Basal differentiation and expression status of SOX2 and KLF4 in basal layers are prognostic factors for disease-specific survival in oral squamous cell carcinoma.

BACKGROUND: Basal differentiation in oral squamous cell carcinoma is usually detected at invasive sites. However, its significance as a prognostic value has been poorly investigated. METHODS: COL17 was selected as a basal differentiation marker because of its stable expression in the basal-like cells of oral squamous cell carcinoma. Sixty-five cases of oral squamous cell carcinoma were subclassified into COL17-high (30 cases) and -low (35 cases) types, and the prognostic value was analyzed by Cox regression analysis. In addition, the stem cell markers such as SOX2, KLF4, MYC as well as the stem cell-related markers BMI1, EZH2, and YAP and its paralog TAZ, were immunohistochemically analyzed. Their prognostic values were investigated along with their COL17 status by Cox regression analysis. RESULTS: No significant difference was observed between the COL17-high and -low groups in the disease-specific survival and recurrence-free survival in oral squamous cell carcinoma. When the COL17-high and -low categories were combined with the SOX2, KLF4, EZH2, or YAP/TAZ status in the basal layers, together with gender and age as covariates, the hazard ratios reached 3.3, 3.7, 2.8, and 3.1, respectively. In addition, multivariate analysis, including COL17, SOX2, and KLF4, with gender and age as covariates, showed a significantly poor prognosis for disease-specific survival. CONCLUSION: Based on the relatively high hazard ratios, it is indicated that basal differentiation and the expression status of SOX2 and KLF4 in the basal layers are prognostic factors for oral squamous cell carcinoma.

RNA-binding protein 14 promotes phase separation to sustain prostate specific antigen expression under androgen deprivation in human prostate cancer.

Castration-resistant prostate cancer (CRPC) is a big challenge in managing prostate cancer patients. The androgen receptor (AR) pathway is a major driver even in CRPC under androgen deprivation. The mechanism in maintaining of the AR pathway under androgen deprivation remains elusive. The recent discovery of biomolecular condensate, a membrane-less intracellular construct formed by liquid-liquid phase separation (LLPS), that facilitate molecular assembly, encouraged the re-screening of our previous microarray data list. We selected Rbm14 as a target molecule for further analysis because it works as a coactivator of nuclear receptors as well as it facilitates formation of biomolecular condensates via its intrinsically disordered region. GFP-tagged Rbm14 transfected into HEK293T cells formed droplet-like puncta, which diminished following treatment with 1,6-hexanediol. Droplet-like structures were also observed in immunofluorescence for endogenous RBM14 of PC-3 and DU145 cells. Luciferase assay revealed that Rbm14 enhanced androgen-responsive element (ARE)-mediated reporter activity in all conditions with or without testosterone and AR. Co-immunoprecipitation confirmed the Rbm14-AR interaction. Long non-coding RNAs, including NEAT1, SRA1, and HOTAIR, were also interacted with Rbm14. Small interfering RNAs of NEAT1 reduced ARE-mediated reporter activity, while transfection of SRA1 and HOTAIR enhance the reporter activity. Treatment with 1,6-hexanediol as well as transfection with a dominant-negative splice variant of Rbm14 reduced expression of prostate specific antigen (PSA), a prototype of androgen-regulated gene, in LNCaP, PC-3, and DU145 cells under androgen deprivation. Immunohistochemically, RBM14 expression was substantially upregulated in prostate cancer tissues after androgen deprivation therapy than in untreated tumors. In conclusion, RBM14 is a novel factor involved in maintenance of PSA expression via phase separation under androgen deprivation in prostate cancer.

Stimulation of cellular senescent processes, including secretory phenotypes and anti-oxidant responses, after androgen deprivation therapy in human prostate cancer.

