Isolation and characterization of a competence operon associated with transformation and adhesion in Helicobacter pylori

Microbes and Infection 8 (2006) 2756e2765 www.elsevier.com/locate/micinf

Original article

Isolation and characterization of a competence operon associated with transformation and adhesion in Helicobacter pylori Tzu-Lung Lin a, Chia-Tun Shun b, Kai-Chih Chang a, Jin-Town Wang a,c,* a

Department of Microbiology, National Taiwan University College of Medicine, Taipei, Taiwan b Department of Forensic Medicine, National Taiwan University Hospital, Taipei, Taiwan c Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan Received 3 January 2006; accepted 11 August 2006 Available online 7 September 2006

Abstract To identify adhesion-associated loci, we screened 1500 mutant strains of a Helicobacter pylori mutant library. A mutant that showed decreased adhesion to two gastric epithelial cell lines was identified. Inverse polymerase chain reaction (PCR) revealed that the interrupted locus of this mutant was an hp0015 homolog of H. pylori strain 26695. DNA sequencing and reverse-transcription PCR revealed that hp0015 and two downstream genes (hp0016 and hp0017) were a transcriptional unit. Deletion and complementation constructs revealed that hp0016 and hp0017 were involved in natural transformation, but not in adhesion. Hp0015 was associated with both adhesion and natural transformation. The reduction of adhesion to human gastric tissues in the hp0015 mutant was similar to that of the babA2 knockout mutants and greater than that of the sabA knockout mutants. Co-culture of a wild-type strain with AGS cells revealed that 19 genes in AGS cells were upregulated; however, five of the 19 genes were not induced by co-culture with hp0015 mutants. These results indicate that hp0015 is associated with adhesion, and hp0015, hp0016, and hp0017 are associated with natural transformation. Hp0017 has been named as virB4/comB4. Therefore, hp0015 and hp0016 are named as comB2 and comB3, respectively. Ó 2006 Elsevier Masson SAS. All rights reserved. Keywords: Helicobacter pylori; Mutant library; Adherence; Natural transformation

1. Introduction Helicobacter pylori is an important gastric pathogen in humans. It is a causative agent of type B gastritis and peptic ulcer, and a risk factor for gastric adenocarcinoma and mucosa-associated lymphoid tissue lymphoma (MALToma) of the stomach [1e5]. Several virulence factors have been identified, such as vacuolating cytotoxin and CagA, but the pathogenesis of H. pylori is still not fully understood. The interaction of bacteria and host cells through the binding of adhesin(s) and host cell receptor(s) is the first step in * Corresponding author. Department of Microbiology, National Taiwan University College of Medicine, 1 Jen-Ai Road, Section 1, Taipei 100, Taiwan. Tel.: þ886 2 2312 3456ext8292; fax: þ886 2 2394 8718. E-mail address: [email protected] (J.-T. Wang). 1286-4579/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.micinf.2006.08.006

bacteria colonization and contributes to persistent infection. Adherence may be advantageous to H. pylori, helping to stabilize it against mucosal shedding into the gastric lumen. The most well defined adhesin in H. pylori, BabA, was identified by the interaction with Lewis b (Leb) blood group antigen [6,7]. The interaction between the H. pylori adhesin SabA and the cellular receptor sialylated-Lewis x (sLex) may promote persistent infection [8]. Leb antigens are only abundant in individuals with blood type O. Seroepidemiologic studies have revealed that infection rates are the same in patients with different blood types [9]. Although a recent study further demonstrated that the BabA adhesin of most H. pylori strains can bind to A-Leb and B-Leb glycans [10], there are some clinical H. pylori strains do not bind to the Leb, A-Leb or sLex antigens [8]. These results suggest that different adhesins might be involved in the adhesion of different strains. Therefore,

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

there could be other adhesins besides BabA and SabA in H. pylori. In the current study, we adopted an H. pylori mutant library to identify the genes involved in cell adhesion. 2. Materials and methods 2.1. Bacterial strains and culture conditions An H. pylori clinical isolate, an NTUH-C1 strain that was cagAþ, vacAþ, with a highly natural transformation efficiency and well genetic characterization in our laboratory, was obtained at the National Taiwan University Hospital (NTUH), as previously described [11e14]. The NTUH-C1 strain was used to identify the strain-specific genes [14], rather than the sequenced 26695 and J99 strains. H. pylori strains were grown on Columbia agar plates (Oxoid Unipath Ltd., Hampshire, England) containing 5% sheep blood and chloramphenicol (4 mg/ ml) or kanamycin (10 mg/ml) and incubated for 2 or 3 days in microaerophilic conditions (5% O2, 10% CO2, 85% N2) at 37  C. Escherichia coli strains were grown on LB agar plates or in LB broth containing appropriate antibiotics. 2.2. Cell adherence assay Human gastric cancer epithelial cell lines (SC-M1; 5  105) [15] were grown in a 24-well culture plate with cover slides. Five hundred microliters of H. pylori (1  109 CFU/ml) were labeled with 5 ml FITC (10 mg/ml) for 1 h at room temperature. The bacterium was washed three times with 0.05% Tween-20/1 PBS. The FITC-labeled bacteria were plated to prove that the bacteria were still alive after these experimental procedures. The FITC-labeled H. pylori were incubated with SC-M1 cells for 1 h (MOI ¼ 1:100). Nonadherent bacteria were removed by washing 2  5 min with PBS buffer on a horizontal shaker. The entire cover slides with SC-M1 cells and FITC-labeled adherent H. pylori were observed under a fluorescence microscope. Each mutant strain was compared with the wild-type strain [16,17]. The adherence of H. pylori with another human gastric cancer epithelial cell line (AGS) was also observed. The expression of Leb and sLex antigens on SC-M1 and AGS cells was detected by immunofluorescence assay with Leb (Seikagaku, Tokyo, Japan) and sLex (Chemicon, Temecula, CA) monoclonal antibodies. The SCM1 cell was Leb negative but sLex positive, whereas the AGS cell was negative for both Leb and sLex antigens. 2.3. Inverse PCR and DNA sequencing To identify the loci that were interrupted by the transposon, genomic DNAs of mutant strains were extracted to perform inverse PCR and DNA sequencing as previously described [11]. 2.4. Reverse-transcription PCR Total RNA of H. pylori NTUH-C1 wild-type strain and M17-84 were extracted as previously described [12], and reverse transcribed by a Superscript II reverse transcriptase

