Gene expression analysis of early and advanced gastric cancers

Operating room to bench for gastric cancer

Gastric cancer is the most common cancer in Korea, with an age-standardized rate of 61.2 in males and 23.9 in females (in 2007), one of the highest in the world. Using a large gastric tissue depository and the extensive clinical experience gained from gastric cancer surgery, we work as a 'translational researcher' to apply basic research tools and results to the clinical field. We are also interested in providing answers to the questions in the operating room using the methods of basic research. I would like to introduce our research activities in this review paper.

Matrix metalloproteinase-10 promotes Kras-mediated bronchio-alveolar stem cell expansion and lung cancer formation

Matrix metalloproteinase 10 (MMP-10; stromelysin 2) is a member of a large family of structurally related matrix metalloproteinases, many of which have been implicated in tumor progression, invasion and metastasis. We recently identified Mmp10 as a gene that is highly induced in tumor-initiating lung bronchioalveolar stem cells (BASCs) upon activation of oncogenic Kras in a mouse model of lung adenocarcinoma. However, the potential role of Mmp10 in lung tumorigenesis has not been addressed. Here, we demonstrate that Mmp10 is overexpressed in lung tumors induced by either the smoke carcinogen urethane or oncogenic Kras. In addition, we report a significant reduction in lung tumor number and size after urethane exposure or genetic activation of oncogenic Kras in Mmp10 null (Mmp10^−/−) mice. This inhibitory effect is reflected in a defect in the ability of Mmp10-deficient BASCs to expand and undergo transformation in response to urethane or oncogenic Kras in vivo and in vitro, demonstrating a role for Mmp10 in the tumor-initiating activity of Kras-transformed lung stem cells. To determine the potential relevance of MMP10 in human cancer we analyzed Mmp10 expression in publicly-available gene expression profiles of human cancers. Our analysis reveals that MMP10 is highly overexpressed in human lung tumors. Gene set enhancement analysis (GSEA) demonstrates that elevated MMP10 expression correlates with both cancer stem cell and tumor metastasis genomic signatures in human lung cancer. Finally, Mmp10 is elevated in many human tumor types suggesting a widespread role for Mmp10 in human malignancy. We conclude that Mmp10 plays an important role in lung tumor initiation via maintenance of a highly tumorigenic, cancer-initiating, stem-like cell population, and that Mmp10 expression is associated with stem-like, highly metastatic genotypes in human lung cancers. These results indicate that Mmp10 may represent a novel therapeutic approach to target lung cancer stem cells.

Upregulated INHBA expression is associated with poor survival in gastric cancer

Expression microarrays are widely used for investigating the candidate molecular targets in human cancer. While genome-wide expression signatures screened by gene set enrichment analysis (GSEA) were not performed in Chinese gastric cancer (GC). To gain new molecular targets for GC, GSEA analysis was performed. In the present study, GSEA were used to pick out differentially expressed gene sets of our database. Total RNA of paired tissue samples (n = 48) and a tissue microarray containing 132 paired tissues were used to further validate expression levels of INHBA and its correction with clinicopathological factors. Upregulated INHBA expression in gastric cancer was screened and further confirmed by qPCR and immunostaining analysis. Increased INHBA expression was significantly correlated with the diameter of cancer and depth of tumor invasion. Patients with higher expression levels of INHBA had a shorter disease-free survival rate. It was effective to gain new molecular targets for GC by GSEA analysis. INHBA may be a poor survival indicator of GC.

Multi-cancer computational analysis reveals invasion-associated variant of desmoplastic reaction involving INHBA, THBS2 and COL11A1

Background

Despite extensive research, the details of the biological mechanisms by which cancer cells acquire motility and invasiveness are largely unknown. This study identifies an invasion associated gene signature shedding light on these mechanisms.

Methods

We analyze data from multiple cancers using a novel computational method identifying sets of genes whose coordinated overexpression indicates the presence of a particular phenotype, in this case high-stage cancer.

