Role of Bacterial Chemical Components

Transcription

1 JOURNAL OF BACTERIOLOGY, May, 1966 Vol. 91, No. 5 Copyright 1966 American Society for Microbiology Printed in U.S.A Role of Bacterial Chemical Components in Immunofluorescence WALLIS L. JONES AND VESTER J. LEWIS Department ofhealth, Education, and Welfare, U.S. Public Health Service, Communiicable Disease Center, Atlanta, Georgia Received for publication 19 January 1966 ABSTRACT JONES, WALLIS L. (Communicable Disease Center, Atlanta, Ga.), AND VESTER J. LEWIS. Role of bacterial chemical components in immunofluorescence. J. Bacteriol. 91: The problem of heterologous staining of bacteria in pure culture and in clinical specimens by fluorescein-labeled antisera for Corylnebacterium diphtheriae was examined by observing the presence of particular cellular components in the bacteria involved. Decreases in the degree of staining of C. diphtheriae, diphtheroids, and cocci by the immunofluorescent reagent were shown to accompany hydrolytic removal from the cells of specific compounds, the identities of which varied from those of the treated organisms. Fluorescent-antibody reagents have been prepared and used to identify Corynebacterium diphtheriae in pure cultures and clinical specimens (2, 10). In addition to staining the diphtheria organism, the specific and control conjugates used in this procedure stained many strains of Staphylococcus aureus and streptococci encountered in clinical specimens. Attempts to improve specificity of the reagents for C. diphtheriae by sorption with cross-reacting organisms and by other means (1) were unrewarding. Chemical components of bacterial cell walls have been suggested as a basis for taxonomic classification (3). The present report describes a study of the amino acid, carbohydrate, and lipid composition of cell walls and cells of C. diphtheriae and other bacteria which were stained by fluorescein-labeled antisera for the diphtheria bacillus. Also, the effect on staining that accompanied hydrolytic removal from the cells of various components is described. MATERIALS AND METHODS Strains of bacteria. Six representative strains of C. diphtheriae that were used to develop the original fluorescent-antibody test for this organism (10) were employed. Diphtheroids studied included strains of C. hofmannii, C. xerose, and C. ovis. Staphylococcal and streptococcal strains previously found to be stained with normal conjugates and with conjugated antisera for C. diphtheriae were employed. All strains were grown ovemight at 35 to 37 C in Difco Heart Infusion Broth (HIB) containing 0.05% dextrose. Production offluorescein conjugates. Antisera to the 0 and K antigens of C. diphtheriae were produced in rabbits as previously described by Moody and Jones (10). Globulins separated from the rabbit sera by the addition of (NH4)2SO to a final concentration of 1.56 M were labeled at 25 C with fluorescein isothiocyanate (8). Preparation o whole cells and cell walls. The cells were collected by centrifugation and were thoroughly washed. To obtain cell walls, bacteria were disrupted either by explosive decompression (5) or by ultrasonification. Walls were washed two times in 0.85% NaCl solution, collected by glycerol gradient centrifugation (6), washed two times in 0.85% NaCI solution to remove the glycerol, and then lyophilized. Dinitrofluorobenzene (DNFB) labeling. Washed cells were suspended in a solution consisting of 2 volumes of 4% DNFB in ethyl alcohol and 1 volume of 8% NaHCO3 in water. After being held overnight at room temperature, the cells were thoroughly washed and then hydrolyzed at 105 C with 6 N HCI to obtain the N-terminal dinitrophenyl (DNP) amino acids. Acid hydrolysis. Washed whole cells or cell walls suspended in either 6 N HCI or 2 N H2SO4 were maintained at 105 C for 2 hr. For milder hydrolysis, the washed cells suspended in HCI were heated only to 37 C for 15, 30, or 60 min. Hydrolysates were collected and neutralized. After the mild hydrolysis, smears for fluorescent-antibody staining were made from the washed cell residue. Lipid extraction. Washed whole cells were exposed to 6 N HCl for 30 min and then extracted with acetone for 1 hr. Treatment with HCl before acetone extraction was shown by Salton (11) to increase the yield of lipid. The residue was extracted twice with chloroformmethanol (2:1, v/v) for 2 hr, and all extracts were pooled. Thin-layer chromatography (TLC). TLC was performed on silica gel H layers on glass plates (20 by