2757

(Invitrogen, San Diego, CA) with an HP0017-5R primer (50 -CTTCTTGGTTGTGGAAATTG-30 ). The cDNA was amplified by PCR using HP0015-F (50 -ATGTCCGCTCATT TTTTAAG-30 ) and HP0016-R (50 -CGCTCTATAACATTTGG TTG-30 ) as primers. For each experiment, genomic DNA was used as a positive control, whereas RNA without reverse transcription served as a negative control to exclude DNA contamination. 2.5. Complementation of M17-84 Because hp0015-hp0016-hp0017 appeared to be an operon, we complemented hp0015, hp0015-hp0016, and hp0015hp0016-hp0017 separately into M17-84 by a shuttle vector pHel2 (a gift from Dr. R. Haas, Max-Planck-Institute fu¨r Biologie, Tu¨bingen, Germany) [18] and chromosomal integration. PCR products of hp0015, hp0015-hp0016, and hp0015hp0016-hp0017 contained the predicted promoter region (50 end, 141 base pairs from the hp0015 coding sequence) and were cloned into a pGEM-T easy vector (Promega, Madison, WI). The cat gene (chloramphenicol-resistant gene, a gift from Dr. D.E. Taylor, University of Alberta) [19] was then cloned into the SalI site of these plasmids. The complementation constructs were naturally transformed into a wild-type NTUH-C1 strain first, then the original hp0015 was mutated with M17-84 genomic DNA. The complemented gene was shown to integrate at the same locus by PCR. Hp0015, hp0015-hp0016, and hp0015-hp0016-hp0017 were also cloned into the EcoRV site of the pHel2 vector. Because the M17-84 mutant was profoundly impaired (<2  109) in natural transformation, complementation plasmids were transformed into wild-type strains by electroporation first, and then the hp0015 in chromosome was mutated by recombination with genomic DNA from M17-84. The gene alignments of these complementation strains were confirmed by PCR using different combinations of primers. 2.6. Constructions of hp0015 and hp0016 deletion mutants Using the hp0015-hp0016-hp0017 operon in a pGEM-T easy plasmid, hp0015 deletion construct was generated by inverse PCR with HP0015(1)R (50 -CAAAACCCTT TTGTTTAATT-30 ) and HP0016(þ2)F (50 -TGATTATCCTG TCAGCGA-30 ) primers. Primers HP0015(þ281)R (50 -TA AAAAATTCCCATAAACC-30 ) and HP0017(1)F (50 -AATG TTAGAAAAGCTTTTAAG-30 ) were used to generate hp0016 deletion construct. A blunt-end PCR product of cat coding region was also amplified by a Pfu polymerase with CAT(þ1)F (50 -ATGCAATTCACAAAGATTG-30 ) and CAT(þ624)R (50 -T TATTTATTCAGCAAGTC-30 ) primers and then phosphorylated by a polynucleotide kinase (New England Biolabs, Beverly, MA). The inverse PCR products and blunt-end cat were ligated. The orientation of the cat gene was the same as the direction of the operon. We obtained two plasmids hp0015 and hp0016 that were replaced by a cat coding region. These plasmids were transformed into a wild-type H. pylori

2758

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

NTUH-C1 strain to obtain hp0015 and hp0016 deletion mutants.

compared in more than 10 fields under 100 magnitudes with fluorescence microscopy.

2.7. Complementation of hp0015 deletion mutant

2.10. cDNA microarray for human cell signaling

The PCR product of hp0015 contained the predicted promoter region and the kanamycin resistant gene was cloned into the single HindIII site of pGEM-T easy/HP0405 plasmid. HP0405 encodes a NifS like protein, and the inactivation of this gene did not affect the growth rate [20]. The complementation construct was naturally transformed into a wild-type NTUH-C1 strain first, and then the original hp0015 was deleted as described above. The gene alignments of these complementation strains were confirmed by PCR using different combinations of primers.

A cDNA microarray, which contained 582 human cDNA clones involved in cell cycle, inflammation, signaling, and gastric physiology, was chosen from a human cDNA library with 9600 clones [22] to analyze the mRNA expression of AGS cells upon interaction with H. pylori. AGS cells (1  107) were incubated with the H. pylori wild type and M17-84 strains (1  109) for 2 h. RNA was then extracted using Trizol reagent (Invitrogen) according to the manufacturer’s protocol. AGS cells alone with the same culture conditions were used as controls. Microarray hybridization and detection were performed as previously described [12,22].

