Non-proteolytic ubiquitination of Hexokinase 2 by HectH9 controls tumor metabolism and cancer stem cell expansion

Abstract

Enormous efforts have been made to target metabolic dependencies of cancer cells for developing new therapies. However, the therapeutic efficacy of glycolysis inhibitors is limited due to their inability to elicit cell death. Hexokinase 2 (HK2), via its mitochondrial localization, functions as a central nexus integrating glycolysis activation and apoptosis resilience. Here we identify that K63-linked ubiquitination by HectH9 regulates the mitochondrial localization and function of HK2. Through stable isotope tracer approach and functional metabolic analyses, we show that HectH9 deficiency impedes tumor glucose metabolism and growth by HK2 inhibition. The HectH9/HK2 pathway regulates cancer stem cell (CSC) expansion and CSC-associated chemoresistance. Histological analyses show that HectH9 expression is upregulated and correlated with disease progression in prostate cancer. This work uncovers that HectH9 is a novel regulator of HK2 and cancer metabolism. Targeting HectH9 represents an effective strategy to achieve long-term tumor remission by concomitantly disrupting glycolysis and inducing apoptosis.

Fig. 1

Tumor-associated HectH9 is a novel regulator of glucose metabolism. a cBioportal was used to access TCGA data for HectH9 (Huwe1) gene alteration from four different studies of human prostate cancer. Frequencies of HectH9 gene amplification are shown in red. (b) Cell viability assay in PC-3 cells stably infected with lentiviruses expressing shRNA targeting GFP, Myc or HectH9 in glucose-containing and -deprived medium (n = 3). c The extracellular acidification rate (ECAR), was measured by using a Seahorse Bioanalyzer in PC-3 cells infected with lentiviruses expressing shRNA targeting GFP or HectH9 (n = 15). d–j Relative abundance of ¹³C-labeled glucose-6-phosphate (d), pyruvate (e), lactate (f), α-ketoglutarate (g), succinate (h), fumarate (i) and malate (j) were determined by GC-MS in PC-3 cells infected with lentiviruses expressing shRNA targeting GFP or HectH9 (n = 3). Results in b–j are presented as mean value ± SD; *p < 0.05, **p < 0.01, by Student’s t-test. Experiments were performed at least twice. k Schematic illustration depicts the metabolic changes (labeled in red) in glycolysis pathway and oxidative phosphorylation (OXPHOS) regulated by HectH9

Fig. 2

HectH9 promotes K63-linked ubiquitination of HK2. a PC-3 cell lysates were harvested for immunoprecipitation (IP) with IgG or anti-HectH9 antibody, followed by immunoblot (IB) analysis with indicated antibodies. The ratio of immunoprecipitated HK2 to total HK2 levels in the input was quantified using ImageJ software. Arrowhead indicates the full-length of HectH9 and signals underneath represent HectH9 fragments. b PC-3 cell lysates were harvested for IP with IgG or anti-HK2 antibody, followed by IB analysis with indicated antibodies. c and d In vivo ubiquitination assay of HK2 by HectH9. HEK293T cells transfected with indicated plasmids were harvested for ubiquitination assay. Cells were lysed by denatured buffer. The ubiquitinated proteins were precipitated with nickel (Ni)-NTA beads and subjected to IB analysis. NTA indicates nickel bead precipitate; WCE indicates whole-cell extracts. e HEK293 cells with Luciferase or HectH9 knockdown were transfected with indicated constructs and harvested for in vivo ubiquitination assay. f Purified catalytically active form (WT) or defective mutant (C4341A) of GST-HectH9 proteins were incubated with adenosine triphosphate, E1 (Ube1) and E2 (Ubc13), with or without purified Flag-HK2 for in vitro HK2 ubiquitination assay. g HEK293T cells transfected with indicated constructs were harvested for in vivo ubiquitination assay. h IB analysis of HK2 and HK1 expression in PC-3 cells with GFP or HectH9 knockdown. Two HectH9 lentiviral shRNAs were used in this assay. The relative intensity of protein expression as indicated was quantified with ImageJ software and normalized to α-Tubulin expression. Immunoblots were performed three times

