dimensional fit, and wider epitope recognition. The multispecificity of antibody molecules, i.e., the ability to combine with a variety of epitopes containing similar molecular structures,is depen-

Antibodies

15

1

Antigen

Water excluded Fig. 6. Attractive forces binding antigen to antibody. A close approach of interacting groups IS needed for these forces to interact. Hydrogen bonding results from formation of hydrogen bridges between atoms; electrostatic forces are the result of attraction of oppositely charged groups on two protein side chains. Van der Waals forces are generated by interaction between electron clouds, and hydrophobic bonds rely on the association of nonpolar, hydrophobic groups, so that contact with water molecules is minimized. Half the total binding may be the result of the hydrophobic bonding.

dent not only on the heterogeneityof the epitope in question, but also on the molecular construction of the antigen-reactive sites (paratope) of the anti-

body molecules. Since the binding of antibody to antigen is mediated by several types of noncovalentbonds (e.g.,electrostatic,hydrogen bonds, Van der Waals forces, and hydrophobic forces) (seeFig. 6), the strongestbinding must occur when the paratope matches the epitope perfectly (best fit).

Basic Immunology

16

Good tit High attraction Low repulsion

High affinity Fig. 7. A good fit betweenantrgenicsitesand antibody-combining sitescreates an environment for the intermolecular attractive forces to be createdand limits the changesof repulsive forces.The strengthof the single antigen-antibody bond is the affinity that reflects the summationof the attractive andrepulsive forces.

4.2.1. Affinity

and Avidity

The binding energy between an antibody molecule and antigen determinant is termed affinity. Thus, it can be seen that antibodies with paratopes that recognize epitopes perfectly will have high affinity (good fit) for the antigen in question, whereas antibodies with paratopes that recognize epitopes imperfectly will have low affinity (poor fit) for the antigen in question. Low-affinity antibodies, where the fit to antigen is less than perfect, will have fewer noncovalent bonds established between the complex, and the strength of binding will be less, as shown in Fig. 7. With simple immunogens containing few epitopes, it is seen that as the antibody response develops (in response) to this immunogen, its recognition by antibody will become better/closer, e.g., low-affinity antibodies will be replaced by high-affinity antibodies, which will cause the interaction between antigen and antibody to be more stable. Antibodies produced later on during infection are generally of higher affinity than

Antibodies

17

those produced early on during infection. Hence, the IgG antibodies produced in response to re-exposure will be of higher affinity than those produced in response to initial exposure. In a serum sample, where there has been polyclonal stimulation of antibody production by antigen, there will be a variety of affinities present within the antibodies. The match (fit) between antibodies to that antigen will be variable, and the antibodies present in that serum sample will bind to antigen differentially. Thus, it can be seen that not only can an antigen stimulate different antibody isotypes, but also antibodies with different affinities for the antigenic determinant. Avidity can be regarded as the sum of all the different affinities between the heterogeneous antibodies contained in a serum and the various antigenic sites (epitopes). We have already discussed the reasons for affinity variation from the points of view of the antibodies and antigens, each interaction between an antibody population and a specific antigenic site has an individual affinity or equilibrium constant for the defined reaction. Thus, the avidity can be regarded as the average of all the affinity constants for all interactions between the serum and antigen(s). It is important to realize that the avidity of a serum may change on dilution, since one may be diluting out particular populations of antibodies. As an example, one could have a serum containing a low quantity of antibodies showing high-affinity for a particular complex antigen and a high quantity of low-affinity antibody. Under immunoassay conditions where that serum is not diluted greatly, one would have “competition” for antigenic sites between the high- and low-affinity antibodies, and the high-affinity antibodies would react preferentially. On dilution, however, the concentration of the high-affinity antibodies would be reduced until one could only be left with low-affinity antibodies. Such problems are important where one is using immunoassays to compare antigens by their differential activity with different antisera. The dilution of any serum can affect its ability to discriminate between antigensowing to the dynamics of the heterogeneous antibody population (relative concentrations and affinities of individual antibody molecules). Such problems of quality and quantity do not apply to monoclonal antibodies (MAb), since by definition the immunoglobulin molecules in the population are identical. They all have the same affinity and, therefore, the avidity equals affinity, Thus, the population reacts identically to any individual molecule in that population. After diluting the monoclonal population, there is no alter-

Basic Immunology

18 Specific reaction

AgX Homologous Antigen

Cross-reaction

No reaction

AgZ Determinant shared

B

No determinants shared

Fig. 8. Specificity, crossreactivity, and nonreactivity. Antisera contain populations of antibodies. Each population is directed against a different determinant (A, B, and C above). Antigen X and Y share a determinant (B). Thus, antiserum against X will react with antigen Y (crossreact) as well as reacting specifically with antigen X. Antiserum against antigen X does not react with antigen Z, since no determinants are shared. ation in the affinity/avidity of the serum, and a change noted for reaction between the MAb and antigen must be the result of changes on the particular antigen. Crossreactions between sera and different antigens are illustrated in Fig. 8. Here specific reactions occur where all the antibodies have “best fit.” Where two antigens share a similar antigen, crossreactions will be observed. The two nonidentical sites may also contribute to the crossreaction. Where all the antibodies show no recognition of the antigens available, no reaction will be seen. It is important to understand the

Antibodies

19

concepts of variability in (1) isotype production, and (2) affinity and affinity maturation when developing immunoassays for infectious disease agents that are normally more chronic than acute in duration. Because antigens introduced into different compartments of the mammalian body can stimulate the production of different antibody isotypes, local antibody responses in the gastrointestinal tract and the respiratory three are predominantly IgA isotypes, whereas those in the other major compartments are predominantly IgG (IgM). Certain sites in the body (e.g., testes) are immunologically privileged and stimulate lower antibody responses to immunogens. Most infectious diseasesare transmitted by aerosolization, close contact, or vectors. Thus, their route of transmission is variable. In addition, their final location may be distant from their point of deposition. Similarly, whereas some pathogens are capable of division within the host, others are incapable of division within the mammalian host (e.g., helminths). 4.3. Antibody Production in Response to Immunization/Vaccination

Individuals can be rendered resistant to infectious agents by either passive immunization or active immunization. In general, the beneficial effects of immunization are mediated by antibodies. Thus, the effects of immunization can be monitored by the immunoassayist. 4.3.1. Passive Immunization This is accomplished by transferring antibodies from a resistant to a susceptible host. Passively transferred antibodies confer a temporary, but immediate resistance to infection, but are gradually catabolized by the susceptible host. Once passive protection wanes, the recipient becomes susceptible to infection again. Passively transferred antibodies can be acquired by the recipient either transplacentally or transcolostrally, as in neonates, or by injection of purified antibodies from a resistant donor into a susceptible recipient. 4.3.2. Active Immunization This is accomplished by administering antigens of infectious agents to individuals so they respond by producing antibodies that will neutralize the infectious disease agents, once contracted. Re-exposure to such agents, following active immunization, will result in an anamnestic immune response, whereby the antibodies or effector cells produced will

Basic Immunology

20

be capable of neutralizing the effect of or destroying the inciting agents. Such antibodies are known as protective antibodies, and their complementary antigens are called protective antigens. The protection conferred by active immunization is not immediate, as in passive immunization, because the immune system requires a length of time in order to process such antigens and produce protective antibodies. However, the advantage of active immunization is that it can be long-lasting, and restimulation by the same antigens present in pathogens leads to an anamnestic response. It is important to recognize that the immunity produced to pathogens, following active immunization, is only as broad as the antigenie spectrum of the preparation used for immunization. Protection can be afforded using different approaches in the formulation of vaccines. 4.3.3. Live Vaccines

These may be attenuated by passageof agent, e.g., viruses, in unusual hosts so that they become nonpathogenic to animals that are vaccinated. Usually these are good vaccines, since they supply the same antigenic stimulus as disease agent. There can be problems of reversion to pathogenic agent. Some replication of the agent usually occurs. 4.3.4. Modified

Vaccines-Whole-Disease

Agent

These can be grown and then chemically modified, e.g., heat-killed, nucleic acid-modified (mutagens), formaldehyde-treated. These are potentially good vaccines, in that a full antigenic spectrum is given. The antigenic mass has to be high, since there is no replication to challenge the immune system. Repeat vaccinations are common to elevate antibody levels. 4.3.5. Purified

Antigens

“Protective” antigens can be identified and used as protein, polypeptide, and peptide immunogens, usually with adjuvants. These vaccines are usually not as good as those in which the total antigenic spectrum is used. They have the advantages of being able to synthesize products on a large scale by chemical methods, e.g., as peptides, and are noninfectious. 4.3.6. DNA Technology

Products

Genes producing particular immunogens can be inserted into replicating agents, so that their products are expressed. This is a novel approach, e.g., in vaccinia, where more than one gene can be inserted.

21

Antibodies 4.3.7. Generalities

Most vaccines are administered by either subcutaneous or intrarnuscular injections. When vaccinating large herds of animals, other techniques, such as high-pressure jet injections, may be employed. It is obvious that the risk of administering unwanted/contaminating organisms/antigens should be minimal. Hence, sterile administration of vaccines is indicated. Subcutaneous or intramuscular vaccination should induce all antibody isotypes given the fact that the inciting antigens are capable of doing so. Hence, the immunoassayist must consider whether total antibody assays, isotype-specific assays, or assays to detect antigen clearance are to be utilized to assessthe effects of vaccination. Some antigens may be administered orally, e.g., poliomyelitis vaccines in humans, by incorporation in food or drinking water, e.g., in poultry flocks, or by inhalant exposure of aerosolized vaccine, e.g., diseases of respiratory tract. In these instances, the production of local antibodies to prevent the ingress of pathogens through the gastrointestinal or respiratory tree barriers is sought. In such situations, the immunoassayist must decide whether an assay for isotype-specific antibodies, notably IgA, may provide a deeper insight into the benefits of vaccination than an assay for total antibody. In some instances, where an infectious disease agent is endemic and vaccination, especially of newborns, is indicated, it may prove difficult or impossible to differentiate the beneficial effects of vaccination because residual levels of antibody may be present in nonvaccinated stock. Such factors must be borne in mind !when assays are developed to determine the immunological status of large groups of mammals. 4.4. Antibody Production in Response ‘to Infectious Agents It is not the function of this chapter to catalog the humoral immune responses produced in mammals in response to the variety of infectious disease agents, such as viruses, bacteria, fungi, protozoa, helminths, and arthropods. Such information may be obtained from textbooks and specialized review articles. This section will deal with the general considerations of the host-parasite relationships, with specific reference to the production of antibodies to pathogens. As already mentioned, most infectious diseases are transmitted by aerosolization, close contact, or vectors, and their final location may be

22

Basic Immunology

distant from their point of deposition. Therefore, these pathogens can have multiorgan involvement. Similarly, many pathogens, but not all, have the capacity to divide within the mammalian body, and in such instances, the numbers and amounts of antigens produced will increase over time and will be proportional to the number of pathogens at the time of sampling. Where pathogens do not divide or reproduce in the mammalian host, the amount of antigens produced may be directly proportional to the infective dose. Hence, in devising assays for infectious disease agents, the immunoassayist must take into account whether high concentrations of antigens and/or antibodies, or low concentrations of antigens and/or antibodies are to be sought. Where antibody titers of c 150 are anticipated, serum dilutions of cl:50 or possibly c 1:10 for the test serum must be employed. Previous knowledge of specific host-parasite systems will prove invaluable in devising more specific and sensitive enzyme immunoassays. Although many nonimmunological mechanisms exist for the removal of pathogens from the body (e.g., lysozyme, iron-binding proteins, myeloperoxidase, lactoperoxidase, compliment, basic peptides and proteins, and so on), it is generally recognized that the immune system plays a vital role in the control and destruction of pathogens. For this reason, the measurement of antibody or antigen by sensitive assays, such as ELISA, provides a useful indicator for the assessment of immune status. When an infectious agent enters the mammalian body, the first components recognized as foreign are surface components of that pathogen. This host-pathogen interface plays a vital role in the control of infectious diseases, not only in its involvement in stimulating the early humoral immune response, but also in its involvement in mediating protective immune responses.Immune responsesthat reduce pathogen numbers by lysis, agglutination and/or phagocytosis, or reduce the antigen load, and so forth, are normally regarded as protective responses, and such antibodies directed against specific epitopes on the pathogens can be sought by the immunoassayist in an effort to correlate protective responses with clinical betterment. However, insight into the molecular basis of such interactions is necessary before immunoassays can be developed to demonstrate protective responses (e.g., knowledge of the immunochemistry of the surface-exposed molecules and their epitopes, knowledge of specific antibody isotypes that mediate these responses). Since infectious disease agents stimulate antibody production, these

23

Antibodies

antibodies can prove useful to the immunoassayist for detecting exposure to pathogens. We have seen that when the antigens of pathogens are recognized by the host, an antibody response ensues, initially of the IgM isotype and followed by the IgG isotype, together with an increase in antibody affinity over time. Where a variety of antibody isotypes are produced in response to infection, use can be made of this isotypic variation in order to determine the chronicity of the infection, since IgM antibody isotypes normally appear before IgG antibody isotypes. Similarly, increasing levels of antibodies can indicate current infections or exacerbations of infections, whereas decreasing antibody titers can indicate past infections or successful control of current infection, In the absence of detectable free circulating antibody, either free antigen or circulating immune complexes can be detected by ELISA. When antibodies specific to antigens of a pathogen are used to detect the presence of free antigen in the test sample (see trapping/capture assays), a direct correlation can be made between ELISA positivity and current infection, Where protective mechanisms occur, destruction of the pathogen is the outcome. This is accompanied by the release of previously internal components, which, if antigenic, will stimulate the production of specific antibodies. Thus, the destruction of pathogens will lead to the production of antibodies against the antigen repertoire, both surface-exposed and internal, of that pathogen. Owing to the commonness of some internal antigens (e.g., enzymes, and so on), the consensus of opinion indicates that the more specific antigens of pathogens (excluding endotoxins) are surface-expressed at one time or another during development. The surface-exposed antigen mosaic is normally less complex than the internal antigen mosaic of pathogens. 4.4.1. Effect ofAntibody

in Viral Infection

Viruses as a group must enter a cell to proliferate, since they lack the biochemical machinery to manufacture proteins and metabolize sugars. Some viruses also lack the enzymes required for nucleic acid replication. The number of genescarried by viruses varies from 3 to about 250, and it is worth noting how small this is compared to the smallest bacterium. The illnesses caused by viruses are varied, and include acute, recurrent, latent (dormant but can recur), and subclinical. The immune response ranges from apparently nonexistent to lifelong immunity, The acute infection is probably most encountered by the immunoassayist interested

24

Basic Immunology

in animal diseases, but it must be borne in mind that the total knowledge of a specific disease is needed in order to devise assays relevant to specific problems. Since the outer surfaces (capsids) of viruses contain antigens, it is against these antigens and the envelope that the antiviral antibodies are mounted. The first line of defense (excluding interferon) is either IgM and IgG antibodies, where viruses are present in plasma and tissue fluids (vector transmitted), or secretory IgA antibodies, where viruses are present on epithelial surfaces (airborne, close contact). Some viruses that replicate entirely on epithelial surfaces (e.g. respiratory tree, gastrointestinal tract, genitourinary tract) and do not have a viremic phase will be controlled by secretory IgA. Antibodies may destroy extracellular viruses, prevent virus infection of cells by blocking their attachment to cell receptors, or destroy virus-infected cells. 4.4.2. Effect of Antibody

in Bacterial

Infection

The role of antibody in combating bacterial infection is diverse. Antibody to bacterial surface antigens (fimbriae, lipotechoic acid, and some capsules) prevents the attachment of the bacterium to the host cell membrane by blocking receptor sites. Antibody can neutralize bacterial exotoxins (possibly by blocking the interaction between the exotoxin and the receptor site). Normally, IgG antibodies are responsible for toxin neutralization. Antibody to capsular antigens can neutralize the antiphagocytic properties of the capsule, or in organisms lacking a capsule, antibodies to somatic antigens may serve a similar function. IgG anti
in Protozoan Infection

In general, antibodies serve to regulate parasites that exist in the bloodstream and tissue fluids, but are ineffective once the parasite has become intracellular. Hence, the importance of antibody varies with the infection under study. The presence of encysted morphological forms in the host

25

Antibodies

also reduces the efficacy of antibodies (e.g., Toxoplasma, Entamoeba, Giardia). Many protozoan parasites undergo development in the mammalian host, which is manifested in a morphologically distinct form. Such developmental forms often have associatedstage-specific surface antigens. Antibodies can damage parasites directly, induce lysis, activate complement, agglutinate extracellular forms, stimulate antibody, dependent cellular cytotoxicity, and block their entry into their host cells. IgM, IgG, and IgA antibody isotypes are involved in the above reactions. The isotype-specificity not only depends on which host compartment the parasite is resident in (e.g., respiratory tree, gastrointestinal tract, genitourinary tract [IgA], bloodstream, lymphoid tissue [IgM, IgG]), but also on the antigens expressed during the different developmental stages and their chronicity. Protozoan parasites can induce chronic infections in mammals and, therefore, large amounts of antibody (IgM + IgG) are produced in response to infection. The external surfaces, and the antigens contained therein, are important in the control of protozoan infections, and many effective/ protective antibody mechanisms are directed against them. Protozoan parasites are capable of the polyclonal stimulation of Bcells, and hence, in some infections (e.g., trypanosomiasis), large quantities of nontrypanosomal IgM and IgG antibodies are produced. In such instances, the immunoassayist should have the capability of distinguishing parasite specific from nonspecific antibody responses. Protozoan parasites are also capable of depressing the immune response (immunodepression), e.g., Babesia, and in such instances, circulating antibody levels are reduced. The above phenomena must be borne in mind when devising ELISAs for protozoan parasites. 4.4.4. Effect of Antibody

in Helminth

Infections

Helminths (trematodes, cestodes, nematodes, Acanthocephalans) normally have complicated life cycles, and their larger size and increased complexity cater to the increase in antigenic diversity found within them. In addition, where many developmental forms exist in the host, a stagespecificity of antigens (restricted to each stage) has been demonstrated. Helminth parasites induce chronic infections, and such long-term antigen challenge exposure can produce elevated levels of circulating antibodies. Since helminth parasites are normally transmitted by close contact, vectors, water, and possibly aerosolization, they affect numer-

26

Basic Immunology

ous tissues and organs of the host. Some helminths have a minimal migratory phase, but the majority are capable of migration throughout the host’s soft tissues. Antibodies of IgM, IgG, IgA, and IgE isotypes are produced in response to helminth antigens, and dependent on which physiological compartment of the host the parasite is resident in, each isotype can be utilized in ELISAs for detecting antibody recognition (exposure) of parasite antigens. IgE immunoglobulins are elevated in helminth infections, and although only a proportion of the IgE immunoglobulin produced has antibody activity against parasite antigen(s), this class of immunoglobulin has been implicated in resistance to helminth. Both IgG and IgE antibodies are involved in antibody-dependent cellular cytotoxicity mechanisms against helminth parasites. Helminth parasites are the most complex infectious disease agents infecting mammals, and hence, the host’s antibody responses to them is the most diverse. Unlike other infectious diseases,most helminths do not divide in the body of the final host, and hence, the antigenic load is dependent on the infective dose. Exceptions to this are Echinococcus spp. and Stongyloides spp. Antibodies are produced against antigens present within helminths (somatic antigens), antigens on helminth outer surfaces (surface antigens) and antigens in helminth excretions and/or secretions (excretory-secretory [ES] antigens). The ES antigens of helminths have been shown to confer the most specificity in immunoassays. Not all stages of the helminth life cycle occur in the mammalian host, and therefore, it is vital that the immunoassayist considers the pathogen of interest with reference to this. Antigens collected/extracted from stages of the parasite that do not occur in the final host are unlikely to be of use for detecting antibodies to the parasite in that host. Similarly, since helminths develop within the host, the earlier mammalian stages of development stimulate antibodies against antigens of that stage only. Later developmental stages stimulate antibodies not only to their respective stages, but also to previous stages if homologous or crossreacting antigens occur in such stages. It can be seen that the humoral immune responses to helminths are complex, and care is necessary in developing ELISAs for helminth infections in order that crossreactions are minimized. Helrninths are capable of modulating the immune response, and immunosuppression of antibody responses occurs in some diseases (e.g., Hemonchus) .

Antibodies

27

4.5. Overview on Antibody Production The complexity of the humoral response to infectious disease agents has been demonstrated in the previous section. It can be seen that as the agents become structurally more complex, a greater degree of antigenic diversity arises that can complicate immunodiagnosis. However, it is apparent that infectious disease agents stimulate the production of specific antibodies in hosts, and these antibodies can be used in ELISAs to determine the exposure of a host to a pathogen. Similarly, interactions between host and pathogen causes the release of pathogen antigens into the host’s bodily fluids. These antigens can be used in ELISAs to detect current infections. In endemic areas, mammals may be infected with more than one pathogen. Where antigens of each pathogen are noncrossreactive with antibodies produced in response to each pathogen, diagnosis by immunoassay is straightforward. Difficulties will inevitably arise where crossreacting antigens (possibly from closely related pathogens) and antibodies occur. Care must be taken when developing ELISAs for such systems, and basic research to define the problems of crossreactions must be undertaken. In considering antigens of potential usefulness in immunodiagnosis, it is apparent that the greaterthe number or variety of antigens used to detect circulating antibodies, the wider the antibody diversity detected will be, and hence, the greater likelihood of detecting positive cases. However, owing to the complexity and crossreactivity of antigens and antibodies, the immunoassayist must select the antigen or group of antigens that stimulate antibody production, but do not produce crossreactions. 4.6. Relationship Between Antigen and Antibody In Vivo We have seen that the ingress of antigens into the host’s body eventually leads to the production of circulating antibodies to them. In the case of complex antigens (e.g., infectious disease agent antigens), each antigen can stimulate the production of a variety of antibody isotypes with a variety of affinities, and these antibodies can be used in ELISA for the immunodiagnosis of infectious diseases. In certain infectious diseases (e.g., viruses, bacteria, protozoa) where the pathogens divide in the host’s body, as the pathogen burden increases, so does the antigenic load. Similarly, when pathogens are dealt with effectively (e.g., by lysis) the release of previously internalized antigens into the surrounding tissue occurs, thus increasing the free antigenic load in the host.

28

Basic Immunology

One function of antibodies is to mop up free antigen and to cause the antigen-antibody complex produced to be removed from the system, normally by phagocytic cells. It is important to consider the kinetics of antigen and antibody appearance and disappearance when dealing with antibody responses to infectious disease agents. In the early stagesof a primary infection, there is very little circulating antibody present. When the infectious disease antigens are processed by the host, antibody synthesis occurs, and these antibodies, predominantly of the IgM isotype, recognize specific antigens. In this stage of the disease process, antigen levels would be expected to be higher than antibody levels in the fluid sampled, i.e., antigen excess. Since there are relatively higher concentrations of antigen than antibody, all available antibody will be bound to antigen to form immune complexes, leaving the residual antigen free to circulate in bodily fluids. As more antibody is produced in response to continuous antigen challenge, the relative concentrations of antibody and antigen become similar, and as this antibody binds to antigen, less free antigen is present. At a point where the amount of circulating antibody equals the amount of free antigen, following antigen-antibody interaction to produce immune complexes, neither free antibody nor free antigen remains, and at this point, both reactants are at equivalence. Where the concentration of circulating antibody exceeds the concentration of circulating antigen, no free antigen will be present, but a residual amount of antibody will be present, i.e., antibody excess. This is a simplistic account of the interaction between antibody and antigen, but demonstrates that whenever specific antibodies to an antigen and that antigen interact, immune complexes are produced. The antibody isotype, the antibody affinity, the number of epitopes on an antigen molecule, and the chronicity of antigen and antibody production are all important in determining the fate of immune complexes. In general, large complexes, i.e., those found by interaction of antibodies and antigen with numerous epitopes, are removed by phagocytic cells. Smaller complexes (antibodies and antigen with few epitopes) are removed slowly by phagocytes and remain in the circulation for longer. The main site for immune complex removal is the liver (Kupfer cells), spleen, and lungs. Low-affinity immune complexes are smaller than high-affinity complexes, and therefore, persist longer in the circulation. Antigen excess and antibody excess immune complexes are normally smaller than anti-

Antibodies

29

gen-antibody equivalence immune complexes and therefore remain in the circulation longer. Some of the pathology associated with infectious diseases is the result of immune complex deposition. Thus, in certain instances, the concentration of soluble (circulating) immune complexes in bodily fluids can provide insight into immune complex pathology. The immunoassayist should be capable of devising assays for soluble immune complexes. For example, in antigen excess complexes, where an antigen-specific antibody (preferably from a different species to the host) is available, the antigen can be trapped by adsorbing the specific antibody onto the ELISA plate. The reaction can be developed with an antihost species-specific antibody enzyme conjugate. Since infectious diseases are often chronic, it would be expected that immune complexes would be produced owing to persistent antigen and antibody production, In the early stages,excessantigen would be expected, and hence, the immunoassayist could measurecirculating antigen. Remember that the antigen chosen for measurement has be to specific for the pathogen in question. As the disease state progresses to chronicity, less free antigen will be available owing to increased antibody production, and the immunoassayist should measure circulating antibodies or circulating immune complexes. It should be borne in mind that infectious disease organisms are antigenically complex, may or may not divide within the host, may reinfect the host, or may vary antigenically, and hence, the balance described above between antigen and antibody will vary from pathogen to pathogen, Since Ipathogens produce numerous antigens of different immunogenicity, many antigen-antibody interactions involving both low- and high-affinity antibodies will be present. The most useful assays will be those that take into account the above considerations. The pathogen-host interactions should be examined as fully as possible before ELISAs are developed. Finally, as already mentioned, many infectious disease agents have immunomodulatory effects varying from immunodepression, including reduction in humoral antibody production, to the induction of T- or B-cell tolerance. Tolerance, whereby an organism becomes unresponsive to a particular antigen, many occur in a variety of ways, but the outcome is that antibodies fail to be produced. Infectious disease agents can blockade receptors on antibody-forming cells or mature antigen-specific lymphocytes, and make them unresponsive to that antigen. Where high levels of antigen, such as

30

Basic Immunology

pneumococcal polysaccharide, induce the event, a state of high zone tolerance exists. Where low doses of monomeric antigen induce the event, a state of low zone tolerance exists. Thus, in designing ELISAs, the effect of parasite antigens on the induction of tolerance must be borne in mind. 4.7. Diagnostic and Antibodies

Usefulness of Antigens in Infectious Diseases

Detection of both antigens and specific antibodies can prove useful in the laboratory diagnosis of infectious diseases.As we have seen,the production of specific antibodies in response to antigen challenge is multifactorial. Not only antigens and antibodies, but also immune complexes can be detected by ELISA, and each will aid diagnosis. In general, free circulating antigen is present for shorter periods of time than free circulating antibody. Reinfection with the same pathogen or exacerbation will cause transient increases in free circulating antigen (antigenemia) following by increased production of circulating antibodies. Both these events can be detected by ELISA, and the immunoassayist must bear this in mind. Similarly, soluble immune complexes can be detected by ELISA, although the assay design is somewhat more complicated. Since antigens of infectious disease agents stimulate the production of specific antibodies, the latter can be used in ELISAs as an indicator of infection. Normally, the antibody response is long-lived and often is present in the absenceof the inciting antigens. For this reason, the detection of antibodies in the host does not indicate the presence of a current infection, but does indicate exposure. Antibodies to some antigens are more persistent than antibodies to other antigens, and in some instances, may persist for the lifetime of the individual. In such instances, the question as to whether the host is immune or refractory to reinfection cannot be answered by detecting antibodies to antigen mosaics of pathogens. Such questions as whether the residual antibody is effective in the prevention of reinfection can only be answered when the biological effects of these antibodies are known. Such effects as neutralization, agglutination, attachment onto receptors to prevent intracellular localization, and lysis are well-known biological effects of antibodies to infectious disease agents. If the mechanism of immunity is known, i.e., the effect of antibody on the target antigen(s), and if the target antigen(s) can be isolated, they can be used in ELISAs to monitor the rate of production and duration of protective antibody. Few

31

Antibodies

of these assays are available at present. Thus, the immunoassaysit must utilize other general phenomena associated with the development of the immune response. Some situations can be expanded. 4.7.1. Neonates

Maternal antibody of the IgG isotype is transmitted passively to neonates transplacentally and in the colostrum. Other isotypes will also be found in colostrum and milk (IgA, IgM), but are not transmitted across the neonatal gastrointestinal tract. Both IgM and IgG antibodies to antigen can be assayedfor, and where IgG antibodies arepresent in the absence of IgM antibodies, the likelihood of them being passively transmitted is high. Where humoral IgM antibodies occur, the organism must be synthesizing them de novo. If antibodies are assayed for the other physiological compartments of the body (e.g., gastrointestinal tract) the above conclusions become invalid, since both maternal IgM and IgA are secreted in milk and may still be biologically active as coproantibodies. 4.7.2. Immunologically

Competent Mammals

We have seen that IgM is the first isotype of humoral antibody produced in response to antigenic challenge, and that if the challenge persists, it is replaced by IgG isotype. If the infection if chronic and the antigen persists, or re-exposure occurs, higher levels of IgG will be produced. This information is valuable to the immunoassayist, becausewhen total antibody (all isotypes) is assayed, an increase in antibody titer over time is indicative of a current infection. This is also true when isotypespecific second antisera are used. The comparison IgM and IgG levels is also a useful indicator. Persistent antigen induces an isotype switch in T-cell-dependent antigens, stimulating the production of IgG rather than IgM. If IgM and IgG levels to the same antigen are compared, the following can be deduced: IgM antibody in the absence of IgG antibody means the recent acquisition of infection where no isotype switch has occurred as yet. The presence of IgM and IgG antibody means prior acquisition of infection where isotype switch has occurred, with IgM response present but possibly declining. IgG antibody in the absence of IgM antibody could mean the previous acquisition of infection, where isotype switch has occurred, and the IgM response is below the assay detection level. Here the infection would be expected to be chronic,

Basic Immunology

32

Where antigens are localized in the respiratory tract, the gastrointestinal tract, or genitourinary tract, stimulating local antibody responses, the detection of IgA antibody isotypes would be of value. Similarly, where it is known that pathogens stimulate isotype-specific responses (e.g., helminth), IgE-specific antibody responses can be assayed. 4.7.3. Herd Immunity

Each individual in a group will have a varying potential to respond to antigens, and hence, to mount a protective immune response. The range of immune responses in a group will follow a normal distribution pattern whereby the majority of individuals will mount an average immune response, and a small,proportion will mount either a very effective immune response of an ineffective immune response. Even where vaccination or exposure to a pathogen is concerned, less than 100% of the population will be adequately protected, and for those individuals concerned, this lack of protection could be serious. This lack of protection may be the result of the inability to recognize antigens or the inability to produce antibodies (agammaglobulinemia). These individuals can act as reservoirs of pathogens for future transmission to other susceptible individuals, and the seriousness of the situation will depend on the mode of transmission, the number of susceptible individuals, and where vaccination is common practice, the efficacy of vaccination. Less than 100% protection may be sufficient to prevent the spread of the disease within the population, since the likelihood of a susceptible individual encountering an infected individual becomes reduced. This phenomenon is known as herd immunity and is an important consideration in assessing the potential of individuals in a population to succumb to infection. There are many factors that can affect the quantity and quality of an immune response, most of which have been mentioned previously. Other factors that affect the immune response are stress, pregnancy, surgery, concomitant infections, extremes of temperature, and especially malnutrition. All these factors reduce the quantity and quality of the immune response, and when the immunoassayist performs seroepidemiological surveys to assessthe herd immunity between various populations, these factors must be taken into account before determining levels of adequate protection within a population.

Antigenic

33

Commonness

For each infectious disease, antibody levels (if any) in naive, subclinitally infected, clinically infected, and immune individuals in the endemic population should be sought. Only in this way can the immunoassayist relate antibody levels to infection. Once these parameters are known, the effect of vaccination in these individuals can be assayed, as well as the relationship between protective antibody and immunity. One major problem is that individuals in an endemic area may have had previous exposure of an infectious disease agent, or antigens thereof, and therefore ascribing an antibody titer or threshold above which protection occurs, or below which reinfection occurs, is difficult, if not impossible. The situation becomes more complex when assessing the protective effects of vaccination in such a group of individuals. Often individual antibody levels are static owing to chronic exposure and variable owing to previously mentioned factors. In the instances where naive individuals become immune, seroconversion occurs, but in the majority of instances where individuals have had previous exposure, increases in antibody levels are sought. These may be so small that they become statistically nonsignificant. 5. Antigenic

Commonness

Since there are a limited number of amino acids, carbohydrates, and so forth, produced by living organisms, the number of combinations of these components is limited. Therefore, it is likely that different organisms will produce similar proteins, glycoproteins, and so on. When partial or complete similarities occur in the products of living organisms, a commonness in terms of antibody recognition will occur. Structural proteins appear to be well conserved and would be expected to demonstrate immunological crossreactivity. The nature of this crossreactivity will vary between similar antigenic molecules and will depend on the ability of the paratope to fit the epitope. We have already seen that the paratopes of antibodies will bind to epitopes that are either recognized completely or partially recognized. The immunoassayist must take into account the possibility that the same or similar antigens may exist in a variety of infectious agents. For example, many cell-surface antigens are carbohydrate in nature, and the same epitopes may be present on two cell types, one of which may be pathogenic, whereas the other may be a commensal. Similarly, some carbohydrates may be present in widely differing infectious disease agents,

34

Basic Immunology

e.g., blood group-like ascarone antigens, and so on. When polyclonal antibodies are produced in response to infection, antibodies to noncrossreactive as well as to crossreactive epitopes will be produced, and hence, the problem of crossreactivity have to be minimized. The problem of antigenic commonness becomes more obvious when using MAbs in diagnosis. In the developmental stage of an assay, the immunoassayist must consider which other antigens, including antigens from other infectious disease agents, might be involved in crossreactions and develop the assay accordingly, bearing in mind the type of problems associated with such crossreactions. Reference 1. Roitt, I. M., Brostoff, St. Louis, MO.

J., and Male, D. K. (1993) Immunology,

3rd ed., Mosby,

CHAPTER2

Basic Principles

of ELISA

1. Reaction Schemes This chapter introduces the basic test formats for the performance of solid-phase heterogeneous ELISA. This form of ELISA has been so useful because: 1. Substances,e.g., antibodies or antigens, may be passively adsorbed to solid surfaces, such as plastics. Microtiter plates in a 96-well format are commercially available for use in ELISA, along with suitable equipment for easy manipulation and dispensing of reagents. This allows use of small volumes and gives the ELISA the potential of handling high numbers of samples rapidly. 2. Smce one of the reactants in the ELISA is attached to a solid-phase, the separation of bound and free reagents is easily made by simple washing procedures. 3. The result of an ELISA is a color reaction that can be observed by eye and read rapidly using specially designed multichannel spectrophotometers. This allows data to be stored and analyzed statistically. The use of passive adsorption also allows a great deal of flexibility in assay design. This chapter describes some basic test schemes. These assays are dealt with in later chapters at the practical level. Diagrams of the schemes are also included to reinforce the principles. ELISAs may be classified under four headings: direct, indirect, sandwich, and competition. In the following, the various ELISAs are represented using letters, as well as diagrams. Ag = Antigen Ab = Antibody directed against antigen AB = Antibody from another animal species as compared to Ab Anti-Ab = Species-specific antiserum, e.g., if Ab was raised in a mouse, antiAb is anttmouse serum *E = Enzyme attached to a particular antibody (Anti-Ab*E = antispecies antibody linked to enzyme, e.g., antimouse) 35

Basic Principles

of ELISA

I = Solid-phase to which reagent is attached passively S = Substrate addition and color development + = Addition of reagents and incubation Read = Measurement of color using spectrophotometer W = Separation of bound and free reagents by washing

2. Direct

ELISA

2.1. Direct-Labeled

I-Ag

+

Ab*E

W

+

Antibody

S + Read

W

Stage (i): Passive adsorption of antigen to plate by incubation in defined buffer. Stage (ii): Wash. Stage (iii): Addition and incubation of enzyme-labeled antibody. Stage (iv): Wash. Stage (v): Addition and incubation of color development system. Stage (vi): Read.

