- Email: [email protected]
Elevated expression of FGF7 protein is common in human gastric diseases
BBRC Biochemical and Biophysical Research Communications 350 (2006) 825–833 www.elsevier.com/locate/ybbrc
Elevated expression of FGF7 protein is common in human gastric diseases q Ron Shaoul a
a,1
, Liat Eliahu b,1, Ifat Sher b,2, Yaheli Hamlet a, Ines Miselevich c, Orit Goldshmidt b, Dina Ron b,*
Department of Pediatrics, Bnai Zion Medical Center, Faculty of Medicine, Technion, Israel Institute of Technology, Haifa, Israel b Department of Biology, Technion, Israel Institute of Technology, Haifa, Israel c Department of Pathology, Bnai Zion Medical Center, Haifa, Israel Received 29 August 2006 Available online 2 October 2006
Abstract Growth alterations within the gastric mucosa during chronic gastric inflammation are key steps in gastric cancer development. FGF7, a specific mitogen for epithelial cells, is implicated in epithelial tissue repair and cancer. We investigated FGF7 expression in normal human stomach, and in 35 cases from various gastric pathologies including 23 gastritis and 8 adenocarcinoma cases. Modest FGF7 protein levels were detected in the normal mucosal gland epithelium and in stromal fibroblasts. FGF7 protein levels, however, were markedly increased in the mucosal epithelium of all gastric inflammation cases. A similar elevated expression was also observed in gastric adenocarcinoma. Upregulation of FGF7 protein was associated with a modest increase in FGF7 mRNA expression. Interestingly, high levels of FGF7 anti-sense (AS) RNA were observed in the gastric pathologies, at the same sites where FGF7 protein was upregulated. Altogether, these findings suggest a role for FGF7 in maintaining gastric mucosa integrity, and that FGF7 protein levels are regulated mainly by posttranscriptional mechanisms. The elevated FGF7 protein levels in gastric inflammation and gastric cancer, together with the known oncogenic potential of FGF7, implicate excessive FGF7 signaling in gastric tumorigenesis, and point to FGF7 as an attractive target for gastric cancer prevention and treatment. 2006 Elsevier Inc. All rights reserved. Keywords: Anti-sense RNA; Fibroblast growth factor; FGF7; Helicobacter pylori; Gastric diseases
Gastric cancer remains the second major cause of cancer-related mortality worldwide. Despite the major improvements in diagnosis and treatment, less than 20% of patients survive up to 5 years [1,2]. Gastric cancer develops along a multistage process beginning with gastritis which progresses to mucosal atrophy (atrophic gastritis) followed by intestinal metaplasia, dysplasia, and carcinoma [2–4]. Chronic inflammation, initiated by Helicobacter
q
Abbreviations: AS, anti-sense; FGF, fibroblast growth factor; FGFR, fibroblast growth factor receptor; HP, Helicobacter pylori. * Corresponding author. Fax: +972 4 8225 153. E-mail address: [email protected] (D. Ron). 1 These authors contributed equally to this work. 2 Present address: Department of Biomedical Sciences, University of Guelph, Guelph, Ontario, Canada. 0006-291X/$ - see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.08.198
pylori (HP) infection as well as environmental factors, is considered the driving force of gastric carcinogenesis. Bacterial virulence factors and environmental factors are recognized as modifiers of key signaling pathways within the gastric mucosa and as such leading to growth alterations [3,4]. Several receptor tyrosine kinase (RTK) ligands, such as members of the epidermal growth factor family and vascular endothelial cell growth factor, are thought to contribute to gastric tumorigenesis. Little is known, however, about the involvement of fibroblast growth factors (FGFs) in gastric diseases [5]. FGF7 is a mesenchymally derived mitogen that acts specifically on cells of epithelial origin, stimulating their migration, differentiation, and proliferation [6]. FGF7 and its high affinity receptor, FGFR2IIIb, are expressed throughout the gastrointestinal (GI) tract, indicating that
826
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
the gut can both produce and respond to FGF7 [6–8]. FGF7 is upregulated in tissues with inflammatory bowel diseases (IBD) and in a variety of GI tract tumors [6,9,10]. Numerous studies suggested that FGF7 plays a key role in repair and preservation of epithelial tissues in the GI tract (reviewed in [6]). In rat stomach, FGF7 transcripts were detected in submucosal and muscular layers but not within the mucosal epithelium [7,11,12]. In addition, FGF7 protein was detected in human gastric fibroblast cell lines and in extracts from human fetal stomach segments [13,14]. The expression pattern of FGF7 within the human stomach and its expression in human gastric pathologies have not been reported. To address this question, we utilized immunohistochemistry as well as in situ hybridization techniques. We found that FGF7 protein is expressed in normal adult human stomach, in both the mesenchyme and epithelium, and its expression is markedly elevated in gastric inflammation, as well as in gastric cancer. Materials and methods Materials. Recombinant FGF7, and FGF7 specific antibodies were produced as described [15]. Restriction enzymes were from NEB, Pharmacia, and Roche. Blocking reagent, Digoxigenin (DIG)-RNA Labeling Kit, 5-bromo-4-chloro-3-indolyl-phosphate, 4-toluidine salt, nitroblue tetrazolium chloride, maleic acid, T7 RNA polymerase, proteinase K, yeast tRNA, and anti-digoxigenin (Fab) conjugated to alkaline phosphatase were from Roche. M-MLV reverse transcriptase and random hexamer primers were from Gibco-BRL. Mouse IgG anti-insulin was from Santa Cruz Biotechnology. GVA-mounting solution, aminoethyl-carbazole solution, enzyme-conjugated HRP–streptavidin solution, antibody diluent reagent solution, serum blocking solution, and biotinylated secondary antibody were from Zymed. BSA from ICN, recombinant human epithelial growth factor (EGF) and recombinant human insulin were from R&D, and all other chemicals were from Sigma. Tissue samples and statistical analysis. Parffin-embedded surgical and biopsy specimens of human stomach were obtained from five healthy donors (normal stomach), four patients with gastric remodeling, 21 with different stages of gastritis, 2 with atrophic gastritis, and 8 with adenocarcinoma of the stomach. The degree of gastritis was according to Dixon et al. [16]. Remodeling was considered as foveolar hyperplasia, cytologically reactive epithelial cells, edema and congestion of lamina propria, without inflammatory component. All sections were stained routinely with Giemsa stain for H. pylori status. The study was approved by the Bnai Zion Medical Center and the Israeli Ministry of Health Ethical Committees. t-test was performed to compare means of two groups and Mann–Whitney test to compare the distributions between two groups when appropriate. Probe preparation and in situ hybridization. cDNA fragments corresponding to nucleotides 17–726 (probe A) or 537–1031 (probe B) of the human FGF7 [17], and nucleotides 104–459 of human perlecan mRNA were generated by the polymerase chain reaction (PCR) using FGF7 cDNA as template or RNA extracted from human keratinocytes for perlecan probe. Each cDNA fragment was cloned in both orientations relative to the T7 RNA polymerase promoter in a pBluescript II plasmid (Promega). The synthesis of DIG-labeled RNA probes, analysis of their specific activity, and in situ hybridization were performed according to manufacturer’s instructions (Roche: in situ Hybridization Application Manual). Paraffin-embedded sections (5 lm) were processed as described in [18]. Prehybridization (3 h, 45 C) and hybridization (16 h, 42 C, humid chamber) were carried out with the indicated DIG-labeled ribo-
probes. Sections were then washed (2· SSPE, 10 min, 24 C and 0.2· SSPE, 1 h, 60C) and incubated (16 h, 24 C) with anti-digoxigenin (Fab) alkaline phosphatase-conjugated antibodies (1:2000). Sections were visualized by a Nikon eclipse E400 microscope. Where indicated, sections were counterstained with Mayer’s hematoxylin. RT-PCR. RNA was extracted from human tissues and cell lines, as described [19]. cDNA was synthesized using 2 lg RNA with primers specific for FGF7 mRNA (p1:5 0 -TAGCTGATGCATATGTGTTGTA ATGG), or the AS-RNA (p2: 5 0 -GGTCAATGACCTAGGAGTAACA ATC). PCR was performed with primers from two different coding exons of FGF7. The primer pairs for the detection of FGF7 anti-sense transcript were p3 (5 0 -AGTGAGAAGACTCTTCTGTCGAACA) and p1, and for the detection of FGF7 mRNA p2 and p4 (5 0 -TGCTCTGGAGTCATG TCATTGCAAG). RT-PCR for GAPDH expression was used as an internal control for RNA levels. Immunohistochemistry. Preparation of rabbit polyclonal antibodies directed against human recombinant FGF7 was previously described [15]. Paraffin-embedded tissue sections (5 lm) were subjected to immunostaining using the streptavidin-peroxidase technique according to the manufacturer’s protocol (LAB-SA detection system, Zymed Labs, San Francisco, CA, USA). Endogenous peroxidase activity was blocked by incubation with 0.3% hydrogen peroxide in methanol (15 min, 24 C). Antigen retrieval was performed by microwave boiling in sodium citrate buffer (10 mM NaCitrate, pH 6.0, 10 min, 95 C), followed by incubation with blocking solution (1 h, 25 C). Incubation with first antibody was carried out for 16 h at 4 C. Anti-FGF7 antibody was diluted 1:250, and anti-insulin MAB was diluted according to manufacturer’s (Santa Cruz) recommendation. Color reaction was performed using streptavidin–peroxidase complex, with aminoethyl-carbazole solution as substrate. Mayer’s hematoxylin was applied as counterstaining where indicated. Staining intensity of the glands and foveole was graded from 0 to 5 by an experienced Pathologist (IM).
