A Fraction Enriched in a Novel Glucocorticoid Receptor-interacting Protein Stimulates Receptor-dependent Transcription in Vitro *

Glucocorticoids influence numerous cell functions by regulating gene activity. The glucocorticoid receptor (GR) is a ligand-activated transcription factor and, like any other transcription factor, does not modulate gene activity just by binding to DNA. Interaction with other proteins is probably required to enhance the establish-ment of a functional transcription initiation complex. To identify such proteins, we analyzed the in vitro interaction of the glucocorticoid receptor bound to a dou- ble glucocorticoid response element with nuclear proteins and describe here three interacting proteins with different molecular weights. One of them, which we named GRIP 170 (GR-interacting protein), was purified and microsequenced, and it turned out to be an unknown protein. When tested in a cell-free transcription assay, the fraction highly enriched for GRIP 170 does not influence basal promoter activity but does enhance GR induction. Glucocorticoids are active in inducing gluconeogenesis in liver, promote the development of various organs, cause apo-ptosis of lymphoid cells, and are necessary for the growth of many cell types in vitro . The glucocorticoid receptor (GR) 1 belongs to the superfamily of steroid hormone receptors and was the first transcription factor to 2 , 1 m M DTT, 1 m M EDTA, 1 m M EGTA, 10% (v/v) glycerol) for 90 min at a flow rate of 2 ml/min. The GR-binding activity of the fractions was assayed in cl-FSW experiments. Positive fractions containing GRIP 170 were pooled (in QS), diluted, adjusted to pH 6, and loaded on Mono S columns for cation exchange chromatography. The proteins were separated using a linear gradient from 0 M NaCl to 1 M NaCl in MS buffer (20 m M MES, pH 6.0, 20 m M KCl, 2 m M MgCl 2 , 1 m M DTT, 1 m M EDTA, 1 m M EGTA, 10% (v/v) glycerol) for 50 min at a flow rate of 0.8 ml/min. The fractions were assayed with the cl-FSW, and the fractions expressing GRIP 170 activ- ity were pooled, diluted, adjusted to pH 7.4, and applied on a heparin-Sepharose column. Protein separation was carried out with a linear gradient between 0 M NaCl and 0.5 M NaCl in HS buffer (25 m M Hepes, pH 7.6, 5 m M MgCl 2 , 1 m M EGTA, 10% (v/v) glycerol) for 80 min at a flow rate of 2 ml/min. This gradient was followed by a one-step elution with 1.5 M salt under the same conditions. Fractions containing GRIP 170, as detected by cl-FSW, were pooled, diluted, and concentrated on Mono Q column by anion exchange chromatography. Therefore, the proteins were eluted by applying a steep linear gradient of 0.1–0.8 M NaCl in QS buffer for 40 min at a flow rate of 1 ml/min. This resulted in three positive fractions containing GRIP 170 as by cl-FSW. Digestion, Reversed Phase HPLC, and Microsequencing— Protein digestion, reversed phase HPLC, and microsequencing of peptides

Glucocorticoids influence numerous cell functions by regulating gene activity. The glucocorticoid receptor (GR) is a ligand-activated transcription factor and, like any other transcription factor, does not modulate gene activity just by binding to DNA. Interaction with other proteins is probably required to enhance the establishment of a functional transcription initiation complex. To identify such proteins, we analyzed the in vitro interaction of the glucocorticoid receptor bound to a double glucocorticoid response element with nuclear proteins and describe here three interacting proteins with different molecular weights. One of them, which we named GRIP 170 (GR-interacting protein), was purified and microsequenced, and it turned out to be an unknown protein. When tested in a cell-free transcription assay, the fraction highly enriched for GRIP 170 does not influence basal promoter activity but does enhance GR induction.
