Molecular and Cellular Probes (1990) 4, 321-334

Use of DNA restriction endonuclease digest and ribosomal RNA gene probe patterns to fingerprint Helicobacter pylori and Helicobacter mustelae isolated from human and animal hosts David D . Morgan and Robert) . Owen* National Collection of Type Cultures, Central Public Health Laboratory, London NW9 5HT, UK (Received 10 January 1990, Accepted 22 January 1990)

Variation amongst strains of Helicobacter pylori and Helicobacter mustelae was examined by DNA restriction endonuclease digestion and rRNA gene patterns generated using a non-radioactive probe . Chromosomal DNA was extracted from 30 cultures of H . pylori from human, Rhesus monkey and pig gastric mucosa, and from three H . mustelae isolates from ferret gastric mucosa . DNA fingerprinting with Hae III and Hind III showed H . mustelae was relatively homogeneous but revealed genomic heterogeneity within H . pylori with at least 18 different DNA patterns identifiable amongst the 30 isolates . Five sets of strains other than duplicates with matching DNA fingerprints were identified within H . pylori. The Peruvian isolates were the largest identical set and comprised eight isolates from four different patients with five consecutive isolates from one patient . The Rhesus monkey strains were a relatively homogeneous set as were several Australian human isolates . The study demonstrates that rRNA gene restriction patterns provide a simple but highly discriminatory electrophoretic fingerprint for H . pylori with potential for use as a novel epidemiological marker in addition to total DNA digest analysis .

KEYWORDS : Helicobacter pylori, H . mustelae, DNA, restriction endonuclease digests, rRNA probes, hybridization.

INTRODUCTION In the past 5 years, Helicobacter pylori, previously named Campylobacter pylori' has become well established as the cause of active chronic gastritis and is suspected to play a role in peptic and duodenal ulcer disease in man ." Although the organism has a worldwide distribution, little is known about source of infection and transmission patterns, and whether strains differ in their pathogenicity or cause different clinical presentations 4 As serotyping and phagetyping are not yet appli* Author to whom correspondence should be addressed . 0890-8508/90/040321 +14 $03 .00/0

321



32 2

D . D. Morgan and R .) . Owen

cable to H . pylori, alternative sensitive molecular methods have been used to discriminate between individual strains . These included one-dimensional SDS-PAGE of proteins, 5-9 DNA fingerprinting by use of the restriction endonuclease Hind 111, 10-12 and immunoblot fingerprinting . 13 Differences between strains have been detected also in plasmid content, 14 cytotoxic activity," ," haernaglutinating activity, 17 and preformed enzyme profiles . 18,19 The aim of the study reported here was to investigate the use of DNA fingerprinting by comparisons of total digest patterns and Southern blot hybridization band patterns obtained with a ribosomal (r) RNA cistron probe (rRNA gene patterns) . The genomic DNA fingerprinting approach is highly sensitive and applicable to a wide range of micro-organisms," so could provide the basis of a novel method of detecting strain variations within H . pylori. Results are presented on H . pylori gastric mucosal isolates from man, pig and Rhesus monkey, and their DNA fingerprints are compared with those of H . mustelae, which is an allied species isolated from ferret gastric mucosa . 21

MATERIALS AND METHODS Bacterial strains and growth conditions The 33 isolates used in this study are listed in Table 1, with their reference numbers and sources . They comprised 26 human isolates, three Rhesus monkey isolates and one pig isolate (all of gastric mucosal origin) of H. pylori, and three isolates of H. mustelae from ferret gastric mucosa . All isolates were grown on Oxoid brain-heart infusion agar (BHI) containing 5% horse blood and supplemented with 1 % isovitalex (BBL Microbiology Systems, Becton Dickinson, Cowley, Oxford, UK) . Cultures were incubated for 48 h at 37°C under microaerobic conditions (5%, 0 2; 5%, C0 2 ; 2%, H2 ; 88%, N 2) in a Variable Atmosphere Incubator (Don Whitley Scientific Ltd, Shipley, Yorks) . Stocks of all cultures were preserved in 10% v/v glycerol in Oxoid nutrient broth over liquid nitrogen .

Chromosomal DNA extraction Chromosomal DNA was extracted and rapidly purified using the guanidium thiocyanate method . 22 The concentration and purity of the DNA sample was determined by absorbance readings at 230, 260 and 280 nm .

