Phosphorylation Selectively Increases Triiodothyronine Receptor Homodimer Binding to DNA*

Thyroid hormone receptors (TRs) are ligand-regulated transcription factors that bind to thyroid hormone response elements (TREs) as monomers and homodimers, and as heterodimers with nuclear proteins such as TR auxiliary proteins and retinoid X receptors. Recently, bacterially expressed human TR beta-1 (hTR beta-1) was shown to be phosphorylated in vitro by HeLa cytosolic extract. However, little is known about the consequences of phosphorylation on the nature of TR complexes. Therefore, we studied the effect of phosphorylation on TR binding of TREs. Bacterially expressed hTR beta-1 was phosphorylated in vitro with ATP by HeLa cytosolic extract. The ratio of phosphoserine to phosphothreonine was approximately 5:1. We then analyzed phosphorylated hTR beta-1 binding to several TREs by electrophoretic mobility shift assay. Phosphorylated hTR beta-1 bound better as a homodimer to the TREs than hTR beta-1 incubated with preheated cytosolic extract. Alkaline phosphatase treatment of the phosphorylated hTR beta-1 eliminated the enhanced homodimer binding to DNA. In contrast, phosphorylation did not affect TR/TR auxiliary protein or TR/retinoid X receptor heterodimer binding to DNA. Triiodothyronine decreased both phosphorylated and unphosphorylated hTR beta-1 homodimer binding to several TREs, and the addition of okadaic acid did not alter this triiodothyronine effect. These results indicate that phosphorylation, in addition to ligand binding, modulates TR dimer binding to TREs. As such, it is possible that phosphorylation may also participate in TR-mediated regulation of transcription.

In addition to the effect of ligand, phosphorylation may be involved in the transcriptional activity andor DNA binding ability of members of the steroidthyroid hormone receptor superfamily. Recently, it has been reported that phosphorylation enhances transcriptional activity of the progesterone receptor (23,24), and both transcriptional and DNA binding activities of the glucocorticoid receptor (25). Chick TRa, v-erbA, and bacterially expressed human TRP-1  are also known to be phosphorylated (26-29). However, the physiological significance of these phosphorylation events is still unclear. The chick TRa and v-erbA are phosphorylated at serine residues that are not conserved in mammalian TRs such as human and rat TR isoforms. The hTRP-1 was phosphorylated in vitro by HeLa cytosolic extract, which resulted in an increase in hTRP-1 DNA binding using an avidin-biotin complex DNA binding assay (29). However, that study did not address the effect of phosphorylation on the nature of TR complexes. Accordingly, we have examined this question using electrophoretic mobility shift assays (EMSA) and in vitro phosphorylated hTRP-1. Surprisingly, phosphorylation of hTRP-1 increased homodimer binding to DNA, without changing TR/RXR or TR/TRAP heterodimer binding to DNA.

MATERIALS AND METHODS
HeLa Cytosolic Extract Preparation-HeLa or HeLa thymidine kinase-negative (tk-; kindly provided by Dr. M. Pfahl, La Jolla Cancer Research Foundation, La Jolla, CAI cells were grown in monolayer culture at 37 "C, 5% COz in Dulbecco's modified Eagle's medium with 10% fetal calf serum, 100 unitdm1 penicillin, and 100 pg/ml streptomycin. Cytosolic extracts were prepared as described previously (29).
In Vitro Phosphorylation of TR-Bacterially expressed full-length hTRP-l* (98% pure, 20 pmoV30 pl) was incubated with 10 pCi of [y-32PlATP, 50 p~ ATP, and 15 pg of HeLa cytosolic extract in the presence of 200 y sodium orthovanadate (Na,VO,) and 1 y okadaic acid in a final volume of 50 pl at 30 "C for 50 min as previously described (29). As a control, HeLa cytosolic extract preheated at 90 "C for 5 min was incubated under the same conditions. In the case of samples used for stoichiometry determination and phosphoamino acid analysis, the reaction was stopped with an equal volume of 10% (w/v) trichloroacetic acid, and 40 pl of 0.1% bovine serum albumin was added. Samples were processed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (10% acrylamide gel) and autoradiography as previously described (30). Stoichiometry was determined by excising J. W. Apriletti, manuscript in preparation.

