Vol. 29, No. 11

JOURNAL OF CLINICAL MICROBIOLOGY, Nov. 1991, p. 2636-2638

0095-1137/91/112636-03$02.00/0 Copyright C 1991, American Society for Microbiology

Gas-Liquid Chromatographic Analysis of Cellular Fatty Acids for Identification of Gram-Negative Anaerobic Bacilli L. STOAKES,' T. KELLY,2 B. SCHIEVEN,l D. HARLEY,' M. D. GROVES,2 AND Z. HUSSAIN1*

RAMOS,l R.

LANNIGAN,'

Department of Clinical Microbiology, Victoria Hospital, London, Ontario N6A 4G5,' and Department of Laboratory Medicine, St. Joseph's Hospital, Hamilton, Ontario L8N 4A6,2 Canada Received 8 April 1991/Accepted 23 August 1991

A commercially available, computer-assisted microbial identification system (MIS) employs gas-liquid chromatographie analysis of cellular fatty acids for bacterial identification. MIS was compared with conventional identification systems. Of 225 gram-negative anaerobes tested, MIS identified 72.4% of the strains to the species level, 88.9% to the appropriate group, and 93.3% to the correct genus.

The pathogenic role of anaerobes in human disease is well established. The definitive identification of anaerobes relies on the use of biochemical reactions in combination with gas-liquid chromatography of metabolic end products as outlined in the manuals from the Anaerobe Microbiology Laboratory of the Virginia Polytechnic Institute (VPI) (6, 7), the Wadsworth Anaerobic Laboratory (17), and the Anaerobe Laboratory of the Center for Disease Control (3). Since the reference methods are rather labor-intensive, this has stimulated interest in the development of automated methods for the identification of anaerobes. Microbial Identification System (MIS) (Microbial ID, Inc., Newark, Del.) employs computerized high-resolution gas-liquid chromatography to determine the fatty acid composition of a bacterial isolate. The fatty acid profile is then compared by the system with the reference data base developed by VPI. The performance of MIS for identification of gram-negative anaerobes was evaluated by comparing the results of MIS with those of conventional methods, including gas-liquid chromatography of metabolic fatty acids (6, 7). A total of 225 clinical strains were examined (Table 1). The following organisms were used for quality control: Bacteroides fragilis ATCC 25285, Bacteroides thetaiotaomicron ATCC 27941, Clostridium perfringens ATCC 13124, Clostridium sordellii ATCC 9714, Peptostreptococcus magnus ATCC 29328, and Clostridium sporogenes VH1904. Conventional methods. For conventional methods, pure growth of the isolates was obtained on enriched blood agar. Colonial morphology, Gram stain reaction, microscopic morphology, hemolytic reaction, pigment production, fluorescence of colonies on exposure to UV light (365 nm), and H202 production of subcultures were examined. Chocolate agar plates (Columbia agar base; Oxoid) were used to determine the aerotolerance of organisms. Gas-liquid chromatography analysis of metabolic short-chain fatty acids was performed. Identification of isolates was obtained by using prereduced anaerobically sterilized media and biochemical tests as outlined in the VPI manual (6, 7). MIS method. For the MIS method, each organism to be tested was subcultured twice in 10 ml of PYG broth (CarrScarborough Microbiologicals, Stone Mountain, Ga.) or in PYG-formate-fumarate broth if the organism was suspected to be Bacteroides ureolyticus. The broth cultures were *

incubated anaerobically at 35°C. The second set of tubes were incubated until heavy growth was obtained, usually after overnight incubation. For strains that produced only scant turbidity, two to four tubes were inoculated to obtain a satisfactory cell pellet. Lysis of cells through saponification, methylation of fatty acids, and extraction of the methyl esters into the organic phase was achieved as previously described (4, 5). The reagents were obtained from vendors recommended in the MIS instruction manual (4). The washed organic extract was then placed in an autosampler vial with an aluminum crimp cap. Specimens were processed on a Hewlett-Packard 5890A gas chromatograph with a Hewlett-Packard 7673A automatic sampler and integrator. The chromatography unit is coupled to a computer with MIS automated software. The parameters of chromatography have been previously described (4, 5). For identification, the VPI broth-based anaerobe library version 3.2 of MIS automated software was employed. The genus Bacteroides has recently undergone major taxonomic revision (15, 16). The MIS uses the current nomenclature. The results of the MIS were acceptable only if the total peak area in the chromatogram was between 45,000 and 500,000. If the area was under the limit, either the extract was concentrated or more PYG tubes were used to get an adequate amount of cells. If the peak area was above the upper limit, it was diluted. In both instances the samples were rerun. In most cases, only one PYG tube had to be used to obtain an adequate pellet. The strains of B. ureolyticus required three to four PYG-formate-fumarate broth tubes. MIS correctly identified 163 of 225 (72.4%) of the strains to the species level. Another 47 (20.9%) strains were designated with the correct genus, and 15 (6.6%) were assigned to the wrong genus. Results for all species are listed in Table 1. All strains of bacteroides and fusobacteria were designated with the correct genus. However, all 12 strains of B. thetaiotaomicron were misidentified at the species level, with 10 strains designated Bacteroides caccae. Two strains of Porphyromonas asaccharolyticus were identified as Porphyromonas gingivalis, and the other two were incorrectly placed in the genus Prevotella. Misidentifications were most common for isolates of Prevotella spp. Two strains of Prevotella species were misidentified as Porphyromonas gingivalis, and the other Prevotella strains misidentified at the genus level were assigned to the genus Bacteroides. Some Prevotella species are closely related to each other, and the members of this genus can be arranged into groups

