- Email: [email protected]
The expression of hepatocyte nuclear factor-4α, a developmental regulator of visceral endoderm, correlates with the intestinal phenotype of gastric adenocarcinomas
Pathology (December 2006) 38(6), pp. 548–554
ANATOMICAL PATHOLOGY
The expression of hepatocyte nuclear factor-4a, a developmental regulator of visceral endoderm, correlates with the intestinal phenotype of gastric adenocarcinomas KAZUHIRO KOJIMA*{, TAKASHI KISHIMOTO*, YUICHIRO NAGAI*, TOHRU TANIZAWA{, YUKIO NAKATANI{, MASARU MIYAZAKI{ AND HIROSHI ISHIKURA*
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
Departments of *Molecular Pathology, {General Surgery, and {Diagnostic Pathology, Chiba University Graduate School of Medicine, Chiba, Japan
Summary Aims: Hepatocyte nuclear factor (HNF)-4a is a developmental regulator of the visceral endoderm, which is expressed in the embryonic gut and later in the adult intestine and colon. However, adult gastric mucosa does not express HNF-4a. We investigated the possible involvement of HNF-4a in intestinal metaplasia and the intestinalisation of gastric adenocarcinomas. Methods: Thirty-five cases of adenocarcinomas and 46 cases of adjacent non-neoplastic mucosa with (22 lesions) or without (24 lesions) intestinal metaplasia were immunostained for HNF-4a. The gastric or the intestinal phenotype was also examined using immunohistochemistry for MUC5AC, MUC2, CD10, and gastric-type mucin (GTM). Adenocarcinomas were classified into the gastric type (Gtype, 42.9%), the mixed gastric and intestinal type (GI-type, 31.4%), and the intestinal type (I-type, 25.7%). Results: The HNF-4a expression was exclusively seen in glandular cells with intestinal metaplasia, which was correlated with MUC2 expression (p,0.05) and inversely correlated with MUC5AC expression (p,0.05). All adenocarcinomas more or less expressed HNF-4a, with an intense expression being seen in the I-type (p,0.01) and in welldifferentiated adenocarcinomas (p,0.03). Conclusions: HNF-4a expression is associated with the intestinal phenotype of non-neoplastic and neoplastic gastric glandular cells, suggesting a possible involvement in the establishment and/or maintenance of the intestinal phenotype of the gastric mucosa and adenocarcinomas. Key words: Hepatocyte nuclear factor-4a, intestinal metaplasia, adenocarcinoma, MUC2, MUC5AC, immunohistochemistry. Received 30 May, revised 5 July, accepted 5 September 2006
INTRODUCTION Intestinal metaplasia of the gastric mucosa are defined as the ectopic occurrence of intestinal-type epithelial cells such as goblet, absorptive, and Paneth cells in the adult stomach. It is not so difficult to detect intestinal metaplasia lesions morphologically on H&E sections, and periodic-acid Schiff (PAS) or alcian blue stain is useful in recognising goblet cells. Recently, expressions of mucin core protein genes have been estimated to be in the normal gastrointestinal
tract. Among these mucins, MUC5AC and MUC2 are widely used as helpful markers for the gastric or intestinal phenotype of gastric epithelial cells,1,2 respectively. Gastric carcinomas have been classified by Lauren3 into two major histological groups: the intestinal type and the diffuse type. This classification divides tumours basically by the degree of glandular formation. On the other hand, the classification that is based on the epithelial mucin phenotype (i.e., the intestinal phenotype or gastric phenotype) is especially meaningful for the study of the histogenesis of gastric adenocarcinomas. A sequence from intestinal metaplasia through dysplasia to adenocarcinoma, which is called Correa’s cascade, has been proposed.4 This sequence has been supported by epidemiological findings that there is an increased risk of cancer development in patients with intestinal metaplasia,5,6 but these findings are still controversial.7,8 Mucin phenotypes between microscopic adenocarcinoma and its surrounding mucosa are not always identical, suggesting that intestinal metaplasia is not a necessary precancerous condition.9 Although the molecular mechanisms of phenotypical regulation in gastric mucosa or gastric adenocarcinoma have yet to be clearly established, the importance of several signal pathways and transcriptional regulation has become evident. These involve molecules that have a pivotal role in the morphogenesis of the organs of endodermal origin, such as small intestine, colon, and stomach. Sonic hedgehog (Shh) is expressed in gastric parietal cells in adult humans,10 and Shh-deficient mice generate intestinal metaplasia in their stomachs.11 Caudal-type homeobox transcriptional factor (Cdx)1 and Cdx2 are expressed in the glandular cells of the small intestine and colon, where they play important roles in the intestinal phenotype.12,13 These transcriptional factors play a key role also in the development of intestinal metaplasia;14,15 direct evidence has been demonstrated by Cdx2-transgenic mice which developed intestinal-type adenocarcinoma from intestinal metaplasia.16 Hepatocyte nuclear factor (HNF)-4a, a hormone receptor-type transcriptional factor,17 contributes a variety of gene expressions in many cell functions, such as liver morphogenesis,18 glucose homeostasis,19 and lipid absorption.20 HNF-4a is considered to be a developmental regulator of the visceral endoderm.21–23 In embryogenesis, the expression of HNF-4a mRNA first appears in the whole primitive endoderm; it is later restricted to the
ISSN 0031-3025 printed/ISSN 1465-3931 # 2006 Royal College of Pathologists of Australasia DOI: 10.1080/00313020601024011
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
HNF-4a EXPRESSION IN STOMACH
visceral endoderm, followed by expression in the embryonic liver, kidney, pancreas, and gut.24 HNF-4a has nine isoforms that are generated by alternative promoter (P1 and P2) usage and alternative splicing.25,26 Six isoforms, HNF4a1–a6, are generated by the usage of the P1 promoter, and three isoforms, HNF4a7– a9, by the P2 promoter. The P1 and P2 promoter-driven HNF-4a is distributed differently in adults; the expression of P1 promoter-driven HNF-4a is observed in the absorptive epithelium of the small intestine, crypt of colon, renal tubules, and pancreas. Interestingly, P1 promoter-driven HNF-4a cannot be detected in the normal gastric mucosa,27,28 although the stomach is of endodermal origin. Also interestingly, as Tanaka et al. have reported, restricted expression of P1 promoter-driven HNF-4a is observed in intestinal metaplasia in the adult gastric mucosa and in 57.1% of gastric adenocarcinomas.28 Their data suggest the possibility that HNF-4a, a developmental regulate of visceral endoderm, contributes to the intestinalisation or the maintenance of the intestinal phenotype of gastric mucosa. However, the correlation between P1 promoterdriven HNF-4a expression and the intestinal phenotype of the gastric mucosa or gastric adenocarcinoma has not been analysed in detail. Therefore, we examined the expression of P1 promoter-driven HNF-4a in adenocarcinoma tissues to test whether its expression is correlated with the intestinal phenotype of gastric adenocarcinomas.
549
diagnosis of carcinoma was based on the criteria of the Japanese Classification of Gastric Carcinoma.29 There were 26 cases of tubular adenocarcinomas (9 of the well-differentiated type and 17 of the moderately-differentiated type), and nine cases of poorly-differentiated adenocarcinoma. As defined according to these criteria, tubular and poorly-differentiated adenocarcinomas roughly correspond to Lauen’s classification of the ‘intestinal type’ and the ‘diffuse type’, respectively.1 From formalin-fixed, paraffin-embedded tissues, 4 mm thick sections were used for H&E stain and immunohistochemical analysis. We evaluated a series of 46 non-neoplastic lesions in the vicinity of the 35 cases of adenocarcinomas. The lesions included gastric mucosa with and without intestinal metaplasia, which were 22 and 24 lesions, respectively. Immunohistochemistry Immunohistochemical study was performed using primary antibodies against HNF-4a (K9218, 1:100; Perseus Proteomics, Japan), MUC2 (1:100; Medac Diagnostika, Germany), MUC5AC (1:50; Novocastra, UK), a gastric-type mucin (GTM; HIK1083, 1:25; Kanto Chemical, Japan), and CD10 (1:80; Novocastra). Anti-HNF-4a antibody, K9218, recognises HNF-4a1, 2, 3, 4, 5, and 6 isoforms, which are derived by P1 promoter usage. The immunoperoxidase method with a LSAB kit (Dako Cytomation, Japan) was used. Each section was deparaffinised by xylene and rehydrated in 100% ethanol. Endogenous peroxidase activity was blocked by 3% H2O2 for 5 min, and slides were washed in phosphate-buffered saline (PBS) three times for 5 min. All sections, except for HIK1083, were treated with microwaves for 20 min in a 10 mM sodium citrate buffer (pH 6.0). After blocking, the sections were incubated with each primary antibody overnight at 4uC. The slides were washed in PBS three times for 5 min. They were incubated with biotinylated secondary antibodies and then with avidinhorseradish peroxidase according to the manufacturer’s protocol. After they were washed in PBS, diaminobenzidine tetrahydrochloride (DAB) substrate was used for visualisation, and haematoxylin was used for the counterstain.