Results Detection of FGF protein in gastric diseases Expression of FGF7 protein was investigated by immunohistochemistry (IHC) employing rabbit polyclonal antibody directed against FGF7 [15]. Normal pancreas tissues that are known to express FGF7 [18] were utilized to determine FGF7 antibody staining specificity. Sections were incubated without or with the anti-FGF7 antibodies, and with the antibodies in the presence of recombinant human FGF7, recombinant human insulin or EGF. FGF7 immunoreactivity was observed in pancreatic acinar cells and to a lesser extent in ductal cells (Fig. 1A, and data not shown) in agreement with Ishiwata et al. [18]. No signal was observed in the presence of recombinant FGF7 or in the absence of the FGF7 antibodies (Fig. 1B, and data not shown). Positive FGF7 immunoreactivity was obtained in sections incubated with FGF7 antibodies and recombinant insulin (or EGF) whereas no insulin immunoreactivity was observed in sections incubated with anti-insulin antibodies and recombinant insulin (Fig. 1C–E, and data not shown). These results clearly indicate that the antibody specifically recognizes the FGF7 protein. The expression of FGF7 in normal stomach and gastric diseases was assessed using specimens from 5 normal cases, and 35 various gastric pathologies of which 22 were positive for HP infection. Representative results are shown in Fig. 1. In the normal stomach, weak FGF7 immunoreactivity
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
827
Fig. 1. Specificity of the FGF7 antibody (A–E). Human normal pancreas sections were incubated with FGF7 antibody in the absence (A) or presence of 10 lg/ml human recombinant FGF7 (B) or insulin (C). Pancreatic sections were incubated with anti-insulin antibody without (D) or with (E) recombinant insulin. IHC was carried out as described under the Materials and methods section. Arrowheads point to FGF7 immunoreactivity in acinar cells, and arrows to insulin immunoreactivity in islet cells surrounded by the negative acinar cells. Strong induction of FGF7 protein in gastric diseases (F–O). (F,G) normal stomach from two different cases. (H,I) Moderate gastritis from two different cases (J) Severe gastritis. (K) Gastric adenocarcinoma. (L,M) Mucosal glands and foveole of a third moderate gastritis case, respectively. (N,O) Mucosal glands and foveole of a second severe gastritis case, respectively. Note the strong immunoreactivity in the cytoplasm of epithelial cells of glands and foveole, and in infiltrating lymphocytes. lp, gl, fo and mm denote for lamina propria, glands, foveole, and muscularis mucosae, respectively. Insets represent magnification of foveole (H) or mucosal glands (J) from the same section. Magnification, 100· (A–C), 40· (D–K). Counterstaining with hematoxylin (A–K).
828
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
was detected in the fibroblasts and the infiltrating lymphocytes of the lamina propria, and a modest signal was observed in the mucosal gland epithelium, but not in foveolar epithelium. Little or no staining was observed in the muscularis mucosae (Fig. 1F–G). Marked induction of FGF7 protein levels was observed in the various gastric diseases. In all gastritis cases, intense FGF7 staining was observed throughout the lamina propria and in the mucosal epithelial cells in both the foveole and glands. Strong immunoreactivity was present in infiltrating lymphocytes, and little or no staining was observed in the muscularis mucosae (Fig. 1H–J and L–O). Similarly, strong and uniform FGF7 immunoreactivity was observed in gastric adenocarcinoma (Fig. 1K). A comparison of FGF7 protein levels in foveole and glands in the various gastric pathologies and in stomach from healthy donors is presented in Table 1. Nearly comparable levels of FGF7 protein were detected in the different histopathological conditions. Six to eightfold increase in FGF7 protein levels was observed in the mucosal glands in mild, moderate, and severe gastritis (total of 21 cases). Similar increase was detected in tissues undergoing remodeling without an overt gastritis, as well as atrophic gastritis (4 and 2 cases, respectively). Strong induction of FGF7 was observed in the foveolar epithelium as compared to little or no expression in normal stomach specimens. Similarly, high expression levels were observed in tumor cells of gastric adenocarcinoma (8/8 cases). These results suggest that upregulation of FGF7 protein is common in gastric pathologies. Detection of FGF7 transcripts in gastric diseases In situ hybridization, using probes complementary to the first exon of the FGF7 gene (Fig. 2A), was carried out in order to assess whether the high induction of FGF7 protein in gastric pathologies is accompanied with a similar induction in FGF7 mRNA. For optimization of the in situ conditions, we initially utilized tissue sections from gastritis
samples. Unexpectedly, a strong signal was observed with the sense probe in all the tested gastritis samples. The signal obtained using the anti-sense (AS) probe, under comparable conditions, was relatively weak, but could be intensified by increasing probe concentration and/or the incubation time during the colorimetric reaction (see representative results in Fig. 2B–D). Under these conditions, no staining was observed in control sample treated in the absence of either probes (data not shown). These results were confirmed with an additional probe complementary to a different region along the FGF7 mRNA (designated probe B, Fig. 2A, E and F). Moreover, in situ hybridization performed with AS and sense probes for a different gene expressed in the GI-tract (perlecan [20]) revealed a positive signal only with the AS-probe (Fig. 2G and H). These results revealed the existence of FGF7 AS-transcript at levels that exceed those of the FGF7 mRNA. To further assess the induction of FGF7 transcripts in gastric diseases, in situ hybridization was carried out with specimens from normal stomach (5 cases), moderate and severe gastritis (5 and 7 cases, respectively), and gastric adenocarcinoma (5 cases). Hybridization was carried out using two probe concentrations (0.2 and 0.5 lg/ml). At the lower AS-riboprobe concentration the results were similar to those described above. FGF7 mRNA was barely detectable while FGF7 AS-RNA was readily detected with the sense riboprobe (data not shown). FGF7 mRNA was detected with the higher probe concentration, and its levels were elevated in gastric diseases, but to a modest degree. FGF7 AS-RNA was not detected in normal stomach whereas it was strongly induced in all gastric pathologies (representative results are shown in Fig. 2). We did not observe a significant difference in staining intensity between the different gastric pathologies indicating that upregulation of FGF7 transcripts, similar to FGF7 protein, is an early event in gastric inflammation. Both RNA types were detected in epithelial cells in both mucosal glands and foveole, as well as in fibroblasts and infiltrating lymphocytes in
Table 1 Comparison of FGF7 protein expression levels in the different gastric histopathological states Diagnosis
Normal Remodeling Gastritis Mild Moderate Severe Atrophic gastritis Adenocarcinoma Poorly differentiated Well differentiated
Number n = 40
5 4 23 1 11 9 2 8 5 3
HP positive
0 1 17 0 10 7 0 4 4 0
FGF7 protein levels (0–5) Glands
Foveolesa
Other
0.5 3.5 3 3 3.1 3.3 4 3 2.5 4
ND 3.25 3.1 5 2.8 3 3 2.5 2.3 3
3 2 3.5
The biopsies from each histopathological state were microscopically examined to estimate FGF7 expression level in the foveole and gland or tumor cells, in a scale between 1 and 5, where 5 is the highest and 0 is the lowest expression level. The numbers representing the average score of expression level. Foveolar and glands stain was significantly higher in all gastric pathologies, including intact gastric tissue from adenocarcinoma surgical specimens, compared with the normal tissues. ND denotes, not detected. a Since FGF7 was not detected in the gastric pits of stomach from healthy donors (normal), the expression levels in foveole were estimated relative to the levels observed in mucosal glands from normal stomach.
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
A FGF7 mRNA
I/E 738
446
1
1030
3850
ORF
5' UTR 1
I/E 849
829
680 Probe A 552
1030 Probe B
Fig. 2. Induction of FGF7 transcripts in gastric inflammation. (A) Schematic representation of the FGF7 mRNA and localization of the probes used for the in situ hybridization. I/E denotes for exon/intron junction. UTR, untranslated region; ORF, open reading frame. (B–H) Detection of FGF7 and perlecan transcripts in gastric tissues. Gastritis specimens were used for the initial detection of FGF7 and perlecan transcripts. In situ hybridization was carried out with FGF7 probes A or B, or perlecan probe (0.2 lg/ml for each probe), in both the anti-sense and sense orientation. (B–D) Serial sections of moderate gastritis specimen hybridized with anti-sense and sense riboprobe A. (B,C) Hybridization with the anti-sense probe. The sections in (B,C) were incubated for 1.5 and 3 h during the colorimetric reaction, respectively. (D) Hybridization with sense probe (1.5 h incubation during the colorimetric reaction). (E,F) Serial sections of a different gastritis sample hybridized with anti-sense and sense probe B, respectively. (G,H) Hybridization with antisense and sense probes for perlecan, respectively. (I–N) The expression of FGF7 transcripts is elevated in gastric diseases. In situ hybridization was carried out with 0.5 lg/ml probe B, in both the anti-sense (I–K) or sense (L–N) orientation. (I,L) normal stomach. (J,M) Moderate gastritis. (K,N) Severe gastritis; similar results were obtained with riboprobe A. Magnification, 25· (B–F), 40· (G–N).
the various gastric pathologies. In addition, FGF7 ASRNA and protein/mRNA were co-localized (compare panels J and K to panels M and N in Fig. 2, and see Fig. 3). These results show that FGF7 transcripts are upregulated in gastric diseases, and that induction of FGF7 AS-RNA exceeds the induction of its mRNA.
