Soluble Antigens for Immunofluorescence Detection

Transcription

1 APPLIED MICROmOLOGY, Mar., 1967, p American Society for Microbiology Vol. 15, No. 2 Printed In U.S.A. Soluble Antigens for Immunofluorescence Detection of Histoplasma capsulatum Antibodies WILLIAM A. HOOK AmN EARL H. FIFE, JR. Department of Serology, Walter Reed Army Institute ofresearch, Walter Reed Army Medical Center, Washington, D.C. Received for publication 20 October 1966 Soluble antigens from Histoplasma capsulatum in the mycelial and yeast phase were purified by gel filtration, fixed onto paper discs, and employed in an indirect immunofluorescence procedure to detect antibody in sera from individuals infected with H. capsulatum. The elution patterns of crude histoplasmin passed through Sephadex G-200 revealed two minor peaks of protein showing immunofluorescence, complement fixing, and precipitating-antigen activity. A large peak containing the pigment and other low molecular weight materials showed no serological activity. A polysaccharide antigen obtained from fragmented, deproteinized yeast-phase cells was reactive in the fluorescent-antibody test but showed no antigen activity in complement fixation or precipitin tests. Although certain sera from culturally proven cases of blastomycosis, coccidioidomycosis, and cryptococcosis reacted with the purified Histoplasma antigens, preliminary evaluation indicated that the immunofluorescence technique may be of value as a screening procedure for the serodiagnosis of histoplasmosis. Histoplasmin, a soluble culture filtrate of mycelial-phase Histoplasma capsulatum, has been widely used in complement fixation (CF), precipitation, and latex agglutination tests for the serodiagnosis of histoplasmosis (9). However, histoplasmin has not been commonly employed in fluorescent-antibody techniques which generally require the use of whole-cell antigen. When histoplasmin was adsorbed to a cellulose acetate filter paper matrix and dried according to the soluble antigen fluorescent antibody (SAFA) technique of Toussaint (hla), it showed no serological activity (Toussaint, unpublished data). This problem was resolved in the present study which revealed that removal of inhibitory substances by gel filtration yielded a sensitive antigen for immunofluorescence detection of Histoplasma antibodies. Intact yeast-phase cells of H. capsulatum have been used as antigen in fluorescent-antibody tests to detect Histoplasma antibodies in anticomplementary sera (4). However, the results necessarily are dependent on antigens which reside on the insoluble outer surface of the cells (8). This restriction precludes immunofluorescence studies with purified, soluble preparations of yeast-phase cells which may be necessary for definitive studies on the problems of sensitivity and specificity of serological tests for histoplasmosis. In the present work, a polysaccharide antigen extracted from fragmented yeast-phase cells reacted well in the 350 SAFA procedure with sera from patients with histoplasmosis. Preliminary evaluation of both mycelial- and yeast-phase antigens in the SAFA procedure indicated that this method was worthy of consideration by those laboratories requiring a sensitive procedure for the detection of Histoplasma antibodies but not equipped to perform the technically more complex CF test. MATERIALS AND MErHODS Sera. Blood was obtained from humans infected with H. capsulatum, Blastomyces dermatitidis, Coccidioides immitis, Cryptococcus neoformans, or Mycobacterium tuberculosis, and the sera were collected by centrifugation. These sera were used to evaluate the sensitivity and specificity of the standard and experimental Histoplasma antigens. Equal volumes of serum from six individuals with chronic pulmonary histoplasmosis were combined, and the pool was used to determine antigen activity in the various fractions extracted from H. capsulatum. In addition, 12 specimens of sera collected from healthy adult males and shown to be nonreactive in SAFA tests with Histoplasma antigens were combined in equal amounts to provide a negative serum pool control for the SAFA procedure. All sera were stored at -60 C until tested. Mycelial-phase antigens. The histoplasmin was a pooled filtrate of H. capsulatum strains G-8, G-92, and G-95 grown for 6 months at 20 C in the synthetic medium described by Smith et al. (11). A 25-ml amount of the crude histoplasmin was passed through a column (35 by Z.5 cm) of Sephadex G-200 dextran

