CUDC‐907, a novel dual PI3K and HDAC inhibitor, in prostate cancer: Antitumour activity and molecular mechanism of action

Abstract Targeting the androgen receptor (AR) signalling pathway remains the main therapeutic option for advanced prostate cancer. However, resistance to AR‐targeting inhibitors represents a great challenge, highlighting the need for new therapies. Activation of the PI3K/AKT pathway and increased expression of histone deacetylases (HDACs) are common aberrations in prostate cancer, suggesting that inhibition of such targets may be a viable therapeutic strategy for this patient population. Previous reports demonstrated that combination of PI3K inhibitors (PI3KIs) with histone deacetylase inhibitors (HDACIs) resulted in synergistic antitumour activities against preclinical models of prostate cancer. In this study, we demonstrate that the novel dual PI3K and HDAC inhibitor CUDC‐907 has promising antitumour activity against prostate cancer cell lines in vitro and castration‐resistant LuCaP 35CR patient‐derived xenograft (PDX) mouse model in vivo. CUDC‐907‐induced apoptosis was partially dependent on Mcl‐1, Bcl‐xL, Bim and c‐Myc. Further, down‐regulation of Wee1, CHK1, RRM1 and RRM2 contributed to CUDC‐907‐induced DNA damage and apoptosis. In the LuCaP 35CR PDX model, treatment with CUDC‐907 resulted in significant inhibition of tumour growth. These findings support the clinical development of CUDC‐907 for the treatment of prostate cancer.


| INTRODUC TI ON
Prostate cancer is the second most commonly diagnosed cancer and the second leading cause of death from cancer in men in the United States, with 174,650 new cases and 31,620 deaths estimated in 2019. 1 The androgen receptor (AR) signalling pathway is critical for the growth and survival of prostate cancer cells, including those of lethal castration-resistant prostate cancer (CRPC), demonstrating that AR is a valid therapeutic target for treating prostate cancer. Accordingly, androgen-deprivation therapy has been the standard treatment for patients with metastatic disease, and the next-generation AR-targeting inhibitors, such as abiraterone and enzalutamide, have been the standard of care for patients with CRPC. [2][3][4] However, resistance to these ARtargeting therapies represents a great challenge for the treatment of these patients, 5 highlighting the need for new therapies.
The PI3K pathway plays an important role in cancer cell proliferation, growth and survival 6 and is frequently activated in different types of cancer, including prostate cancer. Approximately 15% of primary prostate cancer and 50% of CRPC harbour alterations in this pathway, as determined by sequencing analysis. [7][8][9] In addition, aberrant activation of the PI3K pathway is associated with poor prognosis, metastasis and resistance to therapy in prostate cancer patients. [10][11][12] Although small molecule inhibitors targeting the PI3K pathway have demonstrated promising preclinical activity, clinical efficacy has been limited.
These disappointing results are likely due to cross-talk between the PI3K pathway and other survival pathways, suggesting that PI3K inhibitors must be used in combination with inhibitors targeting other survival pathways to ensure successful eradication of cancer cells. [13][14][15][16] Previous studies have shown that histone deacetylases (HDACs) are up-regulated in prostate cancer and their expression levels correlate positively with Gleason scores and cell proliferation. 17,18 Preclinical studies have revealed that combination of PI3K inhibitors with HDAC inhibitors (HDACIs) show promising antitumour activity against solid tumours, including prostate cancer. [19][20][21][22] As a result, CUDC-907, an orally available small molecule that dually inhibits HDAC (class I, II and IV, including HDAC1, 2, 3, 6, 10 and 11) and PI3K (class I α, β and δ) enzymes, was rationally designed and synthesized. 23 CUDC-907 has demonstrated substantial antitumour activity in multiple preclinical models of both solid tumours and haematological malignancies and clinical efficacy in relapsed/refractory lymphoma and multiple myeloma. [23][24][25][26][27][28][29] Furthermore, CUDC-907 is being tested in Phase I and II clinical trials for the treatment of multiple myeloma, lymphoma and advanced/relapsed solid tumours in adults and children (www.clini caltr ials.gov). CUDC-907 was reported to be well tolerated at a paediatric recommended phase II dose administered orally, with mean T max of 2.5 hours and T 1/2 of 1.4 hours. 30 In this study, we investigated CUDC-907 in prostate cancer cells. We found that CUDC-907 induces apoptosis in prostate cancer cells and that this is at least partially mediated by Mcl-1, Bcl-xL, Bim and c-Myc. We also found that CUDC-907 treatment induces DNA damage associated with down-regulation of CHK1, Wee1 and the two subunits of ribonucleotide reductase (RR), RRM1 and RRM2. Further, our in vivo studies revealed that CUDC-907 has potential for the treatment of prostate cancer.

