DNA Methylation Patterns After Weight Loss Reveal Druggable Cancer Targets
Bariatric surgery-linked epigenetic changes in skeletal muscle point to four key genes with therapeutic potential in cancer cachexia.
Summary
Researchers used the Illumina EPIC methylation array to identify conserved DNA methylation changes in skeletal muscle before and after bariatric surgery. By integrating data from multiple public datasets and the CPTAC cancer proteomics consortium, they pinpointed four key genes — KCNB1, PEAK1, SCG5, and TNIK — whose methylation patterns are linked to weight loss. These findings were validated using mouse knockout phenotype data and druggability resources. A pathway enrichment score derived from these probes correlated positively with chromosomal instability scores in tumor tissues, suggesting shared epigenetic mechanisms between obesity-related weight loss and cancer progression. The study also confirmed SCG5 protein expression differences across lung cancer cell lines via ELISA.
Detailed Summary
Cancer cachexia — the severe, involuntary weight loss seen in many cancer patients — remains poorly understood at the molecular level, with few reliable biomarkers or therapeutic targets. This study takes an epigenomic approach, leveraging DNA methylation (DNAm) data from skeletal muscle tissue collected before and after bariatric surgery to identify conserved methylation signatures associated with weight loss.
The researchers analyzed paired skeletal muscle DNAm profiles from two GEO datasets (GSE135063, n=32; GSE272137, n=26) using the Illumina EPIC array, which covers over 850,000 CpG sites. Differential methylation analysis focused on individuals without type-2 diabetes, finding that changes were largely conserved across sexes (Jaccard index ~0.26), with the majority of probes showing hypermethylation before surgery and most changes occurring in 'open sea' genomic regions outside CpG islands. Probe consistency was rigorously evaluated using GTEx multi-tissue data and technical replicates, filtering candidates by intraclass correlation coefficients above 0.4.
From approximately 600 coding genes associated with the candidate probes, four emerged as high-priority targets: KCNB1 (a voltage-gated potassium channel), PEAK1 (a pseudokinase involved in cell signaling), SCG5 (a neuroendocrine secretory protein), and TNIK (a kinase implicated in Wnt signaling). These were validated through mouse knockout phenotype data focusing on growth, adipose, and lean body mass phenotypes, and cross-referenced against druggability databases showing existing small-molecule or approved drug coverage.
To generalize these findings to cancer, the team applied a weighted pathway enrichment score — computed via elastic net regression using the gelnet R package — to CPTAC pan-cancer data spanning five tumor types (LSCC, LUAD, HNSCC, ccRCC, PDAC). The enrichment score showed a positive correlation with chromosomal instability (CIN) scores in tumor tissues, suggesting that the same epigenetic dysregulation driving weight-loss-associated methylation changes may also reflect genomic instability in tumors. Co-methylated blocks on chromosome 19q12-13.12 were identified in samples with copy number gains, further linking DNAm patterns to somatic genomic alterations.
As a functional validation step, SCG5 protein levels were measured by ELISA in four lung cancer cell lines (H292, A549, H226, H520), confirming detectable and variable protein abundance. The study acknowledges limitations including a training cohort restricted to white European adults who remained obese at 52 weeks post-surgery, and the exploratory nature of the chromosome 19 co-methylation analysis. Nonetheless, the integration of epigenomics, proteomics, and mouse phenotypic data provides a multi-layered framework for identifying actionable targets at the intersection of obesity, weight loss, and cancer biology.
Key Findings
- KCNB1, PEAK1, SCG5, and TNIK identified as key methylation targets linked to weight loss and druggable in cancer.
- Pathway enrichment score from weight-loss methylation probes positively correlates with chromosomal instability in tumors.
- Most weight-loss-associated methylation changes occur in open sea regions, not CpG islands, challenging common assumptions.
- Co-methylated blocks on chromosome 19q12-13.12 identified in tumor samples with copy number gains.
- SCG5 protein expression confirmed across lung cancer cell lines, supporting translational relevance.
Methodology
Paired skeletal muscle DNAm data from two bariatric surgery cohorts (n=58 total) were analyzed using the Illumina EPIC array. Probe consistency was evaluated via intraclass correlation in GTEx multi-tissue data; a weighted enrichment score was generated using elastic net regression and applied to CPTAC pan-cancer methylation data across five tumor types.
Study Limitations
The training cohort was limited to white European adults who remained obese at 52 weeks post-surgery, limiting generalizability. The chromosome 19 co-methylation analysis is exploratory, and causal relationships between methylation changes and weight loss phenotypes remain to be established experimentally.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.