DNA dependent recruitment of DDX17 and other interacting proteins by the human androgen receptor

https://doi.org/10.1016/j.bbapap.2008.11.001Get rights and content

Abstract

An oligonucleotide-based assay (OBA) was used to identify novel co-factors that can be recruited by the deoxyribonucleic acid (DNA)-bound androgen receptor (AR). Nuclear extracts obtained from LNCaP cells, after incubation with R1881, were incubated with biotinylated oligonucleotides bound to streptavidin coated beads. The oligonucleotides contain 3 copies in tandem of the androgen responsive element ARE1 from the prostate specific antigen (PSA) gene promoter. As control incubation, a scrambled version of the tandem ARE1 was used. Immunoblots of the eluents revealed that the AR was bound to the ARE1 oligonucleotide and to a much lesser extent to the scrambled oligonucleotide. Proteins eluted from the oligonucleotides, were separated on a 5–15% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gradient gel, followed by identification using mass spectrometry. Identified proteins were scored for having one or more of the following known properties: nuclear localization, involved in transcription regulation, involvement in steroid hormone receptor (SHR) function, or specifical involvement in AR function. A total number of 85 nuclear proteins were found in two separate OBAs. Based on peptide counting, we found enrichment of 7 proteins eluted from the ARE1 oligonucleotide, compared to the scrambled oligonucleotide. Taken together with the obtained scores, these proteins are considered putative AR co-factors. One of these proteins, DDX17, is known to be a co-factor for estrogen receptor α (ERα), but has never been associated with AR function. The results indicate that the ARE oligonucleotide-based assay may allow enrichment of new candidate DNA-bound AR interacting proteins.

Introduction

Androgens (testosterone and dihydrotestosterone) are powerful hormones, exerting actions which are essential for male development and functioning [1]. The androgenic steroid hormones exert their functions by activating the androgen receptor (AR) after binding to the ligand binding domain (LBD). Once activated, the AR uses its DNA binding domain (DBD) to bind to specific DNA sequences, the so-called hormone or androgen responsive elements (HRE or ARE), which results in induction or repression of transcription of target genes. However, androgen mediated transcription regulation does not take place without the recruitment of several proteins from a repertoire of co-factors by the AR [2], [3], [4]. These co-factors can be divided into 2 groups of chromatin modifying enzymes. One group is the family of histone modifying enzymes (HME) which influence the DNA–histone interaction by (de)acetylation and/or (de)methylation of core histones [5]. The other group of chromatin modifying enzymes is the ATP-dependent chromatin remodeling complexes (CRC) which can displace or remove the histone complexes from the DNA. This group consists of the SWI/SNF, ISWI/hSNF2h, and Mi-2/NURD complexes [6], [7], [8], [9], [10], [11], [12]. Besides chromatin modifying enzymes three other groups of co-factors can be recruited by the AR. The first group is the Mediator (MED)-complex (TRAP, DRIP, ARC and SMCC) which is essential for transcription activation by facilitating RNA-Pol II recruitment [13], [14], [15]. The second group consists of co-activators, which can enhance transcriptional activation, but do not belong to any of the above-mentioned categories [16], [17]. Finally, a last group represses transcription, the co-repressors. Well known co-repressors in the nuclear receptor field are N-CoR and SMRT [18], [19]. An overview of co-repressors which are able to inhibit the transactivation of AR and other nuclear receptors has been published [20].

Co-factors have been isolated by different isolation methods, such as yeast two-hybrid, mammalian two-hybrid, glutathione S-transferase (GST) pull-down and co-immunoprecipitation (co-IP) assays. The protein interactions in yeast two-hybrid and mammalian two-hybrid assays take place in vivo, in yeast or mammalian cells, respectively. The protein (fragment) of interest is expressed as a fusion protein containing a GAL4–DNA binding domain. If an interaction takes place with a possible partner containing a transactivation domain, a functional transcription factor is formed, which functions as the interaction indicator [21], [22]. Interacting proteins in co-immunoprecipitation assays are obtained by using antibodies against the protein of interest. Glutathione–agarose beads are used to capture GST-tagged proteins. The protein of interest and the interacting proteins are either captured as an in vivo formed complex, or the complexes are formed in vitro when cellular extracts are added to immobilised tagged proteins or antibodies.

In the early nineties, several studies showed the effects of DNA on the conformation of DNA binding transcription regulators [for an overview: [23]]. It was suggested that DNA alone, or together with another DNA binding protein on or near the binding site, has allosteric effects on DNA binding proteins. One study involving ERα and ERβ clearly showed the allosteric effects of different response elements on the conformation of these steroid hormone receptors [24]. Remarkably, these response elements together with the conformational change of the ERα and ERβ resulted in a differential recruitment of co-factors [24]. Recently, it has been shown that different response elements from one or from different AR responsive promoters can recruit a different repertoire of proteins [25], [26]. Furthermore, different AR binding response elements modified the hormone response of the AR in the presence of different co-factors [27]. These effects on the AR are probably caused by the response element dependent conformational changes, which in turn influence the co-factor recruitment and finally the transcription of genes.

