ELISA Using Synthetic Peptides

Several ELISA formats can be used with peptides to delineate linear and conformational epitopes of a protein [51-53]. In a classical approach, the capacity of a candidate peptide tested in the liquid-phase to inhibit the interaction between a protein adsorbed on the plastic surface of a microtiter plate and antibodies contained in the patients serum is measured. Bound antibodies are subsequently revealed by adding a second antibody against human IgG and/or IgM conjugated to an enzyme, generally horseradish peroxidase, alkaline phosphatase, or -galactosidase. Finally, the absorbance of colored product is measured. It is important to understand that this type of immunoassay will select autoantibodies reacting with the solid-phase immobilized parent protein only, to the exclusion of any other antibody subsets, which may be important for diagnosis. Also, it is important to bear in mind that this approach reveals only the epitopes presented by the immobilized protein and mimicked by the peptide in solution.

Alternatively, an indirect assay can be set up, wherein peptides are directly adsorbed to the wells of a microtiter plate, serial dilutions of patients' sera are incubated in the peptide-coated wells, and bound antibodies are revealed by a second antibody against human Ig linked to a selected enzyme. For B-cell epitope mapping, a complete set of overlapping peptides covering the whole length of the protein is used. This strategy has been applied, for example, to histones [54-56]; 52-kD SSA/Ro protein (Ro52, [57]); D1 protein of Sm antigen (SmD1, [58-60]); A, C, and 70K proteins of the U1 small nuclear ribonucleoprotein (snRNP) antigen (U1A, U1C, U1-70K; [61-63]); hnRNP A2/B1 protein [64]; and human thyroprotein receptor [65]. This ELISA format presents numerous advantages insofar as it is simple, fast, and easily automatable, but it also has a number of conceptual limitations directly related to peptide adsorption to plastic. Direct adsorption of a peptide can affect its conformation and mask a portion of its surface. Furthermore, the efficacy of adsorption of a peptide can vary widely according to its length, charge, and solubility; the pH and composition of coating buffer; and the type of plastic of the microtiter plates (polyvinyl, polystyrene with or without plastic activation) (Fig. 9.2; Dali and Muller, unpublished). Based on our experience, peptides that comprise at least 15 residues are

Fig. 9.2 Relative coating efficacy of various synthetic peptides on polystyrene microtiter plates. Biotinylated peptides of different lengths and charges were allowed to adsorb onto polystyrene microtiter plates (MaxiSorp, Nunc, Roskilde, Denmark) overnight at 37°C. For coating the plates, peptides were suspended in 0.1 M carbonate buffer pH 9.6 at a concentration of 0.5 ^M (black) and 2 ^M

(gray). Adsorption of biotinylated peptides was evaluated by incubating peptide-coated plates with streptavidin conjugated with peroxidase. The final reaction was visualized by adding 3,3',5,5' tetramethylbenzidine in the presence of H2O2. The peptides are arranged by length in the figure (the shortest are on the left and the longest are on the right) (Dali and Muller, unpublished data).

Fig. 9.2 Relative coating efficacy of various synthetic peptides on polystyrene microtiter plates. Biotinylated peptides of different lengths and charges were allowed to adsorb onto polystyrene microtiter plates (MaxiSorp, Nunc, Roskilde, Denmark) overnight at 37°C. For coating the plates, peptides were suspended in 0.1 M carbonate buffer pH 9.6 at a concentration of 0.5 ^M (black) and 2 ^M

(gray). Adsorption of biotinylated peptides was evaluated by incubating peptide-coated plates with streptavidin conjugated with peroxidase. The final reaction was visualized by adding 3,3',5,5' tetramethylbenzidine in the presence of H2O2. The peptides are arranged by length in the figure (the shortest are on the left and the longest are on the right) (Dali and Muller, unpublished data).

in general efficiently adsorbed to plastic after incubating the peptide solution in the wells overnight, at room temperature or 4 °C. However, even for experts, it remains difficult to predict whether a peptide will be a "good" or a "moderate" plastic binder, and when a careful internal calibration of the tests has not been performed, it is often impossible to compare quantitatively the various absor-bance values measured with a panel of peptides.

To overcome such drawbacks, some authors (see, e.g., [66]) recommended the use of peptides conjugated to a carrier protein (bovine serum albumin, ovalbumin) to increase the coating efficiency and diminish the amount of peptide used per assay. Others proposed the use of multi-presentation systems such as MAPs, TASPs, and SOCn (see Sections 9.3.4 and 9.5) to avoid coupling the peptides with a carrier. When unconjugated peptides are used, their effective binding to the plate must be checked, e.g., by using anti-peptide antibodies or by incubating with enzyme-conjugated streptavidin when biotinylated peptides are used (Fig. 9.2). Likewise, when peptide conjugates are used as antigens, the yield of effective peptide coupling must be controlled and the stability of conjugates checked over time [67, 68]. Moreover, the choice of a conjugation procedure is crucial since the antigenic activity of a peptide can be dramatically affected by different coupling procedures [68, 69]. Consequently, the chemical agent used to conjugate the peptide with an appropriate carrier must be carefully chosen according to each specific sequence.

To ensure reliability of peptide-based ELISAs used to monitor autoimmune sera, several important points must be routinely checked:

• Unrelated peptides or, preferably, scrambled peptides containing the same amino acid residues in a different order compared to the parent sequence should always be tested in parallel as controls; because patients' sera can contain a diversity of autoantibodies whose range of specificities is not known with certainty, control peptides need to be selected with care.

• When conjugated peptides are used, it is necessary to verify the absence of reaction with the carrier or, preferably, with the carrier presenting the control peptide; false reactions have been found with patients antibodies cross-reacting, for instance, with bovine serum albumin or ovalbumin.

• The absence of a reaction of enzyme-labeled second antibodies with peptides in the absence of antibodies must be verified.

• A large number of sera from normal donors must be tested in the same ELISA conditions to define the cutoff value for positivity for each peptide. It is known that certain classes of autoantibodies are surprisingly common in the normal population; for example, low levels of anti-SSA/Ro antibodies are present in 5-15% of the normal population [70].

• In isotyping studies, particular attention should be paid to the specificity of secondary enzyme-labeled antibodies with regard to their ability to reveal equally well all IgG subclasses, as certain minor subclasses can be increased in autoimmune situations, e.g., peptide-reacting antibodies of the IgG3 subclass in lupus mice [71], and are poorly detected with most commercial enzyme-labeled anti-IgG antibodies [72].

• Finally, it should be stressed that several factors can generate high ELISA background. False positivity may be due to aggregates present in sera from collections that have been stored for a long time, even in good conditions. Many laboratories routinely heat sera to 56 °C, which can cause major problems in nonspecific binding in ELISAs due to the presence of aggregates in sera. False positivity can also result from cross-reaction of antibodies with nonfat bovine milk or normal serum used as blocking agents to prevent nonspecific adsorption of proteins to wells. These added reagents can contain self-antigens such as DNA and histones [73]. Bovine IgG can contaminate bovine serum albumin [51]. Patients' sera can also contain circulating self-antigens. For example, the presence of nucleosomes and proteinase-3 (a neutro-phil primary granule that is recognized by cytoplasmic anti-neutrophil cyto-plasmic antibodies, cANCA) in the serum of normal and/or autoimmune patients has been demonstrated. Rheumatoid factors also can cause a high level of false positives. Finally, anti-albumin autoantibodies have been detected in the serum of autoimmune and infected patients [74, 75]. Many of these recommendations are also useful when other types of solid-phase immuno-assays are utilized.

0 0

Post a comment