2 VoL. 91, 1966 BACTERIAL CHEMICAL COMPONENTS 1701 cm) for lipid and amino acid analysis. Kieselguhr G layers and layers of silica gel H mixed with 0.1 N boric acid were used for carbohydrates. All plates were dried and activated in a hot-air oven at 105 to 110 C for 30 to 45 min. Amino acids were separated with propanol-water (80:36, v/v). For two-dimensional studies, the plates were resolved as above, then rotated 90 degrees and chromatographed with butanol-acetic acid-water (4: 20: 26, v/v). DNP.labeled amino acids were resolved on silica gel H plates by use of chloroform-amyl alcohol-acetic acid (70:30:3, v/v). Solvents for resolving carbohydrates on kieselguhr G plates were 65 ml of ethyl acetate plus 35 ml of a mixture of 2 volumes of isopropanol and 1 volume of water. Carbohydrates on plates of silica gel H with boric acid were resolved with propanol-ammonium hydroxide-water (6:2:1, v/v). Lipids were resolved with chloroform-acetic acid (99.5:1, v/v). Mercuric acetate derivatives of lipids were resolved with 2-propanol-acetic acid (99: 1, v/v). Samples on undecane-impregnated silica gel H plates were resolved with acetic acid-acetonitrile (1:1, v/v). Amino acids and hexosamines were located on the chromatogram by spraying with ninhydrin. Carbohydrates were detected by anisidine phthalate or anisaldehyde sulfuric acid. Amino acids and carbohydrates were identified further by comparison of their RF values with those of known compounds added to the hydrolysate being tested on two-dimensional chromatograms. Lipids were detected by one or more of the following: 10% solution of phosphomolybdic acid in ethyl alcohol, iodine vapor, 0.2% solution of 2', 7'-dichloro-fluorescein in ethyl alcohol, and potassium dichromate-sulfuric acid. Diphenylcarbazone was used to detect the mercury adducts, the preparation of which was carried out as outlined by Mangold and Kammereck (9). Fluorescent microscopy studies. Smears of bacteria were made on glass slides and fixed with 95% ethyl alcohol for 2 min. The fixed smears were stained for 30 min with fluorescein conjugates, after which they were washed for 10 min in 0.85% saline, buffered at ph 7.5 with 0.01 M sodium phosphate, and mounted with buffered glycerol saline and a no. 1 cover slip. Examination of smears was accomplished by use of a fluorescence assembly with an Osram HBO-200 TABLE 1. mercury vapor lamp, Schott BG-12 exciter filter, and OG-1 barrier filter. The stained bacterial smears were assigned values of negative to 4+, according to intensity of fluorescence observed. RESULTS Carbohydrates detected by TLC from H2SO4 hydrolysates of both whole cells and isolated cell walls are shown in Table 1. Mannose, galactose, and arabinose were detected from all strains of C. diphtheriae and C. ulcerans tested. The C. hofmannii strain yielded arabinose and galactose. All three streptococcal groups tested showed a rhamnose spot under the conditions of hydrolysis employed, and galactose was obtained also for the group G streptococcus. Lipid patterns varied to some degree from strain to strain within each type of C. diphtheriae. Gravis types had patterns similar to strains of C. ulcerans. C. hofmannii showed a pattern closely resembling some mitis strains of C. diphtheriae. No lipids were detected from the coccal strains by the TLC techniques employed. Ninhydrin-reacting compounds released from whole cells and isolated cell walls of various species of bacteria by hydrolysis at 105 C for 2 hr with 6 N HCl are shown in Table 2. Alanine and glutamic acid were demonstrated in all strains tested. In addition, diaminopimelic acid was found in all strains of diphtheria and diphtheroid organisms tested. Glycine was noticed in the staphylococcal chromatograms. Lysine was found in all groups of streptococci and also in staphylococcal preparations. Close agreement was found between results from hydrolysates of whole cells and hydrolysates of cell walls as prepared in this study. Two exceptions were noted: glucosamine was found in the whole-cell hydrolysate and not in the hydrolysate from the isolated cell wall from one strain of C. diphtheriae, and was detected only in the cell wall hydrolysate from a second strain. Whole cells not thoroughly washed before hydrolysis Carbohydrates detected by thin-layer chromatography of H2SO4 hydrolysates from whole bacterial cells and cell walls Organism Rhamnose Glucose Mannose Galactose Arabinose Corynebacterium diphtheriae (6 strains) Corynebacterium ulcerans (2 strains) Corynebacterium hofmannii Group A streptococci (3 strains)... + Group C streptococci (2 strains) _ - Group G streptococci Staphylococcus aureus (2 strains)... * Indicates that the component was not always detected.