Glucocorticoids are active in inducing gluconeogenesis in liver, promote the development of various organs, cause apoptosis of lymphoid cells, and are necessary for the growth of many cell types in vitro. The glucocorticoid receptor (GR) 1 belongs to the superfamily of steroid hormone receptors and was the first transcription factor to be isolated and studied in detail (1,2). Most glucocorticoid-regulated genes contain a short palindromic DNA sequence known as glucocorticoid response element (GRE). Depending on the type of GRE and on the promoter context, binding of GR to these elements can result in either gene induction or repression.
For regulating gene expression, the glucocorticoid receptor, like any other transcription factor, has to interact either directly with the transcription/initiation complex or indirectly via adapters or intermediary factors (3,4). Nuclear receptors have been demonstrated to interact with both types of factors. For some nuclear receptors, but not for GR, interaction with TF II B has been shown (5)(6)(7). The estrogen receptor was demonstrated to interact with TAF II 30, a component of the TF II D complex (8) and with transcription intermediary factor 1, a putative mediator of the ligand-dependent activation function (9). Other, unknown factors have been demonstrated to interact with nuclear receptors in vitro (10,11), or in vivo (12) or to enhance receptor-DNA binding in vitro (13). In addition, receptor interaction with proteins involved in alterations of chromatin structure or function, Swi3 and Spt6, has been shown (14,15).
We have been interested in analyzing mechanisms for a synergistically increased steroid response mediated by multiple binding sites for different or identical transcription factors. Cooperative DNA binding has been shown for the glucocorticoid and the progesterone receptor in the case of duplicated binding sites (16 -19). GR can also synergize with various transcription factors (20). Synergistic activation with the ubiquitous transcription factor OTF-1/Oct-1 is found in the context of the MMTV promoter, and this effect seems to be mediated by cooperative DNA binding (21).
Replacement of the DNA binding domain of GR with the GAL4-DNA binding domain abolished cooperative binding to duplicated upstream activating sequence elements in contrast to functional synergy, which could still be observed (19). Functional synergy may involve two transcription factors simultaneously (and cooperatively) touching a target protein. Therefore, we designed a strategy for the search of GR-interacting factors, which upon isolation should enhance GR mediated in vitro transcription. Our strategy is based on the use of DNAbound GRs to a double GRE to avoid interactions irrelevant to transcriptional control and to possibly increase the binding affinity to the target protein. Using this procedure we identified three GR-interacting proteins (GRIPs).

EXPERIMENTAL PROCEDURES
Materials-Oligonucleotides 53/54 and 51/52 containing two GREs or a single GRE, respectively (20) were purchased from Eurogentec (Belgium). Endoproteinase Lys-C was from Boehringer Mannheim; [␣-32 P]dATP was from Amersham Corp. Fast protein liquid chromatography system and columns for protein separation were used from Pharmacia Biotech Inc. Sykam HPLC (Mü nchen, Germany) was used for peptide separation. HPLC columns and all solvents for automatic protein sequencing were from Applied Biosystems (Weiterstadt, Germany). PVDF Immobilon membrane was from Millipore (Eschborn, Germany). Dexamethasone was obtained from Sigma. All other chemicals were purchased from Merck.
Cell Culture-Human HeLa S 3 -cells were grown in spinner flasks in Dulbecco's modified Eagle's medium supplemented with 10% newborn calf serum, sodium pyrovate, glutamine, nonessential amino acids, vitamins, and antibiotics. Dexamethasone was added 45 min prior harvesting to a final concentration of 10 Ϫ7 M at a density of 8 ϫ 10 5 cells/ml. Preparation of Nuclear Extracts for Protein Purification-Nuclear extracts were prepared by the method of Shapiro et al. (22) with minor modifications. The ammoniumsulfate precipitated proteins were resuspended in buffer D (20 mM Hepes, pH 7.9, 100 mM KCl, 0.2 mM EDTA, 1 mM DTT, 20% (v/v) glycerol) (23). After extended dialyzation against buffer D the dialysate was centrifuged at 100,000 ϫ g for 30 min at 4°C, and the supernatant was stored at Ϫ80°C.
Electrophoresis and Electroblotting-Crude nuclear protein extracts, bacterial protein extracts, or chromatographically purified fractions were precipitated as described elsewhere (24) and separated on a 7% SDS-PAGE according to the method of Laemmli (25).
