Endocrine-Related Cancer (2009) 16 401–413

Azacitidine improves antitumor effects of and in aggressive prostate cancer models

Claudio Festuccia*, Giovanni Luca Gravina1*, Anna Maria D’Alessandro2, Paola Muzi, Danilo Millimaggi 2, Vincenza Dolo2, Enrico Ricevuto3, Carlo Vicentini2 and Mauro Bologna4

Department of Experimental Medicine, Chair of General Pathology, University of L’Aquila, Via Vetoio, Coppito-2, 67100 L’Aquila, Italy, 1Division of Radiotherapy, Department of Experimental Medicine, 2Department of Health Science, 3Division of Clinical Oncology, Department of Experimental Medicine and 4Department of Basic and Applied Biology, University of L’Aquila, L’Aquila, Italy (Correspondence should be addressed to C Festuccia; Email: [email protected]) *(C Festuccia and G L Gravina contributed equally to this work)

Abstract One of the major obstacles in the treatment of hormone-refractory prostate cancer (HRPC) is the development of chemoresistant tumors. The aim of this study is to evaluate the role of as chemosensitizing agent in association with docetaxel (DTX) and cisplatin using two models of aggressive prostate cancer, the 22rv1, and PC3 cell lines. Azacitidine shows antiproliferative effects associated with increased proportion of cells in G0/G1 and evident apoptosis in 22rv1 cells and increased proportion of cells in G2/M phase with the absence of acute cell killing in PC3 cells. In vivo, azacitidine (0.8 mg/kg i.p.) reduced tumor proliferation and induced apoptosis in both xenografts upmodulating the expression of p16INKA, Bax, Bak, p21/WAF1, and p27/KIP1, and inhibiting the activation of Akt activity and the expression of cyclin D1, Bcl-2, and Bcl-XL. In vitro treatments with azacitidine lead to upregulation of cleaved caspase 3 and PARP. BCl2 antagonists, such as HA-14-1, enhanced the effects of azacitidine in these two prostate cancer models. In addition, azacitidine showed synergistic effects with both DTX and cisplatin. In vivo this agent caused tumor growth delay without complete regression in xenograft systems. Azacitidine sensitized PC3 and 22rv1 xenografts to DTX and cisplatin treatments. These combinations were also tolerable in mice and superior to either agent alone. As DTX is the standard first-line for HRPC, the development of DTX-based combination therapies is of great interest in this disease stage. Our results provide a rationale for clinical trials on combination treatments with azacitidine in patients with hormone-refractory and chemoresistant prostate tumors. Endocrine-Related Cancer (2009) 16 401–413

Introduction prostate cancers, the results from two recent Prostate cancer represents a global public health randomized clinical trials indicated that docetaxel problem. Worldwide, it is the second most common (DTX)-based chemotherapy improved survival in noncutaneous cancer in men, accounting for w10% of patients with hormone-refractory prostate cancer, male cancers (Haas et al. 2008). In recent years, 5 year HRPC (Patel et al. 2005, Armstrong et al. 2007). survival rates for prostate cancer have been ranked Docetaxel and are two food and drug the third highest of all cancers (Desireddi et al. 2007, administration – approved chemotherapeutic agents for Svatek et al. 2008). Although chemotherapy histori- clinical treatment of advanced HRPC and DTX is the cally has had limited utility in treating advanced standard first-line chemotherapy in this disease stage.

Endocrine-Related Cancer (2009) 16 401–413 Downloaded from Bioscientifica.comDOI: 10.1677/ERC-08-0130 at 09/25/2021 07:04:29AM 1351–0088/09/016–401 q 2009 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.orgvia free access C Festuccia and G L Gravina et al.: Azacitidine effects in advanced prostate cancer models

Thus, the development of DTX-based combination regression as single agent, whereas improved anti- therapies is of great interest for HRPC. Some clinical tumor effects of DTX and cisplatin resulting these studies supported -based therapy, a promising combinations well tolerable in mice. Therefore, use of platform to develop new combinational schedule, DNA methylation inhibitors might be a suitable which remains a high priority for many ongoing therapeutic tool for the amplification of pharmacologic studies (Patel et al. 2005, Armstrong et al. 2007 and for responses of HRPC versus first- and second-line a review, see Calabro` and Stemberg 2007). In addition . a recent report shows that single-agent improves progression-free survival (PFS) as second- line chemotherapy for patients with HRPC and Materials and methods provides further evidence that platinum salts may Reagents have a role in the treatment of HRPC (Nakabayashi All the materials for tissue culture were purchased et al.2008, Ross et al. 2008). Nevertheless, improvement from Hyclone (Cramlington, NE, USA). Antibodies in the efficacy of chemotherapy is urgently needed for when not otherwise specified were purchased from patients with chemotherapy-resistant HRPC. Santa Cruz (Santa Cruz, CA, USA). Plasticware was It is now well established that cancer cells exhibit a obtained from Nunc (Roskilde, Denmark). Azacitidine number of genetic and epigenetic defects in the (Vidaza) was obtained in collaboration with Celgene machinery that governs programmed cell death and Corporation (Summit, NJ, USA). DNA methyltransfer- that sabotage of apoptosis is one of the principal factors ase activity was evaluated in nuclear cell extracts by a aiding in the progression and treatment of neoplastic colorimetric EpiQuik DNA methyltransferase Activity diseases. Epigenetic modifications, specifically DNA Assay Kit (BioVision, Mountain View, CA, USA). hypermethylation, are believed to play an important role in the downregulation of genes important for Cell lines protection against apoptosis. In addition to classical genetic abnormalities, epigenetic modifications have The two aggressive prostate cancer models (22rv1 and emerged as a central driving force in the molecular PC3 cell lines) were obtained from American Tissue pathology of prostate cancer (McCabe et al. 2006, Culture Collection (ATCC, Rockville, MD, USA) and Murphy et al.2008). Several years ago, Plumb et al. grown as recommended. (2000) showed that resistance to in ovarian cancer cells was mediated by hypermethylation and Cell proliferation inhibition assay loss of function of the MLH1 mismatch repair gene. Cell proliferation was evaluated by 3-(4, Of interest, decitabine, a deoxyderivative of aza-CR, a 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro- cytosine analogue with hypomethylating properties, mide (MTT; Sigma) assay. Briefly, cells were seeded reverses the downregulation of expression of some in 96-well tissue culture plates (Costar, Corning, NY, membrane transporters in vitro (Plumb et al. 2000), USA) at 3000 cells per well per 150 ml. At 24 h after raising the untested possibility that repeated treatment seeding, cells were treated with 50 ml serial concen- could lead to a progressive increase in efficacy. trations of azacitidine, DTX, cisplatin or the combina- In this study, we used azacitidine and its pharma- tion of azacitidine plus DTX and cisplatin. After 72 h cological formulation (Vidaza) to verify if this agent of drug exposure, MTT (100 ml) solution (2 mg/ml) was able as single agent to alter tumor cell proliferation was added to each well to incubate for 2 h. Absorbance and apoptosis, as well as to improve the sensitivity at 570 nm was recorded using a 96-well plate reader. against DTX and cisplatin in vitro and in vivo using Inhibition curves were drawn by means of values two models of aggressive prostate cancer. 22Rv1 is an obtained by OD percentages versus control for each androgen-responsive human prostate carcinoma cell concentration and IC50 values were calculated by the line derived from a primary prostate tumor that GraFit method (Erithacus Software Limited, Staines, expresses mutant (H874Y) androgen receptors (AR), UK). The Chou and Talalay (1984) combination index whereas PC3 is an androgen irresponsive, AR negative (CI) analysis is a well-established index to determine prostate carcinoma cell line. Our results indicate that the pharmacologic interaction of two drugs. When azacitidine has antiproliferative effects in both cell CIZ1 (represents an additive effect of the two drugs), lines and showed pro-apoptotic effects in 22rv1 cells. the CI formula is in the same form as a traditional In vitro showed synergistic effects with DTX and isobologram equation. Synergism is defined as more cisplatin. Its pharmacological preparation (Vidaza) than the expected additive effect with CI !1 and caused tumor growth delay without complete stasis or antagonism is defined as CI O1.

