ELISA in small animal hosts, rodents, and birds

ELISA in small animal hosts, rodents, and birds

B2 ELISA IN SMALL ANIMAL HOSTS, RODENTS, AND BIRDS M. B. Brown^ J. M. Bradbury^ and J. K. Davis General Introduction Basic ELISA Technology Althoug...

638KB Sizes 1 Downloads 14 Views

B2 ELISA IN SMALL ANIMAL HOSTS, RODENTS, AND BIRDS M. B. Brown^ J. M. Bradbury^ and J. K. Davis

General Introduction

Basic ELISA Technology

Although enzyme-Hnked immunoassay (ELISA) has been developed for both antigen and antibody detection, this chapter is restricted to applications for detecting a specific antibody. In its simplest form, the ELISA is performed by allowing the mycoplasma antigen to absorb nonspecifically to a solid phase (usually a 96-well microtiter plate), followed by the blocking of unbound protein sites and the sequential addition (with adequate washing between reagents) of the serum sample, the secondary conjugated antibody, and the enzyme substrate. The assay is completed by developing the colored product of the enzyme reaction, and reactions are read either on a spectrophotometer or by visual inspection. Although the "kit" forms of mycoplasmal ELIS As are usually interpreted subjectively, the use of semiquantitative data obtained by determining the intensity of the color change spectrophotometrically provides better discrimination of positive and negative results. The latter application is especially important in monitoring barrier-maintained animal colonies. A semiquantitative assay is relatively inexpensive and can easily be established in any laboratory with a spectrophotometer. Recent modifications of the ELISA technique include dot-blot ELISA proce93 Molecular and Diagnostic Procedures in Mycoplasmology, Vol. II

Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

94

M. B. Brown et al.

dures which rely on nitrocellulose or other membranes as the solid support. Results are usually recorded visually, although scanning densitometers can provide more quantitative data. Considerable variation in ELISA techniques exists among laboratories working with moUicutes, including differences in mollicute strains selected as antigens, antigen preparation methods, solid-phase supports, reagents, and incubation times and temperatures. All of these factors affect the test and can lead to a lack of agreement in test results among laboratories. In most cases, high quality, commercially available, conjugated secondary antibodies, substrates, and developing reagents have eliminated the need for preparation of these reagents. Availability of standard reference serum reagents for quality control purposes could also reduce variability among laboratories. This is especially important when the ELISA is used for the clinical diagnosis of mycoplasmosis. Use of ELISA in Clinical and Experimental Laboratories

ELISA methodology for the detection of specific antibody to Mycoplasma and Ureaplasma species has proven to be a useful tool in the clinical diagnostic laboratory as well as the research laboratory. In a clinical setting, antibody detection is most often used to confirm exposure to the infectious agent. This is especially true in both poultry and laboratory rodent populations where the establishment of mycoplasma-free populations is a priority. Since individual animals are rarely followed in rodents or poultry, except under experimental circumstances, monitoring of acute and convalescent titers in individual animals is rare and of limited use clinically. However, development of specific antibody can be used as a tool for seroepidemiology in monitoring the spread of disease in populations and in providing additional understanding of the dynamics involved in the interaction between the host and the infectious agent. Changes in antibody levels could precede the appearance of clinical disease and provide an early warning of potential disease outbreaks in various populations. This could be especially important in wild animal hosts, where little is known of the natural occurrence of mollicutes or their possible role in disease. In experimental situations, ELISA technology is particularly useful. The ELISA is flexible in that both serum and secretions can be analyzed for specific antibody. Not only can seroconversion be monitored, but immunoglobulin classspecific responses can be determined. Because of differences in biological activity, the presence or the induction of specific subclasses may be of particular interest. ELISA in Laboratory Rodent Infections

Traditional methods of serologic diagnosis of mycoplasmal infections, such as complement fixation, hemagglutination, hemagglutination inhibition, and meta-

