Gregory G. Fuller, M. Freeberg, Ting Han
Apr 1, 2020
The FASEB Journal
Cells deprived of oxygen rely on glycolysis alone for energy production, generating far less energy per molecule of glucose than aerobic respiration. While modulating levels of glycolysis enzymes and their activity can increase the rate of glycolysis, spatial reorganization of pathway components may also contribute to enhancing rates of glycolysis. We have demonstrated that in S. cerevisiae, glycolysis enzymes, including the rate limiting phosphofructokinase (PFK), coalesce into single non‐membrane‐bound cytoplasmic foci called “glycolytic (G) bodies 1 ” in hypoxia. Cells lacking the AMP activated protein kinase, SNF1, fail to form single foci and are enriched for upstream glycolysis metabolites, while being depleted for the downstream metabolite, acetyl CoA. Furthermore, cells preconditioned in hypoxia (thereby forming G bodies) deplete glucose from media at a faster rate than those cultured in normoxia (lacking G bodies). Thus, the presence of G bodies is correlated with greater rates of glycolysis. However, we know very little about how G bodies form. Many glycolysis enzymes moonlight as RNA binding proteins despite lacking known RNA binding domains, and proteomic analysis of purified G bodies revealed an enrichment of RNA binding proteins, suggesting that G bodies are ribonucleoprotein (RNP) granules. Tethering PFK to nonspecific endoribonucleases to degrade RNA at sites of G body formation resulted in complete loss of G body formation, indicating that RNA nucleates nascent G bodies. Induction of PFK‐RNase in cells with existing G bodies leads to fragmentation into multiple foci, indicating that RNA is required to maintain G body integrity. Consistent with RNA nucleating G body formation, RNAs bound to glycolysis enzymes in normoxic conditions identified by PAR‐CLIP strongly overlapped with RNAs that copurify with G bodies in hypoxia. Many RNP granules form via phase separation of protein and RNA. Although G bodies are distinct from stress granules, G bodies exhibit similar physical properties to stress granules, which are thought to form via phase separation. Proteomic analysis revealed similar protein components as stress granules including HSP70 chaperones. In addition, we have shown that G bodies in mating yeast cells have the ability to fuse and do so during their biogenesis. Furthermore, FRAP revealed that G body components exchange with the cytoplasm. G body fusion, however, takes multiple minutes to occur and FRAP recovery rates occur with half times of greater than 10 minutes, indicating that G bodies behave more akin to gels than a liquid‐liquid phase condensate. Taken together, we propose that G bodies are novel RNP granules that function to enhance rates of glycolysis in response to energy limiting hypoxic stress. Future studies will clarify the mechanisms of G body function and further test their physiological roles in cells.