Abstract
Fifty-two clinical strains and 22 type and reference Aeromonas strains, previously genetically characterized by 16S rRNA gene restriction fragment length polymorphism, were identified in parallel with the MicroScan Walk/Away and BBL Crystal Enteric/Nonfermenter systems. The former identified only 14.8% of the isolates correctly, and the latter identified only 20.3% correctly, which indicates that neither of these systems is useful for this purpose.
Members of the genus Aeromonas are rod-shaped gram-negative bacteria assigned to the family Aeromonadaceae. Thirteen species of the 15 included in the genus have been reported from human infections (9, 11). They include gastroenteritis, bacteremia, cellulitis, meningitis, peritonitis, and soft-tissue and bronchopulmonary infections (12, 13). However, the prevalence of the different species in clinical samples is not well known because the techniques routinely used for species identification are less than perfect (1, 4, 5). They are usually based on biochemical characteristics giving a false predominance of Aeromonas hydrophila (11, 22). When clinical strains are identified by molecular methods, A. caviae and A. veronii are more common than A. hydrophila (9, 15). Even though biochemical tests have proved to be less than accurate for Aeromonas identification (2, 7, 18), they are still broadly used. Recently, Abbott et al. (1) reported several biochemical schemes that can be useful for the phenotypic identification of Aeromonas spp. However, the use of these procedures in the clinical setting is difficult and, at present, laboratories still rely on easy-to-use miniaturized methods. Some of the most commonly used methods in clinical laboratories are the miniaturized BBL Crystal Enteric/Nonfermenter (E/NF; BBL Microbiology Systems, Cockeysville, Md.) and MicroScan Walk/Away (W/A; Dade MicroScan Inc., West Sacramento, Calif.) methods. We have evaluated the accuracy of these two methods in the identification of clinical Aeromonas isolates previously identified genetically by 16S rRNA gene (rDNA) restriction fragment length polymorphism (RFLP) (4, 10).
Fifty-two clinical isolates and 22 type and reference Aeromonas strains (Table 1 and 2) were included in this study. The isolates were grown on Trypticase soy agar (Difco, Barcelona, Spain) at 30°C for 24 h. These cultures were used to inoculate the BBL Crystal E/NF and MicroScan W/A Combo Negative 1S type panels, which were incubated at 36 ± 1°C. As recommended by the manufacturers, oxidase testing was performed as a complement for both systems, while the indole test was used to complement the BBL Crystal method. In the latter, the panel reading gave a 10-digit number that was compared to the corresponding database. A confidence rating (CR) of 0.6000 to 1.0000 was considered a correct identification (20). When the CR was <0.6000 but all the given options were Aeromonas species, the one with the higher CR was given as the valid identification. The chi-square test was used to compare the results obtained with both methods by using the Statistical Package for Social Sciences (SPSS 9.0 Inc., Chicago, Ill.). When the P value was <0.05, differences were considered statistically significant.
TABLE 1.
Comparison of the BBL Crystal and MicroScan systems for the identification of 22 Aeromonas type and reference strains
Reference strain | BBL Crystal result | MicroScan result |
---|---|---|
A. hydrophila CECTc 839Ta | A. hydrophila | A. hydrophila group |
A. bestiarum CECT 4227T | A. hydrophila | A. hydrophila group |
A. salmonicida LMGd 13451 | A. hydrophila | V. fluvialis |
A. salmonicida subsp. salmonicida CECT 894T | Vibrio fluvialis | NGb |
A. salmonicida subsp. masoucida CECT 896 | A. hydrophila | A. hydrophila group |
A. salmonicida subsp. achromogenes CECT 895 | A. hydrophila | Pasteurella multocida |
A. salmonicida subsp. smithia NCIMB 13210 | Misclassified gram-negative bacilli | P. multocida |
A. caviae CECT 838T | A. hydrophila | A. hydrophila group |
A. media CECT 4232T | A. hydrophila | A. hydrophila group |
A. eucrenophila CECT 4224T | A. hydrophila | V. fluvialis |
A. sobria CECT 4245T | A. veronii | P. multocida |
A. veronii bv. sobria CECT 4246 | A. sobria | A. hydrophila group |
A. jandaei CECT 4228T | A. hydrophila | A. hydrophila group |
A. veronii bv. veronii CECT 4257T | A. hydrophila | A. hydrophila group |
Aeromonas sp. (GH11) strain CECT 4253 | V. cholerae | Pseudomonas fluorescens/putida |
Aeromonas group 501 CECT 5178 | A. hydrophila | A. hydrophila group |
Aeromonas group 501 CECT 4254 | Chromobacterium violaceum | V. damsela |
A. schubertii CECT 4240T | A. hydrophila | A. hydrophila group |
A. trota CECT 4255T | A. hydrophila | A. hydrophila group |
A. popoffii LMG 17541T | A. hydrophila | V. damsela |
A. allosaccharophila CECT 4199T | A. hydrophila | A. hydrophila group |
A. encheleia CECT 4342T | A. hydrophila | V. parahaemolyticus |
T, type strain.