| Cell culture
LNCaP, C4-2, C4-2B, LAPC-4, PC-3, DU145, 22Rv1 and HEK-293T cells were purchased from the American Type Culture Collection (ATCC). LNCaP95 cells were kindly provided by Dr Alan Meeker at Johns Hopkins University. The cell lines were cultured in RPMI 1640 with 10% foetal bovine serum (FBS, Thermo Fisher Scientific) and 2 mmol/L L-glutamine, plus 100 U/mL penicillin and 100 µg/mL streptomycin (except LNCaP95, which was cultured in RPMI 1640 F I G U R E 1 CUDC-907 treatment decreases viable cells and induces apoptosis in prostate cancer cell lines, and shows promising antitumour activity in the LuCaP 35CR patient-derived xenograft (PDX) model. A, Prostate cancer cell lines were treated with variable concentrations of CUDC-907 in 96-well plates for 72 h. The SRB assay was performed to determine viable cells. Viable cells in the vehicle control wells were set at 100%. Data are graphed as means of duplicates ± standard error from three independent experiments. B, IC50 values of CUDC-907 from three independent experiments were calculated and graphed as means ± standard error. C, 22Rv1 and LNCaP cells were treated with variable concentrations of CUDC-907 for 48 h and then subjected to Annexin V/PI staining and flow cytometry analysis. Data of apoptosis (left and middle panels) and viable cells (right panel) are presented as means of triplicates ± standard error from one representative experiment. ***indicates P < .001. D, 22Rv1 and LNCaP cells were treated as indicated in panel C. Whole cell lysates were subjected to Western blotting and probed with the indicated antibodies. cf-PARP, cleaved PARP. E, LNCaP cells were infected with NTC-, Bax-or Bak-shRNA lentivirus for 5 h, then washed and cultured in fresh media for 19 h before treated with CUDC-907 for 48 h. Knockdown of Bax or Bak was confirmed by Western blotting (left panel). The NTC-shRNA, Bax-shRNA and Bak-shRNA knockdown cells treated with CUDC-907 were subjected to Annexin V/PI staining and flow cytometry analysis (right panel). ***indicates P < .001 compared to the no drug treatment control, while ### indicates P < .001 compared to CUDC-907-treated NTC-shRNA cells. F&G, LuCaP 35CR tumour bits were inoculated into castrated male nude mice. When the tumours reached ∼250 mm 3 , the mice were treated with 100 mg/kg CUDC-907 via oral gavage for 19 d. Mean mouse bodyweights (panel F) and mean tumour volumes (panel G) are shown. *indicates P < .05, while **indicates P < .01 compared to the vehicle control group F I G U R E 2 CUDC-907 treatment inhibits PI3K and HDACs, but fails to inactivate ERK in prostate cancer cells. A&B, 22Rv1 (panel A) and LNCaP (panel B) cells were treated with variable concentrations of CUDC-907 for 6-48 h. Whole cell lysates were subjected to Western blotting and probed with the indicated antibodies. The fold changes for the densitometry measurements, normalized to β-actin and then compared to no drug control, are shown below the corresponding blot. C-F, 22Rv1 and LNCaP cells were treated with 100 nmol/L CUDC-907 with or without 500 nmol/L ERK inhibitor SCH772984 for 48 h. Whole cell lysates of 22Rv1 (panel C) and LNCaP (panel E) cells were subjected to Western blotting. The CUDC-907 treated 22Rv1 (panel D) and LNCaP (panel F) cells were subjected to Annexin V/PI staining and flow cytometry analyses. ***indicates P < .001 compared to no drug treatment control, while ### indicates P < .001 compared to individual drug treatment with 10% charcoal-stripped FBS), in a 37°C humidified atmosphere containing 5% CO2/95% air. Cell lines were regularly tested for the presence of mycoplasma utilizing the PCR method as described by Uphoff and Drexler. 31 AR expression status is summarized in Table S1.