Antigen attached to the solid phase is reacted directly with an enzymelabeled antiserum. This has the disadvantage that sera raised against different antigens all have to be labeled. Thus, this is a poor assay if used to detect antigen from “crude” samples (containing a high concentration of contaminating substances), since low levels of antigen attach to wells owing to competition for plastic sites by such contaminants. This is a typical assay for use in the estimation of the titer of enzyme-labeled antispecies conjugates. This is illustrated in Fig. 1. Thus, the Ag in this case might be the IgG from an animal of the appropriate species against which the Ab*E is made. The interaction in this assay is made use of mainly in assays involving monoclonal antibodies (MAbs), which are directly labeled with enzyme. In this case, a very important epitope might be detected by the MAb, so that the labeling of a single antibody might be justified and form the basis of other assays as described below. 2.2. Direct-Labeled I-Ab

+ W

Ag*E

+

Antigen

S + Read

W

Stage (i): Passive adsorption of antibody to plate. Stage (ii): Wash.

Stage(iii): Addition and incubation of enzyme-labeledantigen.

Direct

ELISA

37

Fig. 1. Direct ELISA. Antigen is attached to the solid phase. After washing, enzyme-labeled antibodies are added. After an incubation period and washing, the substrate system is added and the color allowed to develop. Stage (iv): Wash. Stage (v): Addition and incubation of color development system. Stage (vi): Read.

Antibodies are adsorbed to the solid-phase usually after crude fractionation to obtain the immunoglobulin fraction (IgG). Enzyme-labeled antigens are then added and react with antibodies of suitable specificity. This has a poor applicability to diagnostic problems. Antigens are rarely labeled. However, a modification of this assay is used to measure progesterone, using competitive conditions. The scheme is illustrated in Fig. 2.

38

Basic Principles

of ELISA

Fig. 2. Direct-labeled antigen. Antigen is labeled with enzyme and can be captured with antibodies attached to the solid phase. This system can form the basis of other assays,e.g., competitive techniques where either antibody or antigen can be incubated with the labeled antigen and the degree of inhibition of binding of the labeled antigen with the solid phase measured. This form of assay has been used in the estimation of hormone concentrations and is analogous to many radioimmunoassay methods. 3. Indirect I-Ag

ELISA

+ Ab + Anti-Ab*E + S + Read W W W Stage (i): Passive adsorption of antigen. Stage (ii): Wash. Stage (iii): Addition of antibody directed against Ag. Stage (iv): Wash. Stage (v): Addition of enzyme-labeled antispecies antibody. Stage (vi): Wash.

Sandwich ELISA

39

Stage(vii): Addition of color developmentsystem. Stage(viii): Read. This is extensively used for the detection and/or titration of specific antibodies from serum samples. The specificity of the assay is directed by the antigen on the solid-phase, which, may be highly purified and characterized or relatively crude and noncharacterized. After addition and incubation of the antigen, the wells are washed to get rid of unbound antigen. Serum containing antibodies against this antigen can then be added and diluted in a buffer that prevents the nonspecific adsorption of protein for any free sites on the solid-phase not occupied by the antigen (blocking buffer). Sera, may be added as a single dilution (common in epidemiological testing of large number of sera)or as a dilution range. After incubation, the wells are washed to get rid of unbound antibodies. Bound antibody is then detected after incubation, with a single dilution of antispecies antibody conjugated to an enzyme. This is diluted in blocking buffer. The amount of specific antibody binding to the antigen is quantified after addition of color development reagents (enzyme substrate or substrate/ dye combination). Such assaysoffer immediate advantagesover the direct tests since only a single antispecies enzyme conjugate is needed to titrate antisera from many animals of a single species. The scheme is shown diagrammatically in Fig. 3. 4. Sandwich ELISA 4.1. Direct Sandwich I-Ab

+

W

Ag

W

+

AB*E

+

S + Read

W

Stage (i): Passive adsorption of antibody. Stage (ii): Wash. Stage (iii): Addition of antigen. Stage (iv): Wash. Stage (v): Addition of enzyme-labeled antIbody* against antigen. Stage (vi): Wash. Stage (vii): Addition of color development system. Stage (viii): Read.

Enzyme-labeled antibody can be produced in the same animal (can be the same serum) that produced the passively adsorbed antibody, or from a different species immunized with the same antigen that is captured.

40

Basic Principles

of ELISA

Fig. 3. Indirect ELISA. Antibodies from a particular speciesreact with antigen attached to the solid phase. Any bound antibodies are detected by the addition of an antispecies antiserum labeled with enzyme. This is a widely used system in diagnosis. This is similar to Section 2.2., except that antibody is attached to the solid-phase, usually as an IgG fraction of the whole serum. A constant dilution of antibody is attached to the solid phase, and after incubation, unadsorped antibody is washed away. Antigen at a single dilution or as a dilution range is then added, in a buffer that prevents nonspecific bind-

Sandwich

ELBA

ing of the serum proteins to any available plastic sites. After incubation, unbound antigen is washed away. Bound antigen is then detected by the addition of enzyme-labeled antibody specific for the “trapped” or “captured” antigen. This antibody can be: 1. The same as that used on the solid phase. 2. Produced in the same species as the trapping serum. 3. Produced in a different speciesto that in which the trapping antibody was made.

After incubation and washing away of unreacted conjugate, the color detection system is added and color is measured. Note that certain anti-

gen preparations cannot be directly attached to microplates, since they are at low concentration and/or they are contained in high concentrations of contaminating protein. The antibody attached to the plastic should bind antigen specifically, so that selection and concentration of the antigen takes place in the sandwich conditions. Such assayshave been described

as capture or trapping assays,referring to the property of the bound antibody to bind the antigen to the plastic surface. Note also that antigens must contain at least two antigenic sites, capable of binding to antibody, since at least two antibodies act in the sandwich. This is important where lowerTmo1 wt antigens are being used of limited antigenic potential, or where antigenic sites are concentrated on one surface. Some difficulties can be encountered using the same MAb in sandwich assays, since the capture step may bind to the only epitope expressed on a “small” antigen. As in the direct ELISA, this test has the disadvantage that all the detecting antisera have to be conjugated. Figure 4 shows

the scheme diagrammatically. 4.2. Indirect I-Ab

+ w

Ag

+ w

AB

Sandwich +

w

Anti- AB *E + W

S + Read

Stage (i): Passive adsorption of Ab. Stage (ii): Wash. Stage (iii): Addition of antigen. Stage (iv): Wash. Stage (v) Addition of antibody from different species vs antigen. Stage (vi): Wash. Stage (vii): Addition of enzyme-labeled antispecies (directed against AB).

42

Basic Principles

of ELISA

Fig. 4. Sandwich ELISA-direct. This systemexploits the antibodies attachedto the solid phase to capture antigen. This is then detectedusing an enzyme-labeled serum specific for the antigen. The detecting antibody is labeled with enzyme. The capture antibody and the detecting antibody can be the same serum or from different sources. The antigen must have at least two different antigenic sites. Stage (viii): Wash. Stage (ix): Addition of color development system. Stage (x): Read. Figure 5 shows the scheme diagrammatically. This is the same as Section 4. l., except that the second antibody is produced in a different species from the trapping antibody. Thus, the second antibody can be detected using a species-specific antiserum con-

Competition

ELISA

43

Fig. 5. Sandwich ELBA-indirect. The detecting-antibody is from adifferent species than the capture antibody. The antispecies enzyme-labeled antibody binds to the detecting antibody specifically and not to the capture antibody. jugate that does not react with the antibody on the plastic. The advantage is that many second antibodies, may be titrated with a single conjugate.

5. Competition ELISA Competition assays imply that two reactants are trying to bind to a third. Proper competition assays involve the simultaneous addition of the two competitors.

44

Basic Principles

of ELISA

Competition

A Incubate

Inhibition/Blocking

B

Wash

+

Incubate

Incubate

C Incubate

Incubate No wash

Fig. 6. Comparison of competition and inhibition ELISA. The test scheme involves the reaction of two antibodies with an antigen attached to the solid phase. Where one of the antibodies is incubated first, the assaysare called blocking or inhibition assays.Competition implies simultaneous addition of reagents.

Inhibition or blocking assays are similar, except that one of the reagents being examined for “competing” ability of a system is added and incubated before the second “competitor” is added. This can be illustrated simply as shown in Fig. 6. 5.1. Direct I-Ag

+ Ab*E

+ W

W +AB

Antibody S + Read

Competition

Competition

ELISA

45

Stage (i): Passive adsorption of antigen. Stage (ii): Wash. Stage (iii): Addition of competing antibody at various dilutions. Stage (iv): (Optional washing step after incubation with AB alone) Stage (v): Addition of enzyme-labeled antibody (pretitrated to give optimal color development). Stage (vi): Wash. Stage (vii): Addition of color development system. Stage (viii): Read.

assays are defined as assays where the detecting (pretitrated labeled) antibodies are added with the sample simultaneously. Competition

When the sample is added for a period of incubation before the pretitrated antibodies, a blocking assay or inhibition assay results. The blocking assay can involve incubation of the test sample followed

by washing and then addition of the labeled antibodies or their addition without washing away nonbound antibodies from the sample. Direct antibody competition involves the adsorption of antigen to the

solid phase. After washing away unadsorbed antigen, a specific antibody labeled with enzyme is added. This is pretitrated so that the antigen is saturated, and no free antigenic sites are available for further antibody combination.

This interaction of antigen and pretitrated antibody (identical to the direct ELBA) is upset if the labeled antibody is mixed with another antibody (competing antibody) that is able to react with the solid-phase bound antigen. The competing antibody is added as a dilution range. Thus, this replaces all or some of the pretitrated conjugated antibody. After washing and development of the assay, the replacement is observed

as a decrease in the color expeated (that found in control wells containing no competing serum). Such assays are of increasing importance particularly where MAbs are used. Note that any serum from any species can be used as competitor. The scheme is illustrated in Fig. 7. 5.2. Direct I-Ag

Antigen

+ Ab*E W

+ S + Read W + Ag as dilution range

Stage (i): Passive adsorption of antigen, Stage (ii): Wash.

Competition

46

Basic Principles Pre-titration

I

of antigen

and labelled

of ELISA

antibodies

Competition-addition of serum containing common to pretitrated conjugate

antibodies

Addition of samples with antibodies in common with pre-titrated labelled serum will bind to same antigenic sites E

> c >

Pre-titrated enzyme-labelled antibodies are blocked by the test antibodies reacting with E common antigenic sites

Where all sites are blocked there is no colour development since no enzyme-labelled antibodies are attached to antigen.

Fig. 7. Competition ELISA-direct antibody. The degree of inhibition by binding of antibodies contained in a serum for a pretitrated enzyme-labeled antiserum reaction is determined. Stage (iii): Simultaneous incubation of free antigen with enzyme-labeled antibody (pretitrated): directed against antigen on plastic. Stage (iv): Wash. Stage (v): Addition of color development system. Stage (vi): Read.

Competition

47

ELISA Pre-titration

Competition

of labelled

antibodies

with sample possibly

Addition of antigen with same antigenic sites as antigen on solid phase

and antigen

containing

same antigen(s)

Addition of antigen with no common antigen sites with solid phase antigen

No colour 4 100°AOCompetition i-00 “2 ’o@00000@

0%

Competition 0

0

Fig. 8. Competition ELBA-direct antigen. Reaction of antigen contained in samples with the enzyme-labeled antibody directed against the antigen on the solid phase blocks its binding to the solid phase. If the antigen has no crossreactivity with the solid-phase antigen, then the labeled antibody binds, and a color reaction is observed. This is as described in Section 5. l., except that the competing substance for the pretitrated conjugated antibody is antigen. If the labeled antibody reacts with the dilution range of added antigen (competitor) in the liquid phase, it is washed away after the incubation step. Thus, labeled antibody is unavailable to react with the solid-phase antigen, and a reduction in expected color is observed. Such assays can be used to quantify antigens or to compare the relative affinity of binding of two antigens for the same serum. The scheme is illustrated in Fig. 8.

Basic Principles

48 5.3. Indirect I-Ag

+

Ab

+

Antibody

Competition

Anti-Ab*E

w +AB

w

of ELISA

+ S + Read w

Stage (i): Passive adsorption of antigen. Stage (ii): Wash. Stage (iii): Addition of test antibody AB at various dilutions. Stage (iv)*: (Optional washing step after incubation with AB alone). Stage (v): Addition of antibody (pretitrated): standard serum. Stage (vi): Wash. Stage (vii): Addition of antispectesconjugate against standard antiserum (Ab). Stage (viii): Wash. Stage (ix): Addition of color development system. Stage (x): Read. This is essentially the same as the indirect ELISA, except that a competing antibody is added to the solid-phase antigen either before or simultaneously with pretitrated specific antibody. The level of antibody used is usually about 70% maximal reactivity (solid-phase antigen excess). The competing antibody must be from a different species from the pretitrated antibody, since the antispecies conjugate must not react with both. If the competing antibody is able to bind to the antigen, then it prevents the pretitrated antibody reacting, and this is observed as a decrease in the expected color as compared to controls without competitor. The scheme is illustrated in Fig. 9. 5.4. Indirect I-Ag

+ W

Ab

+ W

Antigen

Competition

Anti-Ab*E

+ S + Read W

+Ai3 Stage (i): Passive adsorption of antigen. Stage (ii): Wash. Stage (iii): Simultaneous incubation of free antigen (test sample): with antibody directed against antigen on plastic at pretitrated dilution. Stage (iv): Wash. Stage (v): Addition of enzyme-labeled antibody against Ab. Stage (vi): Wash. Stage (vii): Addition of substrate. Stage (viii): Read.

Competition

ELBA

49 Pre-titration

Competition-addition

of indirect system

of samples containing antibodies?

Serum contains antibodies which bind to antigen &,

Serum contains NO antibodies which bind to antigen

These block pre-titrated antibodies binding

4 On addition of anti-species enzyme labelled conjugate A

Conjugate does not bind

No color COMPETITION

Conjugate binds

Color NO COMPETITION

Fig. 9. Competition ELISA-indirect antibody. A pretitrated system for antibody binding to antigen is challenged by the addition of another serum (test) sample. If antibodies bind to the sites in common with the pretitrated antibodies, they block (if added before pretitrated antibody) or compete with (if added simultaneously) this reaction. Since an antispecies conjugate is used, the competing sample serum cannot be from the same species. This is an indirect ELISA where antibody is pretitrated against the solidphase bound antigen by the use of antispecies conjugate, which is challenged by the addition of dilution ranges of antigen in the liquid phase. Again, the amount

Basic Principles

50 Petitration

Addition of same or similar to that on solid phase

of indirect

antigen

of ELISA

system

Addition of antigen on solid phase

different

to that

Wash and add conjugate

Add substrate

No color-100%

competition

Fig. 10. Competition ELBA-indirect antigen. The pretitrated indirect ELISA is competed for by antigen. If the antigen shares antigenic determinants with that of the solid-phase antigen, it binds to the pretitrated antibodies preventing them from reacting with the solid-phase antigen. If there is no similarity, the antibodies are not bound and can react with the solid-phase antigen. Addition of the antispecies enzyme conjugate quantifies the bound antibodies. of pretitrated antibody should be about 70% of the maximal reaction (solidphase antigen excess). Competition is reflected by a decrease in the expected color obtained without competitor. The scheme is illustrated in Fig. 10. 6. Choice of Assays The most difficult question to answer when initiating the use of ELISAs is which system is most appropriate. This section will attempt to

51

Choice of Assays

C

D

Fig. 11. Basic ELISA methods. (A) Direct, (B) Indirect, (C) Sandwich (direct), (D) Sandwich (indirect) investigate the relationships between the various systems to aid in assessing their suitability. Questions that must be addressed are: 1. What is the purpose of the assay? 2. What reagents do I have? 3. What do I know about the reagents? 4. Is the test to be developed for a research purpose to be used by me only, or for applied use by other workers? 5. Is the test to be used in other laboratories? 6. Is a kit required? These questions have a direct affect on the three phases that might be put forward as a general rule for the development of any assay, i.e.: 1. Feasibility-proof that a test system(s) can work. 2. Validation-showing that test(s) is “stable” and that it is evaluated over time and under different conditions. 3. Standardization-quality control, establishment that the test is precise and can be used by different workers in different laboratories. Figures 11-15 summarize the exploitation of ELISA methods, highlighting the relationship of assays to the relative purity and concentration of reactants, and indicating the use of direct and indirect methods for competition assays. These will be relevant when examining the possibilities outlined in Section 6.1.

Basic Principles

52

of ELISA

*ENZ t

A

+ENZ t s-

B

Fig. 12. Competition for direct ELISA. 6.1. Phase

1 in Developing

an ELISA

Feasibility

Phase 1 involves the trial of various systems of ELISA with existing and newly prepared reagents to be able to obtain the desired aim. This phase includes identification of needs based on preliminary experiments and a good knowledge of the biology of the system. The latter point may become more important when attempts at using ELISA fail because of lack of knowledge. Thus, as an example, we may wish to estimate the antibody titer in cattle sera against a particular antigen. The possibility of performing all the ELBA systems and obtaining the most appropriate system will depend on the availability of various reagents and their specificities.

Choice of Assays

53

INDIRECT

A

Fig. 13. Competition for indirect ELBA. 6.1.1. Assessing What Is Available As examples: We may have only the relevant antigen. Figure 16 shows different types of antigens of ,varying complexity. Thus, we may know a great deal about the antigen or very little. We may have a high concentration of a defined protein/polypeptide/peptide of known amino acid sequence or have a thick soup of mixed proteins containing the antigen at a low concentration contaminated with “host cell” proteins. We may have an antiserum against the antigen. This could be against purified antigen or against the crude soup. The antibody may have been raised in a given species, e.g., rabbit. We may have an IgG fraction of the antiserum (or

54

Basic Principles

of ELISA

AG COMPETITION

AB COMPETITION

+

0

+#

+ YI I, 2 d 3 ALTRRNATNE

Yf

Y llbii?S

+o

FOR ADDI’KION

Fig. 14. Competition for sandwich ELISA-same detection.

OF COMPETITORS

antibody for capture and

could easily make one). We may have field sera against the antigen (bovine sera). We may have an MAb. We may have antisera from different species, e.g., rabbit and guinea pig sera. ELISAs for similar systems may have been developed and can be found in the literature. We will require an enzymic reaction in the assay. Thus we will need an antispecies conjugate (commercial, most probably) or will have to label an antigen-specific serum with enzyme (facilities to do this?). We have to decide which commercial conjugate to buy. This will depend on the desired specificity of the conjugate (antiwhole molecule IgG, anti-Hchain IgG, anti-H chain IgM, and so forth). The choice is somewhat determined by the aims of the assay and its design. Thus, we may wish to determine the IgM response of cattle to our antigen, which will require an anti-IgM (specific) somewhere in the ELISA protocol. Obviously the basic needs for performing the ELISA must be addressed in terms of plates, pipets, buffers, reader, and so on.

55

Choice of Assays I &3ENZ +s-+

*ENZ

AG COMPETITION

I,2 & 3 ALTERNATIVE

TIIUES FORADDITION

OF COMPET1mR.S

J

AB COMPETITION

Fig. 15. Competition for sandwich ELBA-different and detection.

antibody for capture

6.1.2. Examination of Possible Assays with Available Materials Obviously, the reagents available must be examined first as previously stated. This section will deal with some extremes in order to illustrate the relationship of the assays available and their particular advantages. As for Section 6.1.1.) perhaps we have to examine the level of antibodies in bovine serum. Some scenarios are described with different available reagents. These will probably cover most of those that are met in practice. This assumes that there are sera to test from infected and noninfected animals. Further subtleties can be examined by defining the specificities of the conjugates (anti-IgG, IgM, or whether they are H-chain-specific). The increase in choice of reagents and the possibilities for performing different ELBA configurations follow: 1. a. Crude antigen (multiple antigenic sites) b. Antibody raised against crude antigen in rabbits c. Anticow conjugate

Basic Principles

56

Large multivalent antigen, sequential and conformational

Smaller multivalent

of ELISA

epitopes

antigen

Smaller antigen, univalent. Polypeptide, linear sequential epitopes. Polypeptide, sequential and conformational

epitope.

Peptide. Peptide, linked to carrier protein with conformational epitope.

Fig. 16. Different forms of antigens for use in ELBA and for antiserum production. These antigens could be contaminated with “host cell” proteins. d. Postinfected and d 0 (uninfected) cow sera 2. a. Purified antigen (small amount, e.g., 100 pg) b. Crude antigen (large amount) c. Antibody raised in rabbits against pure antigen d. Antirabbit conjugate e. Anticow conjugate f. Postinfected and d 0 (uninfected) cow sera 3. a. Crude antigen (as in 1.) b. Antibody against pure antigen (rabbit) c. Antibody against pure antigen (guinea pig) d. Antiguinea pig conjugate e. Postinfected and d 0 (uninfected) cow sera f. Anticow conjugate g. Antirabbit conjugate.

Choice of Assays

57

6.1.2.1. SITUATION 1 Here the use of crude antigen directly on an ELISA may well be unsuccessful, since it may be at a low concentration compared to other proteins and thus only attach at a low concentration. This does not allow the ELISA approaches as shown in Fig. 1 lA,B and, thus, competitive methods based on these, as in Figs. 12B and 13B. Since a rabbit serum against the antigen is available, this may be used as a capture serum (or a capture IgG preparation) coated on wells to capture the crude antigen to give a higher concentration of antigen to allow the binding of antibody as in Fig. 1lC,D. This also allows competitive techniques as shown in Fig. 15B. The bound antibody would be from cows and would be detected using the antibovine conjugate. There may be problems, since the crude antigen was used to raise the rabbit serum. Thus, antibodies against the contaminating proteins may be produced in the rabbit. The cow sera being tested may react with such captured contaminants. However, where the antigen is an infectious agent, antibodies against the contaminating proteins may not be produced, thus eliminating the problem. Where the antigen is used as a vaccine, whereby relatively crude preparations similar to the crude antigen are used to formulate the vaccine, then this problem will be present. Attempts can be made to make the rabbit serum specific for the desired antigenic target. Solid-phase immunosorbents involving the contaminating crude elements (minus the desired antigen) can be used to remove the anticrude antibodies from the rabbit serum, which could then be titrated as a capture serum. An example can be taken from the titration of foot-and-mouth disease virus (FMDV) antibodies. The virus is grown in tissue culture containing bovine serum. Even when virus is purified from such a preparation, minute amounts of bovine serum contaminate the virus. When this “purified” virus is injected into laboratory animals as an inactivated preparation, there is a large amount of antibovine antibodies, as well as antivirus antibodiesproduced.This serum cannot be used in a capture system for specifically detecting virus grown as a tissue-culture sample (containing bovine serum), since it also capturesbovine serum. The capture serum is also unsuitable for capturing relatively pure virus for the titration of bovine antibodies from bovine serum samples, since the capture antibodies react strongly with the detecting cow serum. Thus, the capture serum has to be adsorbed with solid-phase immunosorbents, e.g., those produced through the attachment of bovine serum to agarosebeads,

Basic Principles

58

of ELISA

Solid Phase

Coating with Guinea pig IgG

1

Wash

Addition

1

of anti-guinea

pig IgG conjugate

1 Wash

Addition

1

of substrate/chromogen

Scheme 1.

Once the specificity of the capture serum is established, the optimization of the crude antigen concentration can be made using a known or several known positive cow sera using full dilution ranges. Inclusion of dilution ranges of negative sera allows an assessment of the difference between negative and positive sera at different dilutions of serum. The diagram below illustrates the use of the reagents to set up a sandwich ELISA. The assay is made possible through the specific capture of enough antigen by the solid-phase rabbit serum as in Fig. 11D. I-Ab + Ag + AB + Anti-AB*Enz Rabbit Crude Cow Anticow

+ Substrate + Read

6.1.2.2. SITUATION 2 This situation is not very different from the first. However, we have more reagents! We have the antigen purified and used to raise antibodies in rabbits (see Scheme 1). Thus, with due reference to the reservations

Choice of Assays

59

already described in Section 6.1.2. l., we have the basis of setting up a capture ELISA, since the rabbit antibodies may capture the antigen at a high concentration from the crude antigen, which we have in a large amount. The development of the capture ELISA as shown in Fig. 1 lC,D is as described above. The availability of the antirabbit conjugate may allow development of competitive assays if enough specific antigen binds to plates, although this is unlikely, as indicated above. The antigen and rabbit serum could be titrated in an indirect ELISA (Fig. 11B) in a chessboard fashion, enabling the optimization of the antigen and serum. These optimal dilutions could be used to set up competitive ELISAs, whereby cow sera would be competed for the pretitrated antigen/rabbit/antirabbit conjugate system, as in Fig. 13B. Again it must be emphasized that this is unlikely since the antigen is crude and some form of capture system will be needed to allow enough antigen to be presented on the wells. Since this situation has some purified antigen, this could be used in the development of a similar competitive assay. This will depend on the availability of this antigen, which can be determined after the initial chessboard titrations where the optimal dilution of antigen is calculated, The chief benefit of obtaining purified antigen was to obtain a specific serum in rabbits, allowing specific capture of antigen from the crude sample. In many cases, there is enough antigen of sufficient purity to be used in such assays. Another alternative, as shown in Fig. 14B, is available if the rabbit serum can be conjugated with enzyme. 6.1.2.3. SITUATION 3 Here we have all the possibilities of the first two situations plus the production of a second species (guinea pig) serum against the purified antigen (seeScheme2). This allows the development of competitive assays as in Fig. 15B using either the rabbit or guinea pig as capture serum or detector with the relevant antispecies conjugate. Different species may have better properties in acting as capture reagents and also show varying specificities. This can be assessed in chessboard titrations. This is relevant since we require results on the detection and titration of cattle sera so that the competitive phase relies on the interruption of a pretitrated antibody as close to the reaction of cattle serum with antigen as possible. The rabbit or guinea pig serum may differ in their specificities as compared to cattle sera.

60

Basic Principles

of ELISA

Solid Phase

Coating with dilution of sheep anti-Guinea pig IgG Wash

Addition of guinea pig IgG Wash

Addition of rabbit anti-guinea pig serum

Wash Addition of sheep anti-rabbit enzyme conjugate

Wash

Addition of substrate/chromogen

Scheme 2.

6.1.2.4.

FURTHER

COMMENTS

Assays shown in Fig. 12B (competition for direct ELISA) are probably inappropriate owing to the possession of crude antigen (ior reasons described above). However, if it can be shown that enough antigen can attach and that cattle sera react specifically (and not through excess antibodies directed against contaminants in the crude antigen), then we can

Choice of Assays

61

set up assays based on this system. This requires identification of a positive cow serum and labeling of this serum with an enzyme. Of more practical value could be the use of positive cow serato develop a system as shown in Fig. 14B. Here a positive cow serum, as identified from other tests (or samples of different positive cow sera), can be selected and labeled with enzyme. The serum can then be used for both capture (particularly as an IgG fraction) and detection. In this way, the competitive assay shown in Fig. 14B is feasible and may have an advantage in that the reaction being competed against is homologous (cow antibody against antigen), thus avoiding any complications of the use of second species antisera produced by vaccination. The system is suitable to measure the competition by other cow sera, since the detecting antibody is labeled. Thus, a worker with relatively few reagents and the ability to label antibodies with an enzyme may have enough materials to develop assays. In this brief description of system possibilities, we have concentrated on antibody detection. It should be noted that most of these comments are relevant to antigen detection.

CHAPTER3

Stages in ELISA This chapter gives general information on common practical features of the ELISA, featuring the main elements of: 1, The adsorptionof antigenor antibody to the plastic solid-phase. 2. The addition of the test sampleand subsequentreagents. 3. The incubation of reactants. 4. The separationof bound and free reactantsby washing. 5. The addition of enzyme-labeledreagent. 6. The addition of enzymedetectionsystem(color development). 7. The visual or spectrophotometricreading of the assay. 1. Solid-Phase

By far the most widely used solid-phase is the 96-well microtiter plate manufactured from polyvinyl chloride (PVC, flexible plates) or polystyrene (inflexible “rigid” plates). Many manufacturers supply plates designed for ELISA and provide a standardized product. The use of a wide variety of plates from different manufacturers has been reported for a broad spectrum of biological investigations. It is impossible to recommend one product as a universally accepted plate. Where specific assays have been developed, it is prudent to use the recommended plate. However, since there is, in practice, relatively little difference between plates, it is possible to perform the sametest using different plates provided that suitable standardization is performed. In this respect, laboratories that deal with large numbers of EIJSAs involving different antigens and antibodies can perform standardized assaysusing the same type of plate. Ideally, flat-bottomed wells are recommended where spectrophotometric reading is employed to assesscolor development. However, round-bottomed wells can be used where visual (by-eye) assessmentof the ELISA is made. Such plates can be read by spectrophotometer, but are not ideal. 1.1. Immobilization of Antigen on Solid-Phase-Coating A major feature of the solid-phase ELISA is that antigens or antibodies can be attached to surfaces easily by passive adsorption, This process 63

Stages in ELISA

is commonly called coating. Most proteins adsorb to plastic surfaces,probably as a result of hydrophobic interactions between nonpolar protein substructures and the plastic matrix. The interactions areindependent of the net charge of the protein, and thus, each protein has a different binding constant. The hydrophobic&y of the plastic-protein interaction can be exploited to increase binding, since proteins have most of their hydrophilic residues at the outside and most hydrophobic residuesorientated toward the inside. Partial denaturation of some proteins results in exposure of hydrophobic regions and ensures firmer interaction with the plastic. This can be achieved by exposing proteins to low-pH or mild detergent, and then dialysis against coating buffers before coating. The rate and extent of the coating depend on: 1. The diffusion coefficient of the attachingmolecule. 2. The ratio of the surface areabeing coated to the volume of the coating solution. 3. The concentrationof the substancebeing adsorbed. 4. The temperature. 5. The time of adsorption. These factors are linked. It is most important to determine the optimal antigen concentration for coating in each system by suitable titrations. A concentration range of l-10 l.tg/mL of protein in a volume of 50 PL is a good guide to the level of protein needed to saturate available sites on a plastic microtiter plate. This can be reliable where relatively pure antigen (free of other proteins other than the target for immunoassay) is available. Thus, the concentration can be related to activity. However, where coating solutions contain relatively small amounts of required antigen(s), the amount of this attaching to a well is reduced according to its proportion in the mixture. The other contaminating proteins will take up sites on the plastic. Since the plastic has a finite saturation level, use of relatively crude antigens for coating may lead to poor assays. Care must be taken to assesseffects of binding proteins at different concentrations, since the actual density of binding may affect results. High-density binding of antigen may not allow antibody to bind through steric inhibition (antigen molecules are too closely packed). High concentrations of antigen may also increase stacking or layering of antigen, which may allow a less stable interaction of subsequent reagents (I). Orientation and concentration of antibody molecules must also be considered. Figure 1 illustrates the possible effects on assays.

Solid-Phase

65

I

Fig. 1. Effects on antibodies of coating. (A) Antibody molecules packed evenly, orientation Fc on plate, monovalent interaction of multivalent Ag. (B) Antibody molecules packed evenly, orientation Fc and Fab on plate, monovalent binding of multivalent Ag. (C) Antibody binding in all orientations, monovalent binding of multivalent Ag. (D) Antibody binding via Fab, no binding of Ag. (E) Antibody spaced with orientation to allow bivalent interaction between adjacent antibody molecules. (F) Antibody spaced too widely to allow adjacent molecules to bind bivalently via Fc. (G) As in (E) except that orientation is via Fc or Fab. (H) More extreme caseof(C) with less antibody and more molecules inactive because of their orientation. (I) Multilayered binding in excessleads to binding, but elution on washing.

66

Stages in ELISA

1.2. Coating Time and Temperature The rate of the hydrophobic interactions depends on the temperature. The higher the temperature, the greater the rate. There are many variations on incubation conditions. It must be remembered that all factors affect the coating. Thus, a higher concentration of protein may allow a shorter incubation time as compared to a lower concentration of the antigen for a longer time. The most usual regimens involve incubation at 37°C for l-3 h, overnight at 4OC, or a combination of the two, or incubation (more vaguely) at room temperature from l-3 h, see ref. 2 for a typical study. There are many more variations. Ultimately each scientist has to titrate a particular antigen to obtain a standardized regimen. Increasing the temperature may have a deleterious effect on antigen(s) m the coating stage, and this may be selective, so that certain antigens in a mixture are affected, whereas others are not. Rotation of plates can considerably reduce the time needed for coating by increasing the rate of contact between the coating molecules and the plastic. 1.3. Coating Buffer The coating buffers most used are 50-mm carbonate, pH 9,6, 20 mM Tris-HCI, pH 8.5, and 10 rnMPBS, pH 7.2 (2). Different coating buffers should be investigated where problems are encountered or compared at the beginning of assay development. From a theoretical view, it is best to use a buffer with a pH value l-2 U higher than the p1 value of the protein being attached. This is not easy to determine in practice, since antigens are often complex mixtures of proteins. By direct study of the effects of different pHs and ionic strengths, greater binding of proteins may be observed. An increase in ionic strength to 0.M NaCl in combination with an optimal pH was found to give better results for the attachment of various Herpes simplex viral peptides (3). Proteins with many acidic proteins may require a lower pH to neutralize repulsive forces between proteins and the solid-phase as shown in (3), where the optimal coating for peptides was pH 2.5-4.6. Phosphate-buffered saline, pH 7.4, is also suitable for coating many antigens. Coating by drying down plates at 37°C using volatile buffers (ammonium carbonate) and in PBS is often successful, particularly where relatively crude samples are available. Some antigens pose particular problems. These include some polysaccharides, lipopolysaccharides, and glycolipids. Where it proves impossible to coat wells directly with reagent, initial coating of the well with a specific antiserum

67

Solid-Phase Table 1 Properties of Some Antigens Used in ELISA Antigen Crude mixture with other host proteins and agents, e.g., virus in feces

Relatively crude mixture of antigens, whole organism plus soluble proteins, limited host material, e.g., virus in tissue culture. Semrpurified preparations

Highly purified proteins, such as viruses, polypeptides, peptides, and immunoglobulins

Properties Unsuitable for direct adsorption to plastic, since contaminating proteins at very high protein concentration compete for sites on plastic, Sandwich ELISA needed to capture antigen selectively Cannot relate weight of antigen to protein content. May be sufficient antigen for coating plastic for indirect assay Irregular adsorption possible. High backgrounds if contaminants react with ELISA reagents Enriched antigen preparation. May be possible to relate desired specific protein concentration to antigenic activity. Direct and indirect ELISA possible owing to reduced contamination Possibility to be used as pure reagent with characterized adsorption properties

may be required. Thus, sandwich (trapping) conditions have to be set up. Table 1 gives a general picture of the types of material encountered and highlights some of the problems in coating. Passive adsorption has several theoretical, although not necessarilypractical, drawbacks. Theseinclude desorption, binding capacity, and nonspecific binding. 1.4. Desorption Owing to the noncovalent nature of the plastic-protein interactions, desorption (leaching) may take place during the stagesof the assay. However, if conditions are standardized, then this does not affect the viability of the majority of tests. 1.5. Binding Capacity It is important to realize that plastic surfaces have a finite capacity of adsorption. The capacity for proteins to attach to microplate wells is

Stages in ELBA

influenced by the exact nature of the protein adsorbed to the specific plate used. Saturation levels of between 50 and 500 rig/well have been found valid for a variety of proteins when added as 50-PL volumes. The effective weight of protein per well can be increased if the volume of the attaching protein is increased, effectively increasing the surface area of the plastic in contact with coating antigen. 1.6. Nonspecific Binding Unlike antigen-antibody interactions, the adsorption process is nonspecific. Thus, it is possible that any substance may adsorb to plastic at any stage during the assay. This must be considered in assay design, since reagents may react with such substances. 1.7. Covalent Antigen Attachment A variety of chemicals that couple protein to plastic have been used to prevent desorption, the antigen being covalently bound. These include water-soluble carbodimines, imdo- and succinimidyl-esters, ethanesulfonic acid, and glutaraldehyde. Precoating of plates with high-mol-wt polymers, such as polygluteraldehyde and polylysine, is another alternative (4,5). These bind to plates with a high efficiency and act as nonspecific adhesive molecules. This method is particularly useful for antigens with a high carbohydrate content, since these normally bind poorly to plastic. Generally, successful assays can be obtained without the need to link antigens to plates covalently. Specially treated activatable plates are now available and have to be proven. The use of covalently attached proteins does offer the possibility that plates could be reused. After an assay, all reagents binding to the solid-phase attached protein could be washed away after using a relatively severe washing procedure, e.g., low-pH. The covalent nature of the bonds holding the solid-phase antigen would prevent this from being eluted. Provided this procedure did not destroy the antigenicity of the solid-phase attached reagent, the plates might be exploited after equilibration with normal washing buffers. 2. Washing The purpose of washing is to separatebound and unbound (free) reagents. This involves the emptying of plate wells of reagents followed by the addition of liquid into wells. Such a process is performed at least three times for every well. The liquid used to wash wells is usually buff-

Washing

69

ered, typically PBS (0. lM, pH 7.4), in order to maintain isotonicity, since most antigen-antibody reactions are optimal under such conditions. Although PBS is most frequently used, lower molarity phosphatebuffers (O.OlM) may be used provided that they do not influence the performance of the assay. In this way, a considerable saving on chemicals and money can be made. In some assays tap water has been used for washing. This is not recommended, since tap water varies greatly in composition (pH, molarity, and so on). However, assays may be possible provided the water does not drastically affect the components of the test. Generally, the mechanical action of flooding wells with a solution is enough to wash wells of unbound reagents. Some workers leave washing solution in wells for a short time (soak time) after each addition (1-5 min). Sometimes detergents, notably Tween 20 (O.OS%), are added to washing buffers. This can cause problems where excessive frothing takes place producing poor washing conditions, since air is trapped and prevents the washing solution from contacting the well surface. For most cases, this addition does not contribute significantly to the washing procedure. When using detergents, care has to be taken that they do not affect reagents adversely (denature antigen), and greater care is needed to prevent frothing in the wells. Methods used in washing are as follows. 2.1. Dipping Methods The whole plate is immersed in a large volume of buffer. This method is rapid, but is liable to crosscontamination from different plates. It is also expensive on washing solution. 2.2. Wash Bottles Addition of fluid using a plastic wash bottle with a single-delivery nozzle is easy and cheap. Here the wells are filled individually in rapid succession, and then emptied by inversion of the plate and flicking the contents into a sink or suitable container filled with disinfectant. This process is repeated at least three times. Wells filled with washing solution may also be left for about 30 s before emptying. 2.3. Wash Bottles Plus Multiple-Delivery Nozzles This is essentially as in Section 2.2., except that a multiple-delivery (usually eight) device is attached to the outlet of the bottle. This enables eight wells to be filled at the same time.