B2 ELISA in Small Animals, Rodents, and Birds

95

bolic inhibition tests, have been of little use in laboratory rodents because they are relatively insensitive and depend on antibody function. In well-managed, barrier-maintained rodent colonies, both the prevalence of infected animals and the numbers of organisms per animal are usually low. This situation makes detection by serologic methods very difficult. ELISA for mycoplasmal infections in rodents has proven to be the most suitable serologic test for screening large numbers of rodents (Cassell et al, 1986; Davidson et al., 1981, 1994). A comprehensive review of all aspects of the diagnosis of rodent mycoplasmosis has appeared (Davidson et al., 1994). The test is relatively inexpensive to run, is sensitive, and gives rapid results. In addition, animals do not have to be killed to obtain enough blood for the test. Blood may be collected from anesthetized animals by tail or orbital bleeding. However, the assay currently available cannot distinguish among Mycoplasma species, and quality control of the assay, as with all serological assays, is critical for reliable results. ELISA in Poultry and Other Avian Infections The most common method of diagnosis of either M. gallisepticum or M. synoviae infection in poultry is determination of antibody status. The serum plate agglutination test is commonly used to screen for mycoplasma infections but has been associated with false positives. In the United States, both hemagglutination inhibition and ELISA are approved as confirmatory tests by the National Poultry Improvement Program, but their use also must be specifically approved by each state. Several commercial ELISA kits are available on the market for detecting antibodies to M. gallisepticum and M. synoviae in chicken sera. At least one company markets such kits specifically for use with turkey sera, as well as a kit for detecting M. meleagridis antibodies in turkeys (KPL Inc., Gaithersburg, MD). One commercial kit is designed to react with antibodies to either M. gallisepticum ox M. synoviae (IDEXX Laboratories Inc., Westbrook, ME). Recent advancements in the preparation of ELISA antigens, with removal of nonspecific antigenic components, have improved specificity and decreased cross-reactions. ELISA tests using affinity-purified antibodies are available with good results for both M. gallisepticum and M. synoviae. However, there is no suitable ELISA for detecting M. iowae, an organism found mainly in turkeys. This mycoplasma does not appear to elicit a consistent humoral antibody response, and experimental studies have shown an unacceptably high number of falsepositive reactions in normal turkey sera (Jordan et al., 1987). A "blocking ELISA" has been proposed for detecting M. gallisepticum infection (Czifra et al., 1993). The test uses a monoclonal antibody (MAb) which recognizes an epitope on a 56-kDa polypeptide that appears to be permanently expressed on the mycoplasma. In this system, ELISA plates are coated with M. gallisepticum whole cell antigen, and the serum under test is added undiluted

96

M. B. Brown et al,

(Diagnosticum, Budapest, Hungary; Svanova Biotech, Uppsala, Sweden). The reaction is assessed by the extent of blocking that occurs when the conjugated MAb is added. One assay can be used for sera from any host species without adaptation as long as the relevant antigen is recognized by the individual animal. ELISAs have also been used successfully to detect M. gallisepticum antibodies in diluted or extracted egg yolk (Mohammed et al., 1986; Brown et al., 1991). When the currently available commercial M. gallisepticum ELISAs were compared with other commonly used serological tests, i.e., the rapid serum agglutination (RSA) and hemagglutination inhibition (HI) tests, the RSA detected antibodies slightly earlier than the ELISA or HI tests (Kempf et al., 1994; J. M. Bradbury and J. Lewis, unpublished observations). However, some poultry companies find ELISAs more convenient for the routine screening of large numbers of birds because they already employ ELISA technology in screening for the many important poultry viruses. As with other mycoplasma ELISAs, the sensitivity of the avian mycoplasma ELISAs is determined to some extent by the recommended cutoff levels for positive and suspicious reactions. The sensitivity in some cases may be deliberately "dampened down" to avoid the well-recognized cross-reaction between M. gallisepticum and M. synoviae. This is a dangerous practice as it inherently implies the assay conditions have not been optimized and low level infections may be missed. The development of ELISA antibodies may be suppressed if the birds have received antibiotic treatment. In addition, a serological response will not differentiate birds that have been vaccinated against mycoplasmal infection from those naturally infected. Also, it should be noted that polymerase chain reaction (PCR)-based tests have been developed and these should be evaluated as well as serological tests for their efficacy in detecting infection (see Chapter D9, this volume). EUSA in Other Small Animals

