Immunohistochemical diagnosis of the cagA‐gene genotype of Helicobacter pylori with anti‐East Asian CagA‐specific antibody

Abstract

Cytotoxin‐associated antigen A (CagA) protein produced by Helicobacter pylori is proposed to be associated with the pathogenesis of gastric cancer as well as gastritis and gastroduodenal ulcer. It has been reported that the CagA of H. pylori widespread in East Asian countries, where the mortality rate due to gastric cancer is high, is structurally different from that in Western countries, where the gastric cancer mortality rate is relatively low. In this study, we generated an antibody, East Asian CagA‐specific antibody (α‐EAS Ab), which is specifically immunoreactive with East Asian CagA but not with Western CagA. The CagA was immunohistochemically detected at the surface of the gastric mucosa. Interestingly, positive immunoreactivity was also detected in the nucleus and cytoplasm of the infected gastric epithelium, suggesting that CagA may play some pathogenic role in both the nucleus and cytoplasm. Immunohistochemistry of 47 gastric biopsy specimens detected East Asian CagA‐positive H. pylori in 43 cases. In 46 of the 47 cases examined, the data obtained by immunohistochemistry were completely consistent with those obtained by sequencing of the cagA gene of the isolated strain, suggesting that our immunohistochemical method is reliable and useful for diagnosis of the cagA genotype. (Cancer Sci 2007; 98: 521–528)

Helicobacter pylori is the commonly widespread pathogenic microorganism that plays a causative role in gastritis and gastroduodenal ulcer.^{( 1} , ^{2 )} Furthermore, recent reports have stressed that H. pylori infection may also be associated with gastric cancer.^{( 3} , ⁴ , ⁵ , ^{6 )} On the other hand, although the incidences of H. pylori infection are high in both East Asian and Western countries, the incidences of gastric carcinoma are significantly higher in East Asian countries.^{( 7 )} With respect to this inconsistency, Azuma et al. have reported that the differences in clinical outcome between East Asian and Western countries may be attributed to structural differences in the cytotoxin‐associated antigen A (CagA) protein of H. pylori.^{( 7 )} CagA has a unique domain, the EPIYA motif, which is a 5‐amino‐acid sequence (Glu‐Pro‐Ile‐Tyr‐Ala) located in the carboxyl‐terminal variable region of the protein (Fig. 1).^{( 8} , ⁹ , ^{10 )} The EPIYA motif is found in four distinct EPIYA sites, EPIYA‐A, B, C and D, each of which is defined by the amino‐acid sequence that surrounds the EPIYA sequence.^{( 10} , ^{11 )} The CagA from H. pylori strains circulating in Western countries, designated Western CagA, has EPIYA‐A and EPIYA‐B sites, followed by a single EPIYA‐C site or 1–3 times‐repeated EPIYA‐C sites (Fig. 1), whereas in East Asian countries, the structure of the EPIYA motif of CagA, designated East Asian CagA, is distinct from that of Western CagA.^{( 10} , ^{11 )} East Asian CagA has EPIYA‐A and EPIYA‐B, but lacks EPIYA‐C. The third domain of the EPIYA motif of East Asian CagA has been designated EPIYA‐D (Fig. 1).^{( 10} , ^{11 )} Furthermore, Azuma et al. have reported that the severities of gastritis and atrophy are significantly higher in patients infected with the East Asian cagA‐positive strain than in patients infected with the Western cagA‐positive strain, and that the prevalence of the East Asian cagA‐positive strain is associated with the rate of gastric cancer mortality.^{( 10} , ^{12 )} Thus, these findings suggest that genotyping of the cagA gene may be necessary in order to predict the risk of gastric carcinogenesis in patients with H. pylori‐positive gastritis.

Figure 1.

Structure of cytotoxin‐associated antigen A protein. (a) Schematic representation of East Asian CagA and Western CagA proteins isolated in this study. The EPIYA motif present at the C terminal of the protein is indicated as a black box. Domain structures of Western CagA possessing an A‐B‐B‐B, A‐C or A‐B‐C type EPIYA motif and those of East Asian CagA possessing an A‐B‐D or A‐B‐D‐D type EPIYA motif are shown. The f1 and r1 primers were designed for amplifying the entire region of the EPIYA motifs by PCR. The f2, f3 and r2 primers were designed for sequencing of EPIYA motifs. The f4 and r3 primers were designed for amplifying the full length of the cagA gene. (b) Alignment of the amino acid sequences of Western CagA (A‐B‐C type EPIYA motifs) and East Asian CagA (A‐B‐D type EPIYA motifs). The EPIYA motifs are underlined. The East Asian‐specific (EAS) amino acid sequence ‘AINRKIDRINKIASAGKG’ that was used as the immunogen is shown by an open box.