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access 402 www.endocrinology-journals.org Endocrine-Related Cancer (2009) 16 401–413

Apoptosis analysis by flow cytometry the mean tumor weight at the end of experiments. Treated and untreated cells were collected and fixed Fractional tumor volume (FTV) for each treatment for 30 min by the addition of 1 ml 70% ethanol. After group was calculated as the ratio between the mean 30 min, the cells were pelleted by centrifugation tumor volumes of treated and untreated animals. This (720 g; 5 min), and resuspended in 1 ml DNA staining was performed for treatment a (FTVa), for treatment C C solution (PBS containing 200 mg/ml RNase A, b (FTVb), and for treatment a b(FTVa b). 20 mg/ml propidium iodide plus 0.1% Triton X-100) Combination indices were assayed as described and stained by incubation at room temperature for above. Animals were killed by spinal cord dislocation 60 min. All cells were then measured on a FACScan and tumors were subsequently removed surgically. A flow cytometer (Becton Dickinson, Bedford, MA, part of the tumor was directly frozen in liquid nitrogen USA) and analyzed using Cell Quest software (Becton for protein analysis and the other part was fixed in Dickinson). Resulting DNA distributions were analy- paraformaldehyde overnight for histochemical evalu- zed by Mod.t (Verity Software House, Topsham, ME, ations and immunohistochemical analyses including USA) for the proportion of cells in sub-G0 (apoptotic endothelial (CD31 and vWF; Dako, Glostrup, cells), G1, S, and G2-M phases of the . Denmark), proliferative (Ki67; Dako), apoptosis (bcl2, p21, p27, death receptors and their ligands Santa Cruz), and signaling (p-Akt; Rockland Immuno- In vivo treatments chemicals, Gilbertsville, PA, USA) marker analysis. Male CD1 nude mice (Charles River, Milan, Italy) Anti-DNMT1, DNMT3a, and DNMT3b antibodies were maintained under the guidelines established by were purchased from Biocarta (Hamburg, Germany). our Institution (University of L’Aquila, Medical Indirect immunoperoxidase stain of tumor xenografts School and Science and Technology School Board was performed on paraffin embedded with four Regulations, complying with the Italian government micrometer tissue sections. Briefly, sections were regulation no. 116, January 27 1992, for the use of incubated with primary antibodies for a night at 4 oC. laboratory animals). All mice received s.c. flank Next, avidin–biotin assays were done using the injections of 1!106 PC3 and 22v1 cells. Tumor Vectastain Elite kit obtained from Vector Laboratories. growth was assessed by bi-weekly measurement Mayer’s hematoxylin was used as nuclear counterstain. of tumor diameters with a Vernier caliper Slides were analyzed separately from CF and PM. (length!width). Tumor weight was calculated accord- 3 ing to the formula: TW (mg)Ztumor volume (mm )Z Martius yellow-brilliant crystal scarlet blue 2 d !D/2, where d and D are the shortest and the longest technique diameters respectively. Treatments used in our in vivo study were started when tumor volumes reached about Stains for these techniques were purchased from HD 80 mm3 (Day 0) and were stopped after 28 days. Group Supplies (Aylesbury, UK). This technique was used to 1: 10 mice received i.p. injections of 100 l PBS for analyze the presence of red cells dispersed in tumor consecutive 7 days. Group 2: 10 mice received i.p. tissue and present in blood vessels as well as to injections of 100 l of aza-CR (0.8 mg/kg) for consecu- evaluate better the presence of micro-thrombi and tive 7 days. Group 3: 10 mice received i.p. injections of bleeding zones. Martius yellow, a small molecule dye, DTX (7.5 mg/kg per week) every week for a maximum together with phosphotungstic acid in alcoholic of four administrations. Group 4: 10 mice received i.p. solution stains red cells. Early fibrin deposits may be injections of 100 l of aza-CR (0.8 mg/kg) for consecu- colored, but the phosphotungstic acid blocks the tive 7 days followed by a dose of DTX (7.5 mg/kg per staining of muscle, collagen, and connective tissue week) every week for a maximum of three adminis- fibers. Brilliant crystal scarlet, a medium-sized trations. Group 5: 10 mice received i.p. injections of molecule, stains muscle, and mature fibrin. Phospho- 5.0 mg/kg per day cisplatin for 10 consecutive days. tungstic acid removes any red stain from the collagen. Group 6: mice received i.p. injections of 100 l of aza- The large molecule dye aniline blue stains the collagen CR (0.8 mg/kg) for consecutive 7 days followed by i.p. and old fibrin. injections of 5.0 mg/kg per day cisplatin for 10 consecutive days. Experiments were stopped at 28th Hemoglobin assay day. We replicate the experiments in order to analyze a Tumor hemoglobin levels were quantified as described total of 20 animals for group. The effect of azacitidine elsewhere (14). Tumors were homogenized in double- alone or in combination with cisplatin (CP) or DTX on distilled water. Eighty microliters of homogenates tumor growth was evaluated measuring, for all groups, were mixed with 1 ml Drabkin’s solution and

