Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

The methyl-CpG-binding protein MECP2 is required for prostate cancer cell growth


The incidence of prostate cancer is increasing in western countries because of population aging. Prostate cancer begins as an androgen-dependent disease, but it can become androgen independent at a later stage or in tumors recurring after an antihormonal treatment. Although many genetic events have been described to be involved in androgen-dependent and/or -independent prostate cancer growth, little is known about the contribution of epigenetic events. Here we have examined the possibility that the methyl-CpG-binding protein MECP2 might play a role in controlling the growth of prostate cancer cells. Inhibition of MECP2 expression by stable short hairpin RNA stopped the growth of both normal and cancer human prostate cells. In addition, ectopic expression of the MECP2 conferred a growth advantage to human prostate cancer cells. More importantly, this expression allowed androgen-dependent cells to grow independently of androgen stimulation and to retain tumorigenic properties in androgen-depleted conditions. Analysis of signaling pathways showed that this effect is independent of androgen receptor signaling. Instead, MECP2 appears to act by maintaining a constant c-myc level during antihormonal treatment. We further show that MECP2-expressing cells possess a functional p53 pathway and are still responsive to chemotherapeutic drugs.


Prostate cancer is a leading cause of death and illness for men in western countries. Prostate glands invariably become hyperplastic with age and in one of every six adult men this benign prostatic hyperplasia will evolve to prostate cancer. Early detection allows successful treatment mainly by localized surgery and radiotherapy. Prostate tumor progression depends on androgen receptor (AR) signaling triggered by its ligand dihydrotestosterone, a metabolic subproduct of androgen. At later stages, antiandrogen treatments are available. They induce considerable regression of prostate tumors, but with time recurrent tumors appear and the outcome is eventually fatal. Such recurrent tumors become unresponsive to antiandrogen therapies and practically refractory to any available treatment. In order to develop new prognostic markers and new therapies, it is important to understand the genetic events involved in prostate cancer cell growth and in progression from the hormone-sensitive to -insensitive state (Feldman and Feldman, 2001; Denmeade and Isaacs, 2002; Nelson et al., 2003).

The common features of human cancers, prostate cancer included, are the loss of function of tumor suppressor genes and activation of proto-oncogenes (Hanahan and Weinberg, 2000). Various genetic alterations (e.g. DNA mutations, amplifications, or deletions) affecting both tumor suppressor genes and proto-oncogenes have been described at length (Abate-Shen and Shen, 2000; Visakorpi, 2003; Futreal et al., 2004; Porkka and Visakorpi, 2004). By contrast, and even though epigenetic alterations can regulate tumor suppressor genes and proto-oncogenes, little is known about the involvement of epigenetic alterations such as DNA methylation or post-translational modifications of histone tails in prostate cancer development (Baylin and Herman, 2000; Jones, 2002; Fojo and Bates, 2003; Lund and van Lohuizen, 2004).

Silencing of tumor suppressor genes by DNA methylation during prostate tumorigenesis is now a well-recognized process (Li et al., 2004). As these chromatin modifications are reversible, pharmacological inhibitors targeting DNA methylation appear as attractive potential anticancer agents (Baylin and Herman, 2000; Goffin and Eisenhauer, 2002; Cheng et al., 2004). Histone deacetylation leads to chromatin compaction and may also be involved in silencing tumor suppressor genes. Chemical inhibitors of histone deacetylase (HDAC) can revert this modification and trigger re-expression of tumor suppressor genes (Yoshida et al., 2001). The most effective approach might be to target both DNA methylation and histone acetylation, as combinations of DNA methylation and histone acetylation inhibitors appear to cause higher-level re-expression of tumor suppressor genes than single drugs (Zhu and Otterson, 2003).

Methyl-CpG-binding proteins (MBDs) are part of the epigenetic machinery. They constitute a link between DNA methylation and histone modification in processes leading to stable repression of gene transcription (Dobosy and Selker, 2001, Fuks et al., 2003). On the one hand, they bind both to methylated DNA (Burgers et al., 2002) and to DNA methyltransferases (Dnmts) (Kimura and Shiota, 2003); on the other hand, they can also bind directly to HDACs (Nan et al., 1998b; Ng et al., 1999). Hence, MBDs can be viewed as a molecular bridge between two key epigenetic processes: DNA methylation and histone modifications. Although they might potentially affect cancer development through the regulation of these two pathways, little is known about their role in cancer.