Electroblotting was carried out for 2.75 h in a semidry blotting system with a constant current of 1 mA/cm 2 gel using STGM buffer (0.06% SDS, 12.5 mM Tris, 50 mM glycine, 5% methanol) as electrode buffer.
Electrophoretic Mobility Shift Assay (EMSA) and Western Blotting Analysis-EMSA was performed according to SivaRaman et al. (26). The double-stranded oligonucleotide containing a single GRE was labeled with [␣-32 P]ATP using Klenow polymerase (Boehringer Mannheim) according to the supplier's directions.
The Western blot analysis was carried out as follows. After separation of nuclear proteins by SDS-PAGE using a 7% polyacrylamide gel, the proteins were electroblotted to a PVDF membrane (Immobilon P, Millipore). After blocking, membranes were incubated overnight either with a 1:100 dilution of a polyclonal serum raised against hGR, or with a 1:200 dilution of an affinity-purified polyclonal antibody raised against Hbrm. Purified goat anti rabbit antibody was used as secondary antibody (Aurion, Germany). The immune complexes were visualized by the Immuno Gold detection kit according to the manufacturer (Aurion, Germany).
Renaturation and Blocking-For renaturation the PVDF membrane was transferred to HEMN buffer (20 mM Hepes, pH 7.4, 0.5 mM EDTA, 5 mM MgCl 2 , 150 mM NaCl) plus 0.1% Nonidet P-40 and 3 mM DTT, containing 6 M guanidine hydrochloride for denaturation. The guanidine hydrochloride was diluted out step-wise by adding same buffer without guanidine hydrochloride, allowing the proteins to partially renature. After renaturation the PVDF membrane was blocked at 4°C overnight in HEMN buffer ϩ 5% nonfat dry milk (Carnation).
cl-FSW Assay-The cross-linked far Southwestern analysis was performed as follows. 10 g of partially purified GR fraction was incubated with 0.5 pmol of 32 P-labeled double GRE oligonucleotide (dGRE) in the presence of 0.5-1 g of poly(dI-dC) for 10 min at 4°C in binding buffer (10 mM Tris, pH 7.5, 60 mM KCl, 2.5 mM MgCl 2 , 1 mM DTT, 1 mM EDTA, 5% (v/v) glycerol), followed by UV-cross-link at 286 nm for 5 min at room temperature. The cross-linked mixture had been diluted in 3 ml of HEMN buffer containing 1 g/ml poly(dI-dC), 0.1 mM dexamethasone, 2 mM ATP, and 3 mM DTT. This solution was used as a probe for the renaturated proteins on the milk powder-blocked PVDF membrane. The incubation was carried out for 3 h under gentle shaking at room temperature in a plastic bag. Afterwards the PVDF membrane was washed twice for 10 min in HEMN buffer containing 0.1% Tween 20 and 3 mM DTT. The dry membrane was autoradiographed overnight.
Protein Purification-Crude nuclear protein extracts were separated first on anion exchange chromatography by loading 100 -150 mg of protein on preparative Q-Sepharose column. The proteins were eluted by applying a linear gradient of 0.1 M NaCl to 1 M NaCl in QS buffer (20 mM Tris, pH 7.4, 20 mM KCl, 2 mM MgCl 2 , 1 mM DTT, 1 mM EDTA, 1 mM EGTA, 10% (v/v) glycerol) for 90 min at a flow rate of 2 ml/min. The GR-binding activity of the fractions was assayed in cl-FSW experiments. Positive fractions containing GRIP 170 were pooled (in QS), diluted, adjusted to pH 6, and loaded on Mono S columns for cation exchange chromatography. The proteins were separated using a linear gradient from 0 M NaCl to 1 M NaCl in MS buffer (20 mM MES, pH 6.0, 20 mM KCl, 2 mM MgCl 2 , 1 mM DTT, 1 mM EDTA, 1 mM EGTA, 10% (v/v) glycerol) for 50 min at a flow rate of 0.8 ml/min. The fractions were assayed with the cl-FSW, and the fractions expressing GRIP 170 activity were pooled, diluted, adjusted to pH 7.4, and applied on a heparin-Sepharose column. Protein separation was carried out with a linear gradient between 0 M NaCl and 0.5 M NaCl in HS buffer (25 mM Hepes, pH 7.6, 5 mM MgCl 2 , 1 mM EGTA, 10% (v/v) glycerol) for 80 min at a flow rate of 2 ml/min. This gradient was followed by a one-step elution with 1.5 M salt under the same conditions. Fractions containing GRIP 170, as detected by cl-FSW, were pooled, diluted, and concentrated on Mono Q column by anion exchange chromatography. Therefore, the proteins were eluted by applying a steep linear gradient of 0.1-0.8 M NaCl in QS buffer for 40 min at a flow rate of 1 ml/min. This resulted in three positive fractions containing GRIP 170 as shown by cl-FSW.