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access www.endocrinology-journals.org 403 C Festuccia and G L Gravina et al.: Azacitidine effects in advanced prostate cancer models incubated for 15 min at room temperature. After Results centrifugation at 400 g for 5 min, the supernatants In vitro experiments: effects of azacitidine were subjected to absorbance measurement at 540 nm. as single agent The absorption that is proportional to hemoglobin concentration, was divided by tumor weight. Initially, we analyzed the effects of azacitidine on DNA methylation and expression of DNMT1, DNMT3a, and DNMT3b. After 48 h of treatment Assessment of global methylation levels with azacitidine, the decrease in the DNMT activity was maximal after 10 days and it was maintained up to Proteins were extracted from untreated and in vivo 2 weeks (Fig. 1A). Simultaneously, a significant Aza-CR-treated tumors and cells at 7, 14, 21, and decrement in DNMT1 and DNMT3b expression levels 28 days and DNMT activity were assayed. As occurred with a peak observed at 5–10 days (Fig. 1B). described, frozen tumors were pulverized in a liquid Conversely, azacitidine treatment did not affect the nitrogen-cooled Thermovac pulverizer. The resulting protein expression levels of DNMT3a. Antiprolifera- powders were homogenized in 10 volumes of a 10 mM tive effect of azacitidine treatment was higher in 22rv1 Tris–HCl buffer, pH 7.4, containing 1 mM EDTA, cells than in PC3 cells with IC50 of 0.5 and 2.5 mM 0.5 mM dithiothreitol, 10 mM sodium molybdate, respectively. Treatment of with azacitidine resulted in phosphatase inhibitor cocktail 2 with a dilution 1/100, a dose- and time-dependent inhibition of growth with and protease inhibition cocktail 2 with a dilution 1/100, G1 phase arrest associated to apoptosis in 22rv1 cells, both from Sigma-Aldrich. The homogenates were whereas this drug arrested PC3 cells in G2/M phase centrifuged for 1 h at 105 000 g (C4 8C) and the without acute cell killing. Also, azacitidine increased supernatants (cytosols) were used for protein the levels of p21, p27, p53, Bcl-Xs, and Bax and to determination by immunoblotting. Total protein con- reduce Bcl2 and Bcl-Xl expression in 22rv1 cells tent was measured using the bicinchoninic acid assay. (Fig. 2A). This was also associated with an increased expression of cleaved caspase 3 and PARP. The expression of p21, Bcl2, caspase 3, and PARP was Western blot analysis Cells were washed with cold PBS and immediately lysed with 1 ml lysis buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 50 mM NaF, 1 mM sodium orthovanadate, 30 mM p-nitrophenyl phosphate, 10 mM sodium pyrophosphate, 1 mM phenylmethyl- sulfonyl fluoride, 10 mg/ml aprotinin, and 10 mg/ml leupeptin). Flash-frozen tissue samples were crushed using liquid nitrogen pre-chilled mortar and pestles. Upon addition of lyses buffer, tissues were further homogenized with an electric homogenizer and centrifuged in a microfuge a top spin. Supernatants were collected and analyzed. Cell lysates and tissue extracts were electrophoresed in 7% SDS-PAGE, and separated proteins transferred to nitrocellulose and probed with the appropriate antibodies using the conditions recommended by the antibody suppliers.

Figure 1 DNMT activity (A) and expression of DNMT1, DNMT3a, and DNMT3b (B) measured at different time after a Statistics analysis 48 h 1.0 mM Aza-CR treatment in 22rv1 and PC3 cells. In PC3 cells, a decrement of about twofold in DNMT1 and fivefold in S D Continuous data were expressed as mean and . . and DNMT3b expression was observed after 5–10 days of were compared using an umpired Student’s t-test. treatment with azacitidine. The levels of DNMT1 were restored Categorical data were analyzed by the exact Fisher’s after 15 days, whereas the levels of DNMT3b were maintained lower up to 20 days (about twofold). In 22rv1 cells, the levels of test. P value less than 0.05 were considered statistically DNMT1 and DNMT3b were weakly reduced also if the overall significant. DNMT activity was significantly decreased.

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access 404 www.endocrinology-journals.org Endocrine-Related Cancer (2009) 16 401–413 not changed in PC3 cells following azacitidine with Aza-CR in PC3 (Fig. 3A) and 22rv1 cells treatment, whereas the expression of protein associated (Fig. 3B). These results suggest that the amount of to G2/M arrest such as p16INKA, cyclin B and cdk1, apoptosis is related to Bcl2 levels and the low levels of cdk2 and cdk4 were differentially modulated (Fig. 2B). apoptosis observed in PC3 cells might be partially The treatment with HA-14-1, a Bcl2 antagonist, was able associated with the higher expression of Bcl2. Aza-CR to induce more apoptotic events when co-administrated treatment in PC3 cells was also associated with the phosphorylation of p38 MAP kinase and with the reduced expression of activated Erk (Fig. 5B) which it is required for the induction of cell cycle arrest in the G2/M phase. However, the induction of p38 MAPK activity was able to start the apoptotic program by downmodulation of Bcl2 when Aza-CR treated PC3 cells were grown in polyhema-covered Petri dishes (Fig. 3C). The apoptosis mediated by absence of adhesion (anoikis) was induced by a p38 MAPK- dependent Bcl-2 downmodulation (Fig. 3D) because SB203580, a p38 MAPK inhibitor, treatment caused a significative slow down in Aza-CR/polyhema-dependent Bcl2 downmodulation.

In vitro effects of azacitidine in combination with DTX or cisplatin We analyzed the effects of combined treatments with azacitidine (0.5–1 mM for PC3 and 0.1–0.5 mM for 22rv1), DTX (0.5, 1.0, and 10 nM for both cells) and cisplatin (1, 2.5, and 10 mg/ml for both cells). Azacitidine sensitized versus DTX and cisplatin in PC3 and 22rv1 cells (Tables 1 and 2). In PC3 cells, the stronger antitumor effect was observed with azacitidine (1.0 mM) and DTX (1.0 nM (CIZ0.29)) and with azacitidine (1.0 mM) and cisplatin (10 mg/ml (CIZ0.27)). In 22rv1 cells, the highest antitumor effects were observed with azacitidine (0.5 mM) and

Figure 2 Western blot analysis on the expression of pro- apoptotic and antiapoptotic elements in 22rv1 (A) and PC3 (B) cells treated with 1, 2.5, and 5.0 mM aza-CR. In 22rv1 cells, aza- CR increased the expression of pro-apoptotic elements such as p21 (twofold with 2.5 and 5.0 mM aza-CR), p53 (four- to five-fold with 2.5 and 5.0 mM aza-CR), Bax (two- to four-fold with 2.5 and 5.0 mM aza-CR) and Bcl-XS (twofold with both 2.5 and 5.0 mM aza-CR) as well as the activity of caspase 3 (four- to six-fold with 2.5 and 5.0 mM aza-CR). The antiapoptotic element Bcl2 and Bcl-XL was reduced (about tenfold with both 2.5 and 5.0 mM aza-CR). In PC3 cells, no caspase 3 and PARP activity was showed and this was associated to no significant modulation of Bcl2 and p21 expression. Modulation in check point proteins of cell cycle was, however, shown. P16INK4A levels were significantly increased (about twofold with 5.0 mM aza-CR), CdK1 levels were increased of about two with both 1.0 and 2.5 mM aza-CR and of about fourfold with 5.0 mM aza-CR, CdK2 levels were weakly reduced (1.5-fold with 5.0 mM aza-CR), CdK4 levels were significantly reduced of about fourfold with 2.5 and 5.0 mM aza-CR and cyclin B1 levels were not modified. Each lane was loaded with 40 mg proteins and normalized versus actin. This experiment is representative of three individual experiments.