MATERIALS AND METHODS Samples
Classification of gastric adenocarcinomas by immunohistochemistry
Included in this study were 35 cases of surgically-resected gastric adenocarcinoma from the file of Chiba University Hospital, Japan. The
We used antibodies for MUC2, CD10, MUC5AC and GTM for the classification of adenocarcinomas in relation to the intestinal phenotype.
Fig. 1 Grade of HNF-4a expression by immunohistochemistry. (A) HNF-4a expression in the duodenal mucosa; (B) 1+, tumour cells were less positive than the intensity of the duodenal mucosa; (C) 2+, tumour cells were as equally positive as the intensity of the duodenal mucosa; (D) 3+, tumour cells were more positive than the intensity of the duodenal mucosa (immunostain for HNF-4a, 6200).
550
KOJIMA et al.
Pathology (2006), 38(6), December
TABLE 1 Expression of HNF-4a in a series of 46 non-neoplastic gastric mucosa
IM
MUC2
CD10
MUC5AC
GTM
(+) (+) (+) (2) (2)
(+) (+) (+) (2) (2)
(+) (2) (2) (2) (2)
(2) (2) (+) (+) (+)
(2) (2) (2) (2) (+)
HNF-4a (2)
NHF-4a (+)
n (%)
n (%)
0 0 0 10 10
(0) (0) (0) (71.4) (100)
2 5 15 4 0
Total
(100) (100) (100) (28.6) (0)
2 5 15 14 10
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
IM, intestinal metaplasia; GTM, gastric-type mucin reactive with HIK1083 antibody. MUC2 is a gel-forming mucin that is specifically expressed in the goblet cells.30 CD10, a human membrane-associated neutral peptidase, is prominently expressed in the brush border of intestinal absorptive cells.31 MUC5AC is a gel-forming mucin that is specifically expressed in the foveolar epithelium of gastric mucosa and mucous neck cells.32 HIK1083 recognises gastric-type mucin expressed in cardiac gland cells, mucous cells and pyloric gland cells.33 In accordance with the definition used by Saito et al.34 and Kawachi et al.,35 we classified the adenocarcinomas into four types: (1) the gastric type (G-type), in which more than 10% of cancer cells are positive for MUC5AC and/or GTM and negative for both MUC2 and CD10; (2) the intestinal type (I-type), in which more than 10% of the cancer cells are positive for MUC2 and/or CD10 and negative for both MUC5AC and GTM; (3) the mixed gastric and intestinal type (GI-type), in which more than 10% of the cancer cells are positive for MUC5AC and/or GTM and also more than 10% of the cancer cells are positive for MUC2 and/or CD10; (4) the negative type (N-type), in which MUC2, CD10, MUC5AC, and GTM are all negative.
TABLE 2 Expression of HNF-4a in a series of 46 non-neoplastic lesions, according to the expression of MUC2 and MUC5AC
MUC2 expression (n) Positive (22) Negative (24) MUC5AC expression (n) Positive (39) Negative (7)
HNF-4a (2)
HNF-4a (+)
n (%)
n (%)
p value
0 (0) 20 (83.3)
22 (100) 4 (16.7)
0.045
20 (51.3) 0 (0)
19 (48.7) 7 (100)
0.011
1+, the intensity was less than that of duodenal mucosa; 2+, the intensity was equivalent to that of duodenal mucosa; 3+, the intensity was greater than that of duodenal mucosa (Fig. 1).
Grading of HNF-4a expression
Statistical analysis
Positive staining of HNF-4a was graded as 1+, 2+, or 3+ on the basis of comparisons of the intensity of HNF-4a staining of the duodenal mucosa:
Statistical analysis of the results was carried out using the chi-square test. Values of p of less than 0.05 were considered statistically significant.
Fig. 2 The expression of HNF-4a in intestinal metaplasia. (A) H&E stain (6100). (B) Immunohistochemistry for HNF-4a on the serial section of (A) (6100). (C) Immunohistochemistry for HNF-4a exhibited an increased expression of HNF-4a at the bottom of the metaplastic glands (6200).
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
HNF-4a EXPRESSION IN STOMACH
551
Fig. 3 Representative immunohistochemical findings on gastric mucosa with intestinal metaplasia (6100). (A–C) are serial sections. (A) MUC5AC; (B) MUC2; (C) HNF-4a. Note the almost mirror image between (A) MUC5AC and (B) MUC2 expression. (C) The expression of HNF-4a was seen almost exclusively in glands with MUC2-positive intestinal metaplasia (C). (D–F) are serial sections. (D) GTM; (E) MUC2; (F) HNF-4a. Note that the upper glands are metaplastic, while the lower ones are proper gastric glands. The proper gastric glands were (D) positive for GTM by HIK1083 antibody and (E) negative for MUC2. (F) These glands were negative for HNF-4a.
RESULTS HNF-4a expression in intestinal metaplasia Table 1 shows the relationship between HNF-4a expression, intestinal metaplasia, and immunohistochemical phenotypes of non-neoplastic lesions. Table 2 summarises the data in Table 1 in light of the MUC2 and MUC5AC expression. Twenty-two lesions exhibited intestinal metaplasia morphologically. Intestinal metaplasia glands were characterised by the scattered presence of MUC2-positive goblet cells; goblet cells were always positive for MUC2 antigen. Of the 22 intestinal metaplasia lesions, 15 were mixed with MUC5AC positive cells. Two lesions were positive for CD10 on the luminal surface and were also positive for MUC2 but negative for MUC5AC. Twentyfour lesions with no intestinal metaplasia morphologically were consistently negative for MUC2, and all were positive for MUC5AC. In these mucosa, 10 lesions included GMTpositive glands. Nuclear staining of HNF-4a was observed in all 22 lesions (100%) with intestinal metaplasia (Fig. 2). The expression of HNF-4a tended to be more intense in the lower portion of the IM glands (Fig. 2). In the 24 mucosa with no discernible IM, HNF-4a was positive in 4 lesions (16.7%) and negative in 20 lesions (83.3%). These lesions were immunohistochemically of the gastric-type because they were negative for MUC2 and positive for MUC5AC. In addition, HNF-4a was negative in all proper gastric glands where GTM was positive (Fig. 3). There was a significant association between HNF-4a expression and the presence of MUC2-positive goblet cells, which virtually
indicates an association between HNF-4a expression and intestinal metaplasia (see above description). The expression of HNF-4a was reverse-correlated with that of MUC5AC (Table 2, Fig. 3). HNF-4a expression in adenocarcinoma The immunohistochemical classification of 35 cases of adenocarcinoma by histological type is summarised in Table 3. G-type, GI-type, and I-type represented 42.9, 31.4 and 25.7% of the cases, respectively; they did not correlate with the histological type (p50.33), although there was a tendency for well- and poorly-differentiated adenocarcinoma cells to be equipped with I-type and G-type phenotypes, respectively. There was no N-type adenocarcinoma
TABLE 3 Histological type and mucin classification in a series of 35 gastric carcinomas G-type
GI-type
I-type
Histological type
n (%)
n (%)
n (%)
Well differentiated (n59) Moderately differentiated (n517) Poorly differentiated (n59) Total (n535)
3 (33.3) 6 (35.3)
2 (22.2) 7 (41.2)
4 (44.4) 4 (25.3)
6 (66.7) 15 (42.9)
2 (22.2) 11 (31.4)
1 (11.1) 9 (25.7)
p50.33. G-type, gastric type; GI-type, mixed gastric and intestinal type; I-type, intestinal type.