Stages in ELISA

70 2.4. Multichannel

Pipets

The multichannel pipets used in the ELISA can be used to fill the wells carefully, with washing solution being contained in reservoirs. 2.5. Large

Reservoir

Use of a large reservoir of washing solution is convenient. Here a single or multiple-nozzle can be connected to the reservoir via tubing, so that the system is gravity-fed. Care has to be taken that large volumes of solution do not become microbially contaminated. 2.6. Special

Hand

Washing

Devices

These are available commercially, and involve the simultaneous delivery and emptying of wells by a hand-held multiple-nozzle apparatus. These are convenient to use, but require vacuum-creating facilities. In washing plates manually, the most important factor is that each well receives the washing solution so that, for example, no air bubbles are trapped in the well or a thumb is not placed over corner wells! After the final wash in all manual operations, the wells are emptied and then blotted free of most residual washing solution. This is accomplished usually by inverting the wells and tapping the plate onto an absorbent surface, such as paper toweling, cotton toweling, or sponge material. Thus, the liquid is physically ejected and absorbed to the surface, which is soft to avoid damage to the plate. 2.7, Specialist

Plate

Washers

These are relatively expensive pieces of apparatus that fill and empty wells. Various washing cycles canbe programmed. Theseareof greatadvantage where pathogens are being examined in ELISA, since they reduce aerosol contamination. Most of the methods involving manual addition of solutions and emptying of plates by flicking into sinks or receptacles must be regarded as potentially dangerous if human pathogens are being studied, particularly at the coating stageif live antigen is used.Also remember that live antigens can contaminate laboratories where tissue culture is practiced. The careful maintenanceof such machines is essential,since they areprone to machine errors,such ashaving a particular nozzle being blocked. 3. Addition

of Reagents

Immunoassays involve the accurate dispensing of reagents in relatively small volumes. The usual volumes used in ELISA are in the range of

Addition

of Reagents

71

50-100 pL/well. It is essential that the operator is fully aware of good pipeting techniques and understandsthe relationships of grams, milligrams, micrograms, nanograms and the equivalent for volumes, i.e., liters, milliliters, microliters, and so forth. Thus, assays cannot be performed where there is no knowledge of how to make up O.lM solutions, for example. The ability to make accurate dilutions is also extremely important, so that problems, such as having a l/50 dilution of antiserum and being required to make up a l/3500 dilution in Iafinal volume of 11 mL, should be solvable before you attempt ELBA or any other biological studies! 3.1. Pipets The microtiter plate system is ideally used in conjunction with multichannel microtiter pipets. Such pipets and their use are described later. Essentially, they allow the delivery of reagents via 4, 8, or 12 channels, and are of fixed or variable volumes of the 25-250 PL range. Single-channel micropipets are also required, which deliver in the range 5-250 p.L. Samples are usually delivered by microtiter pipets from suitably designed reservoirs (troughs) which hold about 30-50 mL of solution. General laboratory glassware is needed, such as five 25-r& glass or plastic bottles, and lo-, 5-, and 1 mL pipets. The range of general apparatus can be ascertained from the requirements set out in the following chapters. 3.2. Tips After the microplate, these are the most important aspect of ELISA and also an expensive component. Many thousands of tips might be needed to dispense reagents. There are many manufacturers who supply tips and care is needed to find tips that fit the available microtiter pipets. For multichannel pipets, tips are best accessedby being placed in special boxes holding 96 tips in the microplate format. These can be purchased already boxed (expensive), and then the boxes refilled from tips bought in bulk bags by hand. Sterile tips are available in the box format. Generally, tips should not be handled directly by hand. When restocking boxes or putting them on pipets, plastic gloves should be worn to avoid their contamination, Tips for dispensing in single-channel pipets have to be carefully considered. Where small volumes (5-20 p.L) are pipeted, the pipet manufacturer’s recommended tips should be used. It is essential that the tips

72

Stages in ELISA

fit securely on pipets, and they can be pressed on firmly by hand (avoiding their end). Particular care is needed where multichannel pipets are used to pick up tips from boxes, since often one or two tips are not as securely positioned as the rest, causing pipeting errors. The operator should always give a visual check of the relative volumes picked up. Where there is a problem of economics, tips may be recycled after washing, It is not recommended that tips that have been in contact with any enzyme conjugate be recycled, and these should not be discarded into other tips used for other stagesin ELISA. Washing of tips should be extensive, preferably in acid or strong detergentsolutions, and exhaustive rinsing in distilled water is essential. Damaged tips should be looked for and discarded. Figure 2 illustrates some practical aspects of pipeting in ELISA. 3.3. Other Equipment Several manufacturers supply microtiter equipment to aid multichannel pipeting. These include tube holders and microtip holders. The former consists of a plastic box that carries 96 plastic tubes with a capacity of about 1 mL. The tubes are held in exactly the same format as a microtiter plate, so that samples can be stored or diluted in such tubes and multichannel pipets can then be used for rapid transfer from the tubes. The tip holders involve the same principle, whereby tips for the multichannel pipets are stored in the 96-well format, so that they can be placed onto multichannel pipets rapidly in groups of 8 or 12. Various reservoirs with 8 or 12 channels for separation of reagents are also available. These are useful for the simultaneous addition of separatereagents. 4. Incubation The reaction between antigens and antibodies depends on their distribution, time, temperature, and pH (buffering conditions) at which the incubation step takes place. Intrinsic in any interaction is the actual avidity of the antibodies for the particular antigen(s) in any ELISA. Two types of incubation conditions are common: (1) incubation of stationary plates and (2) incubation of rotating plates (with shaking). These conditions affect the times and temperatures required for successful ELISAs and so will be discussed separately. 4.1. Rotation of Plates While Incubating Reagents The effect of rotating plates is to mix the reactants completely during the incubation step. Since the solid-phase limits the surface area of the

73

Incubation GUIDE TO PIPETTING ONLY USE FIRST STOP! DO NOT DRIP!

DO NOT PRESS HARD INTO WELL!

DO NOT USE TOO ACUTE AN ANGLE!

MAKE SURE TIP TOUCHES SIDE OF WELL AND LIQUID MULTI-CHANNEL PUSH

CHECK CHECK CHECK

PIPETTING

TIPS ON TIGHTLY

t LEVELS FOR BUBBLES FOR BLOCKED

FILL TO FIRST STOP EMPTY TO FIRST STOP HOLD BU’ITON DOWN BETWEEN PLATE RESERVOIR WHEN RE-FILLING PLATE

AND

Fig. 2. Aids to pipeting. adsorbed reactant, the mixing ensures that potentially reactive molecules are continuously coming into contact with the solid phase. During stationary incubation this is not true, and mixing only takes place because of diffusion of reagents. Thus, to allow maximum reaction

74

Stages in ELISA

from reagents in stationary conditions, greater times of incubation may be required than if they are rotated. This is particularly notable where highly viscous samples, e.g., l/20 serum containing antibodies are being examined. This represents 5% serum protein, and diffusion of all antibodies onto the solid-phase may take a long time. This is avoided if mixing is allowed throughout incubation. Similarly, where low amounts of reactant are being assayed,the contact time of the possibly few molecules that have to get close to the solid-phase reactant is greatly enhanced by mixing throughout incubation. Simple and very reliable rotating devices are available with a large capacity for plates. Figure 3 illustrates the advantages. Rotation also allows ELISAs to be performed independent of temperature considerations. The interaction of antibodies and antigen relies on their closeness, which is encouraged with the mixing during rotation. Stationary incubation relies on the diffusion of molecules, and thus is dependent on temperature. Thus, standardization of temperature conditions is far more critical than when rotation is used. This also has implications where many plates are stacked during incubation, since the plates heat up at different rates depending on their position in the stack. The wells on the inside may take longer to equilibrate than those on the outside, and this has a direct effect on the diffusion conditions, which affects the ELISA. This is negated by rotation, since there is the same chance of molecular contact in all wells. 4.2. Stationary

Incubation

Conditions

Assays may be geared to stationary conditions, although the exact times and temperatures of incubation may vary. The temperatures for incubation are most commonly 37”C, room temperature (on the bench), and 4°C. Usually the time of incubation under stationary conditions reflects which incubation temperature is used. Therefore, at 4°C a longer incubation might be given (overnight). In general, most incubation steps for stationary assays involving the reaction of antigen and antibodies are of l-3 h at 37°C. Sometimes these conditions are combined so that one reagent is added for, say, 2 h at 37OC,followed by one overnight at 4OC, usually because this produces a convenient work schedule. Where incubation is performed at room temperature, care is needed to monitor possible seasonal variation in the laboratory, since there can be very different temperatures, particularly in nontemperate countries. Direct sun-

Incubation

75

Fig. 3. Rotation favors maximum contact of reactants in solid and liquid phase. Continuous mixing enables maximum contact of molecules in liquid phase with those on solid phase. This overcomes problems of: 1. Temperature-the closeness of Ag and Ab is important to achieve immunological binding. Increasing temperature under stationary conditions increases diffusion. Any factors altering temperature of incubation will cause variation in diffusion, and hence affect variation in test results, e.g., stacking plates (unequal equilibration of wells), 2. Variability of treatment whereby plates might be moved in a test more than others by different operators (tapped), thus mixing reactants unequally between plates. 3. Vicosity of some samples may be high; therefore, chances of molecules in such samples interacting with solid-phase would be lower than in less viscous samples. 4. Times for incubation can be reduced as compared to nonmixed plates. 5. The detection of low concentrations of liquid phase reactants is increased.

light should also be avoided, as also must other sources of heat, such as machinery in the laboratory. As stated above, attention to how plates are placed during stationary incubation should be given. Ideally plates should be separated and not stacked. The plates should also be handled identically in assays, with no tapping or shaking of plates (including accidental nudging or movement by other personnel), since this will allow more mixing and interfere with the relative rate of diffusion of molecules in different plates. Under mixing

Stages in ELISA

conditions most antigen-antibody reactions are optimum after 30 min at 37”C, so that assays can be greatly sped up with no loss in sensitivity. This is not true under stationary conditions. Care must be taken to consider the types of antibodies being measured under various conditions, since ELISAs rarely reach classical equilibrium conditions. 5. Blocking

Conditions

and Nonspecific

Reactions

In general, measures have to be taken to prevent nonspecific adsorption of proteins to wells from samples added after the coating of the solidchase before, during, or a combination of both stages. Nonspecific adsorption of protein can take place with any available plastic sites not occupied by the solid-phase reagent. Thus, if one is assessing bovine antibodies in bovine serum, and bovine proteins other than specific antibodies bind to the solid-phase, on addition of an antibovine conjugate, these will bind and give a high background color. There are two methods used to eliminate such binding. One is the addition of high concentrations of immunologically inert substances to the dilution buffer of the added reagent. Substances added should not react with the solid-phase antigen or the conjugate used. Commonly used blocking agents are listed below, and they act by competing with nonspecific factors in the test sample for available plastic sites. The concentration used often depends on the dilutions of the test samples. Thus, if l/20 serum is being tested (5% protein), then blocking agents have to be at high concentration to compete successfully or have an increased binding potential as compared to the nonspecific substance. Such blocking agentscan also be added as a separatestep before the addition of the sample. This increases the competing ability of the blocker. Nonionic detergents have also been used to prevent nonspecific adsorption. These are used at low concentration, so as to allow interaction of antigen and antibody. Occasionally, both detergents and blocking substances are added together. Table 2 shows some of the commonly used blocking agents in ELISA. The best conditions for individual assays are only assessedin practice. However, the cost of such reagents should be taken in to account. Skim milk powder has been used successfully in many assays and is very cheap. Note, however, that certain blocking agentsmay be unsuitable for different enzyme systems, e.g., skim milk cannot be used in ureasedirected ELISA, or where biotin-avidin systems are used. Contaminat-

Blocking

Conditions

and Nonspecific

Reactions

77

Table 2 ELISA Blocking Agents (Representative Samples) References

Proteins Normal rabbit serum Normal horse serum Human serum albumin Bovine serum albumin Fetal calf serum Casein Casein hydrolysate Gelatin Detergents Tween 20 Tween 80 Triton X-100 SDS Other Dextran sulfate Coffee mate Nonfat dried milk

(9) (d4) (15) (15) (7) (16)

(17)

ing substances,e.g., bovine IgG in BSA, may eliminate the use of certain blocking agents from different suppliers. Most assays are validated under stated blocking conditions. However, workers may adapt assays for use with other blocking reagents where prescribed substances prove unobtainable or expensive! 5.1. Nonspecific

Immunological

Mechanisms

Reactions between solid-phase positively charged basic proteins and added reagents owing to ionic interactions have been described (15). This was removed by the addition of heparin or dextran sulfate in the diluent. The positive charges could also be removed by the addition of a low concentration of an anionic detergent (SDS). Such interactions have been noted for conjugates that, although not binding to uncoated plates, do bind strongly to those containing antigens. Addition of a variety of blocking buffers, e.g., containing BSA, Tween 20 casein, does not overcome the problem. Commonly, a high concentration of nonimmune serum from the same species as that in which the conjugate was prepared is necessary to prevent such a reaction.

78

Stages in ELISA

5.2. Immunological Mechanisms There are a large number of reports where antibody-antigen reactions have been noted where they should not occur. These can be termed aspecific reactions and are of an immunological nature. These are antibodies that are naturally present in serum and bind to antibodies from other species. They are not present in all sera and consequently cause problems in ELISA. As an example, human heterophilic antibodies have been demonstrated against a common epitope on the F(ab’), fragment of IgG from bovine, ovine, equine, guinea pig, rat, and monkey species (18). Ways of overcoming such antibodies include the use of F(ab’), as capture antibody where the heterophilic reaction is against the Fc portion of IgG or the use of high levels of normal serum obtained from the same species as the ELISA antibody in the blocking buffer. A review of heterophilic antibodies is also contained in ref. 18. 5.3. Rheumatoid Factor Interference This factor (RF) can cause a high level of false positives in the indirect ELISA. The factors are a set of the IgM class of antibodies that are present in normal individuals, but are usually associated with pathological conditions. They bind to the Fc portion of IgG antibodies, which are either complexed with their respective antigen or are in an aggregated form. Thus, any solid-phase/IgG/RF will be recognized by conjugates that recognize IgM and produce a false positive. Conversely, the binding of the RF to the antigen-IgG complex has been shown to interfere with the binding of IgG-specific conjugates producing a lower or false reaction in ELISAs. 5.4. Miscellaneous Problems Many sera contain antibodies specific for other animal serum components, e.g., antibovine antibodies are commonly present in human sera (19). Care must be taken dealing with conjugates that may have unwanted crossreactions of this type. Many conjugates are pretreated to adsorb out such unwanted cross-species reactions. Reagents are available for this purpose where various species serum components are covalently linked, for example, to agarose beads. These are added to sera, incubated for a short time, and then centrifuged into a pellet. Such beads can be reused after a treatment that breaks the immunological bonds between the antigen and serum component with which it reacted. Such solid-phase reagents have the advantage over methods whereby normal sera is added to

79

Enzyme Conjugates

absorb out activities, since the antibody molecules are totally eliminated from the solution. 5.5. Treatment of Samples Many laboratories routinely heat serato 56°C. This can causeproblems in ELISA and should not be pursued. Heating can cause large increases in nonspecific binding to plates. Note the study in ref. 8. 6. Enzve Conjugates Intrinsic to the ELISA is the addition of reagentsconjugated to enzymes. Assays are then quantified by the build-up of colored product after the addition of substrate or substrate and dye combination. Usually antibodies are conjugated to enzymes, and some methods are given later. Other commonly used systems involve the conjugation of enzymes to pseudo-immune reactors, such as protein A and protein G (which binds to mammalian IgGs), and indirect labeling using biotinavidin systems. Four commonly used enzymes will be described. Tables 3 and 4 show properties of enzymes, substrates, and stopping conditions. 6.1. Horseradiqh Peroxidase (HRPO) Plus Hydrogep Peroxide Substrate This is widely used. The substrate hydrogen peroxide is also a powerful inhibitor, so that defined concentrations have to be used. The reduction of peroxide by the enzyme is achieved by hydrogen donors that can be measured after oxidation as a color change. The choice of converted substratesthat remain soluble is essentialin ELISA, so that optimal spectrophotometric reading can be made. Commonly used chemicals are as follows. 6.1,l. 0-Phenylene

Diamine

(OPD)

OPD is prepared as a solution1of 40 mg/lOO mL of 0. 1M sodium citrate buffer, pH 5.0. Preweighed tablets are available commercially. 6.1.2.2,2’-Azinodi-Ethylbenzothiazolinesulfonic

Acid (ABTS)

ABTS is prepared at the same concentration, but in 0. 1M phosphate/ citrate buffer, pH 4.0. Tablets are available. 6.1.3. 5-Aminosalicylic

Acid (5-AS)

Commercial 5-AS is dissolved in 100 mL of distilled water at 70°C for about 5 min with stirring. After cooling to room temperature, the pH of the solution is raised to 6.0 using a few drops of 1M sodium hydroxide.

Table 3 Enzyme/Substrate Systems for ELISA Enzyme label (mol wt) Horseradish peroxidase (40,000)

Alkaline posphatase uwow P-galactosidase (540,000) Urease (483,000)

Substrate H,Oz H202 H202

(0.004%) (0.004%) (0 002%)

HzOz H,O,

(0 006%) (0 02%)

pnpp (2.5 mM) ONPG (3 mkf) Urea

Buffer

Dye Ortho-phenylene diamine (OPD) Tetra-methylbenzidine (TMB) 2,2’-Azino di-ethylbenzothiazolinesulfonic acid (ABTS) 5-Ammosahcylic acid (5AS) Di-aminobenzidine (DAB) Para mtrophenyl phosphate (pnpp) O-Nitrophenyl (ONW Bromocresol

P-D-galactopyranoside

Phosphate/citrate (0. lM), pH 5 0 Acetate buffer (0. lM), pH 5.6 Phosphate/citrate (0. lit!), pH 4.2 Phosphate (0.2&Q pH 6.8 Tris or PBS, pH 7.4 Diethanolamine (10 mM) and MgC12 (0 5 mM), pH 9.5 MgCl, and 2ME (O.OlM)/PBS, pH 7.5 pH 4.8

Enzyme/Substrate Color change

2

Alkaline phosphatase P-galactosidase Urease

Reading wavelength,

nm

Nonstopped

Stopped

Nonstopped

Stopped

Stopping solution

Green/ orange Blue

Orange/ brown Yellow

450

492

1.25M HzS04

650

450

SDS (1%)

ABTS

Green

Green

414

414

No stop

5AS DAB pnw

Brown Brown Yellow/ green Yellow

Brown Brown Yellow/ green Yellow

450 N/A 405

450 N/A 405

420

420

Purple

Purple

588

588

No stop No stop 2M sodium carbonate 2M sodium carbonate 1% merthiolate

Enzyme label Horseradrsh peroxidase

Table 4 Systems for ELISA (2)

System OPD

ONPG Urea bromocresol

Notes Possibly carcinogen, soluble product Nonmutagenic, sohrble product Mutagenic, soluble product Safe, soluble product Insoluble product, safe Safe, soluble product Safe, soluble product Safe, soluble product

82

Stages in ELISA 6.1.4. Tetra-Methylbenzidine

(TMB)

The optimum substrate (hydrogen peroxide) concentration depends on the hydrogen donor and the solid-phase. This is usually established in preliminary tests, but concentrations between 0.010 and 0.0005% are adequate. Hydrogen peroxide is available as 30% commercially. The development of the colored product is measured at different wavelengths. The optimum wavelength may also shift if the reaction is stopped by a blocking reagent to prevent change in the optical density after a reaction period. The stopping reagents involving HRPO are: solutions of hydrochloric or sulfuric acid for OPD and TMB; and sodium dodecyl sulfate (SDS) for ABTS. The optimal wavelengths for reading are: 415 nm for ABTS; 492 nm for acidified OPD (420 nm for nonacidified OPD); 492 nm for 5-AS; and 655 nm for TMB (unstopped)and 450 nm (acidified). 6.2. Alkaline

Phosphatase

Plus

p-Nitrophenylphosphate

This is assayed in buffer depending on the source of the enzyme. For bacterial enzyme, 0. 1M Tris-HCl buffer, pH 8.1, containing 0.01% magnesium chloride is used. For intestinal mucosal enzyme, a 10% (w/w) diethanolamine (97 mL in 1 L of a 0.01% magnesium chloride solution) buffer, pH 9.8 (adjusted with HCl), is used. The p-nitrophenylphosphate is added just before use (available as preweighed pellets) to 1 mg/mL. The production of nitrophenol is measured at 405 nm. The reaction is stopped by the addition of 0.1 vol of 2A4 sodium carbonate. Note that inorganic phosphate has a strong inhibitory effect on alkaline phosphatase, and therefore, PBS or similar buffers are avoided. 6.3. PGaZactosidase Plus 0-Nitrophenyl-PGaZactopyranoside

This is determined after addition of a solution containing 70 mg o-nitrophenyl P-b-galactopyranoside/lOO mL of O.lM potassium phosphatebuffer, pH 7.0, containing 1 mM magnesium chloride and O.OlM 2-mercaptoethanol. The reaction may be stopped by the addition of 0.25 vol of 2A4 sodium carbonate.

Availability

83

of Conjugates

6.4. W-ease, pH Change, and Bromocresol Purple Indicator This is determined by addition of a weakly buffered solution of urea (pH 4.8) in presence of bromocresol purple. The urea is hydrolyzed to liberate ammonia in the presence of urease, and this raises the pH of the solution resulting in a color change from yellow to purple. The reaction can be stopped by the addition of 10 ltL of a 1% solution of merthiolate (thiomersal) to each well. Tables 3 and 4 summarize properties of various enzyme systems. 7. Availability

of Conjugates

Conjugates may be obtained commercially or made in individual laboratories. Great care must be exercised in using the appropriate reagent in any assay. Thus, the immunological implications of various reagents must be considered and knowledge sought. Many conjugates are directed against serum components of different species. The following list illustrates some of their features. 1. The species in which the antiserum is produced can be important. Thus, donkey anticow IgG denotes that a donkey has been used to prepare antiserum against cow immunoglobulin G. As other examples we have rabbit antipig, pig antidog, and so forth. 2. The above description must be refined, since both the donating serum and the immunogen are probably subjected to immunochemical treatments. Thus, the donating serum can be crudely fractionated before conjugation (which is usual) or may be affinity-purified, so that the conjugate is 100% reactive against the immunogen. Thus, we might have: donkey IgG anticow IgG or donkey IgG (affinity-purified) anticow IgG. These descriptions deal with the processing of the donor serum after production and before conjugation. The description of the immunogen is also sometimes critical. Using the last example, the cow IgG may have itself been affinity-purified before injection of the donkey, or specific parts of the immunoglobulin purified to raise a subclass-specific antiserum, e.g., heavy-chain IgG may have been injected into the animal so that the conjugate only reacts with IgG. Remember that all immunoglobulin classesshare antigens, so that it is highly likely that an antiserum raised against IgGs of any species will detect IgA, IgM, and IgG. Thus, the specificity of the conjugate has to be considered in any assay, 3. Different sources and batches of conjugates from the same manufacturing source may vary. Thus, in large-scale applications, it is good practice to

84

Stages in ELISA

obtain a successful batch sufficient for all future testing. The anti-IgG one obtains from commercial company A may give different results than that obtainedfrom company B. 4. Conjugatesmust be titrated to optimum conditions and not usedin excess. This is vital to obtaining reliable results. 8. Conjugation with Enzymes The sensitivity of the ELISA depends on the ability of the antibody to bind and the specific enzyme activity of the labeled immunoreactant, the conjugate. The linkage of an enzyme to an antigen or antibody may affect the specificity of an assay if any chemical modification of the moieties involved alters the antigenic determinants or the reactive sites on antibody molecules. Thus, chemical methods that do not affect these parameters have been chosen. Most of the techniques are straightforward and can be readily used by nonspecialists interested in developing their own enzyme immunoassays. Not only the immunoreactivities, but also the catalytic activity of the enzyme must be maintained after conjugation, Following conjugation, it is necessary to test the immunoreactivity as to whether it has the desired specification. Before use in ELISA, it may be necessary to purify the conjugates to remove unconjugated antigen or antibody and free enzyme. Reagents used to produce conjugates are numerous, and their mode of action is to modify the functional groups present on proteins. Antigens that are nonproteinaceous, e.g., steroids, can be conjugated with different means and are not dealt with here. Enzymes are covalently bound to reagents either directly by reactive groups on both enzyme and reagent or after introduction of reactive groups (e.g., thiol or maleimid groups) indirectly via homo- or heterobifunctional reagents in two-step procedures (20). Requirements for optimal conjugation are: 1. Simplicity and rapidity; 2. Reproducibility (obtaining constantmolar ratio of enzyme and reagent); 3. High yield of labeled reagent,and low yield of polymers of enzyme and reagent; 4. Low-grade inactivation of reagentand enzyme; 5. Simple proceduresfor separationof labeled and unlabeledreagents;and 6. Long-termstability without loss of immunological andenzymaticactivities. 9. Development

of Label The substrate is usually chosen to yield a colored product. The rate of color development will be proportional, over a certain range, to the

Development

85

of Label

amount of enzyme conjugate present. On a kinetic level, reactions are distinguished by their kinetic order, which specifies the dependence of reaction rate on the concentration of reactants. Under the conditions generally employed in ELISA, the reaction exhibits zero order with respect to the substrate. It can be seen that too little substrate will limit the rate of product production. Thus, sufficient substrate must be present to prevent the substrate and/or cofactors from being rate-limiting. Where substrate and chromogenic hydrogen donors are necessary for color development, the concentrations of both must be assessedto obtain optimum conditions. The product must be stable within a defined time, and products that are unstable in bright light or at temperatures at which the assay is performed should be avoided. The physicochemical parameters that affect the development of color include: 1, Buffer composition and pH; 2. Reaction temperature; 3. Substrate

and/or cofactor

concentration

and stability,

4. Product stability; 5. Enzyme stability; and 6. Substrate and product stability.

Horseradish peroxidase is active over a broad pH range with respect to its substrate, hydrogen peroxide. However, the optimum pH for the development of label in the ELISA will vary depending on the chromogenic donor. Changing the pH will reduce the reaction rate, but will not affect the reaction kinetics, e.g., increasing the pH to 5.0 for ABTS will slow down the rate of reaction (pH optimum 4.0), but does not affect the linearity of the kinetics. The majority of the buffers used in substrate formulation are of low molarity citrate base. Since the reaction kinetics are dependent on pH, a stable buffering capacity is essential. The stability of HRPO varies in different buffers, being more stable in 0. 1M citrate than 0. 1M phosphate-buffers. High molarity phosphate buffer can be particularly damaging to HRPO at low-pH. Nonionic detergents exert a stabilizing effect on the enzymic activity of HRPO, and this can be enhanced by increasing reaction temperatures. The detergents have also been demonstrated as having a stabilizing effect on the enzymes.

Alkaline phosphatase is active at alkaline pH and optimum above pH 8.0. The buffers used with the substrate pnpp are diethanolamine/I-ICl, pH 9.6. Inorganic Mg2+ is essential for enzymic activation. Nonionic detergents appear to have no effect on the enzyme activation, substrate

Stages in ELISA

catalysis, or product development. Inactivation of the enzyme on contact with microplates does not occur. Urease isenzymatically active over a broad pH range. The specificity of urease for its substrate (urea) is almost absolute. The urease substrate solution contains urea and a pH indicator, bromocresol purple, at pH 4.7. The urease catalyzes the urea into ammonia and bicarbonate. The released ammonia causes an increase in pH, which changes the color of the indicator from yellow to purple. The generation of color is not directly related to the amount of urea catalyzed. Since the color development is dependent on pH, it is essential to check that the pH is accurate before addition. It is also essential that no alkaline buffers remain after, for example, washing (pH 7.4, PBS), since this will cause a change in color, and plates must be washed finally in water if PBS is the usual washing buffer. 9.1. Reaction Temperature Between-well variation in an assay can cause differential rates of color development. Similarly, varying temperatures in the performance of the assay can cause variation. It is advisable, therefore, that substrates be added at a defined temperature and that plates be incubated under uniform conditions. This is normally room temperature. Note that this definition is rather loose and that this should be assessed in individual laboratories, since there can be great variations in different countries. The best practice is to add substrate solutions at a defined temperature obtained by using solutions heated (or cooled) to that defined temperature. This is particularly important when attempting to standardize assaysbetween operators and laboratories where a fixed time for stopping an assay is used. 9.2. Substratefcofactor

Concentration and Stability As already stated, optimization of substrate concentrations must be made. This is usually stated for particular systems (literature, kits, and so on). Certain solutions can be made and stored. As an example, OPD can be made up in buffer and stored frozen in well-sealed vials. It can then be thawed and used (after the addition of HzOz). This negates the need to weigh out small amounts of OPD for small volumes of substrate solution and aids standardization of assays. The use of preweighed chemicals in the form of tablets available commercially also greatly improves the accuracy and convenience of producing substrate solutions, although these tablets are expensive.

Stopping

Reactions

87

9.3. Product Stability Once the substrate has been catalyzed and a colored product achieved, it is essential that the color remains stable. In the majority of ELISAs, positive results are read by eye or by spectrophotometer, since the intensity of color (optical density) is compared to a series of previously worked out negative values. An unstable colored product would affect the build-up of color. For spectrophotometric reading of results, it is vital that the product color remains stable without shifti’ng the absorption spectrum, since the microplate readers assessthe absorbance of the colored product at a preset wavelength. Generally, enzymic activity is prevented from proceeding further at a predetermined time by the addition of a reagent, preventing further enzymic activity. This is dealt with below. 9.4. Enzyme Stability The enzymes used in the ELISA are stable with respect to their activity with defined substrates. Thus, a high degree of consistency is found using the same batch of conjugate under defined conditions. 9.5. Substrate and Product Stability As already indicated, where substrates are only soluble to a limited extent in aqueous buffers, the use of mixed aqueous/organic buffers is possible. These solvent systems can allow significantly greater amounts of substrates to be incorporated into solution and allow their use in microplate ELISAs. Partially or totally insoluble products have their uses in variants of ELBA, e.g., in the staining of sections in immunohistochemistry where insoluble products localize the area of antigen or antibody reaction. 10. Stopping Reactions Reagents are added to prevent further enzymic reaction in ELISA. This is performed at a time as determined in the specific assay. This process is usually called “stopping,” and the reagent that is used the “stopping reagent.” The stopping is usually made at a time when the relationship among the enzyme-substrate-product is in the linear phase. Molar concentrations of strong acids or strong bases stop enzymic activity by quickly denaturing enzymes. Other stopping reagents are enzyme-specific. Sodium azide is a potent inhibitor of HRPO, whereas EDTA inhibits alkaline phosphataseby the chelation of metal ion cofactors, Since addi-

Stages in ELISA

tion of stopping agents may alter the absorption spectrum of the product, the absorption peak must be known. Thus, e.g., sulfuric acid-stopped OPD/ELISAs are read at 492 nm (450 nm before stop). The wavelengths for reading the appropriate substrates before or after addition of stopping agentsare shown in Table 4. The addition of stopping agents can also increase the sensitivity of an ELISA. In the addition of stopping reagent, the volumes must be kept accurate, since photometric readings are affected if the total volume of reactants varies. 11. Reading Since the product of substrate catalysis is colored, it can be read in two ways, namely, (1) by-eye inspection or (2) using spectrophotometers. II. 1. By-Eye Reading ELISAs can be designed for use with either system, although different conditions and controls may have to be included. It is essential that the principles of ELISA be thoroughly understood before either system is adopted. In particular, the by-eye test is not necessarily simpler to standardize. However, where correct standardization is used, it offers sensitive assays. When a correct plate template is used, the range of color product will be from full color through partial color to no color. Known strong positive samples will give strong color. Weak positives will give partial color, and negatives will give no color, or that of negative wells. It is essential that controls of this sort be incorporated in the intended assays. Some difficulties arise in differentiating weak positives from negatives by-eye. The interpretation of by-eye tests can vary from operator to operator, and hence, results are more subjective than by spectrophotometer. Some substrate/enzyme combinations favor by-eye reading. Where tests have to be read by-eye (where instrumentation is not available), the best assayscan be produced in other laboratories that can quantify reagents using machine reading and evaluate the parameters of the by-eye reading. As an example, a negative population of seracan be examined, and control negative sera, reflecting different parts of the negative OD distribution, can be adopted for by-eye controls. Thus, a serum having the highest OD value may be selected as the negative control. Any sera giving by-eye discernible results higher than this serum would therefore be assessed with high confidence as being positive. Assays that require comparison of closely related data, such as competition assays,

Practical

Problems

89

are not suitable for by-eye interpretation, e.g., where the competition slope is compared. 11.2. Spectrophotometric Reading The product of the substrate catalysis by enzyme is measured by transmitting light of a specific wavelength through the product and measuring the amount of adsorption of that light, if any, by a machine. Since different products are produced in ELISA, care is taken to select appropriate filters for the detection of the correct wavelengths. Although microcuvets and conventional spectrophotometers can be used for this purpose, this is laborious where large numbers of samples are measured. Special machines are available for the reading of colored products in microplates. These read the absorbance of each well at a preselected wavelength of light. Either one well can be read at a time (manual readers) or more suitably, a column of eight wells is read simultaneously (semiautomatic or automatic multichannel spectrophotometers). For the semiautomatic readers, the wavelength filters are added manually, whereas for the automatic readers,the wavelength filter(s) (dual-wavelength machines are available) can be selected from a control panel. In the main the basic results from such a machine are expressed as absorbance units and are recorded on paper rolls. Various (limited, but useful) processing of the data is usually available, such asthe expression of the absorbance values as a matrix or as + and - against control wells or values given to the machine. Most readers can be connected to computers, and a range of software (commercial and private) is available to manipulate and store data. This is important in large-scalesample handling or where complicated arithmetic routines are performed on the data. An important feature of the ELISA having a colored product that can be examined by-eye is that tests can be rapidly assessedbefore machine reading. Thus, one can seethat a test has “worked” or not at a glance.Extensive readingtime is not wastedif a silly mistake hasoccurred,unlike RIA, were it is essentialto count samplesbefore results are obtained. Such by-eye assessment is also convenient when “sighting” experiments are being made during development of assays, 12. Practical Problems This section is intended to discuss problems associated with the practical aspects of ELISAs that have been observed under different laboratory conditions. Some aspects are mentioned in other parts of the manual. However, repetition will do no harm.