To examine the possible involvement of the MBD MECP2 in controlling the growth of normal and cancer prostate cells, we have studied the effect of inhibiting MECP2 expression in these cells. For this we have used stable short hairpin RNA (shRNA) to silence MECP2 expression. Our experiments show that this results in growth arrest of both normal and transformed prostatic cells. They further demonstrate that ectopic MECP2 expression can induce androgen-independent growth of prostate cancer cells by maintaining the level of c-myc oncoprotein without affecting the p53 pathway.


MECP2 is required for growth of normal and transformed prostatic cells

To explore the possible links between MECP2 and cell growth, we used shRNA to inhibit MECP2 expression in normal and tumor-derived prostatic cells. To this end, we first constructed the retroviral vector pRS/MECP2 (see Methods) and checked the ability of the shRNA it encodes to inhibit endogenous MECP2 expression. This was acheived by infecting prostate cancer cell lines with either pRS or pRS/MECP2. MECP2 shRNA was found to inhibit synthesis of the endogenous MECP2 protein in all the cell lines examined (Figure 1a–c).

Figure 1

Prostate cell growth requires MECP2. (ac) LNCaP in (a), PC-3 in (b), or DU-145 in (c) cells were infected with the pRS or pRS/MECP2 retroviral vectors. At 2 days after infection, cells were selected with puromycin. Protein extracts were prepared on day 4 for LNCaP or PC-3 cells, or on day 6 for DU-145 cells. Western blotting analysis was performed against MECP2 or actin as a loading control. (d) Inhibition of MECP2 expression blocks the growth of normal prostate cells. Cells were infected with the pRS or pRS/MECP2 retrovirus. They were stained after 15 days with Crystal Violet. (eg) MECP2 RNA interference blocks growth of LNCaP, PC-3, and DU-145 cells. Cells were infected with the pRS or pRS/MECP2 retroviral vector. Cells were split and selected with puromycin. This was followed by colony formation assays (left) or growth curve analysis (right).

Next we examined the effect of MECP2 depletion on the growth of normal human prostatic epithelial cells (Hudson and Masters, 2003). Cells infected with pRS/MECP2 showed a strong decrease in growth ability as compared to pRS-infected control cells (Figure 1d). When the same experiment was performed on three human prostate cancer cell lines (LNCaP, PC-3, and DU-145), MECP2 downregulation was again found, in all three cases, to cause significant growth inhibition, as demonstrated by colony formation assays and growth curve analysis (Figure 1e–g). These growth inhibition effects were also observed when MECP2 was targeted by another shRNA construct (Supplementary Figure 1) strongly supporting that the growth inhibition effect is specifically due to MECP2 downregulation. Our results suggest that MECP2 synthesis is necessary for growth of both normal and tumor-derived prostatic cell lines.

Ectopic MECP2 expression promotes androgen-independent growth

To further analyse the impact of MECP2 on prostate cancer cell growth, we examined the effect of ectopic MECP2 expression on proliferation of LNCaP cells. To this end, we generated a recombinant retroviral vector encoding MECP2 (pWB3/MECP2) and used the empty pWB3 as a control retroviral vector. To check for constitutive synthesis of MECP2 in infected LNCaP cells, we performed Western blotting and immunofluorescence experiments. Ectopic expression of MECP2 was indeed detected in LNaCP/MECP2 cells, where its product was found to localize to the nucleus, as expected (Figure 2a and b).

Figure 2

Ectopic MECP2 expression allows androgen-independent growth of human prostate cancer cells. LNCaP cells were infected with the retrovirus pWB3 or pWB3/MECP2, selected and maintained with blasticidin. (a) Cell extracts were prepared and tested for MECP2 expression by Western blotting. (b) Immunofluorescence analysis was performed to detect MECP2. (c) Growth curve assays. Cells were split and treated every 2 days with or without the AR antagonist. Every week they were counted and seeded into fresh medium. The number of PDs was calculated. Statistics were performed using VisualStat and *means a significant difference with P<0.05. (d) Colony formation assays. Cells were seeded at low density and treated every 2 days with AR inhibitor. After 12 days, cells were fixed and stained with Crystal Violet. Representative dishes are shown. (e) DNA content analysis. Cells were treated as described in (d) and processed for DNA content analysis after 10 days. (f) Soft agar assays. Cells were plated in soft agar medium with or without the AR antagonist. AR antagonist treatment was renewed every 3 days. The number of foci was counted after 3 weeks.

