The glycolytic shift in fumarate-hydratase-deficient kidney cancer lowers AMPK levels, increases anabolic propensities and lowers cellular iron levels

Abstract

Inactivation of the TCA cycle enzyme, fumarate hydratase (FH), drives a metabolic shift to aerobic glycolysis in FH-deficient kidney tumors and cell lines from patients with hereditary leiomyomatosis renal cell cancer (HLRCC), resulting in decreased levels of AMP-activated kinase (AMPK) and p53 tumor suppressor, and activation of the anabolic factors, acetyl-CoA carboxylase and ribosomal protein S6. Reduced AMPK levels lead to diminished expression of the DMT1 iron transporter, and the resulting cytosolic iron deficiency activates the iron regulatory proteins, IRP1 and IRP2, and increases expression of the hypoxia inducible factor HIF-1α, but not HIF-2α. Silencing of HIF-1α or activation of AMPK diminishes invasive activities, indicating that alterations of HIF-1α and AMPK contribute to the oncogenic growth of FH-deficient cells.

Figure 1. Mutations in FH compromised TCA cycle function and drove an oncogenic shift to aerobic glycolysis

(A) GC/MS analysis of TCA cycle intermediates (+/− SD) in HK-2, UOK262, and UOK262-WT cells. (B) Growth curves of UOK262 and UOK262-WT xenografts in nude mice (n = 10 for each group) as assessed by tumor volume. (C) Oxygen consumption rates (OCR) and extracellular acidification rates (ECAR, an indicator of glycolytic rate) in UOK262 and UOK268 cells compared to HK-2 cells upon changes from 0.5 mM to 25 mM glucose. The vertical blue lines A and B signify the time at which 25 mM glucose was infused. The ECAR plots were shown in percentage to demonstrate the fold enhancement in response to 25 mM glucose infusion. (D) Respiratory complex I activity and protein levels (+/− SD) in UOK262 cells. Results from non-HLRCC renal cell lines, HEK293, 786-0WT and UOK117C4, are presented as fold-differences compared to those of UOK262 cells. (E) Electron microscopy of mitochondrial morphology (arrows) in UOK262 cells and in non-HLRCC cells (UOK117C4 and 786-0WT) treated with the SDH inhibitor, 3-NPA. Scale bar, 0.5 µm. (F) Positron emission tomography (PET) with fluorodeoxyglucose (¹⁸FDG) in an HLRCC patient with advanced renal cancer demonstrated glucose uptake (black) in multiple metastases (arrow points to a representative hepatic metastasis). All data are representative of 3 or more independent experiments. Statistical analyses were performed by Student's t test for paired samples. See also Table S1.

Figure 2. Evidence for glycolysis-associated cytosolic iron deficiency and IRP activation in FH-deficient cancer cells

(A) Schematics for iron-dependent regulation of IRP1 and IRP2. When iron is abundant, IRP2 is degraded, whereas IRP1 is converted to cytosolic aconitase (c-ACO). When iron is scarce, IRP2 is stabilized whereas IRP1 loses its Fe-S cluster, and both IRPs bind to IREs with high affinity. (B) IRP activities in UOK262 and non-HLRCC cells (786-0WT and UOK117C4) cultured for 16–20 hr with or without 50 uM deferoxamine (Dfo) or 100 ug/mL ferric ammonium citrate as assessed by gel-shift assays. Immunoblots of tubulin were used as loading controls (C) Mitochondrial and cytosolic aconitase (m-ACO and c-ACO) activities in UOK262 and non-HLRCC cells (786-0WT and UOK117C4) were assessed by in-gel aconitase activity assays. (D) Immunoblots of IRP1 and IRP2 proteins in UOK262 and non-HLRCC cells (786-0WT and UOK117C4). (E) Immunoblots of TfR1 in HEK293, 786-0WT and UOK262 cells. (F) IRP activities in UOK262 cells cultured with different glucose and pyruvate levels. (G) Immunoblots of IRP2 protein in UOK262 cells and 786-0WT cells cultured in different concentrations of D-glucose. See also Figure S1.

Figure 3. Contribution of cytosolic iron deficiency and IRP activities to the misregulation of HIFα proteins in FH-deficient cells

(A) Immunohistochemistry of HIF-1α (dark brown staining) in healthy kidney tissue and HLRCC renal tumors. Scale bar, 30 µm (B) Immunoblots of HIF-1α and HIF-2α in UOK262 compared to HEK293, 786-0WT and UOK117C4 cells. (C) Immunoblots of HIF-1α and HIF-2α in UOK268 compared to HEK293 and 786-0WT cells. (D) Immunoblots of HIF-2α in IRP1+/+ and IRP1−/− MEF grown in normal media or treated with iron chelator DFO. (E) HIF-2α mRNA levels (+/− SD) in iron-replete (−Dfo) or iron-depleted (+Dfo) HEK293, 786-0WT, UOK117C4 and UOK262 cells as assessed by qRT-PCR. Data were normalized to actin expression. See also Figure S2.

