System. App!. !--.1icrobio!' 18,285-302 (1995) © Gustav Fischer Verlag, Stuttgart· Jena . New York

Identification and Typing of Vibrio anguillarum: A Comparison of Different Methods B. AUSTIN" M. ALSINA b , D. A. AUSTIN', A. R. BLANCHh, F. GRIMONT", P. A. D. GRIMONTc, J. JOFRE b , S. KOBLAVI c , J. L. LARSEN d , K. PEDERSEN d , T. TIAINEN d , L. VERDONCKc, and J. SWINGS" From the Department of Biological Sciences, Heriot-Watt University, Riccarton, Edinburgh EHI4 4AS, Scotland'; Departament de Microbiologia, Universitat de Barcelona. Av. Diagonal 645, 08028 Barcelona, Spain"; Centre National de Typage Moleculaire Enterique, Unite des Enterobacteries, Institut Pasteur, 28 rue du Docteur Roux. F-75724 Paris Cedex \5, France\ Department of Veterinary Microbiolo~, Laboratory of Fish Diseases, Royal Veterinary and Agricultural University, Bulowsvej 13, DK-1870 Frederiksberg C, Denmark ; and Laborarorium voor Microbiologie, Faculteit Wetenschappen, Universiteit Gem, K. L. Ledeganckstraat 35, B-9000 Gent, Belgium
3 h at 37°C with Bgn (Boehringer, Mannheim, Germany) (Grimont et aI., 1989). Restriction fragments were separated e1ectrophoretically through 0.8°;', (w/v) agarose (Appligene, IIIkirch, France) in Tris-borate buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA buffer, pH 8.3) for 16 h at 1.5 VI cm, before transfer to nylon membranes (Hybond-N; Amersham International, Amersham, England) using a VacuGene System (Pharmacia). The DNA was fixed (15 min at 80°C) before hybridization with an acetylaminofluorene (AAF)-labelled ribosomal probe (Eurogenrec, Seraing, Belgium), as described by Grimont et al. (1989). After blocking, the membranes were incubated at room temperature for 2-3 h with mouse anti-acetylaminofluorene monoclonal antibodies, and washed three times for 10 min in buffer (0.1 M maleic acid, 0.15 M NaCI, 0.3% Tween 20; pH 7.5) to eliminate excess antibody. Antimouse IgG alkaline phosphatase-conjugated anribody (1 Ilg/ml) was added for 1 h, then, the systems were washed three times for 10 min. Revelation was in 0.1 M Tris HCI, 0.1 M NaCI, 50 mM MgCI 2 ; pH 9.5, after addition of nitroblue tetrazolium (Sigma) and 5 -bromo-4-chloro3-indolyl phosphate (Sigma) as substrate. Purplish-blue bands appeared within a time span of a few minutes to one hour. The data were examined using the Dice (SAB) coefficient with clustering by single linkage (Sneath and Sokal, 1973; Priest and Austin, 1993). Serotyping. The preparation of absorbed pol}'donal antisera and O-antigen have been described previously (Maniatis et aI., 1982; Larsen et aI., 1994). A loopful of undiluted antiserum was mixed with an equal volume of O-antigen on a glass microscope slide, and agglutination reactions were recorded after visual inspection.