3 1702 JONES AND LEWIS J. BACTERIOL. TABLE 2. Ninhydrint-reacting compounds detected by thin-layer chromatography of HCI hydrolysates from whole bacterial cells and cell walls GltmcDiamino- Glucos- Organism Alanine acid pimelic Glycine Lysine amine Coryniebacterium diphtheriae (6 strains) Diphtheroids (5 strains) _ - + Group A streptococci (3 strains) _ + + Group C streptococci (2 strains) _ + + Group G streptococci _ + + Staphylococcus aureus (2 strains) TABLE 3. Intensity of bacterial fluorescent-antibody (FA) staining after hydrolytic removal of certain ninhydrin-reactintg compounds Duration of hydrolysis (min) Organism FA FA Compounds FA staining* Compounds removed FA staining* Compounds removed staining* staining* removed Corynebacterium Alanine Negative Alanine Negative Alanine diphtheriae Glucosamine Glucosamine Diaminopimelic acid Staphylococciis Glycine Negative Glycine Negative Alanine aureus Alanine Glycine Lysine Group A strepto None 1+ Alanine Negative Alanine cocci detected Group C strepto None 1+ Alanine Negative Alanine cocci detec- ted Group G strepto None 1+ Alanine Negative Alanine cocci detec- Lysine Lysine ted * FA staining: 4+, maximal fluorescence; 3+, positive fluorescence; 2+, slight fluorescence; 1+, faint fluorescence; negative, no fluorescence. TABLE 4. Fluoresceitt-antibody (FA) stainiing ofcells after blockage of N-terminal aminio acids by dinitrofluorobenzene Organism Amino acids labeled by DNFB FA staining* Corynebacterium diphtheriae Alanine 3-4+ (5-10%, 1+) Diaminopimelic acid Staphylococcus aureus Glycine 3-4+ (1-5%, 1+) Alanine Lysine Group A streptococci Lysine 3-4+ (1%, 1+) Group C streptococci Lysine 3-4+ (1%, 1+) Group G streptococci Lysine 3+ (1%, 1+) * FA Staining: 4+, maximal fluorescence; 3+, positive fluorescence; 1+, faint fluorescence. All cells exhibited 4+ intensity before exposure to DNFB.

4 VOL. 91, 1966 BACTERIAL CHEMICAL COMPONENTS 1703 showed additional components after the 2-hr hydrolysis. Also. additional components were found on the chromatogram from whole cells when the hydrolysis was carried out for longer than 2 hr. Cells exposed to 6 N HCI at 37 C for short time intervals released ninhydrin-reacting components, the nature of which depended upon the length of hydrolysis. Table 3 shows components detected in these hydrolysates and the corresponding fluorescent staining intensity of the cells at each time interval of hydrolysis. Staining of S. aureus was inhibited completely when glutamic acid and alanine appeared in the chromatogram, in addition to the earlier appearing glycine. The removal of glycine alone was accompanied by a marked lowering of fluorescent-staining intensity. Streptococcal staining was greatly reduced upon the appearance of alanine in the hydrolysates. The streptococcal strains required longer hydrolysis for the components to be released in amounts detectable on a chromatogram than did the other bacterial species. Specific staining of C. diphtheriae was reduced in intensity concomitantly with the removal of alanine from the cell. When glutamic acid and glucosamine also were detected on the chromatogram, specific staining no longer occurred. DNP amino acids detected by TLC of hydrolysates from cells treated with DNFB are shown in Table 4. When smears prepared from the cells after treatment with DNFB were stained before hydrolysis, it was observed that the intensity of fluorescence exhibited by most of the bacteria was undiminished. However, the fluorescence of a few cells in each preparation was reduced to an intensity of 1+. Usually, only a slight variation in staining intensity within a microscopic field occurred among cells not treated with DNFB. DISCUSSION Insight into the chemical nature of the cellular components essential for staining of homologous and heterologous bacteria with the immunofluorescent reagent for C. diphtheriae was sought by determining