Activation of a metabolic gene regulatory network downstream of mTOR complex 1

Abstract

Aberrant activation of the mammalian target of rapamycin complex 1 (mTORC1) is a common molecular event in a variety of pathological settings, including genetic tumor syndromes, cancer, and obesity. However, the cell-intrinsic consequences of mTORC1 activation remain poorly defined. Through a combination of unbiased genomic, metabolomic, and bioinformatic approaches, we demonstrate that mTORC1 activation is sufficient to stimulate specific metabolic pathways, including glycolysis, the oxidative arm of the pentose phosphate pathway, and de novo lipid biosynthesis. This is achieved through the activation of a transcriptional program affecting metabolic gene targets of hypoxia-inducible factor (HIF1alpha) and sterol regulatory element-binding protein (SREBP1 and SREBP2). We find that SREBP1 and 2 promote proliferation downstream of mTORC1, and the activation of these transcription factors is mediated by S6K1. Therefore, in addition to promoting protein synthesis, mTORC1 activates specific bioenergetic and anabolic cellular processes that are likely to contribute to human physiology and disease.

Figure 1. A genomics approach to identify mTORC1-regulated transcripts

A) Model of mTORC1 regulation downstream of growth factors. B) Growth factor-independent mTOR signaling in Tsc1^−/− and Tsc2^−/− MEFs. MEF lines were serum starved for 24 h in the presence of vehicle (−) or 20-nM rapamycin (12 or 24 h) and, where indicated, were stimulated with 10% serum for the final 1 h. C) Hypothetical Venn diagram of the four major changes in gene expression detected in this study. mTORC1-regulated transcripts were classified as those that met all four of these criteria at a statistical cut of p<0.01 (*). D) Scatter plot of the expression levels (log₂) of the 39,000 gene probes comparing vehicle-treated Tsc2^−/− cells to vehicle-treated wild-type controls (x-axis) and vehicle-treated Tsc2^−/− cells to rapamycin-treated (24 h) Tsc2^−/− cells (y-axis). The larger blue dots depict the pattern of the 239 gene probes meeting the criteria described in (B). The gray dots depict the expression pattern of the whole dataset. E) Heat map of the 239 probes found to be regulated by mTORC1 signaling in this study, showing their expression levels and response to a rapamycin time course in Tsc2^−/− cells. Expression levels shown are representative of the log₂ average obtained from independent replicates per time point. The brightness of green and red represents the degree to which expression is respectively lower or higher in the Tsc2^−/− cells relative to vehicle treated wild-type cells. The gene names and values are provided in supplemental Table S1. (See also heat map for the same genes in Tsc1^−/− cells in Figure S1)

Figure 2. Transcriptional induction of metabolic genes downstream of mTORC1 activation

A–C) mTORC1 activation leads to the induction of genes encoding the enzymes of glycolysis (A), the pentose phosphate pathway (B), and lipid and sterol biosynthesis (C) in Tsc1^−/− (upper graphs) and Tsc2^−/− (lower graphs) cells. The log₂ expression levels shown are the average obtained from independent replicates per time point of rapamycin treatment normalized to the expression levels in vehicle-treated wild-type cells. Values provided as supplemental Table S3. D–F) Confirmation of mTORC1-induced metabolic genes. The expression levels of representative genes from (D) glycolysis, (E) the pentose phosphate pathway, and (F) lipid and sterol biosynthesis was measured by qRT-PCR. MEFs were serum starved for 18 h in the presence of vehicle (−) or 20-nM rapamycin (+) and, where indicated, were subjected to a proliferation arrest via double thymidine block (DTB). Expression levels are presented as mean±SD relative to vehicle-treated Tsc2^+/+ cells and are representative of at least three independent experiments. (See supporting data in Figure S2)