Determination of lipopolysaccharide (LPS) profiles. LPS was extracted after enzymic digestion of lysed cells with proteinase K (Hitchcock and Brown, 1983), polyacrylamide electrophoresis using Mini-Protean" Ready Gels (Bio-Rad, J-krcules, California, USA) without 5DS, and with LPS silver stain (I\.tttelberger and Hilbink,1993). Determination of outer membrane protein (OMP) profiles. OMP was extracted from young cells (overnight broth cultures) which were disrupted by alternate freezing and thawing in TrisHCI (10 mM) buffer prepared in 0.3% (w/v) sodium chloride with the sarkosyl technique (tambert. 1986), polyacrylamide electrophoresis using Mini-Protean" Ready Gels (Bio-Rad) without 5DS, and later with Coomassie Brillianr Blue stain. The gels were stained with protein-silver stain (Merril et aI., 1981), as necessary. Gas chr01l1atogr,lphic analysis of cellular fatty acid methyl esters (FAMEs). A subset of 147 V. anguillartlm cultures was examined by this method. Quanritative analysis of cellular fatty acid compositions was performed using the gas-liquid chromatographic procedure, as described by De Boer and Sasser (1986). Cultures were grown for 24 h at 28°C on TSA supplemented with 1.5% (w/v) sodium chloride. Approximately 70 mg of cells were added to 1 ml of 3.75 M NaOH in 50% (v/v) aqueous methanol, and heated for 30 min in a boiling water bath for saponification. Methylation was achieved by adding 2 ml of 6 N hydrochloric acid in 50% (v/v) aqueous methanol and heating for 10 min at 80°C. After cooling to room temperature, fatty acid methyl esters (FAMEs) were extracted with a 1: I mixture of hexane and methyl-iso butylether. The FAMEs were analysed with a Hewlett-Packard model 5898A gas chromatograph and identified using the Microbial Identification System software package (MIS version no. 3.7, obtained from Microbial ID Inc., Newark, Delaware, USA). Clustering was by numerical analysis using the Euclidean distance coefficient and the unweighted pairgroup method of averages (UPGMA; Sneath and Sakal, 1973; Priest and Austin, 1993). Clusters were delineated at Euclidean distance 69. Plasmid profiling. Overnight cultures in Luria-Berrani broth base (LB; Gibco, Paisley, Scotland) or veal infusion broth (Difco) supplemented with 0.5% (w/v) sodium chloride were used for the determination of plasmid profiles. The plasmids were extracted from the bacteria by the method of Kado and Liu (1981). Plasmids were separated by gel electrophoresis in 0.6-1.0% agarose gels (SeaKem GTG; FMC, Bio Products, USA) in T AE buffer, at pH 8.0. The size of the plasmids was determined by the length of migration, according to the method of Rochelle et al. (1985). All cultures were examined at least four rimes. Examination of ClIltllres by the API 20E rapid idmtifiwtion s)'stem. Inocula were prepared from 24-28 h cultures incubated at 20 DC on marine 2216E agar (Difco). Cells were harvested in phosphate buffered saline (PBS) supplemenred with 2% (w/v) sodium chloride, and the suspensions adjusted with fresh PBS to achieve a standard opacity equivalent to McFarland No.3. The API 20E rapid identification sysrems (BioMerieux, Marcyl'Etoile, France) were inoculated using the modification for marine bacteria described by Kent (1982) and Grisez er al. (1991) with incubation at 20°C for 24 and 48 h. BIOLOG fingerprints. Strains were grown for 24 h at 25 DC on brain heart infusion agar (Difco) supplemented with 1.5% (wI v) sodium chloride. Inocula were prepared in 1.5% (w/v) sodium chloride, and the cell density standardized (0 berween 0.261 and 0.3 OD using a spectrophotometer at 590 nm. Each well in the BIOLOG GN microplate (Biolog Inc., Hayward, CA., USA) was inoculated with 150 fd of cell suspension, and the microplates were incubated at 25°C for 24 h. Changes in colour were measured using a Multiscan Multisoft filter photometer (Labsystems, Helsinki, Finland) at 590 nm. The BIOLOG profiles were com-

Identification and Typing of Vibrio angllillarum pared by numerical analysis (PhenoCompar, L. Vallterin) with similarities calculated by the Pearson product moment correlation coefficient between strains_ The strains were grouped by UPGMA. Phenotypic clusters were delineated at 80% r. Differentiating characters between pairs of clusters were calculated using the BIONUM software package (L. Vallterin et al.), and were defined as those tests yielding >90% and < 10% of positive responses, respectively. In order to extract differentiating characters, optical density values were recalculated between maximum and minimum values, whereby the cut-off factor to obtain positives and negatives was set at 30%. The internal homogeneity within each delineated cluster was calculated using the same software package. Biotyping. Carbon source utilization tests were carried out using Biotype-IOO strips (BioMerieux; La Balme-Ies Grones, France) that contained 99 pure carbon sources. Volumes (60 ml) of Biotype medium No.1 were supplemented with 2% (w/v) of sodium chloride and 100 mM MgCl b and inoculated with 2 ml amounts of a 100- Klett unit suspension of bacterial cells in 2 % (w/v) sodium chloride. The strips were completely filled so as ro obtain a slightly convex meniscus, and incubated at 30°C for 2 and 4 days. The cupules were examined visually for the presence of growth, and the data were examined by the Jaccard (S)) coefficient with UPGMA clustering (Sneath and Sokal, 1973; Priest and Austin, 1993), using the Recognizer program (Institut Pasteur Taxolab, Paris, France). Fish pathogenicity experiments. Bacterial cultures were grown overnight at 22°C in tryptone soya broth (Oxoid) supplemented with 1% (w/v) sodium chloride, centrifuged at 5,000 x 'g' for 10 min and resuspended in 10 ml volumes of 0.9% (w/v) saline to approximately ION cellslml. Ten fold dilutions were prepared to 10 4 cells/ml. A salmonid fish model (Atlantic salmon [Salma salar]) was used to assess pathogenicity, as follows: Groups of 10 fish wre infected by intraperitoneal injection with 0.1 ml volumes of the washed bacterial suspensions to achieve doses of 104 _10 6 cells/fish. The infected animals were maintained for up to 14 days in covered polypropylene tanks supplied with dechlorinated, aerated static freshwater (the water was changed daily) at a temperature of -15°C. Dead and moribund fish were removed, and subjected to standard bacteriological and pathological examination (Austin and Allstin, 1989). Any survivors at the end of the experiment were sacrificed and examined, as above. The disease signs were recorded, and attempts made to recover and confirm the identity of the pathogen by the methods described by Allstin and Austin (1993). Test reproducibility. The test results for 20 duplicate cultures were compared for all methods, and an estimate of individual test variance (S,2) was calculated (formula 15; Sneath and Johnson. 1972). The average test variance was calculated to obtain a pooled variance (52), which was used to estimate the average probability (p) of an erroneous test result (formula 4; Sneath and Johnsorl, 1972).

