A Conserved Helicobacter pylori Gene, HP0102, Is Induced Upon Contact With Gastric Cells and Has Multiple Roles in Pathogenicity

Abstract

Helicobacter pylori, a gram-negative, microaerophilic bacterium, colonizes the gastric mucosa of at least half the world's population [1]. Most H. pylori–infected individuals have chronic-active superficial gastritis throughout their lives, and a significant fraction progresses to peptic ulcer, mucosa-associated lymphoid tissue lymphoma, or even gastric cancer [2–4]. The clinical outcomes of H. pylori infection may be attributed to 2 major factors, the cytotoxin-associated gene CagA and vacuolating cytotoxin VacA, and to a host of accessory virulence determinants [5]. CagA is translocated to the host cell cytoplasm through a type IV secretion system and is phosphorylated by the Src family of tyrosine kinases [5]. CagA can interfere with cell signaling pathways, resulting in altered spreading, migration, and adhesion of epithelial cells, commonly known as the hummingbird phenotype [5]. VacA is a secretory protein that exerts multiple effects on host cells and can induce apoptosis [5].

Many bacterial pathogens have evolved to recognize environmental factors characteristic of the host physiological site of infection as signals for correct spatiotemporal expression of pathogenicity elements [6–10]. Consistent with this common theme, acidic pH and salt concentration have been reported to induce expression of the major virulence gene, cagA, in certain H. pylori strains [11–14]. In addition to environmental parameters, contact with eukaryotic cells is also recognized as a stimulus that modulates virulence gene expression in a number of pathogens [15–23]. The H. pylori virulence genes, cagA and vacA, are strongly induced in bacteria following adherence to epithelial cell lines [22, 23].

Successful infection of a host by a pathogen depends not only on classical virulence factors like toxins but also, very importantly, on efficient colonization of appropriate host niches by the pathogen. Colonization of the stomach by H. pylori is a complex process, which depends on a wide range of bacterial and host factors. Motility of H. pylori is considered to be necessary for colonization of the stomach and enhances pathological outcomes in humans [24]. Bacteria generate propulsion by flagellar rotation but also need to modulate the direction of flagellar rotation to control their trajectory [25, 26]. The chemotaxis system couples environmental sensing to the control of flagellar rotation to enable bacteria to move toward chemoattractants or away from chemorepellents [27]. H. pylori uses chemotaxis to escape the acidic conditions of the gastric lumen and colonize the gastric epithelium. In this context, chemotaxis is considered an essential colonization factor for H. pylori, and chemotaxis mutants have been shown to be colonization defective [28, 29]. The chemotaxis signaling system in H. pylori is similar in many ways to the chemotaxis signaling system of other bacteria [30, 31]. However, in recent studies new members of chemotaxis regulators have been identified that are unique to the family of Epsilonproteobacteria, suggesting that this group of bacteria might have evolved additional regulators for controlling chemotaxis [32].

A study using a genome-saturating library of H. pylori transposon mutants in a mouse infection model identified 223 genes that encode potential colonization factors [28]. One of these genes, HP0102, encodes a predicted glycosyltransferase and was reported to be important for colonization. In this study, we demonstrate that expression of the HP0102 gene is strongly induced following adherence to AGS cells and is necessary for upregulation of cagA expression following adherence to AGS cells. We also demonstrate that HP0102 is necessary for normal chemotaxis and acid-escape response of H. pylori.

MATERIALS AND METHODS

All experiments were performed following standard procedures and are described in the Supplementary Data.

RESULTS

Expression of the HP0102 Gene Is Upregulated in H. pylori Following Adherence to AGS Cells

Semiconfluent monolayers of AGS cells were infected with H. pylori strain G27, and expression of the HP0102 gene was examined in the adhered bacteria at different times after adherence. HP0102 expression was also examined in unadhered bacteria grown under identical conditions without the cell line or isolated from the supernatant of the H. pylori–infected AGS monolayer. A significant increase in HP0102 expression was observed in strain G27 by 2 hours after adherence to AGS cells as compared to the unadhered controls (Figure 1A). HP0102 expression was consistently upregulated in AGS-adhered G27 during the 8-hour examination period (Figure 1A). Upregulation of HP0102 following adherence to AGS cells was also observed in several other H. pylori strains, including Indian clinical isolates (Figure 1B and Table 1).

