1, 9-Pyrazoloanthrones downregulate HIF-1α and sensitize cancer cells to cetuximab-mediated anti-EGFR therapy

Abstract

Cetuximab, a monoclonal antibody that blocks the epidermal growth factor receptor (EGFR), is currently approved for the treatment of several types of solid tumors. We previously showed that cetuximab can inhibit hypoxia-inducible factor-1 alpha (HIF-1α) protein synthesis by inhibiting the activation of EGFR downstream signaling pathways including Erk, Akt, and mTOR. 1, 9-pyrazoloanthrone (1, 9 PA) is an anthrapyrazolone compound best known as SP600125 that specifically inhibits c-jun N-terminal kinase (JNK). Here, we report 1, 9 PA can downregulate HIF-1α independently of its inhibition of JNK. This downregulatory effect was abolished when the oxygen-dependent domain (ODD) of HIF-1α (HIF-1α-ΔODD, the domain responsible for HIF-1α degradation) was experimentally deleted or when the activity of HIF-1α prolyl hydroxylase (PHD) or the 26S proteasomal complex was inhibited, indicating that the 1, 9 PA downregulates HIF-1α by promoting PHD-dependent HIF-1α degradation. We found that the combination of 1, 9 PA and cetuximab worked synergistically to induce apoptosis in cancer cells in which cetuximab or 1, 9 PA alone had no or only weak apoptotic activity. This synergistic effect was substantially decreased in cancer cells transfected with HIF-1α-ΔODD, indicating that downregulation of HIF-1α was the mechanism of this synergistic effect. More importantly, 1, 9 PA can downregulate HIF-1α in cancer cells that are insensitive to cetuximab-induced inhibition of HIF-1α expression due to overexpression of oncogenic Ras (RasG12V). Our findings suggest that 1, 9 PA is a lead compound of a novel class of drugs that may be used to enhance the response of cancer cells to cetuximab through a complementary effect on the downregulation of HIF-1α.

Figure 1. 1, 9 PA downregulates HIF-1α independent of its JNK inhibitory function.

(A) Temporal order and correlation of 1, 9 PA-induced HIF-1α downregulation and JNK inhibition, and the treatment-induced compensatory activation of Akt and Erk. A431 cells were treated with 5 µM 1, 9 PA in 0.5% FBS culture medium for the indicated time intervals. Cell lysates were then prepared for Western blotting with the antibodies shown. (B) Dose-dependent effects of 1, 9 PA on downregulating HIF-1α, inhibiting JNK, and activating Erk. A431 cells were exposed to increasing concentrations of 1, 9 PA for 16 h in 0.5% FBS culture medium at 37°C. Cell lysates were then prepared for Western blotting with the antibodies shown. (C) Independence of 1, 9 PA-induced HIF-1α downregulation from its JNK inhibitory function. A431 cells were treated with increasing concentrations of 1, 9 PA or 1-methyl-1, 9 PA for 1 h in 0.5% FBS culture medium. Cell lysates were then prepared for Western blotting with the antibodies shown. (D) No changes in HIF-1α level after activation or inhibition of JNK. A431 cells wert transiently transfected with a control vector or one of the constructs containing a constitutively active SEK1 S220E/T224D mutant (SEK1-CA), a dominant-negative SEK1 K129R mutant (SEK1-DN), or a dominant-negative MEKK1 K432M mutant (MEKK1-DN) overnight. Cell lysates were then prepared for Western blotting with the antibodies shown. (E) Compensatory activation of Erk after HIF-1α silencing in comparison with treatment by 1, 9 PA or 1-methyl-1, 9 PA. A431 cells were subjected to knockdown of HIF-1α with specific or control siRNA, or treated with 10 µM 1, 9 PA or 1-methyl-1, 9 PA for 16 h. Cell lysates were then prepared for Western blotting with the antibodies shown. (F) Downregulation of HIF-1α by 1, 9 PA in various cancer cell lines. Indicated cell lines were exposed to 10 or 40 µM 1, 9 PA for 1 h in 0.5% FBS culture medium. Cell lysates were then prepared for Western blotting with the antibodies shown.

Figure 2. 1, 9 PA downregulates HIF-1α by enhancing HIF-1α protein degradation.

(A) Reduced HIF-1α protein stability in the presence of 1, 9 PA (1, 9 PA). A431 cells were treated with 10 µM cycloheximide (CHX) plus 10 µM 1, 9 PA or DMSO vehicle control in 0.5% FBS culture medium for up to 60 min at 37°C. Cell lysates were prepared after treatment at each time point and analyzed by Western blotting with the antibodies shown. (B) Inhibition of 1, 9 PA-induced HIF-1α degradation by the proteasomal inhibitor MG132. A431 cells were treated with 10 µM 1, 9 PA±10 µM MG132 for the indicated time intervals. Cell lysates were prepared after treatment at each time point and analyzed by Western blotting with the antibodies shown.

Figure 3. 1, 9 PA downregulates HIF-1α in a PHD- and HIF-1α ODD-dependent manner.