2. Ribotypillg The majority (236/260 = 91 %) of the cultures were ascribed to a single ribotype cluster defined at or above the 80% similarity level (Table 1; Fig. 1). The cultures in this cluster were characterised by three groups of bands. Thus, there were 2-7,4-9 and 2 bands between 10 kb and 20 kb, 4 kb and 8 kb, and 2 kb, respectively. The remaining 24 cultures were recovered in 19 separate clusters. Of these, clusters 2, 3, 4, 6, 8, 9, 10, 11, 12, and 15-20 comprised single isolates; clusters 5, 13 and 14 each contained two isolates, and cluster 7 comprised 6 isolates (Table 1; Fig. 1).

%SIMILARITY

40 !

100

rlU~rn ~O Of mm ij~ oour~ IDENTITY

mmrt

236

VIB66

1

2

6 7

6

8 9 \0

1. Test reproducibility Inclusion of duplicate isolates in the study enabled experimental test error to be estimated. The average probability (p) of an erroneous result was 5_6%, calculated from the pooled variance (52 = 0.0558) of all the methods scored for the duplicate cultures. Values of ranged from zero for some of the test regimes, e.g. ribotyping, through 0.03 for OMP and pathogenicity, 0.05 for serotyping, 0.15 for LPS profiles, 0.44 for the API 20E system, and 0_9 for BIOLOG.

5,z

12 \3 14 IS 16

17

18 11)

20

v. anguillarum

VIB 225

2

3 4 5

11

Results

287

2 2

VIB 76 VIB 189. VIB 190 VlB 239 VIB 192. VIB 193. VIB 195. VIB 210. VIB 21I, VIB 223 VIB 224 VlB 212 VIB 194 VlB 215 VIB 23 VlB 186, VIB 213 VlB 144. VIB 145 VlB 187 VlB 218 VIB 191 VIB 214 VIB 185 VlB IR4

Fig. 1. Simplified dendrogram based on the examination of ribotyping data by the SAS coefficient and single linkage clustering.

HWU VA75

NCMB 828 NCMB 572 UB 01/91 UB 28/89 UB 76/91 UB 109/90 UB 178/90 UB 191/90 UB 207190

VIB 30

VIB 31 VIB 32 VIB 34 VIB 35 VIB 36 VIB 37 VIB 39 VIB 40 VIB41

8

Sp. aurata, Italy

19

1 1

2 2

3

2

2

3

4

1

1 1

2

08

2

2

2 2

1

01 02b 01 crossreactions between 02a&05

10 10 10 2

2 1 3 3 2 3 2 4 2 2

3

.3

1 3 2 3 3 1 1 3

2

OMP profile

1 I 1 24 24 24 1

2 2

Sa. safar, U.K. O. mykiss. Japan ~Iugilidac, Italy D.labrax, Iraly Mugilidae, Italy Sp. aI/rata, Italy D.lahrax, Ital), D.labrax, Ital),

1 I I 12

023 02b 01 01 01 01 01 01 01 01 05 05

9 10 1

8

2 3 4 5 6 7

LPS profile

reactions between 02a&05 02a 01 01 01 01 04 01 01

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 28 29

V1B VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB

02a 03 04 05 06 07 08 09 10 01

01

Sero· group

(ross-

11

Ribotype cluster

Sp. aurata, Greece

ATCC 43306 ATCC 43307 ATCC 43308 ATCC 43309 ATCC 433\0 ATCC 43311 ATCC 43312 ATCC 43313 ATCC 43314 AVL 90-9-22

VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB

AVL 89-2-54 AVL 68-28-C HWU 44 HWU 48 HWU 49 HWU 50 AVL 27.2 AVL 2&.2 AVL 28.3 AVL 28.4 AVL 20/90 AVL 33/91 AVL 18/91 HWU63 HWU 64 HWU V02 HWU VA73

Ollcorh)'IIc/ms lIIykiss, Denmark Gadus morima, Denmark O. mykiss, Denmark G.lllorhua, Denmark G. morhua, Denmark G.lllorhua, Denmark Anguilla anguilla, Denmark G. morl",a, Denmark G. morhua, Denmark C. morhua, Denmark Dicentrarchus labm.>:, Greece D.labrax, Greece D.labrax, Greece O. mykiss, Italy D.lahra.>:, Greece D.lahrax, Greece D. lahrax, Greece O. mykiss, Denmark O.III),kiss, Denmark O. m),kiss, Denmark O. m),kiss, Denmark D.labrax, Greece Sparus tlurata, Greece Sp. aurata, Greece Sallllo salar, Chile fish Sa. salar, Norway D.lahrax, Greece

ATCC 43305

VIB 1

2 3 4 5 6 7 8 9 10

Source and/or country of origin

Laboratory Collection'/ reference no. strain no.

