Vol. 22, No. 1
INFECTION AND IMMUNITY, Oct. 1978, p. 161-170
0019-9567/78/0022-0161$02.00/0 Copyright i 1978 American Society for Microbiology
Printed in U.S.A.
Detection of Antibodies to Mycoplasma pulmonis by an Enzyme-Linked Immunosorbent Assay SHULAMITH A. HOROWITZ AND GAIL H. CASSELL* Department ofMicrobiology, University of Alabama in Birmingham, Birmingham, Alabama 35294 Received for publication 4 August 1978
The enzyme-linked immunosorbent assay, which entails the use of antigencoated tubes and enzyme-labeled anti-immunoglobulins, was applied for the detection of antibodies against Mycoplasma pulmonis in mice. A lysate of M. pulmonis was used as the antigen, and anti-mycoplasmal antibodies of the different immunoglobulin classes were detected by class-specific anti-immunoglobulin labeled with alkaline phosphatase. The optimal conditions for the test were determined, the specificity was evaluated, and the assay was compared with other procedures for detection of mycoplasmal infection. The enzyme-linked immunosorbent assay was found to be a specific, highly sensitive, and reliable procedure for detecting anti-mycoplasmal antibodies in mice. Respiratory disease caused by Mycoplasma pulmonis is the most common infectious disease encountered in laboratory rats and mice. In fact, this disease greatly complicates much of the biomedical research with these animal species (15). At present there is no rapid, reliable diagnostic procedure for detection of this infection, thus limiting critical evaluation of control methods. An ideal diagnostic test would be exquisitely sensitive, yet environmentally safe, and the reagents would be inexpensive, stable, and simple to use. In addition, the method would be amenable to automation (to accomodate large numbers of specimens) and not require sacrifice of animals for definitive diagnosis. The enzymelinked immunosorbent assay (ELISA), which has been used for detecting antibodies against various other organisms (1, 4, 5, 8, 13, 23-26), appears to meet these criteria. It is a modification of the radioallergosorbent technique in which enzyme-labeled anti-immunoglobulin is used to quantitate specific antibody bound to a solid-phase antigen (9). This assay correlates significantly with other methods of antibody detection (6, 18, 21, 25), yet is much more sensitive (4-6, 8, 13, 18, 24, 25), and, furthermore, allows the class of antibody to be identified by the use of labeled anti-heavy chain antibodies (5, 8, 13, 22). The present study evaluates the efficacy of the ELISA for measuring class-specific antibodies against M. pulmonis in mice. MATERIALS AND METHODS M. pulmoni& The M. pulmonis stock culture used in this study was the same as that described previously (16). The M. arthritidis strain used was no. 19611
obtained from the American Type Culture Collection, Rockville, Md. Both species were cultivated in liquid Hayflick medium (pH 7.8) consisting of PPLO broth (Difco Laboratories) supplemented with 10% (wt/vol) yeast extract, 20% heat-inactivated filtered horse serum, 0.5% glucose, 500 U of penicillin per ml and 0.0005% thallium acetate (12). The medium for M. arthritidis was supplemented with arginine (final concentration, 0.2%). M. pulmonis antigen. An inoculum of 112 ml containing 1.8 x 108 colony-forming units (CFU) per ml of M. pulmonis stock culture was propagated in 6 liters of the above medium for approximately 72 h. When the pH of the culture had fallen to 7.2, the cells were harvested by centrifugation at 10,000 x g for 20 min at 4VC. The organisms were washed three times with sterile, phosphate-buffered saline (PBS), pH 7.3, suspended in the same buffer to give a concentration of 5 mg of protein per ml, and added to 0.05 M carbonate-bicarbonate buffer (CBB), pH 10.0, at 370C in a 1:20 proportion (organisms:CBB) (11). The optical density at 500 nm was measured to determine lysis and compared with a control consisting of cells suspended in PBS. After 5 min of incubation (at least), the lysis was terminated by adding boric acid (2.2 g/100 ml of CBB), and the pH was reduced to 7.4. A decrease of 50% in the optical density of the cell suspension was chosen as the end point of lysis. Electron micrographic analysis of the lysate showed its composition to be predominately membrane fragments. This lysate, which contained 166 ,ug of protein per ml (by Lowry protein determination), was used as the ELISA antigen. M. arthritidis antigen. An inoculum of 5 ml containing 1.8 x 108 CFU/ml was propagated in 100 ml of the above medium for 48 h. Cells were harvested, washed, and lysed by the same procedure as M. pulmonis. The lysate contained 130 jtg of protein per ml. Myeloma proteins. Purified
ji-, yr-, and y2-
heavy chains (MOPC 104-j, J588-a, MOPC 300--y,, BTC l-y2) were prepared by J. Kearney, University of 161
HOROWITZ AND CASSELL
Alabama in Birmingham (14) and were used as the antigen in class specificity experiments. Antisera. Anti-M. pulmonis hyperimmune sera were prepared as follows. Sixty pathogen-free CD-1 mice derived and maintained as described previously (15) were infected intravenously with 0.1 ml of M. pulmonis broth culture containing 108 CFU. At 21 days post-immunization, the mice were challenged intranasally with 0.5 x 108 CFU and were exsanguinated on day 28 post-immunization. Sera were pooled and used for establishing the optimal conditions for ELISA. Two pools of hyperimmune (I and II) sera were tested. Mouse anti-M. arthritidis sera known to have complement-fixing antibodies were kindly supplied by B. C. Cole, University of Utah College of Medicine, Salt Lake City, Utah. Normal sera were obtained from 20 noninfected mice. For affinity experiments, antibodies of the immunoglobulin GI (IgGJ) and IgG2 classes were separated from those of IgM and IgA by a column of protein A (Staphylococcus aureus)-Sepharose CL-4B supplied by AB Pharmacia, Uppsala, Sweden.