Table 1.

Open in new tab

HP0102 Sequence Conservation Analysis in Helicobacter pylori Indian Clinical Isolates

H. pylori Strain .	Isolated From Patients With .		cag PAI .	HP0102 Sequence Conservation,a % .	HP0102 Gene Sequence Accession No. .
	DU .	NUD .
I-121	Yes	No	Present	96	KT932667
135 (1)	Yes	No	Present	96	KT932666
I-30	Yes	No	Present	96	KT884537
I-90	Yes	No	Present	96	KT884534
I-230	Yes	No	Present	96	KT884538
San52	No	Yes	Present	96	KT884542
San77	No	Yes	Present	96	KT884535
San10	No	Yes	Present	95	KT884539
San64	No	Yes	Present	95	KT884540
162	Yes	No	Present	95	KT884533
Osc51	No	Yes	Present	95	KT884545
I-10	Yes	No	Present	94	KT884541
Osc30	No	Yes	Present	94	KT884543
152 1(A)	Yes	No	Present	94	KT884536
155 1(A)	Yes	No	Present	94	KT884544

H. pylori Strain .	Isolated From Patients With .		cag PAI .	HP0102 Sequence Conservation,a % .	HP0102 Gene Sequence Accession No. .
	DU .	NUD .
I-121	Yes	No	Present	96	KT932667
135 (1)	Yes	No	Present	96	KT932666
I-30	Yes	No	Present	96	KT884537
I-90	Yes	No	Present	96	KT884534
I-230	Yes	No	Present	96	KT884538
San52	No	Yes	Present	96	KT884542
San77	No	Yes	Present	96	KT884535
San10	No	Yes	Present	95	KT884539
San64	No	Yes	Present	95	KT884540
162	Yes	No	Present	95	KT884533
Osc51	No	Yes	Present	95	KT884545
I-10	Yes	No	Present	94	KT884541
Osc30	No	Yes	Present	94	KT884543
152 1(A)	Yes	No	Present	94	KT884536
155 1(A)	Yes	No	Present	94	KT884544

Abbreviations: DU, duodenal ulcer; NUD, nonulcer dyspepsia; PAI, pathogenicity island.

^a Sequence conservation analysis was done considering the HP0102 gene sequence of H. pylori strain 26695 as a query.

Table 1.

Open in new tab

HP0102 Sequence Conservation Analysis in Helicobacter pylori Indian Clinical Isolates

H. pylori Strain .	Isolated From Patients With .		cag PAI .	HP0102 Sequence Conservation,a % .	HP0102 Gene Sequence Accession No. .
	DU .	NUD .
I-121	Yes	No	Present	96	KT932667
135 (1)	Yes	No	Present	96	KT932666
I-30	Yes	No	Present	96	KT884537
I-90	Yes	No	Present	96	KT884534
I-230	Yes	No	Present	96	KT884538
San52	No	Yes	Present	96	KT884542
San77	No	Yes	Present	96	KT884535
San10	No	Yes	Present	95	KT884539
San64	No	Yes	Present	95	KT884540
162	Yes	No	Present	95	KT884533
Osc51	No	Yes	Present	95	KT884545
I-10	Yes	No	Present	94	KT884541
Osc30	No	Yes	Present	94	KT884543
152 1(A)	Yes	No	Present	94	KT884536
155 1(A)	Yes	No	Present	94	KT884544

H. pylori Strain .	Isolated From Patients With .		cag PAI .	HP0102 Sequence Conservation,a % .	HP0102 Gene Sequence Accession No. .
	DU .	NUD .
I-121	Yes	No	Present	96	KT932667
135 (1)	Yes	No	Present	96	KT932666
I-30	Yes	No	Present	96	KT884537
I-90	Yes	No	Present	96	KT884534
I-230	Yes	No	Present	96	KT884538
San52	No	Yes	Present	96	KT884542
San77	No	Yes	Present	96	KT884535
San10	No	Yes	Present	95	KT884539
San64	No	Yes	Present	95	KT884540
162	Yes	No	Present	95	KT884533
Osc51	No	Yes	Present	95	KT884545
I-10	Yes	No	Present	94	KT884541
Osc30	No	Yes	Present	94	KT884543
152 1(A)	Yes	No	Present	94	KT884536
155 1(A)	Yes	No	Present	94	KT884544

Abbreviations: DU, duodenal ulcer; NUD, nonulcer dyspepsia; PAI, pathogenicity island.