FGF7 AS-RNA is expressed in other human tissues and cultured cells RT-PCR analysis was carried out to determine whether FGF7 AS-RNA is expressed in other human cell types and tissues. The cDNA was generated using either primer
830
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
Fig. 3. FGF7 anti-sense transcript and protein are co-localized. (A–C) In situ hybridization, with probe A in the sense orientation, for the detection of FGF7 anti-sense RNA. (D–F) Detection of FGF7 protein by IHC. (A,D) Serial sections from a moderate gastritis case with intestinal metaplasia. (B,E) Serial sections from a severe gastritis case demonstrating FGF7 protein and AS-RNA expression in a foveole. (C,F) Mucosal glands area from a moderate gastritis case. Counterstaining with hematoxylin (D–F).
specific to FGF7 mRNA or to the FGF7-AS transcript. Among the normal tissues that were examined FGF7 ASRNA was not detected in normal fetal or adult human brain, thyroid, liver, kidney or adenoids, and a weak signal was observed in tonsils (Fig. 4, and data not shown). ASFGF7 transcript was highly expressed in primary human dermal fibroblasts and lung fibroblasts, primary dermal fibroblasts from psoriatic lesions, fibroblast cell line derived from human embryos carrying dyssegmental dysplasia, Silverman–Handmaker type (DDSH) syndrome [21,22], and in polyp and mucosa specimens from a familial polyposis case. Weak expression of the anti-sense RNA was observed in prostate carcinoma. Samples from various areas of colonic adenocarcinoma were negative. The expression patterns of FGF7 AS-transcript and mRNA were similar, except for cecum carcinoma where FGF7 mRNA, but not AS-RNA was detected.
Fig. 4. Expression of FGF AS-RNA in human tissues and cell lines. RNA extraction and specific amplification of FGF7 AS-RNA or mRNA, by RT-PCR, were described under the Materials and methods section. P, fibro, and ca denotes for primary, fibroblasts, and carcinoma, respectively. DDSH fibro is an established fibroblast cell line from human embryos carrying dyssegmental dysplasia, Silverman–Handmaker type (DDSH) syndrome [21,22].
Discussion Chronic inflammation in various sites, including the GI tract, is associated with an increased risk of human cancer [3,4]. FGF7 is a specific mitogen for epithelial cells, from which the majority of cancer types arise [6]. Elevated FGF7 levels stimulate epithelial cell hyperplasia along the GI tract in experimental animals and are associated with inflammation and cancer of the intestine and the upper GI tract [6–10,23]. In the present study, we report, for the first time, that FGF7 protein is expressed in the mucosal epithelial cells in the normal adult human stomach, and that marked induction of FGF7 protein and its anti-sense transcript is a common theme in human gastric diseases. FGF7 is generally considered a mesenchyme derived growth factor that acts on adjacent epithelial cells [24,25]. In normal human stomach, FGF7 protein was expressed at low levels in the lamina propria and mucosal gland epithelial cells. As the FGF7 receptor is expressed in stomach epithelial cells [24,25], our findings indicate that FGF7 can act both in a paracrine manner, and as part of an autocrine loop, in human stomach. Such an autocrine loop may compromise the tight regulation of the directional epithelial– mesenchymal signaling, promoting the growth of epithelial cells. In addition, our findings further support the previously suggested role of FGF7 in preservation of GI epithelial tissue [6,7,13]. Similar to the current findings, several studies indicated that FGF7 is expressed in certain normal epithelial cells, including endometrial epithelium of porcine uterus [26], pancreatic acinar and ductal cells [18], and primary lens epithelial cells [27]. FGF7 protein levels were markedly elevated in the gastric mucosa in each of the gastric pathologies we examined including mild–severe gastritis, atrophic gastritis, and tissues undergoing remodeling without an overt gastritis (total of 27 cases). Strong induction of FGF7 protein levels
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
was also observed in the foveolar epithelium that was negative for FGF7 expression in the normal human stomach. Similar increase was observed in 8/8 gastric adenocarcinoma cases. These results indicate that FGF7 protein upregulation is common in gastric diseases, probably occurring early in gastric inflammation. Enhanced FGF7 protein and RNA levels were similar in HP-positive (22/35) as well as HP-negative (13/35) cases. These findings suggest that factors common to HP negative/HP positive gastric inflammations may be responsible for FGF7 upregulation. Likely candidates are proinflammatory cytokines, such as TNFa, IL1b, and IL6, that are known to be produced in the inflamed stomach and have been shown to upregulate FGF7 expression [4,28–30]. In addition, these proinflammatory cytokines are thought to contribute to carcinogenesis by influencing the survival, growth, mutation, differentiation, and migration of tumor cells and by regulating angiogenesis [31,32]. On the basis of its induction at early stages of gastric inflammation, and its continued expression during the different pre and malignant stages of gastric cancer, it is tempting to suggest that FGF7 may be central in mediating many effects of proinflammatory cytokines in gastric inflammation and gastric carcinogenesis. Numerous studies support a role for FGF7 in gastric cancer development. Expression of FGF7 receptor was implicated in gastric cancer [33–36], and studies in cultured cells indicated that the ability of overexpressed FGF7-receptor to induce malignant transformation is ligand dependent [37]. Thus, elevated levels of FGF7 should accelerate the proliferation of premalignant and malignant gastric epithelial cells. This is consistent with reports that FGF7 stimulates the proliferation of gastric cancer cells in vitro and its ability to induce epithelial cell hyperplasia along the GI tract when administered to rats [7,14]. Besides its specific effects on proliferation, survival, and migration of epithelial cells, FGF7 also induces the expression of angiogenic growth factors (reviewed in [6]) which are essential for solid tumor growth and metastasis [38]. Our results suggest that FGF7 mRNA levels in the normal human stomach are quite low and are only modestly upregulated in stomach diseases as compared to the marked induction of FGF7 protein or FGF7 AS-RNA. These observations imply that FGF7 protein levels are regulated mainly by posttranscriptional mechanisms. In agreement with our findings, Kinoshita et al. reported that FGF7 mRNA levels are rather low in rat stomach [12]. In rat stomach, however, FGF7 mRNA was not detected in the epithelium. These results suggest that the expression pattern of FGF7 in the human stomach differs from that of rat stomach. Another plausible explanation is that similar to human stomach, rat FGF7 mRNA levels in the rat gastric mucosal epithelium are extremely low, and therefore, were below detection under Kinoshita et al. experimental conditions. The observed low levels of FGF7 mRNA may also explain why it has not been detected in gastric cancer cell lines [8,14]. Alternatively, FGF7 expression in
831
gastric cancer cell lines may be switched off under in vitro growth conditions. We discovered the existence of endogenous FGF7 ASRNA in human gastric diseases, as well as other tissues, and cultured cells mainly from pathological conditions. To the best of our knowledge this is the first demonstration for the existence of FGF7 AS-RNA. AS-RNA was reported for other FGFs as well [39,40]. Since FGFs represent a highly conserved gene family, it is possible that a parallel FGF-AS family exists, whose biological function remains to be elucidated. The FGF7 AS-RNA was complementary to at least the 5 0 UTR and the open reading frame (ORF) of FGF7 mRNA, but no coding capacity was detected within this region of the AS-RNA. In the gastric diseases, FGF7 AS-RNA was co-localized with FGF7 mRNA. Moreover, in other tissues and cultured cells positive for FGF7 AS-RNA, we always detected FGF7 mRNA. These results suggest that transcription of FGF7 mRNA and ASRNA may be co-regulated. Endogenous AS-RNAs are thought to play diverse roles, including regulation of gene expression at the level of transcription, mRNA processing, mRNA stability or translation. Anti-sense regulation generally refers to regulation of expression of an RNA target through direct base pairing with a complementary RNA and is usually negative (reviewed in [41]) . In some cases, however, anti-sense RNA positively modulated gene expression and protein translation [42,43]. The observed co-localization of high FGF7 protein and AS-RNA levels in gastric diseases (see Fig. 3), together with the finding that FGF7 mRNA levels are rather low in these diseases, supports an attractive role for FGF7 AS-RNA in positively regulating FGF7 translation efficiency. Further studies are required to address this possibility. In summary, this study suggests a role for FGF7 in the preservation of the gastric mucosa acting both in a paracrine and an autocrine manner. The marked induction of FGF7 protein in gastritis and gastric adenocarcinoma, together with the known oncogenic potential of FGF7, suggest that excessive FGF7 signaling may contribute to gastric cancer development and progression. Gastric cancer represents a major health problem, and despite the worldwide decline in incidence, and the vast improvements in diagnosis and treatment, most patients survive only few years [2,44]. Targeting FGF7 signaling may be an attractive mode for prevention and treatment of gastric cancer. Acknowledgments We thank Natasha Kipnis for her kind help in processing biopsy sections and Sharon Lubinsky-Mink for technical assistance. This work was supported by a grant from the Technion and Bene Zion Hospital research foundation to Dina Ron and Ron Shaoul (#140–614) and partially supported by grants from Israel Cancer Research Fund (ICRF #2004973) Israel Science foundation (ISF #2006261) and Israel Ministry of Health (# 907134; 1-6305) to Dina Ron.
