SEARCH

SEARCH BY CITATION

Keywords:

  • TAFII-17;
  • Paget's disease;
  • 1,25-dihydroxyvitamin D3;
  • osteoclast formation;
  • bone marrow cultures

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

In contrast to normal OCL precursors, pagetic OCL precursors express MVNP and form OCL at physiologic concentrations of 1,25(OH)2D3, as do normal OCL precursors transfected with the MVNP gene. Using a GST-VDR chimeric protein, we identified TAFII-17 as VDR binding protein expressed by pagetic OCL precursors and MVNP transduced normal OCL precursors. TAFII-17 was in part responsible for the increased 1,25(OH)2D3 responsivity of pagetic OCL precursors.

Introduction: Pagetic osteoclasts (OCLs) and their precursors express measles virus nucleocapsid protein (MVNP) and form large numbers of OCLs at low concentrations of 1,25-dihydroxyvitamin D3 [1,25(OH)2D3]. Similarly, normal OCL precursors transfected with MVNP also form OCLs at low concentrations of 1,25(OH)2D3. These results suggest that expression of MVNP in OCL precursors enhances vitamin D receptor (VDR)-mediated gene transcription.

Materials and Methods: To determine the mechanism for the increased OCL formation capacity of pagetic OCL precursors in response to 1,25(OH)2D3, lysates from pagetic and MVNP-transduced normal OCL precursors were incubated with a GST-VDR chimeric protein.

Results: A 17-kDa peptide that bound VDR was detected in MVNP-transduced cells and pagetic OCL precursors treated with 1,25(OH)2D3. This peptide was identified as TAFII-17, a component of the TFIID transcription complex. Expression of increased levels of TAFII-17 in cells allowed TAFII-17 to bind to VDR at low concentrations of 1,25(OH)2D3. An antisense oligonucelotide (AS-ODN) to TAFII-17 significantly decreased OCL formation in response to 1,25(OH)2D3 in pagetic but not normal marrow cultures by ∼40%. Transfection of TAFII-17 or MVNP into NIH3T3 cells increased VDR transcriptional activity as measured by DR-3 reporter assays.

Conclusion: These data show that expression of the MVNP gene in OCL precursors results in increased levels of TAFII-17. TAFII-17 can bind VDR at low concentrations of 1,25(OH)2D3. These results suggest that MVNP expression in Paget's OCL precursors increases expression of a component(s) of the VDR transcription complex that can increase OCL formation.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

PAGET'S DISEASE OF bone is the most exaggerated example of disordered bone remodeling with the primary cellular abnormality residing in the osteoclast (OCL). OCLs are increased in number and size and contain many more nuclei per cell compared with normal OCLs.(1) Immunocytochemical studies have shown that pagetic OCLs contain paramyxoviral-like nuclear inclusions that cross-react with antibodies to measles virus (MV), respiratory syncytial virus, and canine distemper virus nucleocapsid antigen.(2–4) In situ hybridization studies have detected expression of measles virus nucleocapsid protein (MVNP) transcripts in OCLs from patients with Paget's disease.(5) In addition, OCL precursors and circulating peripheral blood cells from patients with Paget's disease express MVNP transcripts.(6) However, the physiologic role that MV infection plays in the abnormal OCL activity in Paget's disease is unknown.

We previously reported that OCL precursors from Paget's patients formed large numbers of OCLs at low concentrations of 1,25(OH)2D3, forming OCLs in vitro at 1,25(OH)2D3 concentrations that are one to two logs lower than that required for normal OCL formation.(7,8) Furthermore, we have shown that normal OCL precursors transduced with the MVNP gene form OCLs that were very similar to OCLs from patients with Paget's disease and formed OCLs at low concentrations of 1,25(OH)2D3.(9) This increased OCL formation in response to 1,25(OH)2D3 was not caused by increased numbers of vitamin D receptors (VDRs).(10) These data suggested that the enhanced capacity of MVNP transduced OCL precursors and OCL precursors from Paget's disease patients to form OCL at low 1,25(OH)2D3 may result from enhanced VDR mediated transcription caused by either increased levels of one or more components of the VDR transcription complex(10) or decreased expression of a corepressor of VDR.

To test this hypothesis, we used a GST-VDR chimeric protein with lysates from Paget's bone marrow cells and MVNP-transduced normal OCL precursors to isolate and characterize VDR binding proteins for their effects on OCL formation.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Chemicals

1,25(OH)2D3 was synthesized in Dr Ishizuka's laboratory as described previously.(11) FBS was purchased from GIBCO-BRL (Grand Island, NY, USA). All other chemicals and media were purchased from Sigma Chemical (St Louis, MO, USA), unless otherwise noted.

Subjects and cell preparation

Bone marrow cells were aspirated under 2% xylocaine anesthesia from the iliac crest of healthy normal donors or patients with Paget's disease into heparinized α-MEM containing 5% FBS, as previously described.(8) Bone marrow mononuclear cells were isolated by separation on hypaque-ficoll gradients (density, 1.077 g/ml), centrifuged at 400g for 30 minutes, and washed three times with α-MEM, as previously described.(8) The Institutional Review Board of the University of Pittsburgh approved these studies.

Transduction of human bone marrow cells

Human bone marrow mononuclear cells were cultivated for 2 days in α-MEM-10% FBS that contained 10 ng/ml each of interleukin-3 (IL-3), interleukin-6 (IL-6), and stem cell factor (SCF; Immunex Research and Development, Seattle, WA, USA). The bone marrow cells were then cultured for an additional 48 h at 37°C in a humidified atmosphere of 5% CO2-air at a density of 1-2 × 105/ml with supernatant (10% vol/vol) containing the vector.(9) Cultures were supplemented with 4 μg/ml of polybrene, 20 ng/ml of IL-3, 50 ng/ml of IL-6, and 100 ng/ml of SCF. In preliminary experiments, we determined that this was the cytokine combination that supported the highest transduction efficiency. After 24 h, the cells were centrifuged, the spent supernatant was removed, freshly prepared viral supernatants supplemented with 4 μg/ml of polybrene and growth factors were added, and the cultures were continued for 24 h. After 48 h, cells were harvested for short-term clonogenic assays in methylcellulose as described below, and an aliquot of the cells was tested for MVNP expression by immunostaining the cells with an anti-MVNP monoclonal antibody (generously provided by Dr Don Forthal, University of California at Irvine, Irvine, CA, USA) as previously described.(2,3) As a control, some of the cells were maintained under the same culture conditions except the viral supernatant was omitted or viral supernatants containing the empty vector were added.