2.8. Constructions of babA2 and sabA knockout mutants 3. Results The babA2 gene [PCR amplified by babA2(1)-F (50 -CAT GAAAAAACACATCCTTTC-30 ) and babA2(810)-R (50 -GTT GGTGATTTCGGTGTAG-30 ) primers] was cloned into the pILL570 plasmid to perform transposon mutagenesis as previously described [11], and the cat gene was cloned into the single XbaI site of pGEM-T easy/sabA [PCR amplified by sabA(389)-F (50 -CAGGAATTGAAAACTGCGC-30 ) and sabA(1059)-R (50 -CCGGTTACCATAATAACTC-30 ) primers] plasmid. These plasmids were naturally transformed into an H. pylori NTUH-C1 wild-type strain to obtain babA2, sabA knockout mutants, and double mutants. Because a family of highly related outer membrane proteins was present, the different combinations of primers outside and in the babA2 and sabA were designed to confirm the correct knockout. The knockout of babA2 was confirmed by babA2-out-F (50 -CATT CTCCGCAACCAATCTC-30 ) and babA2-out-R (50 -ATTAAC GCCAAGTTTCCCAC-30 ) or babA2-R (50 -TTAATAAGCGA ACACATAG-30 ) primers. The correct knockout of sabA was checked by PCR with sabA-F (50 -GTGAGCGCCGGCTAT CAAATC-30 ) and sabA-out-R (50 -CTTAGGCATTAGCGCTA ATC-30 ) primers. 2.9. Human gastric tissue adherence assay Formalin-fixed and paraffin-embedded sections of stomach tissue were obtained from patients at NTUH. The morphology of tissue sections was stained with H&E staining. For tissueadherence assay, tissue sections were deparaffinized and rehydrated in xylene (10 min  2) and isopropanol (5 min  3), rinsed in water followed by PBS (5 min  3), and then incubated for 30 min in a blocking buffer (0.2% bovine serum albumin/0.05% Tween 20/1 PBS). Fifty microliters of FITC-labeled H. pylori in a blocking buffer (1  109 CFU/ ml) were added to the tissue section at room temperature for 1 h in a humidified chamber. Slides were then washed with PBS (5 min  6) and observed using fluorescence microscopy. The H. pylori NTUH-C1 wild-type strain served as a positive control, and Campylobacter jejuni as a negative control [17,21]. The adherence of each strain was observed and

3.1. Screening the mutant library by adherence assay An H. pylori mutant library consisting of 1500 mutants was screened. The library had been proven before to have 20% redundancy [11]. In this study, one mutant strain, M17-84, showed significantly decreased adherence with an SC-M1 cell (Fig. 1A,B). The adherence of M17-84 with MOI ¼ 1:100 was similar to that of the wild-type strain with MOI ¼ 1:10. Therefore, the adherence of M17-84 was decreased to about 10% of that of the wild-type strain. Because the adherence was too scattered at the MOI of 1:10 for mutants, further quantifications of adhesion with lower MOI were difficult to perform. The adherence of H. pylori with another human gastric cancer epithelial cell line (AGS) also showed similar results (w10-fold decrease of adhesion in mutants). We also confirmed the morphologic integrity of this mutant using light microscopy. The wild-type H. pylori strain of bacteria seemed to clump very frequently on SC-M1 or AGS cells, but not the M17-84 mutant. However, neither the wild-type H. pylori nor the M17-84 mutant strains clumped or aggregated in the cell culture medium without gastric cells; therefore the M17-84 mutant was considered to be defect in bacterial-cell rather than in inter-bacterial adhesion. Because Cag organelle encoded by cag pathogenicity island is a surface organelle [23] that would affect the cell adhesion, the adhesion of cagA and cagE mutants [13] was determined. The adherences of cagA and cagE mutants were the same as those of the wild-type strain (data not shown). 3.2. Identifying the interrupted gene loci by inverse PCR and sequencing The mini-TnKm insertion site for the mutant 17-84 was determined and compared with the GenBank databases (http://www.ncbi.nlm.nih.gov/BLAST/) as well as the H. pylori genome database (http://www.tigr.org). The nucleotide sequences interrupted by the mini-TnKm insertion of this mutant showed 97% similarity with a hypothetical gene, hp0015, of

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

2759

Fig. 1. Adherence of H. pylori with SC-M1 cells observed under a fluorescence microscope (400). The green rod-shape bacteria are H. pylori cells. After coculture FITC-labeled bacteria with SC-M1 cells, the cells were counterstained with rhodamine-phalloidin. Comparisons of SC-M1 cells stained with rhodamine-phalloidin and bacteria stained with FITC showed that bacteria are adhered to SC-M1 cells rather than to the glass slides. (i) Light field; (ii) bacteria stained with FITC; (iii) cells stained with rhodamine-phalloidin; (iv) merge of ii and iii. (A) H. pylori NTUH-C1 wild-type strain; (B) M17-84; (C) M17-84 complemented with hp0015 by chromosomal integration; (D) M17-84 complemented with hp0015 gene by pHel2 shuttle vector; (E) M17-84 complemented with hp0015-hp0016 by integration; (F) M17-84 complemented with hp0015-hp0016 by pHel2; (G) M17-84 complemented with hp0015-hp0016-hp0017 by integration; (H) M17-84 complemented with hp0015-hp0016-hp0017 by pHel2; (I) M10-75; (J) Hp0015 deletion mutant; (K) Hp0015 deletion mutant complemented with hp0015 in hp0405; (L) babA2 mutant; (M) sabA mutant.

the H. pylori strain 26695 [24]. This open reading frame contains 282 base pairs and encodes a 93-amino-acid protein. The transposon insertion site is located in the 50 -end 73rd nucleotide of hp0015. Hp0015 did not have any other homolog

in the DNA and protein databases. A protein domain search (http://smart.embl-heidelberg.de/) revealed that HP0015 had a predictive signal peptide and two transmembrane domains, and was most likely a membrane-associated protein.

2760

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

3.3. Analysis of adjacent region and reverse-transcription PCR

Table 1 Frequency of DNA transformation by natural transformation of H. pylori mutants

Sequences of NTUH-C1 revealed that two downstream ORFs of hp0015 were hp0016 and hp0017, as in 26695 [24]. The stop codon TAA of hp0015 overlaps with the start codon ATG of hp0016, and there is only one base pair between hp0016 and hp0017. RTePCR showed that hp0015-hp0016hp0017 was cotranscribed (Fig. 2). However, the RNA transcript of 17-84 was undetected. Therefore, these three genes were an operon.