To evaluate the effect of ectopic MECP2 expression on androgen-dependent and -independent growth, we analysed the growth curves obtained for cells cultured in the usual medium in the presence or absence of the AR antagonist. The cells were split and counted every week for 6 weeks, and the number of population doublings (PDs) was calculated. In the absence of the AR antagonist, ectopic MECP2 expression was found to confer a growth advantage: in the course of six passages, LNCaP/MECP2 cells underwent 5.5 more PDs than control cells (Figure 2c). In the presence of the AR antagonist, constitutive MECP2 expression overcame the growth arrest caused by this treatment (Figure 2c): control cells doubled only five times in 42 days of culture, while the MECP2-expressing cells doubled 19 times (Figure 2c). We controlled that MECP2 constitutive expression was sustained during all the experiments by checking its expression at passages 1, 3, and 5 (Supplementary Figure 2).

To confirm the ability of MECP2 to promote androgen-independent growth, we performed colony formation assays under conditions of AR inhibition. While LNCaP/WB3 control cells formed few colonies, MECP2-expressing cells formed multiple colonies (Figure 2d). Analysis of cell cycle distribution during AR inhibition showed 17.5% of MECP2-expressing cells in the S and G2/M phases. The proportion of LNCaP/WB3 control cells in these phases was only 7.5% (Figure 2e).

To see whether ectopic MECP2 expression contributes to maintaining the tumorigenic phenotype of LNCaP cells under conditions of AR inhibition, we performed soft agar assays in the presence and absence of AR inhibitor. As expected, both LNCaP/WB3 and LNCaP/MECP2 cells grew and formed foci in the absence of bicalutamide. MECP2-expressing cells formed more colonies than the LNCaP/WB3 cells (Figure 2f). In the presence of the AR inhibitor, control cells expectedly formed practically no colonies, while MECP2-expressing cells did multiply and form colonies (Figure 2f). Together, these data show that MECP2 synthesis enables LNCaP cells to keep growing and to maintain their tumorigenic properties during antiandrogen treatment.

MECP2 induces androgen-independent growth via the c-myc oncoprotein

Different mechanisms have been proposed to explain the appearance of androgen-independent prostate cancer after an antihormonal treatment. The main mechanism proposed relies on the AR acquiring the ability to sustain its activity independently of ligand binding or of the presence of AR antagonists (Feldman and Feldman, 2001; Chen et al., 2004; Taplin and Balk, 2004). To test the possible involvement of this mechanism in androgen-independent cell growth promoted by MECP2, we analysed AR signaling activity in LNCaP/MECP2 cells during antihormonal treatment. As expected (Yeap et al., 1999), AR antagonist treatment resulted in a decrease in AR protein (Figure 3a). This decrease was also observed in cells constitutively expressing MECP2 (Figure 3a). In addition, antihormone treatment led in both control and MECP2-expressing cells to decreased expression of the prostate-specific antigen (PSA), a well-characterized AR target gene (Wolf et al., 1992). These findings suggest that AR activity is not sustained in MECP2-expressing cells during antiandrogen treatment (Figure 3a). To further examine the state of AR signaling during antihormone treatment in MECP2-expressing cells, we analysed the state of neuroendocrine (NE) differentiation, which increases during AR inhibition (Wright et al., 2003, Hirano et al., 2004). Phenotypically, NE differentiation results in the appearance of neuron-like structures, as observed in control cells (Figure 3b, top right panel). At the molecular level, NE differentiation results in increased synthesis of the neuron-specific enolase (NSE) marker (Figure 3a). In agreement with the results presented above, the extent of NE differentiation during antihormone treatment was similar in MECP2-expressing and control cells, as indicated by both neuron-like structure formation (lower right panel, Figure 3b) and NSE marker accumulation (Figure 3a). Thus, AR activity does not appear to be sustained in MECP2-expressing cells during antihormone treatment.

Figure 3

MECP2 does not act via the AR pathway. LNCaP cells stably infected with a retroviral vector (pWB3/MECP2 or the control vector pWB3) were treated with AR antagonist for 0, 5, or 10 days. (a) Cell extracts were prepared and resolved by SDS–PAGE, transferred, and analysed for MECP2, AR, PSA, and NSE. (b) Cells stably expressing MECP2 or not were cultured with and without an AR antagonist for 5 days. Representative pictures are shown.