^a Sequence conservation analysis was done considering the HP0102 gene sequence of H. pylori strain 26695 as a query.

Figure 1.

Induction of the HP0102 gene in AGS-adhered Helicobacter pylori strains. A, RNA was extracted from H. pylori strain G27 (A) or different H. pylori strains (B) after adherence to AGS cells (adhered) and from unadhered H. pylori incubated under identical conditions without cell line (unadhered - AGS) or collected from the supernatants of infected AGS cells (unadhered + AGS). HP0102 expression was estimated by quantitative reverse transcription polymerase chain reaction analysis. RNA was extracted 2, 5, and 8 hours after adherence of strain G27 to AGS cells and 2 hours after adherence of the other strains. Relative expression levels were calculated as fold induction of gene expression in adhered bacteria, considering the expression in unadhered bacteria grown without AGS at 2 hours arbitrarily as 1. Bar diagram represents the mean value ± SD of at least 3 experiments. The statistical significance of differences (P values) in expression of the HP0102 gene between the unadhered (−AGS) and adhered H. pylori cells were obtained by the Student t test and are indicated. No statistical significance was obtained in differences of HP0102 expression between unadhered (−AGS) bacteria and unadhered (+AGS) bacteria. *P ≤ .05 and **P ≤ .01. Abbreviation: NS, not significant.

Open in new tabDownload slide

Parenthetically, no difference in HP0102 expression was observed between unadhered bacteria grown without cell line and those isolated from the supernatant of infected AGS monolayers (Figure 1), suggesting that HP0102 induction required direct contact of the bacteria with the AGS cells and was not due to a component secreted by the cell line.

Prevalence and Conservation of HP0102 Gene in H. pylori Strains

To detect the prevalence of the HP0102 gene in H. pylori strains, the gene sequence of strain 26695 was used as a query to search for homologous sequences in H. pylori genomes (available at: http://blast.ncbi.nlm.nih.gov/Blast.cgi). The gene was present in all 69 H. pylori strains whose genome sequences were available, and a high degree of sequence conservation of the gene (>93%) was observed among the different strains (Supplementary Data). Fifteen additional H. pylori clinical strains isolated from India (Table 1) were also examined for the presence of the HP0102 gene by polymerase chain reaction (PCR) analysis, using primers flanking the gene. Sequencing of the PCR products revealed a high degree of conservation of the gene sequence (Table 1). These results indicated that the HP0102 gene is almost ubiquitously present and highly conserved in H. pylori strains. Furthermore, an HP0102 homolog was present in at least 11 different Helicobacter species examined (Supplementary Data). To investigate the function of HP0102, a H. pylori G27ΔHP0102 strain was constructed and analyzed. Expression of the genes HP0103 and HP0101, located upstream and downstream, respectively, of HP0102, was comparable between G27ΔHP0102 and the parental G27 strain (Supplementary Data), indicating that the mutation in HP0102 did not affect expression of the neighboring genes.

Role of HP0102 in Induction of the Hummingbird Phenotype in AGS Cells

H. pylori G27 wild-type (WT) and G27ΔHP0102 strains were added to AGS monolayers at different multiplicities of infection (MOIs), and the number of adhered bacteria was enumerated at different times after adherence (Supplementary Data). With increasing MOIs, the number of adhered bacteria increased progressively. No significant difference in the number of AGS-adhered bacteria was observed between G27WT and ΔHP0102 strains at any MOI during the 24-hour period of examination (Supplementary Data), indicating that HP0102 might have no role in the adherence of H. pylori to gastric cells. However, a remarkable difference in morphology was observed between AGS cells infected with the WT strain and those infected with the ΔHP0102 strain.