Corresponding author. 2636

VOL. 29, 1991

NOTES

TABLE 1. Strains and accuracy of MIS identification No. of strains Correctly identified at the:

Genus and

organism

Tested

TABLE 2. Accuracy of MIS identification of Prevotella strains when analyzed by group No. of isolates

Incorrectly Icret'

Species level Genus level identified Genus Bacteroides B. distasonis B. fragilis B. ovatus B. thetaiotaomicron B. uniforms B. vulgatus B. ureolyticus

93 9 44 il 12 3 10 4

71 8 44 5 0 0 10 4

22 1 0 6 12 3 0 0

0 0 0 0 0 0 0 0

Genus Fusobacterium F. mortiferum F. necrophorum F. nucleatum F. varium

21 1 5 il 4

19 1 4 10 4

2 0 1 1 0

0 0 0 0 0

Genus Porphyromonas P. asaccharolyticus P. gingivalis

13 il 2

9 7 2

2 2 0

2 2 0

Genus Prevotella P. bivia P. buccae P. corporis P. denticola P. disiens P. intermedia P. levii P. loescheii P. oralis P. oris P. melaninogenica

98 13 7 9 5 13 17 2 4 3 4 21

64 il 7 6 1 il il 0 0 0 4 13

21 0 0 2 2 1 4 0 4 2 0 6

13 2 0 1 2 1 2 2 0 1 0 2

Total

225

2637

163 (72.4%) 47 (20.9%) 15 (6.6%)

(16). Of 21 Prevotella strains that were misidentified at the species level, 11 were identified correctly as being members of the appropriate group (Table 2). Overall, in 200 of 225 (88.9%) instances the first or the only choice of MIS was the correct species or a closely related species (member of the same group). Composition of cellular fatty acids is genetically controlled and a highly stable trait. Cellular fatty acid profiles have been used to establish taxonomic relationship among bacteria (7, 9, 12). Chromatography and mass spectrometry of cellular fatty acids have also proved to be useful in the classification of anaerobic gram-negative organisms (1, 2, 9, 11, 14). MIS is the first commercial system that makes use of gas-liquid chromatographic profiles of cellular fatty acids. The system takes into account the presence and absence of fatty acids, the quantity of each acid present, and the ratio of the acids in the extract. In this study, gram-negative strains belonging to the four genera most frequently isolated from clinical material were tested. MIS could correctly designate 210 (93.3%) strains with the correct genus. Only 15 of 225 strains were designated with the wrong genus. This included two strains of Prevotella levii (by conventional methods) that were identified by MIS as Porphyromonas gingivalis. There is some uncertainty as to the taxonomic status of human isolates previously described as Bacteroides levii (7, 16). Further,

Prevotella group

Pigmented P. melaninogenica group P. intermedia groupb Nonpigmented, saccharolytic, proteolyticc Nonproteolytic pentose fermenterd Pentose nonfermenter'

Total

Correctly identified the: at Incorrectly Tested Tested ___________ identified Group Genus level level

32 26 26

22 18 22

4 5 1

6 3 3

il

il

0

0

3

2

0

1

98

75

10

13

P. denticola, P. levii, P. loescheii, and P. melaninogenica. b p. intermedia and P. corporis. C P. bivia and P. disiens. d p. buccae, P. oris, P. ruminicola, and P. zoogleoformans. e P. buccalis, P. oralis, and P. veroralis. a

Mayberry et al. have previously shown that the fatty acid profiles of some Prevotella species and Porphyromonas gingivalis are very similar (11), although their methods of analysis were somewhat different from those of MIS. Elimination of these two strains from the analysis of results would result in only a modest improvement in the system's performance. Identification to the species level by the system was less satisfactory. This could be much higher if the strains of B. thetaiotaomicron were correctly identified. Ten strains were misidentified as B. caccae, a species closely related to B. thetaiotaomicron. Overall correct identification to the group level was 88.9%. Some of the species grouped together are so closely related that phenotypic reactions are unable to distinguish them reliably or can do so only with difficulty. In some cases, therefore, identification to the group level may be adequate. The system needs refinement and improvement of its data base before it can be recommended for general use. Improvements in this regard already have been achieved. In our hands, the rate of identification at the species level was merely 61.2% with the previous VPI broth-based anaerobe library, version 2.0 (data not shown). Certain advantages of this system over other commercial systems are obvious. Shelf life and inventory of supplies are not an issue. The system is not restricted by the number of tests that can be performed or accommodated in a panel. In general, supplementary tests are not needed. Updating of data and revision of bacterial taxonomy are easy, as the system is completely computer driven. After extraction of cellular fatty acids, the system can run unsupervised. The system, which includes both software and hardware, is expensive and beyond the reach of most routine diagnostic microbiology laboratories. Manipulation of the strains, including the extraction of fatty acids, is time consuming. However, subculturing of test organisms in PYG broth twice is not necessary. It is sufficient to isolate the organism on an enriched plate medium with or without blood and then subculture it only once in the PYG broth (12). Also, a chopped-meat culture can be used to inoculate PYG broth, provided care is taken to avoid transferring meat particles (12).