Results

We conclude that there is one shared "core" metastasis-associated gene expression signature corresponding to a specific variant of stromal desmoplastic reaction, present in a large subset of samples that have exceeded a threshold of invasive transition specific to each cancer, indicating that the corresponding biological mechanism is triggered at that point. For example this threshold is reached at stage IIIc in ovarian cancer and at stage II in colorectal cancer. Therefore, its presence indicates that the corresponding stage has been reached. It has several features, such as coordinated overexpression of particular collagens, mainly COL11A1 and other genes, mainly THBS2 and INHBA. The composition of the overexpressed genes indicates invasion-facilitating altered proteolysis in the extracellular matrix. The prominent presence in the signature of INHBA in all cancers strongly suggests a biological mechanism centered on activin A induced TGF-β signaling, because activin A is a member of the TGF-β superfamily consisting of an INHBA homodimer. Furthermore, we establish that the signature is predictive of neoadjuvant therapy response in at least one breast cancer data set.

Conclusions

Therefore, these results can be used for developing high specificity biomarkers sensing cancer invasion and predicting response to neoadjuvant therapy, as well as potential multi-cancer metastasis inhibiting therapeutics targeting the corresponding biological mechanism.

Gene expression profiling of metaplastic lineages identifies CDH17 as a prognostic marker in early stage gastric cancer

BACKGROUND & AIMS

Intestinal metaplasia (IM) and spasmolytic polypeptide-expressing metaplasia (SPEM) are precursors to gastric carcinogenesis. We sought to identify molecular biomarkers of gastric metaplasias and gastric cancer by gene expression profiling of metaplastic lesions from patients.

METHODS

Complementary DNA microarray analysis was performed on IM and SPEM cells isolated from patient samples using laser capture microdissection. Up-regulated transcripts in metaplastic lesions were confirmed by immunostaining analysis in IM, SPEM, and gastric cancer tissues. Proteins that were highly expressed specifically in gastric cancer tissues were analyzed for their association with survival in a test set (n = 450) and a validation set (n = 502) of samples from gastric cancer patients.

RESULTS

Compared with normal chief cells, 858 genes were differentially expressed in IM or SPEM samples. Immunostaining was detected for 12 proteins, including 3 new markers of IM (ACE2, LGALS4, AKR1B10) and 3 of SPEM (OLFM4, LYZ, DPCR1). Of 13 proteins expressed in IM or SPEM, 8 were expressed by 17%-50% of human gastric cancer tissues (MUC13, OLFM4, CDH17, KRT20, MUC5AC, LGALS4, AKR1B10, REG4). Expression of CDH17 or MUC13 correlated with patient survival in the test and validation sets. Multivariate analysis showed that CDH17 was an independent prognostic factor in patients with stage I or node-negative disease.

CONCLUSIONS

We identified several novel biomarkers for IM, SPEM, and gastric cancer using gene expression profiling of human metaplastic lesions. Expression of CDH17 and MUC13 was up-regulated in gastric cancer tissues. CDH17 is a promising prognostic marker for early stage gastric cancer.

Differential gene and protein expression in primary gastric carcinomas and their lymph node metastases as revealed by combined cDNA microarray and tissue microarray analysis

OBJECTIVE

To gain insight into the molecular events of lymph node metastasis of human gastric carcinoma.

METHODS

The gene expression profile of five matched primary gastric carcinomas and their lymph node metastases was analyzed by complementary DNA (cDNA) microarray. Differential genes were identified in the metastatic and corresponding primary tumor pairs. Among the differentially expressed genes, carbonic anhydrase II (CAII) and insulin-like growth factor binding protein 4 (IGFBP 4) genes were detected by RT-PCR. CTTN protein expression was examined by tissue microarray.

RESULTS

There was a high expression (over twofold) of 44 genes and a low expression (under twofold) of 32 genes in lymph node metastasis compared with primary gastric carcinoma, respectively. CAII mRNA was downregulated and IGFBP 4 mRNA was upregulated in paired lymph node metastases of gastric carcinomas. The overexpression of CTTN protein was related to the lymph node metastasis and the clinical stage of gastric carcinomas.

CONCLUSION

This study showed that there is a low expression of genes relative to growth signal and immune response in lymph node metastases, and a high expression of genes relative to growth factor, cell cycle, cell motility and adhesion in lymph node metastases compared with primary gastric carcinomas. The expression of CTTN was related to the invasion and metastasis of gastric cancer.

