Sugar makes fat by talking to SCAP

Abstract

Elevated blood glucose level promotes lipogenesis via activating SREBP transcription factors. Tumors exhibit increased glucose uptake and lipogenesis, but the mechanisms controlling this are poorly understood. In this issue of Cancer Cell, Cheng and colleagues find that glucose activates SREBP by stabilizing SCAP, a central regulator of the SREBP pathway.

Alteration of cellular metabolism is a hallmark of cancer. Tumors exhibit increased aerobic glycolysis and elevated lipogenesis to supply energy and lipids, such as fatty acids and cholesterol, required for cell growth (Menendez and Lupu, 2007). In addition to its role in energy production, glucose is a precursor for lipid synthesis. In tissues such as liver and adipose, insulin transduces a signal from glucose to the sterol regulatory element-binding proteins (SREBPs) to induce lipogenesis and promote energy storage (Shao and Espenshade, 2012). Although known signaling cascades downstream of oncogenic drivers such as RAS or TOR stimulate lipogenesis (Ricoult et al., 2015), whether glucose supply independently activates lipogenesis is unclear.

SREBP transcription factors are central regulators of cellular lipogenesis (Shao and Espenshade, 2012). These endoplasmic reticulum (ER) membrane-bound transcription factors bind to the SREBP cleavage activating protein (SCAP), which controls their activity (Fig. 1). Under repressed conditions, SCAP binds to the ER-resident protein INSIG, and the SCAP-SREBP complex is retained in the ER. When lipid supply is low, the SCAP-SREBP complex traffics to the Golgi where two proteases release the N-terminal transcription factor domain of SREBP allowing it to enter the nucleus and stimulate gene expression. SREBP family members activate synthesis and uptake of fatty acids, triglycerides, and cholesterol. Although SCAP is essential for SREBP activation, few regulatory inputs are known (Shao and Espenshade, 2012). SCAP responds directly to levels of cholesterol in the ER, which serves as an end product, feedback signal from cholesterol synthesis (Fig. 1). Cholesterol binds SCAP, promoting INSIG binding and ER retention. In addition, SCAP levels decrease when SREBP cleavage is blocked in the Golgi through a lysosome-dependent pathway (Shao and Espenshade, 2014). However to date, no major nutritional regulation of SCAP has been reported.

Figure 1. Glucose-induced activation of SREBP and lipogenesis.

Glucose promotes lipogenesis by (1) generating acetyl-CoA, the substrate for lipogenesis, (2) stimulating pancreatic insulin secretion, which activates SREBP-dependent lipogenic gene expression, and (3) generating UDP-GlcNAc required for SCAP N-glycosylation. Glycosylation of SCAP prevents its degradation and promotes disassociation from the ER-reside protein INSIG, resulting in SREBP activation. Abbreviations are as follows: insulin-induced gene, INSIG; uridine diphosphate N-acetylglucosamine, UDP-GlcNAc; acetyl coenzyme A, acetyl-CoA.

SREBP function is best studied in the liver where SREBPs contribute to whole body lipid homeostasis. More recently, knowledge that tumors are highly lipogenic has led investigators to examine the function of SREBP in cancer with efforts initially focusing on glioblastoma (GBM). These studies demonstrate that the EGFR-Akt pathway stimulates SREBP-dependent lipogenesis and that SREBP is required for GBM cell survival and tumor growth (Griffiths et al., 2013; Guo et al., 2009). Together with work in breast cancer cells (Ricoult et al., 2015), these results highlight an important role for the SREBP pathway in tumor growth, making it an interesting therapeutic candidate.

In this issue of Cancer Cell, Cheng et al. build on their previous studies and look for connections between glucose supply and elevated lipogenesis observed in GBM (Cheng, 2015). The authors discover that glucose has an insulin-independent pathway to control SCAP and promote lipogenesis. Their initial observation was that SREBP activation requires glucose in GBM cells and that SCAP levels decrease dramatically in low glucose. Mechanistic experiments show that glucose controls SCAP due to its requirement as a precursor of N-acetylglucosamine in N-glycosylation and that low glucose reduced SCAP glycosylation. Inhibitors of N-linked, but not O-linked, glycosylation decrease SCAP levels and prevent SREBP activation. Earlier studies by the Brown and Goldstein laboratories showed that SCAP is an 8 transmembrane protein that contains three N-glycosylation sites in its luminal loops (Fig. 1). The presence of any one site is sufficient to support SREBP activity (Nohturfft et al., 1998). Here, the authors demonstrate that mutation of all three sites leads to proteasome-dependent degradation of the non-glycosylated SCAP mutant (SCAP-QQQ) and a failure to activate SREBP. Interestingly, wild-type, glycosylated SCAP showed reduced binding to the negative regulator INSIG compared to SCAP-QQQ, suggesting that glycosylation status also may influence INSIG retention of SCAP (Fig. 1).