Granulocyte-macrophage colony-forming unit formation

Transduced marrow mononuclear cells were cultured at 5 × 105 cells/well in α-MEM containing 1.2% methylcellulose, 30% FBS, 1% deionized bovine serum albumin (BSA; Sigma), and 100 pg/ml recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF; Immunex Research and Development) with 250 μg/ml G418. Transduced cells were plated in a volume of 1.0 ml in 35-mm culture dishes (Corning; New York, NY, USA) as reported previously.(9) The dishes were incubated at 37°C in humidified atmosphere of 5% CO2-air for 7 days. Colonies were scored after 7 days of culture using an inverted microscope, and G418-resistant colonies were individually picked, using finely drawn pipettes, and used for OCL formation assays.

Long-term cultures for osteoclast formation

G418-resistant granulocyte-macrophage colony-forming unit (CFU-GM)-derived cells (1 × 105) or human marrow mononuclear cells (1 × 105), obtained as described above, were cultured in 96-well plates (Becton Dickinson Labware, Franklin Lakes, NJ, USA) in α-MEM containing 20% horse serum and various concentrations of 1,25(OH)2D3 or IL-6. The cultures were fed every 3 days by replacing one-half the media, and after 14 (for CFU-GM) or 21 days (for marrow mononuclear cells) of culture, the cells were fixed with 1% formaldehyde and tested for cross-reactivity with the monoclonal antibody 23c6, which recognizes the OCL vitronectin receptor (generously provided by Michael Horton, Rayne Institute, Bone and Mineral Center, London, UK), using a Vectastatin-ABC-AP kit (Vector Laboratories, Burlingame, CA, USA). The 23c6+ multinucleated cells (≥3 nuclei/cell) were scored using an inverted microscope. The bone resorption capacity of the multinucleated cells was assessed by measuring the number of resorption lacunae formed on dentin slices as previously described.(9)

Western blot analysis

To evaluate VDR protein levels, MVNP- or empty vector (EV)-transduced CFU-GMs were isolated as described above. After 4 days of culture with 1,25(OH)2D3, cell lysates were prepared as described.(10) The protein concentration of the lysates was determined by the Bradford method, and the same amount of protein from each lysate was loaded onto 15% SDS-PAGE. Rat anti-human vitamin D receptor monoclonal antibody (Affinity BioReagents, Golden, CO, USA) was used to detect VDR. VDR content was measured by enhanced chemiluminescence (ECL) following the manufacturer's protocol (Amersham Biosciences, Piscataway, NJ, USA).

Binding affinity of 1,25(OH)2D3 to the VDR

Pooled MVNP- or EV-transduced CFU-GM (108 cells) and EV- or MVNP-transduced NIH3T3 cells (27 10-cm culture dishes each) were washed with PBS three times, and the cells were homogenized in phosphate buffer A (25 mM KH2PO4, 0.1 M KCl, 1 mM dithiothreitol) by means of 20 passes of a Potter-Elvehjem Teflon glass homogenizer kept in ice. The homogenate was centrifuged at 105,000g for 1 h in a Hitachi RP72 ultracentrifuge. This supernatant was used as the VDR-containing fraction. The saturation analysis of 1,25(OH)2D3 binding to VDR was performed as follows. Varying amounts of [3H]1,25(OH)2D3 (specific activity 168 Ci/mmol, 3.26-182.82 fmol), with or without a 500-fold excess of unlabeled 1,25(OH)2D3, were dissolved in 50 μl of absolute ethanol in 12 × 75-mm polypropylene tubes (Walter Sarstedt, Numbrecht, Germany). One milliliter of the VDR-containing fraction was diluted to 1.8 mg protein/ml in phosphate buffer A, and 1 mg gelatin was added to each tube kept in an ice bath. The assay tubes were incubated in a shaking water bath for 1 h at 25°C and chilled in an ice bath. One milliliter of 40% (wt/vol) polyethylene glycol 6000 in distilled water was added to each tube, and the tubes were mixed vigorously and centrifuged at 2260g for 60 minutes at 4°C. After the supernatant was decanted, the bottom of the tube containing the pellet was cut off into a scintillation vial containing 10 ml of dioxane-based scintillation fluid containing 10% naphthalene and 0.5% Omnifluor (DuPont, Boston, MA, USA) in 1,4-dioxane. The radioactivity was measured in a Beckman liquid scintillation counter (Model LS6500) using an external standard. Total binding was calculated as the amount of bound [3H]1,25(OH)2D3 in the absence of unlabeled 1,25(OH)2D3, and nonspecific binding was calculated as the amount of bound [3H] 1,25(OH)2D3 in the presence of 500-fold excess of unlabeled 1,25(OH)2D3. Specific binding was calculated by subtracting the nonspecific binding from the total binding. The dissociation constant (Kd) was calculated by Scatchard analysis of specific binding of 1,25(OH)2D3.

Gene reporter assays

The promoter region of the human 24-hydroxylase gene (−186 to −5), which contains two vitamin D responsive elements (VDRE; a gift from Dr E Eguchi, Teijin Institute for Bio-Medical Research, Tokyo, Japan), was cloned into a luciferase reporter vector pGL3-Basic Vector (Promega, Madison, WI, USA).(12) This plasmid construct was cotransfected with the β-galactosidase expression plasmid into CFU-GM-derived cells or NIH3T3cells transduced with MVNP or EV using the DMRIE-C Reagent (Invitrogen Life Techno). Sixteen hours after transfection, vehicle (0.1% ethanol) or 1,25(OH)2D3 (10−10–10−8 M) was added. Twenty-four hours later, the cells were harvested and lysed in the cell lysate solution provided with the luciferase assay kit (Promega). The luciferase activities of the cell lysates were measured with the luciferase assay kit according to the manufacturer's instructions and were standardized by comparing the galactosidase activities of the same cell lysates as determined with a β-galactosidase enzyme assay system (Promega). The DR-5 reporter gene was a generous gift from Dr MJ Tsai (Baylor College of Medicine, Houston, TX, USA). The DR-5 reporter assay was performed using the same protocol for the DR-3 gene reporter assay, except that all trans-retinoic acid (Sigma) was substituted for 1,25(OH)2D3.

