Adiponectin receptor expression in gastric carcinoma: implications in tumor development and progression
- First Online:
- Cite this article as:
- Shin, E., Park, D.J., Kim, HH. et al. J Cancer Res Clin Oncol (2013) 139: 709. doi:10.1007/s00432-013-1379-3
- 295 Downloads
Adiponectin, an adipocyte-secreted endogenous insulin sensitizer, appears to play an important role in progression of several malignancies. Expression of adiponectin receptors—AdipoR1 and AdipoR2—has been documented in gastric cancer (GC) cell lines, but its role in GCs is still controversial. We investigated expression level of 2 adiponectin receptors and correlated their expression with prognosis in GC patients.
We immunohistochemically evaluated AdipoR1 and AdipoR2 expression in 59 non-neoplastic gastric mucosas, 48 gastric adenomas, 250 GCs, and 58 lymph nodes involved by metastatic GC and assessed its association with clinicopathologic characteristics.
Expression rates of both receptors increased stepwise in non-neoplastic gastric mucosa, gastric adenoma, intestinal-type GC, and metastatic GC (p < 0.001). AdipoR1 and AdipoR2 expression was observed in 85 (34.0 %) and 118 (47.2 %) GC cases, respectively. Expression rates were higher in intestinal-type GC than in diffuse-type GC (p < 0.001 and 0.016, respectively). AdipoR1 and AdipoR2 expression was more frequent in advanced GC than in early GC (p < 0.001, each) and was associated with lymphatic invasion (p = 0.046 and 0.001, respectively). AdipoR2 expression was associated with poor overall and disease-free survival (p = 0.001 and 0.007, respectively). AdipoR1 expression was associated with poor disease-free survival for intestinal-type GC patients (p = 0.046). In multivariate analysis, AdipoR2 was an independent prognostic factor for intestinal-type GC (p = 0.017).
Adiponectin receptor expression is related to GC development and progression, especially intestinal-type GC. Thus, adiponectin receptor expression can serve as a prognostic marker in GC patients.
KeywordsGastric carcinoma Adiponectin receptor Prognosis Immunohistochemistry
Adipose tissue is involved in the regulation of energy homeostasis and is an important endocrine organ that secretes a number of biologically active adipokines (Kadowaki and Yamauchi 2005). Adiponectin, also known as gelatin-binding protein 28 (GBP28), AdipoQ, ACRP30, or apM1, is a 30-kDa protein. It is the most abundant adipokine and has insulin-sensitizing, anti-inflammatory, and anti-atherogenic activities. In contrast to most adipokines, plasma adiponectin levels are lower in obese than in lean subjects. Thus, it has been suggested that adiponectin might serve as the basis for the treatment of obesity-linked diseases such as diabetes and metabolic syndrome. However, recent studies have also focused on the potential role of adiponectin in the development and progression of malignancies (Kadowaki and Yamauchi 2005; Nishida et al. 2007; Chou et al. 2010; Kelesidis et al. 2006). Many groups have reported that circulating adiponectin levels are inversely associated with the risk of some malignancies associated with obesity and insulin resistance such as colorectal, breast, endometrial, and renal cancers (Chen et al. 2006; Petridou et al. 2003; Wei et al. 2005; Goktas et al. 2005), and the anti-proliferative effects of adiponectin have been demonstrated in several in vitro and in vivo studies (Kelesidis et al. 2006; Saxena et al. 2010; Sugiyama et al. 2009; Arditi et al. 2007; Kim et al. 2010). Additionally, other studies described a role for adiponectin as a motility inducer (Tang and Lu 2009; Chiu et al. 2009) or an angiogenic factor (Ouchi et al. 2004; Denzel et al. 2009). Together, such studies indicate that investigation into the biological activities of adiponectin is an active area of scientific pursuit.
The functions of adiponectin, including its anti-proliferative effects, are known to be mediated directly, through a hormonal mechanism, as well as indirectly, via a cell signaling pathway that includes the adiponectin receptors, 5′-AMP-activated protein kinase (AMPK), and mammalian target of rapamycin (Kelesidis et al. 2006). Interestingly, adiponectin receptors have been reported to be present in cancers of various organs including breast, prostate, lung, and pancreas (Korner et al. 2007; Michalakis et al. 2007; Petridou et al. 2007; Dalamaga et al. 2009; Chou et al. 2010). As for gastric carcinoma, several studies have shown that adiponectin receptors are expressed in gastric cancer (GC) cell lines or tissues, and although they attempted to establish a relationship between adiponectin receptor expression and the clinicopathologic characteristics of the disease, their results were not reproduced in other studies (Ishikawa et al. 2007; Barresi et al. 2009; Tsukada et al. 2011). Therefore, in the present study, we attempted to clarify the clinicopathologic implications of the expression of the 2 adiponectin receptors, AdipoR1 and AdipoR2, in GC tissue using tissue array and immunohistochemistry methodologies. In addition, we also evaluated the expression of adiponectin receptors in the premalignant lesions of GC and metastatic GC tissue in order to clarify the role of adiponectin receptors in GC pathogenesis and progression.