The above-mentioned assays for co-factor identification (such as yeast and mammalian two-hybrid, GST pull-down and co-IP) lack the involvement of AR DNA binding during or after the procedure to capture the interacting proteins. These assays may miss interacting proteins normally attached to DNA-bound AR. Therefore, in the present study DNA-bound AR interacting proteins are isolated by using biotinylated oligonucleotides containing ARE1, one of the strongest AR binding AREs of the prostate specific antigen (PSA) gene promoter. In this oligonucleotide based assay (OBA), AR-co-factor complexes from nuclear extracts were reconstituted in vitro and isolated for further identification by mass spectrometry.

Section snippets

Cell culture and nuclear extract preparation

AR expressing LNCaP cells were cultured in RPMI-1640 medium (Invitrogen, Carlsbad, CA, USA), supplemented with 7.5% (v/v) dextran-coated charcoal-treated fetal calf serum (FCS; Hyclone, Logan, UT, USA), 100 IU/ml penicillin, and 100 μg/ml streptomycin (BioWhittaker, Vervier, Belgium). This medium was replaced by phenol-free RPMI-1640 containing similar supplements, 3 days before harvesting. The synthetic androgen R1881 was added to a final concentration of 10 nM, 16 h before harvesting. For

Isolation of DNA-bound AR interacting proteins

Searching for novel AR co-factors, an oligonucleotide based assay (OBA) was used for the isolation and purification of candidate proteins. The oligonucleotides were biotinylated and contained the strong ARE1 of the PSA gene in triplicate in tandem [29]. An oligonucleotide containing the scrambled version of the ARE1 was used as a control. Streptavidin-agarose beads were first incubated with one of the oligonucleotides, followed by incubation with the nuclear extract from LNCaP cells. After

Discussion

To identify possible novel AR co-factors which associate with the AR in a hormone and DNA dependent fashion, an oligonucleotide-based assay (OBA) was set up and applied in this study. The AR together with interacting proteins was isolated via binding to oligonucleotides containing either three ARE1s from the PSA gene promoter in tandem or a scrambled version of the ARE1s. After careful selection of the data obtained from two OBAs, in total 85 proteins were identified, of which 7 proteins were

Acknowledgement

We thank the Nijbakker-Morra Foundation for providing financial support to purchase the SpeedVac equipment.

References (38)

  • N.D. Zegers et al.

    Epitope prediction and confirmation for the human androgen receptor: generation of monoclonal antibodies for multi-assay performance following the synthetic peptide strategy

    Biochim. Biophys. Acta

    (1991)
  • J. Veldscholte et al.

    A mutation in the ligand binding domain of the androgen receptor of human LNCaP cells affects steroid binding characteristics and response to anti-androgens

    Biochem. Biophys. Res. Commun.

    (1990)
  • H. Faus et al.

    The ubiquitin-specific protease USP10 modulates androgen receptor function

    Mol. Cell. Endocrinol.

    (2005)
  • C.A. Quigley et al.

    Androgen receptor defects: historical, clinical, and molecular perspectives

    Endocr. Rev.

    (1995)
  • W.L. Kraus et al.

    Nuclear receptor-dependent transcription with chromatin. Is it all about enzymes?

    Eur. J. Biochem.

    (2002)
  • B. Lemon et al.

    Orchestrated response: a symphony of transcription factors for gene control

    Genes Dev.

    (2000)
  • H.V. Heemers et al.

    Androgen receptor (AR) coregulators: a diversity of functions converging on and regulating the AR transcriptional complex

    Endocr. Rev.

    (2007)
  • H. Kwon et al.

    Nucleosome disruption and enhancement of activator binding by a human SW1/SNF complex

    Nature

    (1994)
  • W. Wang et al.

    Purification and biochemical heterogeneity of the mammalian SWI–SNF complex

    EMBO J.

    (1996)
  • Cited by (7)

    • The DEAD box proteins DDX5 (p68) and DDX17 (p72): Multi-tasking transcriptional regulators

      2013, Biochimica et Biophysica Acta - Gene Regulatory Mechanisms
      Citation Excerpt :

      In this study p72 did not coactivate AR, either alone or in conjunction with p68, and specific p68 siRNA knockdown reduced endogenous PSA mRNA expression. However, p72 has been reported to be recruited to the PSA promoter [28], although it was not specifically shown to coactivate AR. It is possible that p72 will coactivate AR in certain contexts; alternatively it is also possible that the alternative translation isoform p82 (see above) may be important in this case.

    • Chromatin binding by the androgen receptor in prostate cancer

      2012, Molecular and Cellular Endocrinology
      Citation Excerpt :

      In the burgeoning era of systems biology smarter methods to fish out protein–DNA complexes and define the identities of the associated proteins are needed. This is starting to happen but is still very much rooted in established cell-line models rather than in a true clinical context (Wong et al., 2009). Undoubtedly advances are being made in all of these areas and we can look forward to great strides in unraveling the contribution of AR-bound chromatin to prostate cancer in the future.

    • Perturbations of RNA helicases in cancer

      2013, Wiley Interdisciplinary Reviews: RNA
    View all citing articles on Scopus
    View full text