We have recently demonstrated that constitutive synthesis of the c-myc oncoprotein is sufficient to induce androgen-independent prostate cancer cell growth (Bernard et al., 2003). Consequently, we studied how antihormone treatment affects the c-myc protein level in MECP2-expressing versus control cells. As described previously, levels of both c-myc and its target ODC decreased during AR inhibition. After 10 days of AR antagonist treatment, both proteins were almost undetectable (Figure 4a). In contrast, cells ectopically expressing MECP2 maintained constant c-myc and ODC levels during antihormonal treatment (Figure 4a). Accordingly, cells with a downregulated MECP2 had lower c-myc protein level in contrast to control cells (Figure 4b). Colony formation assays revealed a similar capacity of MECP2- and c-myc-expressing cells to grow in the presence of AR inhibitor (Figure 4c). As c-myc can be regulated at different levels, we next examined whether MECP2-expressing cells have an increase in c-myc mRNA expression and/or increase in c-myc protein stabilization. Myc mRNA steady-state levels were similar in MECP2-expressing cells when compared to control cells (Figure 4d). In addition, no change in c-myc mRNA stability was observed between MECP2-expressing cells and control cells (Figure 4e). By contrast, c-myc protein was stabilized in MECP2-expressing cells (Figure 4f) and destabilized in MECP2-depleted cells (Figure 4g). These data strongly suggest that maintenance of the level of c-myc protein by constitutively expressing MECP2 is sufficient to induce androgen-independent prostate cancer cell growth.

Figure 4

MECP2 maintains the c-myc level to allow androgen-independent growth. (a) LNCaP cells stably infected with the pWB3 (control) or pWB3/MECP2 retroviral vector were treated with AR antagonist for 0, 5, or 10 days. Cell extracts were prepared and analysed for MECP2, c-myc and ODC. (b) PC-3 cells were infected with pRS or pRS/MEP2 vectors and puromycin selected. After 4 days, cell extracts were prepared and levels of c-myc protein were analysed by Western blotting. (c) Cells stably expressing MECP2 or c-myc or infected with the control vector were cultured for 12 days in the presence of the AR antagonist. Cells were fixed and stained with Crystal Violet. Representative dishes are shown. (d) RNA from control or MECP2-expressing cells was extracted and analysed by RT–PCR for expression of c-myc. Actin is used as a loading control. (e) LNCaP cells with or without MECP2 were treated with 2 μg/ml of actinomycin D during the indicated time. RNA was prepared and analysed by RT–PCR for c-myc expression. (f) Control or MECP-2 LNCaP-expressing cells were treated with 10 μg/ml of cycloheximide during the indicated time. Cell extracts were prepared and analysed for level of c-myc protein. (g) PC-3 cells infected with pRS or pRS/MECP2 vectors were treated with cycloheximide during various times. C-myc protein level was analysed by immunoblotting.

The p53 pathway remains functional in MECP2-expressing cells

As MECP2 is required for human prostate cell growth and can induce androgen-independent growth, we next investigated the response of MECP2-expressing cells to chemotherapeutic drugs. These constitute the major alternative after failure of antihormone therapy (Denmeade and Isaacs, 2002). One key pathway that contributes to determining responsiveness to chemotherapy is the p53 pathway (Kellen, 1994, Valkov and Sullivan, 2003). We therefore examined the status of the p53 pathway in MECP2-expressing cells compared to control cells. Stability of p53 protein after a cycloheximide treatment was similar in LNCaP cells expressing or not expressing MECP2 (Figure 5a). In addition, treatment of control and MECP2-expressing cells with chemotherapeutic drugs, camptothecin or etoposide, resulted in stabilization/activation of the p53 protein and a subsequent increase in the p21 protein level (Figure 5b). Accordingly, treatment with either etoposide or camptothecin resulted in growth arrest of both LNCaP control cells and LNCaP cells ectopically expressing MECP2 (Figure 5c and d). Thus, although MECP2 can confer insensitivity to androgens, it does not seem involved in the resistance to chemotherapy.

