Skip to main content
NCBI home page
Endocrinology logoLink to Endocrinology
. 2009 Oct 9;150(11):4950–4957. doi: 10.1210/en.2009-0358

Modulation of Vitamin D Receptor Activity by the Corepressor Hairless: Differential Effects of Hairless Isoforms

Peter J Malloy 1, Jining Wang 1, Kristin Jensen 1, David Feldman 1
PMCID: PMC2775984  PMID: 19819974

Abstract

The vitamin D receptor (VDR) and its corepressor Hairless (HR) are thought to regulate key steps in the hair cycle because mutations in VDR or HR cause alopecia in humans and mice. Many mammalian cells express two major HR isoforms due to alternative splicing of exon 17. HR isoform-a encodes an 1189-amino acid protein (full-length HR), and isoform-b encodes an 1134-amino acid protein (HRΔ1072-1126). We demonstrated that both HR isoforms are expressed in primary human keratinocytes and in the human keratinocyte cell line HaCaT. In transfected COS-7 cells, the full-length HR repressed VDR-mediated transactivation. In contrast, HRΔ1072-1126 failed to suppress and even stimulated VDR-mediated transactivation. In coimmunoprecipitation, both HR isoforms interacted with the VDR, but only the full-length HR interacted with histone deacetylase 1 (HDAC1). Alanine mutagenesis of two conserved glutamic acids residues (E1100A/E1101A) encoded by exon 17 completely eliminated HR corepressor activity and interactions with HDAC1. When the two HR isoforms were coexpressed in COS-7 cells, the corepressor activity of the full-length HR was not antagonized by the HRΔ1072-1126 isoform. When transfected into HaCaT cells, the full-length HR inhibited endogenous CYP24A1 basal gene expression as well as 1,25-dihydroxyvitamin D3-stimulated CYP24A1 expression. HRΔ1072-1126 failed to suppress basal or 1,25-dihydroxyvitamin D3-stimulated CYP24A1 gene expression. In conclusion, we have demonstrated that both HR isoforms are expressed in keratinocytes and that the HRΔ1072-1126 isoform lacks corepressor activity and is unable to bind HDACs. HRΔ1072-1126 may function as a coactivator in some settings by inhibiting HDAC recruitment to the VDR transcriptional complex.

Hairless (HR) is a VDR corepressor expressed as two isoforms, but the alternatively spliced short isoform is unable to bind histone deacetylases and lacks corepressor activity.

The active form of vitamin D, 1,25-dihydroxyvitamin D3 [1,25(OH)2D3 or calcitriol], regulates calcium homeostasis as well as many other biological processes (1). 1,25(OH)2D3 binds to the vitamin D receptor (VDR) leading to either the activation or the down-regulation of many genes. The VDR is a member of the steroid-thyroid-retinoid receptor superfamily of nuclear transcription factors that regulate gene transcription. Mutations in the VDR gene cause the disease known as hereditary vitamin D-resistant rickets (HVDRR) (2). Patients with HVDRR exhibit early-onset rickets, hypocalcemia, secondary hyperparathyroidism, and elevated 1,25(OH)2D3 levels. Some HVDRR patients also exhibit total body alopecia. Occasionally, HVDRR patients develop dermal cysts that are a phenocopy of atrichia with papular lesions (APL), a disease caused by mutations in the Hairless (Hr) gene (3,4). The Hr gene product HR is located in the nucleus and has been shown to interact with histone deacetylases (HDACs) and function as a corepressor of thyroid hormone receptor (TR) (5,6), retinoic acid receptor-related orphan receptor α (RORα) (7,8), and the VDR (9,10). It has been hypothesized that HR and VDR converge to control a common pathway in hair cycling and epidermal differentiation (3,9). In humans, two major HR isoforms are expressed due to differential splicing of exon 17 (11,12). HR isoform-a encodes the full-length HR, and isoform-b encodes an HR protein with a 55-amino acid deletion, HRΔ1072-1126. Keratinocytes are thought to be the major site for HR-VDR interactions in the regulation of the hair cycle and were shown to express the full-length HR (10). The full-length human and rat HRs were shown to repress VDR-mediated transactivation in transfected cells including keratinocytes (9,10). A previous report using RT-PCR showed that both HR isoforms are coexpressed in many tissues with the exception of kidney and intestine, which exclusively express HR isoform-a, and the skin, which exclusively expresses HR isoform-b (11).

In this report, we show that both HR isoforms are expressed in human keratinocytes. We demonstrate that both the full-length HR and HRΔ1072-1126 interact with the VDR, but only the full-length HR interacts with HDAC1 and represses VDR-mediated gene transactivation. HRΔ1072-1126 may function as a coactivator by preventing HDAC recruitment to the VDR transcriptional complex.