Fig. 3

HectH9 regulates HK2 localization to mitochondria and HK2-mediated glycolysis. a Immunofluorescence image for HK2/VDAC colocalization in Luciferase-knockdown and HectH9-knockdown PC-3 cells (from three biological replicates). VDAC and DAPI was used as markers for mitochondria and nucleus, respectively. White arrowheads show co-localization of HK2/VDAC at mitochondria. Scale bar represents 10 μm. b IB analysis of HectH9 expression in PC-3 cells with Luciferase or HectH9 knockdown (upper). Quantification analyses of HK2/VDAC co-localization in PC-3 cells with Luciferase or HectH9 knockdown (lower). The lower left panel shows the percentage of counted cells with HK2/VDAC colocalization. 17 random fields of view in shLuc and shHectH9 groups that account for 218 and 226 cells, respectively, were counted. The lower right panel shows that the HK2/VDAC correlation per cell was determined using the Pearson’s correlation coefficient. 102 cells were counted in each group from three biological replicates. c Biochemical fractionation for HK2 subcellular localization in Luciferase-knockdown and HectH9-knockdown PC-3 cells. VDAC and α-Tubulin were used as markers for mitochondrial and cytosolic fractions, respectively. The ratio of mitochondrial HK2 to total HK2 levels in WCE was quantified by ImageJ. d Lactate production in PC-3 cells with GFP, HectH9 or HK2 knockdown. (n = 3). e IB analysis for HectH9 and HK2 expression in PC-3 cells with GFP, HectH9, or HK2 knockdown. The relative intensity of protein expression as indicated was quantified with ImageJ and normalized to GRP78 or α-Tubulin expression. f Relative abundance of ¹³C-labeled glucose-6-phosphate, pyruvate and lactate were determined by GC-MS in PC-3 cells with GFP, HK2 or HectH9 knockdown (n = 4). The relatively abundance was normalized to that in GFP-knockdown cells. g and h Lactate production (g) and IB analysis (h) were carried out in HeLa cells transfected with indicated plasmids. i Lactate production was measured in HK2-depleted HeLa cells, followed by restoration of vector control (Vec), WT HK2 or WT HK2 plus HectH9-knockdown. (n = 3). Results in b, d, f, g and i are presented as mean value ±SD; *p<0.05, **p<0.01, by Student’s t-test. Experiments were performed at least twice in triplicates. Immunoblots were performed three times

Fig. 4

HectH9 deficiency triggers cellular apoptosis under glucose deprivation. a Biochemical fractionation assay for Bax mitochondrial localization in PC-3 cells with Luciferase or HectH9 knockdown is shown in the left panel. VDAC and α-Tubulin were used as markers for mitochondrial and cytosolic fractions, respectively. The relative intensity of mitochondrial-bound Bax quantified with ImageJ software and normalized to mitochondrial VDAC levels in the corresponding cells is shown in the right. The experiment was performed twice. b IB analysis of cleaved PARP1 expression in PC-3 cells with GFP or HectH9 knockdown. The relative intensity of cleaved PARP1 expression was quantified with ImageJ software and normalized to α-Tubulin expression. The experiment was performed twice. c and d The percentage of apoptotic cells was determined by annexin V and propidium iodide (PI) staining in Luciferase- knockdown and HectH9-knockdown PC-3 cells (c) or MDA-MB-231 cells (d) cultured in media with or without glucose (Glc) (n = 3). e The percentage of apoptotic cells was determined by annexin V and PI staining in Luciferase-knockdown and HectH9-knockdown PC-3 cells in the presence of vehicle or 2-DG (n = 3). f–h Cell growth assays in PC-3 (f), HeLa (g) and MDA-MB-231 (h) cells with Luciferase or HectH9 knockdown (n = 3). Results in c–h are presented as mean value ± SD; *p < 0.05, **p < 0.01, by Student’s t-test. These experiments were performed three times in triplicates. Immunoblots were performed three times

Fig. 5

K63-linked ubiquitination regulates the mitochondrial localization and functions of HK2. a In vivo ubiquitination assay in HEK293T cells transfected with or without His-Ub, along with WT, K21R, K104R, and K21/104 R mutants of GFP-HK2. b MDA-MB-231 cells infected with viruses expressing vector control (Vec), WT or the K21/104 R mutant of Flag-HK2 were fixed and subjected to immunofluorescence staining with indicated antibodies and DAPI. Representative images of HK2 and VDAC subcellular localization are shown in the upper panel. White arrowheads indicate that the K21/104 R mutant of HK2 diffuses in cytosol and is not co-localized with VDAC on mitochondrial outer membrane. Scale bar represents 10 μm. Quantification results for HK2/VDAC co-localization are shown in the lower panels. The lower left panel shows the percentage of counted cells with co-localized HK2 and VDAC. At least 100 cells were counted in each group from three biological replicates. The lower right panel shows that the HK2/VDAC correlation per cell was determined using the Pearson’s correlation coefficient. 102 cells were counted in each group from three biological replicates. c HK2 kinase activity was measured in HEK293T cells transfected with indicated plasmids. KD represents kinase dead (n = 3). Schematic illustration and quantification of the HK2 activity assay are shown in the upper and lower panels, respectively. IB analysis for protein expression of vector control, WT and various mutants of GFP-HK2 is shown in the right panel. The relative intensity of GFP-HK2 protein expression was quantified with ImageJ software and normalized to α-Tubulin expression. d and e Lactate production was measured in MDA-MB-231 (d) and HeLa cells (e) with HK2 knockdown, followed by ectopic expression of vector control, WT or the K21/104 R mutant of HK2 (n = 3). f The apoptotic cell percentage was determined by annexin V and PI staining in the absence or presence of glucose (Glc) in MDA-MB-231 cells with HK2 knockdown, followed by ectopic expression of vector control, WT or the K21/104R mutant of HK2 (n = 3). Results in b, d–f are presented as mean value ± SD; *p < 0.05, **p < 0.01, by Student’s t-test. Experiments were performed at least twice in triplicates. Immunoblots were performed three times