Figure 5

The p53 pathway is functional in MECP2-expressing cells. (a) LNCaP/WB3 and LNCaP/MECP2 cells were treated with cycloheximide during the indicated time. Analysis of p53 protein steady-state level was performed by immunoblotting. (b) In all, 500 000 LNCaP/WB3 and LNCaP/MECP2 cells were seeded into 10-cm dishes. Etoposide (20 μ M) or Camptothecin (20 μ M) was added the next day and treatment was continued for 10 h. Cell extracts were then prepared and MECP2, p53, and p21 or actin proteins were detected by immunoblotting. (c) In all, 400 000 LNCaP/WB3 or LNCaP/MECP2 cells were seeded into 10-cm dishes and treated once with 4 μ M etoposide. Plates were stained with Crystal Violet after 7 days. (d) In all, 400 000 control or MECP2-expressing cells were seeded into 10-cm dishes and treated with 2 μ M camptothecin. Plates were stained with Crystal Violet after 9 days.


With the constant increase of the human lifespan in western countries, the incidence of age-dependent cancers such as prostate cancer is rising. It is essential, therefore, both to understand the genetic bases of such cancers and to provide therapeutic options. Despite major advances in knowledge of genetic events involved in prostate cancer development, little is known about the epigenetic control of this disease (Abate-Shen and Shen, 2000). Here we have investigated the possibility that the MBD MECP2, a known link between DNA methylation and histone modifications, might contribute to controlling prostatic cell growth.

We demonstrate that interference against MECP2 expression in normal as well as cancer prostate cells results in growth arrest. MECP2 expression was detected in normal as well as in cancer prostate samples (data not shown). The different cell lines tested differ in their sensitivity to MECP2 depletion: MECP2-depleted DU-145 cells did not appear to die (as determined on the basis of morphological criteria), but only to multiply more slowly. PC-3 cells, on the other hand, responded to MECP2 depletion by massive cell death (data not shown). These differential effects may be due to different genetic backgrounds in these cell lines. Genetic differences are bound to exist between normal cells freshly prepared from normal human prostate tissues and human prostate cancer cells with multiple and various genetic alterations (Carroll et al., 1993, Carlson and Ethier, 2000). MECP2 depletion could also result in growth inhibition of other cell types apart from prostate (see Supplementary Figure 3).

We show that stable ectopic expression of MECP2 enables androgen-dependent cells to grow in an androgen-independent manner. This is not due to sustained AR activity during antihormone treatment, since ectopic MECP2 expression fails to sustain AR target synthesis and to prevent the appearance of an NE phenotype (Wright et al., 2003, Ahlgren et al., 2000). Thus, the growth-maintaining effect of MECP2 does not rely on AR signaling. This is not totally surprising: although sustained AR signaling appears as an important factor in the appearance of androgen-independent cancer (Taplin and Balk, 2004, Feldman and Feldman, 2001), it is known that some androgen-refractory tumors, and particularly ones with a NE phenotype, do not express AR (Shah et al., 2004, Abrahamsson et al., 1989, Cohen et al., 1990, Tetu et al., 1987, Miyoshi et al., 2001). MECP2 might thus be involved in the occurrence of these particular recurrent tumors, or it might synergize with AR signaling to help overcome antigrowth signals caused by antiandrogen treatment.

Next we examined whether the c-myc pathway, whose role in androgen independence we have recently highlighted (Bernard et al., 2003), might play a role in the process through which MECP2 expression promotes androgen-independent growth. Surprisingly, we found that MECP2 expression results in a sustained c-myc level during antihormone treatment, whereas a downregulation of MECP2 results in a lower level of c-myc. Since constitutive expression of either MECP2 or c-myc results in androgen-independent cell growth, and since the effect is of similar intensity in both cases, we hypothesize that MECP2 may act via c-myc. We observed a c-myc protein stabilization in MECP2-expressing cells, which is a well-known mechanism to regulate c-myc activity (Sears, 2004). MECP2 is probably indirectly regulating c-myc protein level as MECP2 is primarily described as a transcriptional repressor (Nan et al., 1998a) and as no interaction was detected between MECP2 and c-myc (data not shown). C-myc stability is modified through multiple pathways (Sears, 2004). It is then possible that MECP2 represses expression of a particular gene in these pathways and thus increases c-myc protein stability.