NG, numerous genera.
CECT, Colección Española de Cultivos Tipo, Universidad de Valencia, Valencia, Spain.
LMG, Culture Collection of the Laboratorium voor Microbiologie, Universiteit Gent, Ghent, Belgium.
TABLE 2.
Comparison of the BBL Crystal and MicroScan systems for the identification of 52 clinical Aeromonas isolates identified by 16S rDNA RFLP
16S rDNA RFLP result | No. of strains tested | No. of strains, result
|
|
---|---|---|---|
BBL Crystal | MicroScan | ||
A. hydrophila | 10 | 10, A. hydrophila | 10, A. hydrophila group |
A. caviae | 19 | 16, A. hydrophila | 17, A. hydrophila group |
1, A. sobria | 2, Vibrio fluvialis | ||
1, A. veronii | |||
1, A. caviae | |||
A. veronii | 14 | 6, A. hydrophila | 13, A. hydrophila group |
4, A. sobria | 1, V. fluvialis | ||
3, A. veronii | |||
3, A. veronii | |||
1, Burkholderia cepacia | |||
A. media | 4 | 4, A. hydrophila | 3, A. hydrophila group |
1, V. fluvialis | |||
A. jandaei | 2 | 1, A. hydrophila | 1, V. fluvialis |
1, Vibrio cholerae | 1, A. hydrophila group | ||
A. bestiarum | 1 | 1, A. hydrophila | 1, A. hydrophila group |
A. salmonicida | 2 | 2, A. hydrophila | 1, V. fluvialis |
1, A. hydrophila group |
Of the 74 strains tested, the BBL Crystal method correctly identified 20.3% of them and MicroScan correctly identified 14.8% of them (Tables 1 and 2). When the 52 clinical Aeromonas isolates were tested, the former method identified 50 isolates (96%) correctly to the genus level and the latter identified 44 isolates (84.6%) correctly to the genus level. The difference was statistically significant (P = 0.008). At the species level, BBL Crystal correctly identified 14 (26.9%) and MicroScan correctly identified 10 (19.3%) of the 52 isolates (Table 2). The BBL Crystal method correctly identified 100% (10 of 10) of the A. hydrophila isolates with a CR of 0.8631 to 0.9993, 21.4% (3 of 14) of the A. veronii isolates with a coincident CR of 0.3604, and 5.2% (1 of 19) of the A. caviae clinical isolates with a CR of 0.7663. MicroScan also correctly identified all of the A. hydrophila isolates, although only to the A. hydrophila group level (Table 2). This is not acceptable because species other than A. hydrophila are included in this group and some of them have been isolated, although rarely, from clinical samples (1, 9, 15). Sixteen of the 19 isolates of A. caviae tested were incorrectly identified as A. hydrophila by the BBL Crystal method. This was probably due to the positive response in the lysine test, which was expected to be negative (3). Six of the 14 isolates of A. veronii were misidentified as A. hydrophila. This was due to their positive responses in the esculin hydrolysis test, which was expected to be negative (3). In the case of MicroScan, most of the isolates of A. caviae and A. veronii were assigned to the A. hydrophila group.
Of the 22 type and reference strains tested, only the type strain of A. hydrophila was correctly identified to the species level by BBL Crystal and MicroScan (Table 1). In addition, 4 and 10 strains could not even be properly assigned to the genus Aeromonas by the respective methods.
BBL Crystal and MicroScan incorrectly identified 60.8 and 63.5% of the 74 isolates as A. hydrophila and A. hydrophila group, respectively. If these results were correct, this would agree with those of Vivas et al. (21), who stated that this is the most common clinical species. However, by using the 16S rDNA RFLP method, we found that A. hydrophila only represented 8.1% of the 490 clinical isolates tested (unpublished data). Other authors have also confirmed this low occurrence of A. hydrophila in clinical samples (15, 16). This tendency of most commercial systems to identify clinical strains as A. hydrophila has led to an overestimation of the clinical relevance of this species (8, 15, 16, 17, 18) and has masked the true incidence of other species. This has been taken to such an extreme that in some recent papers, A. hydrophila has been considered the only important species (22).
