FEMS Microbiology Letters 69 (1990) 329-336 329 Published by Elsevier

FEMSLE 04032

Identification of catalase negative/weak Campylobacterjejuni from human blood and faecal cultures by numerical analysis

of electrophoretic protein patterns

R.J. Owen t, M. Costas ~, L. Sloss ~, A. Lastovica 2 and E. Le Roux 2

s National Collection of Type Cultures, Central Public Health Laboratory, London, U. K. and z Red Cross War Memorial Children's Hospita~ Cape Town, Republic of South Africa

Received 15 February 1990 Revision received and accepted 6 March 1990

Key words: CNW campylobacter; Electrophoresis; Protein pattern

1. SUMMARY

Twenty-eight isolates of catalase-negative/ weak (CNW) thermophilic campylobacters from human blood and faecal cultures were char- aeterized by one-dimensional ( l-D) high-resolu- tion SDS-PAGE of cellular proteins. A further 11 Campylobacter strains were included for reference purposes. Partial protein patterns were used as the basis for numerical analysis, which showed that all of the hippurate-positive strains had a high simi- larity to C. jejuni. Two subclusters were formed within C. jejuni corresponding to C. jejuni subsp. doylei (15 strains) and C. jejuni subsp, jejuni (4 strains). Most of the paediatric strains from South Africa were members of C. jejuni subsp, doylei. Hippurate-negative CNW thermophilic strains were identified as "C. upsaliensis". The analysis demonstrated that the catalase-negative C. jejuni strains were quite distinct from "C. upsaliensis" and that eleetrophoretie protein patterns provide

Correspondence to: R.J. Owen, National Collection of Type Cultures, Central Public Health Laboratory, Colindale Avenue 61, London NW9 5HT, U.K.

an excellent criterion for the identification of sub- species within C. jejuni.

2. INTRODUCTION

Members of the human pathogenic species, Campylobacter jejuni, are conventionally consid- ered to produce catalase [1] except for some strains recently classified within C. jejuni subsp, doylei [2]. Strains of C. jejuni subsp, doylei also differ from C. jejuni subsp, jejuni in fairing to reduce nitrate and in susceptibility to cepbalothin [2]. Strains of C. jejuni subsp, doylei are uncommon but have been isolated from the faeces of children in both Australia [2,31 and South Africa [4,51, and from gastric biopsies of adults in West Germany [61 and the United Kingdom [71. Their pathogenic role is uncertain. Catalase-negative strains resem- bling C. jejuni have also been described from paediatric blood cultures in South Africa [8], and from an enteritis case in Canada [9,10].

Lack of, or weak, catalase activity is a feature of another group of thermophilic campylobacters widely referred to as the CNW group or "C.

0378-1097/90/$03.50 © 1990 Federation of European Microbiological Societies

330

Table 1

Bacterial strains studied

Species/group Strain no. other Received from b Reference no. as received relevant designations" in Figs. 1 and 2

Field strains (faecal) Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW

Field strains (blood) Campylobacter CNW Campylobacter CNW Campylobacter CNW Campylobacter CNW

Reference strains C. jejuni subsp, doylei C. jejuni subsp, doylei C. jejuni subsp, doylei C. jejuni subsp, doylei C. jejuni subsp, jejuni C. jejuni subsp, jejuni C. coil C. laridis C. fetus subsp, fetus "C. upsaliensis" "C. upsaliensis "" "C. upsaliensis " "C. upsaliensis " "C. upsaliensis" "C. upsaliensis" "C. upsaliensis '" "C. upsaliensis " "C. upsaliensis" "C. upsaliensis" "C. upsaliensis " "C. upsaliensis" "'C, upsaliensis" "C. upsaliensis " "C. upsaliensis "

(EPJI) (EP-I) (EP-III) (EP-IV)