2 VOL. 15, 1967 DETECTION OF H. CAPSULATUM ANTIBODIES gel (Pharmacia Fine Chemicals, Inc., Piscataway, N.J.), in 0.15 M NaCl (saline) to separate serologically reactive, high molecular weight antigen fractions from lower molecular weight medium constituents and metabolic products. Relative protein contents of the eluates were estimated by optical-density measurements at 280 m,u. Fractions of 5 ml were collected and used to sensitize 0.25-inch (0.6 cm) circular discs punched from grid-lined cellulose acetate filter sheets (HAWG 00010; porosity, 0.45,; Millipore Filter Corp., Bedford, Mass.). After drying, antigen-coated discs were fixed for 10 min at 4 C in 1% acetic acid in 95% ethyl alcohol. Fixative was removed by washing discs in three changes of 0.05 M tris(hydroxymethyl)aminomethane (Tris) buffer in 0.85% NaCl adjusted to ph 8 (12). Fractions showing maximal SAFA activity (tubes 8 to 19, Fig. 1) were pooled and used to sensitize a large number of discs to evaluate the mycelial-phase SAFA antigen. Unfractionated histoplasmin, diluted 1:10, was used as antigen in CF tests. For use in immunodiffusion studies, the histoplasmin was concentrated fivefold by passing the crude -antigen through an ultrafiltration membrane (UM-1, Amicon Corp., Cambridge, Mass.) which retained molecules above a molecular weight of 10,000. Yeast-phase antigens. Soluble antigens were obtained from yeast-phase strains G-8, G-66, G-92, and G-95 grown for 48 hr on Brain Heart Infusion Agar (Difco) at 37 C. Harvests were suspended in 0.5% Formalin, and the cells were washed three times in sterile distilled water, and resuspended to a concentration of S X 108 cells per milliliter in sterile distilled water containing 0.01% Merthiolate. Fragmentation was achieved by twice passing a 40-ml amount of the suspension through a French pressure cell (American Instrument Co., Inc., Silver Spring, Md.) at a pressure of 18,000 psi. After rupture of the cell walls, protein was precipitated at 4 C by adding an equal volume of 0.5 M trichloroacetic acid and allowing the mixture to stand for 3 hr. After centrifugation at 2,000 X g for 15 min, the clear supernatant fluid was decanted and treated with 2 volumes of absolute ethyl alcohol. The precipitate that formed upon standing overnight at 4 C was centrifuged at 2,000 X g for 15 min, resuspended to 5 ml in saline, and adjusted to ph 7; a 2.5-mil quantity was fractionated by passage through a column (1.5 by 22 cm) of Sephadex G-200 in saline. The effluent was collected in 2-ml fractions. Discs then were dipped in each fraction, and the antigen was fixed according to the procedure used for the histoplasmin fractions. Those showing maximal immunofluorescence activity with homologous antiserum (fractions 5 and 6, Fig. 2) were combined and used to evaluate the yeast-phase SAFA test. For CF tests, whole yeast-phase cells of the same strains were suspended in saline containing 0.01% Merthiolate and were diluted to concentration optimal for the CF test. Antigens and discs sensitized with antigen were stored at 4 C. Serological tests. The SAFA test (l la, 12) was performed in the wells of plastic Disposotrays (Limbro Chemical Co., New Haven, Conn.). Antigen-coated discs, prepared as described above, were immersed in ml of test serum diluted 1: 8 in Tris-saline containing 1% Tween 80, and the plates were agitated by rotation at 120 rev/min for 45 min. The serum then was removed by aspiration, and the discs were washed three times in 1-ml volumes of Tris-saline by rotation for 10 min at 180 rev/min. Fluorescein-conjugated goat anti-human globulin (50-780, Microbiological Associates, Inc., Bethesda, Md.) diluted 1:20 in Trissaline containing 2% Tween 80 and clarified by filtration through a cellulose acetate filter of 0.45-;& porosity was used for the secondary reaction. A 0.2-ml volume of conjugated antiserum was added to each disc. The tests then were agitated for 30 min at 120 rev/min and again washed three times. After drying, fluorercence of discs was measured in a photoelectric fluorometer (G. K. Turner Associates, Palo Alto, Calif.) equipped with a 7-54 primary filter, a 2A-12 secondary filter, and a 10% neutral density filter. CF tests were performed by use of precisely standardized reagents prepared according to the method of Kent and Fife (5). Titer was recorded as the reciprocal of the highest serum dilution giving not greater than 50% hemolysis in the presence of an optimal amount of antigen, five 50% hemolytic units of complement, and 1.5 X 108 optimally sensitized sheep erythrocytes in a total volume of 1.5 ml. Sera showing CF titers less than 1:8 were considered negative. Precipitin tests were performed in agar gel on microscope slides containing a 2.0-ml layer of 0.6% purified agar in 0.85% NaCl and 0.01% Merthiolate. Wells 3 mm in diameter were separated by a distance of 6 mm. Slides were examined for the presence of precipitin bands after 2 and 4 days at 22 C. A commercially prepared tube agglutination