Materials and Methods

Cell culture

Normal human keratinocytes and HaCaT cells were grown in keratinocyte serum-free medium (Invitrogen, Carlsbad, CA). African green monkey kidney cells (COS-7) obtained from the American Type Culture Collection (Rockville, MD) were grown in DMEM containing 10% bovine growth serum (Hyclone Laboratories, Logan, UT). Cells were incubated at 37 C under a 5% CO2 atmosphere.

Plasmids

The human cDNAs for HR isoform-a and HR isoform-b in p3xFlagCMV7.1 (Sigma Chemical Co., St. Louis, MO) were kindly provided by Dr. Axel Hillmer. Mouse HDAC1 cDNA was kindly provided by Dr. Stuart Schreiber. Hemagglutinin (HA)-tagged-HDAC1 was constructed by subcloning the HDAC1 cDNA into the SmaI site of phCMV3 (Genlantis, San Diego, CA).

RT-PCR

RNA was isolated from primary human keratinocytes using RNeasy spin columns (QIAGEN, Valencia, CA) and cDNA prepared using Superscript III cDNA synthesis kit (Invitrogen). PCR-ready first-strand cDNA from human skin and kidney was obtained from Clontech (Mountain View, CA). RT-PCR was performed using Taq DNA polymerase. The primers used to distinguish the HR cDNAs were as described by Cichon et al. (11). The PCR products were cloned and sequenced for verification.

Transactivation assays

COS-7 cells were grown in 12-well tissue culture plates and transfected with the 24-hydroxylase luciferase reporter plasmid (125 ng/well) and VDR (62.5 ng/well) and HR (250 ng/well) cDNA expression vectors using Polyfect transfection reagent (QIAGEN). A control plasmid pRLnull (5 ng/well) was included to control for transfection efficiency. The cells were transfected overnight and then treated with graded concentrations of 1,25(OH)2D3 for 24 h. Cell extracts were prepared in passive lysis buffer and luciferase activity determined using the dual-luciferase assay (Promega, Madison, WI) and a Turner luminometer. Transfections were performed in triplicate, and each experiment was repeated at least three times. The data were analyzed by the Student’s t test, and significant differences were designated as P < 0.05.

Western blot

Immunoblotting was performed as previously described (13). The rabbit anti-HR antibody was from Abcam (Cambridge, MA; ab36682). The HR antibody derived from amino acids 900-1000 was used at a 1:1000 dilution for immunoblotting. Proteins were separated on 3–8% NuPage gels in Tris-acetate buffer (Invitrogen).

Mutagenesis

Site-directed mutagenesis of the full-length human HR cDNA was performed using the QuikChangeII XL site-directed mutagenesis kit (Stratagene, La Jolla, CA) as previously described (13). Clones were sequenced to confirm the presence of the mutation.

Coimmunoprecipitation

COS-7 cells were cotransfected with VDR and HR isoforms using Polyfect as previously described (13). After 48 h, cells were collected and lysed in immunoprecipitation buffer [50 mm Tris-HCl (pH 7.4), 1% Nonidet P-40, 10% glycerol, 150 mm NaCl] containing a protease inhibitor cocktail tablet (Roche, Indianapolis, IN). The samples were centrifuged in a microfuge for 10 min at maximum speed at 4 C. The supernatants were then precleared with protein A/G agarose (Santa Cruz Biotechnology, Santa Cruz, CA). Antibodies were then added and incubated at 4 C overnight with constant rotation. Protein A/G agarose was then added for 4 h. The samples were washed three times with immunoprecipitation buffer and eluted with 2× lithium dodecyl sulfate sample buffer (Invitrogen). Immunoprecipitating proteins were resolved by SDS-PAGE and detected by immunoblotting.

Endogenous gene expression

HaCaT cells were transfected with Flag-vector (p3xFlagCMV7.1) or Flag-HR constructs and pSG5-VDR expression vector using Fugene HD (Roche). Cells were treated with ethanol vehicle or 1 nm 1,25(OH)2D3 for 20 h. RNA was extracted and reverse transcribed using Superscript III reverse transcriptase. CYP24A1 and GAPDH mRNAs were then amplified from the cDNA using SYBR-Green qPCR kit (New England Biolabs, Ipswich, MA) and semiquantified by real-time PCR. Transient transfections were performed in triplicate, and each experiment was repeated at least three times. The data were analyzed by the Student’s t test, and significant differences were designated as P < 0.05.