KOJIMA et al.
Pathology (2006), 38(6), December
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
552
Fig. 4 Representative immunohistochemical findings on gastric adenocarcinomas. (A–C) are H&E findings of G-type, GI-type and I-type adenocarcinomas, respectively (6200). All adenocarcinomas examined in the present study expressed HNF-4a. A strong expression of HNF-4a was frequently observed in I-type adenocarcinomas. (D–F) MUC2; (G–I) MUC5AC; (J–L) HNF-4a (6200).
(Table 3). HNF-4a was positive in the nuclei of most tumour cells (.80%) in all cases, regardless of whether they were G-type, GI-type or I-type (Fig. 4). We evaluated grades of nuclear staining for HNF-4a because the intensity of HNF-4a staining varied from case to case. The intense expression of HNF-4a was significantly more frequent in the I- and GI-types than in the G-type (Table 4) and was also frequent in cases of welldifferentiated adenocarcinomas (Table 5).
DISCUSSION A series of regulator gene expressions controls organogenesis. Some of the genes for gastrointestinal differentiation, such as Shh, Cdx1 and Cdx2, have been shown to be involved in the development of intestinal metaplasia.11,14,15 HNF-4a is a transcriptional factor that is implicated in the differentiation of the gastrointestinal tract during embryogenesis.21–23 HNF-4a transcript has been detected in the fore, mid, and hind guts in the developing mouse embryo.24
HNF-4a EXPRESSION IN STOMACH
TABLE 4 Intensity of HNF-4a staining in a series of 35 gastric carcinomas, according to the mucin phenotype Grade of HNF-4a
Type of adenocarcinoma G-type (n515) GI-type (n511) I-type (n59)
1+
2+
3+
n (%)
n (%)
n (%)
8 (53.3) 1 (9.1) 0 (0)
5 (33.3) 3 (27.3) 3 (33.3)
2 (13.3) 7 (63.6) 6 (66.6)
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
p50.010. G-type, gastric type; GI-type, mixed gastric and intestinal type; I-type, intestinal type.
In adult humans, the expression of HNF-4a is seen in the colon and the small intestine, but not in the stomach.27,28 As described previously, our immunohistochemical study confirmed that P1 promoter-driven HNF-4a was not detectable in the normal, proper gastric epithelium and that the expression of P1 promoter-driven HNF-4a was exclusively observed in intestinal metaplasia glands. The expression regulation of several intestine-specific genes, including intestinal fatty acid-binding protein36 and apolipoprotein B,37 involves HNF-4a. Some functions of intestinal absorptive cells, such as the lipid absorption function, are regulated by the transcription regulational function of HNF-4a.20 Thus, it is most likely that reexpressed HNF-4a could regulate some of the proper intestinal function in the intestinal metaplasia glands. Over 80% of nuclei in all cases of adenocarcinoma were positive for HNF-4a, regardless of G-, GI-, or I-type, although the intensity of HNF-4a staining varied from case to case. Interestingly, a significant tendency for the expression of HNF-4a to be more intense in I- and GItypes than in G-types was observed. More than half of the cases (53.3%) of adenocarcinomas in the G-type showed a very weak staining for HNF-4a (which was more faint than those of duodenal mucosa), while most of the I- and GItype adenocarcinomas exhibited intense stainings for HNF4a (which were equivalent or more intense than those of duodenal mucosa). Thus, the intense HNF-4a staining was associated with the intestinal phenotype of adenocarcinoma cells. However, we did not make a specific correlation between intestinal immunophenotype and HNF-4a expression within a particular tumour. A stronger expression of
TABLE 5 Intensity of HNF-4a staining in a series of 35 gastric carcinomas, according to the histological classification Grade of HNF-4a
Histological type Well differentiated (n59) Moderately differentiated (n517) Poorly differentiated (n59) p50.027.
1+
2+
3+
n (%)
n (%)
n (%)
0 (0) 6 (35.3)
1 (11.1) 6 (35.3)
8 (88.9) 5 (29.4)
3 (33.3)
4 (44.4)
2 (22.2)
553
HNF-4a was observed in MUC2-positive areas than in MUC2-negative areas in several GI-type adenocarcinomas, but it was difficult to evaluate the co-localisation of MUC2 and HNF-4a in particular GI type adenocarcinomas because MUC2-positive cells and MUC5AC-positive cells were intermingled in the same cancer nest. Tanaka et al.28 have reported that 57.1% of gastric adenocarcinomas were positive for P1 promoter-driven HNF-4a and 100% for P1 promoter-driven HNF-4a. The discrepancy between the results might come from the condition for immnunohistochemistry or the criterion for identifying positive cells. We judged cells to be positively stained, even if they showed pale staining. Note that most cases classified as grade 1+, whose expression of P1 promoter-driven HNF-4a was generally pale (Fig. 1), were adenocarcinoma of G-type (8 cases of G-type and 1 of GItype). The pivotal role of HNF-4a in intestinal development suggests a possibility that the establishment and/or maintenance of the intestinal phenotype somehow correlates with the function of HNF-4a. mRNA levels of HNF-4 were increased in the perinatal small intestine38 and activated intestinal alkaline phosphatase promoter.39 It has been demonstrated in vivo in Drosophila that intestinal development requires HNF-4a expression.40 Of the differentiated adenocarcinomas, 24–49% expressed the intestinal-type mucin.34,41 In the present study, 20 cases (76.9%) of 26 of the differentiated adenocarcinomas, which correspond to well- and moderately-differentiated types in our classification, expressed P1 promoter-driven HNF-4a. The positive rate of P1 promoter-driven HNF-4a in the present study is higher than that of intestinal-type mucin in differentiated adenocarcinoma. However, there would be no conflict if HNF-4a expression could initiate the intestinalisation of adenocarcinomas. The functional significance of HNF-4a in tumour biology has been suggested. The gradual loss of HNF-4a expression is observed in the progression of hepatocellular carcinoma;42 it is related to renal cell carcinogenesis.43 Indeed, the over-expression of P1 promoter-driven HNF4a can activate the gene expression of p21/WAF1/CIP1, a cyclin-dependent kinase inhibitor, and inhibit growth of tumour cells.44 The role of HNF-4a in the progression of gastric adenocarcinomas is of interest because all of the adenocarcinomas that we examined expressed HNF-4a. In the present study, we note the significant correlation between the expression of P1 promoter-driven HNF-4a and the intestinal phenotype, both in gastric mucosa and adenocarcinomas. In addition, a high expression was observed in the differentiated type of adenocarcinomas. Our results suggest that the high expression of P1 promoter-driven HNF-4a may be a useful marker for the intestinal differentiation of gastric adenocarcinomas. The possible involvement of HNF-4a in the intestinalisation of gastric adenocarcinoma is expected to be established in the future. ACKNOWLEDGEMENTS The authors would like to acknowledge T. Matsui and K. Kawashima for excellent technical assistance. Address for correspondence: Dr T. Kishimoto, Department of Molecular Pathology, Graduate School of Medicine, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8670, Japan. E-mail: [email protected]
554
KOJIMA et al.