Stages in ELISA

90

12.1. Overall Observations on Running ELBA The major cause of problems is the scientist(s) involved. This has been demonstrated graphically by the author’s involvement in training and supplying kit reagentsto many laboratories all over the world. The main problem is lack of close-contact training in the fundamentals of ELISA, so that the scientist has the experience to identify andthen solve problems in the use of reagents.There is no substitute for good training. This manual attempts to be highly practical, and it is hoped that somegood principles will be learned. 121.1. Problems Caused by Lack of General Scientific Knowledge This is obvious evenwith referenceto the manual, since the biological impli-

cations of results cannot be assessedwithout general knowledge of several fields of science,e.g.,epidemiology, immunochemistry, biochemistry, immunology, and so on. This should not be too depressing,since the ELISA should be a tool for the investigation of specific problems rather than an end in itself. 12.1.2. Problems Caused by Sloppy Technique

This can be associatedwith the above factors. The reproducibility of any assaysrelies in part on the accuracy of the workers involved. This is complicated considerably where many operatorsperform the assay,e.g., in a laboratory concernedwith large-scaletesting of seraon a routine basis. Attempts should be made at individual assessment to improve reliability of techniques (pipeting accurately, timing accurately, and so on). Quality control (QC), and quality assurance (QA) protocols should be developed. 12.2. Common Problems of Instrumentation and Reagents Not all problems can be blamed directly on the operator. Although the individual steps of ELISAs are relatively simple, assayscan be regarded as complex in that several steps with different reagents (all of which have to be standardized) have to be made. This increases the likely problems in any methodology. Reagents also have to be stored, and are subject to contamination by microorganism or from other workers introducing unwanted reagents through the use of contaminated tips. 12.2.1. Water

This can be the major problem in standardization of assays between different laboratories even where identical reagentsare used. Thus, where kits are supplied, water may also be given, at least for the initial dilutions

Practical

Problems

91

of the stock reagents. The reasons why water affects the ELBA have not been extensively examined, and no single factors have emerged as being most important. The recommendation that triple-distilled water be used is good, although this is not always available to less-well-equipped laboratories. The type of problem encountered is that of higher than expected readings using control sera, as well as for plate blanks. The supply of tested water for the preparation of buffers for the initial dilutions of reagents will solve this problem. Workers should also obtain supplies from other laboratories to examine whether it cures observed problems. 12.2.2. Laboratory Glassware This should be clean and well rinsed in glass-distilled water. This avoids the introduction of contaminants or adverse pFI conditions into ELBA reagents, especially where initial dilutions of conjugate are concerned. 12.2.3. Micro&et Tips These are expensive and can be in short supply in some laboratories. They can be washed, but as a rule: 1, 2. 3. 4. 5.

Neverwashandreusetips thathavedeliveredconjugateor conjugatesolution. Check that the endsof the tips arenot damagedduring use and washing. Always rinse the tips very well in distilled water. Dry the tips beforeuse. Get the appropriatetips to fit your micropipets. A poor fit meansfrustration while pipeting and leadsto inaccuracy

12.2.4. Micropipets Since thesearethe instrumentsthatdeliver volumes of liquid, they arefundamentally important to the accuracyof the ELISA. They should be checkedregularly for precision and accuracyof delivery volumes. Instructions on how to do this areincluded when pipetsampurchased.If not, contactthemanufacturer, Limited maintenanceof the pipets should be done with attentionto the plungersin multichannel pipets,which becomecontaminated.Cam shouldbe takento make surethat liquids arenot pulled into the pipet. If so,they must be cleaned, 12.25. Plates 1. If a particular plate is recommended,thenusethat plate unlessyou retitrate given reagentsin anothermanufacturer’splate. 2. Never use a tissue-culture-gradeplate for ELBA. Sometimesthat can be made to work, but it gives much more variability than those specifically madefor ELISA.

Stages in ELISA 3. Always report which plate and what treatment of the plate have been made. 4. Reuseof plates after washing is problematic, and high variability ISobserved. However, if owing to economic and supply reasons, this is necessary, use 2M NaOH overnight after washing plates m tap water. Then rinse thoroughly in distilled water. You should use washed plates with many more controls than for new plates to measure variability. 12.2.6. Troughs (Reservoirs) 1. Use specific troughs for conjugate and substrate only to avoid crosscontamination. 2. After use, wash the trough in tap water, then distilled water, and then leave soaking in a mild detergent. For use, rinse in tap water, distilled water, and then dry using a towel. 3. Never leave reagents in troughs for a long time after they have been used in an assay.Rinse immediately if possible. 12.2.7. Substrate Solution The temperature of the solution is important, since this affects the rate of color reaction, so try to perform the addition with the substrate always at the same temperature. This can easily be achieved if buffer tablets are used by keeping the water used at a constant temperature or preincubating this in a water bath. Where substrate solution is kept frozen, you must ensure that on thawing the same temperature is achieved for every test (again by using a water bath). A range of 20-30°C is recommended. The variation in temperature of the substrate solution will be the greatest factor in causing differences between assays performed with the same reagents. 12.2.8. Timing of Steps Generally, individual steps should be timed accurately. Thus, for a l-h incubation step, no more than 5 min either way should be tolerated, For assays that specifically recommend times, there is no reason why they cannot be met. Timing is less important where rotation of plates is made, although it is good practice to follow protocols accurately. 12.2.9. Incubation We have already considered stationary against rotated plates. The conclusion that rotation of plates for incubation steps is to be greatly recommended to eliminate: 1. Vicosity effects. 2. Time differences.

Practical

Problems

93

3, Temperature effects, including edge-well differences caused when plates are stacked and incubated in a stationary manner. However, there is no reason to be depressed when a rotator is not avail-

able. Provided that standardization of methods is used, stationary plate assays are no problem. Several tips may be given on the incubation of plates that are not rotated. 1, Avoid stacking the plates. Keep them separated. 2. Incubate at 37OC. 3. Always use the same procedure for addition of reagents, i.e., do not tap one plate, pick another up three times, and examine one or two plates during the incubation and not others. The reason is that this mixes the reagents over the solid-phase to different degrees and this alters the interaction in the wells. Thus, more care is taken with handling the plates identically in one test and from day to day. 4. If incubation has to be made at room temperature, make a note of the temperature, and its variation during the year. This may explain variation in results at different times.

12.2.10. Conjugates Care must be taken with these reagents, since they are the signal sup-

pliers of the whole assay. 1, Make sure you understand what the conjugate is (species made, specific antibody activity, and so on). 2. Store at recommended temperatures. 3. Never store excessively diluted conjugate for use at some later time. 4. Always make up the working dilution of conjugate Just before you need it. 5. Always use clean tips, preferably previously unused, to dispense conjugates, 6. If the recommended dilution or titrated dilution of conjugate is very high (e.g., l/10,000), to make 10 mL at working strength, 1 PL would have to be added to 10 mL. You may have difficulties in making small volumes of working strength. Thus, a small dilution should be made to allow feasible pipeting of the conjugate without waste. Dilute in 50% glycerol/50% PBS to, say, l/10 of original. Store at -20°C if possible. 7. Never leave conjugates on the bench for excessive time.

12.2.11. Stopping Solution Addition Since the multichannel spectrophotometer reads through a thickness of liquid, any change in the volume in a well will result in an alteration of OD reading for the same colored solution, Thus, it is important to add stopping solution accurately to achieve the same volume in each well

and limit the effect of volume changes (this of course is also true of addi-

Stages in ELISA tion of conjugate solution and also concerns the blotting of plates to get rid of residual washing solution, all of which affect the final volume per well). 12.2.12. Addition of Samples Accurate and consistent pipeting technique is a prerequisite for limiting pipeting error. Major problems are caused by: 1. Failure to put sample into buffer in well, leaving it on the side of the plate (particularly when plates are incubated in a stationary manner). 2. Causing frothing on addition of samples. 3. Lack of concentration when adding large number of samples, causing missed wells and duplication of samples in the same well. 4. Poorly

maintained

pipets and tips.

5. Not thawing out sera properly (protein tends to collect at bottom of tubes on freezing) so that adequate mixmg to ensure homogeneity is essential.

12.2.13. Reading PlateslData The advantage of ELBA is that the plates can be read quickly and a large amount of data obtained. This can lead to several problems: 1. Computerization whereby plate data are processed and results given (* 56%, and so on) must be checked quickly from examination of plate data by eye. This is essential, since some programs do not give warnings to check highly suspect results probably caused by a major sampling error. Thus, mean values may be calculated from the plate data by the computer, and these can be used to ascribe positive or negative for particular samples. Unless safety features to screen for wildly different OD values in a pair are included in the program, false results will be obtained (e.g., two values for a serum are .4 and .42, mean = .41 = positive; two values for a serum are

0.02 and 0.80, mean = .41?). Personalexamination of initial plate data would easily spot this serum result as nonsense, but sole reference to the computer printout of pos/neg would not! This is a facile example, but more complex analytical programs have similar hidden problems. 2. There is a problem where large data bases are set up to store data from large-scale screening. This is related to the checking factor above which results read directly into a data base are taken as reliable without by-eye examination of the feasibility of those results. There is a tendency to want results from many laboratories, so programs have been supplied to facilitate this. Such programs can easily dehumanize the diagnostic process with

the loss of control of results and ability to back-checkdata. 3. Cables that connect computers to spectrophotometers, printers that do not work, and so forth. These are general hardware problems that have to be conquered.

Practical

95

Problems Table 5 Problem Solving in ELISA

Problem No color even after 30 min incubation with substrate

Color all over plate

Patchy color

No hydrogen peroxide added Hydrogen peroxide “gone-off’ Added blocking buffer in adsorption step diluent Wrong dilution of hydrogen peroxide Too strong conjugate Antispecies reacts with adsorbed antigen Serum factors in heated sera Poor and variable adsorption of reagents to plate Bubbles in tips Poor pipeting technique Plates faulty Incubation of stacked plates Poor mixing of reagents Dilution series poor

Color develops very quickly

Poor washing Conlugate too strong Contaminating enzymes

Color develops too slowly

Total unexpected results High background color

Solution

Cause

Conjugate too weak Contamination inhibits activity of enzyme, e.g., sodium azide on peroxidase Low temperature of laboratory or substrate solution Plate format incorrect Dilution series wrong Gross error in test reagent dilutions or omissions of reagents Nonspecific attachment of antibodies Antispecies conjugatereacts with reagent coated on plate

Check Retitrate Check Check Check dilution Check suitable controls Do not heat sera Check adsorption buffer and homogeneity of preparation Avoid vigorous pipeting and Tween Practice care Contact manufacturer, note batch number, use only ELISA plates Keep plates well separated if not rotating plates Ensure good mixing on sampling Practice good dilution techniques, examine pipets, recalibrate, ensure tips fit well Avoid detergents, ensure all wells filled Retitrate to get maximum OD of around 1.5 at 10-15 min Common in cells, pretreatment may be necessary; make sure all reservoirs are clean Check drlutions and time when diluted Avoid wrong preservatives Make sure temperature of substrate is correct Check Check Check Unsuitable blocking buffer or omission of blocking buffer Set up controls to assess whether any reagent binds unexpectedly to any reagent

12.3, Trouble-Shooting ELISA Table 5 showssomeof the problemscommonly seenin ELBA development and practice. This is not an exhaustive list of things that can go wrong, but it highlights areasthat should be examined first where assaysprove diffkult.

Stages in ELISA References 1. Cantarero, L. A., Butler, J. E., and Osborne, J. W. (1980) The bindmg characteristics of proteins for polystyrene and their srgnificance in solid-phase immunoassays. Analyt. Biochem. 105,375-382.

2. Kurstak, E., Tijssen, P., Kurstak, C , and Morisset, R. (1986) Enzyme immunoassay in diagnostic medical virology Bull. WHO 64(3), 465479. 3. Geerligs, H. J., Weijer, W J., Bloemhoff, W., Welling, G. W., and Welling-Wester, S. (1988) The influence of pH and ionic strength on the coating of peptides of herpes simplex virus type I in an enzyme-linked immunosorbent assay J. Immunol. Meth.

106,239-244.

4. Rembaum, A., Margel, S., and Levy, A. (1978) J. Immunol. Meth. 24,239. 5. Gabrilovac, J, Pachmann, K., Rodt, H., Gager, G., and Thierfelder, S. (1979) Particle-labeled antibodies I. Anti-T cell antibodies attached to plastic beads by polyL-lysine. J. Immunol. Meth. 30, 161-170 6. Kohno, T., Hashida, S., and Ishikawa, E. (1985) A more sensitive enzyme immunoassay of anti-insulin IgG in guinea pig serum with less non-specific binding of normal guinea pig serum. J. Biochem. 98,379-384. 7. Meegan, J. M , Yedloutscmg, R. J., Peleg, B. A., Shy, J., Peters, C J., Walker, J S., and Shope, R. E. (1987) Enzyme-linked immunosorbent assay for detection of antibodies to Rift Valley Fever Virus in ovine and bovme sera. Am. J. Vet Res. 48, 1138-1141 8. Husby, S., Holmskov-Neilsen, U., Jensenius, J. C , and Erb, K (1982) Increased non-specific binding of heat treated proteins to plastic surfaces analyzed by ELISA and HPLC-fractionation. J. Immunoassay 6,95-l 10. 9. Herrmann, J. E., Hendry, R. M., Collins, M. F. (1979) Factors involved in enzymelinked immunoassay and evaluation of the method of identification of enteroviruses. J. Clin. Microbial. 10,210-217. 10 Harmon, M. W., Russo, L. L., and Wilson, S. 2. (1983) Sensitive enzyme immunoassay with P-o-galactosidase-Fab conjugate for detection of type A influenza virus antigen in clinical specimens. J. Clin. Microbial 17,305-3 11 11. Kenna, J. G., Major, G. N., and Williams, R. S. (1985) Methods for reducing nonspecific antibody binding in enzyme-linked immunosorbent assays. J, Zmmunol. Meth. 85,409-419.

12. Robertson, P. W., Whybin, L. R., and Cox, J. (1985) Reduction m non-specific binding in enzyme immunoassays using casein hydrolysate in serum diluents. J Immunol. Meth. 76,195-197.

13. Gary, W. G. J. R., Kaplan, E. J., Stine, E. S., and Anderson, J. L (1985) Detection of Norwalk Virus antibodies and antigen with a biotin/avidin system. J. Clin. Microbial.

22,274-278

14. Hatfield, R. M., Morris, B. A., and Henry, R. R. (1987) Development of and enzymelinked immunosorbent assay for the detection of humoral antibodies. Auzan Path

16,123-140. 15. Dietzen, R. G. and Fran&i, R. I. B. (1987) Nonspecific binding of immunoglobulins to coat proteins of certain plant viruses in immunoblots and indirect ELISA. J. Viral, Meth.

lS,159-164.

Practical

Problems

97

16. York, J. J. and Fahey, K. J. (1988) Diagnosis of infectious laryngotracheitis using a monoclonal antibody ELISA. Avian Path. 17, 173-182. 17. Vogt, R. F., Phillips, D. L., Henderson, L. O., Whitfield, W., and Spierto, F. W. (1987) Quantitative differences among various proteins as blocking agents for ELISA microtiter plates. J. Zmmunol Meth. 101,43-50. 18. Boscato, L. M. and Stuart, M. C. (1988) Heterophilic antibodies: a problem for all immunoassays. Clin. Chem. 33,27-33. 19. Dise, T. and Brunell, A. P. (1987) Antibovine antibody in human sera as a cause of nonspecificity in enzyme immunoassay. J. Clin. Microbial. 25,987-990. 20 Ishikawa, E., Imagawa, M., Hashide, S., Yoshatake, S., Hagushi, Y., and Ueno, T. (1983) Enzyme labeling of antibodies and their fragments for enzyme immunoassays and immunological staining. J. Immunoassay 4,209-213.

PART 11

Practical

Exercises

The aim of Chapters 5-8 will be to illustrate the principles of ELISA fully by: 1. Showing worked examplesof eachassay,including diagramsof platesand

representationaldata from assays; 2. Analyzing such data in terms of important rules that are learnedat each stage;and 3. Providing full working instructronsfor workers to be ableto perform each assayso that they obtain their own data to be analyzedas describedin (1) and (2). This includes full instructionson the preparationand standardization of reagents.

The chapters can therefore be used in several ways. Workers without accessto reagentswill obtain a working knowledge of ELISA through the examples. The chapterscan also be used in training courses where reagents may be provided (as indicated in the text). The information will also be useful to workers who have already had some experience of the technique and who may have had difficulties in obtaining and analyzing data. Remember that it is the application of the ELISA to specific problems, and not the methodology for its own sake, that is the most important reason the techniques should be mastered. 1. Test Schemes

You may be already familiar with the concepts in ELISA, whereby an antigen binds to an antibody that can be labeled with an enzyme, or be in turn detected with a species-specific antibody (enzyme labeled). All the ELISAs are variations on this theme. Inherent in the methods of ELISA described in these chapters is the fact that one of the reagents is attached to a solid-phase, making the separation of bound (reacted) and unbound (nonreacted) reagents simple by washing. Before performing ELISA on disease agents, it is useful to train using reagents of defined reactivity, which are easily available and which provide security problems. An ideal

115

116

Practical

Exercises

system is to use an imrnunoglobulin (Ig) and more particularly an immunoglobulin G (IgG) as an antigen. Do not get confused here, since you have learned that the antibody population contains high levels of IgG acting as antibody. In the context of learning the principles, we are using IgG as an antigenic protein, since: 1. IgG from one animal species can be injected into another animal species so that a specific antiserum to that IgG is prepared. 2. Such antibodies can be labeled with enzyme, or detected with a second species-specific antibody labeled with enzyme. Such reagents are defined, easy to standardize, stable, and available commercially. The particular IgG system chosen in most of the chapters involves guinea pig, but similar tests can be performed with other species IgG using the appropriate antispecies reagents. The practical elements of all the assays are very similar, i.e., reagents and equipment needed. The systems described are analogous to the ones most commonly

used to examine problems associated with diagnosis. The schemes will be described using symbols where: I- = solid phase microtiter plate well Ag = antigen Agl, Ag2, etc. = particular antigens highlighted in assay I-Ag = antigen passively adsorbed onto wells I-Ab, I-AB = particular antibodies passively coated onto wells Ab = antibody AB = antibody from a different species to Ab Abx, Aby = different antibodies identified by subscript letters Anti-Ab = antispecies specific antibody (against species in which Ab was produced) Anti-Ab*E = antispecies specific antibody labeled with enzyme W = washing step, involving separation of bound and free reagents + = addition of reagents and incubation step S = substratekhromophore addition Read = read test in spectrophotometer at 492 nm Throughout

Chapters 5-8, many of the practical stages are the same.

The conjugates described are all made with horseradish peroxidase and the substratekhromophore is hydrogen peroxide/orthophenylamine diamine (OPD). The preparation and use of this is described in detail below. 1. Substratekhromophore: This is easiest made up from commercial tablets of OPD that are preweighed. Commercial sources also supply citrate/phos-

Test Schemes

2. 3.

4. 5.

117

phate buffer tablets (pH 5.0). Thus, the volume of OPD can be made as required by following the recommendations by the supplier. As an example, 30 mg tablets are available that make 75 mL of chromophore solution in buffer. Unused OPD solution (without added hydrogen peroxide) can be frozen at -20°C. This can then be thawed and used later. Close inspection should be made to ensure that the OPD is not drscolored. Use complete chromophore/substrate as soon as possible. Larger volumes of OPD in citrate/phosphate buffer can be made and frozen in a tightly stoppered brown bottle in small volumes. The OPD solution should be made and frozen as quickly as possible. Do not use solutions that show discoloration after freezing. Hydrogen peroxide (HzO,) is the substrate for horseradish peroxidase enzyme. This is purchased usually as 30 or 6% w/v and should be stored as recommended by the supplier. The hydrogen peroxide should be kept refrigerated and not subjected to heating. The addition of the hydrogen peroxide should be made immediately before the use of the OPD in the test. Add 5 PL of hydrogen peroxide (30% w/v) to every 10 mL of OPD solution (pH 5.0), or 25 pL of 6% hydrogen peroxide to every 10 mL of OPD solution. Use the substrate/chromophore immediately. OPD is a mutagen, so care is needed in its handling and disposal. Washing solution used in washing steps: This is PBS without the addition of Tween 20. Washing requires the flooding and emptying of wells 4 times with PBS. Blocking buffer: This is PBS containing a final concentration of 1% bovine serum albumin (BSA) and 0.05% Tween 20. This should be made in small volumes as required, but can be stored at 4°C. Care should be taken to avoid contaminated buffer. Stopping solution: This 1M sulfuric acid in water. Care should be taken in its preparation and handling. Read: This implies reading plates using a multichannel spectrophotometer at the appropriate wavelength for the color developing in the ELISA. In all cases for Chapters 5-8, this is 492 nm for OPD. Plates should also be assessedby eye to ascertain whether the test results are as expected.

CHAPTER5

Direct Titration

ELISA

of Antigen

1. Learning

and Antibody

Principles

1, 2. 3. 4. 5. 6.

Measurement of optimum concentration of antigen to coat wells. Measurement of optimum dilution of enzyme-linked antibody. Use of multichannel- and single-channel micropipets. Revision of principles of dilution. Makeup and storage of buffers and solutions. Learning about observation of tests by eye, and by using multichannel spectrophotometers. 7. Handling data. 8. Problem solving.

2. Reaction

Scheme

S + Read + Ab*E + W W I = Microplate wells (solid-phase) Ag = Guinea pig IgG adsorbed to wells Ab*E = Rabbit antiguinea pig conjugated with horseradish peroxidase enzyme S = H,Oa + ortho-phenylene diamine (OPD) (chromophore) Read = Observe by eye or read in spectrophotometer (before or after stopping color development with H,SO,) + = Addition of reagent and incubation at 37OC,or room temperature for 1 h W = Wash wells in PBS (four times) I-Ag

3. Basis of Assay The basis of this assay is to dilute the Ag across the plate one way, in a buffer that allows passive adsorption, incubate the plate at 37°C or room temperature for 2 h, wash the plate, and then dilute the conjugate across the plate, the opposite way to the Ag, obtaining a chessboard titration of Ag against Ab*E. The Ab*E is diluted in a buffer to prevent nonspecific adsorption of the Ab*E to any free protein binding sites on the wells. After

119

Direct ELISA

washing, all the wells receive a solution containing the substrate for the enzyme (H202) and a chromophore that can change color if the Hz02 is acted on by the enzyme. Thus, the color developing in each well depends on: (1) the amount of antigen and (2) the amount of conjugate that has bound to that antigen: The more conjugate, the more enzyme, and the more color!

4. Materials

and Equipment

1. Ag = Guinea pig IgG in PBS at 1 mg/mL. 2. Ab*E = Antiguinea pig IgG prepared in rabbits conjugated to horseradish peroxidase. 3. 96-Well microplate for ELISA. 4. 1ZChannel (tipped) micropipet (S-50 pL). 5. Single-channel micropipet (5-50 pL) plus tips and trough. 6. 10 and 1 mL pipets. 7. Carbonate/bicarbonate buffer, pH 9.5,0.05M. 8. PBS containing 1% bovine serum albumin and 0.05% Tween 20. 9. Solution of OPD in citrate buffer. 10. Bottle of hydrogen peroxide (30% w/v, from 4°C). 11. Washing solution (PBS) reservoir. 12. 1M sulfuric acid in water. 13. Paper towels. 14. Small volume bottles. 15. Multichannel spectrophotometer. 16. Clock. 17. Graph paper.

5. Practical

Details

1. Examine a plate, note the position of letters A-H and the numbers 1-12. Place plate with A at the top left hand corner in front of you, as in Fig. 1. The 8 wells labeled by letters (A-H) will be referred to as rows. The 12 wells labeled by numbers (1-12) will be referred to as columns. 2. Use the lZchanne1 pipet with 12 tips to add 50 pL of carbonate buffer to each well of plate. Use a trough to act as a reservoir for buffer. Add 6 mL to give extra volume needed for the whole plate. 3. Take the antigen (1 mg/mL) and dilute this to 10 pg/mL in carbonate/ bicarbonate buffer. Make up 1 mL of the antigen at this concentration, i.e., add 1 mL of buffer to a small bottle. Pipet 10 pL of antigen into this. Mix well by rotating the bottle by hand (do not be overvigorous). 4. Set the single-channel micropipet to 50 pL. Add 50 p.L of diluted antigen to all the wells of column 1, You should now have 100 pL of antigen in column 1.

Practical

121

Details 1

2

3

4

A

000000000000

B

000000000000

c

000000000000

5

6

‘7

8

9

10 11 12

D000000000000 E

000000000000

F

000000000000

G000000000000

HC,Oi710Cj0000C)OO Fig. 1. Plate layout. 5. Put tips into column 1, and mix contents by pipeting up and down eight times, using the first stop of the pipet. Transfer 50 pL to column 2 (A-H), mix, transfer 50 FL to column 3, and so on, to column 11. After the last mixing, discard 50 pL left in the pipet. You should now have 50 PL of a dilution series in each row, ending with column 11. Check by eye that the volumes are similar in all the wells (Fig. 2). 6. Put a lid on the plate, and leave it on the bench (flat surface) for 2 h at room temperature, at 37°C for 2 h, or if more convenient, leave it at 4°C overnight. 7. Wash the plate. The exact method will depend on equipment used. The principle is to discard contents of wells by “flicking” them into a sink (or suitable container, bowl, and so on), then adding PBS, and flicking this away four times. The major concern is that all the wells are filled at each stage, 8. Turn the plates onto absorbent paper (cloth), and remove the majority of the residual PBS by gently tapping the plates against the paper (picking the plate up to do this, well openings down). 9. Addition of dilutions of conjugate: Take the enzyme conjugate from the refrigerator. Check that it is rabbit antiguinea pig IgG, conjugated to horseradish peroxidase (there may be variations in the particular speciesused to prepare the antiguinea pig serum labeled, e.g., it could be sheep antiguinea pig IgG). Make up 1 mL of a l/200 dilution of the conjugate in a 5-mL bottle. Use the 5-50 p,L single-channel pipet to add the conjugate, i.e., add 5 PL of conjugate to 1 mL of blocking buffer (Remember: We do not

122

Direct ELISA

ofwa/LAJLntm/L 7 1 2 3 4 5 6

8

9

10

11 12

000000000000 000000000000 c 000000000000 D 000000000000 E 000000000000 F 000000000000 G 000000000000 H 000000000000

A B

Fig. 2. Dilution steps for antigen. want any nonspecific adsorption of the conjugate to the plastic during the test). Mix well by gentle swirling action; do not shake vigorously. Add 50 pL of blocking buffer to every well of the microplate using a multichannel pipet fitted with 12 tips. This is accomplished by adding about 6 mL of blocking buffer to a trough and pipetmg from this. Wash the trough, after dispensing the blocking buffer, by tap water (or PBS). Dry the trough with paper towels for use with conjugate. Pour the conjugate dilution into a trough. Using the multichannel with 12 tips attached, add 50 pL of the conjugate dilution mto the first row (A 1-12) of the plate. Thus, we have 100 PL of a l/400 drlution of conjugate in this row. Mix as described above using the multichannel (eight times up and down). Transfer 50 p,L of the conjugate from row A to row B (l-12), mix m row B (eight times). Transfer 50 PL to row C (1-12) and mix. Repeat the transfer of dilutions to the end of the plate (row H). There should now be 50 pL of conjugate dilutions in all wells, as a dilution range from l/400 in row A to l/5 1,200 in row H. Thus, a chessboard titration has been performed relating to how the antigen and antibody have been diluted. This is diagrammatically shown in Fig. 3. Put a lid on the plate and leave at room temperature for 1 h, or at 37°C for 1 h.

Data Explained 1

123 2

3

4

A

@0000000(3000

B

000000000000

tG c

000000000000

G D

000000000000

f

000000000000

F E

5

6

7

8

9

10

11 12

G0000000000G0 Hoooooooooooo

Fig. 3. Dilution steps for conjugate. 10. Wash the plate and flick free from excess washing solution (as described above). 11, Substrate/chromophore addition-color development: Take 10 mL of citrate buffer containing o&o-phenylenediamine (OPD). Thaw out in a water bath or at room temperature (slower only), or make up OPD solution from tablets. Ensure that the solution reachesan acceptabletemperature, i.e., room temperature if this is fairly constant in your laboratory, It is a good idea to have a water bath at a temperature of 20°C and use this to equilibrate the OPD solution to achieve a standardized temperature, since this affects the color development rate of the ELISA. Add 5 pL of hydrogen peroxide (30%)/10 mL of OPD. (Put this straight back in the refrigerator with the top screwed on tightly.) Mix gently. Pour the solution into a trough (must be washed). Use the multichannel to add 50 pL of solution to each well (8 or 12 tips used). 12. Leave the plate on the bench and examine color changes at approx 1,3,5, 8, and 10 min after addition. 13. Add 50 ~.LLof a 1M solution of sulfuric acid in water (supplied) to each well after 10 min color development (use clean trough and multichannels again) to stop color development. 14. Read the plate by eye and using a spectrophotometer at 492 nm.

6. Data Explained Figure 4 shows a diagrammatic representation of plates set up using the same reagents as described, at different times before color development

124

Direct ELISA 1 minute

3 minutes

6 minutes

10 minutes

Fig. 4. Representationof the developmentof color on platesat different times following addition of the substrate/chromophore. has been stopped. Table 1 shows a by-eye assessmentof the development of color. Figure 5 shows the plate OD results from a multichannel spectrophotometer, after being stopped at 10 min. Table 2 shows the OD results for the plate stopped at 10 min. These results are analyzed graphically in Fig. 6, which relates the color developing in wells with different antigen coating concentrations for different dilutions of antispecies conjugate. From the graphs, the optimal dilution of conjugate that might be used in ELISA to detect guinea pig IgG can be determined. Also, the respective antigen concentration that might be used to detect antibodies (highly relevant in the indirect ELISA and explained fully in that section) can be observed. These figures are aimed at people who have never seen an ELISA, as an idea of what to expect, and to act as a comparison with their test results. It will be useful also to those who obtain the text without access to reagents in that it allows them to work through the examples without the need for setting up an actual assay. For people who have performed an assay following the protocols above, they might apply

Data Explained

I25 Table 1 By-Eye Assessment of Plates

Plate observed at: 1 min A B C D E F G H 3 min A B c D E F G H 6 min A B C D E F G H 10 min A B C D E F G H

1 + + +

2 + +-

10

11

-

-

-

-

-

3

4

-

5

6 -

7

8

9

-

++ ++ ++ + + +

++ + + + +

++ + + + +

+ + ++

+

-

-

++ ++ ++ ++ + +

++ ++ ++ ++ + +

++ ++ ++ +-

++ ++ ++

++ ++ ++ ++ ++ + +-

++ ++ ++ ++ + +

- = No detectable color. +- = Weak color + = Definite color ++ = Strong color

++ + +-

+ +-

++ ++ + +-

++ + +

+

+-

-

-

++ ++ ++ ++ ++ + + +

++ ++ +

++ ++ ++ + +

++ ++ ++ + +-

t+ +

+

t-

+-

-

12

126

Direct ELISA

Fig. 5. Diagrammatic

representation

of plate stopped at 10 min.

Table 2 Plate Data

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

1.89 1.87 1.68 1.14 0.99 0.66 0.34 0.21

1.88 1.86 1.45 1.03 0.91 0.44 0.20 0.22

1.67 1.63 1.32 0.94 0.74 0.39 0.16 0.15

1.34 1.29 1.14 0.83 0.54 0.33 0.18 0.18

1.10 1.04 0.96 0.57 0.46 0.24 0.16 0.17

0.97 0.93 0.86 0.45 0.36 0.21 0.18 0.15

0.86 0.84 0.64 0.38 0.29 0.19 0.15 0.13

0.57 0.53 0.45 0.29 0.19 0.15 0.16 0.14

0.44 0.35 0.29 0.19 0.18 0.18 0.14 0.15

0.32 0.24 0.19 0.18 0.15 0.16 0.12 0.13

0.31 0.23 0.17 0.15 0.13 0.14 0.14 0.12

0.31 0.21 0.16 0.16 0.14 0.12 0.13 0.12

the observations on their plate to those demonstrated and compare them critically. 7. Aspects of Assay Described 7.1. Color Development

Little happens during the first 30 s to 1 min after addition of substrate. Color then is detected in wells l-3 of rows A, B, and possibly C. Strongest color is detected in the wells containing the highest concentrations of antigen. By 3 min the pattern should be confirmed, with detectable

Aspects of Assay Described

Antigen

127

dilutions

LoglO

r;>

Fig. 6. Titration curves for conjugateagainst dilutions of antigen (IgG) on columns l-11 of plate. color in rows C,D, and E. After 6 min there is strong color in rows A,B,C, and D, all showing a gradual reduction in color as the antigen is diluted across the plate. The well showing no color (no detectable antibody) for row A is well 10, row B is well 9, row C is well 8, row D is well 6, row E is well 4, row F is well 3, and row G is all rows. At 10 min (the time for stopping color development), there is little change in the pattern, although the intensity of the color may have increased. Notice that at the strongest concentration of the conjugate there may be some color in the negative control well (12). 7.2. Stopping Note the color change on addition of the sulfuric acid. 7.3. Plate Reader Data The color changes associated with each well have now been quantified, so that the exact situation can be assessed(Fig. 6). Each line represents the titration of a different dilution of conjugate against the same

Direct ELISA

dilution range of antigen, Note that rows A and B are very similar. For wells 1 and 2, we have similar color, there being no decrease in color when the antigen is supposedly decreased on the plate. This represents a plateau region (constant high color). Thus, the plate wells have a similar amount of antigen asjudged by the color developing. This means that at antigen levels higher than those of well 2, no more antigen can attach to the plastic of the wells. This is a factor of the binding capacity of the plastic and may vary from protein to protein. From well 3 in rows A and B, we see the color decrease, corresponding to the dilution of antigen on the wells. Note that AlO, 11, and 12 show a similar color (around 0.31). This represents the end point of the titration at the respective conjugate dilutions of l/400. Although row B shows a similar titration range, note that the color in rows 10, 11, and 12 are similar to each other, but lower than for row A. In particular, note well 12 for row B. This is the well that had no antigen, so the color developing in this row represents the nonspecific adsorption of conjugate. Note that the color diminishes in column 12 as the conjugate is diluted, e.g., Cl2 = 0.16, and then stays at a similar level. The conclusion here is that the l/400 and l/800 dilutions of conjugate give some problems of nonspecific adsorption. Below these dilutions, there is no further problem. The residual level of color, independent of the dilution of conjugate, is the plate background and is the result of the change in color of the substrate independent of any enzyrnic activity (oxidation owing to the air, effect of light). The following have now been mentioned: 1. Plateau height.

2. Background,nonspecific adsorptionof conjugate. 3. Plate background.

Note that row C shows a good titration (high levels of color where there is antigen to low levels of color on antigen dilution) range of color. The end point of the titration is around well 9 (last dilution showing color above the plate background). This is similar to rows A and B (since their backgrounds are higher). Note that row D also shows a titration of antigen, although the color is weaker. The end point is now around well 8. This indicates that we are losing some sensitivity in the titration of the antigen at this conjugate

Aspects of Assay Described

129

dilution. Wells E, F, and G demonstrate the loss in sensitivity on dilution of the conjugate, in particular, well G, where there is virtually no titration of the antigen. 7.4. Optimal Dilutions We are now in a position to determine: (1) the dilution of conjugate to be used in an ELISA to detect guinea pig IgG and (2) what dilution of antigen (IgG) can be used on a plate in order to be used in other assays. Remember this test is a demonstration of the principles to be used in specific antigen assays. The same test can be used for their standardization (more clearly demonstrated in Section 7.4.1. dealing with an indirect ELISA). 7.4.1. Optimal

Conjugate

Dilution

The l/400 and l/800 dilutions give good titration with high backgrounds. The l/1600 gives a similar titration curve of similar end point to the l/400 and l/800 with a lower background. Thus, we could use this dilution without loss of sensitivity. The l/3200 also gives an adequate titration of antigen, although there is some loss in sensitivity (ability to react with antigen), asjudged by limiting of end point. Thus, the optimal dilution is somewhere between l/l600 and 113200.In practice, a dilution of l/2000 might be used for initial tests. This might be adjusted after later tests using particular antigens (e.g., if this assay were used to titrate antispecies conjugates, which were then used in the indirect ELISA) (see Chapter 6). 7.4.2. Optimal

Antigen

Dilution

This is of relevance in other ELISAs where specific antigens need to be titrated for use, for example, in indirect assays. In the case of the example above, we might wish to use a constant dilution of IgG to detect antibodies against guinea pig IgG. The levels of IgG available on the wells after adsorption are reflected in the color developing. At high dilutions, there is little color and, therefore, little IgG attached. In the plateau region (at plastic saturation levels), there is an excess of IgG. The optimal amount to titrate antibody is around l-l.5 OD. Units of color are obtained using the optimal conjugate dilution. Therefore, the antigen dilutions in wells 3 and 4 are suitable for reaction with antibody. The exact value can again be adjusted after actual assessments in specific assays.