Protein Digestion, Reversed Phase HPLC, and Microsequencing-Protein digestion, reversed phase HPLC, and microsequencing of peptides were carried out as described by Eggert et al. (27). N-terminal microsequencing was performed by running a preparative SDS-PAGE. The gel was blotted to Immobilon membrane, which was stained with Coomassie Blue solution (0.2% Coomassie Blue, 0.5% acetic acid, 20% (v/v) methanol) and washed extensively (30% (v/v) methanol). The GRIP 170 bands were excised and analyzed by microsequencing.
In Vitro Transcription-HeLa cell nuclear extract was prepared essentially after the method of Dignam et al. (23). Recombinant fulllength human GR was expressed in Spodoptera frugiperda Sf9-cells by using the baculovirus expression system and isolated as described by Srinivasan and Thompson (28). DNA templates were derived from p(C 2 AT) 19 and contain either the full-length MMTV promoter from Ϫ240 comprising the HREs (240/380) linked to a 380-bp-long G-less cassette or a truncated MMTV promoter from Ϫ80 lacking the HREs (80/280) fused to a 280-bp-long cassette, serving as internal control. In vitro transcription was carried out as described (29), except that the final polyethylene glycol concentration was at 1% and poly(dI-dC) was used as competitor at a final concentration of 19 ng/l.
Template DNAs were preincubated with GR, GRIP 170, and in each case with an equal amount of the GRIP 170 buffer for 10 min on ice in a volume of 12.5 l. Then HeLa cell nuclear extract (ϳ40 g), nucleotides were added to make a final volume of 25 l, and transcription was carried out for 45 min at 30°C. Transcripts were treated as described (29), separated by 6.5% denaturing polyacrylamide gel electrophoresis, and visualized by autoradiography. Quantitation was made with a PhosphorImager (Molecular Dynamics), and relative transcription was determined with respect to the transcription of the control template.

Identification of Proteins
Interacting with DNA-bound GR-To identify proteins that interact with GR molecules bound to two adjacent GRE sequences, we generated a probe consisting of a 32 P-labeled dGRE UV-cross-linked to GR molecules. For this purpose we enriched GR from HeLa nuclear extracts by Q Sepharose fractionation. This results in a 30-fold enriched GR-fraction compared with the crude extract. The 97-kDa hGR was identified by Western blotting with a rabbit polyclonal antibody directed against a synthetic hGR peptide (amino acids 346 -367). This fraction generated a strong band shift in an electrophoretic mobility shift assay (not shown), and GR could be UV-cross-linked to 32 P-labeled GRE sequences. Binding and UV-cross-linking was GRE-specific since it was competable with unlabeled GRE DNA (not shown). This GR⅐ 32 P-dGRE complex was used as a probe against HeLa nuclear proteins or, as a control, against Escherichia coli proteins, which were separated by SDS-PAGE and blotted on an Immobilon membrane. Fig. 1 shows an autoradiograph of such an experiment. Four bands migrating according to molecular masses of about 95, 115, 120, and 170 kDa can be detected. The 120-kDa band is difficult to resolve from the 115-kDa band, but extract fractionation (see below) clearly identifies and separates the 115-and 120-kDa bands. The 115-kDa band is seen in Southwestern experiments with the labeled GRE in the absence of proteins in the probe (Fig. 1, lane 10) and with almost any DNA sequence tested (not shown) and, therefore, seems to be an unspecific DNA-binding protein. The other bands, however, were specific as demonstrated by the fact that (i) bacterial proteins on the blot are not visualized (Fig. 1, odd-numbered  lanes), (ii) specific GRE competition during probe cross-linking results in reduced 170-, 120-, and 95-kDa bands (Fig. 1, lane 4), (iii) whole nuclear extract, which contains low amounts of GR, labels the 110-and 95-kDa bands only, and (iv) bacterial extract cross-linked to DNA visualizes a single band (115 kDa), which is seen with the protein-free DNA probe alone (Fig. 1,  lanes 8 and 10). Since the 170-kDa band most clearly showed specificity for the GR⅐ 32 P-dGRE complex, we analyzed this protein.