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access www.endocrinology-journals.org 405 C Festuccia and G L Gravina et al.: Azacitidine effects in advanced prostate cancer models

Figure 3 Role of Bcl2 in the absence of aza-mediated PC3 cell apoptosis. We used the Bcl2 inhibitor HA-14.1 to demonstrate that bcl2 overexpression in PC3 cells could be responsible for the absence of acute apoptosis induced by aza-CR treatment. HA-14.1 (10 mM) triggered apoptosis of PC3 cells both in basal culture condition and in the presence of aza-CR pretreatment (1.0 mM). HA-14.1 was synergistic with aza-CR in PC3 cells, whereas this drug was only additive in 22rv1 cells. These results are representative of three individual experiments. Each lane was loaded with 40 mg proteins.

DTX (1.0 nM (CIZ0.50)) and with azacitidine weight respect to controls (P!0.001) with a tumor (1.0 mM) and cisplatin (2.5 mg/ml (CIZ0.44)). growth delay (TGD) of 5.5 days. The proliferation index was also significantly reduced to 42.6% in In vivo effects of azacitidine as single drug or azacitidine treated tumors (P!0.001) with tumor followed by DTX or cisplatin apoptosis of about 5%. The vessel count was also The i.p. administration of Vidaza (0.8 mg/kg for 7 reduced to 18.7% (P!0.05). In 22rv1 xenografts consecutive days) on intact nude male mice inoculated (Table 4), we demonstrated a 65.5% reduction of tumor subcutaneously with 22rv1 and PC3 cells resulted in a weight in aza-CR treated mice respect to controls statistically significant tumor growth decrement (P!0.001), with a tumor growth delay (TGD) of (P!0.001) in both xenograft models (Fig. 4). In PC3 25.5 days. The proliferation index was also reduced to xenografts (Table 3), following azacitidine treatment 74.3% (P!0.001) and apoptosis was evident in about we demonstrated a 36.7% of reduction in tumor 15% of cells. The vessel count was reduced to 70.4%

Table 1 Pharmacological parameters of combination treatment with azacitidine and docetaxel in PC3 and 22rv1 cells

Treatment A (DTX) Treatment B (aza-CR) Combination treatment

Cells Dose (nM) GI Dose (mM) GI Expected GI Observed GI Increment (CI)

PC-3 0.5 0.95 0.5 0.90 0.85 0.54 1.57 (0.64) 1 0.75 0.70 0.25 2.80 (0.36) 1.0 0.90 0.5 0.90 0.80 0.31 2.58 (0.39) 1 0.75 0.70 0.20 3.50 (0.29) 10.0 0.75 0.5 0.90 0.70 0.25 2.80 (0.36) 1 0.75 0.50 0.20 2.50 (0.40) 22rv1 0.5 0.90 0.1 0.75 0.65 0.40 1.62 (0.62) 0.5 0.50 0.40 0.30 1.33 (0.75) 1.0 0.70 0.1 0.75 0.45 0.25 1.80 (0.55) 0.5 0.50 0.20 0.10 2.00 (0.50) 10.0 0.55 0.1 0.75 0.30 0.20 1.50 (0.67) 0.5 0.50 0.05 0.10 0.50 (2.00)

GIZinhibition (percent versus control).

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access 406 www.endocrinology-journals.org Endocrine-Related Cancer (2009) 16 401–413

Table 2 Pharmacological parameters of combination treatment with azacitidine and cisplatin in PC3 and 22rv1 cells

Treatment A (CP) Treatment B (aza-CR) Combination treatment

Cells Dose (mg/ml) GI Dose (mM) GI Expected GI Observed GI Increment (CI)

PC-3 1 1.00 0.5 0.90 0.90 0.60 1.50 (0.67) 1 0.75 0.75 0.45 1.67 (0.60) 2.5 0.90 0.5 0.90 0.80 0.40 2.00 (0.50) 1 0.75 0.65 0.32 2.03 (0.49) 10.0 0.80 0.5 0.90 0.70 0.25 2.80 (0.36) 1 0.75 0.55 0.15 3.67 (0.27) 22rv1 2.5 0.90 0.1 0.85 0.75 0.60 1.25 (0.80) 0.5 0.67 0.57 0.44 1.30 (0.77) 10.0 0.90 0.1 0.85 0.75 0.45 1.67 (0.60) 0.5 0.67 0.57 0.25 2.28 (0.44) 10.0 0.85 0.1 0.85 0.70 0.35 2.00 (0.50) 0.5 0.67 0.52 0.25 2.08 (0.48)

GIZinhibition.

(P!0.001). Similarly to that observed in vitro, the (i) Docetaxel and cisplatin alone reduced tumor levels of DNMT1 and DNMT3b were also lower in weight of PC3 xenografts by 41.0% (P!0.001) and azacitidine treated mice when compared with controls 22.2% (P!0.01), with a tumor growth delay of 5.1 and both in PC3 and 22rv1 xenograft models (data not 2.1 days respectively when compared with controls. shown). The proliferation index was also reduced to 49.9% The treatment with azacitidine followed by DTX or (P!0.001) and 23.6% (P!0.01) respectively in DTX- CP was also studied in vivo. An increase in the and cisplatin-treated tumors. Apoptosis was evident effectiveness of DTX and cisplatin was observed after only in DTX-treated tumors at levels of about 7% 7 days of co-administration with azacitidine both in (P!0.01). In this xenograft model, 7 days of treatment PC3 and 22rv1 xenograft models (Fig. 4 and Tables 3 with azacitidine followed by DTX or cisplatin resulted and 4). The results are as follows: in a significant decrease in tumor weight of 76.4%

Figure 4 Tumor proliferation in PC3 and 22rv1 xenografts in the presence of azacitidine alone or in combination with docetaxel (DTX) and cisplatin (CP). Group 1: 10 mice received (i.p.) injections of 100 ml PBS for consecutive 7 days. Group 2: 10 mice received i.p. injections of 100 ml of aza-CR (0.8 mg/kg) for consecutive 7 days. Group 3: 10 mice received i.p. injections of docetaxel (7.5 mg/kg per week) every week for a maximum of four administrations. Group 4: 10 mice received i.p. injections of 100 ml of aza-CR (0.8 mg/kg) for consecutive 7 days followed by a dose of docetaxel (7.5 mg/kg per week) every week for a maximum of three administrations. Group 5: 10 mice received i.p. injections of 5.0 mg/kg per day cisplatin for 10 consecutive days. Group 6: mice received i.p. injections of 100 ml of aza-CR (0.8 mg/kg) for consecutive 7 days followed by i.p. injections of 5.0 mg/kg per day cisplatin for 10 consecutive days. Proliferation was measured considering the variation of tumor volume (mm3) in the time (days). Controls were administered with saline and treated xenografts were administered with 0.8 mg/kg for the first 7 days of treatment. Animals were killed at 28th day of treatment.