| SRB assay
Viable prostate cancer cells were measured by using the SRB (Sulphorhodamine B, Sigma-Aldrich) assay, as previously described. 32

| Antibodies and western blot analysis
Anti-H4, anti-ac-Tubulin, anti-CDK1, anti-p-CDK1(Y15), anti-CDK2, Western blotting was performed as described previously. 33 Briefly, whole cell lysates were prepared by sonication in 10 mmol/L Tris-Cl, pH 7.0, containing 1% SDS, protease inhibitors and phosphatase inhibitors (Roche Diagnostics). The samples were separated by electrophoresis on SDS-polyacrylamide gels and transferred onto polyvinylidene difluoride membranes (Thermo Fisher Scientific). After blocking in TBS buffer (150 mmol/L NaCl, 10 mmol/L Tris, pH 7.4) containing 5% fatfree milk for 1 hour at room temperature, the blots were incubated with a primary antibody overnight at 4°C and then incubated with a fluorescent-labelled secondary antibody for 1 hour at room temperature. Immunoreactive proteins were visualized using the Odyssey Infrared Imaging System (Li-Cor). Western blots were repeated at least three times and one representative blot is shown.

| Alkaline comet assay
LNCaP and 22Rv1 cells were treated with CUDC-907 for 18 hours and then subjected to alkaline comet assay, as previously described. 34 Slides were stained with SYBR Gold (Thermo Fisher Scientific) and then imaged on an Olympus BX-40 microscope equipped with a DP72 microscope camera and Olympus cellSens Dimension software (Olympus America Inc). Approximately 50 comets per gel were scored using CometScore (TriTek Corp). The median per cent DNA in the tail was calculated and graphed as mean ± standard error.

| Quantification of gene expression by real-time RT-PCR
Total RNA was extracted using Total RNA Kit I (Omega Bio-Tek).

| LuCaP 35CR patient-derived xenograft (PDX) mouse model
Male CB17 SCID mice were obtained from Charles River at 4-6 weeks of age. After 1 week of adaptation, mice were castrated via a scrotal approach. On day 2 after castration, mice were inoculated subcutaneously with LuCaP 35CR tumour bits as described. 37 When the tumours reached ~250 mm 3

| Statistical analysis
Statistical analyses were performed with GraphPad Prism 5.0. Error bars represent ± SEM. Statistical significance was determined with pair-wise two-sample t test (two-tailed). The level of significance was set at P < .05.

| CUDC-907 treatment decreases viable cells and induces apoptosis in prostate cancer cell lines and shows antitumour activity in the LuCaP 35CR PDX model
To  Figure 1G). At the end of CUDC-907 treatment, the average tumour size in the vehicle control group was 835 mm 3 , while that in the CUDC-907 group was 333 mm 3 , demonstrating a ∼60% inhibition of tumour growth by CUDC-907 ( Figure 1G). Taken together, the data shown in Figure 1 demonstrate that CUDC-907 has promising antitumour activity against preclinical models of prostate cancer both in vitro and in vivo.