GST-VDR affinity binding assays and protein sequencing

The GST-VDR chimeric protein was a gift from Dr S Kato (University of Tokyo, Tokyo, Japan).(13) Twenty micrograms of GST-VDR protein was incubated for 2 h at 4°C with 2 mg of lysates from cells treated with 10−10–10−8 M 1,25(OH)2D3. Ten milligrams of glutathione-conjugated Sepharose 4B (Amersham) was added, the suspension was incubated for 1 h at 4°C, and the beads were collected. Protein was eluted in SDS-PAGE buffer, and the samples were resolved by SDS-PAGE (15% polyacrylamide gel; Bio-Rad, Hercules, CA, USA).

For protein sequence analysis, samples prepared as described above were transferred onto PVDF membranes (Bio-Rad). The membranes were stained with Ponceas S, and the 17-kDa band was cut out from the membrane. After being reduced and S-pyridilethylated, it was digested by Achromobacter protease I for 48 h. After sonication, the supernatant was loaded onto a reversed-phase μ-Bondasphere C8 HPLC column (Waters, Milford, MA, USA). Fractionated samples were sequenced using the 492 precise protein sequencing system (PE Applied Biosystems, Foster City, CA, USA). The amino acid sequences were subjected to analysis using Swiss protein.(14)

For Western blot analysis, the peptides that bound GST-VDR were subjected to SDS-PAGE using 15% polyacrylamide gels, and the was blot transferred onto a PVDF membrane (Millipore Corp.) After blocking with 5% skim milk in Tris-buffered saline containing 0.1% Tween-20 (TBST), the membrane was incubated for 1 h with anti-rabbit-TAFII-17 antibody(15) (generously provided by Dr RG Roeder, the Rockefeller University) at 1:3000 dilution in TBST containing 1% BSA. The blot was incubated for 1 h with horseradish peroxidase-conjugated goat anti-rabbit IgG (DAKO, Carpinteria, CA, USA), and the bands were visualized with an ECL system (Amersham Life Science, Arlington Heights, IL, USA).

To determine if protein levels of TAFII-17 were increased in MVNP-transfected cells, the day 4 cell lysates described above from 1,25(OH)2D3 (10−8 M)-stimulated MVNP- or EV-transduced CFU-GMs were mixed with native sample buffer (Bio-Rad) and were loaded using 15% polyacrylamide gels (Bio-Rad) with SDS free Tris/Glysin Buffer (Bio-Rad). The blots were transferred onto PVDF membranes (Millipore Corp.). Western immunoblotting was performed using anti TAFII-17 antibody or anti-VDR antibody as described above. The 72-kDa band that was identified with the anti-TAFII-17 antibody was cut out and eluted from the membrane. The band was electrophoresed under denaturing conditions, and the gel was blotted with anti-VDR.

Modified mammalian two-hybrid assays

The TAFII-17 cDNA was digested by SmaI and SalI and inserted into the pM vector (CLON-TECH; pM-TAFII-17). EcoRI released the full-length cDNA for human VDR from the pSG5 vector and fused in-frame into the pVP16 vector, which contains the activation domain of a herpesvirus (CLON-TECH; pVP16-hVDR).(16) To examine the interaction of TAFII-17 and VDR, 0.5 μg of pM-TAFII-17, 0.5 μg of pVP-16-hVDR and 0.5 μg of pGVP2-GAL4BS together with 0.25 μg of plasmid containing β-galactosidase cDNA were transfected into NIH3T3 cells using Lipofectamine (GIBCO BRL). Twenty-four hours later, vehicle or 10−8 M 1,25(OH)2D3 was added. After 24 h of incubation, the luciferase activity of the cell lysates was examined and standardized using β-gal activity.

Antisense oligodeoxynucleotide synthesis

The TAFII-17 antisense phosphothioate oligodeoxynucleotide (AS-ODN) was composed of 16 bases that targeted the region that spans the translation start codon of human TAFII-17 mRNA (Table 1). The sense and two four-base mismatches (MS1, MS2) of the AS-ODN were also designed as negative controls (Table 1). All these phosphothioate oligonucleotides were synthesized (Integrated DNA Technologies, Coralville, IA, USA) on an automated solid-phase nucleotide synthesizer and subsequently filter-sterilized. The GC content was the same for all of these oligonucleotides

Table Table 1.. Sequences of TAFII-17 Oligodeoxynucleotides
Thumbnail image of

Antisense oligodeoxynucleotide uptake by osteoclast precursors

OCL precursors were incubated (1 × 105 cells) in 96-well tissue culture plates or 5 × 105 cells in 24-well tissue culture plates in the presence or absence of the antisense or either MS-1 or MS-2 in α-MEM containing 20% horse serum and Penetratin-1 (Q. Biogene) as the carrier. The effect of the AS-ODN and Penetratin-1 on MNC formation was investigated to determine the optimal transfection conditions for OCL precursors. Based on these preliminary experiments, AS-ODN in Penetratin-1 carrier was used at a concentration of 3 μM to introduce the AS-ODN, MS-1, and MS-2 into OCL precursors. The OCL precursors were cultured for 4 (for GST-VDR pull-down assays) or 14 days (for OCL formation). The media were replaced every 3 days with fresh media containing oligodeoxynucleotides.

Bone resorption assays

The CFU-GM-derived cells transduced with MVNP (105 cells/ well) were cultured with 1,25(OH)2D3 (10−8 M) and/or AS-TAFII-17 ODN or the MS-1 MS-ODN on mammoth dentin slices (Wako). After 2 weeks of culture, the cells were removed, the dentin slices were stained with acid hematoxylin (Sigma), and the number of resorption lacunae were counted microscopically as previously described.(9)

Generation of modified Hill equations

To determine if cells transduced with MVNP showed increased responsivity to 1,25(OH)2D3 or retinoic acid compared with EV-transduced cells, a modified Hill equation was used. The relationship between 1,25(OH)2D3 or retinoic acid concentrations and reporter activity or osteoclast formation by CFU-GM-EV and CFU-GM-MVNP was described by modified Hill equation:

  • equation image

In this equation, the effect (E) is the change in DR-3 or DR-5 reporter activity or OCL formation, Emax is the maximum effect possible, which was fixed at 100%; C is the concentration of 1,25(OH)2D3 or retinoic acid; C50% is the concentration of 1,25(OH)2D3 or retinoic acid associated with 50% Emax; and H is the Hill constant, an indication of the sigmoidicity of the curve.