In this study, we successfully generated an antibody that is specifically immunoreactive with East Asian CagA but not with Western CagA. We propose that immunohistochemical detection of East Asian CagA‐positive H. pylori is an easy, useful and less expensive diagnostic method applicable to clinical use.

Materials and Methods

Generation of East Asian CagA specific antibody.  East Asian CagA specific peptide, AINRKIDRINKIASAGKG (see Fig. 1), was synthesized (OPERON Biotechnologies, Tokyo, Japan). New Zealand white rabbits were immunized subcutaneously on the back with 1 mg of keyhole limpet hemocyanin (KLH)‐conjugated synthetic peptide emulsified (1/1 v/v) with Freund's complete adjuvant (FCA). The collected antisera were affinity‐purified using a peptide‐coupled HiTrap NHS‐activated column (Amersham Biosciences, UK).

Patients.  Biopsy specimens were obtained from 47 patients by upper gastroduodenal endoscopy at the Department of Gastroenterology, Oita University Hospital. All patients gave informed consent for use of their samples before their endoscopies. Part of each biopsy specimen was subjected to bacterial culture as described below and the other part was used for routine histopathologic examination.

Isolation and in vitro culture of H. pylori.  Each biopsy specimen was homogenized in saline and inoculated onto Mueller Hinton II Agar (Becton Dickinson, NJ, USA) plates supplemented with 7% horse blood without antibiotics. All plates were incubated at 37°C for up to 10 days under anaerobic conditions with a gas mixture of 10% O₂, 5% CO₂, and 85% N₂. Four isolated single colonies were selected per patient, cultured, and then used for infection studies. Isolates were stored at −80°C in Brucella Broth (Difco, NJ, USA) containing 10% dimethylsulfoxide and 10% horse serum.

Infection experiment.  The human gastric epithelial cell line AGS was cultured in F‐12 mixed medium (Gibco BRL, Grand Island, NY, USA) containing 10% (v/v) fetal bovine serum. After AGS cells were grown to 80% confluence on the tissue culture dishes and washed twice with antibiotic‐free cell culture medium, bacterial cells were added to the culture medium at a multiplicity of infection (MOI) of 100. Five hours after infection, infected cells were washed with 1 × phosphate‐buffered saline (PBS) three times, and subjected to confocal microscopy and electron microscopy.

Nucleotide sequence of the EPIYA motif of cagA.  Genomic DNA of each H. pylori isolate was extracted according to the CTAB (hexadecyltrimethylammonium bromide) method,^{( 13 )} and suspended in 40 µL of TE buffer (10 mM Tris HCl and 1 mM EDTA). A DNA fragment of about 1100 bp covering the whole EPIYA motif of the cagA gene was amplified by polymerase chain reaction (PCR) using the following primer sets, f1:5′‐ATGACTAACGAAAC TATTGAT‐3′ (forward), and r1: 5′‐TTAAGATTTTTGGAAAC CACC‐3′ (reverse) (Fig. 1). DNA sequencing of these DNA fragments was performed using an ABI Prism 310 Genetic Analyzer (Applied Biosystems, CA, USA) according to the manufacturer's manual. The primers used for sequencing contained the forward primers, f2:5′‐CCCATTTATGCTCAAGTTA‐3′ and f3:5′‐AACAGAAGCCACAACGCTCA‐3′, both of which were complementary to the sequences of the 5′ region of the cagA gene and the reverse primer r2:5′‐GTTGCCAAAATGACC TGTT‐3′ complementary to the 3′ region (Fig. 1).

GST‐fusion protein of East Asian and Western cagA.  The full‐length cDNAs encoding East Asian cagA and Western cagA, respectively, were amplified by PCR with LA‐taq DNA polymerase (Takara, Kyoto, Japan) using the forward primer, f4:5′‐ATGACTAACGAAACTATTGAT‐3′, and the reverse primer, r3: 5′‐TTAAGATTTTTGGAAACCACC‐3′, followed by directional subcloning into the glutathione‐S‐transferase (GST) expression vector (pGEX‐4T) (Amersham Biosciences, UK). GST fusion proteins were induced and affinity‐purified according to the method described previously.^{( 14 )}

Western blot analysis.  Western blot analysis was performed as described.^{( 15 )} Briefly, proteins were separated by SDS‐PAGE and then transferred to NitroBind nitrocellulose membrane (0.45 µm, Osmonic Inc. Gloucester, MA,USA). After blocking with BlockAce (Snow Brand Milk Products, Japan), the membrane was incubated with the first antibody, α‐EAS Ab, or α‐CagA Ab (Santa Cruz, CA, USA). After washing, the membranes were incubated with horseradish peroxidase‐conjugated secondary antibodies diluted 1:1000 in diluting solution (DAKO, Denmark). After washing, the signals were visualized on Hyperfilm (Amersham Biosciences, UK) using enhanced chemiluminescence (ECL Western blotting detection kit, Amersham Biosciences, UK).