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access www.endocrinology-journals.org 407 C Festuccia and G L Gravina et al.: Azacitidine effects in advanced prostate cancer models

Table 3 Antitumor activity of Vidaza in PC-3 tumors alone or in combination with docetaxel (DTX) or cisplatin (CP)

Treatment

Weight Tumor Dose of mice weight TGD PI Tumor-free Drug mg/kg Mice (grGS.D.) (mgGS.D.) (days) (Ki67 %) Apoptosis Vessels mice

Saline 20 25.0G2.1 585G135 37.3G4.5 !2 30.0G4.3 0/20 Vidaza 0.8 20 23.4G2.3 370G183 5.5 21.4G2.7 5.4G2.6 24.4G2.7 0/20 DTX 7.5 20 25.0G1.7 345G132 5.1 18.7G3.3 7.2G2.7 27.5G4.5 0/20 VidazaCDTX 0.8C7.5 20 21.8G2.6 138G84 25.5 1.3G1.7 18.2G3.2 7.5G2.5 3/20 CP 5 20 25.1G1.5 455G161 2.1 28.5G3.1 !2 28.5G3.5 0/20 VidazaCCP 0.8C5 20 20.8G2.8 244G58 15.2 6.9G0.6 14.3G2.1 13.4G4.4 5/20

TGD (tumor growth delay) was determined by the difference in time taken for the mean tumor volume to double in treated versus control animals. PI (proliferation index) was measured considering the mean of Ki67 positive cell percentageGS.D. measured on five random fields at !100. Apoptosis was measured as the percentage of tunnel positive cells G S.D. measured on five random fields (!400). Vessels were measured as the vessel number per field at !100.

(P!0.001 versus both Vidaza and DTX single The proliferation index was also reduced of 66.1% treatments) and of 58.3% (P!0.01 versus Vidaza (P!0.001) and 53.2% (P!0.001) respectively in and P!0.001 versus cisplatin) respectively. Moreover, DTX- and cisplatin-treated animals. Apoptosis was sequential regimens of azacitidine C DTX or evident in 10% (P!0.05) and 25% (P!0.01) of cells azacitidine C cisplatin were respectively associated in DTX- and cisplatin-treated animals respectively. with tumor growth delay values of 25.5 and 15.2 days: In this xenograft model, 7 days of treatment with reductions in the proliferation index of 96.5% azacitidine followed by DTX or cisplatin resulted (P!0.001 versus both single treatments) and 81.5% in a significant decrease in tumor weight of 68.5% (P!0.001 versus both single treatments); significant (P!0.05 versus DTX treatment, whereas not signi- increases in apoptosis of 18% (P!0.001 versus both ficant versus Vidaza single treatment) and 77.5% single treatments) and 14.3% (P!0.001 versus both (P!0.001 versus both single treatments) respectively. single treatments); and reductions in vessel counts of Additionally, sequential regimens of azacitidine C DTX w 75% (P!0.001 versus both single treatments) and or azacitidine C cisplatin were respectively associated 55% (P!0.001 versus Vidaza and P!0.01 versus with tumor growth delay values of 28.0 and 30.2 days; cisplatin treatments). Additionally, 15.0% (3/20, NS) reductions in the proliferation index of 79.6% (P!0.01 and 25.0% (5/20, (PZ0.056, NS)) of mice were tumor- versus Vidaza and P!0.001 versus DTX single free after 28 days of treatment. treatments) and 83.5% (P!0.001 versus both single (ii) In 22rv1 xenografts, DTX or cisplatin alone treatments); significant increases in apoptosis of reduced tumor weight of 46.8% (P!0.001) and 30.9% 18.2 and 34.3% (P!0.001 versus all single treat- (P!0.001), with a tumor growth delay of 17.3 and ments); reductions in vessel counts of about 7.5 days respectively when compared with controls. 75% (P!0.05 versus Vidaza and P!0.001 versus

Table 4 Antitumor activity of Vidaza in 22rv1 tumors alone or in combination with docetaxel (DTX) or cisplatin (CP)

Treatment

Weight of Tumor Dose mice weight TGD PI Tumor-free Drug mg/kg Mice (grGS.D.) (mgGS.D.) (days) (Ki67 %) Apoptosis Vessels mice

Saline 20 25.3G1.8 652G220 45.5G6.5 !2 38.5G5.0 0/20 Vidaza 0.8 20 24.1G2.0 225G55 25.5 11.7G0.7 15.4G3.0 11.4G0.8 1/20 DTX 7.5 20 25.0G1.7 305G121 17.3 15.4G5.1 10.2G1.5 17.5G3.0 0/20 VidazaCDTX 0.8C7.5 20 21.8G2.6 205G125 28.0 9.3G1.4 18.2G2.5 9.5G1.0 2/20 CP 5 20 25.1G1.5 450G101 7.5 21.3G3.9 25.5G2.5 7.5G1.5 1/20 VidazaCCP 0.8C5 20 20.8G2.8 147G66 30.2 7.5G2.0 34.3G8.0 7.4G4.4 6/20

TGD (tumor growth delay) was determined by the difference in time taken for the mean tumor volume to double in treated versus control animals. PI (proliferation index) was measured considering the mean of Ki67 positive cell percentage G S.D. measured on five random fields at !100. Apoptosis was measured as the percentage of tunnel positive cells G S.D. measured on five random fields (!400). Vessels were measured as the vessel number per field at !100.