A601/89; 17.85 RCWMCH 20 A602/89; 72.85 RCWMCH 6 A603/89; 100.85 RCWMCH 13 A605/89; 110.85 RCWMCH 19 A606/89; 123.85 RCWMCH 1 A607/89; 139.85 RCWMCH 10 A608/89; 58.86 RCWMCH 12 A609/89; 117.86 RCWMCH 16 A670/88; Z90 H. Goossens 3 A655/88; M373 H. Goossens 4 A675/85; 111-84 RCWMCH 8

A604/89; 106.85 RCWMCH 10 A613/87; 1.86 RCWMCH 5 A603/87; 34.85 RCWMCH 2 A615/87; 28.85 RCWMCH 18

NCTC i 19513` NCTC 9 NCTC 11925 NCTC 14 NCTC 11848 NCTC 15 NCTC 12208 NCTC 7 NCTC 11168 NCTC 17 NCTC 11351T NCTC 21 NCTC 11366 T NCTC 22 NCTC 11352 T NCTC 39 NCTC 10842 T NCTC 38 NCTC 11926 NCTC 26 NCTC 12184 NCTC 27 NCTC 12183 NCTC 28 NCTC 12206 NCTC 30 NCTC 11840 NCTC 34 NCTC 11541 er NCTC 35 NCTC I 1540 NCTC 37 A651/88; L195 H. Goossens 23 A671/88; Y59 H. Goossens 24 A660/88; R218 H. Goossens 25 A629/88; A472 H. Goossens 29 A636/88; D391 H. Goossens 31 A681/88; D41C FJ. Bolton 32 A621/88; 778-85 T. Steele 33 A652/88; L196 H. Goossens 36

a Type strains are indicated by the superscript T or PT (provisional type), b RCWMCH, Red Cross War Memorial Children's Hospital, Cape Town, Republic of South Africa; H. Goossens, Sint-Pieter,

Universitair Ziekanhuis, Brussels, Belgium; NCTC, National Collection of Type Cultures, London, U.K4 F.J. Bolton, Public Health Laboratory, Preston, U.K,; 1". Steele, Institute of Medical and Veterinary Science, Adelaide, South Australia.

upsaliensis" [11,12]. T h i s o r g a n i s m , w h i c h i s

e m e r g i n g as a p o t e n t i a l h u m a n p a t h o g e n a s s o c i -

a t e d w i t h a w i d e s p e c t r u m of h u m a n i l l n e s s [10,13],

d i f f e r s f r o m C. j e jun i i n n o t h y d r o l y s i n g h ip - p u r a t e , T h e t w o s p e c i e s h a v e a l s o b e e n s h o w n to

b e d i s t i n c t g e n e t i c a l l y [12]. H o w e v e r , a p o s s i b l e

source of confusion in the identification of CNW campylobacters is that some strains of C. jejuni lack the ability to hydrolyse hippurate [14] and if catalase-negative they could be confused with "(7. upsaliensis ".

In a previous study of "C. upsaliensis" using numerical analysis of gel electrophoretic protein profiles, we identified three CNW strains of C. jejwff [8]. The purpose of the present study was to investigate further the use of high resolution sodium dodecyl sulphate-polyacrylamide gel dec- trophoresis (SDS-PAGE) of proteins combined with computerized analysis of band patterns to identify a larger set of CNW isolates from human faecal and blood cultures. We have also investi- gated whether protein profiles could be used as an additional criterion for identification of the two subspecies within C. jejuni.

331

phoresis was performed in 10% discontinuous SDS-polyacrylamide gels exactly as described previously [161 .

3. 4. Scanning and data analysis The stained protein patterns in the dried gels

were scanned by laser densitometry and the result- ing data analysed numerically to obtain estimates of similarity between all pairs of standardized traces. This was expressed using the Pearson prod- uct moment correlation coefficient (r). Strains were then clustered by the method of unweighed pair group average linkage (UPGMA). Full details of the methods used were described previously [16]. The analysis performed in this study was based on partial protein patterns in which the major bands (in the 38.0-47.8 kDa range) were excluded.