Pathology (2006), 38(6), December
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
References 1. Ho SB, Niehans GA, Lyftogt C, et al. Heterogeneity of mucin gene expression in normal and neoplastic tissues. Cancer Res 1993; 53: 641–51. 2. Reis CA, David L, Nielsen PA, et al. Immunohistochemical study of MUC5AC expression in human gastric carcinomas using a novel monoclonal antibody. Int J Cancer 1997; 74: 112–21. 3. Lauren P. The two histological main types of gastric carcinoma: Diffuse and so-called intestinal type carcinoma. Acta Pathol Microbiol Scand 1965; 64: 31–45. 4. Correa P. Human gastric carcinogenesis: a multistep and multifactorial process. First American Cancer Society Award Lecture on Cancer Epidemiology and Prevention. Cancer Res 1992; 52: 6735–40. 5. Uemura N, Okamoto S, Yamamoto S, et al. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001; 345: 784–9. 6. Conchillo JM, Houben G, de Bruine A, et al. Is type III intestinal metaplasia an obligatory precancerous lesion in intestinal-type gastric carcinoma? Eur J Cancer Prev 2001; 10: 307–12. 7. Egashira Y, Shimoda T, Ikegami M. Mucin histochemical analysis of minute gastric differentiated adenocarcinoma. Pathol Int 1999; 49: 55–61. 8. Saito A, Shimoda T, Nakanishi Y, et al. Histologic heterogeneity and mucin phenotypic expression in early gastric cancer. Pathol Int 2001; 51: 165–71. 9. Kawachi H, Takizawa T, Eishi Y, et al. Absence of either gastric or intestinal phenotype in microscopic differentiated gastric carcinomas. J Pathol 2003; 199: 436–46. 10. Stepan V, Ramamoorthy S, Nitsche H, et al. Regulation and function of the sonic hedgehog signal transduction pathway in isolated gastric parietal cells. J Biol Chem 2005; 280: 15700–8. 11. Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signals regulate multiple aspects of gastrointestinal development. Development 2000; 127: 2763–72. 12. Subramanian V, Meyer B, Evans GS. The murine Cdx1 gene product localises to the proliferative compartment in the developing and regenerating intestinal epithelium. Differentiation 1998; 64: 11–8. 13. Suh E, Traber PG. An intestine-specific homeobox gene regulates proliferation and differentiation. Mol Cell Biol 1996; 16: 619–25. 14. Mutoh H, Sakurai S, Satoh K, et al. Cdx1 induced intestinal metaplasia in the transgenic mouse stomach: comparative study with Cdx2 transgenic mice. Gut 2004; 53: 1416–23. 15. Mutoh H, Hakamata Y, Sato K, et al. Conversion of gastric mucosa to intestinal metaplasia in Cdx2-expressing transgenic mice. Biochem Biophys Res Commun 2002; 294: 470–9. 16. Mutoh H, Sakurai S, Satoh K, et al. Development of gastric carcinoma from intestinal metaplasia in Cdx2-transgenic mice. Cancer Res 2004; 64: 7740–7. 17. Sladek FM, Zhong WM, Lai E, et al. Liver-enriched transcription factor HNF-4 is a novel member of the steroid hormone receptor superfamily. Genes Dev 1990; 4: 2353–65. 18. Parviz F, Matullo C, Garrison WD, et al. Hepatocyte nuclear factor 4alpha controls the development of a hepatic epithelium and liver morphogenesis. Nat Genet 2003; 34: 292–6. 19. Odom DT, Zizlsperger N, Gordon DB, et al. Control of pancreas and liver gene expression by HNF transcription factors. Science 2004; 303: 1378–81. 20. Carriere V, Vidal R, Lazou K, et al. HNF-4-dependent induction of apolipoprotein A-IV gene transcription by an apical supply of lipid micelles in intestinal cells. J Biol Chem 2005; 280: 5406–13. 21. Morrisey EE, Tang Z, Sigrist K, et al. GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev 1998; 12: 3579–90. 22. Duncan SA, Nagy A, Chan W. Murine gastrulation requires HNF-4 regulated gene expression in the visceral endoderm: tetraploid rescue of Hnf-4(2/2) embryos. Development 1997; 124: 279–87. 23. Duncan SA, Manova K, Chen WS, et al. Expression of transcription factor HNF-4 in the extraembryonic endoderm, gut, and nephrogenic tissue of the developing mouse embryo: HNF-4 is a marker for primary
24. 25.
26. 27. 28. 29. 30. 31.
32.
33.
34. 35. 36.
37. 38. 39.
40. 41. 42. 43. 44.
endoderm in the implanting blastocyst. Proc Natl Acad Sci USA 1994; 91: 7598–02. Taraviras S, Monaghan AP, Schutz G, et al. Characterization of the mouse HNF-4 gene and its expression during mouse embryogenesis. Mech Dev 1994; 48: 67–79. Torres-Padilla ME, Fougere-Deschatrette C, Weiss MC. Expression of HNF4alpha isoforms in mouse liver development is regulated by sequential promoter usage and constitutive 39 end splicing. Mech Dev 2001; 109: 183–93. Drewes T, Senkel S, Holewa B, et al. Human hepatocyte nuclear factor 4 isoforms are encoded by distinct and differentially expressed genes. Mol Cell Biol 1996; 16: 925–31. Jiang S, Tanaka T, Iwanari H, et al. Expression and localization of P1 promoter-driven hepatocyte nuclear factor-4alpha (HNF4alpha) isoforms in human and rats. Nucl Recept 2003; 1: 5–16. Tanaka T, Jiang S, Hotta H, et al. Dysregulated expression of P1 and P2 promoter-driven hepatocyte nuclear factor-4alpha in the pathogenesis of human cancer. J Pathol 2006; 208: 662–72. Japanese Gastric Cancer Association, Japanese Classification of Gastric Carcinoma. 13th ed. Japan: Kinbara-Syuppan, 1999. Winterford CM, Walsh MD, Leggett BA, et al. Ultrastructural localization of epithelial mucin core proteins in colorectal tissues. J Histochem Cytochem 1999; 47: 1063–74. Trejdosiewicz LK, Malizia G, Oakes J, et al. Expression of the common acute lymphoblastic leukemia antigen (CALLA gp100) in the brush border of normal jejunum and jejunum of patients with coeliac disease. J Clin Pathol 1985; 38: 1002–6. Machado JC, Nogueira AM, Carneiro F, et al. Gastric carcinoma exhibits distinct types of cell differentiation: an immunohistochemical study of trefoil peptides (TFF1 and TFF2) and mucins (MUC1, MUC2, MUC5AC, and MUC6). J Pathol 2000; 190: 437–43. Nakamura N, Ota H, Katsuyama T, et al. Histochemical reactivity of normal, metaplastic, and neoplastic tissues to alpha-linked Nacetylglucosamine residue-specific monoclonal antibody HIK1083. J Histochem Cytochem 1998; 46: 793–801. Saito A, Shimoda T, Nakanishi Y, et al. Histologic heterogeneity and mucin phenotypic expression in early gastric cancer. Pathol Int 2001; 51: 165–71. Kawachi H, Takizawa T, Eishi Y, et al. Absence of either gastric or intestinal phenotype in microscopic differentiated gastric carcinomas. J Pathol 2003; 199: 436–46. Rottman JN, Gordon JI. Comparison of the patterns of expression of rat intestinal fatty acid binding protein/human growth hormone fusion genes in cultured intestinal epithelial cell lines and in the gut epithelium of transgenic mice. J Biol Chem 1993; 268: 11994–2002. Antes TJ, Levy-Wilson B. HNF-3 beta, C/EBP beta, and HNF-4 act in synergy to enhance transcription of the human apolipoprotein B gene in intestinal cells. DNA Cell Biol 2001; 20: 67–74. Ogura Y, Suruga K, Takase S, et al. Developmental changes of the expression of the genes regulated by retinoic acid in the small intestine of rats. Life Sci 2005; 77: 2804–13. Olsen L, Bressendorff S, Troelsen JT, et al. Differentiation-dependent activation of the human intestinal alkaline phosphatase promoter by HNF-4 in intestinal cells. Am J Physiol Gastrointest Liver Physiol 2005; 289: G220–6. Zhong W, Sladek FM, Darnell JE Jr. The expression pattern of a Drosophila homolog to the mouse transcription factor HNF-4 suggests a determinative role in gut formation. EMBO J 1993; 12: 537–44. Koseki K, Takizawa T, Koike M, et al. Distinction of differentiated type early gastric carcinoma with gastric type mucin expression. Cancer 2000; 89: 724–32. Lazarevich NL, Cheremnova OA, Varga EV, et al. Progression of HCC in mice is associated with a downregulation in the expression of hepatocyte nuclear factors. Hepatology 2004; 39: 1038–47. Sel S, Ebert T, Ryffel GU, et al. Human renal cell carcinogenesis is accompanied by a coordinate loss of the tissue specific transcription factors HNF4 alpha and HNF1 alpha. Cancer Lett 1996; 101: 205–10. Chiba H, Itoh T, Satohisa S, et al. Activation of p21CIP1/WAF1 gene expression and inhibition of cell proliferation by overexpression of hepatocyte nuclear factor-4alpha. Exp Cell Res 2005; 302: 11–21.