Table 1. Bacterial cultures

2

2

2

1 3 2 2

2

0

0 67 kb 67 kb 67 kb 67 kb 0 67 kb 78, 84 kb

0

kb

1

8 8 1

324752457 304752757 324752457 324752457 324752457 324752757 324752447 324752557 324712457

324712657

304751647 324753757 324773757 324752457 304772457 304753757 304673757 304773757 324773757 724753757 534622557 534622557 534622557 304772757 324712757 304752757 324712657

0 0 67 kb 67 kb 67kb 67 kb 50,67 50,67 50,67 50,67 46,76 76 kb 50,70 67 kb 0 67 kb 0

2

304752556

API 20E profile

324752757 324752757 324752757 304712557 304752757 324752757 324752557 124712557 324752757 304773757

I I I 8

BIOLOG profile

0 0 0 0 0 0 0 0 0 67 kb

67 kb

kb kb kb kb kb

Plasmid conterl(

2 2 2 2 2 2 2 2 2

2

FA:VIE profile

25

2 2

+

+ (weak) + + +

+

+"

+

+

+

7

9

32

+ + + + + + + + + + + + + + + + +

+ + + + + + + + + (weak) +

+

Pathogenicity

24 8 23 I 4 2

27 23

15 16

25 15 30

Biotype profile

tv

:-

'"

~

n

5'

~

?' >c

00 00

D.labrax. Greece D./abrax, France D./abrax, Italy O. mykiss, Finland O. mykiss, Italy

LMG 13186 LMG 13187 LMG 13188 RVAU 87-9-117 RVAU 1692

VIS VIB VIS VlB VIB

82 83 84 85 86

D./abrax, Greece

U.S.A. Japan P/ecogloss/4S a/tive/is. Japan P. a/tive/is, Japan

UB A024 UB A044 UB A055 UB A056 UB A07S UB A085 LMG 4411 LMG 4437T LMG 11684 LMG 10939 LMG 12098 LMG 12099 LMG 12100 LMGI21O! LMG 12102 LMG 13185

VIB 65 VIB 66 VIB 67 VlB 68 VIB 69 VIB 70 VIB 71 VlB72 VIB 73 VlB 75 VIB 76 VIB 77 VIS 78 VIS 79 VIS 80 VIB 81

D./abrax, Italy D./abrax, Italy D./abrax, Italy D./abrax, Italy D./abrax, Italy D./abrax, Iral)' D.labrax, Italy D./abrax, Italy fi.h, Italy D./abrax, Italy D./abrax, Italy Sp. aurata, Italy D. /abrax, Italy D./abrax, Italy D./abrax, Iraly D./abrax, Italy D.labrax, Italy Sp. aurata, Italy D.labrax, Italy D./abrax, Italy D./abrax, Italy Scophtha/mus maxim/IS, Spain Sc. maximus, Spain Sc. maxim/IS, Spain Sc. maximus. Spain Sc. maximu5, Spain Sp. allrata, Spain rotifer, Spain Scotland G. morhua, Norway

UB 217/90 UB 261191 UB 295/91 UB 327/90 UB 342191 UB 393191 UB 397/90 UB 417/90 UB 498/90 UB 505/91 UB 258/91 UB 574/91 UB 578/90 UB 60119) UB 861/89 UB 904/89 US 909/89 UB 910/90 UB 960/90 UB 1032190 UB 1075/90 UB A023

Source andlor country of origin

VlB 42 VIB 44 VIB 45 VIB 46 VlB 47 VlB 48 VlB49 VlB 50 VIB5! VIB 52 VIB 53 VIB 54 VIS 55 VlB 56 VlB 57 VIB 58 VIB 59 VIB 60 VIB 61 VIB 62 VIB 63 VIB 64

Laboratory Collection' 1 reference no. srrain no.

Table I. Continued

I 1

1 1 4

3

1

Ribot)'pe clusrer

01 02a cross reactions VaNT2 01 01 01 01

02a 02a 02a 01 02a 02a,b

02a

01 01

01

01 03 03 03 03 01

02a 07 02a 01 01

02a 02a

0) 04 01 01 01

Serogroup

43 1

14 2 2 I 2 2 14 1 2 43

5

8 1 1 2

1

1 13 1 42 42 42 42

2 1 4 I 1 1 42 2 42 46 2 7 2

LPS profile

2 2 2 2

1

6 9 2 9 1

1

2 11 2 2 6 2 3 2 2 2

2 1 12 2 2 2 3 2 3 1 2 2 2 2 2 6 1 13 2 2 2 2

OMP profile

2 2 2

2 2 2 2 2 2 2 2 2 2

2

2

2

2

FAME profile

46,68, 83 kb 68,83 kb 0 67 kb 67 kb 67 kb

0

50 kb 77 kb

0

67kb 0

0

0 0

0

67 kb 0 67 kb 67 kb 0

0 0 0 0 0 0 26 kb

0 0 0 0 0 0 67,84 kb 67 kb

0

0 35,60 kb 0 67 kb 67 b 67, 84 kb

Plasmid content

4 4

4 1

-

304752757

324772577 304752457 304772457 324772657 304752757

324752456 324752757 324752647 320752756 304752547 324752557 304712457 304752657 304732657 324772557