3. MATERIALS AND METHODS

3.1. Bacteria used The clinical isolates of CNW campylobacters

from faeces and blood specimens, and reference strains obtained from the National Collection of Type Cultures (NCTC) are listed in Table 1 with their alternative strain numbers and sources. "C. upsaliensis" is in quotations throughout to indi- Pate it is not a valid name, although :t has come into common usage. Strains were grown on 5% (v/v) defibrlnated horse blood agar. Cultures were incubated for 48 h at 37°C under microaerobic conditions (5~ 02; 5~ CO2; 2~o H2; 88~o N2) in a Variable Atmosphere Incubator (Don Whltley Sci- entific Ltd., Shipley, Yorks, England). Strains were preserved at -70°C on glass beads in Nutrient Broth No. 2 (Oxoid: CMG 67) containing 10~ (v/v) glycerol, and were lyophilized in 5~ (w/v) inositol serum.

3.2. Conventional bacteriological tests All the clinical isolates were examined using

previously described methods [151 in the following bacteriological tests: catalase production, hip- purate-hydrolysis and nitrate reduction.

3.3. SDS-po!yacrylamide gel electrophoresis Protein samples were prepared and electro-

4. RESULTS

4.1. General features of PAGE-protein patterns One dimensional SDS-PAGE of whole cell

protein extracts of the 39 cultures of Campylo- bacter produced patterns containing 40 to 45 dis- crete bands with molecular weights of 20 to 100 kDa (Fig. 1). Differences between strains were evident principally in the major protein bands with molecular weights in the range 38.0 to 47.8 kDa and these accounted for up to 15~ of the total protein content. Although there were few differences in the other protein bands within each taxon, there were differences between taxa.

4.2. Reproducibility The protein patterns of the Campylobacter

strains were highly reproducible both within and between gels. As a check on between-gel repro- ducibility, molecular weight protein standards were included in each of the four gels, and their calcu- lated average similarity was 95.5 + 1.9%. Samples of strain A609/89 (Ref. no. 16), which was also electrophoresed on each of the four gels, gave average similarities with a mean and standard deviation of 97.3 + 0.9%. The phenons established by numerical analysis proved to be extremely sta- ble when the computations were repeated using

332

. . . . . . .~_.~ ~ ~.-..~. ~ ~ , ~ , , ~ ~ - ~

2 5 3 1 4 1 5 7 X 1 6 6 1 3 1 1 2 1 0 3 5 3 4 8 X 1 6 1 7 1 8 2 4 2 3

~ ~i .

3 6 2 2 2 5 1 1 9 X 16 4 1 9 2 0 3 2 3 9 3 0 2 1 2 6 3 3 3 1 X 1 6 3 8 2 9 2 8 2 7 3 7 Fig. 1. Elcctrophoretic protein patterns of CNW isolates and rcfercnce strains. The numbers rcfer to those used in Table 1 and Fig. 2. Molecular weight marker (track labelled X) arc (from top to bottom): ovotransferrin, "/6-78 kDa; albumin, 66.25 kDa: ovaibumin, 45 kDa; carbonic anhydrase, 30 kDa: myoglobin, 17.2 kDa. The bracket indicates the 38 to 47 kDa range is excluded from the

numerical analysis.

333

different levels of trace alignment, background subtraction and dupficate gels.

4.3. Numerical analysis Numerical analysis of partial PAGE-protein

patterns based on the determination of the Pear- son correlation coefficient and U P G M A clustering revealed that at the 70~ S ( r = 0.70) level, the 39 strains of Campylobacter formed five distinct clus- ters (phenons 1-5) as shown in the dendrogram (Fig. 2). Phenon 1 could be further divided at the 78% S level into two separate sub-phenons, l a and lb . The other phenons remained intact except that a single strain of "C. upsaliensis" NCTC 11540 (Ref. no. 37, Table 1) became an outlier to phenon 3 within which it was included at the 70% S level. The composition of pbenons shown in Fig. 2 was as fellows.