ANATOMICAL PATHOLOGY
The expression of hepatocyte nuclear factor-4a, a developmental regulator of visceral endoderm, correlates with the intestinal phenotype of gastric adenocarcinomas KAZUHIRO KOJIMA*{, TAKASHI KISHIMOTO*, YUICHIRO NAGAI*, TOHRU TANIZAWA{, YUKIO NAKATANI{, MASARU MIYAZAKI{ AND HIROSHI ISHIKURA*
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
Departments of *Molecular Pathology, {General Surgery, and {Diagnostic Pathology, Chiba University Graduate School of Medicine, Chiba, Japan
Summary Aims: Hepatocyte nuclear factor (HNF)-4a is a developmental regulator of the visceral endoderm, which is expressed in the embryonic gut and later in the adult intestine and colon. However, adult gastric mucosa does not express HNF-4a. We investigated the possible involvement of HNF-4a in intestinal metaplasia and the intestinalisation of gastric adenocarcinomas. Methods: Thirty-five cases of adenocarcinomas and 46 cases of adjacent non-neoplastic mucosa with (22 lesions) or without (24 lesions) intestinal metaplasia were immunostained for HNF-4a. The gastric or the intestinal phenotype was also examined using immunohistochemistry for MUC5AC, MUC2, CD10, and gastric-type mucin (GTM). Adenocarcinomas were classified into the gastric type (Gtype, 42.9%), the mixed gastric and intestinal type (GI-type, 31.4%), and the intestinal type (I-type, 25.7%). Results: The HNF-4a expression was exclusively seen in glandular cells with intestinal metaplasia, which was correlated with MUC2 expression (p,0.05) and inversely correlated with MUC5AC expression (p,0.05). All adenocarcinomas more or less expressed HNF-4a, with an intense expression being seen in the I-type (p,0.01) and in welldifferentiated adenocarcinomas (p,0.03). Conclusions: HNF-4a expression is associated with the intestinal phenotype of non-neoplastic and neoplastic gastric glandular cells, suggesting a possible involvement in the establishment and/or maintenance of the intestinal phenotype of the gastric mucosa and adenocarcinomas. Key words: Hepatocyte nuclear factor-4a, intestinal metaplasia, adenocarcinoma, MUC2, MUC5AC, immunohistochemistry. Received 30 May, revised 5 July, accepted 5 September 2006
INTRODUCTION Intestinal metaplasia of the gastric mucosa are defined as the ectopic occurrence of intestinal-type epithelial cells such as goblet, absorptive, and Paneth cells in the adult stomach. It is not so difficult to detect intestinal metaplasia lesions morphologically on H&E sections, and periodic-acid Schiff (PAS) or alcian blue stain is useful in recognising goblet cells. Recently, expressions of mucin core protein genes have been estimated to be in the normal gastrointestinal
tract. Among these mucins, MUC5AC and MUC2 are widely used as helpful markers for the gastric or intestinal phenotype of gastric epithelial cells,1,2 respectively. Gastric carcinomas have been classified by Lauren3 into two major histological groups: the intestinal type and the diffuse type. This classification divides tumours basically by the degree of glandular formation. On the other hand, the classification that is based on the epithelial mucin phenotype (i.e., the intestinal phenotype or gastric phenotype) is especially meaningful for the study of the histogenesis of gastric adenocarcinomas. A sequence from intestinal metaplasia through dysplasia to adenocarcinoma, which is called Correa’s cascade, has been proposed.4 This sequence has been supported by epidemiological findings that there is an increased risk of cancer development in patients with intestinal metaplasia,5,6 but these findings are still controversial.7,8 Mucin phenotypes between microscopic adenocarcinoma and its surrounding mucosa are not always identical, suggesting that intestinal metaplasia is not a necessary precancerous condition.9 Although the molecular mechanisms of phenotypical regulation in gastric mucosa or gastric adenocarcinoma have yet to be clearly established, the importance of several signal pathways and transcriptional regulation has become evident. These involve molecules that have a pivotal role in the morphogenesis of the organs of endodermal origin, such as small intestine, colon, and stomach. Sonic hedgehog (Shh) is expressed in gastric parietal cells in adult humans,10 and Shh-deficient mice generate intestinal metaplasia in their stomachs.11 Caudal-type homeobox transcriptional factor (Cdx)1 and Cdx2 are expressed in the glandular cells of the small intestine and colon, where they play important roles in the intestinal phenotype.12,13 These transcriptional factors play a key role also in the development of intestinal metaplasia;14,15 direct evidence has been demonstrated by Cdx2-transgenic mice which developed intestinal-type adenocarcinoma from intestinal metaplasia.16 Hepatocyte nuclear factor (HNF)-4a, a hormone receptor-type transcriptional factor,17 contributes a variety of gene expressions in many cell functions, such as liver morphogenesis,18 glucose homeostasis,19 and lipid absorption.20 HNF-4a is considered to be a developmental regulator of the visceral endoderm.21–23 In embryogenesis, the expression of HNF-4a mRNA first appears in the whole primitive endoderm; it is later restricted to the
ISSN 0031-3025 printed/ISSN 1465-3931 # 2006 Royal College of Pathologists of Australasia DOI: 10.1080/00313020601024011
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
HNF-4a EXPRESSION IN STOMACH
visceral endoderm, followed by expression in the embryonic liver, kidney, pancreas, and gut.24 HNF-4a has nine isoforms that are generated by alternative promoter (P1 and P2) usage and alternative splicing.25,26 Six isoforms, HNF4a1–a6, are generated by the usage of the P1 promoter, and three isoforms, HNF4a7– a9, by the P2 promoter. The P1 and P2 promoter-driven HNF-4a is distributed differently in adults; the expression of P1 promoter-driven HNF-4a is observed in the absorptive epithelium of the small intestine, crypt of colon, renal tubules, and pancreas. Interestingly, P1 promoter-driven HNF-4a cannot be detected in the normal gastric mucosa,27,28 although the stomach is of endodermal origin. Also interestingly, as Tanaka et al. have reported, restricted expression of P1 promoter-driven HNF-4a is observed in intestinal metaplasia in the adult gastric mucosa and in 57.1% of gastric adenocarcinomas.28 Their data suggest the possibility that HNF-4a, a developmental regulate of visceral endoderm, contributes to the intestinalisation or the maintenance of the intestinal phenotype of gastric mucosa. However, the correlation between P1 promoterdriven HNF-4a expression and the intestinal phenotype of the gastric mucosa or gastric adenocarcinoma has not been analysed in detail. Therefore, we examined the expression of P1 promoter-driven HNF-4a in adenocarcinoma tissues to test whether its expression is correlated with the intestinal phenotype of gastric adenocarcinomas.
549
diagnosis of carcinoma was based on the criteria of the Japanese Classification of Gastric Carcinoma.29 There were 26 cases of tubular adenocarcinomas (9 of the well-differentiated type and 17 of the moderately-differentiated type), and nine cases of poorly-differentiated adenocarcinoma. As defined according to these criteria, tubular and poorly-differentiated adenocarcinomas roughly correspond to Lauen’s classification of the ‘intestinal type’ and the ‘diffuse type’, respectively.1 From formalin-fixed, paraffin-embedded tissues, 4 mm thick sections were used for H&E stain and immunohistochemical analysis. We evaluated a series of 46 non-neoplastic lesions in the vicinity of the 35 cases of adenocarcinomas. The lesions included gastric mucosa with and without intestinal metaplasia, which were 22 and 24 lesions, respectively. Immunohistochemistry Immunohistochemical study was performed using primary antibodies against HNF-4a (K9218, 1:100; Perseus Proteomics, Japan), MUC2 (1:100; Medac Diagnostika, Germany), MUC5AC (1:50; Novocastra, UK), a gastric-type mucin (GTM; HIK1083, 1:25; Kanto Chemical, Japan), and CD10 (1:80; Novocastra). Anti-HNF-4a antibody, K9218, recognises HNF-4a1, 2, 3, 4, 5, and 6 isoforms, which are derived by P1 promoter usage. The immunoperoxidase method with a LSAB kit (Dako Cytomation, Japan) was used. Each section was deparaffinised by xylene and rehydrated in 100% ethanol. Endogenous peroxidase activity was blocked by 3% H2O2 for 5 min, and slides were washed in phosphate-buffered saline (PBS) three times for 5 min. All sections, except for HIK1083, were treated with microwaves for 20 min in a 10 mM sodium citrate buffer (pH 6.0). After blocking, the sections were incubated with each primary antibody overnight at 4uC. The slides were washed in PBS three times for 5 min. They were incubated with biotinylated secondary antibodies and then with avidinhorseradish peroxidase according to the manufacturer’s protocol. After they were washed in PBS, diaminobenzidine tetrahydrochloride (DAB) substrate was used for visualisation, and haematoxylin was used for the counterstain.