When infected with the H. pylori G27WT strain at increasing MOIs (10, 25, 50, and 100), progressively larger number of AGS cells exhibited the hummingbird phenotype (Figure 2). At an MOI of 50, a pronounced hummingbird phenotype was observed in a large majority (about 70%) of AGS cells after 5 hours of adherence of the G27WT strain. However, much lower scattering and elongation was observed in AGS cells after adherence of the ΔHP0102 strain (Figure 2). Complementation of the ΔHP0102 strain with the full-length HP0102 gene integrated into the chromosome (strain ΔHP0102/HP0102) caused reversal of the mutant phenotype and produced the hummingbird phenotype, similar to the WT strain (Figure 2), confirming that HP0102 has a role in induction of the hummingbird phenotype.

Figure 2.

Morphology of AGS cells following adherence of Helicobacter pylori strains. AGS monolayers (uninfected) were incubated with H. pylori strains G27 (wild type [WT]), ΔHP0102, and ΔHP0102 complemented with the HP0102 gene (ΔHP0102/HP0102) at different multiplicities of infection (MOIs) for 5 hours, washed to remove unadhered bacteria, and observed on an inverted microscope in phase contrast at 20× original magnification. The number of elongated cells were counted and expressed as the percentage of the total cells in each sample. Bar diagram represents the mean value ± SD of 3 experiments. The statistical significance of the differences (P values) in number of elongated cells between WT and the ΔHP0102 mutant or complemented strain (ΔHP0102/HP0102) is indicated. *P ≤ .05 and **P ≤ .01. Abbreviation: NS, not significant.

Open in new tabDownload slide

One of the consequences of H. pylori infection is induction of the proinflammatory signaling cascades that ultimately lead to host cell secretion of interleukin 8 (IL-8) [5]. Consistent with the decreased scattering and elongation induced in AGS cells by the G27ΔHP0102 strain, the amount of IL-8 detected in culture supernatants of the G27ΔHP0102-infected AGS cells was lower than that in supernatants of AGS cells infected by the G27WT bacteria (Supplementary Data). Complementation of the G27ΔHP0102 strain with the full-length HP0102 gene restored IL-8 secretion to WT levels (Supplementary Data). Of note, there was a significant difference in IL-8 production between uninfected and infected AGS cell cultures (Supplementary Data). Since the hummingbird phenotype and IL-8 secretion are associated with production and subsequent translocation of CagA from adhered H. pylori to the gastric cells, cagA expression and translocation was next examined in AGS-adhered G27WT and G27ΔHP0102 cells.

Role of HP0102 in cagA Induction in AGS-Adhered H. pylori

Upregulation of cagA expression by approximately 4-fold was consistently observed in the G27WT strain within 2 hours of adherence to AGS cells, as reported earlier (Figure 3A and Supplementary Data) [22, 23]. However, consistent with the lack of hummingbird induction and low IL-8 secretion, practically no cagA upregulation was observed in the AGS-adhered G27ΔHP0102 strain during the 8-hour examination period (Figure 3A and Supplementary Data). In the complemented ΔHP0102/HP0102 strain, cagA expression was similar to that in the G27WT strain (Figure 3A). It was noted that, in unadhered bacteria, there was no difference in cagA expression between the WT and ΔHP0102 strains (Figure 3A).

Figure 3.

Gene expression and Western blot analysis in AGS-adhered and unadhered Helicobacter pylori strains. Expression of cagA (A), fliA (B), and flaA (C) was estimated in unadhered and AGS-adhered G27 wild-type (WT), ΔHP0102 mutant, and ΔHP0102/HP0102 complemented strains by quantitative reverse transcription polymerase chain reaction (PCR) analysis. Transcript levels are expressed relative to that in the unadhered WT strain, arbitrarily taken as 1. Results represent the average of at least 3 independent experiments. Error bars represent SD. The statistical significance (P values) of the differences in gene expression between the unadhered (−AGS) and adhered H. pylori cells were obtained by the Student t test and are indicated. No statistical significance was obtained in differences in gene expression between unadhered (−AGS) bacteria and unadhered (+AGS) bacteria. D, Western blot analysis of CagA in unadhered WT and H. pylori–infected AGS cultures. UreB was used as an internal control. Phosphorylated CagA was estimated in H. pylori–infected AGS cultures, using P-Tyr antibody, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control. E, Schematic depicting transcripts initiating from the σ⁸⁰ and σ²⁸ promoters of cagA, using the rapid amplification of complementary DNA ends (RACE) primers. F, 5′ RACE PCR analysis of cagA transcripts in unadhered WT (a), AGS-adhered WT (b), unadhered ΔHP0102 mutant (c), AGS-adhered ΔHP0102 mutant (d), unadhered ΔHP0102/HP0102 strain (e), and AGS-adhered ΔHP0102/HP0102 strain (f). *P ≤ .05 and **P ≤ .01. Abbreviations: M, molecular size markers; NS, not significant.