Although from our competition tests in the cross-linking reaction we knew that HeLa-GR was bound to the double GRE, we wanted to confirm this result with a glucocorticoid receptor from another source. We used human GR expressed with a baculovirus vector in insect cells. After purification by DNA affinity column (28) we used this protein to generate a GR⅐ 32 P-GRE probe. As seen in Fig. 2A the baculovirus-expressed GR identifies nearly the same bands as the HeLa-GR including the 170-kDa band.
Characterization and Purification of GRIP 170 -Previously, a protein in the range of 170 kDa has been shown to enhance the GR-mediated gene induction. This protein, Swi2 in yeast or Hbrm in humans, with a molecular mass of 180 kDa (30) is not the 170-kDa protein found here, since a Western blot with an antibody against Hbrm clearly labels a band migrating at a molecular mass position close to 200 kDa (Fig. 2B). Similarly, it was ruled out that one of the bands is GR, which is detectable with an antibody (Fig. 2B) but which elutes from Q-Sepharose with low salt (up to 220 mM), whereas the proteins labeled with the cross-linked GR⅐dGRE probe eluted at higher salt concentrations (380 -530 mM; see below). Since the 170-kDa protein is identified by interaction with GR we refer to it as the "glucocorticoid receptor interacting protein 170" (GRIP 170). In order to further substantiate the specificity of the cross-linked probe we added a 25-fold excess of unlabeled cross-linked GR⅐dGRE to the incubation (Fig. 2C, lanes 1 and 3). Clearly, binding to the blotted proteins is inhibited. This was seen for unfractionated nuclear extract as well as for a Q-Sepharose fraction enriched for GRIP 170 (Fig. 2C, lanes 2 and 4). Since binding of the probe to GRIP 170 depends on the presence of GR, we wondered whether this binding requires GR bound to a double GRE or whether GR cross-linked to a single GRE (sGRE) might bind as well. Fig. 2C, lanes 5 and 6, shows the use of a GR⅐ 32 P-sGRE complex with identical specific radioactive activity as the GR⅐ 32 P-dGRE complex (Fig. 2C, lanes 1 and 2). Binding of the sGRE probe to GRIP 170 is not detectable; thus, efficient interaction with GRIP 170 requires GR bound to a double GRE.
