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 The addition of the test sampleand subsequentreagents The incubation of reactants The separationof bound and free reactantsby washing The addition of enzyme-labeledreagent The addition of enzymedetectionsystem(color development) The visual or spectrophotometricreading of the assay 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: The diffusion coefficient of the attachingmolecule The ratio of the surface areabeing coated to the volume of the coating solution The concentrationof the substancebeing adsorbed The temperature 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 illustrates the possible effects on assays Solid-Phase 65 I Fig 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 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 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 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 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 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 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 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 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 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 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 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 or 12 Various reservoirs with or 12 channels for separation of reagents are also available These are useful for the simultaneous addition of separatereagents 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 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 Conjugatesmust be titrated to optimum conditions and not usedin excess This is vital to obtaining reliable results 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 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: Simplicity and rapidity; Reproducibility (obtaining constantmolar ratio of enzyme and reagent); High yield of labeled reagent,and low yield of polymers of enzyme and reagent; Low-grade inactivation of reagentand enzyme; Simple proceduresfor separationof labeled and unlabeledreagents;and Long-termstability without loss of immunological andenzymaticactivities 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; Reaction temperature; Substrate and/or cofactor concentration and stability, Product stability; Enzyme stability; and 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 1M citrate than 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 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 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 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, 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 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 If a particular plate is recommended,thenusethat plate unlessyou retitrate given reagentsin anothermanufacturer’splate 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 Always report which plate and what treatment of the plate have been made 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) Use specific troughs for conjugate and substrate only to avoid crosscontamination 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 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 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: Vicosity effects 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 Incubate at 37OC Always use the same procedure for addition of reagents, i.e., 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 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) Store at recommended temperatures Never store excessively diluted conjugate for use at some later time Always make up the working dilution of conjugate Just before you need it Always use clean tips, preferably previously unused, to dispense conjugates, If the recommended dilution or titrated dilution of conjugate is very high (e.g., l/10,000), to make 10 mL at working strength, 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 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: 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) Causing frothing on addition of samples Lack of concentration when adding large number of samples, causing missed wells and duplication of samples in the same well Poorly maintained pipets and tips 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: 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 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 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 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 Cables that connect computers to spectrophotometers, printers that not work, and so forth These are general hardware problems that have to be conquered Practical 95 Problems Table Problem Solving in ELISA Problem No color even after 30 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 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 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 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 Kurstak, E., Tijssen, P., Kurstak, C , and Morisset, R (1986) Enzyme immunoassay in diagnostic medical virology Bull WHO 64(3), 465479 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 Rembaum, A., Margel, S., and Levy, A (1978) J Immunol Meth 24,239 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 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 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 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 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 (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 [...]... 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... detection of type A influenza virus antigen in clinical specimens J Clin Microbial 17 ,30 5 -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... 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... 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... 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... 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... 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. .. 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... 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... enzyme immunoassay of anti-insulin IgG in guinea pig serum with less non-specific binding of normal guinea pig serum J Biochem 98 ,37 9 -38 4 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, 1 138 -1141 8 Husby, S., Holmskov-Neilsen,