324752757 304652556 324772657 304652457 324752757

324752757 324752757 324752577 324752457 324752457 324752657 324752767 324752446 324752757 324752577 324732657 324752557 324732657 324752657 324752757 324752757 324752457 324752777 324752657 304752757 734752757 304752757

BIOLOG API20E profile profile

10 1

16

19

24 15

23

23

24 23

31 27

15

6

15 16

23

1

1,3

Biot)'pe profile

+

+ +b

+ (weak)

+

+

+

+ + + + + + +

+ + +" + + +"

+

+

+ (weak) + +"

+

+ + + +

+

+" +" + + +

+

+ +" +" +b

Pathogenicit)'

00 \0

1'-'

...~ ~

~::

t:.

a


LMG LMG LMG LMG LMG LMG LMG LMG

VIB VIB VIB VIB VIB VIB VIB VIB

LMG 13230

LMG LMG LMG LMG LMG LMG LMG

VlB 203

204 205 206 207 208 209 210

211 212 213 214

VIB VIB VIB VlB VIB VIB VIB

VIB VIB VIB VIB

LMG 13238 LMG 13239 LMG 13240 LMG 13241

13231 13232 13233 13234 13235 13236 13237

LMG 13227 LMG 13228 LMG 13229

VIB 200 VIB 201 VIB 202

13219 13220 13221 13222 13223 13224 13225 13226

LMG 13216 LMG 13217 LMG 13218

VIB 189 VIB 190 VIB 191

192 193 194 195 196 197 198 199

LMG 13215

VIB 188

water, Greece

water, Greece D./abrax, Greece water, Spain

18

13

9

7

1 7

6 6 6 19 20 19 18

6

19 21 20 21 0 I 1

17 17 18

17

I 19 16 19 17

LPS profile

04

26

weak cross reactions 05 (weak) 20 19 19

06 (weak) 06

06

Sp. aurata, Spain D.labrax. Spain water, Spain rotifers, Spain rotilers, Greece D./abrax, Greece D. /abrax, Greece water, Greece

VaW7 VaNT7 VaNT7

Sp. aurala. Greece Sp. aumta, Greece D./abrax, Greece

VaNT1

cross reactions

reaction~

weak cross reactions cross

04

01 01 01 01 01 01

Serogroup

VaNT7 VaNT7 VaNT7

7 7 10 7

5 5 17

13 15

19

29

Ribotype cluster

rotifers, Greece rotifers, Greece rotifers, Greece water, Greece

Arlemia, Spain

rori fers, Greece

rotifers, Greece rotifers, Greece

rotifers, Greece rotilers, Greece roti fers, Greece

D./llbrax, Spain

O. mykiss, Canada O. m~'kiss, Canada O. Inykiss, Canada O. nerka, Canada SII. sa/ar, Canada O. mykiss. Canada Artemia, Spain Artemia. Spdin Artemia, Spain D./abrax, Spain

RVAU VAI2 RVAU VA41 RVAU VA43 RVAU VA61 RVAU VA70 RVAU VA72 LMG 13211 l.MG 13212 LMG 13213 LMG 13214

VIB VIB VIB VIB VIB VIB VIB VIB VIB VIB

175 176 178 180 181 182 184 185 186 187

Source and/or country of origin

Llboratory Collection'/ reference no. strain no.

Table I. Continued

2 9 17 15

2

2 2

6 6 6 2

3 3 2 2 2

2

2

2

2

2 2 2 2 2 2 5 16 5 3

OMP profile

2

2 2 2

2

2

2

2 2

3

2

2 2 2

2 2 2 2 2 2 2 2

2 2 2

2

2 2

2

2

2

fAME profile

°

120 kb 37 kb 4.4 kb 49 kb

16 kb 16 kb 16,49, 68 kh 16 kb 16 kb 16,49, 68,80 kb 3.2,5.3, 50, 70 kb 0 0 0 36 kb 36 kb 0 0

-190 kb -200 kb 0 0

67 kb 67 kb 67 kb 67 kb 67 kb 67 kb 0 32 kb 4.4 kb 2.2.3.4, 4.2,5.8 6.8, 53 kb 3.3, 123 kh 88 kb 88 kb 50 kb

Plasmid content

3

6 6

6 6 6 6

6 6 6 6 1 I

3 3

.3

5

5

.3

1 3 8

BIOLOG profile

324612556 324610556 724413756 515612557

324752657 324752757 324752657 324610556 304612556 324610556 324610556

324752757

324772557 324772557 324772557

304610556 324610556 324610556 324610556 324772557 324772557 324772557 324772557

324453757 124413757 324412757

100412456

324772757 324772757 304772757 324772757 324772757 324772757 304413757 414712556 724413756 000411456

API20E profile

22

22 22

16

10 10 11

11

Biotype profile

+"

+"

+

+ (weak)

+ + + + + +

+

+ + +

+ + + + + + + +

+ +

+

+

+ (weak) +" + + + + + + + +

Pathogcnicit)'

:-

~

.,

5'

~

?'