H. pylori strain

Natural transformation efficiencya

NTUH-C1 wild-type NTUH-C1 wild-type with pHel2 M17-84 mutant M10-75 mutant M17-84 complementation with pHel2

3  104 4  104

3.4. Complementation of M17-84

M17-84 complementation by chromosomal integration Deletion

M17-84 was shown to have decreased transformation frequency preliminary in a previous study [11]. Because hp0015, hp0016, and hp0017 were proved to be an operon in this study, the adhesion and natural competence roles of each of these three genes were studied. We independently complemented hp0015, hp0015-hp0016, and hp0015-hp0016-hp0017 into an M17-84 mutant by a shuttle vector (pHel2) [18] or chromosomal integration. Complementation with hp0015 alone restored the ability of adhesion, but not natural competence (Table 1 and Fig. 1C,D). The adhesion ability of the hp0015-hp0016 complemented strain of M17-84 was the same as that of the wild-type strain (Fig. 1E,F). The efficiency of natural transformation was only partially recovered by shuttle vector and chromosomal complementation (7.5  107 and 2.5  106, respectively) (Table 1). Complementation of hp0015-hp0016-hp0017 by both shuttle vector and chromosomal integration completely restored the efficiency of natural competence and adhesion (Table 1 and Fig. 1G,H). Complementation by empty pHel2 vector revealed the same abilities of natural transformation (Table 1) and adherence (data not shown) as the wild-type strain. These results suggested that only hp0015 was associated with adhesion, whereas hp0015, hp0016, and hp0017 were all involved in natural competence. 3.5. Deletion of hp0015 and hp0016 To further exclude the possible polar effect, hp0015 and hp0016 deletion mutants were constructed, and alignments

Fig. 2. RTePCR of hp0015-hp0016-hp0017. Arrows indicate the location and direction of each gene. The figure is not drawn to scale. RNAs were reversely transcribed with primer HP0017-5R and then amplified with primers HP0015F and HP0016R. Lane 1, PCR products of genomic DNA of H. pylori NTUH-C1 (positive control); lane 2, RTePCR products of NTUH-C1 RNA; lane 3, RTe PCR products of M17-84 RNA; lane 4, PCR using RNA of NTUH-C1 without RT (negative control); lane M, 1 kb DNA marker.

NTUH-C1 wild-type hp0015 deletion mutant a

hp0015 hp0015-hp0016 hp0015-hp0016-hp0017 hp0015 hp0015-hp0016 hp0015-hp0016-hp0017 hp0015 hp0016 Complementation of hp0015 in hp0405

<2  109 <2  109 <2  109 3  106 4  104 <2  109 8  107 3  104 <2  109 <2  109 6.7  105 4.2  105

Means of three experiments. <2  109: no transformed colony.

were confirmed by PCR. Adherence and natural transformation frequency were characterized in the deletion mutants. An hp0017 knockout mutant (M10-75) has been proven to impair natural competence previously [7]. However, the adhesion ability of M10-75 was the same as that of the H. pylori wildtype strain (Fig. 1I). In the current two deletion mutants, the natural competence of both was impaired (Table 1). The Hp0015 deletion mutant revealed similar adhesion results to those of mutant 17-84 (Fig. 1J). 3.6. Complementation of HP0015 deletion mutant We further complemented hp0015 in the hp0015 deletion mutant by integration into the chromosome site of HP0405 [20]. Integration of an additional hp0015 in the wild-type strain showed similar natural transformation efficiency to that of the wild-type strain (6.7  105). Complementation of hp0015 in the chromosome significantly, but not fully, restored the ability of adhesion and natural transformation (4.2  105) (Table 1 and Fig. 1K). These deletion and complementation results further demonstrated that hp0016 and hp0017 were involved in natural competence, whereas hp0015 was involved both in adhesion and natural competence. Hp0017 was named as virB4/comB4 before [11,25]. Therefore, we defined hp0015 and hp0016 as comB2 and comB3, respectively. Previous study provided the evidence for a conjugation-like mechanism of DNA transfer in H. pylori [18,26]. The effect of HP0015 on conjugation was determined. The NTUH-C1-clarithromycin resistant strain and NTUH-C1streptomycin resistant strain were mated on the plates. Double resistant progeny were observed when the strains were grown together in the absence of DNase I (w1  106), but not in the presence of DNase I (<1  108). The efficiency of DNaseresistant conjugation in the NTUH-C1 strain was under the detection limit of our experiments. The M17-84-streptomycin resistant strain and the M17-84-spectinomycin strain were also

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

2761

mated on the plates. When the strains were grown together with or without DNase I, double resistant progeny were not observed (<1  108). These results indicated that NTUHC1 wild-type strain had a very low DNA transfer frequency or unable to transfer DNA by conjugation. Therefore, the effect of HP0015 mutation on conjugation could not be assayed in our NTUH-C1 strain. 3.7. Human gastric tissue adhesion assay BabA2 and sabA knockout mutants of NTUH-C1 were used as positive controls. Because AGS cells were both negative for Leb and sLex antigens, the adherence of babA2 and sabA mutants with AGS cells did not decrease. However, the adherence of babA2 and sabA mutants with SC-M1 cells, which were Leb negative and sLex positive, did not decrease either (Fig. 1L,M). Therefore we used stomach tissue sections from four patients (two with blood type O, one with blood type A, and one with blood type B) without H. pylori infection to perform tissue adherence assay. Compared with the babA2 and sabA double mutant, the babA2 mutant showed decreased adherence, whereas the sabA mutant was similar to that of the wild-type strain; however, the babA2 and sabA double mutant almost wiped out the adhesion ability of the human gastric tissue section (data not shown). The decrease of adherence of M17-84 was similar to that of the babA2 mutant (Fig. 3). The results of gastric tissue adherence assay revealed that babA2 and sabA have very important and cooperative roles in adhesion, and that comB2 (hp0015) is also significantly associated with adhesion. 3.8. cDNA microarray for cell signaling After normalization with a GAPDH expression level, the expression ratios between non-infected AGS cells and those infected with the NTUH-C1 strain or the M17-84 mutant were calculated. The mean of expression ratio between non-infected AGS cells and those infected with the NTUH-C1 strain was 1.192 (SD ¼ 0.490). There were 19 genes in AGS cells that were significantly upregulated after incubation with the NTUH-C1 wild-type strain (defined as the expression ratio >mean þ 2 SD). These 19 genes were associated with signaling transduction, transcriptional regulation, and apoptosis (Table 2). Among these 19 upregulated genes, three genes (including phosphoinositide-3-kinase, TGFb-inducible early response, and bcl-2 binding component 3 [bbc3]) were also reported to be induced by H. pylori [27]. When the mRNA expression profiles of AGS cells cocultured with M17-84 were examined, the mean of expression ratio between non-infected AGS cells and those infected with M17-84 mutant was 1.290 (SD ¼ 0.504). Five of the 19 genes in AGS cells that were upregulated by adhesion with the wild-type H. pylori were not induced by the M17-84 mutant (defined as the expression ratio
Fig. 3. Adherence of H. pylori with human tissue observed under a light (A, B) or fluorescence microscope (CeG). (A) Tissue section with H&E staining (100). The marked square is the region for subsequent observations. (B) Tissue section with H&E staining (400). (C) H. pylori NTUH-C1 wild-type strain (400). (D) M17-84 (400). (E) babA2 mutant (400). (F) sabA mutant (400). (G) babA2 and sabA double mutant (400).

