Reprimo-like (RPRML) is an uncharacterized member of the Reprimo gene family. Here, we evaluated the role of RPRML and whether its regulation by DNA methylation is a potential non-invasive biomarker of gastric cancer. RPRML expression was evaluated by immunohistochemistry in 90 patients with gastric cancer and associated with clinicopathologic characteristics and outcomes. The role of RPRML in cancer biology was investigated in vitro, through RPRML ectopic overexpression. Functional experiments included colony formation, soft agar, MTS, and Ki67 immunofluorescence assays. DNA methylation-mediated silencing was evaluated by the 5-azacytidine assay and direct bisulfite sequencing. Non-invasive detection of circulating methylated RPRML DNA was assessed in 25 gastric cancer cases and 25 age- and sex-balanced cancer-free controls by the MethyLight assay. Downregulation of RPRML protein expression was associated with poor overall survival in advanced gastric cancer. RPRML overexpression significantly inhibited clonogenic capacity, anchorage-independent growth, and proliferation in vitro. Circulating methylated RPRML DNA distinguished patients with gastric cancer from controls with an area under the curve of 0.726. The in vitro overexpression results and the poor patient survival associated with lower RPRML levels suggest that RPRML plays a tumor-suppressive role in the stomach. Circulating methylated RPRML DNA may serve as a biomarker for the non-invasive detection of gastric cancer.

Restrictions resulting from the COVID-19 pandemic abruptly reversed the slow decline of the diagnosis and mortality rates of gastric cancer (GC). This scenario highlights the importance of developing cost-effective methods for mass screening and evaluation of treatment response. In this study, we evaluated a non-invasive method based on the circulating methylated cell-free DNA (cfDNA) of Reprimo (RPRM), a tumor suppressor gene associated with the development of GC. Methylated RPRM cfDNA was analyzed in three de-identified cohorts: Cohort 1 comprised 81 participants with GC and 137 healthy donors (HDs); Cohort 2 comprised 27 participants with GC undergoing gastrectomy and/or chemotherapy analyzed at the beginning and after three months of treatment; and Cohort 3 comprised 1105 population-based participants in a secondary prevention program who underwent esophagogastroduodenal (EGD) endoscopy. This cohort includes 180 normal participants, 845 participants with premalignant conditions (692 with chronic atrophic gastritis [AG] and 153 with gastric intestinal metaplasia/low-grade dysplasia [GIM/LGD]), 21 with high-grade dysplasia/early GC [HGD/eGC], and 59 with advanced GC [aGC]). A nested case-control substudy was performed using a combination of methylated RPRM cfDNA and pepsinogens (PG)-I/II ratio. The dense CpG island of the promoter region of the RPRM gene was bisulfite sequenced and analyzed to develop a fluorescence-based real-time PCR assay (MethyLight). This assay allows the determination of the absolute number of copies of methylated RPRM cfDNA. A targeted sequence of PCR amplicon products confirmed the gastric origin of the plasma-isolated samples. In Cohort 1, the mean value of GCs (32,240.00 copies/mL) was higher than that of the HD controls (139.00 copies/mL) (p < 0.0001). After dividing this cohort into training–validation subcohorts, we identified an area under the curve of 0.764 (95% confidence interval (CI) = 0.683–0.845) in the training group. This resulted in a cut-off value of 87.37 copies/mL (sensitivity 70.0% and specificity 80.2%). The validation subcohort predicted a sensitivity of 66.67% and a specificity of 83.33%. In Cohort 2 (monitoring treatment response), RPRM levels significantly decreased in responders (p = 0.0042) compared to non-responders. In Cohort 3 (population-based participants), 18.9% %, 24.1%, 30.7%, 47.0%, and 71.2% of normal, AG, GIM/LGD, HGD/eGC, and aGC participants tested positive for methylated RPRM cfDNA, respectively. Overall sensitivity and specificity in distinguishing normal/premalignant conditions vs. GC were 65.0% (95% CI 53.52% to 75.33%) and 75.9% (95% CI 73.16% to 78.49%), respectively, with an accuracy of 75.11% (95% CI 72.45% to 77.64%). Logistic regression analyses revealed an OR of 1.85 (95% CI 1.11–3.07, p = 0.02) and an odds ratio (OR) of 3.9 (95% CI 1.53–9.93, p = 0.004) for the risk of developing GIM/LGD and HGD/eGC, respectively. The combined methylated RPRM cfDNA and PG-I/II ratio reached a sensitivity of 78.9% (95% CI 54.43% to 93.95%) and specificity of 63.04% (95% CI 52.34% to 72.88%) for detecting HGD/eGC vs. three to six age- and sex-matched participants with premalignant conditions. Our results demonstrate that methylated RPRM cfDNA should be considered a direct biomarker for the non-invasive detection of GC and a predictive biomarker for treatment response.

Aberrant expression of multiple microRNAs has been reported in gastric cancer. In particular, microRNA-597 has been associated with poor survival rates but is not yet well characterized. Seventy-five clinical samples, four cell lines, and two patient-derived organoids were evaluated for the expression of microRNA-597 and its target genes. microRNA-597 was transiently transfected for analysis of cell migration, invasion, wound healing, colony formation, and cell viability, and its regulation by long non-coding RNAs was explored using the TCGA-STAD and LncBook tools. In clinical samples, low expression of microRNA-597 was associated with the intestinal subtype (p = 0.002) and stages III and IV (p = 0.048). All functional readouts were reduced after microRNA-597 transfection, including colony formation, in patient-derived organoids. Among target genes, RUNX1 was directly regulated by microRNA-597. Other cell invasion genes were dependent on RUNX1 as a hub for regulation. Analysis of the Intersection between long non-coding RNAs co-expressed with RUNX1 and those with the highest microRNA-597 prediction binding identified KCNQ1OT1 as the top transcript. Silencing of KCNQ1OT1 and co-expression in clinical samples suggest the existence of a KCNQ1OT1/microRNA-597/RUNX1 network. The results indicate that microRNA-597 directly suppresses RUNX1, while KCNQ1OT1 modulates this interaction. Our approach enabled the simultaneous analysis of dysregulation in three families of transcripts in gastric cancer progression.