Statistical analysis

The results of culture assays are reported as the mean ± SEM for typical experiments. Significance was evaluated by a two-sided nonpaired Student's t-test, with differences of p < 0.05 considered significant. A similar pattern of results was seen in experiments using bone marrow cells from 10 separate normal marrow donors transduced with EV or MVNP and 5 Paget's patients.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Expression of the VDR

To determine the mechanism responsible for the increased OCL formation by OCL precursors from pagetic patients to low levels of 1,25(OH)2D3, we tested if MVNP- or EV-transduced normal OCL precursors (CFU-GM) expressed increased amounts of VDR protein or bound increased amounts of 1,25(OH)2D3/mg protein. We previously had reported using competitive PCR that VDR mRNA levels did not differ in Paget's or normal marrow.(10) The amounts of VDR (Fig. 1) and 1,25(OH)2D3 specifically binding to VDR was similar in EV- or MVNP-transduced CFU-GM (3.25 versus 2.97 fmols/mg protein). The amounts of the 80-kDa VDR band, whose occurrence has been reported by others,(17) also did not differ between EV and MVNP transduced CFU-GM.

thumbnail image

Figure FIG. 1.. Expression of the VDR in MVNP- and EV-transduced OCL precursors. The cell lysates (20 μg) were obtained from MVNP- or EV-transduced CFU-GMs cultured for 4 days in α-MEM-10% FBS and separated on 15% SDS-PAGE gels. VDR protein (50 kDa) was detected by immunoblot analysis using the rat monoclonal antibody 9A7.

Download figure to PowerPoint

DR-3, DR-5 reporter activity in MVNP-transduced normal CFU-GM-derived cells

To determine whether VDR-mediated gene transcription was increased in MVNP-transduced cells and whether it was relatively specific to VDR, a luciferase reporter vector containing a DR-3 (VDR) or DR-5 (retinoic acid receptor [RAR]) response element was inserted into EV- or MVNP-transduced early osteoclast precursors, CFU-GMs. VDR-mediated gene transcription was significantly increased in CFU-GM cells transduced with the MVNP cDNA compared with EV-transduced cells. This increase was detectable at concentrations of 1,25(OH)2D3 that were one log less than that required increasing DR-3 reporter activity in EV-transduced CFU-GM cells (Fig. 2A). Using a modified Hill equation to compare the dose response curves, the concentration of 1,25(OH)2D3 that induced a 50% of maximal response (C50%) for the 1,25(OH)2D3 dose response curve (0-10−8 M) for MVNP-transduced cells differed significantly from EV-transduced cells (0.47 × 10−9 versus 2.0 × 10−9 M, p < 0.05). In contrast, transduction of the DR-5 reporter construct into CFU-GM cells transduced with the MVNP or EV construct showed that, although basal transcription was increased in CFU-GM cells transduced with the MVNP gene, the C50% for the retinoic acid dose-response curve (0-10−8 M retinoic acid) for MVNP-transduced cells did not differ significantly from that of the EV-transduced cells (0.7 ± 1.25 × 10−9 versus 1.5 ± 0.72 × 10−9 M, respectively; Fig. 2B). As shown in Figs. 2C and 2D, NIH3T3 cells transfected with MVNP also showed enhanced VDR-mediated transcriptional activity in response to 1,25(OH)2D3 in analogous fashion to OCL precursors expressing MVNP. The transfection efficiency for the reporters in NIH3T3 cells, (85% versus 15%), which are adherent, was much greater compared with CFU-GMs, which are nonadherent hematopoietic cells, as determined by co-transfection of the β-galactosidase gene. NIH3T3 cells normally express VDR.

thumbnail image

Figure FIG. 2.. Effects of transducing MVNP or EV into CFU-GM or NIH3T3 cells expressing a VDR (DR-3) or RAR (DR-5) reporter construct. The reporter plasmid containing the DR-3 or DR-5 promoter and β-gal expression vector were introduced into CFU-GM (CFU-GM-EV and CFU-GM-MVNP) cells or NIH3T3 (NIH3T3-EV and NIH3T3-MVNP) cells. Different concentrations of (A) 1,25(OH)2D3 or (B) all trans-retinoic acid were added 24 h after the transfection of CFU-GM. The same concentrations of (C) 1,25(OH)2D3 or (D) all trans-retinoic acid were added to NIH3T3 cells. Results are expressed as the mean ± SE. *p < 0.01 compared with cells treated with media alone. Similar results were seen in three independent experiments.

Download figure to PowerPoint

Binding affinity of VDR in EV- and MVNP-transduced NIH3T3 cells

A potential mechanism that could account for increased VDR-mediated transcription would be increased affinity of VDR for its ligand. To test this possibility, NIH3T3 cells were used as a surrogate for OCL precursors and were transfected with MVNP or EV, and the VDR binding affinity was determined. It would have been preferable to perform these experiments with MVNP- or EV-transduced OCL precursors, but it is impossible to obtain sufficient numbers of cells required for this assay from human bone marrow cells. As noted above, NIH3T3 cells transfected with MVNP display enhanced VDR-mediated transcription in analogous fashion to OCL precursors expressing MVNP (Figs. 2C and 2D). Scatchard analysis showed that the affinity constants (Kd) and Bmax for 1,25(OH)2D3 binding VDR in EV- or MVNP-transduced NIH-3T3 cells were 2.70 × 10−11 M and 14.0 fmol/mg protein versus 1.90 × 10−11 M and 11.7 fmol/mg protein, respectively, and were not significantly different (Fig. 3). Only one class of VDR was detected.

thumbnail image

Figure FIG. 3.. Scatchard analysis of 1,25(OH)2D3 receptor binding in EV- or MVNP-transduced NIH3T3 cells. Scatchard analysis was performed as described in the Materials and Methods section. Only one class of 1,25(OH)2D3 receptor was detected. Similar results were seen in two independent experiments.

