Typing of the Fish Pathogen Listonella (Vibrio) anguillara by Pyrolysis Mass Spectrometry

Zbl. Bakt. 285, 245-251 (1997) © Gustav Fischer Verlag, Jena

Typing of the Fish Pathogen Listonella (Vibrio) anguillara by Pyrolysis Mass Spectrometry G. P. MANFI0 1,5, M. GOODFELLOW!, B. AUSTIN 2 , D. A. AUSTIN 2 , K. PEDERSEN3, J. L. LARSEN3, L. VERDONCK4, and J. SWINGS 4 Department of Microbiology, The Medical School, University of Newcastle-upon-Tyne, Newcastle-upon-Tyne, UK 2 Department of Biological Sciences, Heriot-Watt University, Edinburgh, UK 3 Department of Veterinary Microbiology, Section on Fish Diseases, The Royal Veterinary and Agricultural University, Frederiksberg C, Copenhagen, Denmark 4 Laboratorium voor Microbiologie, Faculteit Wetenschappen, Universiteit Gent, Gent, BelglUm 5 Present address: Fundac;:ao Tropical de Pesquisas e Tecnologia 'Andre Tosello', Campinas - SP, Brazil 1

Summary Twenty-eight representatives of Listonella (Vibrio) anguillara serovars 01, 02 and 03 were compared by Curie-point pyrolysis mass spectrometry (PyMS). The representatives of serovars 01 and 03 formed discrete, homogeneous groups in ordination plots of the PyMS data. Strains from serovar 02 were recovered in two groups, one of which encompassed six strains including the type strain of the species and the reference strain for serovar 02, and the other included two strains which showed cross-reactions between serovars 02 and 05. The almost complete agreement found between the PyMS and the serological data suggests that pyrolysis mass spectrometry will prove to be an effective method for interstrain comparison within the species Listonella anguillara.

Introduction

Listonella (Vibrio) anguillara, an agent of vibriosis, is a serious pathogen of fin- and shellfish, with worldwide distribution (1). The taxonomy of this pathogen has undergone several revisions with L. anguillara biotype I being retained as L. anguillara (24), and V. anguillarum biotype II being reclassified as V. ordalii (20). It has been proposed that V. anguillarum biotype I be reclassified in a new genus, Listonella, as Listonella anguillara (15, 16), but this classification has not been widely accepted by fish pathologists (1). There is evidence that L. anguillara encompasses considerable phenotypic heterogeneity with the recognition of a plethora of so-called 'types' differentiated largely by biochemical (1) and serological tests (18). Ten major serovars are recognised (22). The serovars cross species boundaries insofar as serovars 01 and 02 are recognised in L. anguillara (1). The speed and reproducibility of Curie-point pyrolysis mass spectrometry (PyMS) and its applicability to a wide range of bacteria make it an attractive method for cla-

246

G. P. Manfio et al.

rifying the taxonomic affinities between closely related organisms (8). Pyrolysis-mass spectrometry is not a typing method per se as a permanent type description is not assigned to the examined organisms, but it has proved to be a quick and effective method of interstrain comparison of bacteria that commonly cause outbreaks of disease and also of environmental isolates (12). It is evident that PyMS can be used to discriminate between strains as accurately as routine typing systems (7, 9, 21), and in some cases has been used to separate isolates beyond the resolution of such systems (6, 9, 21). Similarly, PyMS studies on isolates from outbreaks of disease, due to so-called untypable strains, have been used found to agree with epidemiological data that have subsequently become available (4, 19). The aim of the present study was to examine the potential of pyrolysis mass-spectrometry for typing strains of L. anguillara.

