This article is mentioned in:

Introduction

Thymidine phosphorylase (TP; EC 2.4.2.4) catalyzes the reversible conversion of thymidine, deoxyuridine and their analogs to their respective bases and 2-deoxy-D-ribose-1-phosphate (1). TP is identical to the angiogenic factor, platelet-derived endothelial cell growth factor (PD-ECGF) (2,3). TP stimulates chemotaxis and [3H]thymidine incorporation by endothelial cells in vitro and has an angiogenic activity in vivo(4–7). We previously demonstrated that TP enzymatic activity is indispensable for its angiogenic activity (4,7). Among the degradation products generated by TP enzymatic activity, 2-deoxy-D-ribose, a dephosphorylated product derived from 2-deoxy-D-ribose-1-phosphate, also displays chemotactic activity in vitro and angiogenic activity in vivo. TP is expressed at higher levels in a wide variety of tumors when compared to that in the adjacent non-neoplastic tissues (8–10). Under hypoxic conditions, TP can also enhance the growth of tumor cells and confer resistance to apoptosis induced by hypoxia (11). 2-Deoxy-D-ribose was able to partially prevent hypoxia-induced apoptosis in human leukemia HL-60 cells (12,13). TP also inhibited upregulation of HIF-1α, BNIP3 and caspase-3 under hypoxic conditions (14). These findings indicated that TP has another function apart from angiogenesis.

To elucidate the mechanism by which 2-deoxy-D-ribose suppresses HIF-1α levels under hypoxic conditions, we examined the effect of 2-deoxy-D-ribose on key regulators of HIF-1α: von Hippel-Lindau (VHL) and prolyl hydroxylase (PHD)2.

Materials and methods

Reagents and antibodies

The monoclonal anti-HIF-1α antibody was purchased from Transduction Laboratories (Lexington, KY, USA). The rabbit anti-VHL polyclonal antibody (FL-181) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). A rabbit anti-PHD2 polyclonal antibody was from Bethyl Laboratories (Montgomery, TX, USA). A rabbit anti-hydroxy-HIF-1α (Pro564) was from Cell Signaling Technology (Boston, MA, USA).

Cell lines and induction of hypoxia

Human leukemia HL-60 cells were maintained in RPMI-1640 containing 10% fetal calf serum. 2-Deoxy-D-ribose was added to the culture medium and then hypoxia (1% O2) was induced in a Personal Multigas Incubator (Astec).

Immunoblotting analysis

Samples were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) according to the method of Laemmli. Proteins in the gel were electrophoretically transferred onto polyvinylidene difluoride membranes (Immobilon-P transfer membrane; Millipore, Bedford, MA, USA) using Bio-Rad Trans-Blot SD apparatus. The membrane was treated with buffer A [350 mM NaCl, 10 mM Tris-HCl (pH 8.0), 0.05% Tween 20] containing 3% skimmed milk for 1 h and incubated with the indicated antibody (1:1,000) in buffer A containing 3% skimmed milk for 1 h. Following four washes with buffer A (10 min each), the membrane was incubated with peroxidase-conjugated horse anti-mouse IgG diluted 1:1,000 in buffer A containing 3% skimmed milk for 1 h. Following washing with buffer A, the membrane was developed using the enhanced chemiluminescence western blotting detection system (Amersham Pharmacia, Buckinghamshire, UK).

RT-PCR method

Total cellular RNA was extracted using TRIzol reagent according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA). RT-PCR was performed with the SuperScript One-Step RT-PCR system and gene-specific primers according to the manufacturer’s instructions (Invitrogen). Reaction mixtures containing total RNA (500 ng of each), 0.2 mM dNTPs, 0.2 mM of each primer, 2 units of enzyme mixture including SuperScript II RT, Platinum Taq DNA polymerase, and 1X buffer with 1.2 mM MgSO4 were maintained at 50°C for 20 min, and then at 94°C for 2 min, and PCR was performed as follows. The PCR profile consisted of 30 cycles at 94°C for 15 sec, 55°C for 30 sec, and 70°C for 30 sec. The primers for RT-PCRs were designed based on human sequences in GenBank. These sequences used the following primers: PHD1, 5′-gccagtgctgggctgatggagg-3′ and 5′-cgcagtggcggatgacggcgtc-3′; PHD2, 5′-tgcatgaacaa gcacggcatct-3′ and 5′-atatacatgtcacacatcttcc-3′; PHD3, 5′-ga ctgcgtcctggagcgcgtca-3′ and 5′-atatccgcaggatcccaccatg-3′; VHL, 5′-atgccccggagggcggagaactggg-3′ and 5′-tcaatctcccatcc gttgatgtgca-3′; HIF-1α, 5′-tcgaggcctctgtgatgagg-3′ and 5′-ggc ctctgtgatgaggcttt-3′; GAPDH, 5′-agaacatcatccctgcctctactgg-3′ and 5′-aaaggtggaggagtgggtgtcgctg-3′.