Figure 4. Reduced AMPK levels resulted in activation of anabolic factors ACC and S6, and activation of IRPs in UOK262 and control cells

(A) Levels of AMPKα-p (T172) in healthy kidney tissue and in renal tumors from three patients with HLRCC-associated kidney cancer were assessed by immunohistochemistry (blue-black). Nuclei were stained with Methyl green. Scale bar, 30 µm. (B) Immunoblots of AMPKα-p (T172), ACC-p (S79) (upper panel) and AMPKα, AMPKβ, Akt-p (S473) (lower panel) in UOK262 compared to non-HLRCC cells (HK2, HEK293 and 786-0WT). (C) Levels of AMPKα1 and AMPKβ1 transcripts (+/− SD) in 786-0WT, UOK117C4, HEK293 and UOK262 cells were assessed by qRT-PCR. Data were normalized with actin transcript levels and presented as fold-differences compared to those of UOK262 cells. (D) GC/MS analysis of the AMP:ATP ratio in UOK262 cells and control HK-2 cells. (E) Immunoblots of S6-p (an mTOR downstream effector) in cells that were serum-starved for 2 days and then stimulated with serum for 30 min. (F) Immunoblots of AMPKα-p, ACC-p and IRP activities in 786-0WT cells that were cultured with or without AraA for 48 h or 72h. (G) Immunoblots of AMPKα-p, ACC-p and IRP activities in 786-0WT cells that were cultured with or without metformin (Metf) for 72 h. (H) IRP activities in 786-0WT cells treated with the AMPK activator AICAR for 72 h. (I) Immunoblots of IRP2 in AMPKα-deficient MEFs compared to wild-type MEFs. All data are representative of 3 or more independent experiments.

Figure 5. Reduced AMPK levels in UOK262 cells diminished the expression of the iron transporter DMT1, resulting in activation of IRPs

(A) Immunoblots of DMT1 (glycolysated forms at 70–100kDa) in UOK262 cells compared to HEK293 and 786-0WT cells. (B) Immunoblots of DMT1 in HEK293, UOK117C4 and UOK262 cells that were treated with iron chelator Dfo for 16 h. (C) Comparison of AMPKα and DMT1 levels in HEK293, 786-0WT, UOK262 and UOK262-WT cells. (D) Immunoblots of AMPKα-p and DMT1, and IRP activities in UOK262 cells cultured in different concentrations of glucose. (E) Immunoblots of ACC-p and DMT1, and IRP activities in 786-0WT cells treated with metformin (72 h), Dfo (16h) and/or FAC (16h). (F) Immunoblots of AMPKα-p and DMT1, and IRP activities in 786-0WT cells treated with AraA for 48h. (G) Immunoblots of AMPKα, DMT1 (unglycolysated form at ~55kDa) and IRP2 in AMPKα-deficient MEFs compared to wild-type MEFs. (H) Immunoblots of p53 in UOK262 cells compared to HEK293 and 786-0WT cells. (I) Effect of increased expression of p53 (induced by doxycycline treatment of p53ind cells) on DMT1 protein levels. (J) Effect of siRNA-mediated silencing of p53 on the levels of DMT1 in HEK293 cells. See also Figure S3.

Figure 6. A second glycolytic HLRCC cancer cell line, UOK268, exhibited decreased levels of AMPK, DMT1 and p53

(A) Immunoblots of AMPKα-p, ACC-p (an AMPK effector), DMT1 and IRP2 in UOK268 compared to HEK293 and 786-0WT cells. (B) Immunoblots of AMPKα, AMPKβ and DMT1 levels in UOK268 cells compared to HEK293, 786-0WT and UOK262 cells. (C) Immunoblots of p53 in UOK268 and UOK262 cells compared to HEK293, 786-0WT and UOK262 cells.

Figure 7. Chemical inhibition of SDH in non-HLRCC cells can induce metabolic changes similar to those observed in FH-deficient UOK262 cells

(A) IRP activities were assessed in 786-0WT, UOK117C4 and HEK293 cells that were treated with TTFA, an inhibitor of SDH, for 48 h. (B) Immunoblots of IRP2 and TfR1 in 786-0WT cells treated with TTFA. (C) IRP activities in 786-0WT cells treated with TTFA at different glucose concentrations. (D) Immunoblots of IRP2 and TfR1 in 786-0WT cells treated with TTFA at different glucose concentrations. (E) Time-dependent effect of 1 mM TTFA treatment on ATP levels (+/− SD) in HEK293 cells. (F) Time-dependent effect of 1 mM TTFA treatment on the levels of AMPKα and AMPKα-p in HEK293 and 786-0WT cells as assessed by immunoblot analysis. (G) Effect of TTFA treatment for 48h on the levels of AMPKα-p, AMPKα, AMPKβ, ACC-p, p53, and S6-p in HEK293 cells. (H) Effect of TTFA treatment of HEK293 and UOK117C4 cells for 48h on the levels of AMPKα-p, ACC-p, DMT1 and IRP activities. See also Figure S4.

Figure 8. Silencing of HIF-1α and activation of AMPK reduced the invasive potential of the FH-deficient UOK262 cells, which have a distinctive metabolic profile

(A) Immunoblots of HIF-1α, HIF-2α, AMPKα-p, AMPKβ, DMT1, IRP2 in UOK262 compared to 786-0 cells. (B) Effect of siRNA-mediated silencing of HIF-1α on the invasion activities of UOK262 cells measured using the RT-CIM cell invasion monitoring system. (C) Effect of treatment with the AMPK activators metformin and AICAR on the invasion activities of UOK262 cells. (D) A working model for FH-deficient cells, in which impairment of the TCA cycle imposes a need to shift energy production from respiration to glycolysis, and induces an AMPK-dependent decrease in p53 and activation of anabolic factors ACC and S6. Reduced AMPK levels decrease expression of the DMT1 iron transporter, resulting in reduced iron uptake and activation of IRPs. While reduced cytosolic iron levels and fumarate accumulation are expected to inhibit PHDs and stabilize both HIF-1α and HIF-2α proteins, activation of IRPs would selectively repress HIF-2α translation and attenuate the potential increase in HIF-2α proteins. The increased expression of HIF-1α and glycolytic shift, coupled to decreases of AMPK and p53 levels, and increased activity of anabolic pathways (shown in orange) may all confer growth/survival advantages to FH-deficient RCC.