Methods

Bacterial cultures. Twenty eight strains of Listonella anguillara representing serovars 01, 02 and 03 were examined (Table 1). Authenticity of the cultures was verified after Baumann et al. (3) and Austin and Lee (2). The cultures were stored at -70°C suspended in 1 ml of tryptone soya broth (Oxoid) supplemented with 1 % (w/v) sodium chloride (TN broth) and 15% glycerol (Sigma). Bench cultures were maintained at room temperature on tryptone soya agar (Oxoid) supplemented with 1 % (w/v) sodium chloride (TN agar). Serotyping. The preparation of absorbed polyclonal antisera and a-antigen have been described elsewhere (14, 17). A loopful of undiluted antiserum was mixed with an equal volume of a-antigen on a glass microscope slide, and positive results were indicated by the presence of agglutination. Pyrolysis Mass Spectrometry. Test strains were run and analysed under a code which was broken after data analysis. Strains were grown on TN agar for 48 hours at 22°C, Sample preparation is described in detail elsewhere (12). Briefly, confluent growth was collected from the surface of the agar using disposable plastic loops and thinly spread over Curie point foils (530°C). The sample size was adjusted to allow pyrolysis counts of -3-4 X 106 IOns. Prepared triplicate samples were dried at 80°C for minutes and Curie-point pyrolysis mass spectrometry carried out in a RAPyD 400X instrument (Horizon Instruments, Heathfield, East Sussex, UK). Pyrolysis time was 3 seconds with a temperature rise time of 0.6 of a second. The inlet heater was set at 160°C, The pyrolysate was ionised by a low energy (25 eV) electron beam and separated in the quadrupole mass spectrometer (100 scans at 180 msedscan). Integrated ion counts for each sample at unit mass intervals from 50 to 200 were recorded and stored on hard-disk, together with total ions counts and the sample pyrolysis sequence, without background removal. Data Analysis. PyMS spectra were corrected by iterative re-normalisation (11) and individual masses ranked according to their 'characteristicity' values (5) prior to principal components analysis (PCA). Principal components (PC's) accounting for less than 0.1 % of the total variance were discarded. Canonical variate analysis (CVA) was then used to group the samples on the basis of the retained PC's taking into account the sets of triplicates (25). An ordination diagram of the PyMS data was produced by plotting the eigenvectors from the CVA. The eigenvectors were also used to generate a distance matrix containing information on the Mahalanobis distances. The latter were transformed to a percentage similarity matrix using the SG coefficient (10) and used to construct a dendrogram using the unweighted pair group method with arithmetic averages (UPGMA) algorithm (23). The ordination diagram and corresponding dendrogram were examined in conjunction to determine the relationships between the test strains.

Typing of Listonella anguillara by PyMS

247

Results and Discussion The ordination diagram of the PyMS data for the first two canonical discriminant axes is shown in Fig. 1. Excellent agreement was found between the results of the triplicate analyses on each strain (results not shown) and with the results obtained with the duplicated cultures (VIB29d and VIB72d). It is evident from the interpretation of the ordination diagram that 25 out of the 27 L. anguillara strains can be assigned to four distinct and relatively homogeneous groups. Similar results were obtained when the data were expressed in a dendrogram derived from Mahalanobis distances (Fig. 2). The results of the serotyping analyses are given in Table 1.

Table 1. Source and serological grouping of Listonella anguillara strains Laboratory/ reference no.

Source and/or country of origin

Equivalence in other collections 1

Serovar

VIB1 VIBII VIB14 VIB15 VIB36 VIB46 VIB57 VIB59 VIB65 VIB67 VIB2 VIB12 VIB29 VIB30 VIB31 VIB50 VIB51 VIB55 VIB71 VIB72 VIB3 VIB113 VIB132 VIB136 VIB139 VIB148 VIB152 VIB272

Oncorhynchus mykiss, Denmark Dicentrarchus lahrax, Greece O. mykiss, Italy D.lahrax, Greece Mugilidae, Italy D.lahrax, Italy D.lahrax, Italy D.lahrax, Italy Scophthalmus maximus, Spain S. maximus, Spain Gadus morhua, Denmark D.labrax, Greece D.labrax, Greece Sparus aurata, Greece Salma salar, UK D.lahrax, Italy Fish, Italy D.lahrax, Italy Scotland, UK G. morhua, Norway O. mykiss, Denmark O. mykiss, Denmark O. mykiss, Denmark Anguilla sp., Denmark Anguilla sp., Denmark Water, Denmark Water, Denmark Japan

