Cordycepin (3'-Deoxyadenosine) Suppresses Heat Shock Protein 90 Function and Targets Tumor Growth in an Adenosine Deaminase-Dependent Manner

Abstract

Alterations in metabolism and energy production are increasingly being recognized as important drivers of neoplasia, raising the possibility that metabolic analogs could disrupt oncogenic pathways. 3'-deoxyadenosine, also known as cordycepin, is an adenosine analog that inhibits the growth of several types of cancer. However, the effects of cordycepin have only been examined in a limited number of tumor types, and its mechanism of action is poorly understood. We found that cordycepin slows the growth and promotes apoptosis in uveal melanoma, as well as a range of other hard-to-treat malignancies, including retinoblastoma, atypical teratoid rhabdoid tumors, and diffuse midline gliomas. Interestingly, these effects were dependent on low adenosine deaminase (ADA) expression or activity. Inhibition of ADA using either siRNA or pharmacologic approaches sensitized tumors with higher ADA to cordycepin in vitro and in vivo, with increased apoptosis, reduced clonogenic capacity, and slower migration of neoplastic cells. Our studies suggest that ADA is both a biomarker predicting response to cordycepin and a target for combination therapy. We also describe a novel mechanism of action for cordycepin: competition with adenosine triphosphate (ATP) in binding to Hsp90, resulting in impaired processing of oncogenic Hsp90 client proteins.

Keywords: adenosine deaminase; cordycepin; uveal melanoma.

Figure 5

Cordycepin binds Hsp90 and disrupts its function, resulting in degradation of client proteins. (A) Visualization of possible interactions between the N-terminal of HSP90 (PDB: 6GPT) [52] and cordycepin within the HSP90 hydrophobic pocket (e.g., ATP binding site) observed by docking simulations. Non-polar hydrogens and amino acid residues apart from the key residues known for ATP binding (e.g., N51, D93, G137 and F138) [53,54] were omitted for clarity. (B,C) Competition assay with cordycepin and ATP (B and C left) and with cordycepin triphosphate and ATP ((C) right) using human recombinant full-length Hsp90 (FL), N-terminal and C-terminal domain. (D) 92.1 cells were treated with vehicle, 10 µM MG-132, and 80 µM cordycepin. (E) Co-immunoprecipitation assay using an anti-Akt antibody with vehicle, 80 µM cordycepin, and 80 µM cordycepin/10 µM of MG-132 (combination) then incubated for 6 h. (F) Cells were treated by 80 µM cordycepin only under hypoxic conditions (left) or 20% oxygen condition (right) for 1.5 h. Total cell lysates were prepared and Co-immunoprecipitated (Co-IP) with anti-HIF-1α or Akt antibodies. The interaction between client protein (HIF-1α or Akt) and Hsp90 was analyzed by Western blot analysis. (G,H) Cells were treated with cordycepin, or combination (4–20 µM cordycepin3′-deoxyadenosine and 1 µM EHNA) for 1.5 h under the hypoxic condition. HIF-1α protein level was analyzed by Western blot. (I) Cells were treated with cordycepin in the presence of 1 µM EHNA (left) or pentostatin (right) for 2 days. Total Akt, total ERK, EGFR, and Hsp70 protein expression levels were analyzed by Western blot.

Figure 1

ADA decreases anticancer effects of cordycepin in uveal melanoma. (A) Growth of viable cell mass after cordycepin treatment. Uveal melanoma cells were treated with vehicle, 80, and 160 µM of cordycepin, and after 5 days incubation viable cell mass was measured using fluorescence. (B) Adenosine deaminase (ADA) protein expression levels varied uveal melanoma cells lines on western blots. (C) ADA enzymatic activity in uveal melanoma. (D,E) Immunohistochemical ADA protein expression in human uveal melanomas on tissue microarrays. (F,G), ADA protein expression (F) and cell growth (G) in 92.1 and MP46 uveal melanoma cells after siRNA-based knockdown. Values represent mean ± SD of experiments conducted in sextuplicate. **** p < 0.0001 by one-way ANOVA analysis of variance compared with control group. (H) Antitumor effects of cordycepin in mice bearing xenograft tumors of 92.1 uveal melanoma cells. (I) Pictures of xenografts from 92.1 uveal melanoma cells. (J) Tumor weight in vehicle, 10 mg/kg and 20 mg/kg cordycepin treatment group. Values represent mean ± SEM of experiments. CT, control; 10 mg/kg, cordycepin 10 mg/kg b.w. treatment; 20 mg/kg, cordycepin 20 mg/kg b.w. treatment. ** p < 0.01, *** p < 0.001, **** p < 0.0001 by one-way ANOVA analysis of variance compared with control group.

