All the materials for the tissue culture were purchased from EuroClone S.p.A. (Milan, Italy). Antibodies: Bcl-2 (sc-509), Bax (N20) (sc-493), Bad (C20) (sc-943), p-Bad ser112 (sc-7998), Bcl-xL (H2) (sc-8392), TrkA (C20) (sc-20537), p-TrkA (E6) (sc-8058), p75^NTR (H92) (sc-5634), TRADD (A-5) (sc46653), RIP (H207) (sc 7881), TRAF2 (D-3) (sc-136997), DR4 (C20) (sc-6823) and DR5 (N19) (sc-7192) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). DNMT1, DNMAT3a and DNMT3b antibodies were purchased from BioCarta LLC (San Diego, CA, USA). Plasticware was obtained from Nunc (Roskilde, Denmark). Azacitidine (Vidaza^®) was obtained from Celgene Corporation (Summit, NJ, USA). The p38 MAPK inhibitor SB202190 and ibuprofen were purchased from Sigma-Aldrich Italy (Milan, Italy). Ro 08-2750, a selective p75^NTR inhibitor, was purchased from Tocris Bioscience (Bristol, UK).

Two aggressive PCa models (PC3 and 22rv1 cell lines) were obtained from the American Tissue Culture Collection (ATCC; Rockville, MD, USA) and DSMZ (Braunschweig, Germany), respectively, and were grown as recommended.

Cells were seeded at a density of 2×10⁴ cells/ml into 24-well plates. Cells were left to attach and grow in 5% FCS and Dulbecco's modified Eagle's medium (DMEM). Next, the cells were cultured under appropriate experimental conditions. Morphological controls were performed every day with an inverted phase-contrast photomicroscope (Nikon Diaphot, Tokyo, Japan) before cell trypsinization and counting by the NucleoCounter™ NC-100 automated cell counter system (Chemotec, Gydevang, Denmark). The effect on cell proliferation was measured by taking the mean cell number with respect to the controls over time for the different treatment groups.

Adherent cells were trypsinized, pooled with the culture supernatant containing the apoptotic cells already detached from the dish and centrifuged. Cells (1×10⁶) were washed in phosphate-buffered saline (PBS) and fixed for 30 min by the addition of 1 ml of 70% ethanol. Apoptosis and cell cycle analyses were performed using the Tali Cell Cycle kit and the Tali apoptosis kit, Annexin V-Alexa Fluor 488 and propidium iodide-based staining (Life Technologies Italia, Monza, Italy). Apoptosis was further confirmed by cytofluorimetric analysis following the manufacturer's instructions. Stained cells were then measured on a Tali Image-Based Cytometer. Caspase 8 and −9 activities were evaluated using enzymatic assays from GenScript (Piscataway, NJ, USA).

Cells (1×10⁶) were washed with cold PBS and immediately lysed with 100 μl of 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 phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin and 10 μg/ml leupeptin). Cell lysates were electrophoresed in 7% SDS-PAGE, and separated proteins were transferred to a nitrocellulose membrane and probed with the appropriate antibodies as indicated.

Male CD1 nude mice (Charles River, Milan, Italy) were maintained under the guidelines established by our institution (university of L'Aquila, Medical School and Science and Technology School Board Regulations, complying with the Italian government regulation no. 116, January 27, 1992, for the use of laboratory animals). All mice received s.c. flank injections of 1×10⁶ PC3 and 22v1 cells. Tumor growth was assessed by a bi-weekly measurement of tumor diameters with a Vernier calliper (length × width). Tumor weight was calculated as previously indicated (22). Treatments were started when tumor volumes reached ~80 mm³ (day 0) and were stopped after 28 days. Mice were grouped as follows: group 1, 10 mice received intraperitoneal (i.p.) injections of 100 l PBS for a consecutive 7 days; group 2: 10 mice received i.p. injections of 100 l of azacitidine (Aza-CR) (0.8 mg/kg) for a consecutive 7 days. Tumors were fixed in paraformaldehyde for hystochemical and immunohystochemical analyses. Indirect immunoperoxidase staining of tumor xenografts was performed on paraffin-embedded 4-μm tissue sections. Briefly, the sections were incubated with the primary antibodies overnight at 4°C. Next avidin-biotin assays were carried out using the Vectastain Elite kit (Vector Laboratories, Burlingame, CA, USA). Mayer's hematoxylin was used as a nuclear counterstain.