For the most part, serological assays have been limited to poultry and rodents, although mycoplasmal infections in populations of environmentally threatened wild tortoise populations have been examined using ELISA (Schumacher et al., 1993). In addition, antibody to both M. gallisepticum and M. synoviae has been detected in sera from golden conures (psittacine) located in a private zoological collection (M. B. Brown, unpublished data). The major deterrent to the development of ELISAs for other animals is the availability of a specific antibody which recognizes the animal species. The competitive (blocking) ELISA could be used with any species, but it is imperative that the antigen(s) or epitopes recognized by the control serum are also recognized by all infected animals. Commercial conjugates generally are available for most domestic species. Conjugates for exotic animals may be prepared commercially by specialized companies or by university facilities, such as the Biotechnologies for the Ecological,

B2 ELISA in Small Animals, Rodents, and Birds

97

Evolutionary, and Conservation Sciences Immunological Analysis Laboratory at the University of Florida. Alternatively, a blocking ELISA can be developed.

Antigen Preparation Antigen preparation is a critical step for the success of any ELISA. Antigen preparations may consist of whole cells, lysates, partially purified protein preparations, purified proteins, or even recombinant proteins. Regardless of how the antigen is prepared, certain requirements should be applied. First, the proteins present in the antigen preparation should be present on all (or at least most) field strains of the mycoplasmal species. The preparative steps should be standardized to assure batch to batch reproducibility of the antigen. Given the documented antigenic variation of many mollicutes, it is imperative that purified proteins or recombinant proteins used as antigens should be constitutively expressed and recognized by infected animals at all stages of the infection. If purified proteins are used as antigens, it is important to use a mixture of proteins rather than reliance on a single protein. One of the most common antigen preparations used is a lysate of whole mycoplasma cells to produce the antigen for the assay (Horowitz and Cassell, 1978). This method of antigen preparation is used because it is easily accomplished and yields a reliable, stable antigen for routine use. This general method may be used for most mollicutes, but the harvesting times and protein yield vary considerably from mollicute species to species. Antigens prepared by this method are stable at -70°C for at least 6 months. The effect of storage conditions on antigenicity must be assessed for each antigen preparation and for each mollicute species. Several other ELISA antigen preparation methods have been reported and are described elsewhere in this volume (see Chapter B3). The following technique will usually provide a satisfactory antigen. 1. Grow the test organism in 1-3 liters of broth medium until the culture reaches the peak logarithmic phase. Harvest the organisms by centrifugation (10,000 to 20,000 g), 2. Wash the organisms three times in sterile phosphate-buffered saline (PBS, pH 7.4) by adding PBS at a ratio of 10 times the volume of the pellet, mixing well, and then centrifuging the suspension. Pour off the supernatant. Repeat the wash once again. Resuspend the pellet in a small volume (1 to 10 ml) of PBS. 3. Transfer a small amount of the final suspension into beef heart infusion broth (or other suitable medium) to check for bacterial contamination. Note that some mycoplasmal species will sometimes grow on 5% sheep blood agar plates. 4. Determine the protein concentration of the final suspension. 5. Dilute the suspension in sterile PBS to a final concentration of 5 mg protein/ml. Store aliquots of the suspension at -70°C.

98

M. B. Brown et al,

6. Dilute a small amount of the 5-mg/ml suspension to 1:100 (0.05 ml of suspension to 0.95 ml PBS) and read absorbance at 540 nm. 7. Dilute the remainder of the suspension 1:20 in carbonate-bicarbonate buffer (0.05 M, pH 10.0). Incubate the diluted suspension at 37°C. 8. Check the absorbance of the diluted suspension at 5-minute intervals until the absorbance of the suspension is 50% of the absorbance of the PBS suspension in step 6. This usually takes about 15 minutes. 9. Stop the lysis by adding 2.2 g of boric acid per 100 ml of suspension. 10. Determine the protein concentration of the lysate, and store aliquots at -TOT.