the change in staining that accompanied hydrolytic removal of particular compounds from the cells. As expected, severe hydrolysis abolished staining of all bacteria examined. However, milder treatment showed a certain selectivity. The staining of S. aureus was depressed to a markedly greater extent by 15-min hydrolysis than was staining of the diphtheria organism and streptococci. The species distribution of cell wall components found in this study by use of TLC confirms results from other investigators, who used the more timeconsuming techniques of paper chromatography (3, 4). Application of the simple and rapid technique of TLC in the diagnostic bacteriology laboratory for the identification of specific pathogens by detection of distinctive cellular constituents is a possibility that merits consideration. Such constituents, however, were not found for C. diphtheriae in the present study. Even under identical conditions of bacterial cultivation and extraction, variation in the chromatographic patterns of lipids was apparent more often between different strains of this organism than between a particular strain and other species of corynebacteria. Interspecific patterns of amino acids and carbohydrates were similar. The importance of alanine in the antibodycombining site of antigen was suggested by the observation that the removal of a portion of this amino acid was accompanied by a marked reduction in staining intensity. Results of DNFB labeling of whole cells indicated that alanine occupied an N-terminal position in C. diphtheriae and S. aureus. Alanine was reported earlier by Ingram and Salton (7) to be an important N-terminal amino acid in DNFB-labeled cells of Micrococcus Iysodeikticus, Sarcina lutea, and Bacillus megaterium. The wide variation in fluorescent staining of cells treated with DNFB may be due to variation from cell to cell in the number of amino acids that reacted with the DNFB. Although complete blockage of N-terminal amino acids is difficult to achieve without production of other effects such as denaturation or molecular rearrangement, complete blockage may demonstrate the true role of free amino groups in specific and crossreactions with fluorescent-antibody reagents. ACKNOWLEDGMENTS We wish to thank Walter A. Zygmunt and William B. Cherry for valuable criticism in reviewing this manuscript. LITERATURE CITED 1. CHERRY, W. B., M. GOLDMAN, AND T. R. CARSKI Fluorescent antibody techniques in the diagnosis of communicable diseases. U.S. Public Health Serv. Publ CHERRY, W. B., AND M. D. MOODY Fluorescent-antibody techniques in diagnostic bacteriology. Bacteriol. Rev. 29: CUMMINS, C. S Chemical composition and antigenic structure of cell walls of Corynebacterium, Mycobacterium, Nocardia, Actinomyces, and Arthrobacter. J. Gen. Microbiol. 28: CUMMINS, C. S., AND H. HARRIS The chemical composition of the cell wall in some grampositive bacteria and its possible value as a taxonomic character. J. Gen. Microbiol. 14: FOSTER, J. W., R. M. COWAN, AND T. A. MAAG.

5 1704 JONES AND LEWIS J. BACTERIOL Rupture of bacteria by explosive decompression. J. Bacteriol. 83: FOSTER, J. W., AND E. RIBI Immunological role of Brucella abortus cell walls. J. Bacteriol. 84: INGRAM, V. M., AND M. R. J. SALTON The action of fluorodinitrobenzene on bacterial cell walls. Biochim. Biophys. Acta 24: LEWIS, V. J., W. L. JONES, J. B. BROOKS, AND W. B. CHERRY Technical considerations in the preparation of fluorescent-antibody conjugates. Appl. Microbiol. 12: MANGOLD, H. K., AND R. KAMMERECK Separation, identification and quantitative analysis of fatty acids by thin-layer chromatography and gas-liquid chromatography. Chem. Ind. (London), p MOODY, M. D., AND W. L. JoNES Identification of Corynebacteriuim diphtheriae with fluorescent antibacterial reagents. J. Bacteriol. 86: SALTON, M. R. J The bacterial cell wall. Elsevier Publishing Co., New York.