MATERIALS AND METHODS Samples
Classification of gastric adenocarcinomas by immunohistochemistry
Included in this study were 35 cases of surgically-resected gastric adenocarcinoma from the file of Chiba University Hospital, Japan. The
We used antibodies for MUC2, CD10, MUC5AC and GTM for the classification of adenocarcinomas in relation to the intestinal phenotype.
Fig. 1 Grade of HNF-4a expression by immunohistochemistry. (A) HNF-4a expression in the duodenal mucosa; (B) 1+, tumour cells were less positive than the intensity of the duodenal mucosa; (C) 2+, tumour cells were as equally positive as the intensity of the duodenal mucosa; (D) 3+, tumour cells were more positive than the intensity of the duodenal mucosa (immunostain for HNF-4a, 6200).
550
KOJIMA et al.
Pathology (2006), 38(6), December
TABLE 1 Expression of HNF-4a in a series of 46 non-neoplastic gastric mucosa
IM
MUC2
CD10
MUC5AC
GTM
(+) (+) (+) (2) (2)
(+) (+) (+) (2) (2)
(+) (2) (2) (2) (2)
(2) (2) (+) (+) (+)
(2) (2) (2) (2) (+)
HNF-4a (2)
NHF-4a (+)
n (%)
n (%)
0 0 0 10 10
(0) (0) (0) (71.4) (100)
2 5 15 4 0
Total
(100) (100) (100) (28.6) (0)
2 5 15 14 10
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
IM, intestinal metaplasia; GTM, gastric-type mucin reactive with HIK1083 antibody. MUC2 is a gel-forming mucin that is specifically expressed in the goblet cells.30 CD10, a human membrane-associated neutral peptidase, is prominently expressed in the brush border of intestinal absorptive cells.31 MUC5AC is a gel-forming mucin that is specifically expressed in the foveolar epithelium of gastric mucosa and mucous neck cells.32 HIK1083 recognises gastric-type mucin expressed in cardiac gland cells, mucous cells and pyloric gland cells.33 In accordance with the definition used by Saito et al.34 and Kawachi et al.,35 we classified the adenocarcinomas into four types: (1) the gastric type (G-type), in which more than 10% of cancer cells are positive for MUC5AC and/or GTM and negative for both MUC2 and CD10; (2) the intestinal type (I-type), in which more than 10% of the cancer cells are positive for MUC2 and/or CD10 and negative for both MUC5AC and GTM; (3) the mixed gastric and intestinal type (GI-type), in which more than 10% of the cancer cells are positive for MUC5AC and/or GTM and also more than 10% of the cancer cells are positive for MUC2 and/or CD10; (4) the negative type (N-type), in which MUC2, CD10, MUC5AC, and GTM are all negative.
TABLE 2 Expression of HNF-4a in a series of 46 non-neoplastic lesions, according to the expression of MUC2 and MUC5AC
MUC2 expression (n) Positive (22) Negative (24) MUC5AC expression (n) Positive (39) Negative (7)
HNF-4a (2)
HNF-4a (+)
n (%)
n (%)
p value
0 (0) 20 (83.3)
22 (100) 4 (16.7)
0.045
20 (51.3) 0 (0)
19 (48.7) 7 (100)
0.011
1+, the intensity was less than that of duodenal mucosa; 2+, the intensity was equivalent to that of duodenal mucosa; 3+, the intensity was greater than that of duodenal mucosa (Fig. 1).
Grading of HNF-4a expression
Statistical analysis
Positive staining of HNF-4a was graded as 1+, 2+, or 3+ on the basis of comparisons of the intensity of HNF-4a staining of the duodenal mucosa:
Statistical analysis of the results was carried out using the chi-square test. Values of p of less than 0.05 were considered statistically significant.
Fig. 2 The expression of HNF-4a in intestinal metaplasia. (A) H&E stain (6100). (B) Immunohistochemistry for HNF-4a on the serial section of (A) (6100). (C) Immunohistochemistry for HNF-4a exhibited an increased expression of HNF-4a at the bottom of the metaplastic glands (6200).
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
HNF-4a EXPRESSION IN STOMACH
551
Fig. 3 Representative immunohistochemical findings on gastric mucosa with intestinal metaplasia (6100). (A–C) are serial sections. (A) MUC5AC; (B) MUC2; (C) HNF-4a. Note the almost mirror image between (A) MUC5AC and (B) MUC2 expression. (C) The expression of HNF-4a was seen almost exclusively in glands with MUC2-positive intestinal metaplasia (C). (D–F) are serial sections. (D) GTM; (E) MUC2; (F) HNF-4a. Note that the upper glands are metaplastic, while the lower ones are proper gastric glands. The proper gastric glands were (D) positive for GTM by HIK1083 antibody and (E) negative for MUC2. (F) These glands were negative for HNF-4a.
RESULTS HNF-4a expression in intestinal metaplasia Table 1 shows the relationship between HNF-4a expression, intestinal metaplasia, and immunohistochemical phenotypes of non-neoplastic lesions. Table 2 summarises the data in Table 1 in light of the MUC2 and MUC5AC expression. Twenty-two lesions exhibited intestinal metaplasia morphologically. Intestinal metaplasia glands were characterised by the scattered presence of MUC2-positive goblet cells; goblet cells were always positive for MUC2 antigen. Of the 22 intestinal metaplasia lesions, 15 were mixed with MUC5AC positive cells. Two lesions were positive for CD10 on the luminal surface and were also positive for MUC2 but negative for MUC5AC. Twentyfour lesions with no intestinal metaplasia morphologically were consistently negative for MUC2, and all were positive for MUC5AC. In these mucosa, 10 lesions included GMTpositive glands. Nuclear staining of HNF-4a was observed in all 22 lesions (100%) with intestinal metaplasia (Fig. 2). The expression of HNF-4a tended to be more intense in the lower portion of the IM glands (Fig. 2). In the 24 mucosa with no discernible IM, HNF-4a was positive in 4 lesions (16.7%) and negative in 20 lesions (83.3%). These lesions were immunohistochemically of the gastric-type because they were negative for MUC2 and positive for MUC5AC. In addition, HNF-4a was negative in all proper gastric glands where GTM was positive (Fig. 3). There was a significant association between HNF-4a expression and the presence of MUC2-positive goblet cells, which virtually
indicates an association between HNF-4a expression and intestinal metaplasia (see above description). The expression of HNF-4a was reverse-correlated with that of MUC5AC (Table 2, Fig. 3). HNF-4a expression in adenocarcinoma The immunohistochemical classification of 35 cases of adenocarcinoma by histological type is summarised in Table 3. G-type, GI-type, and I-type represented 42.9, 31.4 and 25.7% of the cases, respectively; they did not correlate with the histological type (p50.33), although there was a tendency for well- and poorly-differentiated adenocarcinoma cells to be equipped with I-type and G-type phenotypes, respectively. There was no N-type adenocarcinoma
TABLE 3 Histological type and mucin classification in a series of 35 gastric carcinomas G-type
GI-type
I-type
Histological type
n (%)
n (%)
n (%)
Well differentiated (n59) Moderately differentiated (n517) Poorly differentiated (n59) Total (n535)
3 (33.3) 6 (35.3)
2 (22.2) 7 (41.2)
4 (44.4) 4 (25.3)
6 (66.7) 15 (42.9)
2 (22.2) 11 (31.4)
1 (11.1) 9 (25.7)
p50.33. G-type, gastric type; GI-type, mixed gastric and intestinal type; I-type, intestinal type.
KOJIMA et al.