832
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
References [1] T.J. Koh, T.C. Wang, Tumors of the stomach, in: M. Feldman, W.O. Tschumy, L.S. Friedman, M.H. Sleisenger (Eds.), Sleisenger & Fordtran’s Gastrointestinal and Liver Disease, 2002, pp. 829–856. [2] M.G. Smith, G.L. Hold, E. Tahara, E.M. El-Omar, Cellular and molecular aspects of gastric cancer, World J. Gastroenterol. 12 (2006) 2979–2990. [3] C. Stoicov, R. Saffari, X. Cai, C. Hasyagar, J. Houghton, Molecular biology of gastric cancer: Helicobacter infection and gastric adenocarcinoma: bacterial and host factors responsible for altered growth signaling, Gene 341 (2004) 1–17. [4] H. Ohshima, H. Tazawa, B.S. Sylla, T. Sawa, Prevention of human cancer by modulation of chronic inflammatory processes, Mutat. Res. 591 (2005) 110–122. [5] J.C. Becker, C. Muller-Tidow, H. Serve, W. Domschke, T. Pohle, Role of receptor tyrosine kinases in gastric cancer: new targets for a selective therapy, World J. Gastroenterol. 12 (2006) 3297–3305. [6] P.W. Finch, J.S. Rubin, Keratinocyte growth factor/fibroblast growth factor 7, a homeostatic factor with therapeutic potential for epithelial protection and repair, Adv. Cancer Res. 91 (2004) 69–136. [7] R.M. Housley, C.F. Morris, W. Boyle, B. Ring, R. Biltz, J.E. Tarpley, S.L. Aukerman, P.L. Devine, R.H. Whitehead, G.F. Pierce, Keratinocyte growth factor induces proliferation of hepatocytes and epithelial cells throughout the rat gastrointestinal tract, J. Clin. Invest. 94 (1994) 1764–1777. [8] S. Iida, O. Katoh, A. Tokunaga, M. Terada, Expression of fibroblast growth factor gene family and its receptor gene family in the human upper gastrointestinal tract, Biochem. Biophys. Res. Commun. 199 (1994) 1113–1119. [9] P.W. Finch, V. Pricolo, A. Wu, S.D. Finkelstein, Increased expression of keratinocyte growth factor messenger RNA associated with inflammatory bowel disease, Gastroenterology 110 (1996) 441–451. [10] M. Brauchle, M. Madlener, A.D. Wagner, K. Angermeyer, U. Lauer, P.H. Hofschneider, M. Gregor, S. Werner, Keratinocyte growth factor is highly overexpressed in inflammatory bowel disease, Am. J. Pathol. 149 (1996) 521–529. [11] Y. Matsubara, M. Ichinose, M. Tatematsu, M. Ichinose, M. Oka, N. Yahagi, K. Kurokawa, T. Kageyama, K. Miki, H. Fukamachi, Stagespecific elevated expression of the genes for hepatocyte growth factor, keratinocyte growth factor, and their receptors during the morphogenesis and differentiation of rat stomach mucosa, Biochem. Biophys. Res. Commun. 222 (1996) 669–677. [12] Y. Kinoshita, H. Nakata, S. Hassan, M. Asahara, C. Kawanami, Y. Matsushima, Y. Naribayashi-Inomoto, C.Y. Ping, D. Min, A. Nakamura, Gene expression of keratinocyte and hepatocyte growth factors during the healing of rat gastric mucosal lesions, Gastroenterology 109 (1995) 1068–1077. [13] P. Chailler, J.R. Basque, L. Corriveau, D. Menard, Functional characterization of the keratinocyte growth factor system in human fetal gastrointestinal tract, Pediatr. Res. 48 (2000) 504–510. [14] K. Nakazawa, M. Yashiro, K. Hirakawa, Keratinocyte growth factor produced by gastric fibroblasts specifically stimulates proliferation of cancer cells from scirrhous gastric carcinoma, Cancer Res. 63 (2003) 8848–8852. [15] D. Ron, D.P. Bottaro, P.W. Finch, D. Morris, J.S. Rubin, S.A. Aaronson, Expression of biologically active recombinant keratinocyte growth factor. Structure/function analysis of aminoterminal truncation mutants, J. Biol. Chem. 268 (1993) 2984–2988. [16] M.F. Dixon, R.M. Genta, J.H. Yardley, P. Correa, Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994, Am. J. Surg. Pathol. 20 (1996) 1161–1181. [17] P.W. Finch, J.S. Rubin, T. Miki, D. Ron, S.A. Aaronson, Human KGF is FGF-related with properties of a paracrine effector of epithelial cell growth, Science 245 (1989) 752–755.
[18] T. Ishiwata, H. Friess, M.W. Buchler, M.E. Lopez, M. Korc, Characterization of keratinocyte growth factor and receptor expression in human pancreatic cancer, Am. J. Pathol. 153 (1998) 213–222. [19] A. Eisemann, J.A. Ahn, G. Graziani, S.R. Tronick, D. Ron, Alternative splicing generates at least five different isoforms of the human basic-FGF receptor, Oncogene 6 (1991) 1195–1202. [20] A.D. Murdoch, B. Liu, R. Schwarting, R.S. Tuan, R.V. Iozzo, Widespread expression of perlecan proteoglycan in basement membranes and extracellular matrices of human tissues as detected by a novel monoclonal antibody against domain III and by in situ hybridization, J. Histochem. Cytochem. 42 (1994) 239–249. [21] E. Arikawa-Hirasawa, W.R. Wilcox, A.H. Le, N. Silverman, P. Govindraj, J.R. Hassell, Y. Yamada, Dyssegmental dysplasia, Silverman–Handmaker type, is caused by functional null mutations of the perlecan gene, Nat. Genet. 27 (2001) 431–434. [22] E. Arikawa-Hirasawa, H. Watanabe, H. Takami, J.R. Hassell, Y. Yamada, Perlecan is essential for cartilage and cephalic development, Nat. Genet. 23 (1999) 354–358. [23] M. Watanabe, T. Ishiwata, K. Nishigai, Y. Moriyama, G. Asano, Overexpression of keratinocyte growth factor in cancer cells and enterochromaffin cells in human colorectal cancer, Pathol. Int. 50 (2000) 363–372. [24] I.J. Mason, F. Fuller-Pace, R. Smith, C. Dickson, FGF-7 (keratinocyte growth factor) expression during mouse development suggests roles in myogenesis, forebrain regionalisation and epithelial–mesenchymal interactions, Mech. Dev. 45 (1994) 15–30. [25] P.W. Finch, G.R. Cunha, J.S. Rubin, J. Wong, D. Ron, Pattern of keratinocyte growth factor and keratinocyte growth factor receptor expression during mouse fetal development suggests a role in mediating morphogenetic mesenchymal-epithelial interactions, Dev. Dyn. 203 (1995) 223–240. [26] H. Ka, T.E. Spencer, G.A. Johnson, F.W. Bazer, Keratinocyte growth factor: expression by endometrial epithelia of the porcine uterus, Biol. Reprod. 62 (2000) 1772–1778. [27] J. Weng, Q. Liang, R.R. Mohan, Q. Li, S.E. Wilson, Hepatocyte growth factor, keratinocyte growth factor, and other growth factorreceptor systems in the lens, Invest. Ophthalmol. Vis. Sci. 38 (1997) 1543–1554. [28] M. Brauchle, K. Angermeyer, G. Hubner, S. Werner, Large induction of keratinocyte growth factor expression by serum growth factors and pro-inflammatory cytokines in cultured fibroblasts, Oncogene 9 (1994) 3199–3204. [29] M. Chedid, J.S. Rubin, K.G. Csaky, S.A. Aaronson, Regulation of keratinocyte growth factor gene expression by interleukin 1, J. Biol. Chem. 269 (1994) 10753–10757. [30] P.W. Finch, C. Lengel, M. Chedid, Cloning and characterization of the promoter region of the human keratinocyte growth factor gene, J. Biol. Chem. 270 (1995) 11230–11237. [31] F. Balkwill, A. Mantovani, Inflammation and cancer: back to Virchow? Lancet 357 (2001) 539–545. [32] L.M. Coussens, Z. Werb, Inflammation and cancer, Nature 420 (2002) 860–867. [33] Y. Hattori, H. Itoh, S. Uchino, K. Hosokawa, A. Ochiai, Y. Ino, H. Ishii, H. Sakamoto, N. Yamaguchi, K. Yanagihara, S. Hirohashi, T. Sugimura, M. Terada, Immunohistochemical detection of K-sam protein in stomach cancer, Clin. Cancer Res. 2 (1996) 1373–1381. [34] H. Ishii, T. Yoshida, H. Oh, S. Yoshida, M. Terada, A truncated K-sam product lacking the distal carboxyl-terminal portion provides a reduced level of autophosphorylation and greater resistance against induction of differentiation, Mol. Cell Biol. 15 (1995) 3664–3671. [35] H. Itoh, Y. Hattori, H. Sakamoto, H. Ishii, T. Kishi, H. Sasaki, T. Yoshida, M. Koono, T. Sugimura, M. Terada, Preferential alternative splicing in cancer generates a K-sam messenger RNA with higher transforming activity, Cancer Res. 54 (1994) 3237–3241. [36] T. Matsunobu, T. Ishiwata, M. Yoshino, M. Watanabe, M. Kudo, K. Matsumoto, A. Tokunaga, T. Tajiri, Z. Naito, Expression of
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
[37]
[38] [39]
[40]
keratinocyte growth factor receptor correlates with expansive growth and early stage of gastric cancer, Int. J. Oncol. 28 (2006) 307–314. E. Shaoul, R. Reich-Slotky, B. Berman, D. Ron, Fibroblast growth factor receptors display both common and distinct signaling pathways, Oncogene 10 (1995) 1553–1561. R.S. Kerbel, Tumor angiogenesis: past, present and the near future, Carcinogenesis 21 (2000) 505–515. R.S. Knee, S.E. Pitcher, P.R. Murphy, Basic fibroblast growth factor sense (FGF) and antisense (gfg) RNA transcripts are expressed in unfertilized human oocytes and in differentiated adult tissues, Biochem. Biophys. Res. Commun. 205 (1994) 577–583. F. Coulier, V. Ollendorff, I. Marics, O. Rosnet, M. Batoz, J. Planche, S. Marchetto, M.J. Pebusque, O. deLapeyriere, D. Birnbaum, The
[41]
[42]
[43]
[44]
833
FGF6 gene within the FGF multigene family, Ann. NY Acad. Sci. 638 (1991) 53–61. S.H. Munroe, Diversity of antisense regulation in eukaryotes: multiple mechanisms, emerging patterns, J. Cell Biochem. 93 (2004) 664–671. R.A. Lease, M. Belfort, A trans-acting RNA as a control switch in Escherichia coli: DsrA modulates function by forming alternative structures, Proc. Natl. Acad. Sci. USA 97 (2000) 9919–9924. E.S. Tasheva, D.J. Roufa, Regulation of human RPS14 transcription by intronic antisense RNAs and ribosomal protein S14 Genes, Development 9 (1995) 304–316. Y. Kodera, M. Fujiwara, M. Koike, A. Nakao, Chemotherapy as a component of multimodal therapy for gastric carcinoma, World J. Gastroenterol. 12 (2006) 2000–2005.