Functional data linking cancer development with MBDs are scarce. In tumor-prone ApcMin/+ mice, the methyl-CpG binding protein MBD2 is essential to efficient tumorigenesis in the intestine (Sansom et al., 2003). Interestingly, these mice with a mutated Apc accumulate β-catenin (Kongkanuntn et al., 1999), and the Apc/β-catenin pathway also results in c-myc activation (He et al., 1998). The status of c-myc in ApcMin/+MBD2+/+ mice versus ApcMin/+MBD2−/− mice has not been examined, but in the light of our results, c-myc appears as a good candidate mediator of MBD2 effects. It would be interesting to check whether MECP2 could also be involved in tumorigenesis in this in vivo model.

In conclusion, we show that, in addition to genetic alterations, proteins involved in the control of epigenetic modifications, such as MECP2, may play an essential growth-promoting role in both normal prostatic cells and prostate cancer cells, whether the cells are androgen-stimulated or not. Our data support the idea that epigenetic modifications may play major roles in cancer development, independently of or in combination with genetic alterations.

Materials and methods

Cell culture and vectors

LNCaP, PC-3, and DU-145 cells (ATCC) were cultured in RPMI 1640 (GibcoBRL) supplemented with 10% FCS (GibcoBRL) and glutamine (GibcoBRL). U2OS cells (ATCC) were cultured in DMEM (GibcoBRL) supplemented with 10% FCS (GibcoBRL). The packaging 293 GP cells (Clontech) were grown in DMEM supplemented with 10% FCS. Human prostatic pluripotent cells were prepared and cultured as described in Hudson and Masters (2003).

The pHygroMarX/c-myc vector has been described in Wang et al. (1998). Human MECP2 cDNA from pcdMP4 (Kudo et al., 2001) was BamHI–EcoRI digested and inserted into the retroviral vector pWZLblasticidin3 (pWB3) to generate pWB3/MECP2. The target sequence used to silence MECP2 was 5′-IndexTermAGTGGAGTTGATTGCGTAC-3′. This sequence was inserted as a short hairpin into the pRetroSuper (pRS) retroviral vector (OligoEngine) according to the manufacturer's recommendations to yield pRS/MECP2. Infected cells were selected with 200–400 ng/ml puromycin (Sigma) or 500 ng/ml to 1 μg/ml blasticidin (Calbiochem).

Transfection and retrovirus-mediated gene transfer

Transfections were performed with PEI (Euromedex) according to the manufacturer's recommendations. Retrovirus production by 293 GP cells was achieved and infection of target cells performed according to Wang et al. (1998).

Growth curves

LNCaP, PC-3, and DU-145 cells were infected with the retroviral vector pRS or pRS/MECP2. Six-well plates were seeded with 5 × 104 LNCaP cells or 4 × 104 PC-3 or DU-145 cells per well and the cells maintained with 200–400 ng/ml puromycin. Every 4 days (LNCaP cells) or every 3 days (PC-3 and DU-145 cells) the cells were counted.

LNCaP cells were infected with the retroviral vector pWB3 or pWB3/MECP2. We then seeded 3 × 105 LNCaP/WB3 or LNCaP/MECP2 cells into 35-mm dishes. The cells were maintained with 500 ng/ml blasticidin, with or without bicalutamide, to inhibit AR activity. Every week the cells were counted and seeded. The number of PDs was calculated at each passage.

Colony formation assays

In the shRNA experiments, LNCaP, PC-3, and DU-145 cells were seeded and infected with the pRS or pRS/MECP2 vector. After 2 days, puromycin was added to the culture medium (final concentration: 200–400 ng/ml) and the cells allowed to grow for 9–12 additional days before fixation with glutaraldehyde (Sigma) and staining with Crystal Violet. In ectopic expression assays, 3 × 105 LNCaP/WB3, LNCaP/MECP2, or LNCaP/myc cells were seeded into 10-cm dishes and treated every 2 days with bicalutamide for 10–12 days. Then, they were stained with Crystal Violet (Sigma) as described in Bernard et al. (2003).

Cell cycle analysis

For cell cycle analysis, cells were prepared as described for the colony formation assays, then fixed in ice-cold 70% ethanol, washed in PBS, and treated with 10 μg/ml RNaseA for 30 min at 37°C. Propidium iodide (Sigma) was added to the samples (final concentration: 10 μg/ml) prior to analysis of 1 × 104 cells with an Epics Elite Cytometer (Coulter).

