TKTL1 is activated by promoter hypomethylation and contributes to head and neck squamous cell carcinoma carcinogenesis through increased aerobic glycolysis and HIF1alpha stabilization

Abstract

Purpose: This study aims to investigate the role of the aberrant expression of Transkelolase-like 1 (TKTL1) in head and neck squamous cell carcinoma (HNSCC) tumorigenesis and to characterize TKTL1 contribution to HNSCC tumorigenesis through aerobic glycolysis and HIF1alpha stabilization.

Experimental design: TKTL1 promoter hypomethylation and mRNA/protein aberrant expression were studied in human HNSCC tumor samples and normal mucosas. Oncogenic functions of TKTL1 were examined in HNSCC cell line panels and tumor xenograft models with TKTL1 expression construct. The metabolite levels of fructose-6-phosphate, glyceraldehydes-3-phosphate, pyruvate, lactate, and the levels of HIF1alpha protein and its downsteam glycolytic targets were compared between the TKTL1-expressing and vehicle-expressing HNSCC cells. Meanwhile, the effects of HIF1alpha/glycolytic inhibitors were evaluated on the TKTL1 transfectants.

Results: TKTL1 exhibits high frequency of promoter hypomethylation in HNSCC tumors compared with the normal mucosas, correlating with its overexpression in HNSCC. Overexpression of TKTL1 in HNSCC cells promoted cellular proliferation and enhanced tumor growth in vitro and in vivo. Overexpression of TKTL1 increased the production of fructose-6-phosphate and glyceraldehyde-3-phosphate, in turn elevating the production of pyruvate and lactate, resulting in the normoxic stabilization of the malignancy-promoting transcription factor HIF1alpha and the upregulation of downstream glycolytic enzymes. Notably, the reduction of TKTL1 expression decreased HIF1alpha accumulation and inhibition with HIF1alpha and/or the glycolysis inhibitor could abrogate the growth effects mediated by TKTL1 overexpression.

Conclusion: TKTL1 is a novel candidate oncogene that is epigenetically activated by aberrant hypomethlation and contributes to a malignant phenotype through altered glycolytic metabolism and HIF1alpha accumulation.

Figure 1. TKTL1 is hypomethylated and overexpressed in HNSCC tumors

A, Hypomethylation of TKTL1 in HNSCC tumor tissues. QUMSP percentage of TKTL1 was conducted in a cohort of HNSCC cancer patients using 58 tumors and 11 mucosal samples to assay promoter demethylation. Each dot represents the relative mean levels of TKTL1 hypomethylation from triplicate experiments in the sample normalized by actin. N, normal mucosas, T, tumors. P<0.001. B, Quantitative RTPCR analysis of TKTL1 on 36 HNSCC tumors and 16 normal samples. N denoted normal mucosa, T, tumor. Each dot represented the relative mean expression of TKTL1 from triplicate experiments in the sample. P=0.047. C, left, Immunoblot of TKTL1 in 16 HNSCC tumor tissues and 7 normal mucosal samples and 6 cancer cell lines using a mouse monoclonal anti-TKTL1 antibody. Blots were probes with mouse monoclonal anti-β-actin antibody as a control for loading and transfer. Right, relative TKTL1 protein levels in all samples examined by western blot. Quantification of bands was performed using NIH ImageJ software. Asterisk, tumor samples with TKTL1 protein overexpression. Overexpression of TKTL1 were found in tumor sample T1, T2, T6, T9, T14 and T16, as well as FaDu and UM 22B HNSCC cells in comparison to that in the normal mucosal samples we examined. T denotes tumor sample, and N normal mucosa.

Figure 2. TKTL1 promoted HNSCC tumor growth in vitro and in vivo

A, TKTL1 promoted cell growth in JHU-O11 cells in vitro. Shown are the MTT proliferation assay in TKTL1-transiently transfected JHU-O11 cells (*P<0.05), western blot displaying the overexpression of TKTL1 protein levels in TKTL1–expressing O11 cells and soft agar assay showing increased anchorage-independent growth in TKTL1-expressing O11 cells compared with the vector-expressing cells (*P<0.05). Representative pictures of soft-agar colonies in TKTL1- and vector-expressing O11 were shown. B, TKTL1 promoted cell growth in JHU-O28 cells in vitro. Shown are the MTT proliferation assay in TKTL1-transiently transfected JHU-O28 cells (*P<0.05), western blot displaying the overexpression of TKTL1 protein levels in TKTL1–expressing O28 cells and soft agar assay showing increased anchorage-independent growth in TKTL1-expressing O28 cells compared with the vector-expressing cells (*P<0.05). Representative pictures of soft-agar colonies in TKTL1- and vector-expressing O28 cells were shown. C, Plot of mean tumor-volume trajectories over time for the mice inoculated with TKTL1- and vector-expressing O11 stable cells. 5 mice per group. Error bars represent the SEMs at each time point. D, left, dissected tumors from the NOD/SCID mice. The representative tumors derived from the O11 cells stably expressing TKTL1 are shown (bottom row), and these tumors were larger than the tumors from the O11 cells stably transfected with vector (upper row). Right, mass of the dissected tumors. Each dot represented the tumor mass from the mice. P=0.016.