Open in new tabDownload slide

Western blot analysis indicated that the level of CagA was higher in the G27WT strain than in the G27ΔHP0102 strain following adherence to AGS cells (Figure 3D). The relative difference in CagA levels between AGS-adhered WT and ΔHP0102 was reflected in the levels of phosphorylated CagA in infected AGS cells, indicating that the activity of the type IV secretion system responsible for translocating CagA remains unaffected in the ΔHP0102 mutant (Figure 3D). Results obtained with the complemented ΔHP0102/HP0102 strain were similar to those for the WT strain (Figure 3D).

HP0102 Influences cagA Expression by Regulating Expression of fliA

We have previously demonstrated that a σ²⁸-RNAP–dependent promoter located downstream of the annotated σ⁸⁰-RNAP promoter of cagA is activated following adherence of H. pylori to AGS cells and that cagA transcripts are initiated from both promoters in AGS-adhered H. pylori [23]. To examine whether HP0102 has a role in activation of the σ²⁸-RNAP–dependent promoter of cagA, expression of fliA, encoding σ²⁸, was examined in the WT, ΔHP0102, and ΔHP0102/HP0102 strains. fliA expression increased about 5.5-fold upon adherence of the G27WT strain to AGS cells, but no upregulation was observed in the AGS-adhered ΔHP0102 strain (Figure 3B). fliA expression in unadhered G27WT and ΔHP0102 strains was comparable, indicating that HP0102 might have a role in upregulation of fliA expression, specifically in AGS-adhered H. pylori. The complemented strain behaved essentially like the WT strain.

To examine the status of functional σ²⁸ in the WT strain and the ΔHP0102 mutant, expression of flaA, known to be dependent on σ²⁸ [33, 34], was compared between the 2 strains. flaA expression increased about 4-fold following adherence of the WT strain and the ΔHP0102/HP0102 strain to AGS cells. However, no upregulation was observed in the AGS-adhered ΔHP0102 strains (Figure 3C). Thus, functional σ²⁸ increased following adherence to AGS cells, and the induction was dependent on HP0102.

Because we previously identified a σ²⁸-RNAP–dependent promoter in addition to the σ⁸⁰-RNAP–dependent promoter of cagA expression [23], we next used 5′ rapid amplification of complementary DNA ends (RACE) PCR to determine which promoter was active in unadhered and AGS-adhered G27WT and ΔHP0102 strains. A single transcript likely to have originated from the σ⁸⁰ promoter was detected in unadhered WT and ΔHP0102 strains. A shorter transcript whose size suggested that it might have originated from the σ²⁸ promoter was detected in the AGS-adhered WT and ΔHP0102/HP0102 strains but not in the AGS-adhered ΔHP0102 strain (Figure 3F).

To determine whether σ²⁸ also has a role in upregulation of HP0102 following adherence to AGS cells, HP0102 expression was examined in unadhered and AGS-adhered H. pylori ΔfliA [23]. No difference in HP0102 expression was observed between the WT and ΔfliA strains (Supplementary Data), indicating that σ²⁸ has no role in upregulation of HP0102 in AGS-adhered H. pylori.