;..

t"

I'" -0

D.labrax, France Sc.max;mIlS, France sea perch, Greece D.labrax, Morroco sea perch, France Sc. maxim liS, Spain O. nerka, U.S.A. O. mykiss, Japan unnamed fish, U.K. Sa. trutta, Iraly salmonid fish, Finland O. mykiss. Finland O. mykiss, Finland O. mykiss, Finland O. mykiss, Tasmania O. myk;ss, Tasmania O. mykiss, Tasmania

CIl'1l164 NCMB 1875

NCMB 571 Lee, VL3355

VIS 242 VIB 243

244 245 246 247 248 249 250 251 252 253

254 255 256 257

VIB VIB VlB VlB VIB VIB VIB VIB VIB VIB

VIB VIB VIB VIB

Carson Carson Carson Carson

89/4046 891066

8912873

89/3620

RVAU 1989 RVAU 1990 RVAU 1990 RVAU 1990 Carson 8513526 Carson 8613831-2 Carson 85/3954-4

HWU BA136

D.labr

t"

Identification and Typing of Vibrio anguillarum

3. Serology

The majoriry (184/260 = 71 %) of the cultures, received as V. anglli//arum, were recovered in serogroups .01 010, of which serogroup 01 with 95 isolates dommated (Table 1). Twenty isolates were considered to represent five previously undescribed serogroups, and are referred to as VaNTl VaNT2, VaNT4, VaNT5 and VaNT7. A further con~plication concerned 16 cultures, of which 15 reacted with antisera representing two serogroups, and one was self-agglutinating (Table 1). The remaining 40 cultures did not react with any of the antisera used or cross reacted with more than two different antisera, and were regarded as untyped. It is noteworthy that the majority (24/40 = 60%) of the untyped isolates were not denved from fish but from anemia, rotifers, sediment and water. Apart from serogroups 01 and 02a and 02b, which accommodated 134 (= 51 %) of the isolates from many fish species and geographical locations, smaller numbers of cultures were associated with serogroups 03 - 010 and VaNTl - VaNT7. Again, serogroups 04 and 05 with 9 and 11 isolates respectively, contained cultures obtained from a wide range of countries and habitats. In cont~ast, the 12 isolates recovered in serogroup 03 were obtamed only from fish in Denmark, Italy and Japan, and water from Denmark (Table 1). Isolates labelled as serogroups 06 and 08, with 6 and 4 isolates respectively, were obtained from many habitats in Europe. In particular, the isolates allocated to serogroup 06 were derived from Denmark and Spain. The cultures regarded as serogroup 08 were obtained from Denmark and Italy. Five isolates, recovered from fish (ayu [Plecoglossus altivelis], eel [Anguilla anguilla], sea bream [Sparus aurata] and turbot [Scophthalmus maximus] in Denmark, Italy and Japan, belonged to serogroup 07. Three isolates from cod (Gadus morhua) obtained from Denmark were equated with serogroup 09 (2 isolates) and 010 (1 isolate) (Table 1). All of the cultures, typed in serogroups 01, 02b, 03, 06,07,08,09,010, VaNT2 VaNT5 and VaNt:7, were recovered in cluster 1 as defined by ribotyping. It IS relevant to note that VIB 76, which was regarded as representing serogroup 02a, was recovered as ribotype cluster 4; two isolates (VIB 185 and VIB 214) from serogroup 04 comprised riborype clusters 19 and 18, respectively; three cultures (VIB 23, VIB 211 and VIB 224) of serogroup 05 were defined as ribotype clusters 12, 7 and 8, respectively; and VIB 195 (serogroup VaNTI) and VIB 215 (serogroup VaNT5) were recovered in ribotype group 7 and 11, respectively. The remaining cultures ~ere unrypeable or demonstrated serological cross reactions between several serogroups. Yet, these were mostly found within ribotype cluster 1. 4. Lipopolysaccharide (LPS) profiles

There was good agreement between the composition of serogroups 01, 02, 04, 06, 07, 08, 09 and 010 and LPS profiles. Thus, all of the isolates in serogroup 01 were homogeneous in terms of their LPS profile, being equated with LPS profile 1. Interestingly, 12 isolates in LPS group 1 were recovered in serogroup 03, 07, VaNT1 and