test, in which histoplasmin was adsorbed onto latex particles (11-58J, Colab Laboratories, Inc., Chicago Heights, Ill.), was used to titrate agglutinating antibody. Titers of 1:8 or greater were considered positive. Chemical analysis. Chemical tests of the yeast- and mycelial-phase SAFA antigens purified by gel filtration included determination of protein by the Folin- Ciocalteau reagent (6), carbohydrate by the Molish reaction (10), deoxyribonucleic acid (DNA) by the method of Dische (3), and lipid by reduction of potassiumi dichromate after extraction with alcohol and petroleum ether (1). Heat stability of the SAFA antigens was determined by placing 0.5 ml of the appropriate purified antigen in a stoppered tube, incubating for 30 min at 70 C or 100 C in a constanttemperature water bath, cooling, and using as antigen to coat discs in the usual manner. RESULTS Purification of histoplasmin. In initial experiments, it was observed that crude histoplasmin lost serological activity when dried on cellulose acetate discs for use in the SAFA procedure (Toussaint, unpublished data). This problem was overcome by passing histoplasmin through Sephadex G-200 dextran gel to separate serologically active, high molecular weight components from inhibitory culture constiuents or metabolic

3 352 HOOK AND FIFE APPL. MICROBIOL. products (Fig. 1). Tubes 8 to 25 contained the antigens which reacted in the immunofluorescence test, with the earlier fractions showing the greatest activity. The characteristic green-brown color of histoplasmin was found to be associated with a serologically inactive major peak of protein material eluted in tubes 33 to 40. Figure 1 also shows two discrete areas of maximal complement-fixing activity (between tubes 10 to 16 and tubes 20 to 22). Precipitating antigen was also found in the same general region. Polysaccharide from yeast-phase cells. A serologically active, soluble polysaccharide was obtained from fragmented yeast-phase H. capsulatwn. These cells, however, were found to be unusually resistant to fragmentation and required at least two passages through the pressure cell to assure rupture. Phase-contrast microscopy of the final effluent from the press revealed broken cells that had lost intracellular constituents, but many of the empty cell walls retained their oval shape. Treatment of the suspension with trichloroacetic acid gave a heavy precipitate. Addition of ethyl alcohol to the trichloroacetic acid supernatant solution produced a precipitate which was highly active as antigen in the SAFA procedure but showed no CF or precipitating activity. Gel filtration of this ethyl alcohol-insoluble material yielded several fractions containing carbohydrate, but immunofluorescence activity was found almost exclusively in the high molecular weight fractions obtained in tubes 5 and 6 (Fig. 2). None of these fractions, however, was active in CF or E 0o cl c: SA FA C F PPTN. Tube Number FIG. 1. Elution pattern ofhistoplasmin passed through a column of Sephadex G-200 in 0.85% NaCI. Fractions showing strongest serological activity against histoplasmosis serum pool are designated by lines of double thickness. 150 _ -I l X i Tube Number FIG. 2. Carbohydrate elution pattern of yeast-phase soluble antigen from Histoplasma capsulatum passed through a column of Sephadex G-200 in 0.85% NaCl. Fractions showing strongest immunofluorescence activity against histoplasmosis serum pool are designated by a line of double thickness. TABLE 1. Chemical characteristics of soluble, mycelial-, and yeast-phase antigens from Histoplasma capsulatum used in immunofluorescence (SAFA) tests Antigen Protein Carbo- DNA Lipid hydrate pg/mi pg/mi pg/mi Mycelial phasė Yeast phase....<2 138 <5 + precipitin tests. All of the above fractionations were performed three times, and the results obtained were nearly identical. Chemical composition of SAFA antigens. The data presented in Table 1 suggest that the purified immunofluorescent antigen obtained from histoplasmin consisted primarily of protein and carbohydrate with some DNA and lipid. On the other hand, the principal component of antigen purified from yeast-phase cells was carbohydrate. Both mycelial- and yeast-phase antigens retained serological activity after heating to 70 C for 30 min, and both contained lipid (Table 1). Although the colorimetric method used for lipid assay has been used by others to quantitate total blood lipids by calculation of the oxidation of a cholesterol standard, use of this standard curve was not considered to be directly applicable to the quantitative measurement of lipids from Histoplasma. Therefore, in the present studies, the methods were used to provide qualitative evidence of the presence or absence of lipid.