Confocal microscopy.  AGS cells were plated and grown on 25‐mm cover slips (Matsunami, Japan) overnight, infected with H. pylori, and then fixed with 4% paraformaldehyde for 15 min. After being permeabilized with 0.3% Triton X‐100 in 1 × PBS for 15 min, the cells were incubated with an α‐EAS Ab diluted 1:200 or α‐CagA Ab diluted 1:20 in diluting solution (DAKO, Denmark) at 4°C for 16 h. After washing, the cells were incubated for 2 h with an Alexa Fluor 488‐conjugated goat antirabbit Ab diluted 1:1000 (Molecular Probes, Eugene, OR, USA) in 1 × PBS. The nuclei were stained with propidium iodide (PI). The mounted coverslips were then observed using a confocal microscope, Zeiss LSM5 PASCAL (Carl Zeiss Inc. Jena, Germany).

Immunohistochemistry.  Immunohistochemistry was performed as described previously.^{( 15 )} Briefly, after antigen retrieval and inactivation of endogenous peroxidase activity, tissue sections were incubated with α‐East Asian CagA specific (EAS) Ab diluted 1:2000 with diluting solution (DAKO, Denmark) or α‐H. pylori Ab (DAKO, Denmark) overnight at 4°C. After washing, the sections were incubated with biotinylated goat antirabbit IgG (Nichirei Co., Japan), followed by incubation with a solution of avidin‐conjugated horseradish peroxidase (Vectastain Elite ABC kit; Vector Laboratories Inc., Burlingame, CA, USA). Peroxidase activity was detected using H₂O₂/diaminobenzidine substrate solution. Positive immunoreactivity was judged as follows. For all cases, we performed Giemsa staining using a serial section to identify the presence of H. pylori. If the H. pylori identified by Giemsa staining was found to be positively immunostained, we judged the case as positive.

Immunoelectron microscopy.  Infected AGS cells and biopsy tissue samples were fixed in 2% glutaraldehyde in 0.2 M Tris‐HCl at pH 7.4 for 16 h, and then incubated in 1% osmium tetroxide for 2 h, dehydrated in ethanol, and embedded in Epok 812. Immunoelectron microscopy was performed as described in detail previously.^{( 16 )} Briefly, thin sections were microwaved for 15 min in Target Retrieval Solution (TRS), pH 10 (DAKO, Denmark), and then incubated with α‐EAS Ab diluted 1:500, α‐CagA Ab diluted 1:50 or without antibody for 30 min at 60°C. Next, after washing, the sections were incubated for 30 min with gold‐conjugated goat antirabbit IgG (Amersham Biosciences, UK). After washing, the ultrathin sections were stained with uranyl acetate and lead citrate, and examined using a transmission electron microscope (JEM‐1200EXII, JEOL).

A‐CagA Ab immunoprecipitation of subcellular extractions of H. pylori‐infected AGS cells.  AGS cells were plated on 10‐cm dishes and grown to 80% confluency, then infected with H. pylori (ATCC43504) for 5 h. Differential extraction of proteins from H. pylori‐infected AGS cells according to their subcellular localization was performed using a ProteoExtract Subcellular Proteome Extraction Kit (Calbiochem, CA, USA) according to the manufacturer's protocol. In short, cytosolic proteins are released with Extraction Buffer I. Subsequently, membranes and membrane organelles are solubilized with Extraction Buffer II, without impairing the integrity of the nucleus and cytoskeleton. Next, nucleic proteins are enriched with Extraction Buffer III. Each of the protein extracts were subjected to Western blotting with α‐acetyl‐Histone H3 (Upstate, VA, USA), α‐GAPDH, and immunoprecipitation with α‐CagA Ab. For immunoprecipitation, α‐CagA Ab was added to each extract under slow rotation at 4°C for 8 h, and then Protein G Plus‐Agarose (Calbiochem, CA, USA) was added at 4°C for 16 h. After washing with PBS containing 0.05% Tween 20, the extracts were heated to 100°C and the supernatants were collect for SDS‐PAGE. Prepared samples were subjected to Western blotting with α‐CagA Ab.