Pathology (2006), 38(6), December
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
552
Fig. 4 Representative immunohistochemical findings on gastric adenocarcinomas. (A–C) are H&E findings of G-type, GI-type and I-type adenocarcinomas, respectively (6200). All adenocarcinomas examined in the present study expressed HNF-4a. A strong expression of HNF-4a was frequently observed in I-type adenocarcinomas. (D–F) MUC2; (G–I) MUC5AC; (J–L) HNF-4a (6200).
(Table 3). HNF-4a was positive in the nuclei of most tumour cells (.80%) in all cases, regardless of whether they were G-type, GI-type or I-type (Fig. 4). We evaluated grades of nuclear staining for HNF-4a because the intensity of HNF-4a staining varied from case to case. The intense expression of HNF-4a was significantly more frequent in the I- and GI-types than in the G-type (Table 4) and was also frequent in cases of welldifferentiated adenocarcinomas (Table 5).
DISCUSSION A series of regulator gene expressions controls organogenesis. Some of the genes for gastrointestinal differentiation, such as Shh, Cdx1 and Cdx2, have been shown to be involved in the development of intestinal metaplasia.11,14,15 HNF-4a is a transcriptional factor that is implicated in the differentiation of the gastrointestinal tract during embryogenesis.21–23 HNF-4a transcript has been detected in the fore, mid, and hind guts in the developing mouse embryo.24
HNF-4a EXPRESSION IN STOMACH
TABLE 4 Intensity of HNF-4a staining in a series of 35 gastric carcinomas, according to the mucin phenotype Grade of HNF-4a
Type of adenocarcinoma G-type (n515) GI-type (n511) I-type (n59)
1+
2+
3+
n (%)
n (%)
n (%)
8 (53.3) 1 (9.1) 0 (0)
5 (33.3) 3 (27.3) 3 (33.3)
2 (13.3) 7 (63.6) 6 (66.6)
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
p50.010. G-type, gastric type; GI-type, mixed gastric and intestinal type; I-type, intestinal type.
In adult humans, the expression of HNF-4a is seen in the colon and the small intestine, but not in the stomach.27,28 As described previously, our immunohistochemical study confirmed that P1 promoter-driven HNF-4a was not detectable in the normal, proper gastric epithelium and that the expression of P1 promoter-driven HNF-4a was exclusively observed in intestinal metaplasia glands. The expression regulation of several intestine-specific genes, including intestinal fatty acid-binding protein36 and apolipoprotein B,37 involves HNF-4a. Some functions of intestinal absorptive cells, such as the lipid absorption function, are regulated by the transcription regulational function of HNF-4a.20 Thus, it is most likely that reexpressed HNF-4a could regulate some of the proper intestinal function in the intestinal metaplasia glands. Over 80% of nuclei in all cases of adenocarcinoma were positive for HNF-4a, regardless of G-, GI-, or I-type, although the intensity of HNF-4a staining varied from case to case. Interestingly, a significant tendency for the expression of HNF-4a to be more intense in I- and GItypes than in G-types was observed. More than half of the cases (53.3%) of adenocarcinomas in the G-type showed a very weak staining for HNF-4a (which was more faint than those of duodenal mucosa), while most of the I- and GItype adenocarcinomas exhibited intense stainings for HNF4a (which were equivalent or more intense than those of duodenal mucosa). Thus, the intense HNF-4a staining was associated with the intestinal phenotype of adenocarcinoma cells. However, we did not make a specific correlation between intestinal immunophenotype and HNF-4a expression within a particular tumour. A stronger expression of
TABLE 5 Intensity of HNF-4a staining in a series of 35 gastric carcinomas, according to the histological classification Grade of HNF-4a
Histological type Well differentiated (n59) Moderately differentiated (n517) Poorly differentiated (n59) p50.027.
1+
2+
3+
n (%)
n (%)
n (%)
0 (0) 6 (35.3)
1 (11.1) 6 (35.3)
8 (88.9) 5 (29.4)
3 (33.3)
4 (44.4)
2 (22.2)
553
HNF-4a was observed in MUC2-positive areas than in MUC2-negative areas in several GI-type adenocarcinomas, but it was difficult to evaluate the co-localisation of MUC2 and HNF-4a in particular GI type adenocarcinomas because MUC2-positive cells and MUC5AC-positive cells were intermingled in the same cancer nest. Tanaka et al.28 have reported that 57.1% of gastric adenocarcinomas were positive for P1 promoter-driven HNF-4a and 100% for P1 promoter-driven HNF-4a. The discrepancy between the results might come from the condition for immnunohistochemistry or the criterion for identifying positive cells. We judged cells to be positively stained, even if they showed pale staining. Note that most cases classified as grade 1+, whose expression of P1 promoter-driven HNF-4a was generally pale (Fig. 1), were adenocarcinoma of G-type (8 cases of G-type and 1 of GItype). The pivotal role of HNF-4a in intestinal development suggests a possibility that the establishment and/or maintenance of the intestinal phenotype somehow correlates with the function of HNF-4a. mRNA levels of HNF-4 were increased in the perinatal small intestine38 and activated intestinal alkaline phosphatase promoter.39 It has been demonstrated in vivo in Drosophila that intestinal development requires HNF-4a expression.40 Of the differentiated adenocarcinomas, 24–49% expressed the intestinal-type mucin.34,41 In the present study, 20 cases (76.9%) of 26 of the differentiated adenocarcinomas, which correspond to well- and moderately-differentiated types in our classification, expressed P1 promoter-driven HNF-4a. The positive rate of P1 promoter-driven HNF-4a in the present study is higher than that of intestinal-type mucin in differentiated adenocarcinoma. However, there would be no conflict if HNF-4a expression could initiate the intestinalisation of adenocarcinomas. The functional significance of HNF-4a in tumour biology has been suggested. The gradual loss of HNF-4a expression is observed in the progression of hepatocellular carcinoma;42 it is related to renal cell carcinogenesis.43 Indeed, the over-expression of P1 promoter-driven HNF4a can activate the gene expression of p21/WAF1/CIP1, a cyclin-dependent kinase inhibitor, and inhibit growth of tumour cells.44 The role of HNF-4a in the progression of gastric adenocarcinomas is of interest because all of the adenocarcinomas that we examined expressed HNF-4a. In the present study, we note the significant correlation between the expression of P1 promoter-driven HNF-4a and the intestinal phenotype, both in gastric mucosa and adenocarcinomas. In addition, a high expression was observed in the differentiated type of adenocarcinomas. Our results suggest that the high expression of P1 promoter-driven HNF-4a may be a useful marker for the intestinal differentiation of gastric adenocarcinomas. The possible involvement of HNF-4a in the intestinalisation of gastric adenocarcinoma is expected to be established in the future. ACKNOWLEDGEMENTS The authors would like to acknowledge T. Matsui and K. Kawashima for excellent technical assistance. Address for correspondence: Dr T. Kishimoto, Department of Molecular Pathology, Graduate School of Medicine, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8670, Japan. E-mail: [email protected]
554
KOJIMA et al.
Pathology (2006), 38(6), December
Pathology Downloaded from informahealthcare.com by University Of Alabama Birmingham on 01/11/13 For personal use only.