Soft agar assays

2 × 104 cells were suspended in 3 ml of 0.35% low-melting-point agarose (GibcoBRL) in RPMI 1640 containing 10% FCS (with 10 μg/ml bicalutamide to inhibit AR when indicated). The suspensions were used to seed six-well plates coated with 0.7% low-melting-point agarose in RPMI 1640 containing 10% FCS or 10% FCS+10 μg/ml bicalutamide. Every 3 days, 1 ml fresh medium containing either 10% FCS or 10% FCS+10 μg/ml bicalutamide was added. The number of macroscopic colonies was scored after 3 weeks.


Immunofluorescence experiments were performed as described in Bernard et al. (2001). The following antibodies were used: anti-MECP2 (RA7) and anti-rabbit TRITC-conjugated (Dako). Staining of nuclei was performed with Hoechst dye (Sigma).

RT–PCR analyses

Cells were homogenized in Tripure (Roche) and total RNA was isolated according to the manufacturer's recommendations. cDNAs were synthesized using Superscript II (Invitrogen). Subsequent PCR reactions were performed with c-myc primers as described in Bernard et al. (2003). The actin primers used have been described (Kasibhatla et al., 1998).

Western blot analyses

Western blotting was carried out as previously described in Bernard et al. (2003). Membranes were incubated with the following primary antibodies: anti-c-myc (C-33, Santa Cruz Biotechnology), anti-MECP2 (RA7), anti-PSA (A0562, Dako), anti-NSE (M0873, Dako), anti-ODC (O1136, Sigma), and anti-AR (sc-7305, Santa Cruz Biotechnology). The corresponding peroxidase-labeled secondary antibody (Amersham) was detected with ECL Western blotting reagents (Perkin-Elmer Life Sciences). Equal loadings were check by red ponceau staining or actin.


  1. Abate-Shen C, Shen MM . (2000). Genes Dev 14: 2410–2434.

  2. Abrahamsson PA, Falkmer S, Falt K, Grimelius L . (1989). Pathol Res Pract 185: 373–380.

  3. Ahlgren G, Pedersen K, Lundberg S, Aus G, Hugosson J, Abrahamsson PA . (2000). Prostate 42: 274–279.

  4. Baylin SB, Herman JG . (2000). Trends Genet 16: 168–174.

  5. Bernard D, Pourtier-Manzanedo A, Gil J, Beach DH . (2003). J Clin Invest 112: 1724–1731.

  6. Bernard D, Quatannens B, Begue A, Vandenbunder B, Abbadie C . (2001). Cancer Res 61: 2656–2664.

  7. Burgers WA, Fuks F, Kouzarides T . (2002). Trends Genet 18: 275–277.

  8. Carlson CA, Ethier SP . (2000). Radiat Res 154: 590–599.

  9. Carroll AG, Voeller HJ, Sugars L, Gelmann EP . (1993). Prostate 23: 123–134.

  10. Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R et al. (2004). NatMed 10: 33–39.

  11. Cheng JC, Yoo CB, Weisenberger DJ, Chuang J, Wozniak C, Liang G et al. (2004). Cancer Cell 6: 151–158.

  12. Cohen RJ, Glezerson G, Haffejee Z, Afrika D . (1990). Br J Urol 66: 405–410.

  13. Denmeade SR, Isaacs JT . (2002). Nat Rev Cancer 2: 389–396.

  14. Dobosy JR, Selker EU . (2001). Cell Mol Life Sci 58: 721–727.

  15. Feldman BJ, Feldman D . (2001). Nat Rev Cancer 1: 34–45.

  16. Fojo T, Bates S . (2003). Oncogene 22: 7512–7523.

  17. Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T . (2003). J Biol Chem 278: 4035–4040.