Results

Generation of an antibody that specifically recognizes East Asian CagA.  It has already been reported that the amino acid sequences between the EPIYA‐B and EPIYA‐D motifs of the East Asian CagA (AB267237) differ from those of the Western CagA (ATCC43504).^{( 11} , ^{12 )} We compared the sequence of the 18 amino acids located between the EPIYA‐B and EPIYA‐D motifs with that of the whole length of the Western CagA, and confirmed that these 18 amino acids were East Asian CagA‐specific, similar sequences being absent in Western CagA (Fig. 1). Therefore, we immunized rabbits with synthetic polypeptides corresponding to the East Asian specific (EAS) sequence to raise the anti‐EAS antibody (α‐EAS Ab), as described in Materials and Methods. To confirm the specificity of the α‐EAS Ab, Western blot analysis with this antibody was performed for E. coli expressing GST‐Western CagA and that expressing GST‐East Asian CagA, Western cagA‐positive H. pylori and East Asian cagA‐positive H. pylori. The SDS‐PAGE gel stained with Coomassie blue showed that the amount of protein loaded to each lane was comparable (Fig. 2; lanes 1–4). As shown in Fig. 2 (lanes 5–8), α‐EAS Ab was found to be immunoreactive with GST‐East Asian CagA and East Asian CagA expressed in H. pylori, but not with GST‐Western CagA and Western CagA expressed in H. pylori. Multiple bands were detected below the main band of GST‐East Asian CagA, and these appeared to be degeneration products. No bands were detected when α‐EAS Ab was preincubated with an excess amount of the peptides that had been used as the immunogen (Fig. 2; lanes 9–12). On the other hand, when α‐CagA Ab was used for Western blot analysis, GST‐Western CagA, GST‐East Asian CagA, Western CagA expressed in H. pylori, and East Asian CagA expressed in H. pylori were all detectable (Fig. 2; lanes 13–16). Thus, these results suggest that α‐EAS Ab is specifically immunoreactive with East Asian CagA.

Figure 2.

Detection of East Asian cytotoxin‐associated antigen A by Western blot analysis with anti‐East Asian CagA‐specific antibody. Coomassie‐stained SDS‐PAGE gel (lanes 1–4) and Western blot analysis with α‐EAS Ab (lanes 5–8), antigen‐preabsorbed α‐EAS Ab (lanes 9–12), and α‐CagA Ab (lanes 13–16). SDS‐PAGE and Western blot analysis with α‐EAS Ab was performed on lysates extracted from E. coli in which GST‐Western CagA fusion protein (lanes 1, 5, 9, 13) or GST‐East Asian CagA fusion protein (lanes 2, 6, 10, 14) had been induced, and lysates extracted from the Western cagA‐positive H. pylori (lanes 3, 7, 11, 15) or East Asian cagA‐positive H. pylori (lanes 4, 8, 12, 16). The presence of CagA protein in both types of H. pylori was confirmed by Western blot analysis with α‐CagA Ab, which is immunoreactive with both East Asian and Western CagA (lanes 13–16).

Immunocytochemical detection of East Asian CagA expressed in H. pylori.  Immunocytochemistry of AGS cells with α‐EAS Ab revealed that those infected with East Asian cagA‐positive H. pylori were positively immunostained, whereas those infected with Western cagA‐positive H. pylori were not (Fig. 3a,b). Positive signals with α‐EAS Ab were not detected when α‐EAS Ab was pretreated with an excess amount of the peptides that had been used as the immunogen (Fig. 3c). On the other hand, α‐CagA Ab, which is immunoreactive with both East Asian CagA and Western CagA, was found to be immunoreactive with AGS cells infected with East Asian cagA‐positive H. pylori as well as Western cagA‐positive H. pylori (Fig. 3e–g). Strongly positive signals tended to be localized at the cell periphery in a dot‐like pattern (arrows in Fig. 3b,f,g), although it was unclear whether they were present within or outside the cell membrane. In addition to these strong signals, we found that the cytoplasm of AGS cells infected with H. pylori was diffusely stained with α‐EAS Ab as well as with α‐CagA Ab, although their signals were weaker than those at the cell periphery.

Figure 3.

Immunocytochemical detection of East Asian cytotoxin‐associated antigen A in H. pylori‐infected AGS cells. A total 5 × 10⁵ AGS cells were plated on glass slides and cultured overnight, followed by co‐culture for 5 h with East Asian type H. pylori (b) (c) (f) or Western type H. pylori (d) (g) at a multiplicity of infection of 100, or without H. pylori (a) (e). After fixation as described in Materials and Methods, the cells were immunocytochemically stained with α‐EAS Ab (a, b, and d) or antigen‐preabsorbed α‐EAS Ab (c) or α‐CagA Ab (e, f and g) and analyzed by confocal laser‐scanning microscopy. The cell periphery (indicated by arrows in b, f and g) and cytoplasm of AGS cells were positively immunostained with α‐CagA Ab (f and g) or with East Asian CagA‐specific antibody (b). Nuclei were counterstained with propidium iodide (PI). Scale bar represents 50 µm (a–g).