Role of HP0102 in Motility and Chemotaxis

It has been reported that a transposon insertion mutation in the HP0102 gene locus resulted in a mice colonization defect [28]. Successful colonization of the stomach by H. pylori is a complex process that depends on a wide range of host and bacterial factors. We examined urease production, macrophage survival, motility, and chemotaxis in the G27WT and ΔHP0102 strains since these factors are known to be necessary for colonization. No difference in urease production and survival in the macrophage cell line (RAW264.7) were observed between the WT and ΔHP0102 strains (data not shown). However, swarming assays on soft-agar plates indicated a significant reduction in the swarm diameter of ΔHP0102 as compared to the WT or ΔHP0102/HP0102 strain (Figure 4A). The motility defect was not due to reduced growth, as growth rates of the strains were comparable (Supplementary Data). Next, swimming behavior of the WT, ΔHP0102, and ΔHP0102/HP0102 strains in liquid medium was observed by video microscopy and single-cell tracking of individual bacteria (Supplementary Data). Although the strains were all motile in liquid medium, the number of stops and reversals in the swimming trajectory was significantly higher in the ΔHP0102 strain as compared to the WT or complemented strains (Figure 4B). Upon addition of acid, the number of stops and reversals increased immediately in the WT and ΔHP0102/HP0102 bacteria, although no difference was observed in the ΔHP0102 strain (Figure 4B). Furthermore it was noted that, prior to acid exposure, the swimming patterns of all the strains were essentially random. In response to acid, the WT and ΔHP0102/HP0102 bacteria demonstrated an average net movement away from the acid (Figure 4D). However, the ΔHP0102 bacteria maintained their random swimming trajectory even after acid exposure (Figure 4D), thereby suggesting that the ΔHP0102 strain might have a poor acid-escape response (Supplementary Data).

Figure 4.

Motility and chemotaxis. A, Swarm plate assay of Helicobacter pylori G27 wild-type (WT), ΔHP0102 mutant, and ΔHP0102/HP0102 complemented strains. Bar diagram represents the average ± standard deviation (SD) swarm diameter of 3 experiments. The statistical significance of the differences (P values) in swarm diameter between the WT and ΔHP0102 mutant or ΔHP0102/HP0102 complemented strains are indicated. B, Swimming of G27WT, ΔHP0102, and ΔHP0102/HP0102 strains in liquid medium before (neutral) and after addition of acid was observed by video microscopy, and the number of stops of individual bacteria (n = 20) during a 5-second period was counted. The bar diagram represents the mean value ± SD of 3 experiments. The statistical significance of the differences (P values) in the number of stops before and after addition of acid in the different strains is indicated. C, Illustration of a wet mount chemotaxis assay (Supplementary Data). D, Analysis of the swimming trajectories of WT, ΔHP0102, and ΔHP0102/HP0102 strains. Using MTrackJ software, the 2D trajectories of individual bacterial cell were determined. Each trajectory is represented by a number, and the position of the number represents the final point of the trajectory. See Supplementary Data and Supplementary Data for H. pylori G27WT before and after addition of acid, Supplementary Data and Supplementary Data for the ΔHP0102 mutant before and after addition of acid, and Supplementary Data and Supplementary Data for the ΔHP0102/HP0102 complemented strain before and after addition of acid, respectively. ***P ≤ .001. Abbreviation: NS, not significant.

Open in new tabDownload slide

Chemotaxis Per Se Does Not Affect cagA Expression Following Adherence to AGS Cells

Two distinct phenotypes, altered chemotaxis and lack of cagA induction following AGS adherence, were associated with the ΔHP0102 strain. To examine whether there was cross talk between chemotaxis and cagA induction, 2 different chemotaxis mutants were constructed, ΔtlpB (hyperchemotactic) [35] and ΔcheA (nonchemotactic) [36], and cagA expression was examined in the mutants following adherence to AGS cells. cagA induction occurred in the AGS-adhered ΔtlpB and ΔcheA mutants, similar to the WT strain. Also, the expression of fliA and flaA in the ΔtlpB and ΔcheA mutants were similar to expression in the WT strain following AGS adherence, suggesting that chemotaxis per se does not affect cagA induction following host cell contact (Figure 5).

Figure 5.

Induction of cagA, fliA, and flaA in chemotaxis mutants of Helicobacter pylori. Expression of cagA, fliA, and flaA in unadhered and AGS-adhered wild-type (WT), ΔtlpB, and ΔcheA strains was examined by quantitative reverse transcription polymerase chain reaction analysis. Transcript levels are expressed relative to that in the unadhered WT strain, arbitrarily taken as 1. The bar diagram represents the mean value ± standard deviation (SD) of at least 3 experiments. Error bars represent SDs. The statistical significance (P values) of the differences in gene expression between AGS-adhered and unadhered H. pylori strains are indicated. *P ≤ .05 and **P ≤ .01.