To our knowledge, the study of Carnahan et al. (A. Carnahan, S. Lee, D. Watsky, and G. Thomas, Abstr. Annu. Meet. Am. Soc. Microbiol. 1994, abstr. C-245, p. 533, 1994) is the only one that has evaluated the BBL Crystal method for Aeromonas identification. There, all of the isolates were correctly identified to the genus level, in contrast to the 52% obtained with the commonly used API 20E system, which was also evaluated by the same authors.
The poor accuracy of MicroScan in identifying Aeromonas isolates to the species level in our study contrasted with the results of Vivas et al. (21). We tested isolates of all of the Aeromonas species, and only 19.3% of such isolates were correctly identified, while Vivas et al. (21) tested isolates of eight species and correctly identified 78.8% of them. An explanation for this discrepancy could be the fact that those authors compared the results obtained with MicroScan with those obtained with other biochemical procedures, and in general, it has been demonstrated that biochemical identification methods are not reliable for this purpose (4, 6, 10, 14).
Of the 74 Aeromonas isolates tested in the present study, BBL Crystal and MicroScan recognized 8.1 and 21.6% of them, respectively, as not belonging to this genus (Tables 1 and 2). The tendency of miniaturized biochemical identification systems to confuse Aeromonas isolates with Vibrio isolates noticed in our study was already known (2, 19). In our case, MicroScan misidentified eight isolates (10.8%) as Vibrio fluvialis, results similar to those reported by Vivas et al. (21). More troubling is the fact that BBL Crystal misidentified two isolates as Vibrio cholerae, which is of special relevance because of the clinical significance of this microorganism. A method based on colony blot hybridization has recently been proposed to avoid misidentification as Vibrio spp. (7).
Application of the schemes proposed by Abbott et al. (1) could probably demonstrate similar limitations of the miniaturized systems evaluated here for Aeromonas identification. The drawbacks of commercial miniaturized biochemical systems for the identification of Aeromonas spp. lie mainly in inappropriate and incomplete databases that predate current taxonomy. For instance, the BBL Crystal database includes only A. hydrophila, A. caviae, A. veronii, and A. sobria, although none of these species was correctly identified in our study. Why the latter species is added to such databases is unclear since it is known that A. sobria has an environmental origin and is very rarely isolated from clinical samples (9, 13). Although A. sobria is the name classically used by clinical microbiology laboratories to refer to A. veronii bv. sobria (9), its use is no longer justified since it creates confusion. To further confuse, the BBL Crystal method identified the type strain of A. sobria as A. veronii, while a reference strain of A. veronii bv. sobria was identified as A. sobria (Table 1).
In summary, our results demonstrate that BBL Crystal and MicroScan are not useful systems for the identification of clinical Aeromonas isolates and highlight the need to develop systems that are more reliable.
Acknowledgments
This work was supported by grant FIS 99/0944 from the Spanish Ministry of Health, Fundació Ciència i Salut, and a fellowship from the Universitat Rovira i Virgili awarded to L. Soler and M. R. Chacón.
We thank the Colección Española de Cultivos Tipo (CECT) and the Belgium Coordinated Collection of Microorganisms (LMG) for kindly providing isolates.
REFERENCES
- 1.Abbott, S. L., W. K. Cheung, and J. M. Janda. 2003. The genus Aeromonas: biochemical characteristics, atypical reactions, and phenotypic identification schemes. J. Clin. Microbiol. 41:2348-2357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Abbott, S. L., L. S. Seli, M. Catino, M. A. Hartley, and J. M. Janda. 1998. Misidentification of unusual Aeromonas species as members of the genus Vibrio: a continuing problem. J. Clin. Microbiol. 36:1103-1104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Altwegg, M. 1999. Aeromonas and Plesiomonas, p. 507-516. In P. R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, and R. H. Yolken (ed.), Manual of clinical microbiology. American Society for Microbiology, Washington, D.C.