Immunohistochemical detection of East Asian CagA in paraffin‐embedded human tissues infected with H. pylori.  Gastric biopsy specimens embedded in paraffin were analyzed by immunohistochemistry with α‐EAS Ab. As shown in Fig. 4a, strongly positive signals were observed at the surface of the gastric mucosal epithelium infected with East Asian cagA‐positive H. pylori, whereas in patients infected with Western cagA‐positive H. pylori, no positive signals were detected (Fig. 4c). In these cases, the presence of H. pylori was confirmed by immunohistochemistry of serial tissue sections with α‐H. pylori Ab, which is immunoreactive with both East Asian cagA‐positive H. pylori and Western cagA‐positive H. pylori (data not shown). Furthermore, in patients without H. pylori infection, no positive signals were detected (Fig. 4d). These findings suggest that α‐EAS Ab is available for detection of East Asian CagA in paraffin‐embedded tissue sections. Immunohistochemically, strong signals, apparently corresponding to H. pylori itself, were observed mainly at the surface of the gastric mucosa (Fig. 4a,b). Furthermore, we noticed that nuclei as well as cytoplasm of mucin‐producing cells infected with H. pylori was weakly and diffusely immunostained (see arrowheads in Fig. 4b).

Figure 4.

Immunohistochemistry of gastric mucosa biopsy specimens with anti‐East Asian CagA‐specific antibody. Sections from paraffin‐embedded tissue of East Asian cytotoxin‐associated antigen A‐positive gastritis (a and b), Western CagA‐positive gastritis (c) or gastritis without H. pylori infection were subjected to immunohistochemistry with α‐EAS Ab. Surfaces of the gastric mucosa infected with East Asian CagA‐positive H. pylori were positively immunostained with α‐EAS Ab (a). Detailed observations at high magnification revealed that strongly positive signals corresponded to H. pylori on the epithelial surfaces. The cytoplasm and nucleus of the epithelial cells were also positively immunostained with α‐EAS Ab (arrowheads in b). Gastric mucosa infected with Western CagA‐positive H. pylori was not immunostained with α‐EAS Ab (c). Cases without H. pylori infection were not stained with α‐EAS Ab (d). Original magnification, ×100 (a, c, d), ×200 (b).

Immunohistochemical diagnosis of the cagA genotype.  Subsequently, we determined the genotypes of the cagA gene in 47 patients (Table 1). Genomic DNA was extracted from cultured H. pylori that had been isolated from patients and stocked. Then, PCR was performed to amplify the DNA corresponding to the EPIYA motif of the cagA gene, and then sequenced. All the sequence data have been submitted to the DDBJ/EMBL/GenBank databases under accession number AB267217‐AB267264 (Table 1). As shown in Table 1, the cagA gene was identified in all of the 47 cases examined. In 44 of the 47 cases, the East Asian cagA gene was identified, whereas the Western cagA gene was identified in the other three. In one patient infected with East Asian cagA‐positive H. pylori, two independent EPIYA motifs that corresponded to the A‐B‐D and A‐B‐D‐D type, respectively, were detected by sequencing analysis (asterisk in Table 1), suggesting that H. pylori possessing each cagA genotype may co‐infect a single patient.

Table 1.

Clinical features, EPIYA motif typing and immunohistochemistry of the cases

Age	Sex	Disease	Typing of EPIYA motif	Typing of CagA	Immunoreactivity by α−EAS antibody	DDBJ accesion number
79	M	Adenoma	A‐B‐D	East Asian	(+)	AB267217
41	F	AG	A‐B‐D	East Asian	(+)	AB267218
52	F	AG	A‐B‐D	East Asian	(+)	AB267219
53	F	AG	A‐B‐D	East Asian	(+)	AB267220
54	M	AG	A‐B‐D	East Asian	(+)	AB267221
57	F	AG	A‐B‐D	East Asian	(+)	AB267222
60	F	AG	A‐B‐D	East Asian	(−)	AB267223
70	M	AG	A‐B‐D	East Asian	(+)	AB267224
77	F	AG	A‐B‐C	Western	(−)	AB267225
81	M	AG	A‐B‐D	East Asian	(+)	AB267226
82	F	AG	A‐B‐D	East Asian	(+)	AB267227
26	M	DU	A‐B‐D	East Asian	(+)	AB267228
27	F	DU	A‐B‐D	East Asian	(+)	AB267229
41	F	DU	A‐B‐D	East Asian	(+)	AB267230
41	F	DU	A‐B‐D	East Asian	(+)	AB267231
42	F	DU	A‐B‐D	East Asian	(+)	AB267232
45	F	DU	A‐B‐D	East Asian	(+)	AB267233
46	F	DU	A‐B‐D	East Asian	(+)	AB267234
56	M	DU	A‐B‐D	East Asian	(+)	AB267235
57	M	DU	A‐B‐D	East Asian	(+)	AB267236
65	F	DU	A‐B‐D	East Asian	(+)	AB267237
66	F	DU	A‐B‐D	East Asian	(+)	AB267238
67	F	DU	A‐B‐D	East Asian	(+)	AB267239
52	M	EG	A‐B‐D, A‐B‐D‐D*	East Asian	(+)	AB267240, AB267241
55	F	GDU	A‐B‐D	East Asian	(+)	AB267242
74	M	GDU	A‐B‐D	East Asian	(+)	AB267243
44	M	GU	A‐B‐D	East Asian	(+)	AB267244
50	F	GU	A‐B‐D	East Asian	(+)	AB267245
54	F	GU	A‐B‐D	East Asian	(+)	AB267246
56	M	GU	A‐B‐D	East Asian	(+)	AB267247
57	M	GU	A‐C	Western	(−)	AB267248
60	M	GU	A‐B‐D	East Asian	(+)	AB267249
61	M	GU	A‐B‐D	East Asian	(+)	AB267250
64	M	GU	A‐B‐D	East Asian	(+)	AB267251
65	M	GU	A‐B‐D	East Asian	(+)	AB267252
67	M	GU	A‐B‐D	East Asian	(+)	AB267253
70	M	GU	A‐B‐D	East Asian	(+)	AB267254
70	M	GU	A‐B‐D	East Asian	(+)	AB267255
72	M	GU	A‐B‐D	East Asian	(+)	AB267256
73	M	GU	A‐B‐B‐B	Western	(−)	AB267257
76	M	GU	A‐B‐D	East Asian	(+)	AB267258
60	F	MALToma	A‐B‐D	East Asian	(+)	AB267259
22	F	MK	A‐B‐D	East Asian	(+)	AB267260
53	F	MK	A‐B‐D	East Asian	(+)	AB267261
56	F	MK	A‐B‐D	East Asian	(+)	AB267262
57	M	MK	A‐B‐D	East Asian	(+)	AB267263
65	M	MK	A‐B‐D	East Asian	(+)	AB267264