Sub-phenon la contained four reference strains of C. jejuni subsp, doylei (including the type strain NCTC 11951), each of which represented different electropherotypes (EP-types 1-IV) as proposed in a previous study [5] and eleven CNW field strains: the latter strains hydrolysed hippurate bu t failed to reduce nitrate and were thus identifiable as C. jejuni subsp, doylei.

Sub-phenon l b contained three reference strains of C. jejuni subsp, jejuni (including the type strain NCTC 11351) and three CNW field strains: the latter reduced nitrate and hydrolysed hippurate and were therefore identifiable as C. jejuni subsp. jejunL

Phenon 2 contained the type strain of C. coil (NCTC 11366).

Phenon 3 contained 15 strains of "C. upsalien- sis": these included the proposed type NCTC

100 90

P E R C E N T A G E S I M I L A R I T Y

80 70 60 50

l a c , ~ m .

2 ~

3 "C. upe~en~"

4 C .m S C . ~

PL0~lgt m A0~lm A413/er S

NCTC 122N :

N~rc 11~S 14 ~ NI~'C 1 1 m lS

~'rG 11168 17 ~ ~ 122O7 10 I~5/H 19 ~ 1136'1 AII61/M 23 Aqr~/U 24 ~

NCT¢ 11~ NCl'~ 11184 27

N ~ C 1 ~ 1 ~

NCrC 11~1 ~S ~

-- NCTC 104142 -- NCTC 11352 39

Fig. 2. Phenogram of the cluster analysis based on partial protein patterns of strains (vertical axis) listed in Table 1. The numbers on the horizontal axis indicate the percentage similarities as determined by the Pearson product moment correlation coefficient ( × 100)

and unwcighted pair group average linkage clustering.

334

11541 [11] and various isolates previously identi- fied as "C. upsaliensis" by conventional bacterio- logical tests and by protein electrophoresis [17].

Phenon 4 contained the type strain of C fetus subsp, fetus (NCTC 10842).

Phenon 5 contained the type strain of C. iaridis (NCTC 11352).

5. DISCUSSION

The results show that the hippurate-positive CNW campylobacters included in this study were all identifiable as C. jejuni on the basis of their electrophoretic protein patterns, which agreed well with DNA~DNA hybridization data [18]. The above analysis, based on the use of partial protein patterns, revealed an unambiguous separation be- tween the hippurate-positive CNW strains of C jejuni and the hippurate-negative CNW strains belonging to " C upsaliensis". The main protein bands were deleted from the numerical analysis as we found previously that total protein patterns appear to be more suitable in the analysis of intra-species variation [16,19,20]. The present analysis confirmed and extended our previous studies on thermophilic campylobacters and dem- onstrated that high resolution numerical analysis of electrophoretic protein patterns is an excellent method for identifying new or biochemically atypical Campylobacter strains [5,19].

An interesting feature of this electrophoretic analysis was that it enabled the two subspecies within C. jejuni to be distinguished, and it also revealed that three catalase-negative, South Afri- can paediatric strains belonged to C. jejuni subsp. jejunL This cluster (phenon l b in Fig. 2) included the species type strain (NCTC 11351 x) but cata- lase-negative isolates of subsp, jejuni are uncom- mon. The other clinical isolates were identifiable as C. jejuni subsp, doylei, which was consistent with the fact that they were all nitrate-negative and agreed with previous observations that weak or negative catalase activity was a feature of some strains of C. jejuni subsp, doflei [2], It is evident from Fig. 2 that there is some electropboretic heterogeneity within " C upsaliensis" and that at

the 89% similarity level at least eight subgroups were discernible.

It is possible that over-refiance on the catalase test can confuse the identification of some ther- mophilic campylobacters because of the paucity of other diagnostic tests. We conclude that the hip- purate and nitrate tests, which generally appear to be more reliable and accurate than catalase pro- duction, should both be used in routine identifica- tion of C. jejuni subsp, doylei and "C. upsaliensis ". Our identifications of the eight Belgium CNW strains (two strains of C. jejuni subsp, doylei and six strains of "C. upsaliensis') were in excellent agreement with results obtained in an independant analysis using similar PAGE-protein techniques [17], and therefore demonstrated the portability and high inter-laboratory reproducibility of the SDS-PAGE technique when applied to the pro- teins of campylobacters.