Open in new tabDownload slide

Role of HP0102 in H. pylori Strain 26695

The role of HP0102 in regulating cagA expression in AGS-adhered H. pylori was examined in another strain, 26695. Similar to H. pylori strain G27, expression of HP0102 was upregulated >7-fold in strain 26695 following adherence to AGS cells (Figure 1B). Adherence of 26695ΔHP0102 to AGS cells resulted in much less cell scattering and reduced IL-8 production as compared to the 26695WT strain (Supplementary Data). Furthermore, similar to H. pylori G27, expression of cagA, fliA, and flaA were upregulated in AGS-adhered H. pylori 26695 WT but not in the ΔHP0102 mutant (Supplementary Data). 5’ RACE analysis was performed to elucidate the transcription start site of cagA, and the results were similar to that obtained for strain G27 (Supplementary Data). Since strain 26695 is nonmotile because of a mutation in fliP [37], the role of HP0102 in chemotaxis could not be examined in this strain.

Expression of HP0102 and cagA Is Upregulated in H. pylori Following Adherence to Primary Gastric Cell Culture

Because AGS is a gastric cancer cell–derived cell line, we next examined whether HP0102 has a role in cagA upregulation during H. pylori infection of nontransformed primary cell cultures developed from human gastric biopsy samples. Upon adherence of H. pylori G27WT to these primary cells, HP0102 and cagA were both upregulated, but the upregulation was not observed upon infection with the G27ΔHP0102 strain (Figure 6), similar to the results obtained with the AGS cell line (Figures 1 and 3).

Figure 6.

Induction of HP0102 and cagA upon infection of primary gastric cells. A, RNA was extracted from Helicobacter pylori G27 2 hours after adherence to primary gastric cells (adhered) or from H. pylori incubated under identical conditions without cell line (unadhered), and HP0102 expression was estimated by quantitative reverse transcription polymerase chain reaction analysis. B, Expression of cagA was estimated in AGS-adhered or unadhered H. pylori G27 wild type (WT) and ΔHP0102. Relative expression levels were calculated as fold induction of gene expression in adhered bacteria, considering the expression in unadhered bacteria arbitrarily as 1. Bar diagram represents the means ± SD of at least 3 experiments. The statistical significance of differences (P values) in gene expression between the unadhered and adhered H. pylori cells was obtained by the Student t test and are indicated. *P ≤ .05 and **P ≤ .01. Abbreviation: NS, not significant.

Open in new tabDownload slide

DISCUSSION

The ubiquitous presence and high degree of conservation of the HP0102 gene in all 84 H. pylori strains, including 15 Indian isolates, examined, as well as upregulation of expression of the gene following adherence to gastric cells suggested that the gene might have important functions. Indeed, it has been demonstrated in this study that HP0102 has a role in 2 different processes: (1) induction of the hummingbird phenotype in AGS cells following H. pylori adherence and (2) chemotaxis and acid-escape response.

One of the most important virulence factors of H. pylori is CagA that is translocated from adhered bacteria to the gastric cells and produces a characteristic scattering and elongation [5]. We have earlier demonstrated that contact with host cells rapidly upregulated expression of cagA in AGS-adhered H. pylori strains [22, 23]. In our endeavor to identify factors necessary for host cell contact–dependent cagA upregulation, we have reported that cagA expression was specifically upregulated in AGS-adhered H. pylori from a σ²⁸-RNAP–dependent alternative promoter of cagA [23].