AG: atrophic gastritis, DU: duodenal ulcer, EG: erosive gastritis, GDU: gastroduodenal ulcer, GU: gastric ulcer; MALToma: mucosal associated lymphoid tissue lymphoma, MK: gastric cancer.

We immunohistochemically analyzed the 47 cases for their immunoreactivity with α‐EAS Ab, and strong immunoreactivity was detected in 43 of them (Table 1). In all 43 cases, the results of immunohistochemical analysis corresponded completely to the sequencing results, whereas in three of the four cases that were not immunostained with α‐EAS Ab, the Western cagA gene was detected. However, we found that the remaining case did not exhibit any immunoreactivity with α‐EAS Ab, even though sequencing analysis had detected the East Asian cagA gene in this patient. In all four cases that were not immunoreactive with α‐EAS Ab, the presence of H. pylori was confirmed by immunohistochemistry with α‐H. pylori Ab using serial tissue sections.

Ultrastructural localization of CagA protein.  We analyzed biopsy specimens from two patients with chronic gastritis by immunoelectron microscopy with α‐EAS Ab. In both cases, a large number of bacteria were observed at the surface of the gastric epithelium. As shown in Fig. 5a, many attached gold particles were found in the cytoplasm of these bacteria. In addition to the strong immunoreactivity, we also found that gold particles were localized to the cytoplasm and nucleus of H. pylori‐infected epithelial cells (Fig. 5b–d). This observation is consistent with the immunohistochemical results shown in Fig. 4, suggesting that East Asian CagA protein is present in the nucleus as well as the cytoplasm in H. pylori‐infected epithelial cells of patients with gastritis.

Figure 5.

Immunoelectron microscopy with anti‐East Asian CagA‐specific antibody of gastric mucosa and AGS cells infected with East Asian CagA‐positive H. pylori. Biopsy specimens from two patients with gastritis infected with East Asian CagA‐positive H. pylori were subjected to immunoelectron microscopy with α‐EAS Ab (a, b, c and d), α‐CagA Ab (g and h) or without the first antibody (i). (a) Gold particles were found attached to the cytoplasm of the H. pylori (Hp in a) at the surface of the gastric epithelium. (b) Gastric mucosa infected with East Asian CagA‐positive H. pylori. The small open square is shown at higher magnification in c and the large open square is shown at higher magnification in d. Muc, mucin of epithelial cell; Nuc, nucleus. (c) Gold particles were found attached at the epithelial microvilli facing the lumen and in the cytoplasm of epithelial cells. (d) In the nuclei of the infected epithelium, gold particles were also attached. (e) Immunoelectron microscopy of AGS cells infected with East Asian CagA‐positive H. pylori in vitro with α‐EAS Ab. At the surface of AGS cells, gold particles were found to be localized not only to the cytoplasm of H. pylori but also the cytoplasm of the long processes of AGS cells facing H. pylori. (f) Gold particles were also found localized to the nucleus as well as the cytoplasm of the infected AGS cells. (g) Gastric mucosa infected with East Asian cagA‐positive H. pylori. The small open square is shown at higher magnification in (h). Gold particles were found attached to the cytoplasm of the H. pylori (Hp in g) at the surface of the gastric epithelium and nuclei of the epithelium (Nuc. in g and h) with α‐CagA Ab. (i) Without the first antibody, no particles were detectable in gastric mucosa or H. pylori. (j) Immunoelectron microscopy of AGS cells infected with Western cagA‐positive H. pylori with α‐CagA Ab. Gold particles attached to the nucleus are more numerous than those attached to the cytoplasm. (k) Without the first antibody, no particles were detectable in gastric mucosa or H. pylori. Scale bars, 1 µm (a, c–h, k), 4 µm (b), 2 µm (i), 500 nm (j).