2638

NOTES

Even with an improved data base, the MIS will be most useful to large institutions or reference laboratories with large numbers of isolates that can be batched together. This would allow the high initial cost of the system to be spread over a large volume of specimens. REFERENCES 1. Brondz, I., J. Carlsson, M. Sjostrom, and G. Sundqvist. 1989. Significance of cellular fatty acids and sugars in defining the genus Porphyromonas. Int. J. Syst. Bacteriol. 39:314-318. 2. Brondz, I., and I. Olsen. 1991. Multivariate analyses of cellular fatty acids in Bacteroides, Prevotella, Porphyromonas, Wolinella, and Campylobacter spp. J. Clin. Microbiol. 29:183-189. 3. Dowell, V. R., Jr., and T. M. Hawkins. 1977. Laboratory methods in anaerobic bacteriology. CDC laboratory manual. U.S. Department of Health, Education, and Welfare publication no. (CDC)78-8272. Center for Disease Control, Atlanta. 4. Ghanem, F. M., A. C. Ridpath, W. E. C. Moore, and L. V. H. Moore. 1991. Identification of Clostridiuim botulinum, Clostridium argentinense, and related organisms by cellular fatty acid analysis. J. Clin. Microbiol. 29:1114-1124. 5. Hewlett-Packard. 1987. Hewlett-Packard 5898A Microbial Identification System operating manual. Version 3.0. Hewlett-Packard, Avondale, Pa. 6. Holdeman, L. V., E. P. Cato, and W. E. C. Moore (ed.). 1977. Anaerobe laboratory manual, 4th ed. Virginia Polytechnic Institute and State University, Blacksburg. 7. Holdeman, L. V., E. P. Cato, and W. E. C. Moore (ed.). 1987. Anaerobe lab manual update. Virginia Polytechnic Institute and State University, Blacksburg.

J. CLIN. MICROBIOL. 8. Kaneda, T. 1967. Fatty acids in the genus Bacillus. I. Iso- and anteiso-fatty acids as characteristic constituents of lipids in 10 species. J. Bacteriol. 93:894-903. 9. Lambe, D. W., Jr., K. P. Ferguson, and W. R. Mayberry. 1982. Characterization of Bacteroides gingivalis by direct fluorescent antibody staining and cellular fatty acid profiles. Can. J. Microbiol. 28:367-374. 10. Lambert, M. A., D. W. Hillis, C. W. Moss, R. E. Weaver, and M. L. Thomas. 1971. Cellular-fatty acids of non-pathogenic Neisseria. Can. J. Microbiol. 17:1491-1502. 11. Mayberry, W. R., D. W. Lambe, Jr., and K. P. Ferguson. 1982. Identification of Bacteroides species by cellular fatty acid profiles. Int. J. Syst. Bacteriol. 32:21-27. 12. Moore, L. V. H., and W. E. C. Moore. Personal communication. 13. Moss, C. W., S. B. Samuels, and R. E. Weaver. 1972. Cellular fatty acid composition of selected Pseudomonas species. Appl. Microbiol. 24:596-598. 14. Shah, H. N., and D. M. Collins. 1980. Fatty acid and isoprenoid quinone composition in the classification of Bacteroides melaninogenicus and related taxa. J. Apple. Bacteriol. 48:75-87. 15. Shah, H. N., and D. M. Collins. 1988. Proposal for reclassification of Bacteroides asaccharolyticus, Bacteroides gingivalis, and Bacteroides endodontalis in a new genus, Porphyromonas. Int. J. Syst. Bacteriol. 38:128-131. 16. Shah, H. N., and D. M. Collins. 1990. Prevotella, a new genus to include Bacteroides melaninogenicus and related species formerly classified in the genus Bacteroides. Int. J. Syst. Bacteriol. 40:205-208. 17. Sutter, V. L., D. M. Citron, M. A. C. Edelstein, and S. M. Finegold. 1985. Wadsworth anaerobic bacteriology manual, 4th ed. Star Publishing, Belmont, Calif.

Gas-liquid chromatographic analysis of cellular fatty acids for identification of gram-negative anaerobic bacilli.

A commercially available, computer-assisted microbial identification system (MIS) employs gas-liquid chromatographic analysis of cellular fatty acids ...
504KB Sizes 0 Downloads 0 Views