4. Discussion BabA and SabA have been identified as adhesins of H. pylori. BabA was identified through the interaction with the Leb blood group antigen, the putative cellular receptor of H. pylori [6,7]. SabA binds to the sLex antigen and is considered to be responsible for persistent colonization [8]. AlpAB was identified by screening H. pylori mutants that consisted of strains containing surface-related loci constructed by BlaM-transposon shuttle mutagenesis [16,17]. The BlaM

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

2762

Table 2 Cellular genes that were upregulated by co-culture with H. pylori Expression level Non-infect

NTUH-C1

17-84

64.80

139.64

123.88

29.07

69.36

60.51

14.01

36.36

34.28

27.99

87.99

73.12

10.57

24.37

24.30

13.35

32.06

25.48

17.71

44.23

43.20

21.04

46.63

38.00

47.28

172.66

127.91

25.98

159.10

123.38

91.16

198.79

161.82

9.14

37.48

23.41

40.80

92.84

83.50

30.73

151.77

120.39

67.53

168.61

93.12

19.92

55.77

29.39

11.44

27.87

19.87

22.02

53.78

27.68

26.27

87.69

38.14

Gene description

Accession number

Ras-related protein Rab-5A Human Fas-associated death domain protein interleukin-1b-converting enzyme 2 Phosphatase 2A B56-alpha (PP2A) Highly similar to latent transforming growth factor B Highly similar to zinc finger protein HF.12 cAMP-dependent protein kinase regulatory subunit type I Signal transducing adaptor molecule STAM Highly similar to Ras-related protein RAB-10 RalGDS-like 2 (RGL2) mRNA Phosphoinositide-3-kinase, catalytic, delta polypeptidea Cyclin-dependent kinase inhibitor 1C (p57, Kip2) Smad anchor for receptor activation Human K-ras oncogene protein mRNA H. sapiens mRNA for Zinc-finger protein (ZNFpT17) Protein phosphatase 2 (formerly 2A), catalytic subunit, alpha HUS1 (S. pombe) checkpoint homolog TGFB inducible early growth response 2a Human Bcl-2 binding component 3 (bbc3) mRNAa Tumor necrosis factor (ligand) superfamily, member 7

H11455 T96912

R59165 R60197

H17016

R39309

R44706

AA074077

R37588 H16564

AI028513 AI240104 N95249 AI364655

R38106

AI078668 AI183650 AA587620

AI419481

a These three genes were also observed to be induced by H. pylori in a previous study [27].

method is very efficient; however, it may miss some genes that are not closely surface related. We identified an HpyC1I restriction and modification system by screening the adherence ability of the H. pylori mutant library with a different screening method [14]. The decrease of adhesion of the hpyC1I mutant was considered to be a result of aberrant morphology or other indirect effects because the mutant H. pylori

Fig. 4. RNA slot blot of AGS cell infected with NTUH-C1 or M17-84 mutant and non-infected. Panel 1, GAPDH; panel 2, phosphoinositide-3-kinase; panel 3, HUS1 checkpoint homolog.

was elongated w2-fold of the wild type. In contrast, the currently identified mutant was interrupted in a probable membrane-associated protein with intact morphology. The babA2 and sabA were not identified in our screening. Although the SC-M1 cell was Leb negative but sLex positive, the AGS cell was negative for both Leb and sLex antigens. But the adherence of babA2 and sabA mutants with SC-M1 and AGS cells was the same as that of the wild-type strain. These results suggested that the adhesion results may vary with different screening methods. Therefore, confirmation in a tissue section was necessary. Hp0015-hp0016-hp0017 appeared to be an operon. To exclude polar effects, there are two methods currently available [28,29]. Unmarked deletion methods in H. pylori had been developed previously and seemed an ideal method. However, unmarked deletion mutants were generated by two steps of transformation. Because the transformation abilities of mutants in these three genes were impaired, the interrupted marker could not be replaced by a second transformation of markerless deletion constructs. Therefore, the markerless deletion method was not feasible for studying transformation related genes. The second way is to replace the genes with the coding region of an antibiotic marker. In our study, hp0015 was replaced by the coding region of CAT and was proven to be nonpolar by the restoration of the abilities of transformation and adhesion after complementation of hp0015 into the chromosome site of hp0405 [20]. By knockout, complementation, and deletion experiments to exclude the polar effect, hp0015 was confirmed as the gene involved in adhesion, whereas hp0015, hp0016, and hp0017 were all associated with natural transformation. The roles of hp0015, hp0016, and hp0017 in natural transformation were also presented recently [30]. These researchers proved that hp0015, hp0016, and hp0017 were involved in natural transformation by starting with computer-based searches. Their findings on natural transformation were in line with our observations. Hp0017 has been named as virB4/comB4. Therefore, hp0015 and hp0016 were defined as comB2 and comB3 accordingly.