Elevated expression of FGF7 protein is common in human gastric diseases q Ron Shaoul a
a,1
, Liat Eliahu b,1, Ifat Sher b,2, Yaheli Hamlet a, Ines Miselevich c, Orit Goldshmidt b, Dina Ron b,*
Department of Pediatrics, Bnai Zion Medical Center, Faculty of Medicine, Technion, Israel Institute of Technology, Haifa, Israel b Department of Biology, Technion, Israel Institute of Technology, Haifa, Israel c Department of Pathology, Bnai Zion Medical Center, Haifa, Israel Received 29 August 2006 Available online 2 October 2006
Abstract Growth alterations within the gastric mucosa during chronic gastric inflammation are key steps in gastric cancer development. FGF7, a specific mitogen for epithelial cells, is implicated in epithelial tissue repair and cancer. We investigated FGF7 expression in normal human stomach, and in 35 cases from various gastric pathologies including 23 gastritis and 8 adenocarcinoma cases. Modest FGF7 protein levels were detected in the normal mucosal gland epithelium and in stromal fibroblasts. FGF7 protein levels, however, were markedly increased in the mucosal epithelium of all gastric inflammation cases. A similar elevated expression was also observed in gastric adenocarcinoma. Upregulation of FGF7 protein was associated with a modest increase in FGF7 mRNA expression. Interestingly, high levels of FGF7 anti-sense (AS) RNA were observed in the gastric pathologies, at the same sites where FGF7 protein was upregulated. Altogether, these findings suggest a role for FGF7 in maintaining gastric mucosa integrity, and that FGF7 protein levels are regulated mainly by posttranscriptional mechanisms. The elevated FGF7 protein levels in gastric inflammation and gastric cancer, together with the known oncogenic potential of FGF7, implicate excessive FGF7 signaling in gastric tumorigenesis, and point to FGF7 as an attractive target for gastric cancer prevention and treatment. 2006 Elsevier Inc. All rights reserved. Keywords: Anti-sense RNA; Fibroblast growth factor; FGF7; Helicobacter pylori; Gastric diseases
Gastric cancer remains the second major cause of cancer-related mortality worldwide. Despite the major improvements in diagnosis and treatment, less than 20% of patients survive up to 5 years [1,2]. Gastric cancer develops along a multistage process beginning with gastritis which progresses to mucosal atrophy (atrophic gastritis) followed by intestinal metaplasia, dysplasia, and carcinoma [2–4]. Chronic inflammation, initiated by Helicobacter
q
Abbreviations: AS, anti-sense; FGF, fibroblast growth factor; FGFR, fibroblast growth factor receptor; HP, Helicobacter pylori. * Corresponding author. Fax: +972 4 8225 153. E-mail address: [email protected] (D. Ron). 1 These authors contributed equally to this work. 2 Present address: Department of Biomedical Sciences, University of Guelph, Guelph, Ontario, Canada. 0006-291X/$ - see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.08.198
pylori (HP) infection as well as environmental factors, is considered the driving force of gastric carcinogenesis. Bacterial virulence factors and environmental factors are recognized as modifiers of key signaling pathways within the gastric mucosa and as such leading to growth alterations [3,4]. Several receptor tyrosine kinase (RTK) ligands, such as members of the epidermal growth factor family and vascular endothelial cell growth factor, are thought to contribute to gastric tumorigenesis. Little is known, however, about the involvement of fibroblast growth factors (FGFs) in gastric diseases [5]. FGF7 is a mesenchymally derived mitogen that acts specifically on cells of epithelial origin, stimulating their migration, differentiation, and proliferation [6]. FGF7 and its high affinity receptor, FGFR2IIIb, are expressed throughout the gastrointestinal (GI) tract, indicating that
826
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
the gut can both produce and respond to FGF7 [6–8]. FGF7 is upregulated in tissues with inflammatory bowel diseases (IBD) and in a variety of GI tract tumors [6,9,10]. Numerous studies suggested that FGF7 plays a key role in repair and preservation of epithelial tissues in the GI tract (reviewed in [6]). In rat stomach, FGF7 transcripts were detected in submucosal and muscular layers but not within the mucosal epithelium [7,11,12]. In addition, FGF7 protein was detected in human gastric fibroblast cell lines and in extracts from human fetal stomach segments [13,14]. The expression pattern of FGF7 within the human stomach and its expression in human gastric pathologies have not been reported. To address this question, we utilized immunohistochemistry as well as in situ hybridization techniques. We found that FGF7 protein is expressed in normal adult human stomach, in both the mesenchyme and epithelium, and its expression is markedly elevated in gastric inflammation, as well as in gastric cancer. Materials and methods Materials. Recombinant FGF7, and FGF7 specific antibodies were produced as described [15]. Restriction enzymes were from NEB, Pharmacia, and Roche. Blocking reagent, Digoxigenin (DIG)-RNA Labeling Kit, 5-bromo-4-chloro-3-indolyl-phosphate, 4-toluidine salt, nitroblue tetrazolium chloride, maleic acid, T7 RNA polymerase, proteinase K, yeast tRNA, and anti-digoxigenin (Fab) conjugated to alkaline phosphatase were from Roche. M-MLV reverse transcriptase and random hexamer primers were from Gibco-BRL. Mouse IgG anti-insulin was from Santa Cruz Biotechnology. GVA-mounting solution, aminoethyl-carbazole solution, enzyme-conjugated HRP–streptavidin solution, antibody diluent reagent solution, serum blocking solution, and biotinylated secondary antibody were from Zymed. BSA from ICN, recombinant human epithelial growth factor (EGF) and recombinant human insulin were from R&D, and all other chemicals were from Sigma. Tissue samples and statistical analysis. Parffin-embedded surgical and biopsy specimens of human stomach were obtained from five healthy donors (normal stomach), four patients with gastric remodeling, 21 with different stages of gastritis, 2 with atrophic gastritis, and 8 with adenocarcinoma of the stomach. The degree of gastritis was according to Dixon et al. [16]. Remodeling was considered as foveolar hyperplasia, cytologically reactive epithelial cells, edema and congestion of lamina propria, without inflammatory component. All sections were stained routinely with Giemsa stain for H. pylori status. The study was approved by the Bnai Zion Medical Center and the Israeli Ministry of Health Ethical Committees. t-test was performed to compare means of two groups and Mann–Whitney test to compare the distributions between two groups when appropriate. Probe preparation and in situ hybridization. cDNA fragments corresponding to nucleotides 17–726 (probe A) or 537–1031 (probe B) of the human FGF7 [17], and nucleotides 104–459 of human perlecan mRNA were generated by the polymerase chain reaction (PCR) using FGF7 cDNA as template or RNA extracted from human keratinocytes for perlecan probe. Each cDNA fragment was cloned in both orientations relative to the T7 RNA polymerase promoter in a pBluescript II plasmid (Promega). The synthesis of DIG-labeled RNA probes, analysis of their specific activity, and in situ hybridization were performed according to manufacturer’s instructions (Roche: in situ Hybridization Application Manual). Paraffin-embedded sections (5 lm) were processed as described in [18]. Prehybridization (3 h, 45 C) and hybridization (16 h, 42 C, humid chamber) were carried out with the indicated DIG-labeled ribo-
probes. Sections were then washed (2· SSPE, 10 min, 24 C and 0.2· SSPE, 1 h, 60C) and incubated (16 h, 24 C) with anti-digoxigenin (Fab) alkaline phosphatase-conjugated antibodies (1:2000). Sections were visualized by a Nikon eclipse E400 microscope. Where indicated, sections were counterstained with Mayer’s hematoxylin. RT-PCR. RNA was extracted from human tissues and cell lines, as described [19]. cDNA was synthesized using 2 lg RNA with primers specific for FGF7 mRNA (p1:5 0 -TAGCTGATGCATATGTGTTGTA ATGG), or the AS-RNA (p2: 5 0 -GGTCAATGACCTAGGAGTAACA ATC). PCR was performed with primers from two different coding exons of FGF7. The primer pairs for the detection of FGF7 anti-sense transcript were p3 (5 0 -AGTGAGAAGACTCTTCTGTCGAACA) and p1, and for the detection of FGF7 mRNA p2 and p4 (5 0 -TGCTCTGGAGTCATG TCATTGCAAG). RT-PCR for GAPDH expression was used as an internal control for RNA levels. Immunohistochemistry. Preparation of rabbit polyclonal antibodies directed against human recombinant FGF7 was previously described [15]. Paraffin-embedded tissue sections (5 lm) were subjected to immunostaining using the streptavidin-peroxidase technique according to the manufacturer’s protocol (LAB-SA detection system, Zymed Labs, San Francisco, CA, USA). Endogenous peroxidase activity was blocked by incubation with 0.3% hydrogen peroxide in methanol (15 min, 24 C). Antigen retrieval was performed by microwave boiling in sodium citrate buffer (10 mM NaCitrate, pH 6.0, 10 min, 95 C), followed by incubation with blocking solution (1 h, 25 C). Incubation with first antibody was carried out for 16 h at 4 C. Anti-FGF7 antibody was diluted 1:250, and anti-insulin MAB was diluted according to manufacturer’s (Santa Cruz) recommendation. Color reaction was performed using streptavidin–peroxidase complex, with aminoethyl-carbazole solution as substrate. Mayer’s hematoxylin was applied as counterstaining where indicated. Staining intensity of the glands and foveole was graded from 0 to 5 by an experienced Pathologist (IM).