Conjugate and Substrate The availability of high quality conjugated antisera has increased dramatically since the 1980s. Comprehensive listings of anti-immunoglobulins are available in publications such as "Linscott's Directory." If the host species is not commercially available, many private companies or university core facilities can prepare custom antisera at a nominal charge. The most common enzymes used for conjugation to anti-immunoglobulins are alkaline phosphatase and horseradish peroxidase; biotinylation also is common. Although less common in ELISA, additional enzymes including urease and 3-galactosidase or fluorochromes can be used. Most antiimmunoglobulins, especially IgG, are available already labeled, and subclassspecific antibodies are readily available for rodent immunoglobulins. If not available, several methods and commercial labeling kits are available (Harlow and Lane, 1988). In addition to specific anti-immunoglobulins, some assays use labeled protein G or protein A to detect bound immunoglobulins. It should be noted that these proteins may vary in their ability to bind IgG from different host species and the binding capacity for the host IgG of interest should be established if this choice is made. A variety of substrates are available for the enzymes just discussed. For the purposes of this discussion, only soluble substrates will be considered. Insoluble substrates are available for dot-blot or other detection systems which immobilize antigens on nitrocellulose or polyvinyl membranes. Most substrates and their buffers are available from commercial sources. For alkaline phosphatase, the substrate of choice is p-nitrophenyl phosphate (PNPP) which yields a yellow solution. The reaction is stopped by the addition of 3 N NaOH, and the test is read at 405 nm. The substrate buffer is described elsewhere in this volume (see Chapter B5). For horseradish peroxidase, the substrates of choice are 2,2'azinobis (3-ethylbenzothiazoline-6-sulfonic acid (ABTS), o-phenylenediamine

B2 ELISA in Small Animals, Rodents, and Birds

99

dihydrochloride (OPD), or 5-aminosalicylic acid (5AS). ABTS yields a bluegreen color, with the reaction stopped by the addition of 1% sodium dodecyl sulfate and the test read at 405 nm. OPD yields an orange-brown product and is read at 492 nm; sensitivity increases two- to fourfold if the reaction is stopped with 3 N HCl or 3 M H2SO4 and the test is read at 450 nm. 5AS yields a brown color which is read at 450 or 550 nm, after the reaction is stopped with 3 A^ NaOH.

Assay Standardization In developing an ELISA there are numerous factors that can affect results and their interpretation. The first decision is whether or not a single dilution, or multiple dilutions, of the specimen should be tested. Results may be reported as absorbance, end point titration, significant rises in titer, activity as compared to some "standard" serum, or simply positive, suspect, or negative when compared to a known positive or negative serum sample. A single serum sample often is adequate for diagnostic purposes. Because single dilution ELISA assays are heavily dependent on antibody affinity, quantitative standardization in terms of milligrams of antibody present is almost impossible. Although serum is the most common specimen examined, egg yolks have been effectively used in avian and reptile systems (Mohammed et aL, 1986; Brown et al., 1991; M. B. Brown and I. M. Schumacher, unpublished data). As a screening tool in the poultry industry, egg yolk testing may offer an advantage under some circumstances because the eggs can be readily collected, thus avoiding the time-consuming and invasive procedures of drawing blood. Lavages from respiratory and genital tracts have been used, but these are used most often in experimental situations where serum antibody levels are also examined (Simecka and Cassell, 1987; Brown and Reyes, 1991; Elfaki et aL, 1992). In selecting the dilutions of antibody and reagents, it is important that the amount of antiserum in the unknown test sample should be the only limiting factor. Adjusting the concentration of antigen, serum dilution, and conjugate dilution by simultaneous checkerboard titration is likely to cause trouble and often leads to a choice of conditions where some factor besides the specific antibody in the test serum is limiting. Choosing Plates The first step in establishing the assay is choosing an acceptable solid-phase support, which is usually a microtiter plate. The mean, standard deviation.