Figure 2

Cordycepin inhibits colony formation and migration abilities of uveal melanoma cells in an ADA dependent manner. (A,B) Effects of cordycepin treatment on soft agar colony formation of uveal melanoma cells. (C,D) the effect of cordycepin treatment on the anchorage-dependent colony formations. (E,F) Inhibition of migration abilities of uveal melanoma cells by treatment with 160 µM cordycepin. Values represent mean ± SD of experiments conducted in sextuplicate. **** p < 0.0001 by one-way ANOVA analysis of variance compared with control group.

Figure 3

Targeting ADA synergistically enhances anticancer effects of cordycepin in uveal melanoma. (A,B) Synergistic effects (A) and scores (B) of cordycepin and EHNA in 92.1, Mel202 and MP46 cells. (C) Combination treatment of 3′-deoxyadenosine and EHNA in uveal melanoma cells for cell growth assays. (D,E) Effects of cordycepin and EHNA on anchorage independent (D) and dependent (E) colony formation abilities in uveal melanoma. (F,G) Inhibition of cell migrations of uveal melanoma by combination treatment of 1 µM pentostatin and 10 µM cordycepin. The cells were seeded onto the Transwells coated with gelatin. After incubation for 12 h, the migratory cells on the bottom of the membrane were stained with crystal violet solution and counted. Values represent mean ± SD of experiments conducted in quadruplicate, or sextuplicate or octuplicate. **** p < 0.0001 by one-way ANOVA analysis of variance compared with control group. (H) Antitumor effects of cordycepin in mice bearing xenograft tumors of MP46 uveal melanoma cells. (I) Pictures of xenografts from MP46 uveal melanoma cells. (J) Tumor weights in vehicle- or compound treated mice. Values represent mean ± SEM of experiments. CT, control; Pento 2 mg/kg, pentostatin 2 mg/kg b.w. treatment; 3′-DA 20 mg/kg; Combo 1, combination treatment with pentostatin 1 mg/kg b.w. and 3′-deoxyadenosine 2 mg/kg b.w.; Combo 2, combination treatment with pentostatin 2 mg/kg b.w. and cordycepin 2 mg/kg b.w. ** p < 0.01, **** p < 0.0001 by one-way ANOVA analysis of variance compared with control group.

Figure 4

Apoptotic induction in uveal melanoma cells. (A,B) Uveal melanoma cells were treated with cordycepin only, or combination treatment with 3′-deoxyadenosine and ADA inhibitors, and the expression cleaved poly-(ACP-ribose) polymerase (PARP) was analyzed by Western blot analysis. (C,D) Cells were treated with 10 µM cordycepin and 1 µM EHNA for 48 h, and the distribution of Annexin V positive cells was analyzed by flow cytometry. Values represent mean ± SD of experiments conducted in triplicate. **** p < 0.0001 by one-way ANOVA analysis of variance compared with the control group.

Figure 4

Apoptotic induction in uveal melanoma cells. (A,B) Uveal melanoma cells were treated with cordycepin only, or combination treatment with 3′-deoxyadenosine and ADA inhibitors, and the expression cleaved poly-(ACP-ribose) polymerase (PARP) was analyzed by Western blot analysis. (C,D) Cells were treated with 10 µM cordycepin and 1 µM EHNA for 48 h, and the distribution of Annexin V positive cells was analyzed by flow cytometry. Values represent mean ± SD of experiments conducted in triplicate. **** p < 0.0001 by one-way ANOVA analysis of variance compared with the control group.

Figure 6

Anticancer effects of cordycepin in other aggressive cancer cells. (A) ADA protein expression levels were measured by Western blot in retinoblastoma, AT/RT, and DIPG cell lines. (B–D) Cells were treated with for 5 days and then cell titer blue was used to measure viable cells. (E) Cells were treated with cordycepin for 2 days, and then cleaved PARP protein level was evaluated by Western blot. (F–I) Cells were treated with cordycepin and 1 µM of pentostatin for 5 days and then cell titer blue was used to measure viable cells. Values represent mean ± SD of experiments conducted in sextuplicate. *** p < 0.001, **** p < 0.0001 by one-way ANOVA analysis of variance compared with control group.

Figure 7

Graphical abstract for the action mechanism of cordycepin. Proposed mechanism of action.