295

VaNT7 and as untypable. Serogroup 02, which was subdivided into 02a and 02b by serology, was regarded as a single LPS grouping, namely LPS profile. 2. In additio~, some of the isolates (3130 = 10%), whICh clustered 111 serogroup 02a, were equated with ther LPS groups, namely LPS profile 14, 21 and 42. Five Isolates, which were ~quated with LPS profile 2, were recovered in serogroups OS, 06, VaNT2 and as untypable. It should be emphasized that these apparently ab.berant strams were recovered in cluster 1, as defined by nbotypmg (Table 1). Cultures ryped as serogroup 04 were recovered principally in LPS profile 4. Again, a minority (2/9 = 22%) were labelled as LPS profile 16 and 26. Of these, VIB 185 (serogroup 04) was recovered as a distinct single member cluster by ribotyping. Seven isolates, recovered in serogroup 03 (VIB 60-63, VIS 132, VIB 148 and VIB 272), were considered to represent LPS profiles 1, 42 and 45 (Table 1). Four cultures (VIB 22, VIB 23, VIS 211 and VIB 224), regarded as representatives of serogroup 05.' were ~quated with LPS profiles 20, 21 and 24. Indeed, It IS qu~stlonable whether these isolates are bona fide representatives of V. anguillarum. Overall most cultures with LPS profile 1 and 2 corresponded t~ serogroups 01 and 02a and b, respectively. Nevertheless, LPS profile 1 was also recorded for cultures in serogroup 03, 07, VaNTI and VaNT7. Similarly, LPS profile 2 was noted among cultures in serogroups OS, O~ and VaNT2. Also, the distribution of the other LPS profiles traversed serogroup boundaries. Thus, it would appear that typing systems based on LPS profiles do not coincide exactly with serogroups. However, some serogroups, e.? 01 and 02, demonstrated a higher level of homology 111 terms of the LPS profile than others. 5. OMP profiles There was negligible agreement between the composition of the serogroups and groupings resulting from examination of the OMP profiles. It is notable that serogroup VaNT7, in which the component organisms ~ere equated exclusively with OMP group 1, was charactenzed by the presence of a 34 kDa protein. Yet, overall, the 260 cultures were recovered in 19 groups based on OMP profiles. Cultures from serogroup 01 were equated with 4 groups based on the OMP profiles (Table 1). f:Iere, the dominant OMP profiles were 1, 2 and 3, which were characterized by the presence of dominant proteins of 33-35, 36-38 and 39-41 kDa, respectively. Again, the cultures recovered in serogroup 02 (a and b), 03, 04, OS, 06 07 and 08 were heterogeneous in terms of their OMP pro'fiIes. With serogroups 09 and 010 there wereinsuffi7 cient cultures to make meaningful detailed compansons. It should be emphasized that representatives of the dominant OMP groups appeared in a wide range of serogroups. For example, OMP profile 1 was found in cultures from serogroups 01, 02a, 02b, 03, 05,06, 07, 08, VaNTl, VaNT2 and VaNT7 (Table 1). It was apparent that most of the cultures, recovered in ribotype clusters 2-20, represented unique OMP profiles. For example, OMP profile 15 was only recorded from VIB 214, which represented ribotype cluster 10 (Table 1).

296

B. Austin et al.

6. FAME profiles

The major fatty acids were identified as C12: 0, C12: 0 30H, C14: 0, C16: 0 iso, C16: I w7c and C16 : 0 (Table 2). From the numerical analysis of the data, all isolates, except for four, were recovered in cluster 2 as delineated at Euclidean distance .6.9. The farry acid content of the isolates was very variable (Table 2). Within cluster 2, one group of isolates, which belonged to BIOLOG cluster 6, were distinct from the other cultures insofar as they exhibited a significantly higher C12: 01C12: 0 30H ratio. However, no correlation could be found between phenotypic or serological diversity and the FAME profiles. Cluster 1, which comprised two cultures, namely VIB 22 (serogroup 05) and VIB 24 (untyped by serology), were also phenotypically distinct and were recovered in BIOLOG cluster 8. Another two isolates, i.e. VIE 25 and VIB 204, comprised cluster 3, and were aberrant from other isolates insofar as they lacked C16: 0 iso, and possessed a higher C16 : 0 content. Nevertheless, these may be regarded as authentic V. anguillarum cultures when examined by other methods.

late) plasmids of 2.6-120 kb. Two cultures (33%) from serogroup 06 contained one (44 kb) or four plasmids (3.2-70 kb). Isolates from the new serogroups were most100

90

80

70

60

so

40

9

8 7

6

5 4

:3

7. Plasmid profiles Approximately half (137/260 = 53%) of the isolates carried plasmids. Isolates belonging to serogroups 03, 07,08,09 and 010 and VaNT4 were empty of plasm ids (Table 1). However, the majority (86/95 = 90%) of isolates in serogroup 01 contained the -67 kb virulence plasmid, equivalent to pJMl (Crosa et aI., 1980), either singly in most cases, or in association with one or two other plasm ids of 26-105 kb. The 67 kb plasmid was also found in VIE 243 and VIB 253, which demonstrated cross reactions between serogroups 01 and VaNTI, and VIB 269, which was labelled as VaNT1. Although most (32139 = 82%) of the isolates in serogroup 02 (a and b) were empty of plasm ids, 7 cultures contained one or two plasmids of 5->200 kb (Table 1). Again, the minority (2/9 = 22 %) of isolates recovered as serogroup 04 possessed single plasm ids of either 32 or 49 kb. From serogroup OS, 5 isolates (45%) contained one (4 isolates) or two (1 isoTable 2. Major fatty acids of Vibrio anguilla rum Fatty acid