G + C estimation The DNA base composition (mol% G + C content) was estimated from the thermal denaturation temperature (Tm), which was determined in triplicate in 0 . 1 X SSC (1 X SSC is 0 . 15 M NaCl plus 0 . 015 M trisodium citrate, pH 7 .0) . The base composition was expressed relative to a chemically determined value of 51 . 1 mol% G+C for Escherichia coli NCTC 9001 . 23

DNA fingerprinting of H . pylori

Table 1 .

323

Bacterial strains used

Host sourcet

Received from$

Country of isolation

Helicobacter pylori 1 NCTC 11637T 2 NCTC 11638 3 NCTC 11639 4 NCTC 11916 12, 15 23111E 17, 21 241A 14, 19 2411A 13, 18 2411E 24 24111A 20, 23 24111E 16 2511B 22 951E 25 14009 26 13509 27 14007 28 14008 29 14096 30 13867 31 13895 32 14000 33 13962 34 13906 35 14173 36 13944 37 14126 13902 38 9 87008 10 87010 11 87012 5 NCTC 11961

Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human (1) Human Human Human Human Human (1) Human (2) Human (2) Human Human Human Rhesus monkey Rhesus monkey Rhesus monkey Pig

NCTC NCTC NCTC NCTC D . R . Morgan D . R . Morgan D. R . Morgan D . R . Morgan D . R . Morgan D . R . Morgan D . R . Morgan D . R . Morgan S . Goodwin S . Goodwin S . Goodwin S. Goodwin S . Goodwin S . Goodwin S . Goodwin S . Goodwin S . Goodwin S . Goodwin S . Goodwin S . Goodwin S . Goodwin S . Goodwin D . Newell D . Newell D . Newell NCTC

Australia Australia Australia UK Peru Peru Peru Peru Peru Peru Peru Peru Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia UK UK UK UK

Helicobacter mustelae 6 NCTC 12031 7 NCTC 12032 8 NCTC 12198T

Ferret Ferret Ferret

NCTC NCTC NCTC

UK UK USA

Study number

Strain number*

* Type strains are indicated by superscript T . All strains were isolated from gastric mucosa . I, 11 and III, indicates isolates were obtained at 0, 15 and 57 days . A, B and E, indicates the site of isolation was antrum, body or elsewhere . t Numbers in parenthesis indicate if the isolate was from the same patient . $ NCTC, National Collection -of Type Cultures, London, UK; D . R . Morgan, Norwich Eaton Pharmaceuticals, Norwich, New York; C. S. Goodwin, Royal Perth Hospital, Perth, Western Australia ; D . Newell, PHLS Centre for Applied Microbiology and Research, Porton Down, Salisbury, UK .

DNA digestion and electrophoresis The DNA (8 µg) was digested with the restriction endonucleases Eco RI, Hae III and Hind III (ca 1 unit per µg DNA) for 3 h at 37 ° C in the buffer recommended by the manufacturers (Northumbria Biologicals Ltd, Cramlington, Northumbria) . The



324

D . D . Morgan and R . J . Owen

digested DNA (10 samples containing 2 pg DNA) was electrophoresed at 25 V for 16 h in horizontal 0 . 7% (wt/vol) agarose (Gibco-BRL Ltd : ultrapure, electrophoresis grade) gel in a buffer containing 89 mm Tris, 89 mm boric acid, 2 mm disodium EDTA (pH 8 . 3) . After electrophoresis, the gels were stained in ethidium bromide (1 pg ml - ') and photographed for a permanent record .

Preparation of biotinylated probe cDNA The biotinylated cDNA probe was prepared from 1 gg Escherichia coli 16+23 S rRNA (BDH Ltd, Poole, UK) using Moloney mouse leukaemia virus reverse transcriptase (Gibco-BRL Ltd), and was biotinylated by the incorporation of biotin-16-dUTP (Gibco-BRL Ltd) according to previously described methods ."