Results Detection of FGF protein in gastric diseases Expression of FGF7 protein was investigated by immunohistochemistry (IHC) employing rabbit polyclonal antibody directed against FGF7 [15]. Normal pancreas tissues that are known to express FGF7 [18] were utilized to determine FGF7 antibody staining specificity. Sections were incubated without or with the anti-FGF7 antibodies, and with the antibodies in the presence of recombinant human FGF7, recombinant human insulin or EGF. FGF7 immunoreactivity was observed in pancreatic acinar cells and to a lesser extent in ductal cells (Fig. 1A, and data not shown) in agreement with Ishiwata et al. [18]. No signal was observed in the presence of recombinant FGF7 or in the absence of the FGF7 antibodies (Fig. 1B, and data not shown). Positive FGF7 immunoreactivity was obtained in sections incubated with FGF7 antibodies and recombinant insulin (or EGF) whereas no insulin immunoreactivity was observed in sections incubated with anti-insulin antibodies and recombinant insulin (Fig. 1C–E, and data not shown). These results clearly indicate that the antibody specifically recognizes the FGF7 protein. The expression of FGF7 in normal stomach and gastric diseases was assessed using specimens from 5 normal cases, and 35 various gastric pathologies of which 22 were positive for HP infection. Representative results are shown in Fig. 1. In the normal stomach, weak FGF7 immunoreactivity
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
827
Fig. 1. Specificity of the FGF7 antibody (A–E). Human normal pancreas sections were incubated with FGF7 antibody in the absence (A) or presence of 10 lg/ml human recombinant FGF7 (B) or insulin (C). Pancreatic sections were incubated with anti-insulin antibody without (D) or with (E) recombinant insulin. IHC was carried out as described under the Materials and methods section. Arrowheads point to FGF7 immunoreactivity in acinar cells, and arrows to insulin immunoreactivity in islet cells surrounded by the negative acinar cells. Strong induction of FGF7 protein in gastric diseases (F–O). (F,G) normal stomach from two different cases. (H,I) Moderate gastritis from two different cases (J) Severe gastritis. (K) Gastric adenocarcinoma. (L,M) Mucosal glands and foveole of a third moderate gastritis case, respectively. (N,O) Mucosal glands and foveole of a second severe gastritis case, respectively. Note the strong immunoreactivity in the cytoplasm of epithelial cells of glands and foveole, and in infiltrating lymphocytes. lp, gl, fo and mm denote for lamina propria, glands, foveole, and muscularis mucosae, respectively. Insets represent magnification of foveole (H) or mucosal glands (J) from the same section. Magnification, 100· (A–C), 40· (D–K). Counterstaining with hematoxylin (A–K).
828
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
was detected in the fibroblasts and the infiltrating lymphocytes of the lamina propria, and a modest signal was observed in the mucosal gland epithelium, but not in foveolar epithelium. Little or no staining was observed in the muscularis mucosae (Fig. 1F–G). Marked induction of FGF7 protein levels was observed in the various gastric diseases. In all gastritis cases, intense FGF7 staining was observed throughout the lamina propria and in the mucosal epithelial cells in both the foveole and glands. Strong immunoreactivity was present in infiltrating lymphocytes, and little or no staining was observed in the muscularis mucosae (Fig. 1H–J and L–O). Similarly, strong and uniform FGF7 immunoreactivity was observed in gastric adenocarcinoma (Fig. 1K). A comparison of FGF7 protein levels in foveole and glands in the various gastric pathologies and in stomach from healthy donors is presented in Table 1. Nearly comparable levels of FGF7 protein were detected in the different histopathological conditions. Six to eightfold increase in FGF7 protein levels was observed in the mucosal glands in mild, moderate, and severe gastritis (total of 21 cases). Similar increase was detected in tissues undergoing remodeling without an overt gastritis, as well as atrophic gastritis (4 and 2 cases, respectively). Strong induction of FGF7 was observed in the foveolar epithelium as compared to little or no expression in normal stomach specimens. Similarly, high expression levels were observed in tumor cells of gastric adenocarcinoma (8/8 cases). These results suggest that upregulation of FGF7 protein is common in gastric pathologies. Detection of FGF7 transcripts in gastric diseases In situ hybridization, using probes complementary to the first exon of the FGF7 gene (Fig. 2A), was carried out in order to assess whether the high induction of FGF7 protein in gastric pathologies is accompanied with a similar induction in FGF7 mRNA. For optimization of the in situ conditions, we initially utilized tissue sections from gastritis
samples. Unexpectedly, a strong signal was observed with the sense probe in all the tested gastritis samples. The signal obtained using the anti-sense (AS) probe, under comparable conditions, was relatively weak, but could be intensified by increasing probe concentration and/or the incubation time during the colorimetric reaction (see representative results in Fig. 2B–D). Under these conditions, no staining was observed in control sample treated in the absence of either probes (data not shown). These results were confirmed with an additional probe complementary to a different region along the FGF7 mRNA (designated probe B, Fig. 2A, E and F). Moreover, in situ hybridization performed with AS and sense probes for a different gene expressed in the GI-tract (perlecan [20]) revealed a positive signal only with the AS-probe (Fig. 2G and H). These results revealed the existence of FGF7 AS-transcript at levels that exceed those of the FGF7 mRNA. To further assess the induction of FGF7 transcripts in gastric diseases, in situ hybridization was carried out with specimens from normal stomach (5 cases), moderate and severe gastritis (5 and 7 cases, respectively), and gastric adenocarcinoma (5 cases). Hybridization was carried out using two probe concentrations (0.2 and 0.5 lg/ml). At the lower AS-riboprobe concentration the results were similar to those described above. FGF7 mRNA was barely detectable while FGF7 AS-RNA was readily detected with the sense riboprobe (data not shown). FGF7 mRNA was detected with the higher probe concentration, and its levels were elevated in gastric diseases, but to a modest degree. FGF7 AS-RNA was not detected in normal stomach whereas it was strongly induced in all gastric pathologies (representative results are shown in Fig. 2). We did not observe a significant difference in staining intensity between the different gastric pathologies indicating that upregulation of FGF7 transcripts, similar to FGF7 protein, is an early event in gastric inflammation. Both RNA types were detected in epithelial cells in both mucosal glands and foveole, as well as in fibroblasts and infiltrating lymphocytes in
Table 1 Comparison of FGF7 protein expression levels in the different gastric histopathological states Diagnosis
Normal Remodeling Gastritis Mild Moderate Severe Atrophic gastritis Adenocarcinoma Poorly differentiated Well differentiated
Number n = 40
5 4 23 1 11 9 2 8 5 3
HP positive
0 1 17 0 10 7 0 4 4 0
FGF7 protein levels (0–5) Glands
Foveolesa
Other
0.5 3.5 3 3 3.1 3.3 4 3 2.5 4
ND 3.25 3.1 5 2.8 3 3 2.5 2.3 3
3 2 3.5
The biopsies from each histopathological state were microscopically examined to estimate FGF7 expression level in the foveole and gland or tumor cells, in a scale between 1 and 5, where 5 is the highest and 0 is the lowest expression level. The numbers representing the average score of expression level. Foveolar and glands stain was significantly higher in all gastric pathologies, including intact gastric tissue from adenocarcinoma surgical specimens, compared with the normal tissues. ND denotes, not detected. a Since FGF7 was not detected in the gastric pits of stomach from healthy donors (normal), the expression levels in foveole were estimated relative to the levels observed in mucosal glands from normal stomach.
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
A FGF7 mRNA
I/E 738
446
1
1030
3850
ORF
5' UTR 1
I/E 849
829
680 Probe A 552
1030 Probe B
Fig. 2. Induction of FGF7 transcripts in gastric inflammation. (A) Schematic representation of the FGF7 mRNA and localization of the probes used for the in situ hybridization. I/E denotes for exon/intron junction. UTR, untranslated region; ORF, open reading frame. (B–H) Detection of FGF7 and perlecan transcripts in gastric tissues. Gastritis specimens were used for the initial detection of FGF7 and perlecan transcripts. In situ hybridization was carried out with FGF7 probes A or B, or perlecan probe (0.2 lg/ml for each probe), in both the anti-sense and sense orientation. (B–D) Serial sections of moderate gastritis specimen hybridized with anti-sense and sense riboprobe A. (B,C) Hybridization with the anti-sense probe. The sections in (B,C) were incubated for 1.5 and 3 h during the colorimetric reaction, respectively. (D) Hybridization with sense probe (1.5 h incubation during the colorimetric reaction). (E,F) Serial sections of a different gastritis sample hybridized with anti-sense and sense probe B, respectively. (G,H) Hybridization with antisense and sense probes for perlecan, respectively. (I–N) The expression of FGF7 transcripts is elevated in gastric diseases. In situ hybridization was carried out with 0.5 lg/ml probe B, in both the anti-sense (I–K) or sense (L–N) orientation. (I,L) normal stomach. (J,M) Moderate gastritis. (K,N) Severe gastritis; similar results were obtained with riboprobe A. Magnification, 25· (B–F), 40· (G–N).
the various gastric pathologies. In addition, FGF7 ASRNA and protein/mRNA were co-localized (compare panels J and K to panels M and N in Fig. 2, and see Fig. 3). These results show that FGF7 transcripts are upregulated in gastric diseases, and that induction of FGF7 AS-RNA exceeds the induction of its mRNA.
FGF7 AS-RNA is expressed in other human tissues and cultured cells RT-PCR analysis was carried out to determine whether FGF7 AS-RNA is expressed in other human cell types and tissues. The cDNA was generated using either primer
830
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
Fig. 3. FGF7 anti-sense transcript and protein are co-localized. (A–C) In situ hybridization, with probe A in the sense orientation, for the detection of FGF7 anti-sense RNA. (D–F) Detection of FGF7 protein by IHC. (A,D) Serial sections from a moderate gastritis case with intestinal metaplasia. (B,E) Serial sections from a severe gastritis case demonstrating FGF7 protein and AS-RNA expression in a foveole. (C,F) Mucosal glands area from a moderate gastritis case. Counterstaining with hematoxylin (D–F).
specific to FGF7 mRNA or to the FGF7-AS transcript. Among the normal tissues that were examined FGF7 ASRNA was not detected in normal fetal or adult human brain, thyroid, liver, kidney or adenoids, and a weak signal was observed in tonsils (Fig. 4, and data not shown). ASFGF7 transcript was highly expressed in primary human dermal fibroblasts and lung fibroblasts, primary dermal fibroblasts from psoriatic lesions, fibroblast cell line derived from human embryos carrying dyssegmental dysplasia, Silverman–Handmaker type (DDSH) syndrome [21,22], and in polyp and mucosa specimens from a familial polyposis case. Weak expression of the anti-sense RNA was observed in prostate carcinoma. Samples from various areas of colonic adenocarcinoma were negative. The expression patterns of FGF7 AS-transcript and mRNA were similar, except for cecum carcinoma where FGF7 mRNA, but not AS-RNA was detected.
