Abstract
Type 2 diabetes is a risk factor for colorectal cancer (CRC) development and progression. However, metformin-treated diabetic CRC patients tend to have better clinical outcomes than those managed by other means. To better characterize the molecular underpinnings of metformin's protective effects, we performed a targeted transcriptomic analysis of primary CRC tissue samples (n = 272). A supervised learning algorithm pinpointed molecular features that discriminate between metformin-treated and diet-controlled diabetic CRC samples, as well as those that discriminated between non-diabetic samples based on their five-year overall survival status. Our results show downregulation of TMEM132 in metformin-treated samples (p = 0.05) and non-diabetics with good clinical outcomes (p = 0.05) relative to diet-controlled and non-diabetics with poor survival, respectively. Furthermore, upregulation of SCNN1A is observed in metformin-treated samples (p = 0.04) and non-diabetics with good clinical outcomes (p = 0.01) relative to diet-controlled samples and those with poor clinical outcomes, respectively. We also show that the antiapoptotic protein sFas is downregulated in metformin-treated samples relative to diet-controlled samples (p = 0.005). These findings suggest a role for the unfolded protein response in mediating metformin-related CRC-protective effects by enhancing apoptosis and suggest the investigation of these proteins as targets for novel CRC therapies.