Immunoprecipitation

For immunoprecipitation experiments, FLAG tagged VHL or PHD2 and HIF-1α cDNA (1 μg each) were transiently transfected into COS or HL-60 cells plated in 6-cm diameter dishes. Twenty four hours following transfection, the cells were treated for 5 h under hypoxic conditions and cells were harvested. The cells were suspended in 200 μg of whole cell extract buffer [10 mM HEPES (pH 7.9), 400 mM NaCl, 0.1 mM EDTA, 5% (vol/vol) glycerol, 1 mM DTT, 1 mM APMSF], and centrifuged at 12,500 rpm for 30 min at 4°C. Subsequent to protein extraction, 300 μg of total proteins was incubated with anti-FLAG or anti-PHD2 antibodies at 4°C for 1 h. Thirty microliters of a 50% slurry of protein G-Sepharose 4B in TEG buffer [20 mM Tris-HCl (pH 7.9), 1 mM EDTA, 10% glycerol, 1 mM DTT], containing 150 mM NaCl and 0.1% Triton X-100, were then added to reaction mixtures and incubated for 12 h at 4°C with rotation. Following a rapid centrifugation, the resulting pellets were washed three times with TEG buffer, and the immunoprecipitated proteins were analyzed by immunoblotting using an anti-HIF-1α antibody.

Results

Effect of 2-deoxy-D-ribose on the expression of HIF-1α under hypoxic conditions

We determined the effect of 2-deoxy-D-ribose on the HIF-1a protein level under hypoxic conditions. As shown in Fig. 1A, we could not detect HIF-1α in HL-60 cells under a normoxic condition. When HL-60 cells were incubated under hypoxic conditions, the HIF-1α protein level was markedly increased (Fig. 1A). Treatment of the cells with 2-deoxy-D-ribose substantially suppressed the level of HIF-1α in the cells under hypoxic conditions. We next determined the effect of 2-deoxy-D-ribose on the expression of HIF-1α mRNA levels by RT-PCR. HIF-1α mRNA levels were not affected by hypoxia, and treatment of the cells under hypoxic conditions by 2-deoxy-D-ribose did not alter the levels of HIF-1α mRNA (Fig. 1B).

Figure 1 Effect of 2-deoxy-D-ribose on the expression of HIF-1α under hypoxic conditions. (A) The cells were incubated under hypoxic conditions for 18 h, and cell lysates were prepared from the cells. HIF-1α was detected by immunoblotting with an anti-HIF-1α antibody. (B) HL-60 cells were cultured under normoxic or hypoxic conditions for 24 h. Total RNA was extracted and the expression of HIF-1α was evaluated by RT-PCR.

Effect of 2-deoxy-D-ribose on the interaction of HIF-1α and VHL

The HIF-1α protein is continuously synthesized and degraded under a normoxic condition, while it accumulates rapidly following exposure to hypoxic conditions. HIF-1α interacts with VHL and is degraded via the ubiquitin-proteasome pathway under a normoxic condition. As the oxygen concentration decreases, PHDs become inactive and the HIF-1α protein is consequently stabilized. One possible mechanism for the attenuation of HIF-1α caused by 2-deoxy-D-ribose under hypoxic conditions might be the enhanced interactions of HIF-1α with VHL and consequent degradation of HIF-1α. To assess the interactions between HIF-1α and VHL in the presence or absence of 2-deoxy-D-ribose under hypoxic conditions, COS cells were cotransfected with HIF-1α and VHL-expressing plasmids. After 24 h following transfection, the cells were exposed for 5 h to hypoxic conditions in the presence or absence of 2-deoxy-D-ribose. As shown in Fig. 2A, incubation of cells under hypoxic conditions resulted in a marked increase in HIF-1α protein levels, and treatment with 2-deoxy-D-ribose substantially suppressed HIF-1α in the cells under hypoxic conditions. To examine the effect of 2-deoxy-D-ribose on the interaction between VHL and HIF-1α, the cell lysates were immunoprecipitated with an anti-FLAG antibody and the coprecipitated HIF-1α was detected with an anti-HIF-1α antibody. Treatment of COS cells with 2-deoxy-D-ribose enhanced interaction of HIF-1α with VHL under hypoxic conditions (Fig. 2B).