Phosphorylation Increases TR Homodimer Binding to
DNA the phosphorylated hTRp-1 from the gel; solubilizing in 1:2:10 (v/v/v) of water, Soluene, and Ultimagold; and performing liquid scintillation spectrometry. Background was measured in a piece of gel without specifically enhanced radioactivity.
Western Blotting-After SDS-PAGE, proteins were transferred to polyvinylidene difluoride (Immobilon P, Millipore) in 25 m M Tris, 192 m glycine, and 20% (v/v) methanol for 120 min at 250 mA constant current. Membranes were dried, and autoradiography was performed.
Phosphwmino Acid Analysis-Phosphoamino acid analysis was performed by thin-layer chromatography as described (31). Briefly, after transfer to polyvinylidene difluoride, phosphorylated hTRp-1 was identified by autoradiography and hydrolyzed with 6 M HCl at 110 "C for 2 h. Phosphoamino acids were separated on cellulose plates in 7% (v/v) formic acid at 600 V for 7 h and identified by autoradiography. The position of migration of phosphoamino acids was determined by ninhydrin staining of standards (phosphoserine, phosphothreonine, and phosphotyrosine).
Immunoprecipitation of in Vitro Phosphorylated TRs-Polyclonal anti-TRp-1 antibody (32) was incubated with the same volume of immunizing peptide solution (1 pg/pl) or phosphate-buffered saline for 1 h at 4 "C. As a control, preimmune serum was also incubated with the same volume of phosphate-buffered saline. ARer centrifugation at 10,000 x g for 5 min, 10 pl of the supernatant was incubated with 10 pl of in vitro phosphorylated hTRp-1 and 30 pl of 1 x immunoprecipitation buffer (50 m Tris-HC1 (pH 7 3 , 150 m NaCl, 0.1% (v/v) Triton X-100, 0.1% (w/v) SDS, 10 rm NaF, 10 rm Na4P20,, 10 m EDTA, 10 m EGTA, 2 m NaaVO,) for 2 h at 4 "C. Slurries (100 p1) of 10% (v/v) protein A-Sepharose (Calbiochem) in 50 m Tris (pH 7.5),150 m NaCl, 10 mx NaF, 10 rm Na4P207 were added to each sample and mixed gently for 1 h at 4 "C. The samples were washed, boiled, and then analyzed by SDS-PAGE as previously described (32), except that samples were washed well with 50 m Tris-HC1 (pH 7.5), 150 m NaCl, 0.1% Triton X-100, and then with 50 m Tris-HC1 (pH 7 . 9 , 150 m M NaCl. DNA Binding AssaylEMSA-EMSA using F2 (chicken lysozyme TRE) as a probe was performed as described previously (7, 17). EMSA using TREpal (palindrome of two TRE half-sites), rMHCa (rat myosin heavy chain ( I TRE), or DR4 (direct repeat of two TRE half-sites spaced by 4 nucleotides) as probes were performed as previously described (19). In vitro phosphorylation of hTRP-1 for EMSA was performed using conditions identical to those described above, except that unlabeled ATP were performed, and the proteins were used for EMSA as previously described (17). Nuclear extracts of 235-1 pituitary cells used for EMSA were prepared as previously described (33). The rat liver TR auxiliary factor/protein(s) was obtained from fractions eluted out the high performance liquid chromatography sizing column during large scale purification of endogenous TRa (34). They did not contain endogenous TR activity, as evidenced by their lack of T3 (data not shown) and DNA ( Alkaline Phosphatase I).eatment-hTRp-l phosphorylated with unlabeled ATP was incubated with or without 36 units of alkaline phosphatase (Boehringer Mannheim) in a final volume of 10 pl at 37 "C for 30 min and was then used for EMSA.