Endocrine resistance is a major problem in prostate cancer. Recent studies suggest that cellular plasticity plays a key role in therapy resistance. Yet little is known about the cellular changes of human prostate cancer after androgen deprivation therapy (ADT). In this study, we investigated cellular senescence, senescence-associated secretory phenotypes (SASPs), and anti-oxidant responses. Hormone ablation upregulated senescence-associated (SA)-ß-Gal activity in prostate glands, as well as the expressions of p27KIP1 and p53, in a mouse castration model. In line with this, the expressions of p21CIP1 and p27KIP1 were significantly more upregulated in human non-pathological prostatic glands after ADT than in untreated specimens. In a study of SASP markers, the expressions of IL6 and IL8 were also more upregulated in human non-pathological prostatic glands after ADT than in untreated specimens. IL6, IL8, and MMP2 were expressed more strongly in human prostate cancer specimens resected after ADT than in untreated tumors. Of note, treatment with the anti-oxidant reagent NAC significantly suppressed SA-ß-Gal activity in androgen-sensitive human prostate cancer LNCaP cells. In immunohistochemical analyses on anti-oxidant response genes, NRF2 and NQO1 were more upregulated after hormone ablation in human prostate gland and carcinoma specimens after ADT than in untreated specimens or in murine prostate glands after castration. Taken together, these findings suggest that ADT induces cellular senescence processes accompanied by secretory phenotypes and anti-oxidant responses in prostate. These cellular changes may be attractive targets for preventing endocrine resistance in prostate cancer.

RhoC upregulation is correlated with reduced E-cadherin in human breast cancer specimens after chemotherapy and in human breast cancer MCF-7 cells.

Therapy-resistant cancer cells are a major problem in cancer research. Recent studies suggest that the epithelial-mesenchymal transition (EMT) is a key mechanism in therapy resistance. Yet, the expressions of EMT markers, EMT core regulators, and a stem cell marker of BMI1 during chemotherapy have been poorly analyzed in clinical breast cancer specimens. In the present study, we investigated the roles of RhoC under chemotherapy to follow up on earlier findings demonstrating the involvement of RhoC in prostate cancer resistance to endocrine therapy. Immunohistochemically, E-cadherin expression was significantly lower in human breast cancer specimens analyzed after chemotherapy than specimens biopsied before chemotherapy. Significant upregulation of fibronectin, a mesenchymal EMT marker, was found in post-chemotherapy analysis. A study of the EMT core regulators of SNAIL1, SNAIL2, TWIST1, and a well-known stem cell marker of BMI1 revealed no post-chemotherapy upregulation of these molecules. In contrast, RhoC expression was significantly upregulated in post-chemotherapy breast cancer specimens. MCF-7 cells stably transfected with the constitutive active (CA) RhoC plasmid manifested a reduced level of E-cadherin at the peripheries and disorganization of actin fibers, with no accompanying upregulation of SNAIL1, SNAIL2, TWIST1, or BMI1 in Western blots. Exposure of etoposide on MCF-7 cells showed RhoC upregulation together with reduced membranous expression of E-cadherin and disorganization of actin fibers. In MTT assay, however, the CA-RhoC-expressing MCF-7 cells failed to show chemotherapy resistance under etoposide treatment. Taken in sum, RhoC may contribute to an EMT-like process in human breast cancer during chemotherapy.

RhoC and guanine nucleotide exchange factor Net1 in androgen-unresponsive mouse mammary carcinoma SC-4 cells and human prostate cancer after short-term endocrine therapy.

BACKGROUND: Endocrine resistance is a critical issue in managing patients with prostate cancer. This study is undertaken to search for a potential molecular target connected with this process using a model system of androgen-dependent and androgen-unresponsive SC-3 and SC-4 cells. METHODS: Expression profiles, actin stress fiber organization, and the levels of activated Rho GTPases were compared between SC-4 and SC-3 cells using an oligonucleotide microarray, phalloidin staining, and a Rho activation assay. The cell viability was analyzed with a Rho inhibitor or by stable transfection with either a dominant-negative (DN) form of RhoC or a mutant form of NET1 (mutNET1). The expressions of RhoC, NET1, and epithelial-mesenchymal transition (EMT) markers were immunohistochemically analyzed in human prostate cancer specimens after short-term endocrine therapy and in an untreated condition. RESULTS: SC-4 cells exhibited mesenchymal phenotypes with activation of Rho signals. Treatment with a Rho inhibitor suppressed the cell viability in SC-4 cells, but not in SC-3 cells. The cell viability of SC-4 cells stably expressing DN-RhoC and mutNET1 was also attenuated. In the immunohistochemical analysis, NET1 and the EMT marker of N-cadherin were expressed at higher levels in prostate cancers after short-term endocrine therapy than in untreated tumors, and RhoC expression was maintained after short-term endocrine therapy. CONCLUSIONS: Rho signaling is involved in the cell survival of SC-4 cells. The higher expressions of RhoC and NET1 in human prostate cancers after short-term endocrine therapy suggest that RhoC and NET1 may become therapeutic targets during endocrine therapy.