Biological and molecular heterogeneity of breast cancers correlates with their cancer stem cell content

Pathways that govern stem cell (SC) function are often subverted in cancer. Here, we report the isolation to near purity of human normal mammary SCs (hNMSCs), from cultured mammospheres, on the basis of their ability to retain the lipophilic dye PKH26 as a consequence of their quiescent nature. PKH26-positive cells possess all the characteristics of hNMSCs. The transcriptional profile of PKH26-positive cells (hNMSC signature) was able to predict biological and molecular features of breast cancers. By using markers of the hNMSC signature, we prospectively isolated SCs from the normal gland and from breast tumors. Poorly differentiated (G3) cancers displayed higher content of prospectively isolated cancer SCs (CSCs) than did well-differentiated (G1) cancers. By comparing G3 and G1 tumors in xenotransplantation experiments, we directly demonstrated that G3s are enriched in CSCs. Our data support the notion that the heterogeneous phenotypical and molecular traits of human breast cancers are a function of their CSC content.

Gene expression analysis identifies over-expression of CXCL1, SPARC, SPP1, and SULF1 in gastric cancer

To elucidate gene expression signatures associated with gastric carcinogenesis, we performed a genome-wide expression analysis of 46 Finnish and 20 Japanese gastric tissues. Comparative analysis between Finnish and Japanese datasets identified 58 common genes that were differentially expressed between cancerous and non-neoplastic gastric tissues. Twenty-six of these genes were up-regulated in cancer and 32 down-regulated. Of these genes, 64% were also differentially expressed in another unrelated publicly available dataset. The expression levels of four of the up-regulated genes, CXCL1, SPARC, SPP1 and SULF, were further analyzed in 82 gastric tissues using quantitative real-time RT-PCR. This analysis validated the results from the microarray analysis as the expression of these four genes was significantly higher in the cancerous tissue compared with the normal tissue (fold change 3.4-8.9). Over-expression of CXCL1 also positively correlated with improved survival. To conclude, irrespective of the microarray platform or patient population, a common gastric cancer gene expression signature of 58 genes, including CXCL1, SPARC, SPP1, and SULF, was identified. These genes represent potential biomarkers for gastric cancer.

Similar expression profiles of a core of genes and proteins in cells that have acquired a metastatic phenotype, genetically or by in vivo evolution

We compared gene expression patterns related to cancer progression, cell adhesion and cytokine expression in v-Src-transformed isogenic cell lines with different levels of metastatic aggressiveness in syngenic animals: a high (HM) and a low metastatic (LM) cell lines, both predetermined by in vitro oncogenic transformation, and a high metastatic (NM) cell line resulting from in vivo selection of grafted LM cells. While LM and NM shared a common v-Src transforming variant different from that of HM, gene expression patterns were similar in HM and NM but different from that of LM. In this model, the induction of the metastatic phenotype by an oncogene variant or by the host environment led to highly similar expression profiles. If confirmed in other models, these results would be relevant for the diagnosis and treatment of metastasis.

Perineural invasion after preoperative chemotherapy predicts poor survival in patients with locally advanced gastric cancer: gene expression analysis with pathologic validation

BACKGROUND

We examined gene expression profiles and clinicopathologic features (tumor location, stage, graded pathologic response, perineural invasion (PNI), Lauren's classification, and survival) of patients with gastric cancer who received preoperative chemotherapy to identify prognostic markers.

METHODS

Thirty-eight patients with locally advanced gastric cancer received preoperative chemotherapy on a phase II trial. Twelve fresh-frozen tumor samples were available for RNA expression analysis. Differential gene expression between tumors with and without PNI was identified and correlated with clinicopathologic features.

RESULTS

Preliminary hierarchical clustering suggested a separation between long- and short-term survivors. The close association between PNI and overall survival was identified and validated immunohistochemically in 31 completely resected gastric tumors. Five-year survival for patients with PNI and without PNI was 5% and 65%, respectively (P < 0.01). PNI added significant prognostic value to posttreatment pathologic stage, (P < 0.01). Differential gene expression profile for PNI and non-PNI tumors identified 111 potentially relevant genes.

CONCLUSIONS

Our results demonstrate that the presence of PNI after preoperative chemotherapy is associated with poor survival. These results need to be validated in prospective studies, to help establish whether patients with evidence of PNI would be candidates for more aggressive therapy or enrollment into clinical trials. The presence of PNI provides additional prognostic importance to posttreatment pathologic stage and may indicate treatment resistance. Understanding the molecular events associated with PNI, may provide insight into new therapeutic agents for this subset of patients with resistant tumors.