| Mcl-1, Bcl-xL and Bim play important roles in CUDC-907-induced apoptosis
It has been well documented that inhibition of the PI3K pathway and/or HDACs leads to changes of protein levels of the Bcl-2 family members, such as down-regulation of anti-apoptotic Mcl-1 and Bcl-xL and upregulation of pro-apoptotic Bim. 20,[39][40][41] It is conceivable that CUDC-907 exerts its antitumour activity against prostate cancer cells by a similar mechanism. As shown in Figure 3A Figure 4A,B, CUDC-907 treatment resulted in induction of γH2AX (a potential biomarker of DNA double-strand breaks) starting at 12 hours post drug treatment, which was prior to cell death ( Figure 3C,D), suggesting that CUDC-907 treatment-induced DNA damage in prostate cancer cells. This was accompanied by decreased CHK1, p-CDC25C, p-CDK1, p-CDK2 and Wee1 ( Figure 4A,B). In

| CUDC-907 treatment down-regulates c-Myc in prostate cancer cells
The oncoprotein c-Myc is frequently activated and promotes tumour development in prostate cancer, 46 Taken together, these data demonstrate that c-Myc plays an important role in CUDC-907-induced apoptosis in prostate cancer cells.

| D ISCUSS I ON
Resistance to AR-targeted therapy remains a major challenge in the treatment of prostate cancer, underscoring the need for novel therapeutic approaches. 5 Simultaneous inhibition of PI3K and HDACs has shown promising anticancer activity in solid tumours, including F I G U R E 5 CHK1, RRM1, RRM2 and Wee1 play an important role in CUDC-907-induced DNA damage and apoptosis. A, LNCaP cells were infected with Precision LentiORF Wee1, CHK1, RRM1, RRM2 or RFP control lentivirus for 5 h, then washed and cultured in fresh media for 48 h prior to adding 2 μg/mL blasticidin. Whole cell lysates of the antibiotic-resistant cells were subjected to Western blotting and probed with the indicated antibodies. The fold change densitometry measurements, normalized to β-actin and then compared to no drug treatment control, are indicated. OE, overexpression. B&C, The antibiotic-resistant cells were treated with 100 nmol/L CUDC-907 for 18 h and then subjected to alkaline comet assay analyses. Representative images are shown (panel B). Data are graphed as mean per cent DNA in the tail from 3 replicate gels ± standard error (panel C). *indicates P < .05 compared to CUDC-907-treated RFP cells. D, The antibiotic-resistant cells were treated with 100 nmol/L CUDC-907 for 48 h and then subjected to Annexin V/PI staining and flow cytometry analysis. ***indicates P < .001 compared to vehicle control-treated cells, while ### indicates P < .001 compared to CUDC-907-treated RFP cells. E&F, 22Rv1 (penal E) and LNCaP (panel F) were treated with CUDC-907 alone or in combination with the Wee1 inhibitor MK-1775, CHK1 inhibitor LY2603618 or RR inhibitor hydroxyurea for 6 h and then subjected to alkaline comet assay analyses. Representative images are shown (left panels). Data are graphed as mean per cent DNA in the tail from 3 replicate gels ± standard error (right panels). **indicates P < .01 and ***indicates P < .001 compared to the no drug treatment control and individual drug treatments. G&H, 22Rv1 (panel G) and LNCaP ( Transcription factor c-Myc plays a critical role in cancer initiation and progression and is one of the most frequently deregulated oncogenes. 53 c-Myc is also one of the most up-regulated genes in prostate cancer. 48 In line with results in DLBCL and AML cells, 26 Reciprocal inhibition has been reported for the PI3K/AKT pathway and the AR pathway. 11,54 AR can be activated upon PI3K/AKT inhibition. 11,54 Interestingly, we did not observe any significant change in AR activity in either LNCaP or 22Rv1 cells after treatment with CUDC-907. This might be due to CUDC-907 inhibition of HDACs, as HDACs are known to positively regulate AR signalling. 55 In conclusion, we demonstrate that CUDC-907 induces apoptosis

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data supporting the findings of this study are available from the corresponding author upon reasonable request.