Vitamin D3 suppresses the androgen-stimulated growth of mouse mammary carcinoma SC-3 cells by transcriptional repression of fibroblast growth factor 8.

Active metabolites of vitamin A and D are well known to act as growth inhibitors in hormone-related prostate and breast cancers. When various concentrations of 1alpha,25-dihydroxyvitamin D3 (vitamin D3), all-trans-retinoic acid (ATRA) and 9-cis retinoic acid (9-cis RA) were examined, the androgen-stimulated growth of mouse mammary carcinoma SC-3 cells was inhibited by vitamin D3 alone in a dose-dependent manner. A flow cytometer analysis showed that vitamin D3 leads SC-3 cells to relative G1-growth arrest after 72 h. Characterization of vitamin D3-responsive genes using an oligonucleotide microarray demonstrated that 220 genes were upregulated at more than threefold, and 84 genes were downregulated to less than one-third, compared with the testosterone-stimulated SC-3 cells. Neither cyclin-dependent kinase inhibitors (CDKIs) nor the antiapoptotic bcl-2 gene were induced in vitamin D3-responsive genes, with the exception of a slight induction of p15(INK4B). Importantly, fgf8 was markedly repressed in response to vitamin D3. The exogenous addition of FGF8 canceled the growth suppression by vitamin D3 in SC-3 cells, suggesting that the repression of fgf8 is an indispensable step in vitamin D3-mediated growth inhibition. In reporter assays using the ARE-containing artificial construct and the natural androgen-regulated PSA promoter, co-transfection of the vitamin D receptor (VDR) and androgen receptor (AR) suppressed AR-stimulated promoter activity. In addition, vitamin D3 also suppressed androgen-stimulated promoter activity in the stably transfected SC-3 cells. Moreover, VDR repressed the core promoter activity of fgf8 in COS1 cells and in the SC-3 cells. All these findings strongly suggest that vitamin D3 serves as a negative regulator for both androgen-related and fgf8 transcriptions.

Transcriptional repression of fibroblast growth factor 8 by transforming growth factor-beta in androgen-dependent SC-3 cells.

We here characterized the transcriptional profiles of TGF-beta-responsive genes using androgen-dependent mouse mammary carcinoma SC-3 cells. Compared with the testosterone-stimulated SC-3 cells, 165 genes were up-regulated at more than 5-fold, and 78 genes were down-regulated to less than one-third in response to TGF-beta. Of note, fgf8, an androgen-inducible growth factor essential to the androgen-dependent growth of SC-3 cells, was severely repressed in response to TGF-beta. Real-time PCR confirmed that the androgenic induction of the fgf8 transcripts is severely attenuated by TGF-beta. Although a considerable number of growth-suppressive genes were up-regulated in response to TGF-beta, the treatment with TGF-beta was insufficient to lead SC-3 cells to apoptosis within 24h by both the TUNEL method and the caspase 3 activity assay. Flow cytometric analysis rather indicated the cell-static effect of TGF-beta on the androgen-stimulated SC-3 cells. In addition, TGF-beta failed to suppress the FGF8-stimulated growth of SC-3 cells, suggesting that the repression of fgf8 is required for the TGF-beta-mediated growth inhibition in SC-3 cells. In a reporter assay, androgen-responsive promoter activity was suppressed by TGF-beta in SC-3 cells. Based on this finding, it is likely that some of the androgen-inducible genes are physiological targets of the TGF-beta-mediated transcriptional control, and therefore, it is strongly suggested that the repression of fgf8 might be directly or indirectly involved in this transcriptional control by TGF-beta.

Distinct expression patterns of claudin-1 and claudin-4 in intraductal papillary-mucinous tumors of the pancreas.

