Cancer-associated fibroblasts (CAF) constitute a major component of the tumor microenvironment. Recent observations in genetically engineered mouse models and clinical studies have suggested that there may exist at least two functionally different populations of CAFs, that is, cancer-promoting CAFs (pCAF) and cancer-restraining CAFs (rCAF). Although various pCAF markers have been identified, the identity of rCAFs remains unknown because of the lack of rCAFspecific marker(s). In this study, we found that Meflin, a glycosylphosphatidylinositol-anchored protein that is a marker of mesenchymal stromal/stem cells and maintains their undifferentiated state, is expressed by pancreatic stellate cells that are a source of CAFs in pancreatic ductal adenocarcinoma (PDAC). In situ hybridization analysis of 71 human PDAC tissues revealed that the infiltration of Meflin-positive CAFs correlated with favorable patient outcome. Consistent herewith, Meflin deficiency led to significant tumor progression with poorly differentiated histology in a PDAC mouse model. Similarly, genetic ablation of Meflin-positive CAFs resulted in poor differentiation of tumors in a syngeneic transplantation model. Conversely, delivery of a Meflin-expressing lentivirus into the tumor stroma or overexpression of Meflin in CAFs suppressed the growth of xenograft tumors. Lineage tracing revealed that Meflin-positive cells gave rise to a-smooth muscle actin-positive CAFs that are positive or negative for Meflin, suggesting a mechanism for generating CAF heterogeneity. Meflin deficiency or low expression resulted in straightened stromal collagen fibers, which represent a signature for aggressive tumors, in mouse or human PDAC tissues, respectively. Together, the data suggest that Meflin is a marker of rCAFs that suppress PDAC progression. Significance: Meflin marks and functionally contributes to a subset of cancer-associated fibroblasts that exert antitumoral effects.

There have been several clinical studies examining the factors associated with cardiovascular disease (CVD) in patients with primary aldosteronism (PA); however, their results have left it unclear whether CVD is affected by the plasma aldosterone concentration or hypokalemia. We assessed the PA database established by the multicenter JPAS (Japan Primary Aldosteronism Study) and compared the prevalence of CVD among patients with PA with that among age-, sex-, and blood pressure-matched essential hypertension patients and participants with hypertension in a general population cohort. We also performed binary logistic regression analysis to determine which parameters significantly increased the odds ratio for CVD. Of the 2582 patients with PA studied, the prevalence of CVD, including stroke (cerebral infarction, cerebral hemorrhage, or subarachnoid hemorrhage), ischemic heart disease (myocardial infarction or angina pectoris), and heart failure, was 9.4% (stroke, 7.4%; ischemic heart disease, 2.1%; and heart failure, 0.6%). The prevalence of CVD, especially stroke, was higher among the patients with PA than those with essential hypertension/hypertension. Hypokalemia (K ≤3.5 mEq/L) and the unilateral subtype significantly increased adjusted odds ratios for CVD. Although aldosterone levels were not linearly related to the adjusted odds ratio for CVD, patients with plasma aldosterone concentrations ≥125 pg/mL had significantly higher adjusted odds ratios for CVD than those with plasma aldosterone concentrations <125 pg/mL. Thus, patients with PA seem to be at a higher risk of developing CVD than patients with essential hypertension. Moreover, patients with PA presenting with hypokalemia, the unilateral subtype, or plasma aldosterone concentration ≥125 pg/mL are at a greater risk of CVD and have a greater need for PA-specific treatments than others.

HOP homeobox (HOPX) is an unusual homeobox gene encoding three spliced transcript variants, among which the only HOPX-b promoter harbors CpG islands. The characteristics of its promoter methylation was analyzed using bisulfite sequencing and quantitative-methylationspecific polymerase chain reaction (Q-MSP), and the effects of HOPX expression were also examined. HOPX-b expression was silenced in all gastric cancer cell lines tested; its expression could be restored by treatment with demethylating agent. On Q-MSP, HOPX-b hypermethylation (cut-off value of 3.55) was found in 84% (67 out of 80) of primary tumor tissues and 10% (8 out of 80) of the corresponding normal tissues and could discriminate normal from tumor tissues (Po0.0001). The prognosis of the advanced cases with HOPX-b hypermethylation was as poor as those with stage IV disease when cut-off value was set at 11.28. This finding was validated in an independent cohort of 90 advanced gastric cancers. The HOPX-b hypermethylation was also an independent prognostic factor (P ¼ 0.029) on multivariate analysis. Exogenous HOPX expression significantly inhibited cell proliferation, colony formation and invasion as well as enhanced apoptosis. Taken together, HOPX-b promoter methylation is a frequent and cancer-specific event in gastric cancer. Quantitative assessment of HOPX-b methylation has great clinical potential as a marker of tumor aggressiveness.