130

Direct ELBA

8. Conclusions This chapter introduces the worker to the basics of ELISA. Many of the areas covered will need less explanation in the later chapters, so protocols for ELISAs will become shorter. The major use for this direct ELISA is to be able to titrate antispecies conjugates, and thus avoid using too strong or too weak preparations. Some of the major principles of ELISA are also introduced, and the concepts of plateau height, end point, nonspecific reactions, backgrounds, and titration curves will be constantly reviewed in all the assays described.

PART 11

Practical

Exercises

The aim of Chapters 5-8 will be to illustrate the principles of ELISA fully by: 1. Showing worked examplesof eachassay,including diagramsof platesand

representationaldata from assays; 2. Analyzing such data in terms of important rules that are learnedat each stage;and 3. Providing full working instructronsfor workers to be ableto perform each assayso that they obtain their own data to be analyzedas describedin (1) and (2). This includes full instructionson the preparationand standardization of reagents.

The chapters can therefore be used in several ways. Workers without accessto reagentswill obtain a working knowledge of ELISA through the examples. The chapterscan also be used in training courses where reagents may be provided (as indicated in the text). The information will also be useful to workers who have already had some experience of the technique and who may have had difficulties in obtaining and analyzing data. Remember that it is the application of the ELISA to specific problems, and not the methodology for its own sake, that is the most important reason the techniques should be mastered. 1. Test Schemes

You may be already familiar with the concepts in ELISA, whereby an antigen binds to an antibody that can be labeled with an enzyme, or be in turn detected with a species-specific antibody (enzyme labeled). All the ELISAs are variations on this theme. Inherent in the methods of ELISA described in these chapters is the fact that one of the reagents is attached to a solid-phase, making the separation of bound (reacted) and unbound (nonreacted) reagents simple by washing. Before performing ELISA on disease agents, it is useful to train using reagents of defined reactivity, which are easily available and which provide security problems. An ideal

115

116

Practical

Exercises

system is to use an imrnunoglobulin (Ig) and more particularly an immunoglobulin G (IgG) as an antigen. Do not get confused here, since you have learned that the antibody population contains high levels of IgG acting as antibody. In the context of learning the principles, we are using IgG as an antigenic protein, since: 1. IgG from one animal species can be injected into another animal species so that a specific antiserum to that IgG is prepared. 2. Such antibodies can be labeled with enzyme, or detected with a second species-specific antibody labeled with enzyme. Such reagents are defined, easy to standardize, stable, and available commercially. The particular IgG system chosen in most of the chapters involves guinea pig, but similar tests can be performed with other species IgG using the appropriate antispecies reagents. The practical elements of all the assays are very similar, i.e., reagents and equipment needed. The systems described are analogous to the ones most commonly

used to examine problems associated with diagnosis. The schemes will be described using symbols where: I- = solid phase microtiter plate well Ag = antigen Agl, Ag2, etc. = particular antigens highlighted in assay I-Ag = antigen passively adsorbed onto wells I-Ab, I-AB = particular antibodies passively coated onto wells Ab = antibody AB = antibody from a different species to Ab Abx, Aby = different antibodies identified by subscript letters Anti-Ab = antispecies specific antibody (against species in which Ab was produced) Anti-Ab*E = antispecies specific antibody labeled with enzyme W = washing step, involving separation of bound and free reagents + = addition of reagents and incubation step S = substratekhromophore addition Read = read test in spectrophotometer at 492 nm Throughout

Chapters 5-8, many of the practical stages are the same.

The conjugates described are all made with horseradish peroxidase and the substratekhromophore is hydrogen peroxide/orthophenylamine diamine (OPD). The preparation and use of this is described in detail below. 1. Substratekhromophore: This is easiest made up from commercial tablets of OPD that are preweighed. Commercial sources also supply citrate/phos-

Test Schemes

2. 3.

4. 5.

117

phate buffer tablets (pH 5.0). Thus, the volume of OPD can be made as required by following the recommendations by the supplier. As an example, 30 mg tablets are available that make 75 mL of chromophore solution in buffer. Unused OPD solution (without added hydrogen peroxide) can be frozen at -20°C. This can then be thawed and used later. Close inspection should be made to ensure that the OPD is not drscolored. Use complete chromophore/substrate as soon as possible. Larger volumes of OPD in citrate/phosphate buffer can be made and frozen in a tightly stoppered brown bottle in small volumes. The OPD solution should be made and frozen as quickly as possible. Do not use solutions that show discoloration after freezing. Hydrogen peroxide (HzO,) is the substrate for horseradish peroxidase enzyme. This is purchased usually as 30 or 6% w/v and should be stored as recommended by the supplier. The hydrogen peroxide should be kept refrigerated and not subjected to heating. The addition of the hydrogen peroxide should be made immediately before the use of the OPD in the test. Add 5 PL of hydrogen peroxide (30% w/v) to every 10 mL of OPD solution (pH 5.0), or 25 pL of 6% hydrogen peroxide to every 10 mL of OPD solution. Use the substrate/chromophore immediately. OPD is a mutagen, so care is needed in its handling and disposal. Washing solution used in washing steps: This is PBS without the addition of Tween 20. Washing requires the flooding and emptying of wells 4 times with PBS. Blocking buffer: This is PBS containing a final concentration of 1% bovine serum albumin (BSA) and 0.05% Tween 20. This should be made in small volumes as required, but can be stored at 4°C. Care should be taken to avoid contaminated buffer. Stopping solution: This 1M sulfuric acid in water. Care should be taken in its preparation and handling. Read: This implies reading plates using a multichannel spectrophotometer at the appropriate wavelength for the color developing in the ELISA. In all cases for Chapters 5-8, this is 492 nm for OPD. Plates should also be assessedby eye to ascertain whether the test results are as expected.

CHAPTER6

Indirect

ELISA

1. Learning Principles 1. Measure optimal antigenconcentrationto coat wells; 2. Titration of antisera;and 3. Use of antispeciesconjugates. 1.1. Reaction I-Ag

+

Ab

+

Anti-Ab*E

w W I- = Microplate wells

Scheme + S + Read W

Ag = Guinea pig IgG adsorbed to wells Ab = Rabbit antiguineapig serum

Anti-Ab*E = Goat antirabbit serumconjugatedwith horseradishperoxidase S = Hz02 + orthophenylenediamlne (OPD) Read= Observeby eye or readin spectrophotometer + = Addition and incubation at 37’C or room temperaturefor 1 h W = Washwells with PBS 2. Basis of Assay The basis of this assay is to titrate antibodies that have reacted with an antigen by using an antispecies conjugate. The indirect aspect, therefore, refers to the fact that the specific antiserum against the antigen is not labeled with an enzyme, but a second antibody specific for the particular species in which the first antibody was produced is labeled. Such assays offer flexibility and form the bases of other ELISAs. In principle, the optimization of reagents is similar to the direct ELISA. However, three factors have to be considered: 1. The optimal dilution of antigen. 2. The optimal dilutions of antisera. 3. The optimal dilution of conjugate. Point 3. has been dealt with for the direct ELISA. You should now be able to titrate the conjugate (antirabbit in this case).The major use of 131

Indirect

132

ELBA

indirect ELISA is to titrate antibodies against specific antigens. In this case, a constant amount of antigen is adsorbed to wells, and antisera are titrated against this as dilution ranges. Any antibody reacting is then detected by addition of a constant amount of antispecies conjugate. Such assays can be evaluated fully from the diagnostic point of view where numbers of field and experimental antisera (known history) are available. Therefore, they can be used to assay single dilutions of antisera, and tests can be adequately controlled using standard positive and negative antisera. Thus, the indirect ELISA has found many applications in epidemiological studies assessing disease status. 3. Materials

and Reagents

1. Antigen (Ag) = guineapig IgG at 1 mg/mL (1 g/L). 2. Antibody (Ab) = rabbit antiguineapig serum. 3. Anti-antibody*E = sheepantirabbit serum linked to horseradishperoxidase(rabbit IgG neededif conjugatetitration not made),as for titration of antiguineapig conjugate). 4. Microplates. 5. Multichannel and single-channelpipets. 6. 10 and 1 mL pipets. 7. Carbonate/bicarbonatebuffer, pH 9.6,0.05M. 8. PBS containing 1% bovine serumand 0.05% Tween 20. 9. Solution of OPD in citrate buffer. 10. Bottle hydrogenperoxide (30% w/v). 11. Washing solution (PBS) in bottle or reservoir. 12. 1M sulfuric acid in water. 13. Papertowels. 14. Small-volume bottles. 15. Multichannel spectrophotometer. 16. Clock. 17. Graph paper. 4. Practical

The first stage in this assay involves the titration of the antispecies conjugate under the conditions described in the direct ELISA. Remember that the antigen used to titrate the conjugate must be appropriate, e.g., if an antibovine conjugate is to be used, then use bovine serum as the antigen in the original chessboard, If an antibovine IgG detection is required, then use bovine IgG as the antigen in the direct ELISA chessboard titration. The antirabbit conjugate needs to be titrated so that we

Practical

133

know the dilution to use in the indirect assay in order to detect any reacted rabbit serum (the optimal dilution of conjugate may be given in class if this procedure has not been carried out). Thus: 1. Titrate the antirabbit conjugate (optimum dilution may be given). 2. Take microtiter plate with Al at the top left-hand comer. Add 50 FL of carbonate/bicarbonate buffer to each well using a multichannel pipet. 3. Make a dilution range of the guinea pig IgG from 5 l.tg/rnL from column 1 (8 wells) to column 11. This is made exactly as described for the direct ELISA. Add 50 pL of the guinea pig IgG at 10 p.g/rnL (or l/50 if concentration unknown) to column 1. Mix (pipet up and down eight times with the multichannel), then transfer 50 pL to column 2, mix, and continue transfer to column 11. Discard 50 l,tL remaining in tips after mixing in column 11. Thus, we have a twofold dilution range of IgG in each row A-H, excluding column 12 wells. 4. Incubate at room temperature or 37OCfor 2 h. 5. Wash the wells in PBS (fill and empty wells four times). 6. Blot the plates. 7. Take the rabbit antiguinea pig serum, and dilute it to l/50 in blocking buffer (PBS containing 1% BSA and 0.05% Tween 20). Make up 1 mL. Therefore, add 20 p.L to 1 mL of buffer. 8. Add 50 p,L blocking buffer to all wells using a multichannel. 9. Add 50 ~.LLof the l/50 antiguinea pig serum to each well of row A. Mix and transfer 50 l,tL to row B, mix, and transfer 50 l.tL to row C, and repeat this procedure to row H. We now have a twofold dilution series of antibody the opposite way to the IgG antigen. 10. Incubate the plate at room temperature or 37OCfor 1 h. 11. Wash and blot the plate. 12. Make up the antispecies conjugate (kept at -20°C) to the optimal dilution found in the direct ELISA (or as instructed) in,blocking buffer. Make up enough for all the wells of the plate + 0.5 mL (approx 5.5 mL). This might appear wasteful, but is convenient practice since it allows for minor errors in pipeting and avoids having to make up a small volume of conjugate when one “runs-out” on the last row (i.e., when the exact volume to fill the plate wells is made up). Add 50 PL of the dilution to each well using the multichannel and a clean trough. 13. Incubate at room temperature or 37OCfor 1 h. 14. Wash and blot the wells. 15. Thaw out the OPD (10 mL). Add 5 pL of HzOz immediately before use. Mix well. Add 50 pL of this to each well, using multichannel and clean

troughs(make surethat the troughis not contaminatedwith conjugatefrom previous addition to the plate).

Indirect

134

ELBA

Table 1 Plate Data from Section 4.1.

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

1.92 1.94 1.56 1.34 1.14 0.92 0.76 0.45

1.89 1.89 1.43 1.23 1.00 0.83 0.56 0.32

1.92 1.91 1.33 1.14 0.89 0.73 0.42 0.29

1.89 186 1.29 1 09 0.76 0.54 0.36 0.21

1.45 1.47 1.07 0.97 0.56 0.43 0.28 0.17

1.12 1.09 0.89 0 75 0.41 0.32 0.21 0.14

0.89 0.87 0.78 0.68 0.32 0.21 0.19 0.15

0 67 0.59 0.56 0.49 0.23 0.17 018 0.18

0.45 0.39 0.43 0 29 0.19 0.19 0.16 0 16

0.39 0.38 0.32 0.21 0.17 0.16 0.14 0.15

0 40 0.31 023 0.17 0.19 0.16 0.15 0.16

0.39 0.29 0.19 0.15 0.12 0.14 0.15 0.10

16. Incubate for 10 min (note color changes). 17. Stop any color development by adding 50 pL of l.OM sulfuric acid to each well. 18. Read the plate by eye and at 492 nm by multichannel spectrophotometer after titration of antigen (guinea pig IgG) and antibody (antiguinea pig serum). Table 1 shows the microplate reader results. Note that these produce a similar picture to the direct ELISA results in Chapter 5, and you should also have observed that there was a similar development of color throughout the 10 min incubation after addition of the substrate solution. Figure 1 shows the data graphically. Plots relating the concentration (or dilution) of the IgG (Ag) to the OD for all the different dilutions of rabbit anti-IgG are shown. Plot your ELISA data as shown in Fig. 2. Thus, relate the IgG concentration on the plate plotted as a loglo twofold series (pg/rnL/ well, or dilution if actual concentration is unknown) against the OD for each dilution of antibody used. You should end up with eight lines on a single graph, one for each antiserum dilution. You have already observed in the direct assay similar results. Similar areas of reactivity can be identified on the indirect chessboard. 1. Plateaus of similar high color are shown in rows A and B wells 14. 2. There are higher plate background values m rows A and B (possibly C) than for more dilute serum. 3. The serum titration

end points (where

OD value for a particular

IgG con-

centration is the same as plate background), are similar for rows A, B, C, and D. After this dilution of antiserum there is loss in detection of IgG.

Practical

135

0.5

0

1

2

3

4

5

6

7

8

Antibody dilutions L05~(2 fold) + Fig. 1. Titration curves of antibody for different concentrations of antigen (IgG) on columns 1-12 of plate. 4. Loss of end pomt detection is matched by a loss m OD at high concentrations of IgG, e.g., m rows F, G, and H at 5 l.tg/mL of IgG, there is substantial and increasing loss in color, as compared to where maximal color (in antibody excess- row A) is observed. Note that row H hardly titrates the IgG, with very low color being obtained. 4.1. Optimization of Reagents

Rows A and B indicate that antibodies are in excess, and there are some problems of nonspecific attachment to the plate without antigen having been adsorped (well 12). Note that in these rows the plateau regions extend to well 4. Thus, no more of the antigen (IgG) is able to absorb to the plate above the concentration in well 4. Rows C and D give

Indirect

136

01 1

I

I

I

I

I

I

I

I

I

I

2

3

4

5

6

7

8

9

10

11

ELISA

12

Zfold dilution of guinea pig IgG antigen d

-A

+B++C

+D

*E

+F

*G

*H

Fig. 2. Titration of dilution series of guinea pig IgG antigen (l-l 1) against eight different concentrations (rows A-H) of rabbit antiguinea pig IgG. Antirabbit conjugate constant.

optimal titrations of the IgG in that maximum values do not exceed 1.6 ODs, and high end point titers are obtained. Below these dilutions, sensitivity for the detection of IgG is lost. Thus, in order to detect the antigen optimally, to use a single dilution of antiserum under the conditions of the ELISA described, use a dilution of around l/400-1/800. The optimum dilution of antigen that might be used as a single dilution to detect and possibly quantify antibodies is best assessed as the dilution (or concentration) that shows good binding across the whole range of antiserum dilutions. The best way to illustrate this is to draw a graph of the plate data, but this time, plot the dilutions of serum against

Use of Indirect

ELISA

137

to Titrate Antibodies

the OD for the various antigen concentrations (or dilutions). This was done in Fig. 1. At the first four concentrations (dilutions) of antigen (IgG), there is little difference in the end point detection for the dilutions of antiserum. After this, the OD readings and the end point detections are reduced. At the extreme, in column 10, hardly any antibody is detectable even where the serum is most concentrated. The higher values in row A, B, and C correspond to the nonspecific binding to the wells seen in row 12. Thus, the dilution of antigen found in columns 3 and 4 is optimal to detect antibodies. 6. Use of Indirect ELISA 5.1. Learning

to Titrate Principles

Antibodies

1. To titrate antibodies from positive sera using full dilution ranges. 2. To establish ELISA negative antibody levels for control nonimmune sera. 3. To duplicate samples tested.

5.2. Reaction I-Ag

+ w

Ab

+ w

Scheme

Anti-Ab*E

+ S + Read w

I- = microplate Ag = optimum concentration of antigen Ab = test serum + or - in reaction for Ag Anti-Ab*E = antispecies antibody linked to enzyme S = substrate/color detection system W = wash + = addition and incubation of reactants

In this exercise, the Ag and anti-Ab*E are used at optimal dilution. The test or standard Abs are added as dilution ranges. 5.3. Basis of Assay We are now in the position to titrate antibodies, since we know the antigen optimum and the conjugate optimum dilutions for our given system. Thus, if sera are reacted with the antigen on the plate and if they contain antibodies against the guinea pig IgG, they will be picked up by

the subsequent addition of the conjugate. The seropositive serum titration curves may then be compared to each other and to the seronegative curves to establish antibody titers and examine the result of nonspecific reactions at the various dilutions of the negative sera, within the system.

138

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Indirect

ELISA

5.4. Materials and Reagents Ag = guinea pig IgG, 1 mg/mL Ab = 3x rabbit serum samples after injection with guinea pig sera test bled after different times following inoculation and three rabbit sera from antibody-negative animals. Anti-Ab*E = sheep antirabbit serum linked to horseradish peroxidase. Microplates. Multichannel and single-channel pipets. lo- and 1-mL pipets. Carbonate/bicarbonate buffer, pH 9.6,0.05M. PBS containing final concentration 1% BSA and 0.05% Tween 20. OPD. Hydrogen peroxide. Washing solution (PBS). 1M sulfuric acid in water. Paper towels. Small-volume bottles/microdilution system. Multichannel spectrophotometer. Clock. Graph paper. 5.5. Practical 5.5.1. Titration of the Antigen Dilution or Concentration for Use in Measuring Antibodies

This is performed as described in Section 4. l., where we also have titrated the optimum dilution of conjugate. We are concerned now with the titration of antibodies against guinea pig IgG in rabbit sera. Therefore, make the chessboard titration of guinea pig IgG against the positive rabbit antiserum, and use a constant dilution of antirabbit conjugate. Note that when one is setting-up indirect ELISA, a positive serum against the particular antigen being detected is needed. Such sera are often available as determined from other serological assays or from systems whereby specific antibodies are expected, e.g., from experimentally infected or vaccinated animals or from animals during the course of an outbreak. The exact conditions of the ELISA may therefore have to be

altered during the developmental stages when many sera have been examined as compared to the originally used positive serum. This will be further examined below. For now the original “optimal” conditions are determined using a defined (experimentally derived) positive serum.

Use

of Indirect

ELZSA to Titrate Antibodies

139

1 2 3 4 5 6 7 8 9 10 11 12

5.6. Titration of Different Sera 1. Dilute guinea pig IgG (Ag) to optimum concentration in carbonate/bicarbonate buffer, pH 9.5, 0.05M. 2. Add 50 pL to each well of the plate using a multichannel pipet. 3. Incubate at 37°C for 2 h. 4. Wash and blot the plate. 5. Add 50 l.tL of blocking buffer to all wells. Use the multichannel and trough. 6. Take the six sera supplied. Label the three positive sera 1,2, and 3. Label the three negative sera 4, 5, and 6. Dilute each one to l/20 in blocking buffer in small bottles. Make up a final volume of 0.5 mL of each (25 + 475 pL blocking buffer). 7. Take the plate with 50 l.rL blocking buffer/well. Turn so that well Hl is on the left-hand top corner (see Fig. 3). Add 50 pL serum 1 dilution to wells Hl, and H2. Add 50 PL of serum 2 dilution to wells H3 and H4. Add 50 PL of serum 3 dilution to wells H5 and H6. Repeat the process adding sera 4,

140

Indirect

ELISA

FFFFFFF

HGFEDCBA Serum 1 Serum 2 Serum 3

00000000 00000000 00000000 00000000 00000000 00000000

1 2 3 4 5 6 7

Serum 4

8 9

Serum 5

10 11

Serum 6

12

Initial dilution of the sera

Fig. 4. Addition and dilution of sera of plate.

8.

9. 10. 11. 12. 13. 14.

5, and 6, to wells H7, H8, H9, HlO, Hll, and H12. We now have each of the sera diluted effectively to l/40 in 100 l,tL blocking buffer in the left-hand extreme row (H) of the plate. Use the multichannel with 12 tips attached to mix, and dilute the sera across wells G, F, E, D, C, B, and A, transferring 50 FL of each dilution. We now have a twofold dilution range of the sera, in duplicate, i.e., there are two dilution series of each of the sera (see Fig. 4). Incubate at 37OC or room temperature for 1 h (exact conditions used in indirect chessboard titration are best). Wash and blot the plate. Add 50 pL of antu-abbit conjugate per well (diluted m blockmg buffer). Incubate at 37°C (or room temperature) for 1 h (conditions as for 1 h incubation in chess-board titration). Wash and blot the plate. Add substrate and chromophore (50 pL).

Use of Indirect

ELISA

40

141

to Titrate Antibodies

! 80

160

320

a-fold dilution

640

1280

2560

5120

of sera L--v’

Fig. 5. Plots of datain Table 4 showing titrations of 6 seradiluted from l/40 in a twofold seriesagainstconstantantigen.Mean values of OD areused. l-6 are different serum samples. 15. Stop color developmentafter 10 min. 16. Read the plate in the multichannel spectrophotometerat 492 nm. Note: Remember to watch the plate as the color develops and make relevant notes. 5.6.1. Data Explained Figure 5 shows typical results from this assay. (The OD readings are shown in Table 2). Examine serum 1: Note that the values of the duplicate samples are very similar. The titration shows a plateau region where the values are the same (wells Hl and 2, and wells Gl and 2). Thus, there \is a maximum color obtained up to l/80; increasing the concentration of antibodies has no effect on the readings. This represents the region where all the

Indirect

142

ELBA

Table 2 Plate Data from Section 5.5. 12 A B C D E F G H

3

0.34 0.54 0.87 1.16 145 1.68 1.76 179

Sera

0.32 0.56 0.91 1.14 1.49 1.70 1.73 1.76 1

456789

0.19 0.34 054 0.76 0.95 115 1.34 1.56

0.23 0.36 0.57 072 0.91 1.17 1.32 1.54 2

0.14 0.17 0.18 0.28 0.31 0.43 0.65 0.78

0.15 0.19 0.19 0.25 0.32 046 0.66 0.76 3

0.17 0.14 0.14 0.17 0.15 0.13 023 0.31

0.16 0.14 0 17 0.16 0.14 0.15 0.24 0.32 4

0.15 0.16 0.17 0 18 0.17 0.14 0 18 0.28

10

11

12

0.16 0.18 0 16 0 16 0.15 0.15 0.17 0.24

0.19 0.16 0.17 0.17 0.14 0.10 0.15 0.21

0.15 0.14 0.14 0.19 0.17 0.18 0.17 023

5

Dilution 5120 2560 1280 640 320 160 80 40

6

antigen is saturated with antibody. The value of the OD is dependent on the amount of antigen that has attached to the wells, which in turn is dependent on the adsorption characteristics of the plastic and the concentration of antigen. On further dilution, the antibodies are no longer in excess,so they aretitrated as seenby a gradual decreasein the OD observed. Serum 2: Note that the OD levels even at l/40 are not equal to those where we had antibody excess in serum 1. Thus, the antibodies are not saturating the antigen on the wells, thus are not present in excess. The titration of the serum begins immediately on dilution. Note that the last dilutions give low OD values equivalent to the plate background, unlike serum 1. Serum 3: Even at l/40, we have low OD values as compared to the serum samples 1 and 2. There are fewer antibodies in this serum than in the other two! Again the titration begins immediately on dilution and the low OD (around 0.19) is attained at l/1280. Overall, these three positive sera can be seen to have different reactivities in terms of the quantity of antibody titrated. Thus, serum 1 has the highest titer, showing a plateau (is able to saturate the antigen). Serum 2 has the next highest amount of antibodies, since it has an end point around l/5120 (point where OD equals the plate background). Serum 3 has the lowest amount of antibody, with an end point of around l/1280. Serum 4: This is a negative serum (clinically). Therefore, by definition it should contain no antibodies. The color obtained reflects the nonspecific attachment of the serum to the antigen. Most nonspecific “sticking” might be expected in the least dilute sample. This is what is found here, with

Use of Indirect

ELISA

143

to Titrate Antibodies

2.0

1.5 E3 3 3 n

1.0

0

0.5 background

0

OD

I

I

Dilution

I

I

I

I

of serum (Loglo ) e

Fig. 6. Serum titration curve showing sigmoidal nature.

background levels being found at l/40 and l/80 serum dilutions. Note that the levels of nonspecific color are much lower than in the positive sera, but are distinct from the assumed plate background, which can be taken as the backgrounds observed for the negative sera at their highest dilution. Note that such wells as (E, D, C, B, and A) give similar results, and no titration is observed on dilution, Sera 5 and 6 give similar results to serum 4, although there is a lower amount of color in the l/80 wells, reflecting different amounts of nonspecific adsorption of serum proteins. 5.6.2. Curve Shapes

The plotted data, particularly for the positive sera, produces curves rather than straight lines, Generally, there is a region on this curve that contains three to five points that are more linear than the rest. The nonlinear regions occur at the top and the bottom of the graphs (see Fig. 6). Such

Indirect

ELISA

sigmoidal curves are typical of serum titrations. Note that the more linear regions of the positive serum titration curves are parallel. Note also that the end point determinations are difficult to assess exactly, since there is a pronounced “tail” at the low OD end of the results. 5.6.3. Comparison

of Serum Titration

Curves

The amount of specific antibody in each serum has been titrated over a dilution range. The serum containing the most antibodies will have a higher dilution end point (dilution where the OD is the same as the background OD). Thus, as indicated above, the end points may be compared as representing the titer of the sera. This can be assessedby eye, as well as by machine reading. A better estimate of the end point is made by drawing a straight line through the points on the curves that are nearly in a straight line. If this was done statistically, then a regression analysis of the points would be made and the best line of fit would be given mathematically. Graphically, this may be approximated to sufficient accuracy (see Fig. 7). Thus, the end points are assessedwhen the regression lines (or graphically produced lines) cut the measuredbackground OD line. This assumes that the curves are similar shapes (the lines are then parallel). This may not always be the case, since different antibody populations may be responsible for the ELISA color (seeFig. 8). In this case, it must be noted and taken into account when the implications of the titers found are considered. Note that the curves obtained for the negatives above arevery flat, even so they have an end point. Sera may also show differences in maximum plateau heights (seeFig. 8). Figure 9 attempts to explain why there are differences in plateau heights for different sera. Here several sera are reacting maximally with the antigen, since on increasing their concentration, there is no increase in color. The plateau heights are different, however, showing that different weights of antibody have reacted with the same antigen for particular sera. This is a function of the number of reactive antigenic sites on the antigen and the quantities and specificities of the antibody populations in the sera. Although this is uncommon using polyclonal antibodies, it is common when using monoclonal antibodies. Where the curves are parallel, any point can be taken on them for comparison of samples. This is illustrated in Fig. 10. Analysis of as many sera as possible over full dilution series and examination of the curves

Use of Indirect

ELISA

145

to Titrate Antibodies

Points

Background I

X to Y regressed

OD I

I

I

I

I

Dilution of serum(Logt,,) 2 fold

Fig. 7. Regression of points in a serum titration to obtain a titer at the intersection of the background OD.

should be made to establish whether there is parallelism. This is important where “spot tests” are required, so that a single dilution of test sample can be established. The dilution can be taken where samples give results in the parallel regions of curves. A line is drawn at a particular OD, and the dilution of serum giving this OD for all the sera is determined, thus giving relative titers. Such relative titers may be expressed compared to an accepted standard serum, which in turn can be given in any units. The actual activity of the standard serum may be known, e.g., number of j,tg/rnL specific antibody, so that all the sera compared to this can be expressed in the same units. 5.7. Negative Sera and Control Sera The test made involved only one control, that of negative sera. Ideally, a plate background should be included to measure the color in wells with

Indirect

146

ELBA

2

1.5 8 R d

1.0

Q

0.5

0 Antibody

dilution

(LoglO

) 2 fold

4

Fig. 8. Variation in sigmoidal curves for serum titrations. antigen and conjugate only. This should correspondto the readings beyond the titration of the antibodies, observed when a low plateau is obtained even on dilution of the samples. Such backgrounds can be subtractedfrom the whole-plate results before any processing of the data, or used to blank the spectrophotometerbefore reading. The treatment of the negative serum results depends on what is known about the negativity in terms of other tests and clinical findings, e.g., British cattle are ideal as negative sera when studying antifoot-and-mouth disease antisera, since Britain is disease-free. This may not always be possible in countries where disease is endemic. Note also that control negative sera obtained from other countries may not reflect the same negative population of another country,

since there are breed differences, complications owing to other infections, and so on. This could affect the performance of kits where standard negative sera are supplied to act as controls in the ELISA. Kits must be evaluated, wherever possible, in the country where they are to be used.

Use of Indirect

ELISA

to Titrate

Antibodies

147

Antibodies present in serum Serum 1 Serum 2 s>

>

-c CI

Maximum 8 molecules bind

Maximum 4 molecules bind

Fig. 9. Diagram to represent maximum number of molecules of antibody that can bind to antigens. Difference in plateau heights can be attributed to different populations of antibodies in sera. The control value for the negative serum supplied may not reflect the mean value for “negative” sera. The immunological implications are dealt with earlier. 57.1. Selection of a Single Serum Dilution to Perform a “Spot-Test” Examination of the serum titration curves for positive and negative sera can tell us which dilution might be suitable to use in the indirect ELISA so that antibodies may be assayed on single wells (or multiple wells using the same dilution). Thus, as shown in Fig. 5, we observe that there is low nonspecific activity seen in the negative sera at l/40 and l/80. The positive sera still show high OD values at these dilutions, so that the relative sensitivity of the assay (detection of specific anti-

148

Indirect

oL 1

2

3

4

5

Dilution

sera -A

4

Titres obtained

dropping

perpendiculars

to x axis

7

8

9

10

11

ELISA

12

of sera

+ B *C

*D

Fig. 10. Comparison of serum titratton curves to standardserum titration at threepoints (OD values 1, 2, and 3 m parallel regionsof curves).Titers can be readfrom x axis and related(representedby gray lines). bodies) can be made at such dilutions. However, if dilutions greater than l/80 are used, we can still measure antibody in the absence of nonspecific reactions. The sensitivity does drop, however. Remember that we are trying to balance sensitivity with low background in the presence of other serum proteins in the sample. If we had used the sera only at l/160, then we would have had values for the ELISA as shown in Table 3. The negative sera levels are therefore, around 0.15, whereas all the positive sera are above this value. The next exercise will expand on this approach.

Indirect

ELISA

to Determine the Positivity

of Sera

149

Table 3 Mean ODJg2of Antiguinea Pig Seraat l/160 Dilution OD Serum 1 1.69 2 3 4 5 6

1.16 0.45 0.14 0.15 0.17 From Table 2.

6. Use of Indirect ELISA to Determine the Positivity of Sera at Single Dilution 6.1. Learning Principles 1. To examme negative serum populations for establishing OD limits of negativity; 2. To examine antibody-positive serum populations; and 3. To examine frequency of results in a population.

I-Ag

+

6.2. Reaction Scheme + S + Read Ab + Anti-Ab*E W W

w I- = microplate Ag = optimum concentration of antigen Ab = test sera at single dilution AntiAb*E = antispecies antibody linked to enzyme S = substrate/color detection system

In this exercise, we use Ag and anti-Ab*E at optimal dilutions. The test sera are added at a constant dilution. Control-positive antisera can be added at a constant dilution or as a dilution range to produce a standard curve relating color to dilution or concentration of antibodies added. Thus, the test sera can be related to the positive serum titration curve. The samecan be doneby including acceptednegative control serastandards. 6.3. Materials and Reagents 1. Ag = guinea pig IgG 1 mg/mL (or previously titrated). 2. Ab = 48 rabbit sera, including high, moderate, and low titer againstguinea pig IgG (24) and negative sera (24).

Indirect 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

ELISA

Anti-Ab*E = sheep antirabbit serum linked to horseradish peroxidase. Microplates. Multichannel and single-channel pipets. lo- and 1-mI. pipets. Carbonate/bicarbonate buffer. PBS containing 1% BSA, 0.05% Tween 20. OPD solution. Hydrogen peroxide. Washing solution. Paper towels. 1M sulfuric acid in water. Small-volume bottles/microdilution equipment. Multichannel spectrophotometer. Clock. Graph paper. Calculator.

6.4. Practical From earlier exercises, you should have assessed the dilution of test serum that can be used to discriminate between positive and negative nonspecific results, based on the difference noted between the selected positive and negative sera titrated over full dilution ranges. We are going to titrate all the sera at the dilution found as duplicates (2 wells/serum dilution in the indirect ELISA). 1. Add the guinea pig IgG to the wells of a microtiter plate at optimum dilution (as in earlier exercises). Incubate at 37°C for 2 h (or under particular optimal conditions). 2. Wash and blot the plate. 3. Dilute the test serum samples appropriately in blocking buffer. Sera may be diluted into small volume bottles. However, this causes two problems: a. Manipulation (capping, and so on) is laborious; and b. Transfer of serum dilutions must be made with a single-channel prpet. Point (b) is important since it takes a long time to transfer all the sera to the different wells. The initially added samples will therefore receive a longer contact time with the antigen, and this may well affect the results. This can be avoided if the samples are transferred to other plates before dilution, e.g., plastic non-ELISA rnicrotiter plates in volumes that need

not be accurate. The plate can then be sampled using a multichannel pipet if the dilution factor for the serais not too high. The initial dilution could

Indirect

ELISA

to Determine the Positivity

of Sera

151

Samples A 1-12 13-24

2536

37-48

6

C D E F G H

00000000000 00000000000 000000 00000 00000000000 00000000000 00000000000 00000000000 00000000000

Fig. 11. Use of micronics systemfor dilution of samples.Order of samples. be made directly into, say, 100 l.tL of blocking buffer in the non-ELISA plates. The transfer of the required volume of the diluted test sample can then be effected using a multichannel pipet. Thus, the samples are transferred at approximately the same time. Special systems have been developed for use with multichannel pipets. These are ideal for the dilution and storage of test samples. Volumes of about 1 mL can be made up making the accurate dilution of up to l/200 (5 FL sample/ml) easy. The microtiter dilution system should be available for this exercise. Add a volume of blocking buffer to the plastic tubes held in the tube holder, e.g., if a dilution of l/100 is required, add 0.5 mL of blocking buffer/tube, then add 5 PL of test sample. If a l/80 dilution is required, see Fig. 11 for pattern of samples on plate. 6.4.1. Example of Data Typical results are shown in Table 4. The results obtained in your specific assay can be processed in the same way. Figure 12 shows a representation of a stopped plate. Since duplicates have been made, examine the variation between the values. This should not be high, i.e., there should be little difference between the ODs for both test wells of the samesample.Take the mean(averageresult) of the OD from both wells if the difference is not large. Variation in results will be discussedlater in the text. Take the mean value to two decimal places.

152

Indirect

ELISA

Table 4 Plate Data from Section 6.4.

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

1.21 1.19 1.00 0.97 0.13 0.12 0.15 0.13

1.09 1.03 0 23 0.27 0.14 013 0.18 0.16

0.78 0.69 0.45 0.49 0.18 0.16 0.13 0.13

0.32 0.31 0.56 0.54 0.09 0.09 0.14 0.15

0.12 0.16 0.78 0.72 0.07 0.08 0.10 0.12

0.66 0.64 0.13 0.16 0 12 0.11 0.15 0.13

0.65 0.62 0.19 0.20 0.14 0.13 0.13 0.12

0.17 0.16 0.45 0.44 0.09 0.10 0.12 0.14

0 67 0.64 0.56 0.53 0.08 0.09 0.13 0.15

0.34 0.37 0.78 0.75 0.12 0.11 0.13 0.12

1.34 1.28 1.00 1.01 0 16 013 0.12 0.11

1.11 1.17 0.56 0.55 0.14 0.15 0.08 0.09

Duplicates of samples made A 1, B 1, A2, B2, and so on Suspect positive sera (24) rows AB and CD Negative (prebleed sera) rows EF and GH.