Materials and methods
Patients and tissue samples
Tissue specimens from 250 patients who underwent surgical resection for GC between May 2003 and December 2004 and for whom paraffin blocks and follow-up information were available were collected from the files of the Department of Pathology at Seoul National University Bundang Hospital. All patients had histologically proven adenocarcinoma of the stomach. Histopathologic and clinical data, including age, sex, tumor location, tumor size, and lymphatic and vascular invasion, were obtained from the medical records and pathologic reports. Histologic classification was evaluated according to the Lauren’s classification, and pathologic staging (pTNM stage) was determined according to the American Joint Committee on Cancer grading system (7th edition). The study population comprised 159 male patients (63.6 %) and 91 female patients (36.4 %). The median age of the individuals was 60.03 years (range 28–87 years). Among these patients, 142 (56.8 %) had advanced GC and 98 (43.2 %) had early GC. Of the 250 patients, 12 (4.8 %) had distant metastasis. Curative resection (R0 according to AJCC guidelines) was performed for 98.4 % of patients. No patient had received preoperative chemotherapy or radiotherapy. The mean follow-up period was 43.02 months (range 1–66 months). Data concerning cases lost to follow-up and death due causes other than GC were regarded as censored data for survival rate analysis.
In addition to the expression of adiponectin receptors in carcinoma lesions, we investigated the expression of adiponectin receptors in 3 other groups as follows: 59 cases of non-neoplastic gastric mucosa with intestinal metaplasia (NNGM), 48 cases of gastric adenoma (GA), and 58 cases of metastatic GC to the lymph nodes (MGC). Fifty-one among the 59 NNGM samples were taken from the same patients with GC enrolled in the present study. This study was approved by the Institutional Review Boards for Research Using Human Subjects at Seoul National University Bundang Hospital.
Tissue array method
Previously stained hematoxylin and eosin slides were retrospectively reviewed, and one formalin-fixed, paraffin-embedded archival block representative of each case was selected. The arrays were assembled by taking core tissue biopsies (2 mm in diameter) from specific locations in preexisting formalin-fixed, paraffin-embedded blocks (donor blocks) and re-embedding them in recipient paraffin blocks (tissue array blocks) using a trephine apparatus. Each tissue array block contained up to 60 samples. Five array blocks (SuperBiochips Laboratories, Seoul, Korea) containing representative samples from all 250 GC cases and one separate array block for each of the NNGM, GA, and MGC groups were prepared.
Immunohistochemical staining for anti-human AdipoR1 (1:400) and anti-human AdipoR2 (1:500) (rabbit polyclonal antibody: Phoenix Pharmaceuticals Inc., CA, USA) was performed on 8 tissue array slides using a streptavidin peroxidase procedure (labeled streptavidin–biotin) after microwave antigen retrieval. The detailed procedure is as follows. Four-micrometer-thick sections from tissue array blocks were deparaffinized in xylene and hydrated using alcohol (3×). They were then placed in citrate buffer (10 % citrate buffer stock in distilled water, pH 6.0). Microwave pretreatment for antigen retrieval was performed for 10 min. Endogenous peroxidase activity was blocked using 1 % horse serum in Tris-buffered saline (pH 6.0). Antibody binding was detected using avidin–biotin–peroxidase complex (Universal Elite ABC kit PK-6200; Vectastain, Burlingame, CA, USA) for 10 min and diaminobenzidine tetrahydrochloride solution (Kit HK 153–5 K; Biogenex, San Ramon, CA, USA).
Analysis of immunohistochemical staining
Two pathologists (E. Shin and H.S. Lee), who were blinded to the follow-up data on the patients’ medical records, independently evaluated the immunostained slides. After interpretation of the staining results, conflicting findings were discussed between both pathologists. Staining intensities were semi-quantitatively measured at a 200 × magnification, and staining intensity was categorized as negative (score = 0), weak (score = 1), moderate (score = 2), or strong (score = 3). The percentage of immune-reactive cells was also assessed (Supplementary Table 1). Because there are no absolute criteria for the immune-positivity of AdipoR1 and AdipoR2, by testing a series of different values, AdipoR1 was considered to be overexpressed when >20 % of tumor cells had an intensity score of ≥1 and AdipoR2 was considered to be overexpressed when >70 % of tumor cells had an intensity score of ≥2.