Results

In humans, two HR isoforms (a and b) have been described due to alternative splicing of exon 17 (11,12). Isoform-a encodes the full-length 1189-amino acid HR protein, whereas isoform-b encodes an 1134-amino acid HR protein due to the deletion of 55 amino acids caused by alternate splicing of exon 17 (HRΔ1072-1126) (Fig. 1). The HR protein contains a single putative zinc finger structure with a conserved cysteine motif (C2HC4) that has structural homology to the GATA family of transcription factors and to the rat testis-specific gene A (TSGA) (14). A bipartite nuclear localization signal (NLS) identified at the N terminus of the mouse HR is conserved in the human HR (15). HR also contains two conserved LxxLL motifs (where L is leucine and x is any amino acid) that are important in RORα interaction (ROR interacting domain 1 or RID1 and ROR interacting domain 2 or RID2) (7,8) and two ΦxxΦΦ (where Φ is any hydrophobic amino acid) motifs that are important for interaction with TR (TR interacting domain 1 or TID1 and TR interacting domain 2 or TID2) (16). Binding of the VDR to the rat HR has been shown to involve RID2 and TID1 motifs (9). The C terminus of HR is highly conserved between the human, rat, and mouse proteins. This region also exhibits homology to JmjC domain-containing proteins (17,18). In humans, the 55 amino acids encoded by exon 17 of the HR gene are contained within the putative JmjC domain (Fig. 1A).

Figure 1.

Figure 1

Open in a new tab

A, Schematic illustration of the human HR isoforms. The location of the two ROR interacting domains (RID1 and RID2) containing the LxxLL motifs that are important in RORα interactions and the two TR interacting domains (TID1 and TID2) containing the ΦxxΦΦ motifs that are important in TR interactions are shown. The 55-amino acid sequence (amino acids 1072-1126) encoded by exon 17 of the HR gene is depicted as a hatched box in HR isoform-a. Also featured are the nuclear localization signal (NLS), the DNA binding motif (C2HC4), and the JmjC domain. B, Both HR isoforms are expressed in human keratinocytes, skin, and kidney. HR isoforms were amplified by RT-PCR. The individual HR cDNA expression vectors (+17 and −17) were amplified to serve as controls. Left panel, lane 1, standard; lane 2, NHEK; lane 3, HR isoform-a (+17) control; lane 4, HR isoform-b (-17) control. Right panel, lane 5, PCR-ready human skin cDNA from Clontech (Sk); lane 6, PCR-ready human kidney cDNA from Clontech (Kid), lane 7, HR isoform-a (+17) control; lane 8, HR isoform-b (−17) control. C, Immunoblot of HR isoforms (indicated by arrows) in the human keratinocyte cell line HaCaT and NHEK.

Cichon et al. (11) showed that transcripts from both HR isoforms (a and b) are coexpressed in many tissues. Exceptions were kidney and intestine, which expressed only isoform-a (the full-length HR), and skin, which expressed only isoform-b (HRΔ1072-1126) (11). As shown in Fig. 1B, when we examined the expression of the HR isoforms in cDNA prepared from human skin and kidney, we found that both HR isoforms were coexpressed in each tissue. In normal human epidermal keratinocytes, the cell thought to be critical for VDR-HR regulation of gene activity during the hair cycle (19,20), three PCR products were amplified by RT-PCR (Fig. 1B, lane 2). To determine the authenticity of the PCR products, we cloned and sequenced the three PCR products. The longer PCR product encoded the full-length HR sequence and the shortest PCR product encoded the HRΔexon 17 sequence. The third PCR product contained a sequence from the intron preceding exon 16, part of exon 16 and all of exon 19.

A commercial HR antibody that recognizes an epitope between amino acids 900-1000 was used to determine whether the two HR proteins were expressed in keratinocytes. As shown in Fig. 1C, the antibody recognized both the full-length HR protein (lanes 2 and 4) and the HRΔ1072-1126 proteins (lanes 3 and 5) that were expressed individually in COS-7 cells. More importantly, we showed that both the full-length HR protein and the HRΔ1072-1126 protein were coexpressed in the human keratinocyte cell line HaCaT (lane 1) and in normal human epidermal keratinocytes (NHEK) (lane 6). Interestingly, the antibody also detected a very large molecular weight protein (>250 kDa) in both HaCaT cells and NHEK. Whether this represents a modified HR remains to be determined.

We next investigated the functional activities of the two human HR isoforms as corepressors of VDR. We first examined the corepressor activities of the individual HR isoforms on VDR-mediated gene transactivation. As shown in Fig. 2A, in the absence of HR, 1,25(OH)2D3 induced a dose-dependent increase in VDR-mediated gene transactivation of the 24-hydroxylase gene promoter. On the other hand, when the human full-length HR was coexpressed, VDR-mediated gene transactivation was substantially repressed up to 1 nm 1,25(OH)2D3. When higher concentrations of 1,25(OH)2D3 were added, the transactivation activity gradually increased, eventually reaching a maximal activity observed in the absence of HR. In contrast, the HRΔ1072-1126 isoform stimulated VDR-mediated gene transactivation at all concentrations of 1,25(OH)2D3 tested (Fig. 2A). Furthermore, in the absence of 1,25(OH)2D3, VDR basal activity was inhibited by the full-length HR and stimulated by HRΔ1072-1126 (Fig. 2B). Immunoblotting showed that both isoforms were expressed at similar levels in COS-7 cells (Fig. 2C).