Download figure to PowerPoint

Binding of the GST-VDR chimeric protein to a 17-kDa peptide in Paget's marrow cells and MVNP-transduced normal CFU-GM-derived cells:

We used a GST-VDR chimeric protein with lysates from MVNP-transduced normal OCL precursors (CFU-GM-derived cells) and marrow cells from involved bones from patients of Paget's disease to try to isolate potential VDR binding proteins. As shown in Fig. 4A, a 17-kDa peptide was detected that bound VDR and was expressed at high levels in MVNP-transduced cells treated with 1,25(OH)2D3, but was expressed in EV-transduced cells. Bone marrow cells derived from patients with Paget's disease also expressed high levels of this peptide in the presence or absence of added 1,25(OH)2D3 (Fig. 4A). Furthermore, NIH3T3 cells that were transduced with the MVNP construct also expressed a similar 17-kDa VDR binding protein when treated with 1,25(OH)2D3 (Fig. 4A). This 17-kDa band was not present in EV-transduced NIH3T3 cells (Fig. 4A). However, a faint 15-kDa band was present in EV-transduced cells that also co-migrated with the 17-kDa band in MVNP-transduced cells. At present, we have not been able to determine the identity of this band.

thumbnail image

Figure FIG. 4.. Binding of a GST-VDR chimeric protein to a 17-kDa protein expressed at high levels in Paget's marrow cells and MVNP-transduced cells. (A) GST-VDR (20 μg) was incubated for 2 h at 4°C with 2-4 mg of cell lysates from cells treated with 10−8 M 1,25(OH)2D3 for 48 h. Two micrograms of immobilized Sepharose 4B was incubated with labeled proteins for 1 h at 4°C. Samples were resolved by SDS-PAGE and detected by Coomassie blue staining. Similar results were seen in three independent experiments. (B) Western blot analysis showing that the 17-kDa band is TAFII-17. GST-VDR (20 μg) was incubated for 2 h at 4°C with 24 mg cell lysates from cells treated with 10−10–10−8 M 1,25(OH)2D3 for 4 days. Two micrograms of immobilized sepharose 4B was incubated with labeled proteins for 1 h at 4°C. Similar results were seen in four independent experiments. (C) Western blot analysis of total cell lysates from MVNP- or EV-transfected CFU-GMs. TAFII-17 protein levels were detected as described above. Similar results were seen in two independent experiments.

Download figure to PowerPoint

Microsequence analysis of the 17-kDa band in two independent experiments identified the protein (N-terminal; PEPASXPPQG; TAFII-17; 20-29; fragment 21): (K)ACTTXAH; TAFII-20 (140; 141-147). The 17-kDa band in patients with Paget's disease and MVNP-transduced CFU-GM samples was identical to TAFII-17, which is a component of the TFIID transcription complex.(15,18)

To determine if the increased levels of TAFII-17 were present in the cells rather than TAFII-17 having increased interaction with VDR, we measured the relative levels of TAFII-17 in EV- and MVNP-transfected CFU-GMs by Western blot analysis as shown in Fig. 4C. MVNP-transduced cells expressed higher levels of TAFII-17. To determine if the increased levels of TAFII-17 could form complexes with VDR at low concentrations of 1,25(OH)2D3, CFU-GM that had been transfected with the MVNP cDNA or EV were treated with 10−10–10−8 M 1,25(OH)2D3 and the lysates used for GST-VDR pull-down experiments. The identity of the 17-kDa band in these experiments as TAFII-17 was confirmed by Western blot analysis (Fig. 4B). TAFII-17 could complex to VDR in MVNP-transduced cells at very low concentrations of 1,25(OH)2D3 (10−10 M) and could complex to VDR even when the cells were only exposed to the very low concentrations of 1,25(OH)2D3 present in the sera used in the media.

Interaction of 1,25(OH)2D3 and VDR with TAFII-17

To further examine protein-protein interactions between VDR and TAFII-17, the mammalian two-hybrid system was used. An expression vector in which human TAFII-17 was fused to GAL4DBD (pM-TAFII-17) was used as the bait construct and the human VDR (pVP16-hVDR) was used as the prey vector (Fig. 5A). VDR mediated gene transcription was induced when 10−8 M 1,25(OH)2D3 was added with the VDR and TAFII-17 constructs (Fig. 5B), confirming that VDR interacted with TAFII-17 or a complex containing TAFII-17.

thumbnail image

Figure FIG. 5.. Interaction between VDR and TAFII-17. (A) Modified mammalian two-hybrid assay was used to examine the interaction between VDR and TAFII-17. The TAFII-17 cDNA was inserted at the EcoRI site of the pMvector, which contains a GAL4 DNA binding domain (pM-TAFII-17). (B) The interaction of VDR and TAFII-17 was examined by measuring the reporter activity in NIH3T3 cells transduced with pM-TAFII-17, pVP16-hVDR, and the reporter plasmid containing the GAL4 DNA-binding sites (pGVP2-GAL4BS). 1,25(OH)2D3 (10−8 M) was added after transfection. Results are expressed as the mean ± SE. *p < 0.01 compared with cells treated with pGVP2-GAL4BS, pGVP2-GAL4BS + pVP16-hVDR, or pGVP2-GAL4BS + pM-TAFII-17 transfected groups. Similar results were seen in three independent experiments. (C) Western immunoblotting to determine if TAFII-17 directly binds VDR. Day 4 cell lysates from 10−8 M 1,25(OH)2D3-stimulated MVNP-transduced CFU-GMs were mixed with native sample buffer and were loaded onto 15% polyacrylamide gels in non-SDS-containing Tris/Glysine Buffer (Bio-Rad). A 17-kDa TAFII-17 band and a 72-kDa band, which contained TAFII-17, were detected with the anti-TAFII-17 antibody. (D) Western analysis of the 72-kDa band from C. The 72-kDa band was eluted and run on SDS gels, and the gels were transferred to membranes and blotted without VDR. Similar results were seen in two independent experiments.