ATCC 43305 AVL 90-9-22 HWU44 HWU48 UB 76/91 UB 327/90 UB 861189 UB 909/89 UB A024 UB A055 ATCC 43306 AVL 89-2-54 HWUVA73 HWUVA75 NCMB 828 UB417/90 UB 498/90 UB 578/90 LMG 4411 LMG4437T ATCC 43307 RVAU 820806-1/3 RVAU 91-7-143 RVAU 92-8-158 RVAU 92-8-168 RVAUV24/3 RVAUV21/4 LMG 13584

01 * 01 01 01 01 01 01 01 01 01 02* 02 02+ 02+ 02 02 02 02 02 02 03* 03 03 03 03 03 03 03

* Reference strains for serovars 01, 02 and 03; +Also reacts with serovar 05; TType strain. 1 Sources:

ATCC, American Type Culture Collection, Rockville, Maryland, USA; AVL, Aquaculture Vaccines Limited, Saffron Walden, UK; HWU, Heriot-Watt University, Edinburgh, UK; LMG, Laboratorium voor Microbiologie, Rijksuniversiteit, Gent, Belgium; NCMB, National Collection of Industrial and Marine Bacteria, Aberdeen, UK; RVAU, Royal Veterinary and Agricultural University, Copenhagen, Denmark; UB, University of Barcelona, Barcelona, Spain.

~

00

25

113

B

139 0

15

132,136

152 0 3 272 00

P

!'C

·ta;> ~

.-C ~ "0

t:!1

o

...... 5

U

U

~ ::s

148 0

~

55

C

0

u Q)

CI)

29

2

~Ii

46 1 59 __ -11,67

30,29d

-14 57

- - --

50

.. 4l..

-15

.... "

.

. .. ..

31 -5

~

~1

12

15

65

2,72d

-25 -25

-15

-5

5

First canonical variate

15

25

Fig_ 1. Ordination plot of the PyMS analysis of Listonella anguillara strains_ Datapoints represent the mean of triplicate samples. The first two canonical variates account for 82 % of the total variation_ Strain numbers refer to the VIB laboratory codes in Table 1. Key to symbols: • = serovar 01, ... = serovar 02, and 0 = serovar 03.

Typing of Listonella anguillara by PyMS

249

% Similarity 40 !

55 !

70 !

85 !

I

14 15 65 11

67 57 59 36

SerovMOI

46

ig ] SerovM 02 29d 139

113 132

136 272

SerovM03

3

148 152 72 72d 12 55 31 50 2

SerovM02

Fig. 2. UPGMA dendrogram representing the relationships between Listonella anguillara strains. Strain codes as in Figure 1.

The largest PyMS group encompassed all ten representatives of serovar 01. Isolates belonging to this serovar are considered to be highly virulent due to the presence of a virulence plasmid, pJM1, which codes for an iron sequestering system (1,18). The second largest group accounted for all eight strains of serovar 03. Strain VIB148, placed on the periphery of this group, formed a subgroup with strain VIB152 in the UPGMA cluster analysis (Fig. 2). It is interesting to note that both strains were isolated from water samples, contrary to the other strains which were obtained from clinical samples. Ribotyping shows that members of serovars 01 and 03 form distinct and homogeneous groups, though serovar 03 strains isolated from the environment were found to be heterogeneous and distinct from serovar 03 strains isolated from clinical samples (18). The two smaller PyMS groups contained the remaining serovar 02 isolates. The recovery of isolates VIB29 and VIB30 in a distinct group was particularly interesting as these strains showed cross reactions to serovars 02 and 05. Strain VIB71 (serovar 02) was placed at the edge of this group on the ordination diagram and dendrogram. The remaining six isolates of serovar 02 were recovered in a diffuse cluster. Two subgroups can be detected on the ordination diagram and dendrogram: one containing isolates VIB12 and VIB72 (including it's duplicate) and another isolates VIB2, VIB31, VIB50 and VIB55. Further studies are needed to clarify the taxonomic status of these subgroups. Strains from serovar 02 have been found to be more diverse in biochemical reactions and plasmid content than those from serovar 01 (13), indicating the pre-