Fig. 4. Expression of FGF AS-RNA in human tissues and cell lines. RNA extraction and specific amplification of FGF7 AS-RNA or mRNA, by RT-PCR, were described under the Materials and methods section. P, fibro, and ca denotes for primary, fibroblasts, and carcinoma, respectively. DDSH fibro is an established fibroblast cell line from human embryos carrying dyssegmental dysplasia, Silverman–Handmaker type (DDSH) syndrome [21,22].
Discussion Chronic inflammation in various sites, including the GI tract, is associated with an increased risk of human cancer [3,4]. FGF7 is a specific mitogen for epithelial cells, from which the majority of cancer types arise [6]. Elevated FGF7 levels stimulate epithelial cell hyperplasia along the GI tract in experimental animals and are associated with inflammation and cancer of the intestine and the upper GI tract [6–10,23]. In the present study, we report, for the first time, that FGF7 protein is expressed in the mucosal epithelial cells in the normal adult human stomach, and that marked induction of FGF7 protein and its anti-sense transcript is a common theme in human gastric diseases. FGF7 is generally considered a mesenchyme derived growth factor that acts on adjacent epithelial cells [24,25]. In normal human stomach, FGF7 protein was expressed at low levels in the lamina propria and mucosal gland epithelial cells. As the FGF7 receptor is expressed in stomach epithelial cells [24,25], our findings indicate that FGF7 can act both in a paracrine manner, and as part of an autocrine loop, in human stomach. Such an autocrine loop may compromise the tight regulation of the directional epithelial– mesenchymal signaling, promoting the growth of epithelial cells. In addition, our findings further support the previously suggested role of FGF7 in preservation of GI epithelial tissue [6,7,13]. Similar to the current findings, several studies indicated that FGF7 is expressed in certain normal epithelial cells, including endometrial epithelium of porcine uterus [26], pancreatic acinar and ductal cells [18], and primary lens epithelial cells [27]. FGF7 protein levels were markedly elevated in the gastric mucosa in each of the gastric pathologies we examined including mild–severe gastritis, atrophic gastritis, and tissues undergoing remodeling without an overt gastritis (total of 27 cases). Strong induction of FGF7 protein levels
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
was also observed in the foveolar epithelium that was negative for FGF7 expression in the normal human stomach. Similar increase was observed in 8/8 gastric adenocarcinoma cases. These results indicate that FGF7 protein upregulation is common in gastric diseases, probably occurring early in gastric inflammation. Enhanced FGF7 protein and RNA levels were similar in HP-positive (22/35) as well as HP-negative (13/35) cases. These findings suggest that factors common to HP negative/HP positive gastric inflammations may be responsible for FGF7 upregulation. Likely candidates are proinflammatory cytokines, such as TNFa, IL1b, and IL6, that are known to be produced in the inflamed stomach and have been shown to upregulate FGF7 expression [4,28–30]. In addition, these proinflammatory cytokines are thought to contribute to carcinogenesis by influencing the survival, growth, mutation, differentiation, and migration of tumor cells and by regulating angiogenesis [31,32]. On the basis of its induction at early stages of gastric inflammation, and its continued expression during the different pre and malignant stages of gastric cancer, it is tempting to suggest that FGF7 may be central in mediating many effects of proinflammatory cytokines in gastric inflammation and gastric carcinogenesis. Numerous studies support a role for FGF7 in gastric cancer development. Expression of FGF7 receptor was implicated in gastric cancer [33–36], and studies in cultured cells indicated that the ability of overexpressed FGF7-receptor to induce malignant transformation is ligand dependent [37]. Thus, elevated levels of FGF7 should accelerate the proliferation of premalignant and malignant gastric epithelial cells. This is consistent with reports that FGF7 stimulates the proliferation of gastric cancer cells in vitro and its ability to induce epithelial cell hyperplasia along the GI tract when administered to rats [7,14]. Besides its specific effects on proliferation, survival, and migration of epithelial cells, FGF7 also induces the expression of angiogenic growth factors (reviewed in [6]) which are essential for solid tumor growth and metastasis [38]. Our results suggest that FGF7 mRNA levels in the normal human stomach are quite low and are only modestly upregulated in stomach diseases as compared to the marked induction of FGF7 protein or FGF7 AS-RNA. These observations imply that FGF7 protein levels are regulated mainly by posttranscriptional mechanisms. In agreement with our findings, Kinoshita et al. reported that FGF7 mRNA levels are rather low in rat stomach [12]. In rat stomach, however, FGF7 mRNA was not detected in the epithelium. These results suggest that the expression pattern of FGF7 in the human stomach differs from that of rat stomach. Another plausible explanation is that similar to human stomach, rat FGF7 mRNA levels in the rat gastric mucosal epithelium are extremely low, and therefore, were below detection under Kinoshita et al. experimental conditions. The observed low levels of FGF7 mRNA may also explain why it has not been detected in gastric cancer cell lines [8,14]. Alternatively, FGF7 expression in
831
gastric cancer cell lines may be switched off under in vitro growth conditions. We discovered the existence of endogenous FGF7 ASRNA in human gastric diseases, as well as other tissues, and cultured cells mainly from pathological conditions. To the best of our knowledge this is the first demonstration for the existence of FGF7 AS-RNA. AS-RNA was reported for other FGFs as well [39,40]. Since FGFs represent a highly conserved gene family, it is possible that a parallel FGF-AS family exists, whose biological function remains to be elucidated. The FGF7 AS-RNA was complementary to at least the 5 0 UTR and the open reading frame (ORF) of FGF7 mRNA, but no coding capacity was detected within this region of the AS-RNA. In the gastric diseases, FGF7 AS-RNA was co-localized with FGF7 mRNA. Moreover, in other tissues and cultured cells positive for FGF7 AS-RNA, we always detected FGF7 mRNA. These results suggest that transcription of FGF7 mRNA and ASRNA may be co-regulated. Endogenous AS-RNAs are thought to play diverse roles, including regulation of gene expression at the level of transcription, mRNA processing, mRNA stability or translation. Anti-sense regulation generally refers to regulation of expression of an RNA target through direct base pairing with a complementary RNA and is usually negative (reviewed in [41]) . In some cases, however, anti-sense RNA positively modulated gene expression and protein translation [42,43]. The observed co-localization of high FGF7 protein and AS-RNA levels in gastric diseases (see Fig. 3), together with the finding that FGF7 mRNA levels are rather low in these diseases, supports an attractive role for FGF7 AS-RNA in positively regulating FGF7 translation efficiency. Further studies are required to address this possibility. In summary, this study suggests a role for FGF7 in the preservation of the gastric mucosa acting both in a paracrine and an autocrine manner. The marked induction of FGF7 protein in gastritis and gastric adenocarcinoma, together with the known oncogenic potential of FGF7, suggest that excessive FGF7 signaling may contribute to gastric cancer development and progression. Gastric cancer represents a major health problem, and despite the worldwide decline in incidence, and the vast improvements in diagnosis and treatment, most patients survive only few years [2,44]. Targeting FGF7 signaling may be an attractive mode for prevention and treatment of gastric cancer. Acknowledgments We thank Natasha Kipnis for her kind help in processing biopsy sections and Sharon Lubinsky-Mink for technical assistance. This work was supported by a grant from the Technion and Bene Zion Hospital research foundation to Dina Ron and Ron Shaoul (#140–614) and partially supported by grants from Israel Cancer Research Fund (ICRF #2004973) Israel Science foundation (ISF #2006261) and Israel Ministry of Health (# 907134; 1-6305) to Dina Ron.