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

Although complementation of the hp0015 deletion mutant did not fully restore the abilities of natural transformation and adhesion, the restoration was significant. The findings on the natural transformation of hp0015 were in line with a recent study [30]. They also failed to restore the function completely by complementation [30]. The reasons for the incomplete restoration might be the different genome locations that affected the regulation and gene expression level rather than other secondary effects. Hp0015 played the roles of both adhesion and natural transformation. Similar results have been observed in Campylobacter jejuni, in which one gene, comB3, was also involved in both natural competence and adherence [31,32]. The presence of membrane-associated ComB3 might enhance the level of natural transformation by affecting surface changes that promote DNA binding in C. jejuni. The mechanism of association with both competence and adhesion in HP0015 could be similar to that of ComB3 in C. jejuni; however, further investigations are needed. The function of hp0015 in adhesion was reconfirmed by human gastric tissue section adherence assay. In our study, babA2 and sabA were shown to have important and cooperative roles of adhesion by the results of a double knockout mutant as reported previously [7,8]. Hp0015 did not have any other homolog in the DNA and protein databases. A protein domain search revealed that HP0015 has a predictive signal peptide and two transmembrane domains. Therefore, HP0015 was probably a membrane-associated protein. We failed to express HP0015 recombinant protein in several E. coli hosts with different expression vectors. However, we were able to select E. coli clones containing hp0015 with its own promoter in the pHel2 plasmid. The reason HP0015 failed to express in E. coli strains could be that large amount expression of the gene is lethal for E. coli, because there is different promoter recognition between E. coli and H. pylori, or for other unknown reasons. The initial and critical step of cellular expression change is the intimate interaction of bacteria and host cells through the binding of adhesin(s) and receptor(s) [33]. In Pseudomonas aeruginosa, 27 genes were identified to differentially express upon infection by microarray; 16 of these genes were adhesion dependent and required the expression of the adhesin type IV pili [33]. We found that five out of 19 genes in AGS cells that were upregulated by adhesion with the wild-type H. pylori were not induced by the M17-84 mutant. The effects of the M17-84 mutant on the AGS cells’ induction might be directly intimate contact dependent through the binding of HP0015. However, the CagA protein was known to be translocated into host cells by a type IV secretion system encoded by cag pathogenicity island and subsequently to induce several host proteins [23]. The decreased adherence to gastric cells of the M17-84 mutant might decrease the translocation and activity of CagA or other unknown proteins. Therefore, the effects of the M17-84 mutant on the AGS cells’ induction might be direct or indirect. However, these data provided another indirect evidence that HP0015 was required for intimate contact

2763

with human gastric epithelial cells. Adhesion was an important virulence factor and played a role in disease pathogenesis in vivo. Much hostebacteria crosstalk happened after the adhesion. But the adhesion of H. pylori and subsequently hostebacteria crosstalk in vivo are only poorly understood. The HP0015 mutant failed to induce the five genes (protein phosphatase 2A, HUS1 checkpoint homolog, TGFB inducible early growth response 2, human Bcl-2 binding component 3 [bbc3] and tumor necrosis factor superfamily member 7) in AGS cells in vitro by microarray study. The protein phosphatase 2A (PP2A) may be involved in the regulation of metabolism, transcription, RNA splicing, translation, differentiation, cell cycle, oncogenic transformation, and signal transduction [34]. The HUS1 was a DNA damage sensor protein in cell cycle progression and important in the DNA damage checkpoint [35]. The TGFB inducible early growth response 2 (TIEG2 or KLF11) has been implicated in the regulation of cell growth and gene expression [36]. The transcriptional regulation of bbc3, a pro-apoptotic BH3-only gene, contributes to the transduction of diverse cell death and survival signals [37]. The cytokine, tumor necrosis factor superfamily member 7, contributes to T-cell activation and plays a role in regulating B-cell activation, the cytotoxic function of natural killer cells, and immunoglobulin synthesis [38]. These results might shed light on the role of adhesion in hostebacteria crosstalk in vivo. DNA uptake in H. pylori has been identified to use a transport system related to the type IV secretion system encoded by ComB4 and ComB7-10 previously [25,39]. In a recent study, a third putative type IV secretion system (containing orfA to Q) of an H. pylori PeCan18B strain was identified [40]. Hp0015-hp0016-hp0017 was similar to orfD, orfE, and orfG in this region. However, the gene alignment and flanking regions in NTUH-C1 and 26695 were different from those of PeCan18B. Therefore, determining whether hp0015-hp0016-hp0017 was a type IV secretion system awaits further study. By computer prediction modeling and topology analyses, the involvement of HP0015-HP0016 in a type IV secretion system was demonstrated recently [30]. The study predicted that HP0015 has a secondary structure similar to VirB2, and HP0016 has a similar structure to VirB3. HP0015 and HP0016 were identified as essential components of the transformation apparatus, structurally and functionally homologous to VirB2 and VirB3 in a type IV secretion system, respectively. In conclusion, we identified an operon, hp0015-hp0016hp0017 (comB2-comB3-comB4), associated with adhesion and natural transformation. ComB2 is involved in adhesion, whereas ComB2, ComB3, and ComB4 are associated with natural transformation in H. pylori.

Acknowledgments This work was supported by a grant (NSC 93-2321-B-002032-) from the National Science Council in Taiwan.