Cluster: 2

C12:0 C12:0 30H C14:0 C16:0 iso C16:1OJ7c C16:0 l:F3 LF7

3.1-6.4 1.7-3.4 4.1-7.9 2.4-3.7 30.4-39 18.4-25.4 2.2-4.5 9.6--21.5

1.4-7.1 0.1-4.8 3.4-10.1 0.1-10.5 34.1-44.4 17.5-28.6 1.3-5.3 9.3-22.0

3 7.4-8.2 4.5-6.0 8.4-11.8 0 41.2-43.2 41.5-43.2 5.16-6.4 5.6-7.3

2

Ratios: C12:0/14:0 CI2:0/12:0 30H CI6:0/16:1 w7c

0.79 1.88 0.63

0.72 1.62 0.62

0.77 1.42 0.46

Fig. 2. Simplified dendrogram based on the examination of BIOLOG data by Pearson product correlation and UPGMA clustering.

Identification and Typing of Vibrio anguillarum

ly devoid of plasmids. However, two (40%) of VaNTl isolates comprised two (53 and 56 kb) and three (5.3, 11.5 and 67 kb) plasm ids, respectively. Again, 2 (40%) of the isolates from VaNT2 comprised single (11.5 kb) or two (68 and 83 kb) plasmids. A single (50%) isolate from VaNT5 contained a 39 kb plasmid. All the cultures from VaNT7 contained a 16 kb plasmid. In addition, VIB 199 and VIB 202 both contained a 49 and a 68 kb plasmid. The majority of the serologically untyped cultures were empty of plasm ids. Expressed in terms of ribotyping, it was apparent that cultures from ribotypes 3, 4, 6, 8, 10, II, 12, 14, 16 and 20 were empty of plasm ids. However, a wide range of plasmid profiles was recorded from some of the isolates (Table 1).

8. BTOLOG fingerprints Nine clusters were delineated at 78% r among the 260 isolates, using the Pearson correlation coefficient (Fig.2). All isolates tested in the BIOLOG GN-microplates were positive for oxidation of dextrin, glycogen, N-aceryl-Dglucosamine, D-fructose, D-glucose and maltose. Conversely, all isolates were negative for adonitol, D-arabitol, Table 3. Tests differemiating the clusters as defined after numerical analysis using the Pearson correlation coefficient and UPGlvlA clustering of BIOLOG fingerprints

erythrirol, L-fucose, lactulose, D-melibiose, D-raffinose, turanose, xylitol, formic acid, galactonic acid lactone, Dgalacruronic acid, D-glucosaminic acid, y-hydroxybutyric acid, p-hydroxy-phenylacetic acid, itaconic acid, a-ketovalerie acid, malonic acid, D-saccharic acid, sebacic acid, L-leucine, L-phenylalanine, L-pyroglutamic acid, D,L-carnitine, y-amino-butyric acid, urocanic acid, phenyl-ethylamine, putrescine, 2-amino-ethanol and 2,3-butanediol. The differentiating characters (which were set at a discriminating level of 89%) between each cluster has been given in Table 3. The clusters did not possess any unique single features permitting differentiation from all other groups. Nevertheless, the number of characters needed to be used to distinguish between each of the delineated clusters has been presented in Fig. 3. From this diagram, it is apparent that dusters I, 2, 8 and 9 could not be distinguished from each other. Clusters 2 and 9 each comprised two isolates deemed to be from serogroup 01. These clusters were similar to the majoriry of V. anguillarum isolates studied by other methods. In common with the results of riboryping, the majority of the isolates, particularly those regarded as belonging in serogroup 01 (= 90%) and including the type strain, were

Test:

Cluster: 2

3

4

5

6

Negative control Cyclodextrin N-acetyl-D-galactosamine L-arabinose Gentobiose m-inositol a-D-Iactose D-mannose j)-methylglucoside D-sorbitol mono-methyl-succinate Acetic acid D-gluconic acid Succinic acid Bromosuccinic Alaninamide acid L-alanine L-alanylglycine L-aspartic acid Glycyl-L-aspartic acid Glycal-L-glutamic acid L-histidine L-proline L-threonine Uridine Thymidine Glycerol D,L-a-glycerolphosphate Glucose-I-phosphate Glucose-6-phosphate

.

'+' and '-' correspond

297

7

8

9

+

+

.B 54 65 33 47 25

+ + +

85 76

+ +

85 73 65 83

+

68 20 53

+

30 85 63

+

84

+ +

+

50

+

83

50

+

+ + +

50 50

+ +

50

+ + + + +

+

17 83

+

50 17 17 50 17 17

20 20 20

+

60 80 80

20 40

33 67 67 67

+ + +

33 67

+ + + +

33 67 33 67 33

40 60

+ +

67

+ +

+ + + +

33 33

20

50 50 50

+

33

+ + +

+

60 20 87 20

33

+ + + + +

+ + + +

+ + + + +

20 80 80

+ +

60 80 20 20

+

80 60 80

+ + +

80 60

+ + + +

33

60

+

+ + + + +

67

+

67

+ + + +

60

+ +

+ + +

+ + + + + + + + + + + +

50

+ +

+ + + +

[Q >89% and