Figure 2 Effect of 2-deoxy-D-ribose on the interaction between VHL and HIF-1α. (A) COS cells were transiently cotransfected with VHL and HIF-1α in the presence of MG-132, and the cells were cultured under normoxic or hypoxic conditions for 5 h. HIF-1α and VHL were detected by immunoblotting with an anti-HIF-1α and anti-VHL antibody, respectively. (B) COS cells were transiently cotransfected with VHL and HIF-1α plasmids in the presence of MG-132, and the cells were cultured under normoxic or hypoxic conditions for 5 h, and cell lysates were prepared from the cells. Following immunoprecipitation of proteins with an anti-FLAG antibody, HIF-1α was detected by immunoblotting using an anti-HIF-1α antibody.

Effect of 2-deoxy-D-ribose on the expression of PHD1/2/3 and VHL

To determine the effect of 2-deoxy-D-ribose on the expression of PHD1/2/3 and VHL in HL-60 cells, the cells were cultured under normoxic or hypoxic conditions in the presence or absence of 2-deoxy-D-ribose. As shown in Fig. 3A, the expression of PHD2 was detected but those of PHD1 and PHD3 mRNA were not detected by RT-PCR. 2-Deoxy-D-ribose did not affect the expression of PDH2 mRNA and protein levels under normoxic and hypoxic conditions in HL-60 cells (Fig. 3).

Figure 3 Effect of 2-deoxy-D-ribose on the expression of PHD1, PHD2, PHD3 and VHL. (A) HL-60 cells were cultured under normoxic or hypoxic conditions for 24 h. Total RNA was extracted and the expression of PHD1, PHD2, PHD3, VHL and GAPDH was evaluated by RT-PCR. (B) The cells were incubated under hypoxic conditions for 24 h and cell lysates were prepared from the cells.

Effect of 2-deoxy-D-ribose on the interaction between HIF-1a and PHD2

2-Deoxy-D-ribose may decrease HIF-1α protein levels by affecting the interaction between HIF-1α and PHD2 and consequently modulating its degradation. To assess the effect of 2-deoxy-D-ribose on the interaction between HIF-1α and PHD2, HL-60 cells were cultured under normoxic or hypoxic conditions in the presence or absence of 2-deoxy-D-ribose. Whole cell extracts were prepared and immunoprecipitated with an anti-HIF-1α antibody and coprecipitated PHD2 was detected with an anti-PHD2 antibody. Hypoxia attenuated the interaction between HIF-1α and PHD2. Treatment of the cells with 2-deoxy-D-ribose augmented the interaction between HIF-1α and PHD2 under hypoxic conditions (Fig. 4A).

Figure 4 Effect of 2-deoxy-D-ribose on the interaction between PHD2 and HIF-1α, and the levels of hydroxy-HIF-1α. (A) Effect of 2-deoxy-D-ribose on the interaction between PHD2 and HIF-1α. HL-60 cells were cultured under normoxic or hypoxic conditions for 24 h and cell lysates were prepared from the cells. Following immunoprecipitation of proteins with an anti-PHD2 antibody, HIF-1α was detected by immunoblotting using an anti-HIF-1α antibody. (B) Effect of 2-deoxy-D-ribose on the levels of hydroxy-HIF-1α. HL-60 cells were cultured in the presence or absence of 2-deoxy-D-ribose or MG-132 for 5 h, and cell lysates were prepared from the cells. Hydroxy-HIF-1α was detected by immunoblotting using an anti-hydroxy-HIF-1α antibody.

Effect of 2-deoxy-D-ribose on the levels of hydroxy-HIF-1a

HIF-1α can be hydroxylated by proline hydroxylase under normoxic conditions. Hydroxylation of HIF-1α leads to its binding to VHL and ubiquitin-mediated degradation (30,31). To further characterize the mechanisms leading to decreased HIF-1α levels by 2-deoxy-D-ribose, we analyzed the levels of hydroxy-HIF-1α. Direct analysis of the hydroxylation state of HIF-1α was carried out using an antibody specifically directed against hydroxy-HIF-1α. To visualize hydroxy-HIF-1α, the cells were treated with the proteasome inhibitor MG-132 to prevent the proteasomal degradation of HIF-1α. As shown in Fig. 4B, 2-deoxy-D-ribose increased the amounts of hydroxy-HIF-1α in the presence of MG-132. This finding suggests that 2-deoxy-D-ribose enhances proline hydroxylase activity.