We performed in vitro studies to determine whether CagA protein is actually detectable in the cytoplasm of H. pylori‐infected cells. Five hours after infection of AGS cells with East Asian cagA‐positive H. pylori, long foot processes were formed at the cell surface (Fig. 5e), and these increased with time. Gold particles were localized not only in the cytoplasm of H. pylori located outside the AGS cells (Fig. 5e), but also in the cytoplasm and nucleus of the infected AGS cells (Fig. 5f). Especially strong immunoreactivity was found in the cytoplasm of the long processes of H. pylori‐infected epithelial cells facing the bacteria (Fig. 5e). In contrast, when AGS cells were infected with Western cagA‐positive H. pylori (ATCC43504), no particles were detected (data not shown).

To confirm the presence of the CagA protein in the nucleus and cytoplasm, another antibody, α‐CagA Ab, was used for immunoelectron microscopy. As shown in Fig. 5(g,h), many gold particles were found in the nucleus of gastric epithelial cells as well as in the cytoplasm of the bacteria. The cytoplasm of gastric epithelial cells also had attached gold particles, although the immunoreactivity was weaker than that of the nuclei of the epithelial cells and the cytoplasm of H. pylori. False positive immunoreactivity was unlikely because no particle was detectable in gastric epithelial cells or H. pylori when immunoelectron microscopy was performed without the first antibody (Fig. 5i).

Furthermore, when AGS cells that had been infected with Western cagA‐positive H. pylori were analyzed by immunoelectron microscopy with α‐CagA Ab, immunoreactivity was detectable in both the nucleus and cytoplasm (Fig. 5j). In contrast, no particles were detectable without the first antibody in H. pylori‐infected AGS cells (Fig. 5k).

Next, in order to further confirm that the CagA protein is really localized in the nucleus, we performed fractionation of AGS cells infected with H. pylori (ATCC43504) into the membrane fraction, the cytosolic fraction, and the nuclear fraction, as described in Materials and Methods. Subsequently, we performed immunoprecipitation with α‐CagA Ab, followed by Western blot analysis with α‐CagA Ab. As shown in Fig. 6, a 130 kDa band corresponding to Western CagA was detectable in the membrane fraction (Fig. 6; lane 2) and nuclear fraction (Fig. 6; lane 6) of H. pylori‐infected AGS cells, but no such band was detectable in uninfected AGS cells (Fig. 6; lanes 1 and 5). On the other hand, CagA was not detected in the cytosolic fraction (Fig. 6; lanes 3 and 4). Western blot analysis with anti‐GAPDH antibody revealed that GAPDH was present in the cytosolic fraction, whereas in the other fractions it was detected at only a trace level. Furthermore, Western blot analysis with anti‐acetyl‐Histone H3 antibody revealed that acetyl‐Histone H3 was detectable only in the nuclear fraction, but scarcely detectable in the other fractions, suggesting that the subcellular fractionation had been performed appropriately.

Figure 6.

Detection of CagA in subcellular fractions of H. pylori‐infected AGS cells. AGS cells infected with H. pylori or without the bacterium were fractionated into membrane, cytosolic and nuclear fractions. Each fraction was immunoprecipitated with α‐CagA Ab, followed by Western blotting with α‐CagA Ab. No bands were detected in AGS cells without H. pylori infection (lanes 1, 3, and 5). The band corresponding to CagA indicated by the arrow was detectable in the membrane (lane 2) and nuclear fraction (lane 6) of AGS cells infected with H. pylori. No bands were detected in the cytosolic fraction (lane 4). Western blot analysis of each fraction was performed with α‐GAPDH Ab and α‐acetyl‐Histone H3 Ab.