- 4.Borrell, N., S. G. Acinas, M. J. Figueras, and A. Martínez-Murcia. 1997. Identification of Aeromonas clinical isolated by restriction fragment length polymorphism of PCR-amplified 16S rRNA genes. J. Clin. Microbiol. 35:1671-1674. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Carnahan, A. M., S. Behram, and S. W. Joseph. 1991. Aerokey II: a flexible key for identifying clinical Aeromonas species. J. Clin. Microbiol. 29:2843-2849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Castro-Escarpulli, G., M. J. Figueras, G. Aguilera-Arreola, L. Soler, E. Fernández-Rendón, G. O. Aparicio, J. Guarro, and M. R. Chacón. 2003. Characterisation of Aeromonas spp. isolated from frozen fish intended for human consumption in Mexico. Int. J. Food Microbiol. 84:41-49. [DOI] [PubMed] [Google Scholar]
- 7.Chacón, M. R., G. Castro-Escarpulli, L. Soler, J. Guarro, and M. J. Figueras. 2002. A DNA probe specific for Aeromonas colonies. Diagn. Microbiol. Infect. Dis. 44:221-225. [DOI] [PubMed] [Google Scholar]
- 8.Clark, N. M., and C. E. Chenoweth. 2003. Aeromonas infection of the hepatobiliary system: report of 15 cases and review of the literature. Clin. Infect. Dis. 37:506-513. [DOI] [PubMed] [Google Scholar]
- 9.Figueras, M. J., J. Guarro, and A. J. Martínez-Murcia. 2000. Clinically relevant Aeromonas species. Clin. Infect. Dis. 30:988-989. [DOI] [PubMed] [Google Scholar]
- 10.Figueras, M. J., L. Soler, M. R. Chacón, J. Guarro, and A. J. Martínez-Murcia. 2000. Extended method for discrimination of Aeromonas spp. by 16S rDNA RFLP analysis. Int. J. Syst. Evol. Microbiol. 50:2069-2073. [DOI] [PubMed] [Google Scholar]
- 11.Janda, J. M. 2001. Aeromonas and Plesiomonas, p. 1237-1270. In M. Sussman (ed.), Molecular medical microbiology. Academic Press, San Diego, Calif.
- 12.Janda, J. M., and S. L. Abbott. 1996. Human pathogens, p. 151-174. In B. Austin, M. Altwegg, P. J. Gosling, and S. Joseph (ed.), The genus Aeromonas. John Wiley & Sons, Inc., New York, N.Y.
- 13.Janda, J. M., and S. L. Abbott. 1998. Evolving concepts regarding the genus Aeromonas: an expanding panorama of species, diseases presentation, and unanswered questions. Clin. Infect. Dis. 27:453-461. [DOI] [PubMed] [Google Scholar]
- 14.Kozińska, A., M. J. Figueras, M. R. Chacón, and L. Soler. 2002. Phenotypic characteristics and pathogenicity of Aeromonas genomospecies isolated from common carp (Cyprinus carpio L.). J. Appl. Microbiol. 93:1034-1041. [DOI] [PubMed] [Google Scholar]
- 15.Kuijper, E. J., A. G. Steigerwalt, B. C. I. M. Schoenmakers, M. F. Peeters, H. C. Zanen, and D. J. Brenner. 1989. Phenotypic characterization and DNA relatedness in human fecal isolates of Aeromonas spp. J. Clin. Microbiol. 27:132-138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kuijper, E. J., and M. F. Peeters. 1991. Bacteriological and clinical aspects of Aeromonas-associated diarrhea in The Netherlands. Experientia 47:432-434. [PubMed] [Google Scholar]
- 17.Overman, T. L., and J. K. Overley. 1986. Feasibility of same-day identification of members of the family Vibrionaceae by the API 20E system. J. Clin. Microbiol. 23:715-717. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Overman, T. L., J. F. Kessler, and J. P. Seabolt. 1985. Comparison of API 20E, API Rapid E, and API Rapid NFT for identification of members of the family Vibrionaceae. J. Clin. Microbiol. 22:778-781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Park, T. S, S. H. Oh, E. Y. Lee, T. K. Lee, K. H. Park, M. J. Figueras, and C. L. Chang. 2003. Misidentification of Aeromonas biovar sobria as Vibrio alginolyticus by the Vitek system. Lett. Appl. Microbiol. 37:349-353. [DOI] [PubMed]
- 20.Peele, D., J. Bradfield, W. Pryor, and S. Vore. 1997. Comparison of identifications of human and animal source gram-negative bacteria by API 20E and Crystal E/NF systems. J. Clin. Microbiol. 35:213-216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Vivas, J., A. I. Sáa, A. Tinajas, L. Barbeyto, and L. A. Rodríguez. 2000. Identification of motile Aeromonas strains with the Microscan WalkAway system in conjunction with the Combo Negative type 1S panels. Appl. Environ. Microbiol. 66:1764-1766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zhiyong, Z., L. Xiaoju, and G. Yanyu. 2002. Aeromonas hydrophila infection: clinical aspects and therapeutic options. Rev. Med. Microbiol. 13:1-12. [Google Scholar]