832
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
References [1] T.J. Koh, T.C. Wang, Tumors of the stomach, in: M. Feldman, W.O. Tschumy, L.S. Friedman, M.H. Sleisenger (Eds.), Sleisenger & Fordtran’s Gastrointestinal and Liver Disease, 2002, pp. 829–856. [2] M.G. Smith, G.L. Hold, E. Tahara, E.M. El-Omar, Cellular and molecular aspects of gastric cancer, World J. Gastroenterol. 12 (2006) 2979–2990. [3] C. Stoicov, R. Saffari, X. Cai, C. Hasyagar, J. Houghton, Molecular biology of gastric cancer: Helicobacter infection and gastric adenocarcinoma: bacterial and host factors responsible for altered growth signaling, Gene 341 (2004) 1–17. [4] H. Ohshima, H. Tazawa, B.S. Sylla, T. Sawa, Prevention of human cancer by modulation of chronic inflammatory processes, Mutat. Res. 591 (2005) 110–122. [5] J.C. Becker, C. Muller-Tidow, H. Serve, W. Domschke, T. Pohle, Role of receptor tyrosine kinases in gastric cancer: new targets for a selective therapy, World J. Gastroenterol. 12 (2006) 3297–3305. [6] P.W. Finch, J.S. Rubin, Keratinocyte growth factor/fibroblast growth factor 7, a homeostatic factor with therapeutic potential for epithelial protection and repair, Adv. Cancer Res. 91 (2004) 69–136. [7] R.M. Housley, C.F. Morris, W. Boyle, B. Ring, R. Biltz, J.E. Tarpley, S.L. Aukerman, P.L. Devine, R.H. Whitehead, G.F. Pierce, Keratinocyte growth factor induces proliferation of hepatocytes and epithelial cells throughout the rat gastrointestinal tract, J. Clin. Invest. 94 (1994) 1764–1777. [8] S. Iida, O. Katoh, A. Tokunaga, M. Terada, Expression of fibroblast growth factor gene family and its receptor gene family in the human upper gastrointestinal tract, Biochem. Biophys. Res. Commun. 199 (1994) 1113–1119. [9] P.W. Finch, V. Pricolo, A. Wu, S.D. Finkelstein, Increased expression of keratinocyte growth factor messenger RNA associated with inflammatory bowel disease, Gastroenterology 110 (1996) 441–451. [10] M. Brauchle, M. Madlener, A.D. Wagner, K. Angermeyer, U. Lauer, P.H. Hofschneider, M. Gregor, S. Werner, Keratinocyte growth factor is highly overexpressed in inflammatory bowel disease, Am. J. Pathol. 149 (1996) 521–529. [11] Y. Matsubara, M. Ichinose, M. Tatematsu, M. Ichinose, M. Oka, N. Yahagi, K. Kurokawa, T. Kageyama, K. Miki, H. Fukamachi, Stagespecific elevated expression of the genes for hepatocyte growth factor, keratinocyte growth factor, and their receptors during the morphogenesis and differentiation of rat stomach mucosa, Biochem. Biophys. Res. Commun. 222 (1996) 669–677. [12] Y. Kinoshita, H. Nakata, S. Hassan, M. Asahara, C. Kawanami, Y. Matsushima, Y. Naribayashi-Inomoto, C.Y. Ping, D. Min, A. Nakamura, Gene expression of keratinocyte and hepatocyte growth factors during the healing of rat gastric mucosal lesions, Gastroenterology 109 (1995) 1068–1077. [13] P. Chailler, J.R. Basque, L. Corriveau, D. Menard, Functional characterization of the keratinocyte growth factor system in human fetal gastrointestinal tract, Pediatr. Res. 48 (2000) 504–510. [14] K. Nakazawa, M. Yashiro, K. Hirakawa, Keratinocyte growth factor produced by gastric fibroblasts specifically stimulates proliferation of cancer cells from scirrhous gastric carcinoma, Cancer Res. 63 (2003) 8848–8852. [15] D. Ron, D.P. Bottaro, P.W. Finch, D. Morris, J.S. Rubin, S.A. Aaronson, Expression of biologically active recombinant keratinocyte growth factor. Structure/function analysis of aminoterminal truncation mutants, J. Biol. Chem. 268 (1993) 2984–2988. [16] M.F. Dixon, R.M. Genta, J.H. Yardley, P. Correa, Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994, Am. J. Surg. Pathol. 20 (1996) 1161–1181. [17] P.W. Finch, J.S. Rubin, T. Miki, D. Ron, S.A. Aaronson, Human KGF is FGF-related with properties of a paracrine effector of epithelial cell growth, Science 245 (1989) 752–755.
[18] T. Ishiwata, H. Friess, M.W. Buchler, M.E. Lopez, M. Korc, Characterization of keratinocyte growth factor and receptor expression in human pancreatic cancer, Am. J. Pathol. 153 (1998) 213–222. [19] A. Eisemann, J.A. Ahn, G. Graziani, S.R. Tronick, D. Ron, Alternative splicing generates at least five different isoforms of the human basic-FGF receptor, Oncogene 6 (1991) 1195–1202. [20] A.D. Murdoch, B. Liu, R. Schwarting, R.S. Tuan, R.V. Iozzo, Widespread expression of perlecan proteoglycan in basement membranes and extracellular matrices of human tissues as detected by a novel monoclonal antibody against domain III and by in situ hybridization, J. Histochem. Cytochem. 42 (1994) 239–249. [21] E. Arikawa-Hirasawa, W.R. Wilcox, A.H. Le, N. Silverman, P. Govindraj, J.R. Hassell, Y. Yamada, Dyssegmental dysplasia, Silverman–Handmaker type, is caused by functional null mutations of the perlecan gene, Nat. Genet. 27 (2001) 431–434. [22] E. Arikawa-Hirasawa, H. Watanabe, H. Takami, J.R. Hassell, Y. Yamada, Perlecan is essential for cartilage and cephalic development, Nat. Genet. 23 (1999) 354–358. [23] M. Watanabe, T. Ishiwata, K. Nishigai, Y. Moriyama, G. Asano, Overexpression of keratinocyte growth factor in cancer cells and enterochromaffin cells in human colorectal cancer, Pathol. Int. 50 (2000) 363–372. [24] I.J. Mason, F. Fuller-Pace, R. Smith, C. Dickson, FGF-7 (keratinocyte growth factor) expression during mouse development suggests roles in myogenesis, forebrain regionalisation and epithelial–mesenchymal interactions, Mech. Dev. 45 (1994) 15–30. [25] P.W. Finch, G.R. Cunha, J.S. Rubin, J. Wong, D. Ron, Pattern of keratinocyte growth factor and keratinocyte growth factor receptor expression during mouse fetal development suggests a role in mediating morphogenetic mesenchymal-epithelial interactions, Dev. Dyn. 203 (1995) 223–240. [26] H. Ka, T.E. Spencer, G.A. Johnson, F.W. Bazer, Keratinocyte growth factor: expression by endometrial epithelia of the porcine uterus, Biol. Reprod. 62 (2000) 1772–1778. [27] J. Weng, Q. Liang, R.R. Mohan, Q. Li, S.E. Wilson, Hepatocyte growth factor, keratinocyte growth factor, and other growth factorreceptor systems in the lens, Invest. Ophthalmol. Vis. Sci. 38 (1997) 1543–1554. [28] M. Brauchle, K. Angermeyer, G. Hubner, S. Werner, Large induction of keratinocyte growth factor expression by serum growth factors and pro-inflammatory cytokines in cultured fibroblasts, Oncogene 9 (1994) 3199–3204. [29] M. Chedid, J.S. Rubin, K.G. Csaky, S.A. Aaronson, Regulation of keratinocyte growth factor gene expression by interleukin 1, J. Biol. Chem. 269 (1994) 10753–10757. [30] P.W. Finch, C. Lengel, M. Chedid, Cloning and characterization of the promoter region of the human keratinocyte growth factor gene, J. Biol. Chem. 270 (1995) 11230–11237. [31] F. Balkwill, A. Mantovani, Inflammation and cancer: back to Virchow? Lancet 357 (2001) 539–545. [32] L.M. Coussens, Z. Werb, Inflammation and cancer, Nature 420 (2002) 860–867. [33] Y. Hattori, H. Itoh, S. Uchino, K. Hosokawa, A. Ochiai, Y. Ino, H. Ishii, H. Sakamoto, N. Yamaguchi, K. Yanagihara, S. Hirohashi, T. Sugimura, M. Terada, Immunohistochemical detection of K-sam protein in stomach cancer, Clin. Cancer Res. 2 (1996) 1373–1381. [34] H. Ishii, T. Yoshida, H. Oh, S. Yoshida, M. Terada, A truncated K-sam product lacking the distal carboxyl-terminal portion provides a reduced level of autophosphorylation and greater resistance against induction of differentiation, Mol. Cell Biol. 15 (1995) 3664–3671. [35] H. Itoh, Y. Hattori, H. Sakamoto, H. Ishii, T. Kishi, H. Sasaki, T. Yoshida, M. Koono, T. Sugimura, M. Terada, Preferential alternative splicing in cancer generates a K-sam messenger RNA with higher transforming activity, Cancer Res. 54 (1994) 3237–3241. [36] T. Matsunobu, T. Ishiwata, M. Yoshino, M. Watanabe, M. Kudo, K. Matsumoto, A. Tokunaga, T. Tajiri, Z. Naito, Expression of
R. Shaoul et al. / Biochemical and Biophysical Research Communications 350 (2006) 825–833
[37]
[38] [39]
[40]
keratinocyte growth factor receptor correlates with expansive growth and early stage of gastric cancer, Int. J. Oncol. 28 (2006) 307–314. E. Shaoul, R. Reich-Slotky, B. Berman, D. Ron, Fibroblast growth factor receptors display both common and distinct signaling pathways, Oncogene 10 (1995) 1553–1561. R.S. Kerbel, Tumor angiogenesis: past, present and the near future, Carcinogenesis 21 (2000) 505–515. R.S. Knee, S.E. Pitcher, P.R. Murphy, Basic fibroblast growth factor sense (FGF) and antisense (gfg) RNA transcripts are expressed in unfertilized human oocytes and in differentiated adult tissues, Biochem. Biophys. Res. Commun. 205 (1994) 577–583. F. Coulier, V. Ollendorff, I. Marics, O. Rosnet, M. Batoz, J. Planche, S. Marchetto, M.J. Pebusque, O. deLapeyriere, D. Birnbaum, The
[41]
[42]
[43]
[44]
833
FGF6 gene within the FGF multigene family, Ann. NY Acad. Sci. 638 (1991) 53–61. S.H. Munroe, Diversity of antisense regulation in eukaryotes: multiple mechanisms, emerging patterns, J. Cell Biochem. 93 (2004) 664–671. R.A. Lease, M. Belfort, A trans-acting RNA as a control switch in Escherichia coli: DsrA modulates function by forming alternative structures, Proc. Natl. Acad. Sci. USA 97 (2000) 9919–9924. E.S. Tasheva, D.J. Roufa, Regulation of human RPS14 transcription by intronic antisense RNAs and ribosomal protein S14 Genes, Development 9 (1995) 304–316. Y. Kodera, M. Fujiwara, M. Koike, A. Nakao, Chemotherapy as a component of multimodal therapy for gastric carcinoma, World J. Gastroenterol. 12 (2006) 2000–2005.