Figure 3. Alterations in cellular metabolism induced by mTORC1 activation

A) Placement of the mTORC1-induced metabolic genes within their metabolic pathways. Genes found to be regulated by mTORC1 are shown in red, with those in the most stringent set indicated (*). B) mTORC1 signaling stimulates glucose uptake. MEFs were serum starved for 16 hours in the presence of vehicle or rapamycin (20 nM). Glucose uptake was measured as the incorporation of 2-deoxy-D-[³H]-glucose over 4 minutes and normalized to cell number. Levels are presented as mean±SD relative to Tsc2^+/+ cells from four independent experiments. *p<0.002 versus Tsc2^+/+; **p<0.009 versus Tsc2^−/−. C) mTORC1 signaling increases lactate production. Cells were grown as in (B), and lactate secretion into the media was measured over the final hour. Lactate concentrations normalized to cell number are presented as mean±SD relative to Tsc2^+/+ cells from three independent experiments. *p<0.00001 versus Tsc2^+/+; **p<0.006 versus Tsc2^−/−. D) mTORC1 signaling stimulates de novo lipid biosynthesis. MEFs were serum starved for 24 h in the presence of vehicle or rapamycin (20 nM) and were incubated with D-[6-¹⁴C]-glucose for the final 4 h. ¹⁴C incorporation into the lipid fraction was measured and is presented as mean±SD relative to Tsc2^+/+ cells. These data are representative of three independent experiments. *p<0.006 versus Tsc2^+/+; **p<0.004 versus Tsc2^−/−. E) Metabolomic profiling demonstrates that mTORC1 activation increases metabolites of glycolysis and the pentose phosphate pathway. MEFs were grown as in (B) and metabolites were extracted and profiled by LC-MS. Relative levels of specific metabolites, normalized to cell number, from three independent samples for each cell line and treatment are shown in the heat map. Metabolites showing a pattern consistent with mTORC1 regulation are indicated (*). The complete metabolomic profile is provided as supplemental Figure S3 and Table S4. F) mTORC1 signaling increases flux through glycolysis. MEFs were grown as in (B) and incubated with [1,2-¹³C]-glucose for 15 min prior to metabolite extraction and LC-MS analysis. Levels of dually ¹³C-labeled glycolytic intermediates, normalized to cell number, are presented as mean±SD over three independent samples. G) mTORC1 signaling increases flux through the oxidative branch of the pentose phosphate pathway. The ratio of singly (1×) and doubly (2×) ¹³C-labeled to unlabeled (¹²C) pentose phosphate pathway metabolites were measured by LC-MS in the samples from (F). P-values for pair-wise comparisons in (F) and (G) are listed in the accompanying table.

Figure 4. HIF1α and its target genes are upregulated by mTORC1 signaling

A) HIF1 target genes are elevated and rapamycin sensitive in Tsc1 (upper graph) and Tsc2 (lower graph) null cells. The log₂ expression levels shown are the average obtained from independent replicates per time point of rapamycin treatment normalized to the expression levels in vehicle-treated wild-type cells, with those in the most stringent set indicated (*). (Values provided in supplemental Table S3). B) HIF1α protein levels are shown in wild-type and either Tsc1 or Tsc2 null MEFs grown in serum-free media and subjected to a time course of rapamycin (20 nM). *non-specific band. (See also Figures S4A–C) C) Induction of translation from the 5'-UTR of HIF1α is regulated by 4E-BP1. The bicistronic reporter was cotransfected with empty vector (Vec), TSC2, or 4E-BP1-AA (T37A/T46A). 24 h post-transfection, cells were serum starved for 16 h in the presence or absence of rapamycin (Rap, 20 nM), and the ratio of Renilla to firefly luciferase was measured. The ratios are presented as mean±SD relative to Tsc2^+/+ cells and are representative of three independent experiments. (See also Figures S4D–F) D, E) mTORC1-dependent increase in the expression Hif1α (D) and its glycolytic gene targets (E). MEFs were transfected with siRNAs targeting the indicated transcripts or control non-targeting siRNAs (ctl). 32 h post-transfection, the cells were serum starved for 16 hours in the presence or absence of rapamycin (Rap, 20 nM). Transcript levels, measured by qRT-PCR, are presented as the mean±SD relative to levels in Tsc2^+/+ (Ctl) cells over three independent experiments. (See also Figure S4G) F) The stimulation of glucose uptake downstream of mTORC1 is dependent on HIF1α. MEFs were transfected with siRNAs and treated as in (D). Glucose uptake was measured as the incorporation of 2-deoxy-D-[³H]-glucose over 4 min and normalized to cell number. Levels are presented as mean±SD relative to Tsc2^+/+ (Ctl) cells over six independent experiments. *p<0.003 for Tsc2^+/+ versus Tsc2^−/− and Tsc2^−/− versus rapamycin-treated Tsc2^−/−; **p<0.03 for Tsc2^−/− versus Tsc2^−/− with Hif1α knockdown.