Download figure to PowerPoint

To determine if TAFII-17 could directly bind VDR, cell lysates from MVNP transduced CFU-GM were fractionated on native gels and blotted with either anti-VDR or anti-TAFII-17. As shown in Fig. 5C, a 72-kDa complex that contained TAFII-17 (MW = 17 kDa) and VDR (MW = 55 kDa) was detected. To confirm that VDR was present in the 72-kDa complex, the band was cut out, eluted from the membrane, run under denaturing conditions, and blotted with anti-VDR. Figure 5D show that VDR was part of the 72-kDa complex.

Effects of anti-sense TAFII-17-ODN on OCL formation:

To determine the potential physiologic relevance of increased TAFII-17 on OCL formation, the effects of the TAFII-17 AS-ODN on OCL formation by Paget's OCL precursors and MVNP-transduced CFU-GM were investigated. Treatment of Paget's OCL precursors with TAFII-17-antisense ODN markedly decreased the protein expression levels of TAFII-17 compared with MS-ODN-treated osteoclast precursors as assessed by GST-VDR pull-down assays. As shown in Fig. 6, treatment with either MS-ODN did not alter the protein levels of TAFII-17 (100 versus 69 densitometry units). When Paget's patient derived or MVNP-transduced CFU-GMs were cultured for 14 days in the presence or absence of 3 μM TAFII-17 AS-ODN and treated with 10−11–10−7 M 1,25(OH)2D3, OCL formation was inhibited (37 ± 6% Paget's patients and 64 ± 4% MVNP-transduced cells; Fig. 7A). In contrast, 3 μM of TAFII-17 AS-ODN did not block OCL formation in cultures of EV transduced CFU-GM (Fig. 7B). Importantly, TAFII-17 AS-ODN did not affect OCL formation induced by IL-6 in Paget's patients and normal donors. (Fig. 7C). Treatment with either MS-ODN did not significantly decrease OCL formation.

thumbnail image

Figure FIG. 6.. TAFII-17 antisense oligonucleotide decreases the amount of TAFII-17 in Paget's OCL precursors. GST-VDR (20 μg) was incubated for 2 h at 4°C with 2 mg cell lysates from cells treated with 10−8 M 1,25(OH)2D3 for 4 days. Two micrograms of immobilized sepharose 4B was incubated with reaction mixture for 1 h at 4°C. Samples were resolved by SDS-PAGE and detected by silver staining. Similar results were seen in three independent experiments.

Download figure to PowerPoint

thumbnail image

Figure FIG. 7.. Effects of TAFII-17 antisense oligonucleotide (AS-ODN) or missense oligonucleotide (MS-ODN) on OCL formation by CFU-GM-derived cells from Paget's and normal marrow mononuclear cells, or normal CFU-GM-derived cells transduced with MVNP or EV. CFU-GM-derived cells (105 cells/well) from normal or Paget's marrow (A) or CFU-GM-derived cells transfected with MVNP or EV (B) were cultured for OCL cell formation in the presence of 10−11 M to -10−7 M 1,25(OH)2D3. Alternatively, CFU-GM-derived cells from normal or Paget's marrow mononuclear cells were also cultured with 10-1000 pg/ml of IL-6 (C). Three micromolar AS-TAFII-17 ODN or MS-AS ODN were added to these cultures. After 14 days of culture, the cells were fixed and stained with the 23C6 monoclonal antibody, which identifies OCLs. The results are expressed as mean ± SE for quadruplicate cultures from a typical experiment, p < 0.001 compared with results of cultures treated with the same concentration of MS-ODN. A similar pattern of results was seen in five independent experiments. In addition, the CFU-GM-derived cells transduced with MVNP (105 cells/well) were cultured for OCL formation with 1,25(OH)2D3 (10−8 M) with AS-TAFII-17 ODN or MS-ODN on the mammoth dentin slices (D). After 2 weeks of culture, the dentin slices were stained with acid hematoxylin. Magnification, ×100.

Download figure to PowerPoint

To confirm that antisense TAFII-17-ODN affected OCL formation, bone resorption assays were performed with MVNP transduced CFU-GM treated with TAFII-17-AS-ODN. Treatment with 3 and 10 μM of TAFII-17-AS-ODN markedly decreased pit formation compared with MS-ODN- and 1,25(OH)2D3-treated cultures (Fig. 7D). The number of pits formed on dentin slices in cultures containing 1,25(OH)2D3 or 1,25(OH)2D3 and 3 μM of MS-ODN were 132 ± 12 versus 129 ± 20, respectively. Rare resorption pits were detected in cultures treated with 3 and 10 μM of TAFII-17-AS-ODN and low concentrations of 1,25(OH)2D3 (10−9 M).

Effects of transfection of TAFII-17 on VDR-mediated gene transcription:

NIH3T3 cells were stably transfected with TAFII-17 or EV and tested for DR3 reporter activity. Transfection of TAFII-17 increased VDR mediated transcription in response to low levels of 1,25(OH)2D3 in an analogous fashion as MVNP (Figs. 2C and 8). In contrast, transfection of DR-5 reporter construct into these NIH3T3 cells showed that the retinoic acid dose-response curve was similar in EV- and MVNP-transfected cells, although basal transcription was increased (Fig. 8).

thumbnail image

Figure FIG. 8.. Effects of transfecting TAFII-17 into NIH3T3 cells expressing a VDR (DR3) or retinoic acid receptor (DR5) reporter constructs. The reporter plasmid was transfected into NIH3T3 (NIH3T3-EV and NIH3T3-TAFII-17) cells. Different concentrations of (A) 1,25(OH)2D3 or (B) all trans-retinoic acid were added to the cultures for 24 h after the transfection. Results are expressed as the mean ± SE. *p < 0.01 compared with cells treated with media alone. Similar results were seen in three independent experiments.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We have previously reported that OCL precursors from patients with Paget's disease contain MVNP and form OCL at 1,25(OH)2D3 concentrations that are one to two logs less than required by normal OCL precursors.(8–10) We have further shown that MVNP-transduced normal OCL precursors also form OCL that share many of the features of pagetic OCLs and form OCLs at low concentrations of 1,25(OH)2D3.(9) Our hypothesis to explain this phenomenon was that VDR-mediated transcription was enhanced because of increased levels of VDR coactivators or decreased levels of VDR corepressors. In support of this hypothesis, as shown in Fig. 2A, MVNP-transduced OCL precursors showed increased DR-3 reporter activity at 1,25(OH)2D3 concentrations that were one to two logs less than that required for the EV-transduced cells. This increase transcription activity seemed to be relatively specific for VDR because, although basal transcription was increased in MVNP- or EV-transduced cells transfected with a DR-5 reporter construct, the C50% for the retinoic acid dose-response curves of MVNP-transduced cells did not differ from that of the EV-transduced cells (Fig. 2B).