Southern blot hybridization After photography, the DNA in the gel was depurinated by treatment with 0 . 25 N HCI for 30 min, then denatured in 0 . 5 M NaOH-1 . 5 m NaCl for 30 min and neutralized in 0 . 5 M Tris-HCI-1 . 5 M NaCI-1 mm disodium EDTA (pH 7 . 2) for 30 ruin . DNA was transferred to Hybond-N membrane (0 . 45 µm pore size : Amersham International) by capillary transfer (18-20 h) or by vacuum-assisted transfer (Vacublot ; Anderman & Co . Ltd, Kingston-upon-Thames, Surrey) . The membranes were washed once in 2 X SSC, air dried and baked at 80°C for 2 h . Prehybridization (42°C for 3-4 h) and hybridization (42°C for 18 h) were carried out exactly as described previously . 25 The hybridization reactions were visualized colorimetrically with the BIuGENE (Gibco-BRL Ltd) non-radioactive nucleic acid detection system, which contained streptavidin-alkaline phosphatase conjugate and dyes, as recommended by the manufacturer .

Band size estimation Fragment sizes in the total digest and in the Southern blot hybridization patterns were calculated from migration distances by the DNA SIZE program as described previously. 26 Biotinylated lambda phage (Gibco-BRL Ltd) digested with Hind III was used to provide the size markers (X, in figures) .

RESULTS DNA base compositions The DNA base compositions are listed in Table 2 . Six strains of H . pylori had values of 36-39 mol% G+C (mean 37 . 5±1 . 2 mol%) . The three strains of H . mustelae had slightly higher values of 41-42 mol% (mean 41 . 2±0 . 6 mol%) . Previously published values for several of these strains are included in Table 2 for comparison .

DNA fingerprinting of H. pylori Table 2 .

325

DNA base compositions of Helicobacter pylori and H . mustelae strains C+C content (mol%)

Species

Strain number

T m ( ° C)

Present study

Literature

H. H. H. H. H. H.

NCTC 11637T NCTC 11638 NCTC 11639 NCTC 11961 87008 87010 NCTC 12031 NCTC 12032 NCTC 12198T

69. 3 68. 7 68. 5 69 .0 69. 4 70. 0 71 . 0 70. 7 71 . 3

37. 7 36. 5 36 . 1 37 . 1 38. 0 39 . 3 41-3 40. 6 41 . 8

36-3727,28

pylori pylori pylori pylori pylori pylori

H . mustelae H . mustelae H . mustelae

36 27

40-41 21 37-3921

DNA fingerprints of reference strains Chromosomal DNA samples from eight H . pylori and three H . mustelae reference strains were digested with Hae III and Hind III, which cut DNA from all of the strains with a high frequency to give multiple electrophoretic band patterns (> 15 bands) . The fragments with sizes of about 4 kb and larger were generally well resolved (Fig . 1) . The corresponding rRNA gene patterns for the Hae III digest, which comprised two to five bands of 9 kb or less are shown in Fig. 2 . The isolates were relatively heterogeneous and about five main DNA patterns, depending on the quality of resolution, could be discerned by visual inspection with several isolates having

Fig. 1 . Electrophoretic patterns of Hae III digests of DNA from human and animal reference strains of H . pylori and H . mustelae .

326

D. D . Morgan and R. J . Owen

X

1

3

5

X

6

7

9

10

1I

X

kb

23 . 1

9.4 6. 6

4.4

2.3 2.0

Fig. 2 . Schematic diagram of rRNA gene patterns for Hae III digests of DNA from human and animal reference strains of H. pylori and H. mustelae .

matching patterns in one or other (or both) of the fingerprints derived from total digestion and rRNA gene analysis (Table 3) . The cultures with DNA similarities were grouped as follows : Set A comprised H . pylori Australian human isolates NCTC 11637 1 and NCTC 11638 : these two cultures were identical in both total digest and rRNA gene patterns . Set B comprised UK Rhesus monkey isolates 87008, 87010 and 87012 : these cultures appeared identical in total digest analysis but 87010 showed a single 2 . 8 kb band difference in the rRNA gene patterns . Set C comprised H. mustelae NCTC 12198T and two other representatives of the species which were similar to each other in Hae III and Hind III digest patterns and rRNA gene patterns . The H . mustelae strains were completely different from H . pylori as the H . mustelae Hae III digest patterns did not contain any bands with sizes exceeding 6 . 6 kb .