Statistical analysis was performed using the SPSS version 18.0 software package (SPSS, Chicago, IL, USA). The correlation between adiponectin receptor expression and clinicopathologic features was analyzed using Pearson’s χ2 test and the Student’s t test. Differences in protein expression between GCs and matched NNGMs were evaluated through McNemar test. The results were considered statistically significant for p ≤ 0.05. For the analysis of survival data, Kaplan–Meier survival curves were constructed, and the differences between survival curves were determined using the logrank test. Multivariate analysis was performed using Cox’s proportional hazards regression modeling, with p ≤ 0.05 considered statistically significant. The variables analyzed were overall survival (OS), which was calculated from the date of operation until death or last follow-up appointment, and disease-free survival (DFS), which was calculated from the date of operation until the date when recurrence or metastasis was detected.
Adiponectin receptor expression
Correlation between adiponectin receptor expression status and clinicopathologic variables of GC
Correlation analysis of clinicopathologic parameters and expression of adiponectin receptors
58.1 ± 11.7
63.9 ± 9.3
59.0 ± 12.5
60.6 ± 10.6
Tumor size (cm)
5.2 ± 3.4
4.9 ± 2.5
4.9 ± 3.0
5.2 ± 3.2
Depth of invasion
Relationship between AdipoR1 expression and survival in GC patients
Survival analysis of patients with intestinal-type GC indicated that there was no difference in OS between the AdipoR1-positive GC and AdipoR1-negative GC groups (p = 0.201) (Fig. 4c). However, the DFS of AdipoR1-positive GC patients was significantly shorter than that of the AdipoR1-negative GC patients (48.5 and 52.8 months, respectively; p = 0.046) (Fig. 4d); this statistical significance disappeared in multivariate analysis (p = 0.220).
Relationship between AdipoR2 expression and survival in GC patients
Cox regression analysis of prognostic factors in gastric adenocarcinoma of intestinal type
95 % CI
Distant metastasis (absent vs. present)
Lymph node metastasis (absent vs. present)
Lymphatic invasion (absent vs. present)
Vascular invasion (absent vs. present)
AdipoR1 expression (negative vs. positive)
AdipoR2 expression (negative vs. positive)
The presence of two isoforms of adiponectin receptors—AdipoR1 and AdipoR2—was first identified by Yamauchi et al. (2003). These two adiponectin receptors had unique distributions and affinities for the different forms of circulating adiponectin. AdipoR1 is a high-affinity receptor for globular adiponectin and is most abundant in skeletal muscle, whereas AdipoR2 has intermediate affinity for both globular and full-length adiponectin and is predominantly expressed in the liver (Goldstein and Scalia 2004). It is well known that these receptors mediate adiponectin regulation of fatty acid oxidation and glucose uptake. However, like adiponectin, the adiponectin receptors have attracted attention within the field of oncology and have been suggested to be associated with tumorigenesis as well as metabolism. Recent studies have reported the expression of adiponectin receptor mRNA and protein in several malignancies including those of the breast, colorectal, prostate, lung, and pancreas, and generally, the corresponding normal tissues express the receptors at decreased levels. Further, several studies have revealed that the expression of AdipoR1 and AdipoR2 in GC was higher than that in normal gastric mucosa, concordant with our results (Ishikawa et al. 2007; Barresi et al. 2009; Tsukada et al. 2011). Although discordant result was reported in the other report, they showed no statistical parameter and the number of enrolled cases was too small (Otani et al. 2010).