EGFR signaling stimulates glucose uptake and activates SREBP in GBM cells (Guo et al., 2009). The authors next addressed whether these effects required glucose. Glucose activated SREBP and increased SCAP in the absence of growth factor. EGF had no effect on SREBP activity alone, but addition of glucose strongly increased SCAP and activated SREBP. As expected, SCAP was required for EGF activation of SREBP, and inhibition of N-glycosylation both reduced SCAP and prevented EGF-stimulated activation of SREBP. These results suggest that one function of EGFR signaling is to increase glucose uptake to support SCAP glycosylation.

Previous studies showed the importance of lipogenesis and SREBP activation for GBM tumor growth (Griffiths et al., 2013; Guo et al., 2009). The authors investigated the role of SCAP glycosylation in vivo using both subcutaneous and orthotopic xenograft mouse models. Knockdown of SCAP decreased tumor weight and prolonged survival of intracranial tumor-bearing mice. To investigate SCAP glycosylation specifically, the authors overexpressed either wild-type SCAP or the SCAP-QQQ mutant in GBM cells. Interestingly, overexpression of wild-type SCAP increased tumor weight relative to control and decreased survival in the orthotopic model. Overexpression of SCAP-QQQ had dominant effects, decreasing tumor weight and prolonging survival consistent with its inability to activate SREBP-dependent lipogenesis. These findings are significant and demonstrate both that SCAP glycosylation is required to support tumor growth and, perhaps more importantly, that SCAP is limiting for tumor growth and pathogenesis in vivo.

Collectively, these experiments outline a pathway by which the requirement of glucose for SCAP N-glycosylation permits glucose supply to control SREBP-dependent lipogenesis. Growth factor signaling stimulates glucose uptake for energy production, but this study describes another mechanism by which glucose additionally drives lipogenesis, adding to our understanding of why these two processes coexist in tumorigenesis.

This study raises a number of important questions that require further investigation. These experiments show that low glucose regulates SCAP protein levels, but to what extent do physiological changes in glucose influence SCAP glycosylation and function? Blood glucose concentrations are tightly controlled and such low glucose may only occur in poorly vascularized tumors. Does this signaling pathway function in normal tissues and do non-mitotic cells such as hepatocytes respond to glucose in a similar manner? Furthermore, does elevated glucose regulate SCAP in the setting of diabetes? Could this mechanism explain links between diabetes and disease progression, in particular cancers? Finally, cancer cells frequently increase glutamine uptake, and glutamine can also serve as a precursor for N-acetylglucosamine (Metallo and Vander Heiden, 2010). Can glutamine supply also signal to SCAP and lipogenesis?

This paper highlights the importance of lipogenesis in tumor growth, and identifies SCAP as a key regulator in this context. To date, the requirement of the SREBP pathway for tumor growth has been studied in GBM, breast, and prostate cancer cells (Krycer et al., 2012). It will be essential to expand these studies to see how general the requirement for the SREBP pathway is. To this point, the authors demonstrate that glucose controls SCAP levels in cell lines of multiple cancer types.

Finally, this study reports the first nutritional stimulus to control SCAP. From a mechanistic standpoint, it will be important to know the molecular machinery required for proteasome-dependent degradation of non-glycosylated SCAP. Activation of a required E3 ubiquitin ligase could serve as a therapeutic strategy to down-regulate SCAP and inhibit lipogenesis. In addition, while general glycosylation inhibitors are likely toxic, specific inhibitors of SCAP glycosylation may block lipogenesis. Chemical inhibitors of SCAP will impact not only cancer treatment but also other prevalent metabolic diseases such as nonalcoholic fatty liver disease that requires SCAP, making the development of specific SCAP inhibitors a priority.