6.4.2. Mean and Standard Deviation from Mean of Negative Serum Data

Take all the means of the negative sera, and calculate the mean and standard deviation of the negative population using a calculator. Note: Instruction should be taken on the use of the calculator. Noncourse users should obtain a calculator and follow instructions for use to calculate the same parameters. 6.4.3. Frequency Plots of Negative

Serum Results

Plot the results for the negative sera as shown in Fig. 13. These relate the number of samples giving a particular OD. A frequency distribution is obtained so that the distribution of negative results is obtained. Make out a table of OD intervals, and score the numbers of sera falling into the intervals. Add up the score, and plot this against the intervals. The mean of the data for the negative sera and the standard deviation of the data can be found using a calculator. Thus, the population mean of a limited (in this case) negative population has been found. If the population of negative ELISA readings is distributed normally (normal distribution statistics), then the upper limits of negativity can be ascribed with defined confidence limits depending on the number of standard deviations from the mean that are used. The mean value in this case is 0.125, and the SD is 0.026. Thus, if we select 3x SD above this mean value (= 0.084) and add this to the mean value (=0.209), any values equal to or above this value are unlikely to be

Indirect

ELISA

to Determine the Positivity

of Sera

153

0.0.. 000000000000 000000000000 000000000000 000000000000 Fig. 12. Diagrammatic

representation

of plate.

part of the measured negative population as defined by the fact that only approx 0.1% of negative sera examined tended toward this value. Limits using 2x the SD above the measured negative population mean reduced confidence in the results for ascribing positivity (increase the possible sensitivity, but reduce specificity). In practice, such distributions are skewed to the right-hand side, so that a tailing of results is seen at the higher ODs (see Fig. 14). In order to establish an OD reading that reflects the upper limit of negativity (since all negative sera have been studied), a statistical evaluation of the distribution is required. In general, since the distribution is skewed, a value of 2x the mean OD for all the negative sera has been found to determine the upper limit of negativity (which corresponds to the lower limit of positivity) with a 99.0% confidence limit. Thus, we are 99.0% certain that a sample giving an OD value of equal to or greater than the value at 2x the mean of the negative population OD results is positive. 6.4.4. Problems When examining “negative” populations, we are assuming such seronegativity using one or more factors, such as other serological test results, knowledge of the clinical history of the animals, epidemiological factors, and so on. Thus, it may be easy to identify “sero-negative” ani-

Indirect

154

ELISA

10

5

0 0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18 0.20

Interval Fig. 13. Frequency plot relating number of sera giving particular OD values. mals in countries where a particular disease has never been recorded. This may not be true for countries that have endemic disease or where vaccination campaigns have been mounted at various times (with variable amounts of antibody against specific disease agents being elicited). In such conditions, the experimentor might make the best assessment of likely negative animals and follow the exercise as shown above. In this case, after plotting the frequency curves, one of several distributions might be obtained. 1. Figure 15: One peak at low OD end of distribution. Probably negative population; all sera showing low OD. 2. Figure 16: Two peaks fairly well separated at the low OD end and at the higher OD end of distribution. Distinct populations of animals that are positive and negative, recent infection, or vaccination? 3. Figure 17: Two peaks merged.

Indirect

ELISA

to Determine

the Positivity

of Sera

155

Meab OD

Interval Fig. 14. Distribution skewedto the right. There is no clear distinction between populations (the high OD and low ODs overlap a great deal). These curves also illustrate what the picture might be after sampling total populations containing positive and negative animals. Thus, for the example in Fig. 15, there is no problem in ascribing an upper limit for negativity. Obviously the sera show the type of result expected of a totally negative population. Although in Fig. 16 we have a percentage of high OD results, these probably represent positive animals, and we can use the clear difference in the two distributions obtained to suggest strongly that the low OD results represent a negative population. The distribution in Fig. 17 demonstrates a situation where we have a heterogeneous population of animals with respect to their levels of antibodies. Thus, it is probable that the low OD range (which is shown by the situations in Figs. 15 and 16) represents negative animals. The merging of high and low results with high numbers of animals probably indicates that antibody levels have

156

Indirect

ELISA

25

.02

.04

.06

.08

.lO

.I2

.14

.16

.18

.20

.22

.24

.26

.28

Interval Fig. 15. Frequencyplot of OD resultsfrom analysisof sera.Negative population described. been reduced in the population after a past infection. Such a population can be studied using a defined negative population (maybe from another source), but the negative distribution cannot be assessedfrom the study of this type of distribution alone. Thus, the experimentor may obtain serum samples from relevant species from countries where the disease being studied is absent. The negative value(s) obtained from such sera may not always be the same as that of the indigenous stock, but for most exercises will suffice. 6.45. Establishment

of Control-Negative

Sera

It is possible to use a limited number of negative sera to act as controls in any assay of antibodies.This can only be realistically done if a distribution of many negative serum OD levels has been made (approx 100

Indirect

ELISA

to Determine

the Positivity

.04

.24

of Sera

25

.08

.12

.16 .20

.28

.32

.36 .40

.44

.48

.52

.56

Interval Fig. 16. Frequencyplot of OD resultsfrom analysis of sera.Two peaksrepresentingdistinct populations. minimum). Thus, a serum typifying the mean of the population of negative sera can be used. If this is included as a single dilution in the indirect ELBA, the OD value obtained will represent the mean value for the negative serum population. The upper limit of negativity can then be calculated by multiplying this value by 2 (since we know that this is a relevant value after studying the distribution). This approach is relevant where multichannel spectrophotometers are being used to read the color. If by-eye assessment is being used, then control-negative sera giving OD levels at the upper limit of negativity (around 2x mean) might be used. Color development in such assays should then be allowed until color is just detectable in the negative controls. The test should then be stopped. Therefore, any wells showing color more intense than the control wells will be positive for antibody.

158

Indirect

.04

.08

.12

.16

.20

.24

.28

.32

.36

.40 .44

.48

52

ELISA

.%i

Interval Fig. 17. Frequency plot of OD results from analysis of sera. No clear distinction between populations.

6.4.6. Taking Your Data You have calculated the mean OD of the negative population. You have calculated the standard deviation from the mean of the population. Find a serum that characterizes the mean of the population. Find a serum that characterizes the upper limit (2x mean) of the population. 6.5. Relating Single Test Dilutions to Standard Positive Antiserum Curves

If a characterized antiserum is available, then it may be used as a standard in the indirect ELISA. In this case, a full dilution range of the serum is made and titrated under identical conditions to the single dilutions of the test sera. A typical plate format is shown in Fig. 18. At the end of the test, a standard curve relating the OD to the dilution of standard positive

Indirect

ELISA

to Determine

the Positivity

of Sera

159

Duplicates of test sera 000000000000 000000000000 000000000000 000000000000 000000000000 000000 000000 000000000000

000000 000000

0 Standard posltlve serum Fig. 18. Plate layout for comparison of test sera with standard serum titration.

serum is constructed. The titers of the test samples can then be read from this curve so that a relative assessment of activity is obtained. This is demonstrated in Fig. 19. The standard serum may be given an arbitrary activity (units), so that results may be expressed in those units. Such control-positive sera may be useful where standardization between laboratories is required. of Actual Disease Studies on many systems have shown that false-positive results are obtained in a low percentage of animals from a guaranteed noninfected population. It is difficult to determine why such reactions occur, but several reasons have been proposed, such as contamination of the serum with bacteria and fungi, dietary factors, and heating of the sera. The percentage can be on the order of l-2%, and these samples are easily read as very high ODs as compared to the majority of samples giving the typical negative distributions already discussed. This nonspecificity may be 6.6. Complications

Indirect

160

ELISA

Serum B ................ ...

0.5 -

I

0’ 1

2

I 8

3

relatwe

to standard

Dilution of standard antibody+ Fig, 19. Use of standard serum titration curve to assesstiters of test sera. OD values obtained from serum A and B are read from the titration curve of the standard serum. eliminated, e.g., by using different antigenic preparations. However, the number of likely false-positive results can be taken into account when diagnosing disease on a herd basis. Thus, if we know that 2 animals in 100 show this response, and we find that 20 animals out of 100 show high responses, it is likely that disease is diagnosed. However, if we find only one to three animals “positive, ” this could be because of the identified nonspecific reactions.

PART 11

Practical

Exercises

The aim of Chapters 5-8 will be to illustrate the principles of ELISA fully by: 1. Showing worked examplesof eachassay,including diagramsof platesand

representationaldata from assays; 2. Analyzing such data in terms of important rules that are learnedat each stage;and 3. Providing full working instructronsfor workers to be ableto perform each assayso that they obtain their own data to be analyzedas describedin (1) and (2). This includes full instructionson the preparationand standardization of reagents.

The chapters can therefore be used in several ways. Workers without accessto reagentswill obtain a working knowledge of ELISA through the examples. The chapterscan also be used in training courses where reagents may be provided (as indicated in the text). The information will also be useful to workers who have already had some experience of the technique and who may have had difficulties in obtaining and analyzing data. Remember that it is the application of the ELISA to specific problems, and not the methodology for its own sake, that is the most important reason the techniques should be mastered. 1. Test Schemes

You may be already familiar with the concepts in ELISA, whereby an antigen binds to an antibody that can be labeled with an enzyme, or be in turn detected with a species-specific antibody (enzyme labeled). All the ELISAs are variations on this theme. Inherent in the methods of ELISA described in these chapters is the fact that one of the reagents is attached to a solid-phase, making the separation of bound (reacted) and unbound (nonreacted) reagents simple by washing. Before performing ELISA on disease agents, it is useful to train using reagents of defined reactivity, which are easily available and which provide security problems. An ideal

115

116

Practical

Exercises

system is to use an imrnunoglobulin (Ig) and more particularly an immunoglobulin G (IgG) as an antigen. Do not get confused here, since you have learned that the antibody population contains high levels of IgG acting as antibody. In the context of learning the principles, we are using IgG as an antigenic protein, since: 1. IgG from one animal species can be injected into another animal species so that a specific antiserum to that IgG is prepared. 2. Such antibodies can be labeled with enzyme, or detected with a second species-specific antibody labeled with enzyme. Such reagents are defined, easy to standardize, stable, and available commercially. The particular IgG system chosen in most of the chapters involves guinea pig, but similar tests can be performed with other species IgG using the appropriate antispecies reagents. The practical elements of all the assays are very similar, i.e., reagents and equipment needed. The systems described are analogous to the ones most commonly

used to examine problems associated with diagnosis. The schemes will be described using symbols where: I- = solid phase microtiter plate well Ag = antigen Agl, Ag2, etc. = particular antigens highlighted in assay I-Ag = antigen passively adsorbed onto wells I-Ab, I-AB = particular antibodies passively coated onto wells Ab = antibody AB = antibody from a different species to Ab Abx, Aby = different antibodies identified by subscript letters Anti-Ab = antispecies specific antibody (against species in which Ab was produced) Anti-Ab*E = antispecies specific antibody labeled with enzyme W = washing step, involving separation of bound and free reagents + = addition of reagents and incubation step S = substratekhromophore addition Read = read test in spectrophotometer at 492 nm Throughout

Chapters 5-8, many of the practical stages are the same.

The conjugates described are all made with horseradish peroxidase and the substratekhromophore is hydrogen peroxide/orthophenylamine diamine (OPD). The preparation and use of this is described in detail below. 1. Substratekhromophore: This is easiest made up from commercial tablets of OPD that are preweighed. Commercial sources also supply citrate/phos-

Test Schemes

2. 3.

4. 5.

117

phate buffer tablets (pH 5.0). Thus, the volume of OPD can be made as required by following the recommendations by the supplier. As an example, 30 mg tablets are available that make 75 mL of chromophore solution in buffer. Unused OPD solution (without added hydrogen peroxide) can be frozen at -20°C. This can then be thawed and used later. Close inspection should be made to ensure that the OPD is not drscolored. Use complete chromophore/substrate as soon as possible. Larger volumes of OPD in citrate/phosphate buffer can be made and frozen in a tightly stoppered brown bottle in small volumes. The OPD solution should be made and frozen as quickly as possible. Do not use solutions that show discoloration after freezing. Hydrogen peroxide (HzO,) is the substrate for horseradish peroxidase enzyme. This is purchased usually as 30 or 6% w/v and should be stored as recommended by the supplier. The hydrogen peroxide should be kept refrigerated and not subjected to heating. The addition of the hydrogen peroxide should be made immediately before the use of the OPD in the test. Add 5 PL of hydrogen peroxide (30% w/v) to every 10 mL of OPD solution (pH 5.0), or 25 pL of 6% hydrogen peroxide to every 10 mL of OPD solution. Use the substrate/chromophore immediately. OPD is a mutagen, so care is needed in its handling and disposal. Washing solution used in washing steps: This is PBS without the addition of Tween 20. Washing requires the flooding and emptying of wells 4 times with PBS. Blocking buffer: This is PBS containing a final concentration of 1% bovine serum albumin (BSA) and 0.05% Tween 20. This should be made in small volumes as required, but can be stored at 4°C. Care should be taken to avoid contaminated buffer. Stopping solution: This 1M sulfuric acid in water. Care should be taken in its preparation and handling. Read: This implies reading plates using a multichannel spectrophotometer at the appropriate wavelength for the color developing in the ELISA. In all cases for Chapters 5-8, this is 492 nm for OPD. Plates should also be assessedby eye to ascertain whether the test results are as expected.

CHAPTER7

Use of Antibodies on Solid-Phase in Capture 1. Use of Capture

ELISA

to Detect

ELISA

and Titrate

Antigen

In this exercise, the capture antibody, the detecting antibody, and the conjugate are used at optimal dilutions.

1.1. Learning

Principles

1, To optimize amounts of capture antibody attached to wells; and 2. To optimize amount of detecting antibody. 2. Reaction

Scheme

I-AB

+ Ag + Ab + Anti-Ab*E + S + Read W w w W I- = Microplate AB x = Capture antibody (species X) specific for Ag Ag = Antigen Ab v = Detecting antibody (species Y) specific for Ag Anti-Abv*E = Antispecies-Y antibody linked to enzyme S = Substrate/color detection system W = Wash + = Addition and incubation of reactants Read = Read QD at 492 nm using spectrophotometer

3. Basis of Assay Antigens may: 1. Attach poorly to plastics. 2. Be present in low quantity, e.g., in tissure-culture fluids. 3. Be present as a low percentage of total protein in a “dirty” sample, e.g., in feces or in epithelium samples. 4. Be unavailable for purification and concentration, since they are antigenitally unstable when separated from other serum components. 161

162

Use ofAntibodies

on Solid Phase in Capture

ELBA

In these cases, the indirect assayis unsuitable for handling the antigen, since it relies on the attachment of the antigen directly to the wells. The capture assay overcomes many of these problems, since the antigen is attached to the wells via specific antibodies. The test relies on the availability of two antisera from different species, so that the conjugate reacting with the second (detecting) antibody does not react with the capture antibody. It is also essential that the antigen has at least two antigenic sites, so that antibody may bind to allow the sandwich (the antigen being the filling). Thus, where small antigens are being used (e.g., peptides), they may not react in such assaysowing to their limited antigenic targets. The test offers an advantage over the indirect assay in the quantification of antigens, since direct attachment of proteins to wells is often nonlinear, i.e., is not proportional to the amount of protein in the sample. This is exaggerated if contaminating proteins are present with the antigen (e.g., serum components), since thesecompete for plastic sites in a nonlinear way. Since the capture antibody is specific, it binds antigen in a proportional way over a large range of protein concentrations. Thus, such assaysgive reproducible results where quantification is required (explained later more fully). The assay is really identical to the indirect assay, except that an extra step (the capture antibody) is added.Thus, we havethree parametersto optimize. 1. The captureantibody. 2. The detecting antibody. 3. The conjugateagainstthe detecting antibody. 4. Materials 1. Capture antibody (AB,) = sheepantiguinea pig Ig (an IgG preparation [AB] 5 mg/rnL in PBS). 2. Antigen (Ag) = guineapig Ig at 1 mg/mL or as preparedby worker (Ag). 3. Detection antibody (Ab,) = rabbit antiguineapig Ig serum (Ab). 4. Anti-AbY*E sheepantirabbit conjugate(HRPO). 5. Microplates. 6. Multichannel pipets, single-channelpipets, 10 and 1-mL pipets. 7. 0.05M Carbonate/bicarbonatebuffer, pH 9.6. 8. PBS containing 1%BSA, 0.05% Tween 20. 9. Solution of OPD in citrate buffer. 10. Hydrogen peroxide 30%, v/v. 11. Washing solution. 12. 1M sulfuric acid in water. 13. Papertowels.

Methods

14. 15. 16. 17.

163

Small-volume bottles. Multichannel spectrophotometer. Clock. Graph paper 6. Methods 5.1. General Notes Since you are now familiar with the indirect assay, the steps in the optimization of the capture ELISA should be straightforward. The first essential is to determine the amount of capture antibody to be attached to the wells. We have two situations in the laboratory depending on the availability of specific reagents. We can use capture antibody as an IgG preparation or, if sufficiently high titer serum is available against the antigen, as whole serum. The easiest way to avoid serum effects is to prepare the IgG. Salt fractionation is usually adequateand does not affect antibody activity. Care must be taken to assessthe effect of chemical preparation of IgG from monoclonal antibodies (MAbs). 5.2. Use of Ig Preparations The advantage here is that the weight of IgG can be calculated, so that a defined quantity of reagent may be added to the plate. In general, a maximum amount of protein will attach to the wells, so that the IgG at “saturating” level may be added in the knowledge that a maximum possible binding of subsequently added antigen may be expected. Thus, a good estimation of the activity of a capture antibody (the particular dilution/concentration to be used) can be assessed.As an example, if capture antibody is added at 5 pg/mL in 5O-l.rLamounts, this represents the saturating amount of antibody protein that will attach to the wells. The ultimate activity will depend on the concentration of the specific IgG (against the Ag) in the capture antibody and the spacing of the capture molecules. Some assays perform better at lower than saturating levels of capture antibody, so that a titration is needed. Generally, the amount of specific antibodies in a serum as a percentage of the total protein is around l-5%. The preparation of IgG eliminates a large percentage of the serum proteins not involved in the assay (e.g., serum albumins). Therefore, the activity of the IgG protein (relative increase in the IgG fraction that will attach to each well) is increased effectively. In other words, there is a greater proportion of IgG sticking to the wells to act as trapping antibody if IgG preparations are used.

164

Use of Antibodies

0'

I 1234

I

I

on Solid Phase in Capture ELISA

I / 567

I

I 8

9

\ 10

I I 11 12

Serum or IgG dilution (2 fold) I

IgG

-+4-

Serum

Fig. 1. Comparison of capureof IgG using whole serum or IgG as capture antibody. 5.2. Use of Whole Serum

Dilutions of untreated serum can be used. However, as indicated above, the proportion of specific IgG is low, and other serum proteins attach in a competitive manner. One cannot assume that putting on a low dilution of serum will give a good level of capture antibody. The most usual event is that a bell-shaped curve of capture ability is obtained, with little activity at high concentrations of serum and a rise in activity as the serum is diluted. In general, serum has to be diluted to around l/5001/2000. Thus, we have to have fairly high titers to be able to use whole serum. Figure 1 demonstrates the activity of whole and IgG capture antibodies as they are diluted down to illustrate the bell-shaped curve. 6. Titration of Capture Antibody Using IgG 1. Dilute the sheepantiguineapig IgG preparationto 5 ug/mL in carbonate buffer. Add 50 uL to eachwell on the plate except column 1. 2. Add adsorptionbuffer alone to row 12. Incubate at 37°C for 2 h or overnight if more convenient(rememberto put lids on plates,and so forth).

Titration

of Capture

Antibody

Using IgG

165

3. Wash the plates. From now on, we are performing a similar procedure to the indirect ELISA demonstrated earlier. 4. Take microtiter plate with well Al at the top left-hand corner. Add 50 PL of blocking buffer to each well. 5. Make a dilution range of guinea pig IgG (the antigen of mterest) from 5 pg/mL from column 1(8 wells) to column 11 in blocking buffer. Thus, add 50 ltL of guinea pig IgG at 10 pg/mL to first row 1 using a multichannel pipet. Mix, and double dilute across the plate (you should be competent at this now). Remember to discard the last 50 pL in the tips, so that each well should only contain 50 l.tL of fluid (check!). 6. Incubate the plates at room temperature or at 37°C for 1 h. 7. Wash the plates. 8. Add 50 pL of blocking buffer to each well. 9. Take rabbit antiguinea pig serum, and dilute to l/100 in blocking buffer (make up 1.OmL, add 10 l,tL of undiluted serum to 1.OmL of buffer). Mix. Add 50 pL of the dilution to row A using a single-channel pipet. Dilute across rows A-H using a multichannel pipet. We now have a twofold dilution range from l/200 (row A) to l/25,600 (row H). 10. Incubate the plate at room temperature or at 37OCfor 1 h. 11. Wash the plate. 12. Make up optimum dilution of sheep antirabbit conjugate. This can be pretitrated as in Chapter 5. Add 50 l.tL to each well using multichannel pipets. 13. Incubate for standard time as used in optimization of conjugate (1 h at 37OCor room temperature). 14. Wash the plate. 15. Add substrate and stop color at 10 min.

6.1. Data Essentially we have made a chessboard titration of the antigen against the detecting antibody (as in the indirect assay). Thus, we have assumed that the capture antibody, put on the plate as an IgG, is at maximal reactivity. The results are therefore similar to those obtained in the indirect assay and can be treated in a similar way. Each of rows A-H had an identical dilution series of the antigen (guinea pig IgG) being captured by the same amount of antibody. Thus the same amount of guinea pig IgG should be present and attached via antibody to wells l-l 1. The rabbit antibody against the antigen has been titrated at different dilutions, so we can examine which dilution shows the best detection of IgG in rows A-H. The use of 5 pg/mL, of capture IgG was taken as that which from experience saturates the plastic sites available on the plate wells. Once the antigen

166

Use of Antibodies

on Solid Phase in Capture

ELISA

Table 1 Chessboard Titration of Guinea Pig IgG vs Rabbit Antiguinea Pig IgG-Constant Capture Antibody, Constant Conjugate

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

1.67 1.68 1.56 1.12 1.00 0.78 0.54 0.34

1.67 1.68 1.54 1.09 0.97 0.74 0.51 0.34

1.68 1.65 1.52 1.00 0.89 0.71 0.51 0.32

1.65 1.56 1.43 0.94 0.78 0.56 0.42 0.21

1.54 1.51 1.34 0.87 0.67 0.51 0.36 0.18

1.34 1.31 1.23 0.78 0.56 0.43 0.32 0.15

1.09 1.04 0.99 0.67 0.43 0.32 0.21 0.16

0.89 0.84 0.76 0.56 0.34 0.21 0.16 0.09

0.67 0.59 0.52 0.45 0.23 0.19 0.14 0.08

0.54 0.51 0.43 0.34 0.21 0.14 0.09 0.07

0.34 0.32 0.23 0.19 0.17 0.09 0.08 0.08

0.23 0.17 0.09 0.08 0.08 0.10 0.09 0.09

Fig. 2. Diagrammatic

representation

of plate, data from Table 1.

and detecting serum optima have beenestablishedusing this level of capture IgG (as shown below), this can be altered to examine the effect on the assay. Table 1 shows the spectrophotometric plate readings. A representation of the plate is also shown in Fig. 2. 6.2. Plots

of Data

Figure 3 shows the data plotting the OD results obtained at the different Ag dilutions for each dilution of rabbit antiguinea pig serum.

Titration

of Capture Antibody

I

I

Using IgG

Y

I

10

0123456789

Antigen

Rows

167

Dilution

11 12

+

-A+B*C+DxE*FfG*H

Fig. 3, Titration of guinea pig IgG using constant capture conditions. Each line represents titration of the same dilution range of IgG using a different concentration of rabbit antiguinea pig IgG. The conjugate dilution is constant. From data in Table 1,

of Data Column 12 contained no antigen (guinea pig IgG). Therefore, examination of the color here gives a measure of the binding of the detection system to the plate or capture antibody. Thus, rows A and B show higher levels of color than the rest. The value around 0.09 appears to be the plate background expected in the presence of the same dilution of conju6.3. Assessment

gate. Thus, the end point detection of IgG is affected in rows A and B. Examination of the plateau heights indicates that the trapping system is

saturated in columns l-4, since we obtain similar OD values. Thus as an example, we have around 1.67 for the first four wells using the l/200 detecting antibody. Although the actual plateau height value reduces on dilution of the detecting rabbit antiguinea pig serum, examination of Fig.

3, which relates the curves obtained for the detection of trapped Ig for different dilutions of the rabbit antiguinea pig Ig, shows most easily that the last dilution giving an optimal titration is in row C, After this dilution, the effect is to reduce more markedly the OD in the plateau region

168

Use of Antibodies

on Solid Phase in Capture

ELISA

(where the trapped Ig is in excess) and also affect the sensitivity of detection of the Ig at higher dilutions, as indicated by a reduction in the end points where the test background is the same as the plate background. of Capture IgG The optimal dilutions chosen will depend on how the test is to be used. If an antigen is to be detected, then we might require high detection limits in the system, so that we can use a dilution of detecting antiserum to maximize this, We will see later that capture assayswill be used in competitive situations where the amount of antigen to be captured needs to be reduced, so that a variation in reagent concentrations may be necessary and can be read from an exercise as described here. The established optima for the antigen and detecting serum can be reassessedusing lower concentratons of capture IgG. Thus, a full chessboard titration as described above can be performed using 2.5, 1.25, and 0.625 pg/mL of the capture IgG. However, a simpler procedure is to coat plates with a dilution range of the capture IgG, and use constant antigen, detecting antiserum, and conjugate dilutions as found above. Results of a typical titration of this sort are demonstrated in Table 2. Here plates have been coated with capture anti-IgG at 5 pg/mL in a twofold range from rows A-H, columns l-4 only, thus quadruplicate samples are being examined. After incubation and washing, antigen (guinea pig IgG) at 0.625 pg/mL has been added in blocking buffer. Following incubation and washing, the detecting antibody (rabbit antiguinea pig IgG) has been added at l/400 diluted in blocking buffer. After incubation and washing, the antirabbit conjugate has been added at the dilution used to optimize the reagents. Table 2 shows that the capture IgG produces similar results at 5 and 2.5 pg/mL, thus the latter dilution can be used in an assay to capture antigen. Lower concentrations produce lower OD values, indicating that not all the available antigen is being captured. The reduction in ability to bind antigen (when in excess) is accompanied by a loss in ability to detect small amounts of antigen (minimum detection limit is reduced). Similar titrations of the other reagents can be made where only one is diluted and the others are kept constant. Thus, in the case above, we know we have three conditions optimized under experimental conditions with control sera and antigen. The capture IgG can be used at 2.5 pg/mL, the antigen can be used at 0.625 yg/mL, and the rabbit detector at l/400, 6.4. Retitration

Titration

of Capture Antibody

169

Table 2 Titration of Capture IgG Against Optimized Antigen, Detecting Antibody and Conjugate Capture IgG concentration, WmL 5.0 2.5 1.25 0.63 0.32 0.16 0.08 0.08

A B C D E F G H

1

2

3

4

Mean

1.50 1.49 1.25 0.95 0.67 0.36 0.14 0.05

1.48 1.47 1.21 0.94 0.69 0.37 0.12 0.04

1.49 1.51 1.24 0.96 0.69 0.40 0.15 0.04

1.51 1.46 1.27 0.93 0.66 0.37 0.12 0.03

1.50 1.48 1.24 0.95 0.68 0.38 0.13 0.04

with the antirabbit conjugate at a constant dilution as assessedoriginally against the relevant IgG attached to a microplate. We may wish to reassess the conjugate dilution under standardized conditions. Thus, using the capture IgG, antigen, and detecting antiserum optima found above, replicate wells can be used to titrate different dilutions of conjugate. An example is shown in Table 3. Here a dilution of 2X)0 of conjugate gives similar results. Effectively, a dilution of l/800 gives “optimal” results(OD value around 1.45)for an assay. 7. Titration of Capture Antibody when Used as Whole Serum As already stated, whole serum can be used to coat plates and act as a capture reagent. This is not recommended, since we cannot measure the protein Ig because is contaminated with “blocking” serum proteins that compete for plastic binding sites preferentially over the IgG. The simplest method is to perform a chessboard assay relating dilutions of capture serum to dilutions of detecting antibody and keep the antigen constant. The diagram below illustrates this: I-Ab

+ w

Ag

+ w

AB

+

Anti-AB*E

w

I-Ab = Dilution range of trapping antibody Ag = Constant dilution of antigen (high concentration) AB = Dilution range of detecting antibody

+ W

S + Read

170

Use of Antibodies

on Solid Phase in Capture

ELISA

Table 3 Assessment of Constant Capture System with Different Dilutions of Conlugate Conjugate dilution 200 400 800 1600 3200 6400 12,800 none

A B C D E F G H

1

2

3

4

Mean

1.95 1.84 1.45 0.95 0.77 0.36 0.15 005

1.87 1 82 1.41 0.94 0.79 0.37 0 14 0.04

1 87 1.84 1.44 0.96 0.79 0.40 0.15 0 04

1.95 1 82 1.47 0.93 0.76 0.37 0.14 0.03

192 1 83 1.44 0.95 0 78 038 0.15 0 04

Anti-AB*E = Conjugated antispecies antibody S = Substrate/chromophore Read = Read plate in spectrophotometer

This assay will not be described in detail. However, a description of the test will be given with relevant points highlighted. You should now have enough experience to be able to set up the exact practical details yourself with help from the exercise titrating IgG as capture antibody. 1. 2.

3. 4. 5.

7.1. Method The serum containing capture antibody is diluted on plates in carbonate/ bicarbonate buffer (begin at l/100, twofold dilutions). Incubate and then wash plates. Constant (excess) antigen is then added diluted in blocking buffer (difficult to specify here what excess might be for specific systems (e.g., an undiluted tissue culture sample containing vuus might be expected to have a high concentration of antigen). Incubate for 1h using standard conditions. Wash and add dilutions of detecting antibodies in blocking buffer to obtam chessboard titration (dilute in the opposite direction to the capture serum). Incubate and then wash plates. Add optimal conjugate, incubate, and then wash. Add substrate and then stop at 10 mm. 7.2. Data

Typical results are shown in Table 4. Rows A-H contain dilution ranges of the capture serum l/100-l/51,200, i.e., column 1 = l/100, column 12 = l/51,200. Row A received the detecting antiserum at l/200,

Use of Capture ELBA

to Detect Antigens

171

Table 4 Dilutions of Capture SerumMOO-l/51,200

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

0.67 0.68 0.65 0.56 0.45 0.23 0.15 0.15

0.96 0.99 0.98 0.88 0.67 0.43 0.23 0.19

1.34 1.42 1.36 1.23 1.00 0.78 0.34 0.18

1.35 1.37 1.34 1.19 1.09 0.76 0.35 0 17

1 32 1.29 1.15 1.01 0.98 0.56 0.21 0.16

1.11 1.09 0.99 0.88 0 78 0.45 0.15 0.10

0.98 0.89 0.87 0.74 0.56 0.40 0.16 0.09

0.76 0.75 0.72 0.65 0.45 0 33 0.09 0.08

0.56 0.54 0.52 0.43 0.34 0 23 0.08 0.09

0.45 0.36 0.31 0.26 0.21 0 16 0.07 0.07

0.23 0.22 0.17 0.14 0.12 0.12 0.09 0.07

0.12 0.11 0.09 0.09 0.07 0.08 0.09 0.09

row B l/400, and so on, to row H at l/12,800. See Fig. 4 for further clarification. 7.3. Conclusions 1. Optimal

dilution

of capture serum is around column

5 (last column

show-

ing maximum OD). 2. Optimal dilution of detecting second antibody is around row C/D (last showing

maximal

titration

curve of antigen).

3. Bell-shapedcurvesareobtainedwherelow dilutions of captureserumgive low OD values (columns 1 and 2). 8. Use of Capture ELBA to Detect Antigens Once the optimal conditions have been established, the capture assay can be used in several ways as indicated below. 8.1. Diagnosis I-Ab

+ w

Ag

+ w

of Specific AB

+ W

Disease Agents Anti-Ab*E + S + Read W

Here a sample possibly containing antigen is added to a capture system (microtiter plate wells coated with an antiserum against a specific disease). Any bound antigen is then detected by another antibody from a different species. Such assays are important in serotyping where the second antibody may further “divide” the disease agent into a serological grouping, e.g., as is used routinely to serotype foot-and-mouth disease viruses (FMDV) into one of seven distinct serotypes. The use of capture antibody means that relatively crude or contaminated samples can be used.

Use of Antibodies

on Solid Phase in Capture ELBA

16

08 06

Dilution of capture ambody (l/100,2 fold) Rows

-+A

Dhtlon

l/100

+B l/200

*C l/400

*D 11800

*E

*F

l/1600

l/3200

*G l/6400

*H l/12800

Fig. 4. Graph of data in Table 4. Columns 1-12 contain dilutions of capture antibody on wells. Rows A-H have different dilutions of detecting antibody. Quantification of antigens may be made with reference to a standard antigen titration on the same plate. Single dilutions of material containing the Ag can then he titrated in the same system and the developing OD read against the standard titration. Again, the use of the capture antibody ensures an efficient and proportional uptake of the antigen onto the plate, which is unaffected by contaminating proteins.

9. Use of Capture to Detect and Titrate

ELISA Antibodies

9.1. Learning Principles 1. Optimization of capture antibodies; and 2. Optimization of detecting antibody. 9.2. Reaction Scheme + Abv + Anti-Abv*E + I-ABx + Ag w w w w I- = Microplate AB, = Trapping antibody (species X) Ag = Antigen Abv = Test or control sera (species Y)

S + Read

Capture ELBA

to Detect and Titrate Antibodies

173

Anti-Abv*E = Antispecies Y antibody conjugated with enzyme S = Substrate/color detection system W = Wash + = Addition and incubation of reagents Read = Measure OD in spectrophotometer

of Method Essentially the same parameters have to be standardized as for capture ELISA for antigen detection, However, the test is used to measure antibodies against a fixed amount of antigen captured on the plate. Thus, we have to optimize the system to have the correct amount of capture anti9.3. Principle

body and antigen necessary to bind any test or control antisera. The test

offers the ability to capture antigen specifically using a solid-phase antibody. Thus, relatively crude preparations can be used where the required antigen concentration may be low. Care has to be taken to avoid reactions of the conjugate with components of the assay. 1. 2. 3. 4. 5. 6. 7.

9.4. Materials And Methods Capture antibody (ABx) = sheep antiguinea pig Ig at 5 mg/mL in PBS. Antigen (Ag) = guinea pig Ig at 1 mg/mL. Test antisera (Ab,) = 3 antirabbit antiguinea pig Ig (Ab). Also sero-negative rabbit sera (as used in Chapter 6, Section 5.5.). Anti-Abv*E = sheep antirabbit IgG conjugated to horseradish peroxidase. Microplates. Multichannel, single-channel, lo- and 1-mL pipets. 0.05M Carbonate/bicarbonate buffer, pH 9.6.

8. PBS containing 1% BSA, 0.05% Tween 20. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Solution of OPD in citrate buffer. Hydrogen peroxide. Washing solution. Paper towels. 1M sulfuric acid in water. Small-volume bottles. Multichannel spectrophotometer. Clock. Graph paper. 9.5. Optimization

We need to know: 1. What dilution of captureantibody to use. 2. What dilution of antigen to use.

of Test

174

Use of Antibodies

on Solid Phase in Capture

ELBA

3. What dilution of conjugate to use.

The aim is to have a constant system involving capture antibody (AB), antigen (Ag), and conjugate (Anti-Ab*E), which can then be usedto titrate test sera (Ab). The use of capture antibody as whole serum or as IgG has been dealt with. This exercise will deal with the use of sheep antiguinea pig IgG (or the equivalent in an individual’s system). Thus, examination of the data in Table 1 allows an estimation of the optimum capture IgG and antigen levels required to allow detection of antibodies. 9.5.1. Data

Using the titrations established in Section 6.1.) we can obtain the optimal amount of antigen (guinea pig Ig in this case) that gives a high pla-

teau OD where the detecting antiserum is in excess. Turn to the data showing the plate readings for Section 6.1. (Table 1). We can see the plateau height is maintained to around column 4, showing that there is a maximum level of antigen to react with the antibodies in the positive serum. This concentration (or dilution) can be used in the capture assay under the same conditions to titrate antibodies from any sera. From the data obtained in Section 6.1.) we can use: 1. Tbe capture antibody at 2.5 l.tg/rnL if used as an Ig preparation or at the titrated level as found in Table 2. 2. The antigen at the concentration or dilution used in columns 4 and 5 (Table 1). 3. The conjugate as titrated initially in Table 3.