  18. Futreal PA, Coin L, Marshall M, Down T, Hubbard T, Wooster R et al. (2004). Nat Rev Cancer 4: 177–183.

  19. Goffin J, Eisenhauer E . (2002). Ann Oncol 13: 1699–1716.

  20. Hanahan D, Weinberg RA . (2000). Cell 100: 57–70.

  21. He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Costa LT et al. (1998). Science 281: 1509–1512.

  22. Hirano D, Okada Y, Minei S, Takimoto Y, Nemoto N . (2004). Eur Urol 45: 586–592.

  23. Hudson DL, Masters JR . (2003). Methods Mol Med 81: 59–67.

  24. Jones PA . (2002). Oncogene 21: 5358–5360.

  25. Kasibhatla S, Brunner T, Genestier L, Echeverri F, Mahboubi A, Green DR . (1998). Mol Cell 1: 543–551.

  26. Kellen JA . (1994). Anticancer Res 14: 433–435.

  27. Kimura H, Shiota K . (2003). J Biol Chem 278: 4806–4812.

  28. Kongkanuntn R, Bubb VJ, Sansom OJ, Wyllie AH, Harrison DJ, Clarke AR . (1999). Oncogene 18: 7219–7225.

  29. Kudo S, Nomura Y, Segawa M, Fujita N, Nakao M, Dragich J et al. (2001). Brain Dev 23 (Suppl 1): S165–S173.

  30. Li L-C, Okino ST, Dahiya R . (2004). Biochim Biophys Acta (BBA) Rev Cancer 1704: 87–102.

  31. Lund AH, van Lohuizen M . (2004). Cur Opin Cell Biol 16: 239–246.

  32. Miyoshi Y, Uemura H, Kitami K, Satomi Y, Kubota Y, Hosaka M . (2001). BJU Int 88: 982–983.

  33. Nan X, Cross S, Bird A . (1998a). Novartis Found Symp 214: 6–16; discussion 16–21, 46–50.

  34. Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN et al. (1998b). Nature 393: 386–389.

  35. Nelson WG, De Marzo AM, Isaacs WB . (2003). N Engl J Med 349: 366–381.

  36. Ng HH, Zhang Y, Hendrich B, Johnson CA, Turner BM, Erdjument-Bromage H et al. (1999). Nat Genet 23: 58–61.

  37. Porkka KP, Visakorpi T . (2004). Eur Urol 45: 683–691.

  38. Sansom OJ, Berger J, Bishop SM, Hendrich B, Bird A, Clarke AR . (2003). Nat Genet 34: 145–147.

  39. Sears RC . (2004). Cell Cycle 3: 1133–1137.

  40. Shah RB, Mehra R, Chinnaiyan AM, Shen R, Ghosh D, Zhou M et al. (2004). Cancer Res 64: 9209–9216.

  41. Taplin ME, Balk SP . (2004). J Cell Biochem 91: 483–490.

  42. Tetu B, Ro JY, Ayala AG, Johnson DE, Logothetis CJ, Ordonez NG . (1987). Cancer 59: 1803–1809.

  43. Valkov NI, Sullivan DM . (2003). Drug Resistance Updates 6: 27–39.

  44. Visakorpi T . (2003). Urology 62: 3–10.

  45. Wang J, Xie LY, Allan S, Beach D, Hannon GJ . (1998). Genes Dev 12: 1769–1774.

  46. Wolf DA, Schulz P, Fittler F . (1992). Mol Endocrinol 6: 753–762.

  47. Wright ME, Tsai MJ, Aebersold R . (2003). Mol Endocrinol 17: 1726–1737.

  48. Yeap BB, Krueger RG, Leedman PJ . (1999). Endocrinology 140: 3282–3291.

  49. Yoshida M, Furumai R, Nishiyama M, Komatsu Y, Nishino N, Horinouchi S . (2001). Cancer Chemother Pharmacol 48 (Suppl 1): S20–S26.

  50. Zhu WG, Otterson GA . (2003). Curr Med Chem Anti-Canc Agents 3: 187–199.

Download references


We thank members of Yvan de Launoit's laboratory for helpful discussions. We also thank S Kudo for the pcdMP4 vector and the RA7 MECP2 antibody. This work was carried out with the support of the FNRS (Belgium), the ‘Action de Recherche Concertée de la Communauté Française de Belgique’, ‘Fortis Bank assurances’ (Belgium), the ‘Fédération Belge Contre le Cancer’, the CNRS (France), the ARC (France), the FRM Nord Pas de Calais (France), the ‘Conseil Régional Nord/pas-de-Calais’ (France), and the European Regional Development Fund. JG is supported by Human Frontiers Science Program and Cancer Research UK. DB is supported by a Marie Curie Intra-European Fellowship within the sixth European Community Framework Program. FF is a ‘Chercheur Qualifié du FNRS’.

Author information



Corresponding author

Correspondence to D Bernard.

Additional information

Supplementary Information accompanies the paper on Oncogene website (

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Bernard, D., Gil, J., Dumont, P. et al. The methyl-CpG-binding protein MECP2 is required for prostate cancer cell growth. Oncogene 25, 1358–1366 (2006).

Download citation


  • methyl-CpG-binding protein
  • epigenetic
  • cancer
  • prostate
  • growth

Further reading


Quick links