Figure 3. TKTL1 enhanced the production of fructose-6-phosphate and glyceraldehyde-3-phosphate

A, simplified overview of glycolysis, TCA, and PPP. G6P, glucose-6-phosphate; F6P, fructose-6-phosphate; G3P, glyceraldehyde-3-phosphate; PYR, pyruvate; LAC, lactate; R5P, ribose-5-phosphate. B and C, metabolite levels in the O11 cells stably expressing TKTL1 and vector, as determined by LC-MS/MS. Shown are the quantitative changes of F6P and G3P in TKTL1-expressing cells compared with that in vector expressing cells. Error bars denote the SEMs (n = 3). *P<0.05.

Figure 4. TKTL1 increased aerobic glycolysis in HNSCC

A, the production of pyruvate (left) and lactate (right) in the culture buffer was measured in TKTL1- and vector-expressing JHU-O11 cells. B, the production of pyruvate (left) and lactate (right) in the culture buffer was measured in TKTL1- and vector-expressing JHU-O28 cells. The cells were allowed to grow in Krebs media for 6 h, and then media were collected and assayed for concentration of pyruvate and lactate. Data represent mean ± SEM. of three individual experiments. *P < 0.05. C, The glucose uptake of TKTL1 expressing O11 cells was significantly increased compared with that of the control cells after 48 hour culture in RPMI with 11 mM glucose. Triplicate samples were used in each group. *P< 0.05. D, The ATP level in TKTL1-expressing O11 cells was significantly increased compared with that of the vector-expressing cells. The cells were cultured with normal medium for 48 hours and harvested for intracellular ATP content (nmols per million cells ± SEM) measurement by luciferin-luciferase assay. *P<0.05.

Figure 5. TKTL1 contributes to HIF1α accumulation in HNSCC

A, left, the protein levels of HIF1α were determined 4 h following switching from culture medium to Krebs-Henseleit buffer in O11 cells stably expressing TKTL1. HIF1α was overexpressed by 1.5-fold in the O11 cells stably expressing TKTL1 versus its corresponding vehicle cells (n = 3, P < 0.05). Right, quantitative RTPCR analysis for two HIF-regulated downstream genes, GLUT1 and LOX in O11 cells expressing TKTL1. Shown are mean values of triplicate assays ± SEM. *P<0.05. B, left, the protein levels of HIF1α and TKTL1 in tumor xenografts originated from TKTL1-expressing O11 cells compared with those from vector-expressing cells. Representative images from three independent experiments were shown. Right, quantitative RTPCR analysis for GLUT1 and LOX in tumor xenografts originated from TKTL1-expressing O11 cells. Shown are mean values of triplicate assays ± SEM. *P<0.05. C, Western blot analysis on HIF1α and TKTL1 in from FaDu cells after transient transfection of TKTL1 shRNAs. After 48 h of transfection, cells were incubated with glucose-free Krebs buffer for 4 h. The cells then were harvested for Western blotting. Results are representative of three independent experiments. Bar graph shows the inhibition of TKTL1 expression in Fadu cells after TKTL1 shRNA transfection, as quantitated by quantitative real-time RTPCR. *P<0.05. D, the growth of O11 cells stably expressing TKTL1, was inhibited after treatment with NSC134754, a HIF1α small inhibitor. Cell growth was estimated by using MTT assay, after exposure for 48 h to 0.5 µM of NSC134754. Data are mean ± SEM. values from three independent experiments. *P<0.05.

Figure 6. HIF1α provides a feed-forward loop in TKTL1-mediated HNSCC tumorigenesis by regulation of glycolytic enzymes

A, quantitative RTPCR analysis for four HIF1α -regulated downstream genes implicated in glycolysis (HK2, ALDOC, PGK1, and PGM1) in O11 cells stably expressing TKTL1. Student’s t test showed significance for expression of HK2, ALDOC, PGK1, and PGM1 between TKTL1- and VEC-expressing O11 stable cells, whereas there was no significance for expression of HK2, ALDOC, and PGK1 after 48 h treatment with 1µM of NSC134754 (the HIF1α inhibitor). Actin gene served as controls. Shown are mean values of triplicate assays ± SEM. *P<0.05, #P>0.05. B. the growth of O11 cells stably expressing TKTL1, was inhibited after treatment with 3-bromopyruvate, a glycolysis inhibitor through inhibition of hexokinase. Cell growth was estimated by using MTT assay, after exposure for 48 h to 30 µM of 3-bromopyruvate. Data are mean ± SEM. values from three independent experiments. *P<0.05. C, Ribose-5-phosphate (R5P) levels in TKTL1-expressing O11 cells. Cells were cultured with normal RPMI1640 medium for 48 h and harvested for R5P assay using LC-MS/MS as described in Methods. *P<0.05. D, the proposed mechanism of TKTL1-mediated HNSCC tumorigenesis.