4 VOL. 15, 1967 DETECTION OF H. CAPSULATUM ANTIBODIES Preliminary evaluation of mycelial- and yeastphase antigens. Sera from culturauy proven cases of histoplasmosis were tested with both mycelialand yeast-phase SAFA antigens, and the results were compared with those obtained in CF, precipitation, and latex agglutination tests (Table 2). In the SAFA procedure, 11 of 12 serum specimens from individuals infected with H. capsulatum gave fluorometer dial readings (FDR) of 10 or greater with both mycelial- and yeast-phase SAFA antigens, which were considered positive reactions. On the other hand, only six of the specimens had CF titers toward histoplasmin. These latter sera also were positive in precipitin tests with histoplasmin. Positive yeast-phase CF tests were observed with 8 of the 12 sera, and titers seemed generally to parallel the magnitude of the yeastphase SAFA reactions. Results with sera from 12 healthy individuals showed no positive reactions in CF, precipitin, and latex agglutination tests. On the other hand, SAFA tests of 12 normal sera showed 3 positive reactions with mycelial phase antigen and 5 observable FDR with yeast-phase 353 antigen. Of these five, however, only one serum specimen was considered to give a positive reaction, i.e., FDR greater than 10. Table 3 gives results obtained with the same antigens tested against sera from human cases of blastomycosis, coccidioidomycosis, cryptococcosis, and tuberculosis. Four of the six individuals infected with B. dernatitidis showed significant FDR with both mycelial- and yeast-phase antigens from H. capsulatum. Three of these sera also reacted in CF tests with yeast-phase antigen; one was anticomplementary, and one had a histoplasmin latex agglutination titer of 1: 8. The findings presented in Table 3 also suggest that sera from humans infected with C. immitis may give positive reactions with SAFA antigens from H. capsulatum. One of the six coccidioidomycosis sera had a CF titer of 1:8 against the yeast-phase Histoplasma. Of the six cryptococcosis sera tested, four gave positive SAFA values with mycelial-phase H. capsulatum antigen, and one showed a titer of 1:32 in the histoplasmin latex test (Table 3). In TABLE 2. Comparison of immunofluorescence (SAFA) tests with complement fixation, agar gel precipitation, and latex agglutination tests for the detection of antibodies toward Histoplasma capsulatum Diagnostic status Serm no. Mycelial-phase antigens' Yeast-phase antigens SAFA CF PPTN LA SAFA CF Histoplasmosis Hi 28 1:32 + _ 52 H2 99-1:8 77 H3 97 1: H :64 H :16 H :64 H :16 H8 20-1:8 45 1:32 H9 53 1: :16 H : :8 Hit 43 1: :256 H : _ Healthy Ni N N N N N N N N N N Numbers in SAFA column are fluorometer dial readings. Complement fixation (CF) and latex agglutination (LA) values are serum titers. Precipitin tests (PPTN) were recorded as positive when one or more bands of precipitation were observed.