100

M. B. Brown et aL

coefficient of variation, and ratio of immune to normal serum absorbance of the colored product of the enzyme reaction should be determined for each sample. The coefficient of variation should be as small as possible, and the ratio of the immune to normal serum should be as large as possible. Most plates show a significant "edge effect," and avoidance of the outside rows of wells for the test is advisable. The criteria used for selecting a microtiter plate are a coefficient of variation less than 0.5 to 0.75 for both immune and normal serum and the highest possible ratio between hyperimmune and normal serum. Blocking Solutions In almost all ELISA assays, unbound sites that can nonspecifically absorb either immunoglobulin in the sample or conjugated secondary antibody remain on the microtiter plates. These unbound sites should be blocked by inert protein. A variety of blocking solutions have been used, including Tween 20, BLOTTO, 2% w/v bovine serum albumin (BSA), 5% w/v BSA, 2% v/v fetal calf serum (PCS), and 5% v/v PCS. The ability to block both the binding of normal serum and the binding of the conjugated secondary antibody to the microtiter wells should be assessed. The choice of the blocking reagent to be used must be examined for each microtiter plate and for each antigen preparation. Determining Antigen Concentration Dilutions of antigen, test serum, and conjugate can be determined as follows. Arbitrarily select a dilution of a known immune serum and the conjugated secondary antibody that will give a high concentration of these two reagents so that they cannot be the limiting reagents in the assay. Coat the microtiter plates with various concentrations (0.1 to 50 |JLg/ml of protein) of mycoplasma antigen. Run the assay in the routine manner. Plot the absorbance versus the antigen concentration. Pick an antigen concentration that is well on the plateau of the absorbance curve for all future tests. A protein concentration of 2-10 |xg/ml is usually adequate to ensure that the antigen is never the limiting factor in the assay. Determining Conjugate Concentration The working concentration of each conjugate is determined by reacting antigen-coated wells with PBS and then using dilutions (1:500 to 1:10,000) of the affinity-purified conjugates in the assay. Plot the absorbance versus the dilution of the conjugate and choose the lowest dilution that does not show appreciable nonspecific binding to the antigen. As a general rule, the conjugate dilution should be less than 1:5000 to ensure that the conjugate is never the limiting factor in the assay.

B2 ELISA in Small Animals, Rodents, and Birds

101

Determining Serum Dilution

The serum dilution to use is the last of these parameters to be determined because the working concentrations of antigen and conjugate and which microtiter plate and blocking reagent works best should already be established. If multiple dilutions and end point titration are used, then this is not necessary. The plates are coated with antigen, blocked, and reacted with different concentrations of serum (dilutions range from 1:10 to 1:1500; lavages or egg yolk may be used as well), followed by reaction with the conjugate. Plot the absorbance vs the serum dilution. Choose the lowest dilution that shows little nonspecific absorption to the plates. Sometimes a compromise must be reached at this step because a very low nonspecific absorbance and a low serum dilution are both desirable. The former is necessary for maximum sensitivity, but the latter is required to allow the detection of small amounts of antibody. As a final check of the reaction reagents, the conjugate titration should be run again with normal and immune sera as well as the PBS control. When the absorbance is plotted versus the conjugate dilution, the normal serum and PBS control curves should be virtually identical. There should also be a wide separation between the values of the immune serum and the normal serum at the conjugate dilution selected for use in the assay. Incubation Times

A wide variety of incubation conditions have been reported for ELISAs. In some cases, incubation times and temperatures are adjusted for convenience (i.e., overnight incubations or room temperature vs 37°C). Ideally, the conditions for each incubation step should be determined by similar methods to those used to determine which microtiter plates were acceptable for use in the assay. The mean, standard deviation, coefficient of variation, and ratio of the absorbance of the immune to normal serum are determined for each incubation procedure. The coefficient of variation should be as small as possible, indicating reproducibility. Likewise, the ratio of immune to normal serum absorbance values should be as large as possible, indicating the range of the assay and its potential sensitivity. In addition, the ratio of the normal serum to the PBS control values should be determined because this is a measure of the nonspecific activity in the system. This ratio should be as close to 1 as possible, thereby indicating no nonspecific activity.