Figure 2.

Figure 2

Open in a new tab

Corepressor activities of HR and HRΔ1072-1126 isoforms. A, COS-7 cells were transfected with pSG5 and p3XFLAG-CMV-7.1 vectors (control), pSG5-VDR and p3XFLAG-CMV-7.1 expression vectors (VDR), pSG5-VDR and p3XFLAG-HR-CMV-7.1 (HR), pSG5-VDR and p3XFLAG-HRΔ1072-1126-CMV-7.1 (HRΔ17), and the 24-hydroxylase promoter luciferase reporter. Cells were treated with graded concentrations of 1,25(OH)2D3 for 24 h and then assayed for luciferase activity. VDR transactivation was repressed by the full-length HR at low concentrations of 1,25(OH)2D3. The repression was relieved when higher concentrations of ligand were added. VDR transactivation was not repressed by the HRΔ1072-1126 isoform. B, Basal activity of the 24-hydroxylase promoter in the presence of VDR is inhibited by the full-length HR but stimulated by HRΔ1072-1126. C, Immunoblot of full-length HR (HR) and HRΔ1072-1126 (HRΔ17) show that both proteins were expressed at the same level.

Because HRΔ1072-1126 had no corepressor activity, we examined whether the deletion of the 55 amino acids affected interactions with VDR in intact cells. Interaction of the VDR with the rat HR has previously been shown to involve a combination of the RID2 and TID1 interacting domains in HR. Because the RIDs and TIDs are present in HRΔ1072-1126, we expected that its interaction with VDR would be unaffected by the deletion of the 55 amino acids caused by splicing out of exon 17. COS-7 cells were cotransfected with VDR and Flag-tagged HR isoform-specific expression vectors. The Flag-tagged HRs were then immunoprecipitated with Flag antibodies and coimmunoprecipitating VDR detected by immunoblot. In Fig. 3A, we show that the VDR was coimmunoprecipitated with both the full-length HR and with the HRΔ1072-1126 isoform. The results demonstrate that deletion of amino acids 1072-1126 does not alter the ability of HRΔ1072-1126 to interact with the VDR in intact cells.

Figure 3.

Figure 3

Open in a new tab

Interactions of the HR isoforms with VDR and HDAC1. A, The HR isoforms interact with VDR. VDR was coexpressed with Flag-tagged HRs in COS-7 cells. Proteins were immunoprecipitated with anti-Flag antibody (IP:Flag) and detected by immunoblot with anti-VDR antibody (WB:VDR). B, Interactions of HDAC1 with the HR isoforms. HA-tagged HDAC1 was coexpressed with Flag-tagged HRs in COS-7 cells. Proteins were immunoprecipitated with anti-Flag antibody (IP:Flag) and detected by immunoblot with anti-HA antibody (WB:HA). In, Input; IP, immunoprecipitation; NS, nonspecific.

The rat HR has been shown to interact with HDACs (HDACs 1, 3, and 5) in intact cells and in vitro (6). We have also shown that naturally occurring mutations in the human HR that are the cause of APL fail to interact with HDAC1 (13). However, no naturally occurring mutations have been found in the 55-amino acid region. To examine whether the HRΔ1072-1126 isoform interacted with HDACs, we coexpressed HA-tagged HDAC1 and the Flag-tagged HR isoforms in COS-7 cells. Flag-tagged HRs were immunoprecipitated using Flag antibodies and coimmunoprecipitating HA-HDAC1 determined by immunoblot using HA antibodies. In Fig. 3B, we show that the full-length HR coimmunoprecipitated HA-HDAC1. In contrast, HRΔ1072-1126 failed to coimmunoprecipitate HA-HDAC1. The data demonstrate that the 55-amino acid region encoded by exon 17 is required for interaction with HDACs.

We next determined whether the loss of HDAC1 binding by HRΔ1072-1126 was due to a major structural change in the protein caused by the deletion of the 55 amino acids or whether amino acids 1072-1126 were directly involved in HDAC interactions. To resolve this question, we mutated two highly conserved glutamic acid residues to alanine (E1100A/E1101A) in the 55-amino acid region within the full-length HR and assessed its affect on corepressor activity and its ability to interact with VDR and HDAC1. As shown in Fig. 4A, the HR E1100A/E1101A double mutant interacted with the VDR in intact cells as expected; however, in transactivation assays, the double mutant lacked corepressor activity and stimulated VDR transactivation (Fig. 4A). In addition, in coimmunoprecipitation assays, the HR E1100A/E1101A double mutant failed to coimmunoprecipitate HA-HDAC1 (Fig. 4B). These data demonstrate that amino acids E1100 and E1101 are required for HR corepressor activity and are important in interacting with HDAC1. The data also suggest that the 55-amino acid region of HR encoded by exon 17 is critical for its corepressor activity by recruiting HDACs.