Toward the isolation and characterization of GRIP 170 we fractionated HeLa nuclear extract (Fig. 3) by Q-Sepharose chromatography and assayed for the presence of GRIP 170 with the HeLa GR⅐ 32 P-dGRE probe. In contrast to the unfractionated nuclear extract, pooled QS fractions, eluted with 390 -540 mM salt, contained only three out of the four labeled bands, GRIP 170, the 120-kDa protein, and the 95-kDa protein (Fig.  4A). This pool from 2 g of nuclear extract was further purified on a Mono S column, allowing the isolation of GRIP 170 containing fractions eluting between 350 and 450 mM salt. This cation exchange chromatography led to the separation of GRIP 170 from the two other interacting factors, since they show longer retention times (not shown). Further purification was achieved with heparin-Sepharose where GRIP 170 could not be eluted in a linear salt gradient between 0 and 500 mM salt. The following one-step elution at 1.5 M salt revealed the GRIP 170 activity. GRIP 170 was recovered in two fractions, which were diluted to a final salt concentration of 150 mM NaCl and concentrated on Mono Q column. The anion exchange chromatography leads to three positive fractions eluting between 400 and 475 mM salt, respectively. Chromatography by these four columns resulted in about 1000-fold purification of GRIP 170. The final purification was achieved by SDS-PAGE and subsequent blotting on a PVDF membrane. Fig. 4B shows the pattern of protein bands after Coomassie staining (lane 1). In parallel an aliquot was blotted after SDS-PAGE and incubated with baculovirus-expressed GR after UV-cross-linking to a 32 P-dGRE oligonucleotide. Since the 170-kDa band was labeled (Fig. 4B,  lane 2), we excised and analyzed the corresponding band on the PVDF membrane by N-terminal microsequencing. Addition-  ally, a preparative SDS-PAGE was run, and after Coomassie staining the 170-kDa band was excised and digested by endoproteinase Lys-C. The resulting peptides were purified by reverse phase HPLC and sequenced in addition to the full-length protein. Three different sequences could be read and compared with the Swiss Prot computer data base. No homology to known protein sequences could be detected.
A Fraction Highly Enriched for GRIP 170 Stimulates GRinduced Transcription in Vitro-Since enhancement of receptor binding to DNA by receptor-interacting proteins has been demonstrated (13), we tested the purified GRIP 170 fraction in an EMSA with baculovirus-expressed GR (Fig. 4C). The addition of GRIP 170 neither causes a retarded band nor changes the intensity or position of the GR/DNA complex.
Another step toward the functional analysis of a protein interacting with a transcription factor would be the use of an in vitro transcription system. Therefore, we prepared nuclear extracts from HeLa cells (23) that efficiently mediate induction by purified glucocorticoid receptor (29). Human GR was expressed in insect cells with a baculovirus expression system and purified (28). DNA templates used contain the full-length MMTV promoter up to position Ϫ240 bp of the transcriptional start site. This promoter region encompasses the natural GRE sequences (31) and was fused to a G-less cassette of 380 bp in length. As an internal control the reaction contained a second template with a truncated MMTV promoter, lacking the GRE sequences but containing the nuclear factor I and octamer transcription factor I binding sites. This promoter up to Ϫ80 bp was fused to a G-less cassette of 280 bp in length and provides a standard for basal gene transcription. In vitro transcription in the presence of radioactive nucleotides yielded a specific transcript for both templates (Fig. 5A). The upper band in each pair of bands represents read-through transcripts, which are almost not affected by the presence of GR or GRIP 170 fraction, whereas the lower band represents the correctly initiated transcripts for both DNA templates and is enhanced in the presence of GR with the full-length MMTV template. PhosphorImager quantitation was used to determine the relative transcription activity. Addition of the GRIP 170 fraction to the basal transcription reaction results, if in anything, in a marginal increase in transcription on both promoters (Fig. 5, A, lanes 2 and 4, and  B). Similarly, no effect was seen with an adenovirus major late promoter as template (data not shown). Addition of baculovirus-expressed GR generates a transcriptional induction as expected (29). The combined addition of GR and GRIP 170 fraction shows a synergistic increase in transcription (Fig. 5, A,  lane 6, and B). This induction very likely was mediated by GRIP 170 because all of the reactions contained the same amount of GRIP 170 buffer, since a different fraction not containing GRIP 170 did not change transcription efficiency, and finally GRIP 170 "shoulder fractions" are less effective as the peak fraction (Table I). DISCUSSION Here we identified three proteins (170, 120, and 95 kDa) that interact with a specifically designed glucocorticoid receptor probe. The probe consists of a 32 P-labeled DNA harboring a double GRE with UV-cross-linked GR molecules. This probe serves several requirements. First, the GR can be easily and efficiently labeled. Second, the double GRE arrangement provides several DNA bound GR molecules, which increase the binding affinity to the target protein. This assumption is based on the finding that transactivation is stronger with two synergizing GREs as compared with individual GRE sequences and is confirmed by the failure of the GR-32 P-sGRE probe to detect GRIP 170 on a blot. The GR⅐ 32 P-dGRE probe also avoids detection of proteins interacting with the inactive, DNA-unbound GR such as hsp 90, p59, and others (1,2,32). In addition, blotted HeLa GR was not detected, which may be explained by an incorrect folding after denaturation and renaturation or by the occupied dimerization domain of the dimerized DNA-bound GR molecules.