DNA fingerprints of Peruvian human isolates Chromosomal DNA samples from eight isolates of H . pylori from four Peruvian (Lima) patients were digested with Hae III (see Fig . 3) and Hind Ill . All the electrophoretic band patterns were identical but different from those of the



DNA fingerprinting of H . pylori Table 3. cultures

327

Summary of DNA fingerprint matches observed amongst H . pylori and H . mustelae

Number of cultures

Species H. pylori UK human UK pig UK monkey Australia, human

1 1 3 17

Matching set designations*

Peru, human

8

B A E F G D

H. mustelae UK ferret USA ferret

2 1

C C

Strain number

87008, 87010, 87012 NCTC 11637T, NCTC 11638 13509, 13944 14008, 13962 13867, 13906, 14193 23111E, 24A, 24I11A, 2411E, 24111A, 24111E, 25118, 95E NCTC 12031, NCTC 12032 NCTC 12195'

* A match is defined as any two patterns with bands at identical positions.

reference strains above . A digest of NCTC 11638 DNA was included for comparison (sample 2 : see also Fig . 1) . The eight isolates with a common pattern, which comprised set D (see Table 3), were from four patients . The isolates from patient 24 constituted a consecutive multiple (five) isolate set, comprising isolates from two sites (antrum and body) obtained at different times (0, 15 and 57 days) . The strains were also identical to each other in Hind III digest patterns and in Hind III-rRNA gene patterns (Fig . 4), although there were minor differences between isolate 24a (no . 17) and the others in some band intensities .

Fig. 3.

Electrophoretic patterns of Hae III digests of DNA from the Peruvian H . pylori human isolates .



3 28

Fig. 4.

D. D . Morgan and R. J . Owen

The rRNA gene patterns for Hind III digests of DNA from the Peruvian H. pylon human isolates .

DNA fingerprints of Australian human isolates DNA samples from 14 cultures of H . pylori from patients in Perth (Australia) were digested with Hae Ill (Fig . 5a), Hind III (Fig . 5b) and Eco RI (Fig. 5c) . The DNAs of strains 14000 and 14007 (nos . 27 and 32 in Fig . 5) were not cut by Hae III but were cut by the other two endonucleases . The rRNA gene fingerprints of these strains are shown in Fig . 6 (Hae III digests) and Fig . 7 (Hind III digests) . Visual inspection of the various electrophoretic band patterns revealed heterogeneity between strains but positive matches between several patterns were detected (sets E, F, G, see Table 3) . Cultures in the matching sets were as follows : Set E comprised isolates 13509 (no . 26) and 13944 (no . 36), which were identical in Hind III rRNA gene fingerprints but differed in their total digest patterns . These strains were from different patients . Set F comprised isolates 14008 (no . 28) and 13962 (no . 33), which were similar but not completely identical in both total digest and rRNA gene fingerprints . They were from the same patient . Set G comprised isolates 13867 (no . 30), 13906 (no . 34) and 14193 (no . 35) which were identical in the Hind Ill rRNA fingerprint . The latter two cultures were from the same patient and were identical in all fingerprints . Culture 13867 (no . 30) was isolated from a different patient and was not identical to the other cultures in its total digest fingerprint .

Fig . 5 . Electrophoretic digest patterns of DNA from Australian H. pylori human isolates. Restriction endonucleases used : (a) Hae III; (b) Hind III ; (c) Eco RI.



x

25

26 27 X

28

29

30 31 X 32 33 34 35 X

36

37

38

X

kb

(a

(b)

X

25 26 27

X

28 29 30 31

X

32 33 34 35 X 36 37 38 X

kb



330

D . D . Morgan and R. J . Owen

X

Fig. 6.

25 26

27 X 28

29 30 31

X

32

33 34 35

X 36

37

38

X

kb

The rRNA gene patterns for Hind III digests of DNA from the Australian H. pylori human isolates.