Enzyme-conjugated anti-immunoglobulins. Anti-immunoglobulins from two sources were labeled and evaluated as conjugates in the ELISA. Goat antimouse IgM, IgG1, and IgG2 were purchased from Meloy Laboratories, Inc., Springfield, Va. Goat antimouse a, At, yi, and y2 were kindly supplied by J. Kearney, University of Alabama in Birmingham, and their specificity was determined as previously described (14). Each anti-immunoglobulin was labeled with alkaline phosphatase, Sigma type VII (Sigma Chemical Co., St. Louis, Mo.; specific activity, 1,140 U/ml of protein) essentially as described by Engvall and Perlmann (9). A 0.3-ml portion of the enzyme solution was centrifuged for 2 min at 8,000 x g. The supernatant was discarded, and 0.1 ml of the specific anti-mouse immunoglobulin was added to the pellet (1.5 mg of enzyme per 0.5 mg of immunoglobulin). After overnight dialysis against PBS (pH 7.3), 10 !d of glutaraldehyde was added to give a final concentration of 0.2%. The mixture was incubated at room temperature for 2 h, diluted to 1 ml with PBS, and dialyzed overnight against PBS. The dialyzed conjugate was then diluted to 10 ml with 0.05 M tris(hydroxymethyl)aminomethane-hydrochloride (pH 8.0) that contained 5% bovine serum albumin, 0.001 M MgCl2, and 0.02% NaN3. The conjugate was stored at 40C. Immunoassay. The technique described by Engvall and Perlmann was employed (9). Disposable polystyrene tubes (11 by 55 mm, Falcon Plastics, Dickinson & Co., Oxnard, Calif.), each containing 1 ml of mycoplasma lysate dilution in 0.1 M sodium carbonate buffer (pH 9.6) with 0.02% NaN3, were incubated at 371C for 4 h. The tubes were then stored with the antigen solution in the cold until used. Before testing, the coated tubes were washed three times (x5 ml) to remove nonattached antigen with 0.9% NaCl containing 0.05% Tween 20 (NaCl/T). The washed, antigen-coated tubes were then incubated with 1.0 ml of anti-mycoplasma antiserum (normal control serum) diluted in PBS-Tween 20-NaN3, pH 7.3 (PBS/TA), for 5 h at room temperature with
INFECT. IMMUN. rotation at 30 rpm. Nonbound material was removed by washing with NaCl/T as before. Class-specific enzyme-conjugated anti-immunoglobulin (1 ml) was then allowed to react with antigen-bound anti-mycoplasma antibodies. After incubation for 16 h at room temperature, the tubes were again washed three times, and 1 ml of alkaline-phosphate substrate (p-nitrophenylphosphate, Sigma Chemical Co., 1-mg/ml solution in 0.05 M sodium carbonate buffer, pH 9.8, containing 0.001 M MgCl2) was added. The reaction was developed at 370C, and after suitable incubation time it was stopped by addition of 0.1 ml of 1 N NaOH. Absorbance at 400 nm (Ao) was calculated per 100 min. Controls and interpretation of results. In all experiments, control reactions were performed as follows. The assay procedure was carried out in tubes omitting one reactant at a time (substituted by PBS), and at the end the substrate was added and A4w was measured. Because all the other controls gave A400 < 0, we routinely included only the following: (i) substrate control; lysate-coated tube with no serum and no conjugate; (ii) conjugate control, lysate-coated tube with no serum, but with conjugate; and (iii) antiserum control, uncoated tube (no lysate) with antiserum and with conjugate. Unless otherwise stated, all results were calculated as follows. The substrate control served as a blank in absorbance readings, and then the absorbance value of the conjugate control was subtracted from the absorbance value of the sera tested. This difference gives an indication of the quantity of the reaction product that results from the binding of specific antibodies to the lysate antigen. Because at present the evaluation of results can be done only on a relative basis, i.e., in comparison with known negative reference serum, the significant difference between negative (normal) and positive (immune) serum was determined.
RESULTS Determination of optimal antigen concentration. Initially, the effect of the amount of antigen used to coat the tube wall on the sensitivity of the ELISA was examined. Tubes were coated with different concentrations of mycoplasma lysate (0.3 to 40 jig/ml), and to each concentration, different dilutions of antiserum (10-2 to 10-6) were added. Two pools of hyperimmune sera were compared with two pools of normal sera. This checkerboard titration was performed for the four immunoglobulin classes-IgA, IgM, IgG1, and IgG2. Representative results summarized in Fig. 1 indicate that 10 ,gg of antigen per ml resulted in maximum binding of IgA, IgG1, and IgG2 antibodies. However, a twofold lower concentration (5 ,ug/r Ad gave 80 to 90% of the optimal absorbance vai.( and was therefore chosen for routine use to conserve antigen. The absorbance values obtained with normal serum for the four classes of immunoglobulins were