RESULTS AND DISCUSSION
Silver staining of the unmodified bacterially expressed hTRP-1 revealed two TR forms (55 and 49 m a ) , with the latter probably due to translation from an internal methionine (Fig.  IA, lane 1 ). No bands corresponding to hTRp-1 were observed in HeLa cytosolic extract (Fig. lA, compare lane 2 with lane 1 ). We previously reported that co-incubation of HeLa nuclear extract with hTRP-1 formed a TRrIRAP heterodimer band that migrated more slowly than the TR homodimer band on EMSA (35). However, co-incubation with HeLa cytosolic extract did not form additional bands on EMSA, suggesting that the cytosolic extract did not contain any TRAP activity (data not shown). When hTRP-1 was phosphorylated with  Fig. 1C (lane 1 ). These bands were not observed when anti-TRP-1 antibody was preincubated with immunizing peptide (Fig. lC, lane 9) or when preimmune serum was used (Fig. lC, lane lo), indicating that these 32Plabeled doublet bands represent phosphorylated hTRP-1. When cytosolic extract was incubated without hTRP-1 (Fig. lC, lanes  2-41, or hTRP-1 was incubated with preheated cytosolic extract (Fig. lC, lanes 5-71 in the presence of [y-32PlATP, these bands were not observed. These data confirm that hTRP-1 phosphorylation is mediated by a kinase(s) in the HeLa cytosolic extract (as previously reported; Ref. 29) whose activity was inactivated by heating. 32P-Labeled doublet bands with apparent molecular sizes of approximately 95 kDa were also immunoprecipitated (Fig. lC, lane 8). The identity of these 95-kDa bands is unknown. It is possible that they are due to weak antibody cross-reactivity with low abundance, but highly phosphorylated, non-TR proteins contained in the bacterial and/or cytosolic extract (Fig. 1, B and C ) .
We next performed phosphoamino acid analyses of the bands corresponding to TR and the 95-kDa protein. As shown in Fig.   2 A , phosphorylation of serine, tyrosine, and threonine residues in TR was in a ratio of 90:3:7 (lane 1 ) and 80:4:16 (lane 2 ) in the absence and presence of okadaic acid and Na3V04, respectively. The ratio in the presence of okadaic acid and Na3V04 is similar to the ratio observed by Lin et al. (29). In contrast, the relative ratios of phosphoserine, phosphotyrosine, and phosphothreonine in the 95-kDa protein(s) were 61:2:37 ( Fig. 2 B , lane 1) and 55:3:42 ( Fig. 2 B , lane 2 ) in the absence and presence of okadaic acid and Na3V04, respectively. These data suggest that the 95-kDa protein is unrelated to TR. Determination of stoichiometry showed that -0.1 mol of phosphatdmol of hTRP-1 was incorporated. However, since cytosol is reported to contain large amounts of ATP (0.2-1.9 mM) (36,37), the actual stoichiometry is probably higher than 0.1. We next performed EMSA to study the binding of phosphorylated hTRp-1 (incubated with cytosolic extract in the presence of ATP) and unphosphorylated hTRP-1 (incubated with preheated cytosolic extract in the presence of ATP) to several TREs containing half-sites of different orientations. As shown in Fig. 3,panel A (lanes 2 and 3) andpanel B (lanes 2,4,12,14,  16, and 18), phosphorylation of hTRP-1 increased homodimer binding to F2 (inverted palindrome), rMHCa (direct repeat), DR4 (direct repeat), and TREpal (palindrome). In the latter case, unphosphorylated hTRp-1 bound mainly as monomer (Fig. 3B, lane 2 ) , and the monomer binding was not increased by phosphorylation of hTRP-1 (Fig. 3B, lane 4). Cytosolic extract alone showed no binding to F2 (Fig. 3A, lane 1 ).
We next treated phosphorylated hTRP-1 with alkaline phosphatase. As shown in Fig. 3C, phosphatase treatment decreased TR homodimer binding (lanes l and 2 ) to F2, confirming that hTRp-1 phosphorylation enhances homodimer binding to DNA. Similar results were also observed with another TRE, DR4 (data not shown). Comparison of silver-stained hTRP-1 samples treated with or without alkaline phosphatase showed no difference in the protein patterns, indicating that the phosphatase did not contain any protease activity (data not shown).
influence of phosphorylation on DNA binding appears to be selective for homodimers and, therefore, could affect TR function when homodimers are active. This might be more prominent in the absence of T3, since the ligand tends to disrupt homodimers, and unliganded TRs may repress basal transcription. However, since okadaic acid enhances T3-mediated transcriptional activation: it is possible that phosphorylated TR may activate transcription in its liganded state. In any case, our results provide evidence for another mechanism, in addition to ligand binding, by which homodimer binding to TREs is modulated. As such, it is possible that phosphorylation of TRs may participate in TR-mediated regulation of transcription.