9.6. Methods for Titration of Antibodies As described in Chapter 6, we can examine serafor antibodies by using full dilution ranges or as single dilutions. The methodology in the capture ELISA is the same, except that after initial optimization of capture and antigen-coating conditions, capture plates are set up, coated with antigen, and then used to assess antibodies. Thus, optimization of the capture antibody and antigen concentration (as shown above), allows the possibility of determining specific antibodies from test and control sera. Perform the capture assay based on the addition of the same rabbit antiguinea pig sera, as from stage 4, in the direct assay described in Chap-

ter 6, Section 5.5. Add the various dilutions of positive and negative rabbit antiguinea pig sera to plates containing an optimal concentration of captured guinea pig IgG. After incubation and washing of plates, add the antirabbit conjugate (optimal dilution as measured in Table 3), incubate,

Capture

ELLSA to Detect and Titrate

Antibodies

175

wash, and develop color. Plot the data from the spectrophotometer and compare these to the data obtained in Chapter 6, Section 5.5. (Table 2). Repeat the exercise in Chapter 6, Section 6.6., using the capture ELISA, capturing IgG under optimal conditions to determine the positivity of the samerabbit seraat a single dilution. Read and plot the data. Compare these to the indirect assay results and assessthe sera as positive or negative using the statistical criteria outlined before. Note any differences. 9.6. Problems Using Capture Assays 1. Care must be taken to examine whether any of the reagents interact. Unexpected crossreactions can be found with immunological reagents, e.g., the conjugated antibodies might react with species other than those for which they were prepared. There are crossreactionsbetween certain species,so that conjugates against cow proteins will react with sheepand goat proteins. Thus, a system using sheep or goat Ig as a capture antibody will preclude the use of antibovine conjugates to detect the reaction of bovine antibodies with a particular antigen. 2. Where relatively crude antigens are captured, contaminating proteins may also be trapped that interfere with the assay.As an example of difficulties, when purified FMDV is injected into an animal, there is a specific response against the virus, but also a response against contammating bovine serum proteins that are present in extremely low amounts, coming from the tissue-culture medium. Such sera used as capture reagent will capture not only virus, but also bovine proteins. Thus, in typing exercises using tissueculture or bovine epithelial samples, a high quantity of bovine protein is captured. The use of antibovine conjugates to detect bound bovine serum in a trapping assay,therefore, also binds to the trapped bovine protein giving high backgrounds.

In the typing assay proper, guinea pig sera are prepared as the second typing detecting sera.These also bind bovine proteins and therefore, detect bound bovine protein to the capture antiserum. Again, specific typing is affected. However, the second antibody can be treated to remove the crossreactivity either by adding a high concentration of the crossreactiveprotein to the reagent (in this case, 1 mL of normal nonimmune bovine serum is added to 1 mL of typing guinea pig serum), or by using affinity reagents where bovine serum is attached to a solid-phase, e.g., agarose beads, which can be incubated with the serum, so that the crossreactive antibodies are removed after incubation and separation of the beads by centrifugation, or as is most common, the test may be made using blocking buffers containing high levels (around 5%) of the crossreactive protein.

PART 11

Practical

Exercises

The aim of Chapters 5-8 will be to illustrate the principles of ELISA fully by: 1. Showing worked examplesof eachassay,including diagramsof platesand

representationaldata from assays; 2. Analyzing such data in terms of important rules that are learnedat each stage;and 3. Providing full working instructronsfor workers to be ableto perform each assayso that they obtain their own data to be analyzedas describedin (1) and (2). This includes full instructionson the preparationand standardization of reagents.

The chapters can therefore be used in several ways. Workers without accessto reagentswill obtain a working knowledge of ELISA through the examples. The chapterscan also be used in training courses where reagents may be provided (as indicated in the text). The information will also be useful to workers who have already had some experience of the technique and who may have had difficulties in obtaining and analyzing data. Remember that it is the application of the ELISA to specific problems, and not the methodology for its own sake, that is the most important reason the techniques should be mastered. 1. Test Schemes

You may be already familiar with the concepts in ELISA, whereby an antigen binds to an antibody that can be labeled with an enzyme, or be in turn detected with a species-specific antibody (enzyme labeled). All the ELISAs are variations on this theme. Inherent in the methods of ELISA described in these chapters is the fact that one of the reagents is attached to a solid-phase, making the separation of bound (reacted) and unbound (nonreacted) reagents simple by washing. Before performing ELISA on disease agents, it is useful to train using reagents of defined reactivity, which are easily available and which provide security problems. An ideal

115

116

Practical

Exercises

system is to use an imrnunoglobulin (Ig) and more particularly an immunoglobulin G (IgG) as an antigen. Do not get confused here, since you have learned that the antibody population contains high levels of IgG acting as antibody. In the context of learning the principles, we are using IgG as an antigenic protein, since: 1. IgG from one animal species can be injected into another animal species so that a specific antiserum to that IgG is prepared. 2. Such antibodies can be labeled with enzyme, or detected with a second species-specific antibody labeled with enzyme. Such reagents are defined, easy to standardize, stable, and available commercially. The particular IgG system chosen in most of the chapters involves guinea pig, but similar tests can be performed with other species IgG using the appropriate antispecies reagents. The practical elements of all the assays are very similar, i.e., reagents and equipment needed. The systems described are analogous to the ones most commonly

used to examine problems associated with diagnosis. The schemes will be described using symbols where: I- = solid phase microtiter plate well Ag = antigen Agl, Ag2, etc. = particular antigens highlighted in assay I-Ag = antigen passively adsorbed onto wells I-Ab, I-AB = particular antibodies passively coated onto wells Ab = antibody AB = antibody from a different species to Ab Abx, Aby = different antibodies identified by subscript letters Anti-Ab = antispecies specific antibody (against species in which Ab was produced) Anti-Ab*E = antispecies specific antibody labeled with enzyme W = washing step, involving separation of bound and free reagents + = addition of reagents and incubation step S = substratekhromophore addition Read = read test in spectrophotometer at 492 nm Throughout

Chapters 5-8, many of the practical stages are the same.

The conjugates described are all made with horseradish peroxidase and the substratekhromophore is hydrogen peroxide/orthophenylamine diamine (OPD). The preparation and use of this is described in detail below. 1. Substratekhromophore: This is easiest made up from commercial tablets of OPD that are preweighed. Commercial sources also supply citrate/phos-

Test Schemes

2. 3.

4. 5.

117

phate buffer tablets (pH 5.0). Thus, the volume of OPD can be made as required by following the recommendations by the supplier. As an example, 30 mg tablets are available that make 75 mL of chromophore solution in buffer. Unused OPD solution (without added hydrogen peroxide) can be frozen at -20°C. This can then be thawed and used later. Close inspection should be made to ensure that the OPD is not drscolored. Use complete chromophore/substrate as soon as possible. Larger volumes of OPD in citrate/phosphate buffer can be made and frozen in a tightly stoppered brown bottle in small volumes. The OPD solution should be made and frozen as quickly as possible. Do not use solutions that show discoloration after freezing. Hydrogen peroxide (HzO,) is the substrate for horseradish peroxidase enzyme. This is purchased usually as 30 or 6% w/v and should be stored as recommended by the supplier. The hydrogen peroxide should be kept refrigerated and not subjected to heating. The addition of the hydrogen peroxide should be made immediately before the use of the OPD in the test. Add 5 PL of hydrogen peroxide (30% w/v) to every 10 mL of OPD solution (pH 5.0), or 25 pL of 6% hydrogen peroxide to every 10 mL of OPD solution. Use the substrate/chromophore immediately. OPD is a mutagen, so care is needed in its handling and disposal. Washing solution used in washing steps: This is PBS without the addition of Tween 20. Washing requires the flooding and emptying of wells 4 times with PBS. Blocking buffer: This is PBS containing a final concentration of 1% bovine serum albumin (BSA) and 0.05% Tween 20. This should be made in small volumes as required, but can be stored at 4°C. Care should be taken to avoid contaminated buffer. Stopping solution: This 1M sulfuric acid in water. Care should be taken in its preparation and handling. Read: This implies reading plates using a multichannel spectrophotometer at the appropriate wavelength for the color developing in the ELISA. In all cases for Chapters 5-8, this is 492 nm for OPD. Plates should also be assessedby eye to ascertain whether the test results are as expected.

CHAPTER8

Competitive

ELISA

1. General Information The direct, indirect, and capture ELISAs have now been examined. You should be able to optimize the conditions of the tests and be able to use them to measure antigen or antibody in a variety of formats. Competitive ELISAs involve the principles of all these types of assay. Basically they involve methods that measure the inhibition

of a reac-

tant for a pretitrated system. The degree of inhibition reflects the activity of the unknown. We can, therefore, measure antibody or antigen, and

even begin to subtly compare small differences in the binding of antigens or antibodies so that antigenic subtyping may be performed by comparing the relative avidity of one antiserum for two antigens in the same system. As a reminder, let us consider the competitive assays based on the indirect and the capture ELISAs for the detection of antigens or antibodies in a diagrammatic way. The symbols used are: I = Solid phase microtlter plate. I-Ag = Antigen attached to solid phase by passive adsorption. Ab = Antibody against Ag. AB = Antibody produced m a different species to Ab. Anti-Ab or anti-AB = Antispecies serum against particular Ab or AB, Anti-Ab*E or Anti-AB*E = Antispecies Ab or AB conjugated serum. W = Washing step. S = Substrate/chromophore. + = Addition of reagents and incubation.

1.1. Indirect I-Agl W

Assay-Antigen +AB + + Ag2 W

Detection

Anti-AB*E

by Competition

+ S +Read W

Here a pretitrated indirect assay with optimal Agl, AB, and conjugated anti-AB is competed for by Ag2, as a dilution range, in the liquid 177

Competitive

EiX3A

phase. If Ag2 can bind AB, then this prevents Al3 binding, which would normally react with Agl on the plate. The maximum expected OD for the pretitrated system without competitor is therefore reduced in the presence of the competitor Ag2. The degree of inhibition of the pretitrated reaction is proportional to the relative amount of the competitor. 1.2. Indirect

I-Ag

Assay-Antibody

+AB +Ab

+

W

Detection

AntI-AB*E

W

+

by Competition

S + Read

W

Here a pretitrated system is challenged by a dilution range of Ab. The competing antibody has to be from a species that is not the same as that of the AB in the optimized system. The degree of inhibition of the pretitrated system depends on the concentration and interaction of the Ab competitor with the Ag, this time on the solid-phase. The direct ELISA could also be used for both systems 1.1 and 1.2. Note that in the direct assayany speciesof competing antibody can be used since the AB is labeled with conjugate. Such assaysare becoming increasingly relevant where monoclonal antibodies (MAbs) are being used. 1.3. Capture

Assay-Antigen

I-AB-Agl

+Ab + Ag2 W

+

Detection

Anti-Ab*E

W

by Competition

+

S + Read

W

Here the capture assay is optimized to detect the Agl trapped on the plates using Ab. The competition is achieved where Ag2 is mixed with the Ab in the liquid phase. If this reacts, the amount of Ab available for reaction with the trapped Agl is reduced. 1.4. Capture

Assay-Antibody

I-AB

+

+Ag + Aby

Abx

Detection

+

Anti-Abx*E

by Competition

+

S + Read

W W W W Here the capture antibody is optimized to bind Ag, which is detected using a constant amount of Abx (from animal speciesX). The competition involves the reaction of the Ag with antiserafrom species Y, which should not interact with the conjugate Anti-Abx in the liquid-phase. Remaining Ag after the competition phase is then captured and titrated by the Abx

Direct

Competitive

ELISA

and the conjugate. Reduction in the expected color for the system without any AbY represents competition. Three assays will be dealt with practically. 1. Direct assay-antigen detection; 2. Indirect assay-antigen detection; and 3. Indirect assay-antibody detection: (a) full titration curves and (b) spot test assessmentof sera.

2. Direct Competitive ELISA for Antigen Detection and Quantification This has assumed an increased importance with the development of MAbs. A single MAb can be the one reagent that dominates a diagnostic assay and therefore is worth labeling for use in an assay. The specificity of the assay is ensured and relatively crude antigenic preparations can be coated for use in a direct test format (providing enough antigen attaches). This is also relevant to polyclonal antibodies. The demonstrated assays involve IgG/anti-IgG systems. 2.1. Learning

Principles

1. Optimization of homologous system. 2. Competition curves.

2.2. Reaction I-Agl

+ Ab*E + Ag2

+

Scheme

S + Read

w w I-Agl = Microplate with optimum concentration of antigen attached. Ag2 = Competing antigen as a dilution range. Ab*E = Optimum dilution of conjugated Ab specific for the Agl. S = Substrate/color detection system. + = Addition and incubation steps. W = Wash. Read = Spectrophotometric reading at 492 nm.

This exercise will most simply demonstrate the principles involved with competitive assays. 2.3. Materials and Reagents 1. Agl = guinea pig IgG at 1 mg/mL for attachment to solid phase. 2. Ag2 = two samples: (a) guinea pig IgG (known concentration) and (b) rabbit IgG at 1 mg/mL.

Competitive

180

ELISA

Table 1 Data From Chessboard Titration of Guinea Pig IgG and Anti-Guinea Pig Enzyme Conjugate in Exercise 2.4.

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

1.89 1.87 1.68 1.14 0.99 0.66 0.34 0.30

1.88 1.86 1.45 1.03 0.91 0.44 0.20 0.19

1.67 1.63 1.32 0.94 0.74 0.39 0.16 0.15

1.34 1.29 1.14 0.83 0.54 0.33 0.18 0.16

1.10 1.04 0.96 0.57 0.46 0.24 0.16 0.15

0.97 0.93 0.86 0.45 0.36 0.21 0.18 0.17

0.86 0.84 0.64 0.38 0.29 0.19 0.15 0.13

0.57 0.53 0.45 0.29 0.19 0.15 0.16 0.12

0.44 0.34 0.29 0.19 0.18 0.18 0.14 0.13

0.32 0.24 0.19 0.18 0.15 0.16 0.12 0.13

0.31 0.23 0.17 0.15 0.13 0.14 0.14 0.15

0.31 0.21 0.16 0.16 0.14 0.12 0 13 0.16

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Ab*E = rabbit antiguinea pig IgG conjugated to horseradish peroxidase. Microplates. Multichannel and single channel 10 mL and 1 rnL pipets. 0.05M carbonate/bicarbonate, pH 9.6. PBS containing 1% BSA, 0.05% Tween 20, Solution of OPD in citrate buffer. Hydrogen peroxide. Washing solution. Paper towels. Small-volume bottles. 1M sulfuric acid in water. Multichannel spectrophotometer. Clock. Graph paper. Calculator.

2.4. Practical Repeat exercise 5 in Chapter 5 involving the chessboard titration of antigen and enzyme-linked antibody. You should obtain a similar picture. Compare the results. The labeled conjugate dilutions are made from A-H, IgG is diluted l-l 1, and 12 has no antigen. Plot the chessboard titrations of guinea pig IgG against the conjugate as shown in Table 1 and Fig. 1.

2.4.1. Assessment of Data, Choice of Conditions for Competition We are trying to compete the antigen (guinea pig IgG) and a different antigen (IgG from the rabbit) for a pretitrated homologous solid phase reaction. The ultimate sensitivity of the assay depends on the exact rela-

Direct

Competitive

0

181

ELJSA

12

3

4

Dilution +1

+2

5

6

of antigen *3

*4

*5

7

8

9

10

1112

added to wells. *6

*7

*S

Fig. 1. Data from Table 1 relating antigen titrations at different concentrations of conjugate. tionship of the antibody and antigen attached to the solid-phase. If we use too much antibody, so that it is in excess of that required to saturate the Ag, we will have a quantity of free antibody that may bind to the competitor and there will still be an amount left to react with the solid-phase IgG. Thus, competition will only occur where extremely high concentrations of competitor are used. This can be illustrated by examination of the titration curves sketched in Fig. 2. Note that the plateau regions represent antibody excess for any given antigen concentration. The extent of these plateau regions varies according to the exact amount of antigen attached to the solid-phase. As we reduce the antigen, the plateau height values decrease. At the highest concentrations of Ag the titration curves are similar for different antibody concentrations, indicating that the antigen and antibody are behaving at maximum saturating levels. On dilution of the antigen we see (curve 4) that the plateau height is reduced, even where we know that the

182

Competitive

ELISA

2

1.5

8 z * cl 0

l

0.5

-

0 0

I

I

I

1

2

3

I 4

I 5

I

I

6

7

8

Conjugate dilution Fig. 2. Illustration of regions of conjugate excessand nonexcess when titrating conjugate against concentration of antigen.

antibody is available for higher OD values (curves 1 and 2). Here the antigen is the limiting factor in color development. In the competition assay a maximum plateau height, dependent on the amount of antigen attached, of around 1.0-1.5 OD should be selected. That is to say, find out which dilution of antigen produces serum titration curves giving a maximum plateau of these values, e.g., curves 3 and 4. From this titration curve we need to estimate the dilution of antibody yielding about 70% of the maximum plateau OD. Thus, using curve 4, we can illustrate this below. The conditions are now set for competition. We have: 1. Antigen dilution as for a curve. 2. Antibody dilution as shown in Fig. 3.

Direct Competitive

ELISA

183

2

7

1.5

8 m %

t 70%

Maximum

1

8

OS

. .

0

.

.

.

.

.

.

I

0

1

2

4 t

3

Dilution Conjugate

dilution

5

I 6

.

.

.

.

.

I 7

8

for test -

Fig. 3. Estimation of conjugate dilution for use in competition stage.

2.5. Competition

Assay Proper

1. Prepare the optimum antigen plates coated with guinea pig IgG as determined above. 2. Wash the plates. 3. Take the guinea pig IgG (homologous competitor) and the rabbit IgG (heterologous competitor). Dilute each to 40 l.tg/mL in blocking buffer. 4. Add 50 l,tL of blocking buffer to each well of the IgG-coated plate. 5. See plate design in Fig. 4. Make a twofold dilution range of the guinea pig and rabbit IgG by addmg 50 l.tL of the IgGs to row A. Do this in triplicate (3 rows for the guinea pig Ig, lA, B, C, and 3 rows for the rabbit IgG, column lD, E, F). 6. Double dilute the IgGs across the plate (l-l 1).

184

Competitive

ELISA

Dilution range of competitor -> Guinea

pig IgG

000000000000 000000000000 000000000000

Rabbit

000000000000

IgG

000000000000 000000000000 No competitor

000000000000

100%

000000000000

Competition

Constant d&.&ionof conjugate Fig, 4. Plate design for performance of competition assay. 7. Dilute the antiguinea pig conjugate (pretitrated level found above) in blocking buffer. Make up 6 mL. Do not allow tips to touch liquid in the wells. 8. Add 50 pL of the diluted conjugate to rows A, B, C, D, E, F, and G. Do not add to row H. Avoid touching liquid in wells with tips when adding conjugate. Mix the contents of the plates by gentle tapping. Add 50 p.L of blocking buffer to row H. 9. Incubate for 1 h at room temperature (or under conditions you used to titrate the conjugate). Rotate the plate to mix reagents every 10 min. 10. Wash the plates. 11. Add 50 pL/well of the OPD/H202 solution. 12. Stop the reaction after 10 min by addition of 50 pL/well of 1M H,S04. 13. Read OD in spectrophotometer at 492 nm.

2.5.1. Data-Typical

Results

Table 2 shows the results from the spectrophotometric reading of plates at the conclusion of the exercise. Figure 5 relates the OD values to the various concentrations of competitors added. The results are processed by taking the mean value for each of the triplicate dilutions (e.g., mean OD of column 1, wells A, B, C or column 5, wells D, E, F), as shown in Table 3.

2.5.2. Further Processing of Data 1. Take the mean OD reading of row G. This represents the OD resulting from the reaction of the conjugate with the solid-phase IgG and the conjugate

Direct

A B C D E F G H

1 0.04 0.06 0.07 1.12 1.14 1.13 1.35 0.05

Competitive

ELISA

Table 2 Plate Data From the Competition of Samplesof GuineaPig and Rabbit IgG for a Direct ELISA in Exercise2.5 2 3 4 5 6 7 8 9 10 0.05 0.07 0.10 0.23 0.35 0.56 0.78 0.98 1.12 0.06 0.08 0.12 0.25 0.41 0.61 0.79 1.01 1.14 0.05 0.09 0.13 0.21 0.43 0.58 0.81 1.05 1.17 1.23 1.34 1.35 1.34 1.36 1.29 1.37 1.36 1.41 1.24 1.35 1.35 1.36 1.39 1.34 1.36 1.33 1.34 1.25 1.34 1 38 1.38 1.41 1.42 1.35 1.33 1.38 1.34 141 1.35 1.36 1.32 1.29 1.34 137 139 0.06 0.05 0.07 0.08 0.04 0.07 0.05 0.04 0.07

185

11 1.34 1.35 1.36 1.32 1.32 1.34 1.32 0.08

12 1.34 1.38 1.34 1.34 1.38 1.32 1.45 0.04

only. This value should be similar to that obtained when you titrated the conjugate. This representsthe 0% competition OD, where we get most color. 2. Take the mean of the OD from row H. This represents the 100% competition level, i.e., where there is a total inhibition of the binding of antibody (not strictly true 100% control, since the conjugate was excluded from the test, but approximates very well). Thus, we have the 100% competition (degree of inhibition) and the 0% competition OD values. 3. Convert the OD values obtained for the wells that contained the two competitors mto percent competition using the two values calculated above. 2.5.3. Example from the Above Data Mean of row G = 1.35 (equivalent to 0% competition [a lot of color]) Mean of row H = 0.07 (equivalent to 100% competition [little color]) Subtract mean of row H from all values obtained. If value is minus then call this 0.

This determines the 0% competition level, (i.e., the range is from O-1.29 OD). Using a simple formula, we can calculate the percent competition of the samples. % competition = 100 - (Test OD -background x lOO)/Range

As examples, we have for the guinea pig Ig competition: Range = 1.29. Taking row 5 we have: 100 - [(0.16/1.29) x 1001 = 87.7%

Taking row 6 we have: 100 - [(0.33/1.29) x 1001 = 75%

Competitive

186

ELISA

1.6

Rabbit IaG

1.4 1.2 1 0.8 0.6

Guine:

0.4 0.2 0

III 0

III

12

3

4

5

6

I

Ill

7

8

I 9

10

11

12

Dllutlon competmgIgG + Fig. 5. Competition of guinea pig IgG and rabbit IgG for guinea pig system.

Taking row 7 we have: 100 - [(0.51/1.29) x 1001 = 60%

Repeat this exercise for your data! Plot the data relating the percent competition against the concentration or dilution of the IgGs as shown in Fig. 6. 1. 2. 3. 4.

2.5.4. Analysis of Data Notice that as you dilute out the homologous competitor (the guinea pig IgG), the competition reduces. The plateau of 100% competition is where the competing IgG is in large excess over that on the plate. Suggest what is happening at the 50% competition point. Notice that the competition curve IS sigmoidal.

Indirect Assay-Antigen

187

Competition

Table 3 Mean Values of Data in Table 2 for Various Dilutions of Competing Antigens in Exercise 2.5. Mean ABC minus row H guinea pig IgG

Mean DEF minus row H rabbit IgG

1 2 3 4 5 6 7 8 9 10 11

0 0 0 0.01 0.16 0.33 0.51 0.73 093 1.10 1.28

1.17 1.18 1.29 1.35 1.28 1.27 1.27 1.29 1.28 129 126

12

1.29

1.27

Mean row G - Mean row H = 1 29 = the range

5. Notice that the rabbit IgG hardly competes, and ask why. 6. How might the sensitivity of the assay be altered? Clue: What happens if we reduce (1) the amount of antigen on the solid phase or (2) the amount of conjugate

in the test?

3. Indirect

Assay-Antigen

I-Ag 1 + Ab + Ag2

+ W

Anti-AB*E

Competition +

S + Read

W

This exercise is similar to that for the direct competition assayfor antigen in Section 2. except that the antigen is detected by an unlabeled antibody (rabbit antiguinea pig IgG), which in turn is detected using an antispecies conjugate (antirabbit IgG linked to enzyme). The indirect assay conditions are optimized as in Chapter 6, by chessboard titration of antigen and antiserum using a constant dilution of antispecies conjugate. You can use the results of the chessboard titration shown in Table 1 to assess: (a) the best guinea pig concentration/dilution to adsorb to wells, and (b) the optimum amount of antibody to give about 70% binding to the optimum amount of antigen found above.

188

Competitive

ELISA

120

100

g .rl 5 gE

80

60

s s

40

20

0 0

1

Log

2

3

4 .

,,,dh.ztmn

5 .

6

7

competitor

8

9

LO

1112

v

Fig. 6. Percentage competition plots of guinea pig and rabbit IgG competing for guinea pig system.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

3.1. Materials and Reagents Ag = guinea pig IgG at 1 mg/mL for attachment to solid-phase. Ab = rabbit antiguinea pig IgG. Ab*E = swine antirabbit IgG conjugated to horseradish peroxidase. Microplates. Multichannel and single channel 10 mL and 1 mL pipets. 0.05M Carbonate/bicarbonate, pH 9.6. 0.05% PBS containing 1% BSA, Tween 20. Solution of OPD in citrate buffer. Hydrogen peroxide. Washing solution. Paper towels.

Indirect Assay-Antigen

I

00 00 I2 00 *fi 00 g .i 00 2 00 cl 1

189

Competition

0000000000 0000 000000 0000 000000 0000 000000 0000000000 0000000000 0000000000 0000000000 2

3

4

5

0% 100%

Competitors Fig. 7. Plate design for addition of competitors. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Small volume bottles. 1M sulfuric acid in water. Multichannel spectrophotometer. Clock. Graph paper. Calculator. Standard guinea pig IgG (known concentration). Three samples containing unknown concentrations of guinea pig IgG. Bovine IgG at 1 mg/mL.

3.2. Method 1. Coat wells of microplate with 50 pL of guinea pig IgG at optimum concentration found from chessboard titrations. 2. Wash the wells. 3. Take the homologous guinea pig IgG competitor of known concentration and the bovine IgG sample and dilute to 40 pg/mL in blocking buffer. 4. Take the 3 samples of unknown levels of guinea pig IgG and dilute l/10 in blocking buffer. 5. Add 50 pL of blocking buffer to all the wells of the microplate that is coated with the optimum guinea pig IgG. 6. Make twofold dilution range of all the diluted samples. Thus, add 50 PL of the initial dilution as shown in the plate plan in Fig. 7. Prepare duplicate columns of each by 8 wells.

190

Competitive

ELISA

Fig. 8. Representation of plate of competition assay; data in Table 5. 7. Add 50 FL of pretitrated rabbit antiguinea pig IgG to columns l-l 1, Do not allow tips to touch fluid in wells. Add 50 pL of blocking buffer to column 12. 8. Incubate under the conditions in which initial chessboard titrations were performed. Mix contents every 10 min if not rotating plates (unless overnight incubation is being used). 9. Wash the wells and blot. 10. Add 50 pL/well of the swine antirabbit conjugate at optimal dilution. 11. Incubate plates at 37°C for 1 h. Wash plates. 12. Add the substrate/chromophore (50 ltL/well, OPD/hydrogen peroxide solution), stop reaction after 10 min by addition of 50 PL 1M H2S04/well. 13. Read OD using a spectrophotometer at 492 nm. 3.3. Data-Typical Results Figure 8 shows a diagrammatic representation of plates at the conclusion of the assay. Table 4 shows the results read from the spectrophotometer. 3.3.1. Treatment of Data 1. Take the mean value of the OD from column 12 (0.08 in the example). 2. Subtract this from all the ODs of the rest of plate. 3. For each of the duplicate wells, find the mean OD for each competitor dilution. Thus, for example, see Table 5.

Indirect

Assay-Antigen

191

Competition

Table 4 Plate Data for Indirect Competition ELISA to Measure Antigen in Exercise 3 A B

1

2

3

4

5

6

7

8

9

10

11

12

1.31 1.12

130 1.14

1.31 1.32 1.29 1.27 1.25 125 1.21 1.15

133 1.34 1.28 1.19 1.24 1.26 1.22 1.18

1.32 1.28 1.09 0.85 0.67 0.41 0.30 0.13

1.32 1.29 1.10 0.79 0.69 0.45 0.29 0.15

1.26 1.21 1.00 0.76 0.47 0.26 0.16 0.09

1.34 1.18 0.98 0.74 0.48 0.27 0.17 0.07

1.12 0.88 0.68 0.44 0.22 0.09 0.08 0.07

1 10 0.83 0.66 0.43 0.23 0.09 0.09 0.06

134 1.32 1.34 1.32 1.29 1.34 1.33 1.35

006 0.08 0.09 0.06 0.08 0.09 0.08 0.06

C

0.79

0.77

D E F G H

0.57 0.33 0.18 0.09 0.08

0.54 0.36 0.15 0.10 0.07

Guinea

Bovine

pig test

Guinea

Gumea

pig 1

Pii3 2

Guinea Pig 3

0 %

100 %

control

Table 5 Mean Values of Plate Data Shown in Table 4 A B C D E F G H

12

34

56

78

9 10

1.22 1.05

1.24 1.25

1.24 1.21

1.22 1.12

105 0.81

0.70 0.47 0.27 0.08 0.00 0.00

1.21

1.01

1.15 1.15

0.74 0.59 0.34 0.21 0.06

0.91 0.65 0.38

0.59 0.36 0.14

1.24 1.26 1.24 1.21

0.19 0.09 0.00

0.00 0.00 0.00

1.26 1.25 1.27

1.14 1.13 1.07

11 1.26

4. Take the mean result of column 1 1 = 1.26. This is the 0% competition value. Use the formula to work out the percent competition of each IgG dilution. % competition

= 100 - [(OD test/range) x 1001

The worked examples are shown in Table 6. Plot the competition curves relating competition to log,, dilution or concentration as shown in Fig. 9. 3.3.2. Examination 1. Bovine

IgG competition

of Data

is very low, and the slope of the curve is very

different from thoseof homologouscontrol guineapig IgG.

192

Competitive Competition

ELISA

Table 6 Percent Values Calculated From Data Shown in Table 5

G. pig control

G. pig

G. pig

G. pig

W A

Iii@ B

I@ C

56

78

9 10

5 7 20 42 53 75 87 95

7 12 28 49 70 87 100 100

20 36 53 71 100 100 100 100

Bovine IgG

W

% Competition A B C D E F G H

12

34

6 20 35 67 79 93 100 100

5 5 9

1

10 10 11 15 20

2 Log,,

3 dilution

- G pig control x G pig A

4

5

competitor

+ Bovine

* G pig B

6

7

8

(2 fold)+

IgG

* G pig C

Fig. 9. Competition curves for various competitors; data shown in Table 6.

Indirect

Assay-Antigen

1

Competition

2

3

4

Log 1. dilution competitor + G pig control * G pig A

5

6

7

8

(2 fold) ->

+ Bovine IgG

* G pig B * G pig C

Fig. 10. Competition curves for various competitors; data shown in Table 6.

2. The curves for all guinea pig competitors are of similar shape. 3. The curves for guinea pig IgG samples A, B, and C are displaced as compared to the control IgG curve.

A standard curve relating the concentration of guinea pig IgG competitor in the liquid-phase to the competition achieved is shown by the control IgG. The concentration of IgG in the other samples can be determined with reference to this standard curve. Since the general slope of the curves is similar, we can compare the relative concentration at any point on the standard curve. Ideally the best comparison point is at 50% competition. Thus, draw a line across the 50% competition point on your graphs, as shown for the data in Fig. 10. Read the dilution of the test IgGs that give 50% competition, and then relate this to the known IgG concentration necessary to give 50% competition as determined from the standard curve at this point.

Competitive

ELISA

Thus, assuming starting concentration of guinea pig IgG at 2 pg/niL, We have for standard IgG 50% competition = l/64 Dilution for IgG A = l/20 Dilution for IgG B = 1140 Dilution for IgG C = l/140 Multiply the dilution factor by the 2 p&L to get concentration/ml for the test IgG. IgG C = 140/64 x 2 pg/l.tL = 4.4 pg/mL IgG B = 40/64 x 2 ug/pL = 1.25 l@mL IgG A = 20/64 x 2 pg/pL = 0.63 pg/mL Remember that the dilution range is in logic steps, so the antilog of the value has to be taken to obtain the dilution factor at 50%.

3.4. Conclusions 1. We have used a standard curve relating a known concentration of homologous competitor to its competing ability to measureunknown concentrations of the same IgG in samples. This has analogies to the radioimmunoassay approaches used in the quantification of hormones. 2. Note that if it is known that the substance for detection and quantification is the same immunologically (homologous) as the standard substance used to compute the standard curve, a single dilution of test could be used, and their competing ability read from a standard curve. 3. Such competition assayscan be used to determine the similarity of antigens in the same system competing for a single antiserum. The slopes of the competition lines can be compared to obtain a measure of antigenic relatedness.

4. Indirect I-Ag

Competition Assay for Antibody 4.1. Reaction Scheme +Ab +AB

+

Anti-Ab*E

+

Detection

S + Read

W W W I- = Microplate. Ag = Antigen. Ab = Retitrated antibodies against Ag. AB = Competing antibody (from a different species to Ab). Anti-Ab*E = Conjugated antispecies in which Ab was produced. S = Substrate and chromophore.

Indirect

Competition

Assay for Antibody

Detection

195

W = Wash. + = Addmon and incubation of reagents. Read= Read in spectrophotometer. In this exercise, the indirect assay is used to pretitrate the homologous antibody, as for Section 3. The optimized system is then competed with a dilution range of antibodies from another species (the conjugate must not react with the competing antibodies). In this assay, the pretitration of the homologous serum is slightly different than the antigen competition indirect ELISA in that we need to add the amount of homologous antibodies that just saturate the antigen coated on the plate, since we do not wish to leave excess free antigenic sites that could react with the competing antibody and have little influence on the binding of the homologous antiserum. Note that this kind of assay can be made using the direct ELISA using a conjugated homologous serum, as for the direct antigen competition ELISA. Such assays are becoming more common with the advent of the use of MAb reagents. 4.2. Materials and Reagents 1. Ag = Guineapig IgG at 1 mg/mL for attachmentto solid phase. 2. Ab = Pig antiguineapig IgG. 3. Ab*E = Goat antipig IgG conjugatedto horseradishperoxidase. 4. AB = 1X rabbit antigumeapig IgG standardserum. 2 rabbit serafrom animals injected with guinea pig IgG (unknown titer). 2 rabbit serafrom antibody-negativeanimals. 5. Multichannel and single channel 10 mL and 1 mL pipets. 6. 0.05M Carbonate/bicarbonate,pH 9.6. 7. PBS containing 1% BSA, 0.05% Tween 20. 8. Solution of OPD in citrate buffer. 9. Hydrogen peroxide. 10. Washing solution. 11. Papertowels. 12. Small-volume bottles. 13. 1M sulfuric acid in water. 14. Multichannel spectrophotometer. 15. Clock. 16. Graphpaper. 17, Calculator. 18. Microtiter plates.

Competitive

196

ELISA

2

lr+-x

1 to 8 dilutions

of serum

IgG dilutions

Fig. 11. Titration curves relating IgG dilutions on wells against different serum dilutions.

4.3. Data Figure 11 shows a graph relating pig antiguinea pig antibody titration curves to the IgG concentrations on the wells. This was obtained by chessboard titration of captured guinea pig IgG against dilutions of the pig antiserum, with a constant optimal dilution of antipig conjugate. The

conditions for the indirect chessboardtitration were as for those described for the titration of the rabbit antiguinea pig serum in Chapter 6. From these data we can: 1. Assessthat the best antigen concentration for use is the competition assay. Select the IgG concentration that gives a plateau maximum (in antibody excess) of around l-l.5 OD. (curves 4 and 5 in Fig. 11). 2. Select the dilution of serum that just saturates this level of IgG (approx l/100).

Indirect

Competition

Assay for Antibody Detection

197

4.3.1. Increasing the Confidence of the Titration Curve Results Since in the chessboard titration we are only using a single dilution range of antibody against the antigen, it is essential to titrate the antiserum in multiple rows against the antigen level found to be optimal, i.e., we adsorb IgG at a level equivalent to the fourth or fifth dilution used in the above test, then titrate in quadruplicate a dilution series of serum against it. In this way, we can observe the variation in results and assess the confidence in the titer of antibody that just saturates the antigen used in the competition assayproper. This may be necessary where, for example, one obtains poor competition in the test proper with low sensitivity, indicating that too high or very much too low a concentration of antiserum was used. 4.4. Competition Assay Proper 1. Add 50 pL of guineapig IgG to plates at the optimal concentrationfound 2. 3. 4. 5.