Figure 5. mTORC1 signaling activates SREBP1 to induce de novo lipid biosynthesis and growth factor-independent proliferation

A) Levels of full length (fl) and processed (p) SREBP1 are shown from littermate-derived pairs of wild-type and either Tsc1 or Tsc2 null MEFs grown in serum-free media and subjected to a time course of rapamycin (20 nM). (Below) The ratio of processed to full length SREBP1 at the indicated time points was quantified and plotted relative to untreated wild-type cells. B) MEFs were transfected with control non-targeting siRNAs (ctl) or siRNAs targeting Raptor or Rictor, and 48 h post-transfection, cells were serum starved for 24 hours. C, D, E) The mTORC1-induced expression of genes involved in fatty acid (C) and sterol (D) biosynthesis and the pentose phosphate pathway (E) is dependent on SREBP. MEFs were transfected with control non-targeting siRNAs (ctl) or siRNAs targeting Raptor, Srebp1 (Sr1), Srebp2 (Sr2), or both (Sr1/2). 32 h post-transfection, cells were serum starved for 16 h in the presence or absence of rapamycin (Rap, 20 nM). The expression levels of representative genes were measured by qRT-PCR and are presented as the mean±SD relative to levels in Tsc2^+/+ cells over three independent experiments. (See also Figures S5A and B) F) Overexpression of Rheb or the processed form of SREBP1a induces G6PD expression. HEK-293 cells were transiently transfected with empty vector (Vec), FLAG-RHEB, or FLAG-processed SREBP1a. 24 h post-transfection, cells were serum starved for 16 h in the presence of vehicle or rapamycin (20 nM). G6PD mRNA levels, measured by qRT-PCR, are presented as the mean±SD relative to levels in vector-transfected cells over three independent experiments. (See also Figures S5C and D) G) The stimulation of de novo lipid biosynthesis downstream of mTORC1 is dependent on SREBP. MEFs were transfected with siRNAs targeting the indicated transcripts or control non-targeting siRNAs (ctl). 24 h post-transfection, cells were serum starved for 24 h in the presence of vehicle or rapamycin (20 nM) and were incubated with D-[6-¹⁴C]-glucose for the final 4 h. ¹⁴C incorporation into the lipid fraction was measured and is presented as mean±SD relative to Tsc2^+/+ (Ctl) cells. These data are representative of three independent experiments. *p<0.004 for Tsc2^+/+ versus Tsc2^−/− and Tsc2^−/− versus rapamycin-treated Tsc2^−/−; **p<0.008 versus Tsc2^−/− (ctl); ***p<0.002 versus Tsc2^−/− (ctl). A control immunoblot is provided with numbering matching the samples from the graph. (See also Figure S5E) H) The mTORC1-driven proliferation of Tsc2 null cells in the absence of growth factors is dependent on SREBP. Tsc2^+/+ (T2+/+) and Tsc2^−/− (T2−/−) MEFs were transfected with siRNAs targeting indicated genes or non-targeting control siRNAs (siCtl). 24 h post-transfection, 1×10⁵ cells per well were re-seeded and grown in serum-free medium in the presence of vehicle or rapamycin (20 nM, Rap), with daily medium changes. Triplicate samples were counted every 24 h, and the mean cell numbers ±SD for a representative experiment are shown. (See also Figure S5F)

Figure 6. mTORC1 regulates SREBP1 through S6K1 but independent of GSK3 and effects on proteasomal degradation

A) The inhibitory effects of rapamycin on processed SREBP1 (SREBP1(p)) are independent of proteasomal degradation. MEFs were serum starved for 24 h in the presence of vehicle or 20-nM rapamcyin (12 or 24 h) and, where indicated, were treated with ALLN (25 μg/mL) for the final 1 h. Light and dark exposures of the same blot of SREBP1(p) are shown. B) Involvement of S6K1, but not GSK3, in the mTORC1-dependent effects on SREBP1. MEFs were transfected with siRNAs targeting the indicated transcripts or control non-targeting siRNAs (ctl). 24 h post-transfection, cells were treated as in (A) with ALLN. C) S6K1 regulates expression of SREBP1 target genes. MEFs were transfected with the indicated siRNAs, and 32 h post-transfection, cells were serum starved for 16 h in the presence or absence of rapamycin (Rap, 20 nM). Transcript levels were measured by qRT-PCR and are presented as the mean±SD relative to levels in Tsc2^+/+ (Ctl) cells over three independent experiments. D) S6K1 contributes to stimulation of de novo lipid biosynthesis downstream of mTORC1. Cells were transfected with the indicated siRNAs and serum starved for 24 h. For the final 4 h, cells were labeled with D-[6-¹⁴C]-glucose. ¹⁴C incorporation into the lipid fraction was measured and is presented as mean±SD relative to Tsc2^+/+ (Ctl) cells. *p<0.003 versus WT; **p<0.03 versus Tsc2^−/− (ctl). E) Model of the control of specific metabolic pathways downstream of mTORC1. See text for details.