The expression of claudin-4 was investigated in human pancreas, pancreatic ductal adenocarcinomas, and intraductal papillary-mucinous tumors of the pancreas (IPMT), and compared with that of claudin-1. In human adult pancreatic specimens, both claudin-1 and claudin-4 were immunohistochemically found in main and branching pancreatic ducts, terminal ductules and acinic cells, with the exception of endocrine cells. Of 12 cases of pancreatic ductal adenocarcinoma, 11 (92%) had positive immunostaining for claudin-4, and seven (58%) for claudin-1. In 44 lesions of 22 cases of IPMT, including six hyperplastic foci distant from the main lesions, clauidin-1 was positive in three out of six (50%) hyperplastic foci, 14 out of 17 (82%) adenomas, three out of 10 (30%) borderline tumors, two out of six (33%) non-invasive carcinomas, and one out of five (20%) invasive carcinomas, producing a statistically negative correlation with histological tumor grades. In contrast, claudin-4 was negative in the six hyperplastic foci, and positive in four out of the 17 (24%) adenomas, five out of the 10 (50%) borderline tumors, five out of the six (83%) non-invasive carcinomas, and four out of the five (80%) invasive carcinomas, producing a statistically positive correlation with histological tumor grades. On study of IPMT subtypes, claudin-1 was positive in nine out of 10 (90%) clear-cell types, seven out of 20 (35%) dark-cell types, and four out of eight (50%) compact-cell types. In contrast, claudin-4 was positive in two out of the 10 (20%) clear-cell types, 13 out of the 20 (65%) dark-cell types, and three out of the eight (38%) compact-cell types. These distinct expression patterns of claudin-1 and claudin-4 suggest that both claudins serve as useful molecular markers for the tumor classification of IPMT.

Neuroendocrine tissue-specific transcription factor, BETA2/NeuroD, in gastric carcinomas: a comparison with chromogranin A and synaptophysin expressions.

BETA2/NeuroD (NeuroD) is a basic helix-loop-helix type of transcription factor mainly involved in neuroendocrine differentiation. In this study, we evaluated the prevalence of neuroendocrine differentiation in gastric carcinomas by analyzing the NeuroD expression in comparison with those of chromogranin A and synaptophysin. Of the 70 cases of gastric adenocarcinoma, the expressions of NeuroD, chromogranin A, and synaptophysin were detected in 17 (24.3%), four (5.7%), and 24 cases (34.3%), respectively, with preferential expressions in a non-solid type of poorly differentiated adenocarcinoma. The expression pattern of NeuroD was mostly concordant with that of synaptophysin and partly with chromogranin A, indicating that NeuroD serves as a good neuroendocrine marker in gastric adenocarcinomas. On the other hand, no immunoreactivity against NeuroD was detectable in nine cases of gastric neuroendocrine carcinomas, including small cell carcinomas, despite the presence of synaptophysin and chromogranin A expressions. These findings led us to conclude that neuroendocrine differentiation is estimated to be present in 20-30% of gastric adenocarcinomas with preference to a non-solid type of poorly differentiated adenocarcinoma. In addition, the negative expression of NeuroD in neuroendocrine carcinomas suggests that other regulatory mechanisms are possibly involved in the development of neuroendocrine carcinoma.

Fibroblast growth factor 8 expression in breast carcinoma: associations with androgen receptor and prostate-specific antigen expressions.

Recent experimental data have clearly demonstrated that fibroblast growth factor (FGF)8 plays a key role in the development of human prostate and breast cancers. However, little is known about the FGF8 expression profile in human breast cancer specimens. In this study, we analyzed FGF8 expression in 78 surgically resected specimens of breast cancer using an immunohistochemical method. In total, FGF8 expression was found in 40 (51.3%) of the breast carcinomas. FGF8 expression was not associated with any of the general clinicopathological parameters, including age, tumor size, histological grade, and histological type. In addition, there was no correlation between FGF8 expression and either c-erbB-2 overexpression or the status of the axillary lymph-node metastasis, both of which have been established as important prognostic factors in breast carcinomas. While no significant association was found between FGF8 expression and estrogen- or progesterone-receptor status, it is of interest that FGF8 expression was significantly associated with androgen-receptor status and the expression of prostate-specific antigen (PSA), one of the androgen-regulated proteins, in human breast carcinomas. These associations support the reported in vitro data demonstrating the regulation of FGF8 by androgens, and also suggest that PSA may be a useful marker for patients with FGF8-expressing breast carcinomas.