Figure 4.

Figure 4

Open in a new tab

Mutations in 55-amino acid region in the full-length HR disrupt HDAC1 binding. A, COS-7 cells were transfected with pSG5 and p3XFLAG-CMV-7.1 vectors (control), pSG5-VDR and p3XFLAG–CMV-7.1 expression vectors (VDR), pSG5-VDR and p3XFLAG-HRE1100A/E1101A-CMV-7.1 (HR2E2A), and the 24-hydroxylase promoter luciferase reporter. Cells were treated with graded concentrations of 1,25(OH)2D3 for 24 h and then assayed for luciferase activity. The E1100A/E1101A double mutation in the HR cDNA abolishes corepressor activity and disrupts interaction with HDAC1. Inset is immunoblot of HR E1100A/E1101A protein (HR2E2A). B, HA-tagged HDAC1 was coexpressed with Flag-tagged full-length HR or HR E1100A/E1101A in COS-7 cells. Proteins were immunoprecipitated with anti-Flag antibody (IP:Flag) and detected by immunoblot with anti-HA antibody (WB:HA). In, Input; IP, immunoprecipitation; RLU, relative light units.

In addition to keratinocytes, many other tissues have been shown to express both HR isoforms (11). We next determined whether overexpression of HRΔ1072-1126 affected the corepressor activity of the full-length HR. Figure 5A shows the VDR transactivation activity when treated with 0.1 nm 1,25(OH)2D3 in the absence of HR. Coexpression of the full-length HR reduced the transactivation activity by about 60%. Coexpression of increasing amounts of HRΔ1072-1126 had no effect on the corepressor activity of full-length HR (Fig. 5, A and B). These data suggest that the HRΔ1072-1126 isoform does not function as an antagonist of the full-length HR. An alternative interpretation of the findings is that the coactivator activity of the HRΔ1072-1126 isoform to enhance VDR transactivation was inhibited by the full-length HR.

Figure 5.

Figure 5

Open in a new tab

The corepressor activity of the full-length HR is not antagonized by the HRΔ1072-1126 isoform. A, COS-7 cells were transfected with VDR and a constant amount of the full-length HR expression vector and graded concentrations of the HRΔ1072-1126 isoform expression vector and the 24-hydroxylase promoter luciferase reporter. Cells were treated with 0.1 nm 1,25(OH)2D3 for 24 h and then assayed for luciferase activity. B, Immunoblot of HRs and VDR.

Finally, we examined whether the HR isoforms exhibited differential activity on endogenous CYP24A1 gene expression in HaCaT cells. HaCaT cells were transfected with VDR and the HR isoforms. As shown in Fig. 6A, in the absence of 1,25(OH)2D3, the basal CYP24A1 gene transcription was inhibited by about 90% when the cells were cotransfected with VDR and the full-length HR when compared with cells transfected with VDR and empty vector control. When treated with 1 nm 1,25(OH)2D3, the cells transfected with VDR alone exhibited an approximately 58-fold increase in CYP24A1 mRNA (Fig. 6B). Cells transfected with VDR and the full-length HR exhibited an approximately 20-fold increase in CYP24A1 mRNA, an inhibition of about 65% compared with VDR alone (Fig. 6B). On the other hand, cotransfection of VDR and HRΔ1072-1126 failed to suppress both basal and 1,25(OH)2D3-stimulated CYP24A1 mRNA expression (Fig. 6, A and B). It also did not stimulate CYP24A1 expression as was seen in the COS-7 cells transfected with a CYP24A1 promoter luciferase reporter construct (Fig. 2).

Figure 6.

Figure 6

Open in a new tab

HR isoforms and endogenous CYP24A1 gene expression in HaCaT cells. HaCaT cells were cotransfected with VDR and Flag vector or HR expression vectors. Cells were treated with vehicle (A) or 1 nm 1,25(OH)2D3 (B) for 20 h. CYP24A1 and GAPDH mRNA levels were determined by real-time RT-PCR. Values represent mean ± sd of triplicate transfections. Asterisks indicate significantly different from the VDR-Flag vector control: *, P < 0.05.

Discussion

Mutations in either VDR or HR cause alopecia in humans and mice. The discovery that VDR and HR interact with each other and repress gene transactivation has led to the hypothesis that the VDR and HR converge on a common pathway to regulate the hair cycle by a mechanism involving gene repression (9,20). Keratinocytes are thought to play an important role in hair follicle development because targeted expression of the VDR in keratinocytes in VDR knockout mice with alopecia restores hair growth (19). In fact, targeted expression of a mutant VDR that has impaired ligand binding or a transactivation domain mutant VDR that fails to activate gene transcription in VDR knockout mice restored hair growth (20). These data suggest that regulation of the hair cycle by VDR is independent of ligand and gene transactivation. In humans, two HR isoforms (a and b) have been described due to alternative splicing of exon 17 (11). In this report, we examined the properties of the two HR isoforms as corepressors of VDR.