Quality Control One of the disadvantages of the ELISA assay is its extreme sensitivity to changes in procedure. If a minor variation is inadvertently introduced into the

102

M. B. Brown et aL

procedure at any one point, this variation may be amplified in the test results. Therefore, stringent quality control measures are required. The Youden plot has been proven to be very useful (Jeffcoate, 1982). This method requires the use of dilutions of immune serum simulating high, mid-range, and low concentrations of specific antibody to the antigen in question. To determine within run and between run variability in the assay, one sample of each dilution is tested at the beginning of a set of tests and another is included at the end of the assay. The absorbance values obtained on these control samples are plotted against each other, resulting in a linear plot with a slope of 1 at a 45° angle from the X and Y axes. Variability within an assay is indicated by the spread of points away from the 45° line. The maximum divergence from the line should be no more than a 10% deviation. Day to day variation is indicated by the spread of values away from the averages of the last 10 assay runs. If a value for an individual day deviates by more than 10% from the average of the values for the last 10 runs, the run should be repeated. Other indications that a run needs to be repeated are when all three samples fall on one side of the line and are outside of a 5% deviation range.

Limitations and Other Considerations ELISA has many advantages, but it depends on the ability of the animal to produce a specific antibody. It is important to understand the natural history of the host immune response to assess the efficacy of ELISA. For example, in rodents there is a delay in antibody production during subclinical infections, resulting in a 1- to 3-month "lag time" when infected animals cannot be detected by serologic methods (Cassell et aL, 1986). However, on farms with chickens of various ages, the number of seropositive birds increases exponentially with the age of the bird, suggesting a rapid seroconversion to avian mycoplasmas (Brown etal. 1991). In rodents, a crucial problem is cross-reactions among the murine mycoplasmas (Minion et aL, 1984; Cassell et aL, 1986; Davis et aL, 1986; Watson et aLy 1987). All of the murine mycoplasmas share some common antigens; these appear to be particularly troublesome in trying to discriminate M. pulmonis infection from M. arthritidis or M. muris infection (Cassell et aL, 1986; Watson et aL, 1987; Davidson et aL, 1994). In poultry, the inability to differentiate between infected birds and vaccinated birds may also be troublesome. As ELISAs are developed for mycoplasmal infections in other hosts (i.e., cats, dogs, tortoises), the potential for cross-reactive antigens should be of some concern. Such antigens may also pose a particular problem in ELISA tests on exotic animals and other hosts with less well-defmed mycoplasmal flora.

B2 ELISA in Small Animals, Rodents, and Birds

1 03

Optimum serologic detection of mycoplasmal infections depends on the sensitivity and specificity of the test system and on the interpretation of the results obtained. Technological advancements now suggest the possibility that serologic tests may eventually be replaced by a variety of antigen detection methods, including DNA hybridization methods with specific gene probes or the polymerase chain reaction (Kleven et aL, 1988; Lauerman et aL, 1993). Ultimately, stringent field testing combined with careful experimental infection studies will be required to define the reliability of such diagnostic procedures.