250

G. P. Manfio et a!.

sence of several clonal lines. Ribotyping has also demonstrated that the serovar 02 strains are more heterogeneous on the basis of ribotype patterns (18). It can be concluded from these preliminary data that the similarities and differences between the L. anguillara strains, when analysed by PyMS, agree almost completely with the serotyping data. This suggests that PyMS is an effective method for interstrain comparison within the species L. anguillara. In some instances the pyroclassification provided further clues to the existence of subgroups within each of the serovars, but the authenticity of this finer classification needs to be investigated further using additional taxonomic methods. The relative merits of PyMS and other genomic fingerprinting methods of strain comparison, such as DNA restriction patterns, ribotyping and PCR typing (RAPDS), has yet to be established. However, by comparison with such methods PyMS is rapid, cheaper in consumable costs and capable of simultaneously analysing large number of isolates, especially if the PyMS data are analysed using neural networks (8). Inexpensive interstrain comparisons of L. anguillara isolates should help to promote understanding about the pathogenicity and epidemiology of this troublesome fish pathogen. Acknowledgements. Financial support was provided by CEC Grant No.AIR-CT920341. G.P.M. was supported by the Brazilian RHAE-CNPq grant No. 260181/90.0.

References 1. Austin, B. and D. A. Austin: Bacterial Fish Pathogens: Disease in Farmed and Wild Fish. 2nd Ed. Simon & Schuster, Chichester, UK (1993) 2. Austin, B. and J. Lee: Aeromonadaceae and Vibrionaceae. In: Identification Methods in Applied and Environmental Microbiology, pp.163-182, eds. R. G. Board, D.Jones, and F. A. Skinner. Society for Applied Bacteriology Series No.29, Blackwell, London (1992) 3. Baumann, P., A. L. Furniss, and J. Lee: Genus 1. Vibrio Pacini 1854, 411 AL. In: Bergey's Manual of Systematic Bacteriology, Volume 1, pp. 518-538, eds. N. R. Krieg and J. G. Holt. Williams & Wilkins, Baltimore (1984) 4. Cartmill, T. D. I., K. Orr, R. Freeman, and P. R. Sisson: Nosocomial infection with Clostridium difficile investigated by pyrolysis mass spectrometry. ]. Med. Microbio!. 37 (1992) 352-356 5. Eshuis, w., P. G. Kistemaker, and H. L. C. Meuzelaar: Some numerical aspects of reproducibility and specificity. In: Analytical Pyrolysis, pp.151-166, eds. c. E. R. Jones and C. A. Cramers. Elsevier, Amsterdam (1977) 6. Freeman, R., F. K. Gould, P. R. Sisson, and N. F. Lightfoot: Strain differentiation of capsule type 23 penicillin-resistant Streptococcus pneumoniae from nosocomial infections by pyrolysis mass spectrometry. Lett. App!. Microbio!.13 (1991) 28-31 7. Freeman, R., P. R. Sisson, N. F. Lightfoot, and J. McLauchlin: Analysis of epidemic and sporadic strains of Listeria monocytogenes by pyrolysis mass spectrometry. Lett. App!. Microbio!'12 (1991) 133-136 8. Goodfellow, M., J. Chun, E. Atalan, and J.J. Sanglier: Curie-point pyrolysis mass spectrometry and its application to bacterial systematics. In: Bacterial Systematics and Diversity, pp.87-104, eds. F. G. Priest, A. Ramos-Cormenzana, and B. Tindall. Plenum Press, London (1994) 9. Gould, F. K., R. Freeman, P. R. Sisson, B. D. Cookson, and N. F. Lightfoot: Inter-strain comparison by pyrolysis mass spectrometry of Staphylococcus aureus nosocomial infection. J. Hosp. Infect. 19 (1991) 41-48

v.

v.