6. 7. 8. 9. 10. 11. 12. 13. 14.

in Section 4.3.1. Incubate plates under conditions used for coating in Section 4.3.1.) and wash plates. Take the rabbit antiguinea pig sera, label the standardantiserum 1, and label the two unknown titers sera 2 and 3. Take the two sero-negative rabbit sera and label them 4 and 5. Dilute the rabbit sera to l/50 in blocking buffer (make up 0.5 mL of each, i.e., add 10 p.L serum to 0.5 mL of buffer). Add 50 pL of blocking buffer to all the antigen-coated plate wells. Add 50 p.L rabbit serum 1 to wells Hl and H2. Add duplicate rows of other serainrowH(serum2inH3,4;serum3inH5,6;serum4inH7,8;serum 5 in H9, 10). Dilute the sera using a multichannel pipet, transferring and mixing 50 pL in each step. We thus have a dilution range from l/100 (row H) to l/12,800 (row A) for each of the sera. Incubate for 30 min at 37°C. Do not wash the plate. Add 50 pL of the swine antiguinea pig serum at the optimal dilution to each well from columns l-l 1. Do not touch pipet tips in liquid of wells when adding reagent. Add 50 pL blocking buffer to column 12. Incubate for 1 h at 37OC. Wash the wells. Add 50 pL of the optimal dilution antiswine conjugate to each well. Incubate at 37°C for 1 h. Add 50 pL/well of substrate and OPD; incubate for 10 min. Stop the reaction by addition of 50 pL 1M HzS04 to each well.

Competitive

198

ELBA

Fig. 12. Representation of plate showing competition assay; data in Table 7. Table 7 Plate Data From Exercise 4.3. Showing Competition of Indirect Assay by Antibodies

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

1.12 1.07 0.89 0.63 0.42 0.23 0.13 0.08

1.16 1.09 0.91 0.61 0.41 0.26 0.12 0.09

1.21 1.21 1.10 0.87 0.63 0.43 0.23 0.12

1.20 1.19 1.09 0.89 0.65 0.45 0.25 0.10

0.78 0.56 0.34 0.21 0.09 0.08 0.07 0.08

0.84 0.58 0.32 0.19 0.08 0.07 0.08 0.07

1.14 1.15 1.13 1.10 1.16 1.13 1.15 1.14

1.13 1.12 1.09 1.09 1.09 1.14 1.12 1.16

1.14 1.16 1.15 1.13 1.14 1.14 1.16 1.14

1.15 1.14 1.12 1.15 1.13 1.16 1.15 1.15

1.11 1.15 1.17 1.16 1.15 1.15 1.17 1.15

0.07 0.09 0.07 0.06 0.08 0.07 0.06 0.07

4

0%

Standard Serum

1

2

3

100%

4.4.1. Typical Data Figure 12 shows a representation of the ELISA plate after stopping. Table 7 shows the data. 4.4.2. Processing Data This is similar to the other competition assays performed. 1. Column 12 = 100% competition value; take the mean OD = 0.08.

Indirect

Competition

Assay for Antibody

Detection

199

Table 8

Mean Values of Data in Table 7 A B C D E F G H

12

34

56

78

9 10

11

1.05 1.oo 0.90 0.52 0.34 0.16 005 0.00

1.12 1.12 1.01 0.80 0.56 0.36 0 16 0.03

0.71 0.49 0.25 0.12 0.00 0.00 0.00 0.00

1.06 1.06 1.03 1.01 106 1.06 1.06 1.07

1.07 1.07 1.06 1.06 106 1.07 1.07 1.06

1.08 107 1.09 108 107 1.07 1 09 1.07

2. Subtract this from OD values of all wells. 3. Take the mean OD of the duplicates for the competitors. This is shown in Table 8. Plot the data. Relate the loglo dilution of each of the antiserum to the percent competition as illustrated in Fig. 13. 1. 2. 3.

4.

4.4.3. Analysis of Data Note that the curves for the rabbit antisera are similar. All the samples compete. Sample 3 is a strong competitor since it gives high competition at higher dilutions than the standard.Sample 2 is aweaker competttor than the standard. The negative sera give little or no competition even at low dilutions. The activity of each of the two sera can be compared to the standard competing antiserum. Arbitary units can be ascribed to the standard serum so that serum titers could be expressed against this. As an example, let the titer of the standard serum at 50% competition be 1000. The relative titers of the other two test sera can then be related to this. Since the same dilution range was used for the samples we have at 50% competition for serum 2, it is 2x stronger than the standard, so we need twice as much antiserum to compete to the same level as the standard. Therefore the relative titer of the serum 1sl/500. For serum 3 at 50% we require 5x less antiserum to give the same result as the standard, so the titer IS 5000. The difference in the dilution factors necessary to give 50% competition is easily assessed from the graphs shown above. Note that this processing only holds true if the competition curves show similar characteristics (shape). Considerable variation m slopes indicates that there is a different population of antibodies in the competing serum. As m all assays,the general picture of titration curves is best examined by the assay of as many sera as possible.

Competitive

200

3 0 0

ELISA

4

l

+

+

+

*

1

2

3

4

5

6

7

8

Log 1o dilution serum Fig. 13. Competition curves for various competing sera;data in Table 8. 6. Indirect

Assay Competition for Antibody-Detection Using a Single Dilution of Test Serum 5.1. Reaction Scheme The optimization of the antigen, homologous serum, and detecting serum is as described in the last exercise. In this assay we use the standard rabbit antiguinea pig serum as a full titration range in 3 rows of the plate. The rest of the plate contains a number of rabbit sera of high, medium, and low titer against guinea pig IgG as used in the ‘spot-test’ in Chapter 6. Not all the sera can be examined in this exercise since only a single plate is being used. The assay is identical to that in Section 4.4., except that duplicate samples of sera are assessedat a single dilution for their competing ability. The titer of the serum is then read from the standard curve obtained on full dilution of the standard serum. The test therefore has two stages: (1) The titration of the homologous antiserum and solid-phase antigen in a chessboard indirect ELBA, followed by accurate titration of the serum using a constant amount of cap-

Detection Using a Single Dilution

of Test Serum

201

tured IgG and replicate dilutions of the antiserum and (2) the competition assay proper.

5.2. Materials 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

and Reagents

Ag = guinea pig IgG at 1 mg/mL for attachment to solid-phase. Ab = pig antiguinea pig IgG. Ab*E = goat antipig IgG conjugated to horseradish peroxidase. AB = 1X rabbit antiguinea pig IgG standard serum. 32 rabbit sera, including high, medium and low titer sera against guinea pig IgG and seronegative sera. Multichannel and single channel 10 mL and 1 mL pipets. 0.05M Carbonate/bicarbonate, pH 9.6. PBS containing 1% BSA, 0.05% Tween 20. Solution of OPD in citrate buffer. Hydrogen peroxide. Washing solution. Paper towels. Small-volume bottles. lit! sulfuric acid in water. Multichannel spectrophotometer. Clock. Graph paper. Calculator. Microtiter plates.

5.3. Titration-Stage Repeat the chessboard titration and accurate antibody titration of pig antiguinea pig system, as in Section 4.4., or use these results for conditions. From the data the best antigen concentration, and the dilution of swine antibody that just saturates the IgG, can be determined.

5.4. Competition

Assay Proper

1. Add 50 pL of guinea pig IgG to plates at optimal dilution. Incubate using the same regimen as in Section 5.3. Wash plates. 2. Take the rabbit test sera and dilute them to l/50 in blocking buffer. Use the micronics racks for dilutions so that the samples can be added using the multichannel pipet. Take the standard rabbit antiserum and dilute to l/50. 3. Add 50 pL blocking buffer to columns 1,2 and 11 and 12. 4. Add 50 p.L of the diluted standard rabbit serum to Hl and 2. Make a twofold dilution range of the serum to Al and 2. 5. Add the test samples to the wells as duplicates, as indicated in Fig. 14.

202

Competitive

ELISA

Test sample duplicates

T

0% 100% Controls

Standard antibody dilution range Fig. 14. Plate layout for “spot testing” in competition assay. 6. Incubate the plates for 30 min at 37°C. 7. Add 50 PL of the optrmum dilution of swine antiguinea pig serum in blocking buffer, incubate for 1 h at 37OC,and mix contents of plates every 10 min. Do not touch tips in liquid of the wells when adding this serum. Do not add this serum to column 12. Instead add 50 pL blocking buffer. 8. Wash the plates. 9. Add 50 pL/well of the antiswine conjugate diluted in blocking buffer (optimum dilution). Incubate 1 h at 37°C. 10. Wash the plate. 11. Add 50 pL/well of the substrate/OPD solution; incubate for 10 min at room temperature. 12. Stop the reaction with 50 pL/well of 1M HzS04. 13. Read the plate using a spectrophotometer.

Detection Using a Single Dilution

of Test Serum

203

Table 9 Plate Data From Exercise 5.4. ‘Spot-Test’

A B C D E F G H

1

2

3

4

5

6

7

8

9

10

11

12

1.21 1.03 0.91 0.76 0.53 0.31 0.12 0.06

1.24 1.05 0.90 0.73 0.54 0.34 0.13 0.08

0.34 0.19 1.13 0.98 0.06 0.34 1.21 0.06

0.32 0.18 1.15 0.96 004 0.36 1.23 0.09

1.23 0.43 067 0.13 0.34 0.14 1.14 0.15

1.21 0.45 0.69 0.12 0.36 0.16 1.11 0.12

1.12 0.56 1.11 0.16 0.16 1.17 0.09 0.23

1.12 0.58 1.13 0.13 0 18 1.19 0.07 0.27

0.09 0.78 0.12 0.78 1.23 0.08 0.67 0.10

0.09 0.78 0.14 0 80 1.21 0.10 0.69 0.12

1.23 1.21 1.24 1.24 123 1.21 1.26 1.23

0.07 0.09 0.08 0.09 0.08 0.07 0.05 0.06

Fig. 15. Representation

of “spot testing”

competition

assay; data in Table 9.

5.4.1. Typical Data

Table 9 shows data obtained from the spectrophotometer. Figure 15 shows a diagrammatic representation of the stopped plate. 5.4.2. Treatment

of Data

As for the other competition results: 1. Take the mean of column 12, and subtract this from all results from other wells. 2. Take the mean from column 11 (after subtraction above).

204

Competitive

ELISA

Table 10 Mean Values of Test Sera From Table 9 Processed as Percent Competition Values % competition results A B C D E F G

12

34

56

78

9 10

0 16 28 41 59 78 100

73 90 7* 23 100 76 100

2* 68 48 96 76 93 95

16 57 9 93 91 4” 84

99 39 95 38 0 98 97

3. Express the other OD values as a percentage of the range 0 to the mean of column 11, i.e., from O-l. 16 in the example above. 4. Take mean OD of the duplicate wells. Formula is percent competition in each well. 100 - [(Test OD/1.16) x 1001 Plot the standard serum competition activity relating competition to log,, of the dilution. 5. Read the relative titers of the other competition results from the curve. 6. Another approach to evaluation of spot-testing is that whereby accepted negative sera are assessedascontrols. Several sera can be included in a test so that their mean competition value and its variation can be assessed. Thus, sera giving higher values of competition under the same conditions (with prescribed confidence limits), can be assessedfor positivity. Studies on a large number of negative sera give better population data as described for the other assays,so that chosen negative controls may be added from the defined population (see Table 10). In the above example, the sera with asterisks could be the negative controls in order to test whether the system was ideal. The percent value of their mean plus a defined interval as a percent of this mean (as directed from large population studies) could be given. Here we have mean = 3%. Assume that twice this mean is the acceptable upper limit for negative competition values. Therefore, sera could be ascribed as positive with competition values 26%. Actual titers could be read as in Section 5. above. 5.5. Notes

We have used a dilution of l/100 for the test serain the example. This is based on preliminary studies establishing the dilution as being optimal for

Overall Conclusions on Competition

Assays

205

distinguishing positive and negative values. This must be attempted in your laboratories for specific disease studies. The approach to examination of negative populations has already been discussed. In the case of competition assays, a lower dilution of test serum might be used (effectively increasing the sensitivity of the assay), since nonspecific factors detected in the indirect assay do not seem to affect competition assay results. Construction of full-scale serum titration competition curves of many negative and positive sera will nominate the best dilution (with definable confidence limits) of serum to be used. The sources of such sera have already been discussed. Thus, for any particular dilution used in the competition assay,an upper limit of negativity should be definable (as a competition value) above which positivity of antibody will be detected. Once competition assays have been characterized in central laboratories it is usually simple to read the assays by eye, with good levels of precision and sensitivity. In these cases the selection of appropriate negative controls that define upper limits of negativity as determined by eye, is important. 6. Overall

Conclusions

on Competition

Assays

1. They provide a relatively simple method once the homologous systems have been titrated. 2. These assayscan be read by eye, with some loss of sensitivity and reduction in confidence limits. 3. In all the examples above we have used 50 pL of competitor and 50 l.tL of homologous serum as a mixture to compete for only 50 pL of antigen on the solid-phase. You can alter the volumes to suit, for example: a. 100 pL antigen solid phase vs 50 p,L homologous serum and 50 pL competing antigen (or antibody). In this case,the pretitration would be with 100 l.tL solid-phase Ag vs 50 l.tL serum dilutions + 50 pL blocking buffer. b. 50 pL solid-phase antigen vs 25 p.L homologous serum + 25 l,tL competing antigen (or antibody). In this case,the pretitration would be between 50 FL solid-phase antigen and 25 pL antibody dilutions + 25 ltL buffer. c. The competitor and homologous serum can be mixed together in another plate before addition to the solid-phase antigen plate. These types of assayscan be termed Inhibition Assays since the reagents are not directly competing in the same system, Differences in results can be observed by alteration of the sequence of reagents i.e., where true competition and inhibition methods are used, In practice, the mixing of reagents in a true competition assay gives the most sensitive assaysand best reflects avidity differences between reagents.

CHAPTER9

Immunochemical

Techniques

1. Introduction The scope of this book does not allow a complete description of the many techniques available for improving and facilitating immunoassays in general. There is a large amount of literature covering techniques, and these can be consulted for specific problems. Examination of many of the catalogs produced by commercial companies is useful, since they often include good technical sections describing methods using their products. This chapter contains the practical basics of conjugation (a large field in itself) and details of other immediately useful techniques that might be first desired in starting ELISA. The book recommended (I) should be considered definitive in the laboratory since it is extremely “digestable” and covers a large field of methods, all of which are relevant to ELISA. 2. Labeling Antibodies with Enzymes Antibodies can be readily labeled by covalent coupling to enzymes (2-7). The ideal product for any coupling reaction should have a 1: 1 ratio of antibody to enzyme with no loss of specific activity of either reactant, but this is technically unachievable. However, owing to the amplification of the signal by the enzyme action, even relatively poor conjugates have required sensitivities. A large number of enzymes have been used to label antibodies. The most commonly used are horseradish peroxidase, alkaline phosphatase, and pgalactosidase.The ideal enzyme considerations arecost, stability, size, and easeof conjugation. The enzyme should have a high catalytic activity, and a range of substratesthat yield both soluble products and insoluble products (for immunoblotting, immunocytochemical techniques). The purchasing of enzyme-linked reagents from commercial sources is recommended, but for laboratory-produced specific reagents, such as monoclonal antibodies (MAbs) or affinity-purified antibodies, conjugates will need to be prepared in the laboratory.

207

208

Immunochemical

Techniques

2.1. Coupling Antibodies to Horseradish Peroxidase Two general methods are used for the preparation of antibody peroxidase conjugates, the twstep glutaraldehyde method and the periodate method. Good batches of horseradish peroxidase (HRP) can be determined by measuring the ratio of the HRP absorbance at 403 and 280 nm (RZ = OD 403 nm/OD 280 nm). This ratio should be at least 3.0. Good reagents designed for coupling are available commercially.

2.1.1. Gluteraldehyde

Coupling

In the two-step glutaraldehyde method, glutaraldehyde is first coupled to pure HRP via the relatively few reactive amino groups available on the enzyme. By performing this step in high glutaraldehyde concentrations, very few HRP-HRP conjugates are formed. The HRP-glutaraldehyde mixture is then purified and added to antibody in solution. This method has a low coupling efficiency, so the HRP-antibody conjugates need to be separated from unconjugated material for optimum sensitivity. The HRP must be pure to minimize crosslinking of enzyme molecules to contaminating proteins during the first step of the following procedure. 1. Dissolve 10 mg of HRP in 0.2 mL of 1.25% glutaraldehyde (electron microscopic grade) in 100 rnM sodium phosphate, pH 6.8. CAUTION: Glutaraldehyde is hazardous. Work in a fume hood. 2. After overnight incubation at room temperature; remove excessfree glutaraldehyde by gel filtration.* 3. Concentrate the enzyme solution to 10 mg/rnL (1 rnL final volume) by ultrafiltration or by dialysis against 100 mM sodium carbonate/sodium bicarbonate buffer, pH 9.5, containing 30% sucrose. Change the buffer to 100 rnM sodium carbonate-bicarbonate, pH 9.5, either by dialysis or by washing on the ultrafiltration membrane. 4. Add 0.1 mL of antibody (5 mg/rnL in 0.15M NaCl) to the enzyme solution and check that the pH > 9.0. 5. Incubate at 4OCfor 24 h. *Use a gel matrix with an exclusion hmit of 20,~50,000 for globular protems. Use medmmsized beads (approx 100 pm in diameter). Prepare a column with 5 mL of bead volume according to the manufacturer’s instructions. To make the column easier to load and run, first add 20 pL of glycerol and 20 pL of 1% xylene cylanol. The column should be prerun with a minimum of 10 column volumes of 0.15M NaCl. Allow the column to run until the buffer level drops just below thetop of thebedresm.Stopthe flow of thecolumn.Carefullyloadthecolumnwith theglutaraldehyde-treatedHRP. Release the flow, andallowthe HRP to run into the column.Justasthe level of the HRP solutiondropsbelow the top of the column,carefully add0.15M NaCl. Run the columnwith 0.15M NaCI. Poolthe fractionsthat look brown,Thesecontainthe active enzyme

Labeling

Antibodies

with Enzymes

209

6. Add 0.1 mL of 0.2M ethanolamine, pH 7.0. Incubate at 4°C for 2 h. At this stage,there will be present in the solution the uncoupled HRP, the uncoupled antibody, and the HRI-antibody conjugate. For some assays,no further purification is necessary. In these cases, the uncoupled HRP will not bind to any antigen and will be lost during any washes prior to enzyme detection, Further purification will require separation based on the differences between the three species. The easiest separation will be between the uncoupled HRP and the two antibody-containing fractions. If the antibody binds to protein A, the antibodies can be removed simply as described later. Separation between the two antibody fractions can be achieved by gel filtration (a 50 mL S300 or equivalent) or affinity chromatography on a Concanavalin A column (eluted with 0.2M glucose or methylmannoside). Alternatively, the whole separation can be achieved on the basis of size by gel filtration. Column eluates can be assayed by enzyme activity, absorbance at 403 nm, or absorbance at 280 nm. 2.1.2. Periodate Coupling Periodate treatment of carbohydrates opens the ring structure and allows them to bind to free amino groups. Coupling antibodies and horseradish peroxidase with periodate linkage are an efficient method. This method is based on refs. (4) and (5). 1. Resuspend 5 mg of HRP in 1.2 mL of water. Add 0.3 mL of freshly prepared O.lM sodium periodate in 10 mM sodium phosphate, pH 7.0. 2. Incubate at room temperature for 20 mm. 3. Dialyze the HRP solution versus 1 mM sodium acetate, pH 4.0, at 4OCwith several changes overnight. 4. Prepare an antibody solution of 10 mg/mL in 20 mit4 carbonate. 5. Remove the HRP from the dialysis tubingm and add to 0.5 mL of the antibody solution. 6. Incubate at room temperature for 2 h. 7. The Schiff’s bases that have formed must be reduced by adding 100 pL of sodium borohydride (4 mg/mL in water). Incubate at 4OCfor 2 h. 2.1.2.1. NAKANE AND KAWAOI METHOD OF ENZYME ACTIVATION 1. Dissolve the horseradish peroxidase (HRPOSigma Type VI, RZ = 3) in 1.O mL of freshly prepared 0.3M sodium bicarbonate, pH 8.1 (should be this pH on making up). Note the mg/mL on bottle of HRPO. 2. Add 0.1 mL of a 1% solution (v/v) fluorodinitrobenzidine in absolute ethanol. Mix for 1 h (leave on bench and gently swirl every 10 mm). 3. Add 1.0 mL of 0.08M sodium periodate (NalOJ in distilled water. Mix gently for 30 mm at room temperature (swirl every 5 min).

210

Immunochemical

Techniques

4. Add 1.O mL of 0.16M ethylene glycol (ethanediol) in distilled water. Mix gently (as above) for 1 h. 5. Dialyze against O.OlM sodium carbonate/bicarbonate buffer, pH 9.5, at 4°C (three changes vs 500-1000 mL).

Conjugation is as follows: 1. Add the IgG (or other protein) dialyzed against O.OlM carbonate/bicarbonate buffer, pH 9.5) at a ratio of 5 mg IgG (protein) to 1.33 mg of activated enzyme. Note: You know the volume of your activated enzyme and know the original mg/mL. Therefore, you know the effective concentration of the activated enzyme and can add so many milligrams in a certain volume. Mix, and stand at room temperature for not ~3 h (overnight is suitable). 2. Add 1 mg of sodium borohydride (NaBH,)/mg of enzyme used. Make the sodium borohydride up fresh to about 200 mg/mL, and add relevant volume containing the correct number of milligrams. 3. Dialyze against PBS. 4. You may wish to separate the free enzyme by methods already described, but in most ELISAs, this is not necessary.

3. Coupling Antibodies to Alkaline Phosphatase Conjugatation of antibodies to alkaline phosphatasecan be made using a one-step procedure with glutaraldehyde. The conjugates retain good immunological and enzymatic activity, but can be large and heterogeneous in nature. The major drawbacks are the high cost of the enzyme and the need to use very concentrated solutions of enzyme and antibody. 1. Mix 10 mg of antibody with 5 mg of alkaline phosphatase in a final volume of 1 mL. Alkaline phosphatase is usually supplied as a suspension in 65% saturated ammonium sulfate, which should be centrifuged at 4OOOg for 30 min (5 min in microfuge). The antibody solution can then be added to resuspend the enzyme pellet. 2. Dialyze the mixture against four changes of O.lM sodium phosphate buffer, pH 6.8, overnight. This is essential to remove free amino groups present in the ammonium sulfate precipitate. 3. Transfer the enzyme-antibody mixture to a container suitable for stirring small volumes. In a fume hood, add a small stir bar and place on a magnetic stirrer. Slowly, with gently stirring, add 0.05 mL of a 1% solution of EM gradeglutaraldehyde.CAUTION: Glutaraldehyde is hazardous, 4. After 5 min, switch off the stirrer, and leave for 3 h at room temperature. Add 0.1 mL of 1M ethanolamine, pH 7. 5. After 2 h of further incubation at room temperature, dialyze overnight at 4°C againstthree changesof PBS.

Coupling Antibodies

to Alkaline

Phosphatase

211

6. Centrifuge at 40,OOOg for 20 min. 7. Storethe supematantat 4°C in thepresenceof 50% glycerol, 1 mM ZnCl,, 1 mM MgC12,and 0.02% sodium azide. The procedure may be scaled down to the 1-mg antibody level if the antibody and enzyme concentration is reduced by a factor of 10. Here, the time allowed for coupling should be increased to at least 24 h. The yield of conjugate may be reduced. 3.1. AvidbBiotin Systems in ELISA The specific binding between avidin (an egg-white protein) and biotin (a water-soluble vitamin) hasbeen exploited in ELISA. Avidin is a tetramer containing four identical subunits, each of which contains a very highaffinity binding site for biotin. The binding is not disturbed by extremes of salt, pH, or chaotropic agents, such as guanidine hydrochloride (up to 3M). The avidin-biotin system is well suited for use as a bridging or sandwich system in association with antigen-antibody reactions. The biotin molecule can be easily coupled to either antigens or antibodies, and avidin can be conjugated to enzymes (and other immunological markers, such as fluorochromes, colloidal markers, and ferritin). This section will deal briefly with applications of the biotin-avidin system to ELISA. An excellent outline of reagents and biotin-protein-labeling methods (biotinylation) can be found in ref. 8. Three basic systems can be outlined as shown in Fig. 1. 3.1.1. LAB System An antigen immobilized on a microtiter well is detected by incubation with a primary antibody. After washing, this is detected by incubation with an antispecies antibody that is biotinylated (linked to biotin molecule(s). Again after washing, the complex is detected by the addition of avidin, which is linked to enzyme followed by addition of the relevant substrate. 3.1.2. BRAB System This is essentially the same as the LAB system, except that the avidin is not conjugated to an enzyme. Here the avidin acts as a bridge to connect the biotinylated secondary antibody and biotinylated enzyme. Since the avidin has multiple biotin binding sites, this system allows more biotinylated enzyme to be complexed with a resulting amplification of signal, thus making the systempotentially more sensitive than the LAB system.

212

Immunochemical Labelled

Avidin-Biotin

System (LAB)

Bridged

Avidin-Biotin

System (BRAB)

Techniques

c

Substrate

Avidin-Biotin

Complex

b

System (ABC)

Antigen Primary

BiOth antibody

Secondary detecting antibody antibody

Avidin

ElUple Coloar

development

Fig. 1. Different systems for use of avidin and biotin in ELISA.

Preparation

213

of Immunoglobulins Table 1 Compounds Available for Introduction of Biotin Biotinylation

reagent

NHS-LC-Biotin NHS-Biotin SULFO-NHS-Biotin NHS-LC-Biotin NHS-SS-Biotin Photo Activatable Biotm-HPDP Iodoacetyl-LC-Biotin Biotm hydrazide Biotinylated Protein A

Reactive against Primary amines Primary amines Primary ammes Primary amines Primary amines Nucleic acids Thiols Thiols Carbohydrates Antimammalian IgG

3.1.3. ABC System

This is almost identical to the BRAB system except that it requires preincubation of biotinylated enzyme with avidin to form large complexes that are then incubated with the secondary antibody. In this way there is a large increase in signal owing to the increase in enzyme molecules. 3.2. Methods

for Labeling

with

Biotin

There are many biotinylated commercial reagents designed for use in ELISA. Reference 7 illustrates the various methods for the introduction of biotin onto reagents for use in ELISA using a variety of chemicals. A brief survey is shown in Table 1 to illustrate the versatility of labeling methods for proteins, carbohydrates, and nucleic acids. 4. Preparation of Immunoglobulins About 10% of serum proteins are immunoglobulins. After immunization, the specific antibodies produced are about l-25% of this fraction, so that the required immunoglobulin (in ELISA) may be from O. l-2.5% of the total protein in a serum. Some assays are favored by the relatively crude fractionation of serum to obtain immunoglobulins, e.g., for use in binding to plates in trapping (sandwich assays) to avoid competition for plastic binding sites by other serum proteins. There are several methods for separation of immunoglobulins for use in ELISA. These procedures are suitable for polyclonal antibodies, but not necessarily for MAbs. The isolation of total immunoglobulins (Ig) as compared to the purification of specific immunoglobulins is relatively simple.

Immunochemical

214

Techniques

4.1. Salt Fractionation

Two salts are used for selective Ig precipitation, ammonium sulfate and sodium sulfate. The concentration of ammonium sulfate is expressed as a percentage saturation, whereas the concentration of sodium sulfate is expressed as percentage w/v. The concentration of salt at saturation depends on temperature, particularly for sodium sulfate (5x less at +4”C). The isolation of mammalian IgG and IgA by ammonium sulfate precipitation depends on the volume of the serum being processed. For large volumes, the salt is added directly, whereas for small volumes, the salt is added as a concentrated solution. As already indicated, proteins are precipitated by different amounts of ammonium sulfate. This is a method that can be used to obtain samples of sufficient purity for most ELISAs. The initial volume of serum here is given as 10 mL. Adjust the volumes accordingly to suit the starting volume of your serum. 1. To 10 mL. serum add 2.7 g of (NH&S04. Add a small quantity in steps. 2. 3. 4. 5. 6.

Stir constantly at room temeperature. Incubate at room temperature for 1 h while stirring. Centrifuge at approx 5OOOgat 4°C for 10-15 mm. Discard the supernatant fluid. Dissolve the pellet in 2-3 mL of distilled water. Add 0.5 g (NH&SO+ Stir constantly at room temperature.

7. Centrifuge as before. 8. Dissolve pellet in 10 mL distilled water or PBS. 9. Dialyze against appropriate buffer for use in ELISA, and so on, or dialyze against distilled water and then freeze-dry. 4.2. Ion-Exchange

Chromatography

After salt fractionation, IgG can be purified further on DEAE-cellulose, DEAE-Sephadex A-50, or DEAE-Sephacel. Such methods are not described in this volume, but a large amount of literature is available. 4.3. Protein A and Protein G Protein A has some affinity for the Fc of most mammalian IgGs and

can be used for their isolation. Although protein A as used in immunoassays has little practical use in detecting sheep, bovine, and goat IgG, they can be purified when the protein A concentrations are high, as in the commercially available protein A-sepharose or protein A conjugated to Affi-gels (Bio-Rad, Richmond, VA) or glass beads. Such reagents are

Immunosorbents

215

very useful in rapid separation of most mammalian IgGs. Briefly, 5 mL Protein A columns are equilibrated with PBS. Serum or crude IgG is then added and elution with PBS maintained. The IgG attaches to the column (via reaction to the protein A bound to the inert matrix), and the other serum proteins pass through the column. The bound IgG is then eluted using a 0.9% sodium chloride solution containing 0.6% acetic acid or by addition of a solution of sodium thiocyanate (2-W). Such methods are particularly useful in the purification of mouse IgGs from MAb ascites preparations. 5. Immunosorbents A breakthrough in the easy use of immunosorbents has been made with the availability of reagents, such as n-Hydroxysuccinimide-derivatized agarose (Bio-Rad). This gel is can be washed three times in cold distilled water and then be used to covalently attach any protein, merely by incubation of that protein(s) in a wide variety of buffers (such as 0.1 sodium carbonate buffer). Blocking of unreacted active sites on the gel is achieved by the addition of ethanolamine or by merely leaving the gel overnight. Such gels are thus very easy to prepare. Antisera can then be added in neutral buffers, and addition of some detergents (e.g., 0.5% Tween 80) minimizes nonspecific adsorption of serum proteins. Desorption of bound antibodies is then achieved by addition of chaotropic ions (sodium thiocyanate), or organic acids with low surface tension or pH extremes. Thus, such affinity techniques can be used to get rid of unwanted cross reactions from sera, e.g., if a serum has antibovine IgG activity, this can be adsorped out by passing that serum over an affinity column with bound bovine serum or IgG. In this case, the antibodies passing through the column will be free from antibovine activity. Other immunosorbents are available commercially based on beaded agaroseor glass. A wide variety of proteins, such as whole serum or IgG, can be purchased attached covalently to beads and are extremely convenient (but expensive) for the removal (as an example) of unwanted crossreactive antibodies from small volumes of antisera. The beads are simply added to an antiserum and, after a short incubation, are separated by centrifugation (microfuge, 12,000 rpm for lo-30 s). The agarose beads thus capture any unwanted antibodies on the solid-phase, leaving the antiserum free of that “contaminant.” This method has the advantage over blocking by addition of high levels of specific protein (against which

216

Immunochemical

Techniques

the unwanted antibodies react) in that there is complete separation of immune complexes, which may interfere with ELISAs. Such reagents can be reused by eluting the immunologically bound protein using similar methods (low pH, and so on) as described above, followed by extensive washing. The section on immunoaffinity purification in ref. I should be consulted for extensive practical details of methods. 6. Production of Antisera The raising of antisera in laboratory animals would fill a manual in itself. The variability in immune response within and between species and the various antigens used mean that no brief rules can be given, and reading of the relevant scientific literature is essential. Generally, the administration of a nonreplicating agent requires the addition of an adjuvant, whose effect is to stimulate the immune system so that efficient presentation of the antigen takes place. 6.1. Immunization The purpose of immunization is to obtain high-titered antisera that binds strongly to antigen (high avidity). The properties of antisera are determined by the genetic composition of the animal injected (particularly the Ir genes). This means that there can be great variation in the quantitative and qualitative aspects of antisera from between species and even between individuals of the same species. This should be borne in mind when considering the use to which the serum is being made. In preparing sera one should: (1) always obtain a preimmunization serum and (2) never automatically pool sera. Point (2) is particularly important if a defined property of an antiserum is required, e.g., in discrimination of antigens. Up to a certain degree, an increase in the dose (weight) of antigen will increase antibody titer, however this may also increase crossreactivity. Adjuvants also increase the immunogenicity of proteins. Haptens should also be labeled with carrier proteins to elicit an immune response. The carrier protein should also be foreign to the host to be recognized by the T-cells. For most immunogens the interaction of T- and B-cells is essential for antibody production. The animal species chosen can be important, The animal species most often used in laboratories are rabbits, goats, guinea pigs, pigs, sheep, goats, and rats. Commercial companies may favor horses and donkeys for large-scale preparations. Many animals contain crossreactive antibodies in their serum before immunization. This could complicate their

Production

of Antisera

217

use in ELISA, and some simple absorption technique may be required (or may have been performed in commercial preparations) to block such reactions. Another point to remember is that many smaller animals can be immunized as compared to only a few larger animals owing mainly to cost considerations. Because of the variations in sera from animals mentioned above, smaller animals offer advantages where relatively small volumes of serum are required. For most immunization regimens, the immunogen (at around 2 mg/mL) in an isotonic salt solution is mixed vigorously with an equal volume of Freund’s complete or incomplete adjuvant, to obtain an emulsion that is stable in water. It is essential that the antigen be added in a small aliquot in a stepwise fashion to the adjuvant, with vigorous mixing between each addition (e.g., using a vortex mixer). On complete addition of the antigen, the emulsion must be tested for stability. This is easily done by placing a drop of the emulsion onto the surface of some distilled water in a beaker. This should spread out over the surface. However, a second drop added (or sometimes the third) should not spread and remain as a distinct drop. The edges of the drop should show no signs of dissolution. 6.2. Improving Antigenicity of Antigens Where there is no response in animals after multiple injections, alternative animal species should be tried and the dose of antigen(s) increased. If this does not succeed, then attempts to enhance the antigenicity by direct modification methods can be tried. An excellent practical description of these techniques is found in ref. 7. Common methods include: 1. 2. 3. 4.

The addition of small modifying groups,suchas dinitrophenl or arsanate. Denaturationof antigenby heat treatmentand/or SDS treatment. Coupling of antigento that are good immunogens. Coupling of antigen to small synthetic peptides that are sites for T-cell receptor-classII protein binding. 5. Coupling of antigensto large particles, such as sheepred blood cells or agarosebeads. 6. Purification of antigens with other antibodies and injection of immune complexes. References 1. Harlow, E. andLane,D. (eds.)(1988)Antibodies. A Laboratory Manual. Cold Spring Harbor Laboratory,

Cold Spring Harbor, NY.

218

Immunochemical

Techniques

2. Avrameas, S. (1972) Enzyme markers: their linkage with proteins and use in immunhistochemrstry. Histochem. J. 47,321-330. 3. Farr and Nakane, P. K. (1974) Immunochemistry and enzyme labeled antibodies: a brief review. J. Immunol. Meth. 47, 129-144. 4. Nakane, P. K. and Kawaoi, A. (1974) Peroxidase-labelled antibody. A new method of conjugation J. Histochem. Cytochem. 22,1084-1091. 5. Tijssen, P. and Kurstak, E. (1984) Highly efficient and simple methods for the preparation of peroxidase and active peroxidase-antibody conjugates for enzyme immunoassays. Anal. Biochem. 136,451-457. 6. Avrameas, S. and Ternynck, T. (1969) The crosslinking of proteins with gluteraldehyde and its use for the preparation of immunoadsorbents. Immunochemistry 6,53-66. 7. Avrameas, S. and Uriel, J. (1966) Methode de marquage d’antigen et anticorps avec des enzymes et son application en immunodiffusron. Cr. Acad Sci. D 262, 2543-2545. 8. Immunogens, Ag/Ab Pur@cation, Antibodies, Avidin-Biotin, Protein Modifzcatzon: PIERCE Immunotechnology Catalog and Handbook, vol. 1. Pierce and Warriner, UK Published yearly.