References Brown, M. B., and Reyes, L. (1991). Immunoglobulin class- and subclass-specific responses to Mycoplasma pulmonis in serum and secretions of naturally infected Sprague-Dawley female rats. Infect. Immun. 59, 2181-2185. Brown, M. B., Stoll, M. L., Scasserra, A. E., and Butcher, G. D. (1991). Detection of antibodies to Mycoplasma gallisepticum in egg yolk versus serum samples. J. Clin. Microbiol. 29, 29012903. Cassell, G. H., Davis, J. K., Cox, N. R., Brown, M. B., and Minion, F. C. (1986). State of the art detection methods for rodent mycoplasmas. In "Complications of Viral and Mycoplasmal Infections in Rodents to Toxicology Research and Testing" (T. E. Hamm, ed.), pp. 143-160. Hemisphere Press, Washington, DC. Czifra, G., Sundquist, B., Tuboly, T., and Stipkovits, L. (1993). Evaluation of a monoclonal blocking enzyme-linked immunosorbent assay for the detection of Mycoplasma gallisepticumspecific antibodies. Avian Dis. 37, 680-688. Davidson, M. K., Lindsey, J. R., Brown, M. B., Schoeb, T. R., and Cassell, G. H. (1981). Comparison of methods for detection of Mycoplasma pulmonis in experimentally and naturally infected rats. J. Clin. Microbiol. 14, 646-655. Davidson, M. K., Davis, J. K., Gambill, G. P., Cassell, G. H., and Lindsey, J. R. (1994). Mycoplasmas of laboratory rodents. In "Mycoplasmosis in Animals: Laboratory Diagnosis" (H. W. Whitford, R. F. Rosenbusch, and L. H. Lauerman, eds.), pp. 97-133. Iowa State Univ. Press, Ames. Davis, J. K., Cassell, G. H., Gambill, G., Cox, N., Watson, H., and Davidson, M. (1986). Diagnosis of murine mycoplasmal infections by enzyme linked immunosorbent assay (ELISA). Isr. J. Med. Sci. 23, 717-722. Elfaki, M. G., Ware, G. O., Kleven, S. H., and Ragland, W. L. (1992). An enzyme-linked immunosorbent assay for the detection of specific IgG antibody to Mycoplasma gallisepticum in sera and tracheobronchial washes. J. Immunoassay 13, 97-126. Harlow, E., and Lane, D. (1988). "Antibodies: A Laboratory Manual," pp. 319-358. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Horowitz, S. A., and Cassell, G. H. (1978). Detection of antibodies to Mycoplasma pulmonis by an enzyme-linked immunosorbent assay. Infect. Immun. 22, 161-170. Jeffcoate, S. L. (1982). Use of Youden plot for internal quality control on the immunoassay laboratory. Ann. Clin. Biochem. 19, 435-437. Jordan, F. T. W., Yavari, D., and Knight, D. L. (1987). Some observations on the indirect ELISA for antibodies to Mycoplasma iowae serovar I in sera from turkeys considered to be free from mycoplasma infections. Avian Pathol. 16, 307-318.

104

M . B. Brown et al.

Kempf, I., Gesbert, F., Guittet, M., Bennejean, G., and Stipkovits, L. (1994). Evaluation of two commercial enzyme-linked immunosorbent assay kits for the detection of Mycoplasma gallisepticum antibodies. Avian Pathol. 23, 329-338. Kleven, S. H., Morrow, C. J., and Whithear, K. G. (1988). Comparison of Mycoplasma gallisepticum strains by hemagglutination-inhibition and restriction endonuclease analysis. Avian Dis. 32, 731-741. Lauerman, L. H., Hoerr, F. J., Sharpton, A. R., Shah, S. M., and van Santen, V. L. (1993). Development and application of a polymerase chain reaction assay for Mycoplasma synoviae. Avian Dis. 37, 829-834. Minion, F. C., Brown, M. B., and Cassell, G. H. (1984). Identification of cross-reactive antigens between Mycoplasma pulmonis and Mycoplasma arthritidis. Infect. Immun. 43, 115-121. Mohammed, H. O., Yamamoto, R., Carpenter, T. E., and Ortmayer, H. B. (1986). Comparison of egg yolk and serum for the detection of Mycoplasma gallisepticum and M. synoviae antibodies by enzyme-linked immunosorbent assay. Avian Dis. 30, 398-408. Schumacher, I. M., Brown, M. B., Jacobson, E. R., Collins, B. R., and Klein, P. A. (1993). Detection of antibodies to a pathogenic mycoplasma in desert tortoises (Gopherus agassizii) with upper respiratory tract disease (URTD). J. Clin. Microbiol. 31, 1454-1460. Simecka, J. W., and Cassell, G. H. (1987). Serum antibody and cellular responses in LEW and F344 rats after immunization with Mycoplasma pulmonis antigens. Infect. Immun. 55, 731-735. Watson, H. L., Cox, N. R., Davidson, M. K., Blalock, D. K., Davis, J. K., Dybvig, K., Horowitz, S. A., and Cassell, G. H. (1987). Mycoplasma pulmonis proteins common to the murine mycoplasmas. Isr. J. Med. Sci. 23, 442-447.