2764

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

References [1] M.J. Blaser, Hypotheses on the pathogenesis and natural history of Helicobacter pylori-induced inflammation, Gastroenterology 102 (1992) 720e727. [2] B.J. Marshall, J.R. Warren, Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration, Lancet 1 (1984) 1311e1315. [3] A. Nomura, G.N. Stemmermann, P.H. Chyou, I. Kato, G.I. PerezPerez, M.J. Blaser, Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii, N. Engl. J. Med. 325 (1991) 1132e1136. [4] J. Parsonnet, G.D. Friedman, D.P. Vandersteen, Y. Chang, J.H. Vogelman, N. Orentreich, R.K. Sibley, Helicobacter pylori infection and the risk of gastric carcinoma, N. Engl. J. Med. 325 (1991) 1127e1131. [5] A.C. Wotherspoon, C. Doglioni, T.C. Diss, L. Pan, A. Moschini, M. de Boni, P.G. Isaacson, Regression of primary low-grade B-cell gastric lymphoma of mucosa-associated lymphoid tissue type after eradication of Helicobacter pylori, Lancet 342 (1993) 575e577. [6] T. Boren, P. Falk, K.A. Roth, G. Larson, S. Normark, Attachment of Helicobacter pylori to human gastric epithelium mediated by blood group antigens, Science 262 (1993) 1892e1895. [7] D. Ilver, A. Arnqvist, J. Ogren, I.M. Frick, D. Kersulyte, E.T. Incecik, D.E. Berg, A. Covacci, L. Engstrand, T. Boren, Helicobacter pylori adhesin binding fucosylated histo-blood group antigens revealed by retagging, Science 279 (1998) 373e377. [8] J. Mahdavi, B. Sonden, M. Hurtig, F.O. Olfat, L. Forsberg, N. Roche, J. Angstrom, T. Larsson, S. Teneberg, K.A. Karlsson, S. Altraja, T. Wadstrom, D. Kersulyte, D.E. Berg, A. Dubois, C. Petersson, K.E. Magnusson, T. Norberg, F. Lindh, B.B. Lundskog, A. Arnqvist, L. Hammarstrom, T. Boren, Helicobacter pylori SabA adhesin in persistent infection and chronic inflammation, Science 297 (2002) 573e578. [9] J.T. Lin, J.T. Wang, T.H. Wang, M.S. Wu, T.K. Lee, C.J. Chen, Helicobacter pylori infection in a randomly selected population, healthy volunteers, and patients with gastric ulcer and gastric adenocarcinoma. A seroprevalence study in Taiwan, Scand. J. Gastroenterol. 28 (1993) 1067e1072. [10] M. Aspholm-Hurtig, G. Dailide, M. Lahmann, A. Kalia, D. Ilver, N. Roche, S. Vikstrom, R. Sjostrom, S. Linden, A. Backstrom, C. Lundberg, A. Arnqvist, J. Mahdavi, U.J. Nilsson, B. Velapatino, R.H. Gilman, M. Gerhard, T. Alarcon, M. Lopez-Brea, T. Nakazawa, J.G. Fox, P. Correa, M.G. Dominguez-Bello, G.I. Perez-Perez, M.J. Blaser, S. Normark, I. Carlstedt, S. Oscarson, S. Teneberg, D.E. Berg, T. Boren, Functional adaptation of BabA, the H. pylori ABO blood group antigen binding adhesin, Science 305 (2004) 519e522. [11] K.C. Chang, Y.C. Yeh, T.L. Lin, J.T. Wang, Identification of genes associated with natural competence in Helicobacter pylori by transposon shuttle random mutagenesis, Biochem. Biophys. Res. Commun. 288 (2001) 961e968. [12] S. Ang, C.Z. Lee, K. Peck, M. Sindici, U. Matrubutham, M.A. Gleeson, J.T. Wang, Acid-induced gene expression in Helicobacter pylori: study in genomic scale by microarray, Infect. Immun. 69 (2001) 1679e1686. [13] Y.P. Lai, J.C. Yang, T.Z. Lin, J.T. Wang, J.T. Lin, CagA tyrosine phosphorylation in gastric epithelial cells caused by Helicobacter pylori in patients with gastric adenocarcinoma, Helicobacter 8 (2003) 235e243. [14] T.L. Lin, C.T. Shun, K.C. Chang, J.T. Wang, Isolation and characterization of a HpyC1I restriction-modification system in Helicobacter pylori, J. Biol. Chem. 279 (2004) 11156e11162. [15] S.Y. Jiang, R.Y. Shyu, H.Y. Chen, M.M. Lee, K.L. Wu, M.Y. Yeh, In vitro and in vivo growth inhibition of SC-M1 gastric cancer cells by retinoic acid, Oncology 53 (1996) 334e340. [16] S. Odenbreit, M. Till, R. Haas, Optimized BlaM-transposon shuttle mutagenesis of Helicobacter pylori allows the identification of novel genetic loci involved in bacterial virulence, Mol. Microbiol. 20 (1996) 361e373.