Fig. 6

Deficiency in HectH9 or HK2 inhibits CSC self-renewal via ROS production. a Representative image of flow cytometry analysis for ROS production in PC-3 cells with GFP or HectH9 knockdown. Green arrow indicates higher ROX level in HectH9-knockdown PC-3 cells. b Quantitative results for ROS production in PC-3 cells with GFP or HectH9 knockdown. Data were collected from three independent experiments. c Populations of CD44⁺/ALDH⁺ cells were determined by flow cytometry analysis in PC-3 cells with GFP or HectH9 knockdown (n = 3). d and e Tumor sphere formation assays in PC-3 (d) and HeLa (e) cells with Luciferase, HectH9 or HK2 knockdown (n = 3). Scale bar represents 100 μm. f Tumor sphere formation assay in HeLa cells infected with lentiviruses containing shRNA targeting the 3’-UTR of HectH9, followed by restoration with vector control, catalytically active form (WT) or defective mutant (C4341A) of HA-HectH9 (n = 3). Scale bar represents 100 μm. g Tumor sphere formation assay in HeLa cells infected with viruses expressing vector control or HK2 overexpression, followed by Luciferase or HectH9 knockdown (n = 3). Scale bar represents 100 μm. h Tumor sphere formation assay in PC-3 cells with Luciferase or HectH9 knockdown in the absence and presence of NAC (n = 3). Scale bar represents 100 μm. i Tumor sphere formation assay in PC-3 cells incubated with vehicle, 2-DG alone or 2-DG and BI8626 in combination for 12 days (n = 3). j and k Cell growth inhibition assay in PC-3 cells with Luciferase or HectH9 knockdown in the absence and presence of doxorubicin for 96 h (j) or paclitaxel for 48 h (k) (n = 3). Cell numbers were counted by using a hemocytometer. Results are presented as mean value ± SD; *p < 0.05, **p < 0.01, by Student’s t-test. Experiments were performed at least twice in triplicates

Fig. 7

HectH9 deficiency inhibits tumor metabolism and progression. a A schematic illustration depicting the process of metabolite profiling in HectH9-proficeint and HectH9-deficient PC-3-derived prostate tumors by GC-MS. b–d Relative abundances of ¹³C-labeled glucose-6-phosphate (b), pyruvate (c) and lactate (d) were quantified by GC-MS (n = 5). e Representative images of consecutive tissue sections for histological analyses of HectH9 expressions in normal tissues and prostate tumors with moderately or poorly differentiation status (n = 225). f In vivo primary tumor growth derived from PC-3 cells with Luciferase or HectH9 knockdown. Cells were injected subcutaneously into the right flanks of nude mice and tumorigenesis was monitored (n = 5 in each group). g In vivo primary tumor growth derived from MDA-MB-231 cells with HectH9 or HK2 depletion, as well as HK2-depleted MDA-MB-231 cells restored with vector alone (pBabe), WT or K21/104 R of Flag-HK2. Cells were injected into the mammary fat pads of nude mice and tumorigenesis was monitored. Quantitative results of tumor volumes at day 31 are shown in the left (n = 7 in each group). IB analysis of HectH9 and HK2 expression in MDA-MB-231 cells with GFP, HectH9 and HK2-depletion, as well as HK2-depleted cells restored with vector alone (pBabe), WT or K21/104 R of Flag-HK2 is shown in the right. h The working model depicts that hypoxia-induced HectH9, by promoting K63-linked ubiquitation of HK2, contributes to tumor progression via concomitant glycolysis elevation and apoptosis inhibition in cancer cells, as well as enhanced self-renewal of CSCs mediated by ROS blockade in tumors. Results in b–d, f and g are presented as mean value ± SEM; *p < 0.05, **p < 0.01, by Student’s t-test