It was previously reported that a number of human tissues express both HR isoforms, the exceptions being that intestine and kidney expressed only isoform-a and skin expressed only isoform-b (11). HR isoform-a encodes the full-length HR, and isoform-b encodes an HR protein that is 55 amino acids shorter (HRΔ1072-1126) as a result of alternative splicing that deletes exon 17 (11,12). In contrast to the original report, we have demonstrated by RT-PCR that both HR isoforms are coexpressed in human skin and keratinocytes as well as kidney (Fig. 1B). Furthermore, we demonstrated by immunoblotting that both HR proteins are expressed in normal human epidermal keratinocytes (Fig. 1C). Xie et al. (10), using an HR-specific antibody that was generated using a peptide sequence from the C-terminal region of the human HR, showed by immunoblot that normal human keratinocytes express an approximately 130-kDa HR as well as several shorter immunoreactive fragments of 110, 100, and 60 kDa. The approximately 130-kDa migrating band most likely represented the full-length HR (calculated molecular mass 127,494 Da) because the antibody used was targeted to the C terminus of the full-length HR protein. Whether the keratinocytes expressed the shorter HRΔ1072-1126 protein (calculated molecular mass 121,857 Da) could not be ascertained from the blot because it is not known whether the HR antibody used by Xie et al. (10) recognized the shorter HRΔ1072-1126 protein.

Importantly, we demonstrated that the HRΔ1072-1126 isoform lacks corepressor activity. In COS-7 cells, the HRΔ1072-1126 isoform increased both VDR basal and 1,25(OH)2D3-stimulated transactivation. We also showed that HRΔ1072-1126 interacted with the VDR. These latter results were not unexpected because the 55-amino acid deletion did not affect the RID2 or TID1 binding sites for VDR.

HDACs have been implicated in both gene repression and gene activation (21). The full-length HR has been shown to interact with HDACs both in vivo and in vitro (5,6,13,22). We have also shown that the human full-length HR can bind HDAC1 and that mutations in HR that cause APL inhibit HDAC binding as well as abolish suppression of transactivation (13). HRΔ1072-1126 also failed to interact with HDAC1. We also showed that mutation of two highly conserved glutamic acid resides within the 55-amino acid sequence encoded by exon 17 in the full-length HR inhibited HDAC binding and abolished its corepressor activity. These data indicate that the 55-amino acid fragment is critical for HR corepressor activity and interactions with HDACs.

We demonstrated that overexpression of the HRΔ1072-1126 protein does not inhibit the corepressor activity of the full-length HR (Fig. 5). We also showed that the full-length HR repressed both endogenous basal and 1,25(OH)2D3-stimulated CYP24A1 gene expression in HaCaT cells. In the intact cell system, HRΔ1072-1126 again failed to suppress either basal or 1,25(OH)2D3-stimulated CYP24A1 endogenous gene transactivation.

HRΔ1072-1126 may function as a VDR coactivator in some settings because coexpression of HRΔ1072-1126 with VDR enhanced VDR transactivation in COS-7 cells. However, a stimulatory effect of HRΔ1072-1126 on endogenous CYP24A1 gene transcription was not observed in transfected HaCaT cells.

Although it has been demonstrated that HR interacts with HDACs (5,6,13,22), it has not been definitively shown that HDAC recruitment and activity is essential for HR corepressor activity. Whether the inability of HRΔ1072-1126 to interact with HDACs is the basis for the loss of its corepressor activity is unclear. The loss of interaction with HDACs may be secondary to the lack of HRΔ1072-1126 corepressor activity whatever that activity may be. Alternatively, we have shown that HRΔ1072-1126 stimulated VDR transactivation and may function as a coactivator by preventing HDAC recruitment to the VDR transcriptional complex.

Interestingly, the 55 amino acids encoded by exon 17 are located in the region with homology to the JmjC domain (Fig. 1). Proteins containing JmjC domains have recently been shown to exhibit histone demethylase activity, and it has been hypothesized that the JmjC domain encodes this activity (23). Moreover, HR exhibits significant homology to JHDM2A (jumonji-containing histone demethylase 2A also known as testis-specific gene A or TSGA) (24). JHDM2A has been shown to exhibit histone 3 lysine 9 (H3K9) demethylase activity and function as a coactivator of the androgen receptor (25). Demethylation of histones by histone demethylases has been linked to both gene repression and gene activation (26). However, the alignment of the potential iron- and α-ketoglutarate binding sites are different in HR, raising the possibility that HR is devoid of demethylase activity (27). Whether HR exhibits demethylase activity is currently being investigated.