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access 408 www.endocrinology-journals.org Endocrine-Related Cancer (2009) 16 401–413

DTX single treatments) and 81% (P!0.001 versus for the comparisons versus Vidaza or cisplatin Vidaza and NS versus cisplatin single treatments). (PZ0.096 for both comparisons). Moreover, 10.0% (2/20) and 30.0% (6/20) of mice (iii) Sequential treatments with either azacitidine treated with the Vidaza plus DTX and Vidaza plus C DTX or azacitidine C cisplatin resulted in a cisplatin combined treatments were tumor-free after significant reduction in tumor blood vessels compared 28 days, although only the second combined treatment with controls (Tables 3 and 4) both in PC3 and 22rv1 was statistically significant versus control (P!0.05) xenografts. This antiangiogenetic effect was further and the statistic analysis showed not significant values supported by quantification of hemoglobin content in

Figure 5 (A) Western blot analysis showing the increment of DNMT1 and DNMT3b expression without modulation of Akt activity after CP treatment in PC3 and 22rv1 tissues. Forty micrograms of proteins were loaded in each lane. (B). Immunohistochemical appearance of PC3 tumors harvested from control and CP treated mice showing increased DNMT1 expression after CP treatment. (C). Microscopic appearance of PC3 and 22rv1 tumors stained with Martius yellow-brilliant crystal scarlet blue technique. We show the presence of large blood vessels (stained in yellow/orange) near to cancer cell nests of untreated 22rv1 tumors. Untreated PC3 tumors show the presence of numerous and small blood vessels scattered in the tumor mass and delineating single tumor cell nests. Vidaza reduces blood vessel number both in 22rv1 and PC3 xenografts. Vessels appear dilated with unstructured capillary bed dispersed in collagen I deposits (azure staining) primarily in 22rv1 tumors. DTX treated 22rv1 tissues appear with large unstructured capillary bed and abundant hemorrhagic areas. The presence of numerous phagocytes (neutrophilic granulocytes and monocytes) dispensed in collagen I deposit suggests that a previous colliquative necrosis was present. In the combined treatment (Vidaza plus DTX), the above-mentioned appearance was amplified with deposition also of fibrin clots (pale pink-staining) evident in 22rv1 tissues and formation of dense collagen I deposits (fibrosis) which envelop tumor cell nests. Single pictures are at !400. (B) Angiogenesis was evaluated measuring hemoglobin presence in tissue extracts or counting CD34/von Willebrand double positive endothelial cells forming blood vessels. Full colour version of this figure available via http://dx.doi.org/10.1677/ERC-08-0130.

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access www.endocrinology-journals.org 409 C Festuccia and G L Gravina et al.: Azacitidine effects in advanced prostate cancer models tumor tissues, used as a marker of the number of et al. 2008, Ross et al. 2008). The resistance of tumor blood vessels. To specifically assess vascular density cells to anticancer agents remains a major cause of within the tumors, sections were stained with Masson failure in the treatment of patients with cancer. The trichromic and BMBS. Blue-stained spiral strands of classic mechanism for the evolution of a multidrug- collagen were detected in peripheral areas, whereas resistant phenotype was believed to involve a single PC3 cells grow to form large cell masses that were molecular mechanism, such as overexpression of surrounded by red-stained blood vessels that often P-glycoprotein. However, it now seems that the seemed dilated. Control xenografts exhibited numer- multidrug-resistant phenotype represents a complex, ous dilated and even sinusoid-looking vessels as multifactor process in which epigenetic mechanisms compared with treated xenografts. The antiangiogenic may play a crucial role. Epigenetic alterations, effects of azacitidine followed by either DTX or including aberrant promoter methylation, are respon- cisplatin were further confirmed by immunohisto- sible for the silencing of tumor suppressor genes and chemical localization of von Willebrand factor that is for increased resistance to chemotherapy. This high- a marker for endothelial cells and the presence of lights a potential role for DNMT as potential molecular microvessels. However, a reduction in microvessel targets in cancer therapy (Mishra et al. 2008). density was only observed in tumors treated with Exploiting gene reactivation by epigenetic acting azacitidine C cisplatin compared with controls or agents in combination with cytotoxic therapies has with azacitidine alone. Results also showed anti- been a strategy that holds much clinical promise (for angiogenetic effects with the combination of azaciti- review, see Lu et al. 2006). Although azacitidine has dine and DTX, including significant reductions in the shown significant promise in the treatment of hemato- vessels analyzed with the presence of fibrin clots and poietic malignancies, and is approved for the treatment red dispersion in the tumor parenchyma, and the of (MDS; Issa et al. 2005, amount of hemoglobin analyzed in tissue extracts Abdulhaq & Rossetti 2007), research into its activity in (Fig. 5A and B). solid tumors has been less impressive and have been (iv) In addition, we found that treatments with similarly disappointing (Schwartsmann et al. 2000, cisplatin were able to increase the levels of DNMT1 Pohlmann et al. 2002, Appleton et al. 2007). However, in PC3 but not in 22rv1 cells. This is in agreement it is important to note that the majority of these studies with the higher effects of combination azacitidine used dose and schedule combinations that are now plus cisplatin in PC3 when compared with those recognized to be suboptimal. In addition, the major observed in 22rv1. In Fig. 5C, we show western blot issue involves either drug uptake or tumor cell analysis performed on PC3 and 22rv1 tissue extracts proliferation rate because the analysis of tumor tissues whereas in Fig. 5D, we show the expression of showed a disappointingly small decrease of methyl- DNMT1 in PC3 xenografts with or without cisplatin ation (3%). This has been shown to be below the treatment. DTX was not able to modify DNMT threshold necessary for resensitization to chemother- expression in PC3 and 22rv1 xenografts (data not apy (Appleton et al. 2007). Our research demonstrates shown) supporting the high effects of combination that tumor levels of DNMT1 and DNMT3b were lower treatment with azacitidine and cisplatin observed in in azacitidine-treated conditions both in vitro and in both PCa models. PC3 and 22rv1 xenograft models. Our in vitro data indicate that DNA methylation activity reaches maximum inhibition of inhibition at 10 days after a Discussion 48 h of azacitidine treatment. This observed inhibition One of the major obstacles in curing locally advanced remained at high levels for a maximum 2 weeks. and metastatic prostate cancer is the development of In parallel, a significant downregulation in DNMT1 resistance to therapy. Historically, chemotherapy has and DNMT3b expression occurred and peaked at had limited utility in treating castration-resistant 5–10 days, and persisted even after 14 days. This locally advanced and metastatic prostate cancer. reduction of DNMT activity was associated with the Nevertheless, DTX is the standard first-line chemo- in vitro and in vivo re-expression of markers of therapy for this disease. The responses to DTX are, differentiation, including AR (Jarrard et al. 1998, however, partial and associated with increased resis- Gravina et al.2008), and is responsible for the tance to apoptosis, thus indicating a clear need for new restoration of responsiveness to hormonal therapy therapies in HRPC patients. A recent report shows that (Izbicka et al. 1999, Sonpavde et al. 2007, Gravina single-agent satraplatin improves PFS as second-line et al. 2008). This reduction in DNMT activity is also chemotherapy for patients with HRPC (Nakabayashi likely responsible for the delay in progression to