Discussion

In this study, we generated a polyclonal antibody named α‐EAS Ab that is specifically immunoreactive with East Asian CagA, and by using it for immunohistochemistry we were able to successfully determine the cagA genotype of H. pylori infecting gastric epithelial cells. In order to determine the genotype of H. pylori, it has been necessary to sequence the genomic DNA of the bacterium. ^{( 17} , ^{18 )} However, our immunohistochemical method using α‐EAS Ab appears to have a number of merits for the genotype diagnosis of H. pylori. First, our method can be performed using the immunohistochemical techniques routinely available in most hospitals and is undoubtedly easier than sequencing the genomic DNA of H. pylori. Second, our method is less invasive to patients in comparison with H. pylori sequencing analysis, because an additional biopsy specimen for culture of H. pylori is not necessary. Third, our present immunohistochemical data obtained using α‐EAS Ab were proven to be reliable because they were almost consistent with the results obtained by sequencing of the cagA gene. In the present study, in one patient in whom the East Asian cagA gene was detected by sequencing, the tissue specimen was not stained with α‐EAS Ab. One explanation for this discrepancy may be that the cagA gene is mutated upstream of the region that we sequenced, resulting in the production of a truncated protein that is undetectable by α‐EAS Ab. However, we found that the corresponding serial section was positively immunostained with α‐CagA Ab, which recognizes the amino‐terminal 300 amino acids (data not shown), suggesting that such a mutation may not be present at least within the amino‐terminal region, although we cannot exclude the possibility that mutations may be present in other regions of the cagA gene. The second possible explanation is that different strains of H. pylori, i.e. Western cagA‐positive H. pylori, were co‐infecting the gastric mucosa in this patient. The biopsy samples used for bacterial culture and those used for histopathologic diagnosis were taken from different sites of the gastric mucosa in any given patient. Therefore, it is possible that Western cagA‐positive H. pylori and East Asian cagA‐positive H. pylori would have been detected at the different sites of gastric mucosa if both types of H. pylori had been present.

Recently, Azuma et al. have analyzed Japanese patients with gastritis for the genotype of the cagA gene and found that, in all patients except for those living in Okinawa, located at the far south‐west of the Japan archipelago, the cagA genotype of H. pylori corresponded to the East Asian type.^{( 12 )} However, in our present study, we detected Western cagA‐positive H. pylori in 6% (3 out of 47) of the patients, suggesting that patients infected with Western cagA‐positive H. pylori may be present among those infected with East Asian cagA‐positive H. pylori. Thus, we can speculate that the positivity rate for Western cagA‐positive H. pylori may differ between local regions in Japan. Furthermore, in East Asian countries, such as Korea, China, Vietnam and Thailand, the cagA genotypes of H. pylori have not been fully analyzed. Therefore, we propose that our immunohistochemical method is applicable to diagnosis of the H. pylori cagA genotype in these countries as well as in Japan.

In this study, we found that CagA protein was detectable not only in the cytoplasm but also in the nucleus of H. pylori‐infected epithelial cells by immunoelectron microscopy. This observation was confirmed using two distinct anti‐CagA antibodies (α‐EAS Ab and α‐CagA Ab). Furthermore, this nuclear localization was observed not only in H. pylori‐infected AGS cells but also in gastric epithelial cells of patients with gastritis infected with H. pylori. Further Western blot analysis of subcellular fractions of AGS cells infected with H. pylori proved that CagA was detectable in the nuclear fraction. In this Western blot analysis, we were unable to detect CagA in the cytosolic fraction, although it was detectable in the cytoplasm as a weaker signal by immunoelectron microscopy, suggesting that the quantity of CagA localized in the cytoplasm is significantly lower than that in the membrane or nucleus. Thus, these findings suggest that at least a proportion of CagA protein may translocate into the nucleus after H. pylori infection. On the other hand, inconsistent with our present data, Backert et al. on the basis of laser‐scanning immunofluorescence microscopy, have recently reported that CagA was not detected in the nucleus, but only in the cytoplasm of AGS cells that had been infected with H. pylori for 2 h.^{( 19 )} In the present study, we incubated the AGS cells with H. pylori for 5 h before immunoelectron microscopy and laser‐scanning immunofluorescence microscopy. Therefore, we cannot rule out the possibility that the difference in the intracellular localization of CagA in H. pylori‐infected AGS cells between the previous and present studies may have been due to the difference in the duration of infection.

The pathologic function of CagA in the nucleus is still largely unknown. Recent reports have proposed that CagA may affect the signal transduction pathway of host cells.^{( 10 )} It has been reported previously that CagA is injected into gastric epithelial cells via a type IV secretion system,^{( 19} , ²⁰ , ^{21 )} and that the injected CagA is phosphorylated and binds to the Src‐homology 2 (SH2) domain‐containing protein tyrosine phosphatase (SHP2) in the cytoplasm, resulting in activation of the ERK signaling pathway, which may cause abnormal cell proliferation or cell movement.^{( 22} , ^{23 )} Therefore, we can speculate that CagA translocated into the nucleus may also affect the physiologic function of cellular proteins such as transcription factors. Therefore, to clarify the pathologic function of CagA in the nucleus, further studies will be necessary to identify proteins that interact with CagA in this location.