Several mechanisms may be responsible for the increased VDR-mediated transcriptional activity induced by MVNP and the increased OCL formation by pagetic and MVNP-transduced normal OCL precursors treated with low levels of 1,25(OH)2D3. One possibility is that VDR numbers are increased by expression of MVNP in OCL precursors. However, as shown in Fig. 1, VDR protein levels were similar in EV- or MVNP-transduced OCL precursors. These data are consistent with our previous report using competitive RT-PCR, which showed similar levels of VDR mRNA in pagetic and normal OCL precursors.(10)

A second possibility is that, in cells expressing MVNP, the affinity of VDR for 1,25(OH)2D3 is increased. As shown in Fig. 2C, although NIH3T3 cells transfected with MVNP show increased VDR-mediated transcription in response to low levels of 1,25(OH)2D3 in an analogous fashion to OCL precursors expressing MVNP, Scatchard analysis showed that the affinity (Kd) of 1,25(OH)2D3 for VDR and VDR receptor numbers were similar in EV- or MVNP-transfected NIH3T3 cells (Fig. 3).

A third possibility is that MVNP induces expression of a component of the VDR transcription complex or decreases the levels of a VDR corepressor. Experiments using a GST-VDR chimeric protein with lysates from marrow cells from Paget's disease patients or MVNP-transduced normal OCL precursors did not show that any VDR binding protein was decreased in pagetic or MVNP-transduced cells compared with EV-transfected cells. These results suggest that a corepressor of VDR was not decreased by MVNP, although more extensive experiments are required to confirm this preliminary observation. Instead, these experiments showed that TAFII-17 levels were increased in MVNP-transduced CFU-GM and pagetic cells and bound VDR. Interestingly, increased amounts of TAFII-17 were detected in GST-VDR pull-down experiments from cells treated with increasing concentrations of 1,25(OH)2D3, using other MVNP-transduced CFU-GM or pagetic marrow cells. TAFII-17 was only detected in GST-VDR pull-down experiments with lysates from pagetic cells cultured in the absence of 1,25(OH)2D3. This difference most likely reflects the prolonged exposure of pagetic cells to low concentrations of 1,25(OH)2D3 in vivo before isolation and testing (Fig. 4A). Consistent with this possibility, when MVNP-transduced CFU-GMs were exposed to low levels of 1,25(OH)2D3 for 4 rather than 2 days, as shown in Fig. 4B, increased levels of TAFII-17 could be detected in the absence of added 1,25(OH)2D3.

Mammalian two hybrid assays showed that TAFII-17 interacted with VDR (Fig. 5B). Furthermore, Western blot analysis of lysates from MVNP-transduced CFU-GMs (Figs. 5C and 5D) showed the direct association of TAFII-17 and VDR in cells. However, the mechanism for how TAFII-17 increases VDR-mediated transcription is still unclear. Possibly, TAFII-17 may act as a coactivator for VDR-mediated gene transcription or the increased amounts of TAFII-17 permit formation of the VDR transcription complex at low concentrations of 1,25(OH)2D3 bound to VDR. Consistent with the latter hypothesis, high concentrations of TAFII-17 in MVNP-transduced OCL precursors allowed the transcription complex to form at low concentrations of 1,25(OH)2D3 bound to VDR (Fig. 4B). In support of this finding, overexpression of TAFII-17 in NIH3T3 cells increased VDR-mediated gene transcription (Fig. 8).

Further support for a role of TAFII-17 in the enhanced OCL formation by pagetic OCL precursors in response to low levels of 1,25(OH)2D3 are our experiments using TAFII-17 AS-ODN. When MVNP-transfected OCL precursors were treated with an AS-ODN to TAFII-17, the TAFII-17 AS-ODN decreased the amount of TAFII-17 bound to VDR in Paget's patient-derived OCL precursors and MVNP-transduced CFU-GMs and significantly reduced OCL formation in response to 1,25(OH)2D3. Decreased OCL formation was not observed in normal CFU-GM- or EV-transduced CFU-GM cultures treated with TAFII-17 AS-ODN, consistent with the very low levels of TAFII-17 bound to VDR in normal OCL precursors. This effect seems to be relatively specific for VDR-mediated OCL formation by pagetic or MVNP-transduced cells, because OCL formation by IL-6 was not affected by blocking TAFII-17 expression, and the antisense construct had no effect on the expression of IL-6 in these cultures (Fig. 7C). In addition, other components of the VDR transcription complex may also be increased in pagetic cells and MVNP-transduced CFU-GMs to 1,25(OH)2D3 and play a role in the increased OCL formation.

It is currently unknown if the transcriptional effects of TAFII-17 are specific for VDR-mediated transcription because other TFIID subunits can also bind the thyroid hormone receptor.(19,20) However, our experiments using DR3- and DR5-reported constructs in NIH3T3 cells transfected with TAFII-17 suggest that the effects of TAFII-17 seem to be relative-specific for VDR-mediated TAFIID gene transcription. The TFIID transcription factor complex is composed of a TATA binding protein and at least four TAFII subunits. Until recently, TFIID was thought to be responsible only for promoter recognition and directing RNA polymerase II to core promoters in response to activators of transcription.(21,22) This function of TFIID was thought to be universal rather than gene-specific because TFIID is expressed in all tissues. Recent reports have shown that TFIID components can be cell type specific and regulate specific genes. For example, TAFII-105 is highly expressed in testes and ovary and regulates transcription of specific genes in granulosa cells.(23) Furthermore, mice lacking TAFII-105 are sterile.(23) These data show that different components of the TFIID complex can mediate the transcription of specific genes that can have dramatic effects on the function of specific cells.