Interestingly, cagA transcription start site identification by 5’ RACE indicated that HP0102 has a role in activation of the σ²⁸-RNAP–dependent promoter following H. pylori adherence to AGS cells, since transcripts originating from the σ²⁸ promoter of cagA were detected in the AGS-adhered WT strain but not in the ΔHP0102 mutant strain. This may be due to higher σ²⁸ levels in the adhered WT strains as compared to the ΔHP0102 mutant, as suggested by the observation that expression of the σ²⁸-dependent flaA was significantly higher in the AGS-adhered WT strain than in the ΔHP0102 mutant. Thus, HP0102 has a role in increasing the concentration of functional σ²⁸ in AGS-adhered H. pylori. We have recently reported that another factor, the flagellar hook length control protein FliK exerts a similar effect [23]. Similar to HP0102, H. pylori fliK expression was upregulated following adherence to AGS cells, and a ΔfliK mutant, defective in hook basal body assembly, had lower levels of free σ²⁸ than the WT strain following adherence to AGS cells; consequently, expression from the σ²⁸-dependent promoter of cagA was not observed in the AGS-adhered ΔfliK mutant [23]. Thus, in both the ΔfliK and ΔHP0102 mutants, after adherence to AGS cells, levels of active σ²⁸ was lower than in the WT strain; consequently, induction of σ²⁸-dependent transcription of cagA following adherence to AGS cells was not observed. Expression of fliK was, however, not affected in the ΔHP0102 mutant, and the induction of fliK reported in the WT strain following AGS adherence was also observed in the ΔHP0102 mutant (data not shown). Thus, we have now identified 2 factors, FliK and HP0102, that have nonoverlapping roles in cagA expression and are required for upregulation of cagA, specifically in host cell–associated H. pylori. Further studies of ΔfliK and ΔHP0102 infection in mice models are required to establish the role of the gene products in H. pylori pathogenesis.

Recent studies of H. pylori clinical isolates from 2 diverse geographical locations, Colombia [38] and Portugal [39], have suggested a strong association between the presence of a DNA motif (AATAAGATA) upstream of the cagA open reading frame (ORF) and high levels of cagA expression and disease severity. The motif is included within the σ²⁸ RNAP promoter upstream of the cagA ORF [23]. In this context, it might be important to consider that the mere presence of this motif in a H. pylori strain is not sufficient to drive cagA expression to high levels, but it is also important to determine the functional status of the factors FliK and HP0102 to predict whether CagA upregulation would occur following infection with a particular strain. A study to examine the correlation between cagA upregulation, fliK, and HP0102 status in a large number of H. pylori strains isolated from individuals with different clinical symptoms has recently been initiated in our laboratory. In addition, further work is ongoing to determine the correlation between HP0102 and other fundamental processes in H. pylori.

HP0102 has a role in another phenotype important for H. pylori pathogenicity—chemotaxis. ΔHP0102 mutants have normal flagella and are motile, but they exhibit aberrant flagellar rotation. As compared to the WT strain, the ΔHP0102 mutant switches flagellar rotation more frequently even under standard culture conditions. When exposed to acid, the WT strain shows a definite directional movement away from the acid, but the mutant continues to switch flagellar rotation randomly and hence has a poor acid-escape response. H. pylori has 4 chemoreceptors, of which 1, TlpB, is required for chemorepulsive responses to acid [35]. HP0102 is encoded in an operon downstream from tlpB (HP0103). The expression of tlpB was similar between WT and ΔHP0102 strains (Supplementary Data), but whether HP0102 can influence the activity of TlpB either directly or indirectly needs further investigation.

It is intriguing that a glycosyltransferase, HP0102, mediates the diverse effects reported in this study. Glycosyltransferases have been reported to affect virulence, motility, and biofilm formation in pathogens such as Pseudomonas syringae, Xanthomonas citri, and Streptococcus species [40–44], although the molecular basis of these effects is as yet largely unknown. Based on the results reported in this study, we hypothesize that the primary effect of HP0102 is on the upregulation of fliA (σ²⁸) expression. H. pylori has evolved to be able co-opt σ²⁸ to increase cagA expression following adherence to host cells. This integration presumably ensures that 2 essential virulence determinants, motility/chemotaxis and CagA, are coordinately expressed.

Notes

Acknowledgments. We thank all members of the laboratory, for cooperation and helpful discussions; Kalidas Paul, for suggestions and excellent technical support; D. Scott Merrell, Uniformed Services University of Health Sciences (Bethesda, MD), for the generous gift of H. pylori G27 strain; and Dr Apurba K. Pal and Dr Kingshuk K. Dhar, R. G. Kar Medical College and Hospital, for human gastric biopsy materials.

Financial support. The work was supported by the Council of Scientific and Industrial Research–Indian Institute of Chemical Biology (grant number BSC0119) and the Council of Scientific and Industrial Research (research fellowship to S. B.).

Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References