In addition to its corepressor activity, HR has also been shown to stabilize RORα by inhibiting its degradation by the ubiquitin (Ub)-proteasome pathway (8). Stabilization of RORα by HR has been speculated to provide a mechanism for prolonging the action of the HR-RORα complex on promoter regions of repressed genes (8). The VDR also has been shown to be degraded by the Ub-proteasome pathway (28,29,30). Whether HR stabilizes VDR by inhibition of Ub-proteasome-mediated degradation is currently being investigated.

In conclusion, in this report, we showed that the two HR isoforms are expressed in skin and keratinocytes, the primary cell involved in the regulation of the hair cycle. Both HR isoforms interact with VDR, but only the full-length HR acts as a corepressor. HRΔ1072-1126 may function as a coactivator by preventing HDAC recruitment to the VDR transcriptional complex.

Acknowledgments

We thank Dr. Axel Hillmer for the HR expression vectors and Dr. Stuart Schreiber for the mouse HDAC1 cDNA.

Footnotes

This work was supported by a grant from the National Institutes of Health (DK42482) to D.F.

Disclosure Summary: The authors of this manuscript have nothing to disclose.

First Published Online October 9, 2009

Abbreviations: APL, Atrichia with papular lesions; HA, hemagglutinin; HDAC1, histone deacetylase 1; HR, hairless protein; HVDRR, hereditary vitamin D-resistant rickets; NHEK, normal human epidermal keratinocytes; 1,25(OH)2D3, 1,25-dihydroxyvitamin D3; RID, ROR interacting domain; RORα, retinoic acid receptor-related orphan receptor α; TID, TR interacting domain; TR, thyroid hormone receptor; Ub, ubiquitin; VDR, vitamin D receptor.