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access 410 www.endocrinology-journals.org Endocrine-Related Cancer (2009) 16 401–413 androgen independent disease and for the improvement Funding in survival in the transgenic adenocarcinoma of mouse This research did not receive any specific grant from any prostate mouse model of prostate cancer (Zorn et al. funding agency in the public, commercial or not-for-profit 2007). In the present report, we observed also that sector. DNMT inhibition with azacitidine was able to induce the in vitro and in vivo expression of p16INKA, Bax, References Bak, p21/WAF1, and p27/KIP1. This was also responsible for inhibiting the activation of Akt, the Abdulhaq H & Rossetti JM 2007 The role of azacitidine in expression of cyclin D1, Bcl-2, and Bcl-XL associated the treatment of myelodysplastic syndromes. Expert with the processing of caspases-3 with increased Opinion on Investigational Drugs 16 1967–1975. apoptosis. Nevertheless, we observed an increased Appleton K, Mackay HJ, Judson I, Plumb JA, McCormick C, Strathdee G, Lee C, Barrett S, Reade S, Jadayel D et al. Bcl2 mediated apoptosis by culturing aza-CR treated 2007 Phase I and pharmacodynamic trial of the DNA cells in presence of polyhema which impedes cell methyltransferase inhibitor decitabine and carboplatin adhesion. This condition induces a specialized apopto- in solid tumors. Journal of Clinical Oncology 25 tic mechanism termed anoikis that is normally 4603–4609. suppressed in tumor cells (Diaz-Montero et al. 2006, Armstrong AJ, Garrett-Mayer ES, Yang YC, de Wit R, Glinsky 2006, Kupferman et al. 2007) including PC3 Tannock IF & Eisenberger M 2007 A contemporary cells. Although the reactivation of anoikis by aza-CR prognostic nomogram for men with hormone-refractory can represent an important therapeutic approach to metastatic prostate cancer: a TAX327 study analysis. increase tumor cell death present in the blood stream Clinical Cancer Research 13 6396–6403. Brown R & Strathdee G 2002 Epigenomics and epigenetic and could slow down the metastatic process and therapy of cancer. Trends in Molecular Medicine 8 contribute to antimetastatic treatments of advanced S43–S48. tumors, ad hoc experiments are required. Bastola DR, Pahwa GS, Lin MF & Cheng PW 2002 In addition, the downmodulation of p-Akt and Downregulation of PTEN/MMAC/TEP1 expression members of Bcl-2 family could indeed sensitize PC3 in human prostate cancer cell line DU145 by growth and 22rv1 cells versus DTX and cisplatin increasing stimuli. Molecular and Cellular Biochemistry 236 the antiproliferative and pro-apoptotic effects of single 75–81. drugs both in vitro and in vivo. The in vivo experiments Bertram J, Peacock JW, Fazli L, Mui AL, Chung SW, Cox revealed synergistic effects with DTX and cisplatin ME, Monia B, Gleave ME & Ong CJ 2006 Loss of PTEN is associated with progression to androgen independence. both in PC3 and 22rv1 xenografts. Immunocyto- Prostate 66 895–902. chemistry and enzymatic assays revealed also that Calabro` F & Sternberg CN 2007 Current indications for in vivo DNMT1 expression was significantly increased chemotherapy in prostate cancer patients. European in cisplatin treated 22rv1 and PC3 xenografts, whereas Urology 51 17–26. the induction of DNMT1 levels was lower in DTX Chou TC & Talalay P 1984 Quantitative analysis of dose- treated tumors. This is in agreement with previous effect relationships: the combined effects of multiple reports showing that an increased DNMT1 expression drugs or enzyme inhibitors. Advances in Enzyme is a common event in a multimodality-resistant Regulation 22 27–55. Desireddi NV, Roehl KA, Loeb S, Yu X, Griffin CR, phenotype in tumor cells (Misrha et al. 2008), in Kundu SK, Han M & Catalona WJ 2007 Improved cisplatin and taxane (Segura-Pacheco et al. 2006) stage and grade-specific progression-free survival rates resistance mechanisms. Taken together, our study after radical prostatectomy in the PSA era. Urology 70 indicate the DNMT downmodulation represents an 950–955. useful therapeutic approach for enhancing chemosen- Diaz-Montero CM, Wygant JN & McIntyre BW 2006 sitivity of prostate cancers to DTX and cisplatin and PI3-K/Akt-mediated anoikis resistance of human provide a rationale for clinical trials on combination osteosarcoma cells requires Src activation. European treatments with azacitidine in patients with advanced Journal of Cancer 42 1491–1500. prostate tumors. Glinsky 2006 Genomic models of metastatic cancer: functional analysis of death-from-cancer signature genes reveals aneuploid, anoikis-resistant, metastasis-enabling phenotype with altered cell cycle control and activated Polycomb Group (PcG) protein chromatin silencing Declaration of interest pathway. Cell Cycle 5 1208–1216. There is no conflict of interest that could be perceived as Gravina GL, Festuccia C, Millimaggi D, Dolo V, Tombolini V, prejudicing the impartiality of the research reported. de Vito M, Vicentini C & Bologna M 2008 Chronic

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access www.endocrinology-journals.org 411 C Festuccia and G L Gravina et al.: Azacitidine effects in advanced prostate cancer models