[17] S. Odenbreit, M. Till, D. Hofreuter, G. Faller, R. Haas, Genetic and functional characterization of the alpAB gene locus essential for the adhesion of Helicobacter pylori to human gastric tissue, Mol. Microbiol. 31 (1999) 1537e1548. [18] D. Heuermann, R. Haas, A stable shuttle vector system for efficient genetic complementation of Helicobacter pylori strains by transformation and conjugation, Mol. Gen. Genet. 257 (1998) 519e528. [19] Y. Wang, K.P. Roos, D.E. Taylor, Transformation of Helicobacter pylori by chromosomal metronidazole resistance and by a plasmid with a selectable chloramphenicol resistance marker, J. Gen. Microbiol. 139 (1993) 2485e2493. [20] P. Alamuri, R.J. Maier, Methionine sulphoxide reductase is an important antioxidant enzyme in the gastric pathogen Helicobacter pylori, Mol. Microbiol. 53 (2004) 1397e1406. [21] P. Falk, K.A. Roth, T. Boren, T.U. Westblom, J.I. Gordon, S. Normark, An in vitro adherence assay reveals that Helicobacter pylori exhibits cell lineage-specific tropism in the human gastric epithelium, Proc. Natl. Acad. Sci. USA 90 (1993) 2035e2039. [22] J.J. Chen, R. Wu, P.C. Yang, J.Y. Huang, Y.P. Sher, M.H. Han, W.C. Kao, P.J. Lee, T.F. Chiu, F. Chang, Y.W. Chu, C.W. Wu, K. Peck, Profiling expression patterns and isolating differentially expressed genes by cDNA microarray system with colorimetry detection, Genomics 51 (1998) 313e324. [23] S. Odenbreit, J. Puls, B. Sedlmaier, E. Gerland, W. Fischer, R. Haas, Translocation of Helicobacter pylori CagA into gastric epithelial cells by type IV secretion, Science 287 (2000) 1497e1500. [24] J.F. Tomb, O. White, A.R. Kerlavage, R.A. Clayton, G.G. Sutton, R.D. Fleischmann, K.A. Ketchum, H.P. Klenk, S. Gill, B.A. Dougherty, K. Nelson, J. Quackenbush, L. Zhou, E.F. Kirkness, S. Peterson, B. Loftus, D. Richardson, R. Dodson, H.G. Khalak, A. Glodek, K. McKenney, L.M. Fitzegerald, N. Lee, M.D. Adams, E.K. Hickey, D.E. Berg, J.D. Gocayne, T.R. Utterback, J.D. Peterson, J.M. Kelley, M.D. Cotton, J.M. Weidman, C. Fujii, C. Bowman, L. Watthey, E. Wallin, W.S. Hayes, M. Borodovsky, P.D. Karp, H.O. Smith, C.M. Fraser, J.C. Venter, The complete genome sequence of the gastric pathogen Helicobacter pylori, Nature 388 (1997) 539e547. [25] D. Hofreuter, S. Odenbreit, R. Haas, Natural transformation competence in Helicobacter pylori is mediated by the basic components of a type IV secretion system, Mol. Microbiol. 41 (2001) 379e391. [26] E.J. Kuipers, D.A. Israel, J.G. Kusters, M.J. Blaser, Evidence for a conjugation-like mechanism of DNA transfer in Helicobacter pylori, J. Bacteriol. 180 (1998) 2901e2905. [27] A.R. Sepulveda, H. Tao, E. Carloni, J. Sepulveda, D.Y. Graham, L.E. Peterson, Screening of gene expression profiles in gastric epithelial cells induced by Helicobacter pylori using microarray analysis, Aliment, Pharmacol. Ther. 16 (Suppl. 2) (2002) 145e157. [28] M. Copass, G. Grandi, R. Rappuoli, Introduction of unmarked mutations in the Helicobacter pylori vacA gene with a sucrose sensitivity marker, Infect. Immun. 65 (1997) 1949e1952. [29] W. Fischer, D. Schwan, E. Gerland, G.E. Erlenfeld, S. Odenbreit, R. Haas, A plasmid-based vector system for the cloning and expression of Helicobacter pylori genes encoding outer membrane proteins, Mol. Gen. Genet. 262 (1999) 501e507. [30] A. Karnholz, C. Hoefler, S. Odenbreit, W. Fischer, D. Hofreuter, R. Haas, Functional and topological characterization of novel components of the comB DNA transformation competence system in Helicobacter pylori, J. Bacteriol. 188 (2006) 882e893. [31] D.J. Bacon, R.A. Alm, D.H. Burr, L. Hu, D.J. Kopecko, C.P. Ewing, T.J. Trust, P. Guerry, Involvement of a plasmid in virulence of Campylobacter jejuni 81-176, Infect. Immun. 68 (2000) 4384e4390. [32] D.J. Bacon, R.A. Alm, L. Hu, T.E. Hickey, C.P. Ewing, R.A. Batchelor, T.J. Trust, P. Guerry, DNA sequence and mutational analyses of the pVir plasmid of Campylobacter jejuni 81-176, Infect. Immun. 70 (2002) 6242e6250. [33] J.K. Ichikawa, A. Norris, M.G. Bangera, G.K. Geiss, A.B. van’t Wout, R.E. Bumgarner, S. Lory, Interaction of Pseudomonas aeruginosa with epithelial cells: identification of differentially regulated genes by

T.-L. Lin et al. / Microbes and Infection 8 (2006) 2756e2765

[34]

[35] [36]

[37]

expression microarray analysis of human cDNAs, Proc. Natl. Acad. Sci. USA 97 (2000) 9659e9664. K. Lechward, O.S. Awotunde, W. Swiatek, G. Muszynska, Protein phosphatase 2A: variety of forms and diversity of functions, Acta Biochim. Pol. 48 (2001) 921e933. H. Niida, M. Nakanishi, DNA damage checkpoints in mammals, Mutagenesis 21 (2006) 3e9. C.Z. Song, G. Gavriilidis, H. Asano, G. Stamatoyannopoulos, Functional study of transcription factor KLF11 by targeted gene inactivation, Blood Cells Mol. Dis. 34 (2005) 53e59. J. Han, C. Flemington, A.B. Houghton, Z. Gu, G.P. Zambetti, R.J. Lutz, L. Zhu, T. Chittenden, Expression of bbc3, a pro-apoptotic BH3-only

2765

gene, is regulated by diverse cell death and survival signals, Proc. Natl. Acad. Sci. USA 98 (2001) 11318e11323. [38] R.Q. Hintzen, S.M. Lens, M.P. Beckmann, R.G. Goodwin, D. Lynch, R.A. van Lier, Characterization of the human CD27 ligand, a novel member of the TNF gene family, J. Immunol. 152 (1994) 1762e1773. [39] D. Hofreuter, S. Odenbreit, G. Henke, R. Haas, Natural competence for DNA transformation in Helicobacter pylori: identification and genetic characterization of the comB locus, Mol. Microbiol. 28 (1998) 1027e1038. [40] D. Kersulyte, B. Velapatino, A.K. Mukhopadhyay, L. Cahuayme, A. Bussalleu, J. Combe, R.H. Gilman, D.E. Berg, Cluster of type IV secretion genes in Helicobacter pylori’s plasticity zone, J. Bacteriol. 185 (2003) 3764e3772.