DISCUSSION Helicobacter pylori and H. mustelae are phylogenetically associated species with

DNA base compositions of 35-44mol% G+ C.1,27 However, the classification of these species in Campylobacter has been controversial and it was recently proposed that both species should be reclassified in a new genus named Helicobacter . 1 Our DNA base compositions are consistent with those values previously reported for H . pylori. Our estimate of 41 mol% for H . mustelae NCTC 12198T is slightly higher than previous values for this strain yet is similar to values of other members of the species . These minor G + C differences are not significant, however, and are within the range of expected interlaboratory variation . Although H . pylori strains may appear very homogenous with respect to DNA base composition and relatedness measured by DNA-DNA hybridization, 21,28,29 these criteria do not reflect the heterogeneity at the strain level revealed by DNA restriction digest analysis and as our studies also show, by rRNA gene patterns . Earlier investigation s 10-12 demonstrated that H. pylon isolates obtained from the gastric mucosa of different patients all produced unique DNA digest patterns with Hind III even when isolates were from members of closely associated family pairs such as husband and wife, or brother and sister . Reference strains NCTC 11637 T and NCTC 11638 were the only isolates from two unrelated patients previously reported" to have identical profiles . However, Majeswki & Goodwin" attributed the similarity between these isolates to the fact that they were cross-contaminated shortly after isolation and concluded they now represent the same strain . Our DNA results confirm the similarities between these two cultures (see matching set A),



DNA fingerprinting of H . pylori

X

25 26

X

28 29 30 31

X

331

33 34 35 X 36

37 38 X

kb

23. 1

(a)

X

25

26

X

28

29 30 31

X

33

34 35

X

36

37

38

X

kb 23 •I 9 .4

6.6 4.4

2.3 2.0

(b)

Fig. 7. The rRNA gene patterns for Hae III digests of DNA from the Australian H . pylori human isolates. (a) Bands on membrane. (b) Schematic diagram of bands.



33 2

D . D . Morgan and R . J. Owen

which agreed also with data from the numerical analysis of electrophoretic protein patterns ." Our investigation using total digest patterns and rRNA gene probes revealed a number of similarities between strains as well as differences . Most of the H . pylori human isolates were different from each other with 15 different DNA patterns identifiable amongst isolates from 26 unrelated patients . The eight Peruvian isolates constituted the largest single set of strains, which were from four different patients in Lima, all sharing a common DNA fingerprint. The results agreed with the similarities previously observed between these isolates obtained by numerical analysis of 1-D SDS-PAGE protein profiles .8'9 However, there were insufficient epidemiological details available to establish if there was any link between the patients to explain the observed similarities of the isolates . Three different pairs of matching strains (sets E, F and G) were also detected amongst the Perth strains . Clinical details were not available but the set E isolates and one of the set G isolates were known to be from different patients (C . S . Goodwin, pers . comm .) . The pig and rhesus monkey isolates were quite different from those of the H. pylori human strains . Our DNA data showed, however, that the three rhesus monkey strains all had identical DNA profiles except for strain 87010, which had a single band difference in its rRNA gene pattern . These results were in excellent agreement with previous data on numerical analysis of electrophoretic protein patterns which also showed strain 87010 to be slightly different ." The overall resemblance between these isolates suggest they may have been acquired by their respective animal hosts from a common source . The isolates of H. mustelae from ferrets differed quite markedly from H . pylori in their total digest and rRNA gene patterns . That result was in agreement with the low relatedness between the two species indicated by DNA-DNA hybridization, 29 and electrophoretic protein profiling ." The UK strains (NCTC 12031 and NCTC 12032) and the USA strain (NCTC 12198T) had similar DNA profiles which reflected the close relatedness between them in DNA-DNA hybridization 2' and protein profiling." Although our comparisons are based only on a small number of strains, the results suggest the H. mustelae isolates, which were from different ferrets, may be a more homogeneous species than H. pylori at the nucleotide sequence level . In this and earlier studies of H . pylori, restriction endonuclease digest analysis of chromosomal DNA has proved to be a highly sensitive method of strain identification ."" Such patterns appear to be highly stable and apparently unaffected by changes in other characteristics such as colonial morphology, 10 and loss of catalase or urease activity ." We found that rRNA gene patterns produced with an E. coli 16+/ 23S-rRNA gene probe provide an additional means of accurate strain identification in H . pylori. This finding is consistent with conclusions about the method reached in studies of a number of other bacterial species . 30 The method gives concordant results with total digest analysis in most cases . The advantage of the rRNA gene patterns is their relative simplicity (

Use of DNA restriction endonuclease digest and ribosomal RNA gene probe patterns to fingerprint Helicobacter pylori and Helicobacter mustelae isolated from human and animal hosts.

Variation amongst strains of Helicobacter pylori and Helicobacter mustelae was examined by DNA restriction endonuclease digestion and rRNA gene patter...
9MB Sizes 0 Downloads 0 Views