azacitidine treatment results in differentiating effects, Nakabayashi M, Sartor O, Jacobus S, Regan MM, sensitizes against bicalutamide in androgen-independent McKearn D, Ross RW, Kantoff PW, Taplin ME & prostate cancer cells. Prostate 68 793–801. Oh WK 2008 Response to docetaxel/carboplatin-based Hellebrekers DM, Jair KW, Vire´ E, Eguchi S, Hoebers NT, chemotherapy as first- and second-line therapy in Fraga MF, Esteller M, Fuks F, Baylin SB, van Engeland M patients with metastatic hormone-refractory prostate & Griffioen AW 2006 Angiostatic activity of DNA methyl- cancer. BJU International 101 308–312. transferase inhibitors. Molecular Cancer Therapeutics 5 Patel AR, Sandler HM & Pienta KJ 2005 Radiation Therapy 467–475. Oncology Group 0521: a phase III randomized trial of Haas GP, Delongchamps N, Brawley OW, Wang CY & de la androgen suppression and radiation therapy versus androgen Roza G 2008 The worldwide epidemiology of prostate suppression and radiation therapy followed by chemotherapy cancer: perspectives from autopsy studies. Canadian with docetaxel/prednisone for localized, high-risk prostate Journal of Urology 15 3866–3871. cancer. Clinical Genitourinary Cancer 4 212–214. Izbicka E, MacDonald JR, Davidson K, Lawrence RA, Pohlmann P, DiLeone LP, Cancella AI, Caldas AP, Dal Lago L, Gomez L & Von Hoff DD 1999 5,6-dihydro-50- Campos O Jr, Monego E, Rivoire W & Schwartsmann G azacytidine (DHAC) restores androgen responsiveness in 2002 Phase II trial of cisplatin plus decitabine, a new DNA androgen-insensitive prostate cancer cells. Anticancer , in patients withadvanced squamous Research 19 1285–1291. cell carcinoma of the cervix. American Journal of Clinical Issa JP, Gharibyan V, Cortes J, Jelinek J, Morris G, Oncology 25 496–501. Verstovsek S, Talpaz M, Garcia-Manero G & Kantarjian Plumb JA, Strathdee G, Sludden J, Kaye SB & Brown R 2000 HM 2005 Phase II study of low-dose decitabine in patients Reversal of drug resistance in human tumor xenografts with chronic myelogenous leukemia resistant to imatinib by 20-deoxy-5-azacytidine-induced demethylation of mesylate. Journal of Clinical Oncology 23 3948–3956. the hMLH1 gene promoter. Cancer Research 60 Jarrard DF, Kinoshita H, Shi Y, Sandefur C, Hoff D, 6039–6044. Meisner LF, Chang C, Herman JG, Isaacs WB & Nassif N Pulukuri SM & Rao JS 2005 Activation of p53/p21Waf1/ 1998 Methylation of the androgen receptor promoter CpG Cip1 pathway by 5-aza-20-deoxycytidine inhibits cell island is associated with loss of androgen receptor proliferation, induces pro-apoptotic genes and expression in prostate cancer cells. Cancer Research mitogen-activated protein kinases in human prostate 58 5310–5314. cancer cells. International Journal of Oncology 26 Kim KM & Lee YJ 2005 Amiloride augments TRAIL- 863–871. induced apoptotic death by inhibiting phosphorylation of Ross RW, Beer TM, Jacobus S, Bubley GJ, Taplin ME, Ryan kinases and phosphatases associated with the P13K-Akt CW, Huang J & Oh WK 2008 Prostate cancer clinical trials pathway. Oncogene 24 355–366. consortium. A phase 2 study of carboplatin plus docetaxel in Kupferman ME, Patel V, Sriuranpong V, Amornphimoltham men with metastatic hormone-refractory prostate cancer P, Jasser SA, Mandal M, Zhou G, Wang J, Coombes K, who are refractory to docetaxel. Cancer 112 521–526. Multani A et al. 2007 Molecular analysis of anoikis Schwartsmann G, Schunemann H, Gorini CN, Filho AF, resistance in oral cavity squamous cell carcinoma. Oral Garbino C, Sabini G, Muse I, DiLeone L & Mans DR Oncology 43 440–454. 2000 A phase I trial of cisplatin plus decitabine, a new Lu Q, Qiu X, Hu N, Wen H, Su Y & Richardson BC 2006 DNA-hypomethylating agent, in patients with advanced Epigenetics, disease, and therapeutic interventions. solid tumors and a follow-up early phase II evaluation in Ageing Research Reviews 5 449–467. patients with inoperable non-small cell . McCabe MT, Low JA, Daignault S, Imperiale MJ, Wojno KJ Investigational New Drugs 18 83–91. & Day ML 2006 Inhibition of DNA methyltransferase Segura-Pacheco B, Perez-Cardenas E, Taja-Chayeb L, activity prevents tumorigenesis in a mouse model of Chavez-Blanco A, Revilla-Vazquez A, Benitez- prostate cancer. Cancer Research 66 385–392. Bribiesca L & Duenas-Gonza´lez A 2006 Global DNA Mishra MV, Bisht KS, Sun L, Muldoon-Jacobs K, Awwad R, hypermethylation-associated cancer chemotherapy Kaushal A, Nguyen P, Huang L, Pennington JD & resistance and its reversion with the demethylating agent Markovina S 2008 DNMT1 as a molecular target in hydralazine. Journal of Translational Medicine 4 32. a multimodality-resistant phenotype in tumor cells. Sonpavde G, Aparicio A, Guttierez I, Boehm KA, Hutson TE, Molecular Cancer Research 6 243–249. Berry WR, Asmar L & von Hoff DD 2007 Phase II study of Mottet D, Bellahce`ne A, Pirotte S, Waltregny D, Deroanne C, azacitidine to restore responsiveness of prostate cancer to Lamour V, Lidereau R & Castronovo V 2007 Histone hormonal therapy. Clinical Genitourinary Cancer 5 457–459. deacetylase 7 silencing alters endothelial cell migration, a key Svatek RS, Lee JJ, Roehrborn CG, Lippman SM & Lotan Y step in angiogenesis. Circulation Research 101 1237–1246. 2008 Cost-effectiveness of prostate cancer chemopre- Murphy TM, Perry AS & Lawler M 2008 The emergence of vention: a quality of life-years analysis. Cancer 112 DNA methylation as a key modulator of aberrant cell 1058–1065. death in prostate cancer. Endocrine-Related Cancer 15 Thakkar H, Chen X, Tyan F, Gim S, Robinson H, Lee C, 11–25. Pandey SK, Nwokorie C, Onwudiwe N & Srivastava RK

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access 412 www.endocrinology-journals.org Endocrine-Related Cancer (2009) 16 401–413

2001 Pro-survival function of Akt/protein kinase B in Xu J, Langefeld CD, Zheng SL, Gillanders EM, Chang BL, prostate cancer cells. Relationship with TRAIL resist- Isaacs SD, Williams AH, Wiley KE, Dimitrov L, Meyers ance. Journal Biological Chemistry 276 38361–38369. DA et al. 2004 Interaction effect of PTEN and CDKN1B Verhagen PC, van Duijn PW, Hermans KG, Looijenga LH, chromosomal regions on prostate cancer linkage. Human van Gurp RJ, Stoop H, van der Kwast TH & Trapman J Genetics 115 255–262. 2006 The PTEN gene in locally progressive prostate Zorn CS, Wojno KJ, McCabe MT, Kuefer R, Gschwend JE & cancer is preferentially inactivated by bi-allelic gene Day ML 2007 5-aza-20-deoxycytidine delays androgen- deletion. Journal of Pathology 208 699–707. independent disease and improves survival in the Whang YE, Yuan XJ, Liu Y, Majumder S & Lewis TD 2004 transgenic adenocarcinoma of the mouse prostate mouse Regulation of sensitivity to TRAIL by the PTEN tumor model of prostate cancer. Clinical Cancer Research 13 suppressor. Vitamins and Hormone 67 409–426. 2136–2143.

Downloaded from Bioscientifica.com at 09/25/2021 07:04:29AM via free access www.endocrinology-journals.org 413