Continuous data are presented as the mean ± standard deviation (SD), and were compared using an umpired Student's t-test.

In the present study, we demonstrated that the antiproliferative and pro-apoptotic effects induced by azacitidine at non-toxic concentrations (ranging between 0.1 and 0.5 μM) were enhanced in the presence of 10 ng/ml NGF. We observed that NGF-induced growth inhibition was significantly higher when compared to data observed following treatment with the demethylating agent alone both in the PC3 (Fig. 1A) and 22rv1 cells (Fig. 1B). The addition of NGF triggered a significant and early apoptosis in both cell lines (Fig. 1C and D). Next, we analyzed the molecular arrangement induced by treatments using western blotting. We observed that expression of the high affinity p75^NTR, but not expression of the low affinity receptor (TrkA) was increased after in vitro administration of azacitidine in a time-dependent manner in the PC3 and 22rv1 cells (Fig. 2A and B). As a positive control we used the androgen receptor (AR) which was widely demonstrated to be induced after azacitidine treatment (19,25). Western blot analyses were confirmed by immunocytochemical analyses (Fig. 2C) in the aggressive 22rv1 cell line. The same tumor cell models were examined for expression levels of downstream components proximal to p75^NTR (TRADD, RIP, DR4, DR5 and TRAF2) in both the PC3 (Fig. 2D) and 22rv1 cells (Fig. 2E). Previously, we demonstrated that an intraperitoneal administration of azacitidine (Vidaza^®, 0.8 mg/kg for 7 consecutive days) in nude male mice subcutaneously inoculated with 22rv1 and PC3 cells was able to reduce tumor growth both in the PC3 and 22rv1 xenograft models (22). In the present study, we demonstrated that reduction in the tumor mass after Aza-CR treatment was associated with p75^NTR re-expression in vivo (Fig. 2F). Cell death was dependent on activation of the caspase 9 pathway which we also verified in vivo by analyzing FasL and TRAIL protein expression in the control and Aza-CR-treated tumors.

It has been previously demonstrated that p38 MAPK activation, associated with ibuprofen treatment, is related to induction of p75^NTR in PCa cells (26). Therefore, considering that the antitumor effects of azacitidine are associated with increased p38 MAPK activity necessary for the induction of cell cycle arrest in the G2 phase (27,28) and apoptosis (29), we verified whether the p38 MAPK inhibitor, SB202190, was able to prevent the induction of p75^NTR using as a positive control, treatment with ibuprofen (1 μM). In agreement with the literature data, we demonstrated that: i) NGF reduced ERK activation, but not p38 MAPK (Fig. 3A), and ii) ibuprofen induced p75^NTR (Fig. 3B) whereas p38 MAPK inhibition was able to reduce p75^NTR expression (Fig. 3B). Similar results were obtained using P38 MAPK siRNA (data not shown) indicating that p75^NTR expression was under p38 MAPK-dependent epigenetic control. Thus, we analyzed the role of p75^NTR in the azacitidine-induced cell death. For this aim we used a blocking antibody for p75^NTR as well as the non-peptide antagonist of NGF, Ro 08-2750 [able to bind to the NGF dimer and inhibit selectively p75^NTR at submicromolar concentrations whereas both p75^NTR and TrkA receptors were blocked at >5 μM (29)]. Both anti-NGF agents significantly increased the cell viability (Fig. 3C) and apoptosis (Fig. 3D) induced by co-treatment with azacitidine and NGF. Conversely the p75^NTR induction by Aza-CR treatment was partial after SB202190 (data not shown). Increased p75^NTR observed after azacitidine treatment was related to increased expression of Bax and PARP cleavage as well as to reduced levels of Bcl2, XIAP and survivin (Fig. 4A) with caspase 8-dependent caspase 3 activation (Fig. 4B and C) in agreement with our in vivo effects (30).