(A) Requirement of O₂ and Fe²⁺ in the 1, 9 PA-induced downregulation of HIF-1α. A431 cells were untreated or treated with 10 µM 1, 9 PA in 0.5% FBS culture medium under normoxic conditions in the absence or presence of DFO (100 µM) or MG132 (10 µM), and under hypoxic conditions for 16 h at 37°C. Cell lysates were then prepared for Western blotting with the antibodies shown. (B) Role of the ODD of HIF-1α in the 1, 9 PA-induced downregulation of HIF-1α. A431 cells were transiently transfected with the HIF-1α-ΔODD construct or a control vector for 48 h and were then either untreated or treated with the indicated concentrations of 1, 9 PA for 1 h at 37°C. Cell lysates were then prepared for Western blotting with the antibodies shown. (C) Resistance of A431/HIF-1α-ΔODD cells to the 1, 9 PA-induced downregulation of HIF-1α. A431 cells stably expressing the HIF-1α-ΔODD construct were treated as described in (A). Cell lysates were then prepared for Western blotting with the antibodies shown.

Figure 4. 1, 9 PA enhances HIF-1α ubiquitination.

(A). A431 cells were treated with 10 µM 1, 9 PA for indicated time period. Cell lysates were prepared for HIF-1α immunoprecipitation, followed by Western blotting of the immunoprecipitates with antibodies directed against ubiquitin (top), HIF-1α, and β-actin. (B) A431 cells were untreated or treated with 10 µM 1, 9 PA in the presence of 10 µM MG132 for 1 h in 0.5% FBS medium. HIF-1α was immunoprecipitated followed by Western blot analysis with an anti-HIF-1α antibody.

Figure 5. Differential effect of constitutively active Akt on cetuximab and 1, 9 PA-induced downregulation of HIF-1α.

A431 cells were transiently transfected with a control vector or a myristoylated Akt (Myr-Akt) for 24 h in 0.5% FBS medium. The vector- or Myr-Akt–transfected cells were then treated with either 20 nM cetuximab or PBS overnight (A), or with 10 µM 1, 9 PA or DMSO vehicle control for 1 h (B). After treatment, cell lysates were prepared for Western blotting with the antibodies shown.

Figure 6. Combination of 1, 9 PA and cetuximab induces apoptosis through downregulation of HIF-1α.

(A) Induction of PARP cleavage by the combination of 1, 9 PA and cetuximab. A431, HN5, and DiFi cells were untreated or treated with cetuximab (10 nM for A431 and HN5 cells and 2 nM for DiFi cells for 16 h), 1, 9 PA (10 µM or 40 µM added the last hour before cell lysis), or both in 0.5% FBS culture medium. Cell lysates were prepared and analyzed by Western blotting with the antibodies shown. (B) Increased induction of apoptosis by the combination of 1, 9 PA and cetuximab. A431 cells were treated as described in (A). Cell lysates were prepared and analyzed by apoptosis ELISA. The relative absorbance values are plotted. The p value was <0.01 when comparing the level of apoptosis by 1, 9 PA alone (10 or 40 µM) or cetuximab alone with that of apoptosis by combination of the 2 agents (note: only the p values comparing 10 µM 1, 9 PA alone and in combination with cetuximab are shown). (C) Dependence of induction of apoptosis by the combination of 1, 9 PA and cetuximab on HIF-1α downregulation. A431neo and A431/HIF-1α-ΔODD cells were treated as indicated, and the cell lysates were prepared and analyzed as described in (A).

Figure 7. 1, 9 PA enhances responses of cancer cells expressing an oncogenic Ras mutant to cetuximab.

(A) Effect of 1, 9 PA and cetuximab, either alone or in combination, on the HIF-1α level and induction of apoptosis. A431neo and A431/RasG12V cells were untreated or treated with cetuximab (10 nM for 16 h), 10 µM 1, 9 PA (added the last hour before cell lysis), or both in 0.5% FBS culture medium. Cell lysates were prepared and analyzed by Western blotting with the antibodies shown. (B) 1, 9 PA-mediated sensitization to cetuximab-induced growth inhibition. A431neo and A431/RasG12V cells were treated with increasing concentrations of cetuximab ±5 µM 1, 9 PA in 0.5% FBS culture medium for 5 days. After treatment, the cells were subjected to an MTT assay. The optical density values of the treated groups were normalized to the values of the control groups (with or without 1, 9 PA treatment) and expressed as a percentage of respective control. The percentage of surviving cells was plotted as a function of treatment with increasing concentrations of cetuximab. The differences in cell survival between the two groups were statistically significant (p<0.01) when the concentrations of cetuximab were greater than 0.625 nM in A431neo cells and 1.25 nM in A431/RasG12V cells. (C) 1, 9 PA-mediated sensitization to cetuximab-induced inhibition of VEGF production. A431neo and A431/RasG12V cells were untreated or treated with 10 nM cetuximab, 10 µM 1, 9 PA, or both in 0.5% FBS culture medium for 16 h. The VEGF secreted into the conditioned media by the cells was measured by ELISA. The p-values for indicated comparisons were shown. (D) Comparison of A431 and GEO cells to treatment with 1, 9 PA and cetuximab, either alone or in combination. A431 and GEO cells were treated as described in (A). Cell lysates were prepared and analyzed by Western blotting with the antibodies shown. (E). 1, 9 PA-mediated sensitization GEO cells to cetuximab-induced growth inhibition. GEO cells were treated with increasing concentrations of cetuximab ±5 µM 1, 9 PA in 0.5% FBS culture medium for 5 days. After treatment, the cells were subjected to an MTT assay. The data were processed as described in (B). The differences in cell survival between the two groups were statistically significant (p<0.01) at all concentrations of cetuximab tested.