ACKNOWLEDGEMENTS

We thank Dr H. Goossens for kindly providing strains. This work was carried out in the frame- work of Contract no. BAP-133-UK of the Biotech- nology Action Program of the Commission of the European Communities. Research by A. Lastovica was supported by a grant from the South African Medical Research Council.

REFERENCES

[11 Smibert, R.M. (1984) in Bergey's Manual of Systematic Bacteriology, Vol. I (Krieg, N.R. and Holt, J.G., eds.), pp. 111-118, Williams and Wilkins, Baltimore.

[2] Steele, T.W. and Owen, RJ. (1988) Int. J. Syst. BacterioL 38, 316-318.

[3] Steele, T.W. Sansster, N. and Lanser, J.A. (1985) J. Clin. Microbiol. 22, 71-74.

[4] Lastovica, A.J., Kirby, R. and Ambrosio, R.E. (1985) in Campylobacter I!I (Pearson, A.D., Skirrow, M.B., Linr, M. and Rowe, B., eds,), p. 201, Public Health Laboratory Service, London.

[5] Owen, RJ., Costas, M. and Sloss, L.L. (1988) Euro J. EpidemioL 4, 277-283.

[6[ Kasper, G. and Dickgie,,~r, N. (1985) Lancet i, 111-112. [7] Owen, R.J., Beck, A. and Borman, P. (1985) Eur. J.

EpidemioL 1,281-287.

[8] owen, R.J., Morgan, D.D., Costas, M. and Lastovica. A. (1989) FEMS Mierobiol. Left. 58, 145-150.

[9] Borczyk, A. and Lior, M. (1985) in Campylobacter III (Pearson, A.D., Skirrow, M.B., Lior, M. and Rowe, B., eds.), p. 235, Public Health Laboratory Service, London.

[10] Taylor, D.E., Hiratsuka, K. and Mueller, L. (1989) J. Clin. Microbiol. 27, 2042-2045.

[11] SandstedL K. and Ursing, J. (1986) in Abstracts of XIV Int. Congr. Microbiol. abstr, no. P. 138-17, p. 61. IUMS, Manchester.

[12] Sandstedt, K., Ursing, J. and Walder, M. (1983) Curt. Microbiol. 8, 209-213.

[13] Patton, M., Shaffer, N., Edwards, P., Barrett, T3., Lam- bert, M.A., Baker, C., Perlman, D.M. and Brenner, D.J. (1989) J. Clin. Mierobiol. 27, 66-73.

[14] Torten, P.A., Patton, C.M. Tenover, F.C., Barrett, T.J.,

335

Stature, W.E., Steigerwalt, A.(,i., Lin, J.Y., Holmes, ICK. and Brenner, D.J. (1987) J. Clin. Mierobiol. 25,1747-1752.

[15] Lastovica, A.J., Le Roux, E, and Penner. J.L. (1989) J Clin. Mierobiol. 27, 657-659.

[16] Costas, M., Cookson, B.D., Talsania, H.G. and Owen, R.J. (1989) J. Clin. Microbioi. 27, 2574-2581.

[17] Goossens. H., Pot, B., Vlaes, L. et al. (1990) J. Clin. Mierobiol. 28 In press.

[18] Hernande~, J., Owen, R.J., Costas, M. and Lastovica, M. (1990). J. AppL Bact. 69 In press.

[19] Owen, RJ., Costas, M., Sloss, L. and BoRon, FJ. (1988) J. Appl. Bacteriol. 65, 69-78.

[20] Owen, R.J., Costas, M., Morgan, D.D., On, S.L.W., Hill, L.R., Pearson, A.D. and Morgan, D.R. (1989) Antoine van Leeuwenboek 55, 253-267.