Discussion

Previous studies have demonstrated that TP confers resistance to apoptosis induced by hypoxia and that the enzymatic activity of TP is required for this effect. 2-Deoxy-D-ribose, a degradation product of thymidine generated by TP activity, can also prevent hypoxia-induced apoptosis in human KB epidermoid carcinoma cells (11,13) suggesting that it may be a downstream mediator of the TP function. 2-Deoxy-D-ribose prevented the hypoxia-induced activation of caspase-3 and -9 in HL-60 cells (12). TP-overexpressing Jurkat cells were also resistant to hypoxia-induced apoptosis. The induction of caspase-3 activity and the expression of HIF-1α and BNIP3 were found to be suppressed under hypoxic conditions in TP-expressing Jurkat cells (14).

HIF-1α protein is rapidly accumulated under hypoxia and degraded under a normoxic condition (15–17). The accumulated HIF-1α under hypoxia dimerizes with HIF-1β and translocates into the nucleus. HIF-1 binds to the hypoxia response element (HRE) on nuclear DNA, recruits coactivators p300/CBP, and promotes gene transcription of various genes (18,19). Many tumors contain a hypoxic microenvironment, a condition associated with resistance to anticancer agents and poor prognosis. HIF-1α is upregulated in a broad range of tumors and is involved in angiogenesis, invasion and altered energy metabolism (20). There appears to be a delicate balance between the pro- and anti-tumorigenic effects of HIF-1α. Several gene products such as BNIP3, RTP801 and Noxa were identified as HIF-1α-responsive pro-apoptotic proteins. Hypoxia-mediated BNIP3 expression is regulated by HIF-1α that directly binds to a consensus HRE in the BNIP3 promoter (21). Overexpression of BNIP3 has been shown to be cytotoxic in a number of tumor cell lines (22–25). Mitochondria may be a direct target of death signals mediated via BNIP3 under hypoxic conditions. BNIP3 induced by HIF-1α binds to mitochondria and opens the mitochondrial permeability transition pore (26). BNIP3 is a key regulator of mitochondrial function and cell death of ventricular myocytes during hypoxia (26). We previously demonstrated that TP suppressed the level of BNIP3 under hypoxic conditions, and 2-deoxy-D-ribose, a downstream mediator of the TP function, accelerated the proteasome-mediated degradation of HIF-1α by enhancing the ubiquitination under hypoxic conditions (12,14).

O2-dependent regulation of the HIF-1α protein is mediated by a functional domain of 200 amino acids located on the carboxy terminal to the PAS domain, which was named the oxygen-dependent degradation (ODD) domain (27). HIF-1α prolyl hydroxylases utilize O2 and α-ketoglutarate as a substrate to generate 4-hydroxyproline at residues 402 and/or 546 of human HIF-1α (28,29). Three such prolyl hydroxylases 1–3 (PHD1–3) were identified in mammalian cells. Under normoxic conditions, the hydroxylation of specific proline residues of HIF-1α by PHD2 promotes the interaction of HIF-1α with the VHL protein and consequently ubiquitination and proteasomal degradation of HIF-1α (30,31). Although PHD1 and PHD3 can hydroxylate HIF-1α in vitro, HIF does not appear to be their physiological target in the cell (32). During hypoxia, the reduced levels of proline hydroxylation lead to HIF-1α degradation, causing an increase in its levels.

2-Deoxy-D-ribose had no effect on HIF-1α mRNA expression levels indicating that 2-deoxy-D-ribose affected the level of HIF-1α at the protein level. In a previous study, we also demonstrated that 2-deoxy-D-ribose accelerated proteasome-mediated degradation of HIF-1α by enhancing the ubiquitination of HIF-1α (12). In this study, we demonstrated that 2-deoxy-D-ribose increased the interaction of HIF-1α with VHL and PHD2 under hypoxic conditions, and levels of proline hydroxy-HIF-1α (Figs. 2 and 4). A potential mechanism of the enhanced levels of proline hydroxy-HIF-1α by 2-deoxy-D-ribose might be the increased level of α-ketoglutarate. The cosubstrates oxygen and α-ketoglutarate as well as the cofactors Fe2+ and ascorbate are required for PHD activity. 2-Deoxy-D-ribose is phosphorylated to 2-deoxy-D-ribose 5-phosphate which is then cleaved by deoxy-D-ribose phosphate aldolase to acetoaldehyde and glyceraldehyde-3-phosphate, which then produces pyruvate. α-ketoglutarate and alanine are produced by enzymatic transamination reaction between glutamate and pyruvate (33,34). α-Ketoglutarate produced by 2-deoxy-D-ribose may have increased PHD activity.

2-Deoxy-D-ribose is produced by the catalytic action of TP and is the downstream mediator of TP function. TP is expressed at higher levels in a wide variety of solid tumors when compared to adjacent non-neoplastic tissues. TP may thus play an important role in the progression of tumors by producing 2-deoxy-D-ribose from thymidine and decreasing the level of HIF-1α.

References