The prognostic value of CD147/EMMPRIN is associated with monocarboxylate transporter 1 co-expression in gastric cancer

The aim of the present work was to assess the role of monocarboxylate transporters (MCTs), namely MCT1 and MCT4 as well as MCT/CD147 co-expression in gastric tissues and evaluate their clinico-pathological significance in gastric carcinoma. For that, we analysed the immunohistochemical expression of MCT1, MCT4 and CD147, in a large series of gastric samples, including non-neoplastic, tumour and metastatic tissues. A significant decrease in MCT4 plasma membrane expression was observed from non-neoplastic to gastric primary malignant tissues and to lymph-node metastasis and both MCT1 and MCT4 correlated with CD147. Importantly, both MCT4 and CD147 were more frequently expressed in Lauren's intestinal-type tumours and MCT1/CD147 co-expression was associated with advanced gastric carcinoma, Lauren's intestinal type, TNM staging and lymph-node metastasis. Our results showed that the prognostic value of CD147 was associated with MCT1 co-expression in gastric cancer cells, supporting the view that CD147 plasma membrane activity is dependent on MCT co-expression.

Prediction of breast cancer metastasis by genomic profiling: where do we stand?

Current concepts conceive “breast cancer” as a complex disease that comprises several very different types of neoplasms. Nonetheless, breast cancer treatment has considerably improved through early diagnosis, adjuvant chemotherapy, and endocrine treatments. The limited prognostic power of classical classifiers determines considerable over-treatment of women who either do not benefit from, or do not at all need, chemotherapy. Several gene expression based molecular classifiers (signatures) have been developed for a more reliable prognostication. Gene expression profiling identifies profound differences in breast cancers, most probably as a consequence of different cellular origin and different driving mutations and can therefore distinguish the intrinsic propensity to metastasize. Existing signatures have been shown to be useful for treatment decisions, although they have been developed using relatively small sample numbers. Major improvements are expected from the use of large datasets, subtype specific signatures and from the re-introduction of functional information. We show that molecular signatures encounter clear limitations given by the intrinsic probabilistic nature of breast cancer metastasis. Already today, signatures are, however, useful for clinical decisions in specific cases, in particular if the personal inclination of the patient towards different treatment strategies is taken into account.

Endothelial adherens junctions control tight junctions by VE-cadherin-mediated upregulation of claudin-5

Intercellular junctions mediate adhesion and communication between adjoining cells. Although formed by different molecules, tight junctions (TJs) and adherens junctions (AJs) are functionally and structurally linked, but the signalling pathways behind this interaction are unknown. Here we describe a cell-specific mechanism of crosstalk between these two types of structure. We show that endothelial VE-cadherin at AJs upregulates the gene encoding the TJ adhesive protein claudin-5. This effect requires the release of the inhibitory activity of forkhead box factor FoxO1 and the Tcf-4-beta-catenin transcriptional repressor complex. Vascular endothelial (VE)-cadherin acts by inducing the phosphorylation of FoxO1 through Akt activation and by limiting the translocation of beta-catenin to the nucleus. These results offer a molecular basis for the link between AJs and TJs and explain why VE-cadherin inhibition may cause a marked increase in permeability.

Metastasis signatures: genes regulating tumor-microenvironment interactions predict metastatic behavior

The possibility of predicting clinical outcome of cancer patients through the analysis of gene expression profiles in the primary tumor is a kind of ideological revolution as the multistep carcinogenesis model postulates that the proportion of cells within the primary tumor that actually acquire metastasis driving mutation(s) is small; too small to leave its imprint on the gene expression profile. The data collected to date have brought a new paradigm to reality in the metastasis field: metastasis must at least in part rely on mutations and/or gene regulation events present in the majority of cells which constitute the primary tumor mass. By analyses of differential expression of primary tumors versus metastases or by functional analyses of putative metastasis genes in experimental metastasis, many metastasis-associated gene expression events have been identified that correlate with the development of metastases. Among genes "favoring" metastasis, we find many molecules that are expressed not by the tumor cell itself but by the cells of the microenvironment, as well as genes over-expressed in the primary tumor that have a principle role in mediating tumor-host interactions. Here we review these concepts and advance hypotheses on how gene expression of the primary tumor and the microenvironment can favor the spread of the metastasis seeds and how this knowledge can provide tools to secondary prevention.