References

  1. Feldman D, Pike JW, Glorieux FH, eds. 2005 Vitamin D. 2nd ed. San Diego: Elsevier Academic Press [Google Scholar]
  2. Malloy PJ, Pike JW, Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D-resistant rickets. Endocr Rev 20:156–188 [DOI] [PubMed] [Google Scholar]
  3. Miller J, Djabali K, Chen T, Liu Y, Ioffreda M, Lyle S, Christiano AM, Holick M, Cotsarelis G 2001 Atrichia caused by mutations in the vitamin D receptor gene is a phenocopy of generalized atrichia caused by mutations in the hairless gene. J Invest Dermatol 117:612–617 [DOI] [PubMed] [Google Scholar]
  4. Bergman R, Schein-Goldshmid R, Hochberg Z, Ben-Izhak O, Sprecher E 2005 The alopecias associated with vitamin D-dependent rickets type IIA and with hairless gene mutations: a comparative clinical, histologic, and immunohistochemical study. Arch Dermatol 141:343–351 [DOI] [PubMed] [Google Scholar]
  5. Potter GB, Beaudoin 3rd GM, DeRenzo CL, Zarach JM, Chen SH, Thompson CC 2001 The hairless gene mutated in congenital hair loss disorders encodes a novel nuclear receptor corepressor. Genes Dev 15:2687–2701 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Potter GB, Zarach JM, Sisk JM, Thompson CC 2002 The thyroid hormone-regulated corepressor hairless associates with histone deacetylases in neonatal rat brain. Mol Endocrinol 16:2547–2560 [DOI] [PubMed] [Google Scholar]
  7. Moraitis AN, Giguère V, Thompson CC 2002 Novel mechanism of nuclear receptor corepressor interaction dictated by activation function 2 helix determinants. Mol Cell Biol 22:6831–6841 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Moraitis AN, Giguère V 2003 The co-repressor hairless protects RORα orphan nuclear receptor from proteasome-mediated degradation. J Biol Chem 278:52511–52518 [DOI] [PubMed] [Google Scholar]
  9. Hsieh JC, Sisk JM, Jurutka PW, Haussler CA, Slater SA, Haussler MR, Thompson CC 2003 Physical and functional interaction between the vitamin D receptor and hairless corepressor, two proteins required for hair cycling. J Biol Chem 278:38665–38674 [DOI] [PubMed] [Google Scholar]
  10. Xie Z, Chang S, Oda Y, Bikle DD 2006 Hairless suppresses vitamin D receptor transactivation in human keratinocytes. Endocrinology 147:314–323 [DOI] [PubMed] [Google Scholar]
  11. Cichon S, Anker M, Vogt IR, Rohleder H, Pützstück M, Hillmer A, Farooq SA, Al-Dhafri KS, Ahmad M, Haque S, Rietschel M, Propping P, Kruse R, Nöthen MM 1998 Cloning, genomic organization, alternative transcripts and mutational analysis of the gene responsible for autosomal recessive universal congenital alopecia. Hum Mol Genet 7:1671–1679 [DOI] [PubMed] [Google Scholar]
  12. Ahmad W, Zlotogorski A, Panteleyev AA, Lam H, Ahmad M, ul Haque MF, Abdallah HM, Dragan L, Christiano AM 1999 Genomic organization of the human hairless gene (HR) and identification of a mutation underlying congenital atrichia in an Arab Palestinian family. Genomics 56:141–148 [DOI] [PubMed] [Google Scholar]
  13. Wang J, Malloy PJ, Feldman D 2007 Interactions of the vitamin D receptor with the corepressor hairless: analysis of hairless mutants in atrichia with papular lesions. J Biol Chem 282:25231–25239 [DOI] [PubMed] [Google Scholar]
  14. Cachon-Gonzalez MB, Fenner S, Coffin JM, Moran C, Best S, Stoye JP 1994 Structure and expression of the hairless gene of mice. Proc Natl Acad Sci USA 91:7717–7721 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Djabali K, Aita VM, Christiano AM 2001 Hairless is translocated to the nucleus via a novel bipartite nuclear localization signal and is associated with the nuclear matrix. J Cell Sci 114:367–376 [DOI] [PubMed] [Google Scholar]
  16. Thompson CC, Bottcher MC 1997 The product of a thyroid hormone-responsive gene interacts with thyroid hormone receptors. Proc Natl Acad Sci USA 94:8527–8532 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Clissold PM, Ponting CP 2001 JmjC: cupin metalloenzyme-like domains in jumonji, hairless and phospholipase A2β. Trends Biochem Sci 26:7–9 [DOI] [PubMed] [Google Scholar]
  18. Trewick SC, McLaughlin PJ, Allshire RC 2005 Methylation: lost in hydroxylation? EMBO Rep 6:315–320 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Chen CH, Sakai Y, Demay MB 2001 Targeting expression of the human vitamin D receptor to the keratinocytes of vitamin D receptor null mice prevents alopecia. Endocrinology 142:5386–5389 [DOI] [PubMed] [Google Scholar]
  20. Skorija K, Cox M, Sisk JM, Dowd DR, MacDonald PN, Thompson CC, Demay MB 2005 Ligand-independent actions of the vitamin D receptor maintain hair follicle homeostasis. Mol Endocrinol 19:855–862 [DOI] [PubMed] [Google Scholar]
  21. Nusinzon I, Horvath CM 2005 Histone deacetylases as transcriptional activators? Role reversal in inducible gene regulation. Sci STKE 2005:re11 [DOI] [PubMed] [Google Scholar]
  22. Djabali K, Christiano AM 2004 Hairless contains a novel nuclear matrix targeting signal and associates with histone deacetylase 3 in nuclear speckles. Differentiation 72:410–418 [DOI] [PubMed] [Google Scholar]
  23. Tsukada Y, Fang J, Erdjument-Bromage H, Warren ME, Borchers CH, Tempst P, Zhang Y 2006 Histone demethylation by a family of JmjC domain-containing proteins. Nature 439:811–816 [DOI] [PubMed] [Google Scholar]
  24. Knebel J, De Haro L, Janknecht R 2006 Repression of transcription by TSGA/Jmjd1a, a novel interaction partner of the ETS protein ER71. J Cell Biochem 99:319–329 [DOI] [PubMed] [Google Scholar]
  25. Yamane K, Toumazou C, Tsukada Y, Erdjument-Bromage H, Tempst P, Wong J, Zhang Y 2006 JHDM2A, a JmjC-containing H3K9 demethylase, facilitates transcription activation by androgen receptor. Cell 125:483–495 [DOI] [PubMed] [Google Scholar]
  26. Kouzarides T 2002 Histone methylation in transcriptional control. Curr Opin Genet Dev 12:198–209 [DOI] [PubMed] [Google Scholar]
  27. Cloos PA, Christensen J, Agger K, Helin K 2008 Erasing the methyl mark: histone demethylases at the center of cellular differentiation and disease. Genes Dev 22:1115–1140 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Masuyama H, MacDonald PN 1998 Proteasome-mediated degradation of the vitamin D receptor (VDR) and a putative role for SUG1 interaction with the AF-2 domain of VDR. J Cell Biochem 71:429–440 [PubMed] [Google Scholar]
  29. Li XY, Boudjelal M, Xiao JH, Peng ZH, Asuru A, Kang S, Fisher GJ, Voorhees JJ 1999 1,25-Dihydroxyvitamin D3 increases nuclear vitamin D3 receptors by blocking ubiquitin/proteasome-mediated degradation in human skin. Mol Endocrinol 13:1686–1694 [DOI] [PubMed] [Google Scholar]
  30. Jääskeläinen T, Ryhänen S, Mahonen A, DeLuca HF, Mäenpää PH 2000 Mechanism of action of superactive vitamin D analogs through regulated receptor degradation. J Cell Biochem 76:548–558 [DOI] [PubMed] [Google Scholar]

Articles from Endocrinology are provided here courtesy of The Endocrine Society

RESOURCES