Typing of Listonella anguillara by PyMS

251

10. Gower, J. c.: A general coefficient of similarity and some of its properties. Biometrics 27 (1971) 857-874 11. Huff, S. M., H. L. C. Meuzelaar, D. L. Pope, and C. R. Kjeldsberg: Characterisation of leukemic and normal white blood cells by Curie-point pyrolysis mass spectrometry. I. Numerical evaluation of the results of a pilot study. J. Anal. Appl. Pyrolysis 3 (1981) 95-110 12. Kay, H. E., H. L. C. Coutinho, M. Fattori, G. P. Manfio, R. Goodacre, M. P. Nuti, M. Basaglia, and J. E. Beringer: The identification of Bradyrhizobium japonicum strains isolated from Italian soils. Microbiology 140 (1994) 2333-2339 13. Larsen, J. L. and J. E. Olsen: Occurrence of plasmids in Danish isolates of Vibrio anguillarum serovars 01 and 02 and association of plasmids with phenotypic characteristics. Appl. Environm. Microbiol. 57 (1991) 2158-2163 14. Larsen, J. L., K. Pedersen, and 1. Dalsgaard: Vibrio anguillarum serovars associated with vibriosis in fish. J. Fish Dis. 17 (1994) 259-267 15. MacDonell, M. T. and R. R. Colwell: Phylogeny of the Vibrionaceae, and recommendation for two new genera, Listonella and Shewanella. Syst. Appl. Microbiol. 6 (1986) 171-182 16. MacDonell, M. T., D. G. Swartz, B. A. Ortiz-Conde, G. A. Last, and R. R. Colwell: Ribosomal RNA phylogenies for the vibrio-enteric group of eubacteria. Microbiol. Sci.3 (1986) 172-179 17. Maniatis, T., E. F. Fritsch, and J. Sambrook: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, New York (1982) 18. Olsen, J. E. and J. L. Larsen: Ribotypes and plasmid contents of Vibrio anguillarum strains in relation to serovar. Appl. Environm. Microbiol. 59 (1993) 3863-3870 19. Orr, K., F. K. Gould, P. R. Sisson, N. F. Lightfoot, R. Freeman, and D. Burdess: Rapid inter-strain comparison by pyrolysis mass spectrometry in nosocomial infection with Xanthomonas maltophilia. J. Hosp. Infect. 17 (1991) 187-195 20. Schiewe, M. H., T. J. Trust, and J. H. Crosa: Vibrio ordalii sp. nov.: A causative agent of vibriosis in fish. Curro Microbiol. 6 (1981) 343-348 21. Sisson, P. R., R. Freeman, F. K. Gould, and N. F. Lightfoot: Strain differentiation of nosocomial isolates of Pseudomonas aeruginosa by pyrolysis mass spectrometry. J. Hosp. Infect. 19 (1991) 137-140 22. S0rensen, U. B. and J. L. Larsen: Serotyping of Vibrio anguillarum. Appl. Environm. Microbiol. 51 (1986) 593-597 23. Sneath, P. H. A. and R. R. Sokal: Numerical Taxonomy: The Principles and Practice of Numerical Classification. W. H. Freeman, Baltimore (1973) 24. West, P. A., J. v. Lee, and T. N. Bryant: A numerical taxonomic study of species of Vibrio isolated from the aquatic environment and birds in Kent. J. Appl. Bacteriol. 55 (1983) 263-282 25. Windig, w., P. G. Kistemaker, and J. Haverkamp: Interpretation of a set of pyrolysis mass spectra by discriminant analysis and graphical rotation. Anal. Chern. 55 (1983) 387-391

B. Austin, Department of Biological Sciences, Heriot-Watt University, Riccarton, Edinburgh, EH 14 4AS, UK