In the present study, adiponectin receptors were more frequently expressed in GC possessing more aggressive characteristics such as lymphatic invasion, lymph node metastasis, and deep invasion. Moreover, they were associated with poor OS and DFS rates. MGC demonstrated consistently strong positivity for adiponectin receptors. Previously, adiponectin was shown to have potent proangiogenic actions in both in vitro and in vivo studies (Hebbard et al. 2008; Ouchi et al. 2004). Angiogenesis is central to malignancy, and so, high adiponectin receptor expression may result in increased tumor invasion and lymph node metastasis by promoting angiogenesis. Moreover, Tang et al. showed that adiponectin directed migration via adiponectin receptors and the AMPK and nuclear factor-kappa B pathways in chondrosarcoma and prostate cancer cells (Chiu et al. 2009; Tang and Lu 2009). The association between adiponectin receptor expression and tumor invasiveness has been shown in other types of cancers. AdipoR2 expression correlated with nodal metastasis of esophageal adenocarcinoma (Howard et al. 2010), and invasive breast carcinoma displayed greater expression of adiponectin receptors than ductal carcinoma in situ (Jeong et al. 2011). However, in contrast with these results, tumor invasiveness was inversely correlated with adiponectin receptor expression in colorectal carcinoma (Byeon et al. 2010), and adiponectin was shown to suppress liver tumor metastasis in vivo (Man et al. 2010). Anti-proliferative effects of adiponectin were reported in several studies (Saxena et al. 2010; Sugiyama et al. 2009; Kim et al. 2010; Arditi et al. 2007), but opposing data have also been presented (Chen et al. 2012). The discrepancy between these studies may be the result of the different methodologies used by the investigators, but they also reflect the complex influence of adiponectin on tumor biology and tissue-type or cell-type dependency. For example, Wang et al. and Grossman et al. showed that the ability of adiponectin to inhibit the proliferation of breast cancer cell lines is cell-type dependent (Wang et al. 2006; Grossmann et al. 2008). In the present study, we also observed the difference of adiponectin receptors expression between two types of GC and this difference may be somewhat related with the cell-type dependency of adiponectin–adiponectin receptor axis. Because of the anti-proliferative action of adiponectin on tumors, adiponectin per se or adiponectin analogues have been considered as potentially effective anticancer agents with important therapeutic implications. However, the complexity of the biological functions of adiponectin should be fully considered, and more intensive basic and clinical research efforts are essential in order to better define its clinical applications.
A few cell line studies have shown that within the setting of low adiponectin levels, adiponectin receptor mRNA might be upregulated in a compensatory manner as is frequently seen in many hormonal axes (Dos Santos et al. 2008; Konturek et al. 2008). Thus, the association between the upregulation of AdipoR1 and AdipoR2 expression and the aggressive characteristics of some tumors might result from lower serum adiponectin levels in patients. However, in other studies, expression of adiponectin receptors in cancer tissue was not related to circulating adiponectin levels (Gialamas et al. 2011; Howard et al. 2010; Tsuchida et al. 2004; Gonzalez et al. 2010) and evidences have emerged which indicated that expression of adiponectin receptor might be regulated by more complex interaction with other hormones and cytokines in a tissue-specific manner (Tsuchida et al. 2004; Mistry et al. 2006; Gonzalez et al. 2010). Unfortunately, in the present study, the serum sample of enrolled patients was not available, so we could not examine the relation between the serum level of adiponectin and adiponectin receptor expression in GC tissue. Studies are needed to systematically compare the levels of serum adiponectin and adiponectin receptor expression in malignant tissues and to clarify the mechanism of overexpression of adiponectin receptor in cancer tissues.
The significant trend in our data toward a higher frequency of adiponectin receptor expression in the Lauren intestinal-subtype GC than in the diffuse-subtype GC is consistent with the previous reports that showed significantly higher expression of AdipoR1 in intestinal-type GC (Barresi et al. 2009) Histologic subtyping according to the Lauren’s classification not only is based on the morphology of tumors but also reflects the difference in histogenesis and metastatic behavior of both tumors: intestinal-type adenocarcinomas preferentially metastasize hematogenously, whereas diffuse-type adenocarcinomas are mostly associated with peritoneal dissemination (Carneiro and Sobrinho-Simoes 1996; Mori et al. 1995). Interestingly, we found that the correlation between adiponectin receptor expression and tumor stage or patients’ survival was evident especially in intestinal-type GC, and AdipoR2 was identified as an independent prognostic factor for intestinal-type GC patients. These results suggest the possibility of the aggressive behavior of intestinal-type GC being influenced by the angiogenic-related mechanisms of the adiponectin receptors and emphasize the need to subclassify tumors according to the Lauren’s classification when evaluating the clinical significance of adiponectin receptor expression in GC. In addition, their increased expression in premalignant lesions of GC (NNGM, GA, and intestinal-type GC) in a stepwise manner suggests that adiponectin receptors may have a pathogenetic role in the intestinal metaplasia-GA-intestinal-type gastric adenocarcinoma pathway.
In conclusion, adiponectin receptor expression is related to GC development and progression. The adiponectin receptors are suggested to be prognostic markers in GC patients, especially in those with intestinal-type GC. Further intensive study will be indispensible in identifying the exact role of adiponectin and the adiponectin receptors in gastric carcinogenesis and progression.
This study was supported by grant no. 11-2012-008 from the Seoul National University Bundang Hospital Research Fund.
Conflict of interest
The authors declare that they have no competing interests.