Receptor interaction and transcriptional activation has been shown for factors comprising the TATA box-binding protein and its associated factors and TF II B complexes (5,8). There is no direct proof that the GRIP 170 factor, binding to the GR⅐ 32 P-dGRE probe, is identical to the factor within the GRIP 170 fraction causing the increase in GR-mediated in vitro transcription. Nevertheless, the co-purification of these two functions via four different columns and the reduced activity of GRIP 170 "shoulder fractions" argue for the transcriptional activity of GRIP 170. From the fact that the transcription extract is prepared from HeLa nuclei one would predict that GRIP 170 should be present and that the addition of GRIP 170 should not change the transcriptional activity. One explanation might be that GRIP 170 is limiting in the presence of added GR and, therefore, that additional amounts of GRIP 170 optimize the GR-mediated transcription. Comparison of blotted protein from nuclear extracts prepared from glucocorticoid-treated or from untreated cells did not reveal any hormone-dependent change in GRIP binding intensities (not shown).
Our sequence analysis rules out the possibility that GRIP 170 might belong to the TAF or TF II group as far as these are known. Similarly, we would rule out that GRIP 170 belongs to the group of factors associated with the inactive GR, such as hsp 90, hsp 59, or hsp 70 (33)(34)(35)(36). Receptor interaction and/or potentiation of receptor function has been shown for proteins involved in alterations of chromatin structure or function (14,15,30,37). Again, the GRIP 170 sequence is different from these proteins (Swi, Snf, or Spt6).
Other factors have been shown to increase the receptor binding affinity to DNA by in vitro binding tests. Glucocorticoid and androgen receptors require RAF 130 for optimal DNA binding (13); for the estrogen receptor hsp 70 and 48-and 45-kDa proteins increase the DNA association rate (12); and in the case of GR low molecular weight factors have been seen with a similar function (38). Such an activity was not seen with GR and GRIP 170.
Receptor-interacting proteins with molecular weight comparable with that of GRIP 170 have been identified in conjunction with the estrogen receptor and named ERAP 160 or RIP 160 (10,11). ERAP 160 and RIP 160 may be identical and bind to the estrogen receptor only in the presence of estrogen. For GR such a ligand-dependent binding would not be detectable, since GR will be activated through the purification procedure (39,40). Thus, based on the molecular weight, GRIP 170 could be similar or identical to ERAP 160/RIP 160. For the GRIP 170 fraction we found a GR-dependent transcriptional increase in the in vitro transcription assay, an activity not identified for previously described receptor-associated or interacting factors (10 -13). Although in vitro binding to GR depends on the presence of two GREs, it remains to be shown, whether GRIP 170 is required to mediate synergy or to mediate the intrinsic induction from a single GRE in vivo.

TABLE I GRIP 170-dependent increase of GR-mediated transcriptional induction in vitro
For experimental details see also the legend to Fig. 5. Control experiments in GR mediated in vitro transcription using the GRIP 170 peak fraction (peak) and "shoulder" fractions eluting with less salt (shoulder low) or with higher salt (shoulder high) from Mono Q as compared with the peak fraction. A fraction eluting early from Mono Q, which does not contain GRIP 170 is used as well ("no GRIP" fraction).