5 354 HOOK AND FIFE APPL. MICROBIOL. TABLE 3. Diagnostic status Cross-reactions between antigens from Histoplasma capsulatum and sera from blastomycosis, coccidioidomycosis, cryptococcosis, and tuberculosis Serum no. Mycelial-phase antigens" Yeast-phase antigens SAFA CF PPTN LA SAFA CF Blastomycosis RB :32 B B :8 B :8 39 1:16 B B Coccidioidomycosis C C C C C :8 C Cryptococcosis Dl D :32 D D D D : Tuberculosis Ti 7 _ 1: T T T T T a Numbers in SAFA column are fluorometer dial readings. Complement fixation (CF) and latex agglutination (LA) values are serum titers; PPTN = precipitin tests. contrast, only one of the six cryptococcosis sera reacted with the yeast-phase SAFA antigen, and this serum also reacted strongly in the CF test (titer 1: 32). Results in Table 3 also show that none of the six tuberculosis sera gave reactions considered positive in CF or SAFA tests with mycelial-phase antigens. One serum (T-1) had a latex agglutination titer of 1: 8 and a SAFA yeast-phase FDR of 23, whereas another had a SAFA yeast-phase FDR of 19. The remaining four sera were considered negative. DiscussIoN Purification of histoplasmin by gel filtration made it possible to dry this soluble antigen on filter paper discs without loss of serological activity for use in an immunofluorescence test. Apparently the presence of high concentrations of medium or metabolic constituents in the crude histoplasmin caused a loss of the antigenic principle when antigen was dried (Toussaint, unpublished data). The elution pattern of crude histoplasmin passed over Sephadex G-200 revealed two minor peaks of protein material showing SAFA, complement-fixing, and precipitating-antigen activity, and one large peak of lower molecular weight material that had a greenish-brown color but was serologically inert (Fig. 1). Yeast-phase polysaccharide obtained by fragmentation, deproteinization, and concentration was a soluble product that could be further purified and characterized by gel filtration. The elution of antigen reactive in the immunofluorescence test among the earliest fractions collected (Fig. 2) suggested a molecular weight approaching 200,000. Chemical determinations of the purified histoplasmin used as SAFA antigen revealed the presence of protein, carbohydrate, lipid, and some nucleic acid (Table 1). The fractions containing the antigen (tubes 8 to 19, Fig. 1) possessed both complement-fixing and precipitating activity in addition to immunofluorescence activity. On the other hand, the deproteinized polysaccharidelipid antigen extracted from H. capsulatum did not fix complement and showed no precipitin activity (Fig. 2). Perhaps removal of nitrogenous material from this antigen was responsible for the failure to fix complement as has been suggested

6 VOL. 15, 1967 DETECTION OF H. CAPSULATUM ANTIBODIES in the case of antigens from other pathogenic fungi (13). Both mycleial- and yeast-phase antigens were stable to heating (70 C for 30 min), and discs impregnated with these antigens could be kept at 4 C for several months without loss of activity. Preliminary evaluation of the SAFA test for histoplasmosis by use of mycelial- and yeast-phase antigens suggested adequate sensitivity for serodiagnosis, and in some instances the procedure appeared to be superior to the standard tests. This was evidenced by the frequent ability of a single antigen, e.g., purified histoplasmin, to detect antibody when standard histoplasmin CF tests are negative (Table 2). A few of the sera from healthy individuals tested by the SAFA procedure gave positive reactions with both mycelialand yeast-phase antigens. Since the "normal" sera were obtained from residents of Washington, D.C., and surrounding states, an area moderately endemic for histoplasmosis, it is likely that at least some may have had contact with H. capsulatum or its products. Thus, it would appear that the presence of a detectable reaction in the SAFA test does not necessarily indicate active histoplasmosis. Results of Table 3 indicated that sera from proven cases of blastomycosis, coccidioidomycosis, cryptococcosis, and perhaps tuberculosis react in SAFA test with antigens from