Proteins with Evolutionarily Hypervariable Domains are Associated with Immune Response and Better Survival of Basal-like Breast Cancer Patients

Maltase-glucoamylase (MGAM) and MGAM2 both belong to the glycoside hydrolase family 31. MGAM, a therapeutic target for type 2 diabetes, is α-1,4-glucosidase and expressed in the intestine to catalyze starch digestion. MGAM2, however, is largely uncharacterized. By investigating The Cancer Genome Atlas data, we found that among breast cancer subtypes, MGAM2 expression is nearly exclusive to basal-like breast cancers (BLBCs), whereas MGAM tends to express in luminal A breast cancers. Moreover, MGAM2 expression is associated with better patient survival and correlated with immune genes/signatures, unlike MGAM. Both genes have emerged in mammals, but diverged after the placental-marsupial split. In placentals, MGAM2 has likely lost its α-1,4-glucosidase activity due to mutations in key catalytic sites, and has acquired a large domain that is extracellular, threonine-rich and evolutionarily hypervariable (EHV). Guided by MGAM2 findings, our genome-wide search identified >1000 human proteins with EHV regions. These proteins are enriched in immune functions and molecules, including major histocompatibility complex proteins. Their genes are expressed higher in BLBCs and are associated with better patient survival, like MGAM2. Their EHV-coding sequences are rich in simple repeats and harbor more cancer passenger mutations. In conclusion, MGAM2 diverges from MGAM structurally and likely functionally in placentals. MGAM2 is among >1000 human proteins with EHV regions and associated with immune response. We propose that these EHV molecules may have significant implication in cancer immunotherapy and BLBC treatment.