References 1. Ho SB, Niehans GA, Lyftogt C, et al. Heterogeneity of mucin gene expression in normal and neoplastic tissues. Cancer Res 1993; 53: 641–51. 2. Reis CA, David L, Nielsen PA, et al. Immunohistochemical study of MUC5AC expression in human gastric carcinomas using a novel monoclonal antibody. Int J Cancer 1997; 74: 112–21. 3. Lauren P. The two histological main types of gastric carcinoma: Diffuse and so-called intestinal type carcinoma. Acta Pathol Microbiol Scand 1965; 64: 31–45. 4. Correa P. Human gastric carcinogenesis: a multistep and multifactorial process. First American Cancer Society Award Lecture on Cancer Epidemiology and Prevention. Cancer Res 1992; 52: 6735–40. 5. Uemura N, Okamoto S, Yamamoto S, et al. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001; 345: 784–9. 6. Conchillo JM, Houben G, de Bruine A, et al. Is type III intestinal metaplasia an obligatory precancerous lesion in intestinal-type gastric carcinoma? Eur J Cancer Prev 2001; 10: 307–12. 7. Egashira Y, Shimoda T, Ikegami M. Mucin histochemical analysis of minute gastric differentiated adenocarcinoma. Pathol Int 1999; 49: 55–61. 8. Saito A, Shimoda T, Nakanishi Y, et al. Histologic heterogeneity and mucin phenotypic expression in early gastric cancer. Pathol Int 2001; 51: 165–71. 9. Kawachi H, Takizawa T, Eishi Y, et al. Absence of either gastric or intestinal phenotype in microscopic differentiated gastric carcinomas. J Pathol 2003; 199: 436–46. 10. Stepan V, Ramamoorthy S, Nitsche H, et al. Regulation and function of the sonic hedgehog signal transduction pathway in isolated gastric parietal cells. J Biol Chem 2005; 280: 15700–8. 11. Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signals regulate multiple aspects of gastrointestinal development. Development 2000; 127: 2763–72. 12. Subramanian V, Meyer B, Evans GS. The murine Cdx1 gene product localises to the proliferative compartment in the developing and regenerating intestinal epithelium. Differentiation 1998; 64: 11–8. 13. Suh E, Traber PG. An intestine-specific homeobox gene regulates proliferation and differentiation. Mol Cell Biol 1996; 16: 619–25. 14. Mutoh H, Sakurai S, Satoh K, et al. Cdx1 induced intestinal metaplasia in the transgenic mouse stomach: comparative study with Cdx2 transgenic mice. Gut 2004; 53: 1416–23. 15. Mutoh H, Hakamata Y, Sato K, et al. Conversion of gastric mucosa to intestinal metaplasia in Cdx2-expressing transgenic mice. Biochem Biophys Res Commun 2002; 294: 470–9. 16. Mutoh H, Sakurai S, Satoh K, et al. Development of gastric carcinoma from intestinal metaplasia in Cdx2-transgenic mice. Cancer Res 2004; 64: 7740–7. 17. Sladek FM, Zhong WM, Lai E, et al. Liver-enriched transcription factor HNF-4 is a novel member of the steroid hormone receptor superfamily. Genes Dev 1990; 4: 2353–65. 18. Parviz F, Matullo C, Garrison WD, et al. Hepatocyte nuclear factor 4alpha controls the development of a hepatic epithelium and liver morphogenesis. Nat Genet 2003; 34: 292–6. 19. Odom DT, Zizlsperger N, Gordon DB, et al. Control of pancreas and liver gene expression by HNF transcription factors. Science 2004; 303: 1378–81. 20. Carriere V, Vidal R, Lazou K, et al. HNF-4-dependent induction of apolipoprotein A-IV gene transcription by an apical supply of lipid micelles in intestinal cells. J Biol Chem 2005; 280: 5406–13. 21. Morrisey EE, Tang Z, Sigrist K, et al. GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev 1998; 12: 3579–90. 22. Duncan SA, Nagy A, Chan W. Murine gastrulation requires HNF-4 regulated gene expression in the visceral endoderm: tetraploid rescue of Hnf-4(2/2) embryos. Development 1997; 124: 279–87. 23. Duncan SA, Manova K, Chen WS, et al. Expression of transcription factor HNF-4 in the extraembryonic endoderm, gut, and nephrogenic tissue of the developing mouse embryo: HNF-4 is a marker for primary
24. 25.
26. 27. 28. 29. 30. 31.
32.
33.
34. 35. 36.
37. 38. 39.
40. 41. 42. 43. 44.
endoderm in the implanting blastocyst. Proc Natl Acad Sci USA 1994; 91: 7598–02. Taraviras S, Monaghan AP, Schutz G, et al. Characterization of the mouse HNF-4 gene and its expression during mouse embryogenesis. Mech Dev 1994; 48: 67–79. Torres-Padilla ME, Fougere-Deschatrette C, Weiss MC. Expression of HNF4alpha isoforms in mouse liver development is regulated by sequential promoter usage and constitutive 39 end splicing. Mech Dev 2001; 109: 183–93. Drewes T, Senkel S, Holewa B, et al. Human hepatocyte nuclear factor 4 isoforms are encoded by distinct and differentially expressed genes. Mol Cell Biol 1996; 16: 925–31. Jiang S, Tanaka T, Iwanari H, et al. Expression and localization of P1 promoter-driven hepatocyte nuclear factor-4alpha (HNF4alpha) isoforms in human and rats. Nucl Recept 2003; 1: 5–16. Tanaka T, Jiang S, Hotta H, et al. Dysregulated expression of P1 and P2 promoter-driven hepatocyte nuclear factor-4alpha in the pathogenesis of human cancer. J Pathol 2006; 208: 662–72. Japanese Gastric Cancer Association, Japanese Classification of Gastric Carcinoma. 13th ed. Japan: Kinbara-Syuppan, 1999. Winterford CM, Walsh MD, Leggett BA, et al. Ultrastructural localization of epithelial mucin core proteins in colorectal tissues. J Histochem Cytochem 1999; 47: 1063–74. Trejdosiewicz LK, Malizia G, Oakes J, et al. Expression of the common acute lymphoblastic leukemia antigen (CALLA gp100) in the brush border of normal jejunum and jejunum of patients with coeliac disease. J Clin Pathol 1985; 38: 1002–6. Machado JC, Nogueira AM, Carneiro F, et al. Gastric carcinoma exhibits distinct types of cell differentiation: an immunohistochemical study of trefoil peptides (TFF1 and TFF2) and mucins (MUC1, MUC2, MUC5AC, and MUC6). J Pathol 2000; 190: 437–43. Nakamura N, Ota H, Katsuyama T, et al. Histochemical reactivity of normal, metaplastic, and neoplastic tissues to alpha-linked Nacetylglucosamine residue-specific monoclonal antibody HIK1083. J Histochem Cytochem 1998; 46: 793–801. Saito A, Shimoda T, Nakanishi Y, et al. Histologic heterogeneity and mucin phenotypic expression in early gastric cancer. Pathol Int 2001; 51: 165–71. Kawachi H, Takizawa T, Eishi Y, et al. Absence of either gastric or intestinal phenotype in microscopic differentiated gastric carcinomas. J Pathol 2003; 199: 436–46. Rottman JN, Gordon JI. Comparison of the patterns of expression of rat intestinal fatty acid binding protein/human growth hormone fusion genes in cultured intestinal epithelial cell lines and in the gut epithelium of transgenic mice. J Biol Chem 1993; 268: 11994–2002. Antes TJ, Levy-Wilson B. HNF-3 beta, C/EBP beta, and HNF-4 act in synergy to enhance transcription of the human apolipoprotein B gene in intestinal cells. DNA Cell Biol 2001; 20: 67–74. Ogura Y, Suruga K, Takase S, et al. Developmental changes of the expression of the genes regulated by retinoic acid in the small intestine of rats. Life Sci 2005; 77: 2804–13. Olsen L, Bressendorff S, Troelsen JT, et al. Differentiation-dependent activation of the human intestinal alkaline phosphatase promoter by HNF-4 in intestinal cells. Am J Physiol Gastrointest Liver Physiol 2005; 289: G220–6. Zhong W, Sladek FM, Darnell JE Jr. The expression pattern of a Drosophila homolog to the mouse transcription factor HNF-4 suggests a determinative role in gut formation. EMBO J 1993; 12: 537–44. Koseki K, Takizawa T, Koike M, et al. Distinction of differentiated type early gastric carcinoma with gastric type mucin expression. Cancer 2000; 89: 724–32. Lazarevich NL, Cheremnova OA, Varga EV, et al. Progression of HCC in mice is associated with a downregulation in the expression of hepatocyte nuclear factors. Hepatology 2004; 39: 1038–47. Sel S, Ebert T, Ryffel GU, et al. Human renal cell carcinogenesis is accompanied by a coordinate loss of the tissue specific transcription factors HNF4 alpha and HNF1 alpha. Cancer Lett 1996; 101: 205–10. Chiba H, Itoh T, Satohisa S, et al. Activation of p21CIP1/WAF1 gene expression and inhibition of cell proliferation by overexpression of hepatocyte nuclear factor-4alpha. Exp Cell Res 2005; 302: 11–21.