H. capsulatum. The occurrence of such cross-reactions with antigens from Histoplasma is well known (2). Although the SAFA procedure may be somewhat less specific for differentiation between coccidioidomycosis and histoplasmosis than the CF, precipitin, or agglutination tests, this deficiency appeared to be compensated for by the greater sensitivity of the immunofluorescence test. In view of its relative technical simplicity and the ability to obtain results on the day of testing, the SAFA procedure seems to be worthy of special consideration for use in laboratories not equipped to perform complement fixation tests for histoplasmosis. It appears to be more suitable than either the tube latex agglutination or precipitin tests as a screening procedure to identify cases which require additional immunological, histological, and cultural studies. In addition, the SAFA procedure can be used to obtain results from sera that cannot be tested by complement fixation because of anticomplementary activity, or from sera that are not suitable for use in the agglutination test. Experience gained by the testing of more than 1,000 serum specimens submitted for routine fungal serological studies have supported these conclusions. 355 Improvement of specificity of the SAFA tests for histoplasmosis conceivably may be achieved by further chemical purification of the antigens, e.g., use of additional steps of column chromatography or chemical precipitations. However, efforts along these lines necessarily would be restricted to relatively mild procedures since experience has shown that extensive chemical treatment may reduce rather than improve specificity of antigens. For example, Pine et al. (7) reported that antigens extracted from yeast-phase H. capsulatum by a variety of chemical procedures were less specific than the unaltered, whole-cell preparations. Perhaps a more rewarding approach toward improving specificity of SAFA antigens might utilize genetically homogeneous cultures of carefully selected strains of fungi. ACKNOWLEDGMENT We wish to thank Robert L. Taylor for supplying the standard complement-fixing antigens used in these studies. We also thank Louis Barno for his excellent technical assistance and Andre J. Toussaint for his advice concerning the SAFA procedure. LiTERATuRE CITED 1. BRAGDON, J. H Colorimetric determination of blood lipides. J. Biol. Chem. 190: CAMPBELL, C. C The accuracy of serologic methods in diagnosis. Ann. N.Y. Acad. Sci. 89: DiscHE, Z Color reactions of nucleic acid components, p In E. Chargaff and J. N. Davidson [ed.], The nucleic acids, vol. 1. Academic Press, Inc., New York. 4. KAUFMAN, L., B. BRANDT, AND D. MCLAUGHLIN Evaluation of the fluorescent antibody and agar gel precipitin tests for detecting Histoplasma antibodies in anticomplementary sera. Am. J. Hyg. 79: KENT, J. F., AND E. H. FIFE, JR Precise standardization of reagents for complement fixation. Am. J. Trop. Med. Hyg. 12: LowRY, 0. H., N. J. ROSBROUGH, A. L. FARR, AND R. J. RANDALL Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: PINE, L., C. J. BOONE, AND D. McLAUGHLIN Antigenic properties of the cell wall and other fractions of the yeast form of Histoplasma capsulatum. J. Bacteriol. 91: SALVIN, S. B., AND E. RIBI Antigens from yeast phase of Histoplasma capsulatum. II. Immunologic properties of protoplasm vs. cell walls. Proc. Soc. Exptl. Biol. Med. 90: SCHUBERT, J. H., AND G. L. WIGGINS The evaluation of serologic tests for histoplasmosis in relation to the clinical diagnosis. Am. J. Hyg. 77: SHETLAR, M. R., AND Y. F. MASTERS Use

7 356 HOOK A IIN4D FIFE APPL. MICROBIOL. of the thymol-sulfuric acid reaction for determination of carbohydrates in biological materials. Anal. Chem. 29: SMITH, C. E., E. C. WHmNG, E. E. BAKER, H. G. ROSENBERGER, R. R. BEARD, AND M. T. SAiTO The use of coccidioidin. Am. Rev. Tuberc. 57: la. TOUSSAINT, A. J Improvement of soluble antigen fluorescent antibody procedure. Exptl. Parasitol. 19: TOUSSAINT, A. J., AND R. I. ANDERSON Soluble antigen fluorescent-antibody technique. Appl. Microbiol. 13: WILSON, J. W., Am 0. A. PLUNKErr The fungous diseases of man, p Univ. California Press, Berkeley.