It is still unknown if our in vitro observations that pagetic OCL precursors form OCLs at low levels of 1,25(OH)2D3 also occurs in Paget's patients. Currently, there are no animal models of Paget's disease available to rigorously test these findings in vivo. To do this, a VDR antagonist or a 1,25(OH)2D3 antagonist would have to be tested in patients with Paget's disease and normals to determine the effects of these agents on bone turnover and bone resorption. However, our results suggest that enhanced TAFII-17 expression plays an important role in OCL formation by pagetic OCL precursors.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We thank Dr Merrill J Egorin (University of Pittsburgh) for help with the Hill equation calculations and the University of Pittsburgh General Clinic Research Center (GCRC) staff. These studies were supported by National Institutes of Health Grant RO1 AR44603 to GDR and research funds from the Department of Veterans Affairs. We acknowledge the GCRC at the University of Pittsburgh for assistance with obtaining the marrow samples.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Hosking DJ 1981 Paget's disease of bone. BMJ 283: 686688.
  • 2
    Mills BG, Singer FR, Weiner LP, Suffin SC, Stabile E, Holst P 1984 Evidence for both respiratory syncytial virus and measles virus antigens in the osteoclasts of patients with Paget's disease of bone. Clin Orthop 183: 303311.
  • 3
    Basle MF, Russell WC, Goswami KA, Rebel A, Giraudon P, Wild R, Filmon R 1985 Paramyxovirus antigens in osteoclasts from Paget's bone tissue detected by monoclonal antibodies. J Gen Virol 66: 21032110.
  • 4
    Gordon MT, Mee AP, Anderson DC, Sharpe PT 1995 Canine bone marrow cultures infected with canine distemper virus: An in vitro model of Paget's disease. Bone 19: 461S466S.
  • 5
    Basle MF, Fournier JG, Rozenblatt S, Rebel A, Bouteille M 1986 Measles virus RNA detected in Paget's disease bone tissue by in situ hybridization. J Gen Virol 67: 907913.
  • 6
    Reddy SV, Singer FR, Mallette L, Roodman GD 1996 Detection of measles virus nucleocapsid transcripts in circulating blood cells and hematopoietic progenitors from patients with Paget's disease. J Bone Miner Res 11: 16021607.
  • 7
    Yates AJ 1988 Paget's disease of bone. Baillieres Clin Endocrinol Metab 2: 267295.
  • 8
    Kukita A, Chenu C, McManus LM, Mundy GR, Roodman GD 1990 Atypical multinucleated cells form in long-term marrow cultures from patients with Paget's disease. J Clin Invest 85: 12801286.
  • 9
    Kurihara N, Reddy SV, Menaa C, Anderson D, Roodman GD 2000 Osteoclasts expressing the measles virus nucleocapsid gene display a pagetic phenotype. J Clin Invest 105: 607614.
  • 10
    Menaa C, Barsony J, Reddy SV, Cornish J, Cundy T, Roodman GD 2000 1,25-dihydroxyvitamin D3 hypersensitivity of osteoclast precursors from patient with Paget's disease. J Bone Miner Res 15: 228236.
  • 11
    Trost BM, Dumas J 1992 New strategy for the total synthesis of 1α-hydroxyvitamin D derivatives. J Am Chem Soc 114: 19241925.
  • 12
    Umesono K, Murakami KK, Thompson CC, Evans RM 1991 Direct repeats selective response elements for the thyroid hormone, retinoic acid, and vitamin D3 receptors. Cell 65: 12551266.
  • 13
    Yamagisawa J, Yanagi Y, Mashihiro Y, Suzawa M, Watanabe M, Kashiwagi K, Toriyabe T, Kawabata M, Miyazono K, Kato S 2000 Convergence of transforming growth factor-β and vitamin D signaling pathways on SMAD transcriptional coactivators. Science 283: 13171321.
  • 14
    Kuwahara H, Araki N, Makino K, Honda S, Kaibuchi K, Fukunaga K, Miyamoto E, Ogawa M, Saya H 1999 A novel NE-dlg/SAP102-associated protein, p51-nedasin, related to the amidohydrolase superfamily, interferes with the association between NE-dlg/SAP102 and N-methyl-D-aspartate receptor. J Biol Chem 274: 3220432214.
  • 15
    Hoffmann A, Roeder RG 1996 Cloning and characterization of human TAF20/15. J Biol Chem 271: 1819418202.
  • 16
    Ozono K, Saito M, Miura D, Michigami T, Nakajima S, Ishizuka S 1999 Analysis of the molecular mechanism for the antagonistic action of a novel 1,25-dihydroxyvitamin D3 analogue toward vitamin D receptor function. J Biol Chem 274: 3237632381.
  • 17
    Manolagas SC, Yu XP, Girasole C, Bellido T 1994 Vitamin D and the hematolymphopoietic tissue: A 1994 update. Semin Nephrol 14: 129143.
  • 18
    Mengus G, May M, Jacq X, Staub A, Tora L, Chambon P, Davidson I 1995 Cloning and characterization of hTAFII18, hTAFII20 and hTAFII28: Three subunits of the human transcription factor TFIID. EMBO J 14: 15201531.
  • 19
    Mengus A, Gangloff YG, Carre L, Lavigne AC, Davidson I 2000 The human transcription factor IID subunit human TATA-binding protein-associate Factor 28 interacts in a ligand-reversible manner with the vitamin D3 and thyroid hormone receptor. J Biol Chem 275: 1006410071.
  • 20
    Hoffmann A, Oelgeschlager T, Roeder RG 1997 Consideration of transcriptional control mechanisms: Do TFIID-core promoter complexes recapitulate nucleosome-like functions. Proc Natl Acad Sci USA 94: 89288893.
  • 21
    Goodrich JA, Hoey T, Thut CJ, Adamon A, Tjian R 1993 Drosophila TAFII-40 interacts with both a VP16 activation domain and the basal transcription factor TFIIB. Cell 75: 519530.
  • 22
    Verrizer CP, Yokomori K, Chen JL, Tjian R 1994 Dorsophia TAFII-105: Similarity to yeast geneTSM-1 and specific binding to core promoter DNA. Science 264: 933942.
  • 23
    Frieman RN, Albright SR, Zheng S, Sha WC, Hammer RE, Tjian R 2001 Requirement of tissue selective TBP-associated factor TAFII-105 in ovarian development. Science 293: 20842087.