Two independent pathways for transcription from the MMTV promoter

ft. Steroid Biochem. Molec. Biol. Vol. 51, No. 1/2, pp. 21-32, 1994 Pergamon

0960-0760(94)00105-7

Copyright © 1994 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0960-0760/94 $7.00 + 0.00

Two Independent Pathways for Transcription from the MMTV P r o m o t e r C h r i s t i a n C. M 6 w s , 1 T h o m a s P r e i s s , 1. E m i l y P. S l a t e r , 1 X i n a n C a o , 1 C. P e t e r V e r r i j z e r , 2 P e t e r C. V a n D e r V l i e t 2 a n d M i g u e l B e a t o it llnstitut fiir Molekularbiologie und Tumorforschung (IM T), Phillips Universit~t, Emil-Mannkopff-Str. 2, D-35037 Marburg, Germany and 2Laboratorium voor Fysiologische Chemic, Rijksuniversiteit te Utrecht, Vondellaan 24a, 3521 GG Utrecht, The Netherlands The influence o f p r o g e s t e r o n e r e c e p t o r (PR) a n d g l u c o c o r t i c o i d r e c e p t o r (GR) on t r a n s c r i p t i o n f r o m the m o u s e m a m m a r y t u m o u r virus (MMTV) p r o m o t e r was a n a l y z e d using cell-free t r a n s c r i p t i o n o f DNA t e m p l a t e s with a G - f r e e cassette. P r e i n c u b a t i o n of the t e m p l a t e s w i t h e i t h e r P R or GR s t i m u l a t e s the r a t e o f t r a n s c r i p t i o n i n i t i a t i o n 10-50 fold, w h e r e a s the r e c o m b i n a n t DNA b i n d i n g d o m a i n o f GR is inactive. M u t a t i o n s t h a t i n a c t i v a t e the n u c l e a r f a c t o r I (NFI) b i n d i n g site, or NFI d e p l e t i o n o f the n u c l e a r e x t r a c t , decrease basal t r a n s c r i p t i o n w i t h o u t influencing r e c e p t o r - d e p e n d e n t i n d u c t i o n . R e c o m b i n a n t NFI, b u t not its D N A - b i n d i n g d o m a i n , r e s t o r e s efficient basal t r a n s c r i p t i o n o f the d e p l e t e d e x t r a c t . R e c o m b i n a n t OTF1 or OTF2, b u t not the P O U d o m a i n of OTF1, e n h a n c e M M T V t r a n s c r i p t i o n i n d e p e n d e n t l y o f NF1. In a g r e e m e n t with this finding, NFI a n d OTF1 do not cooperate, b u t r a t h e r c o m p e t e for b i n d i n g to the wild type M M T V p r o m o t e r , t h o u g h t h e y have the p o t e n t i a l to b i n d s i m u l t a n e o u s l y to p r o p e r l y o r i e n t e d sites. O u r results i m p l y the existence of two i n d e p e n d e n t p a t h w a y s for M M T V t r a n s c r i p t i o n : one i n i t i a t e d by NFI a n d the o t h e r d e p e n d e n t on o c t a m e r t r a n s c r i p t i o n factors. Only the second p a t h w a y is s t i m u l a t e d by s t e r o i d h o r m o n e r e c e p t o r s in vitro.

ft. Steroid Biochem. Molec. Biol., Vol. 51, No. 1/2, pp. 21-32, 1994

INTRODUCTION

br could be mediated by alterations in chromatin structure [2]. In cells carrying stably integrated copies of the mouse mammary tumor virus (MMTV) DNA, expression from the viral promoter is strictly dependent on glucocorticoid or progesterone treatment and this behaviour can be reproduced by transient transfection experiments [3,4]. Induction by both hormones is mediated by the interaction of their receptors with a set of four imperfect palindromic HREs located between - 190 and - 75 upstream of the transcription start site [4--6]. In minichromosomes the M M T V - L T R region is organized in nucleosomes precisely positioned along the DNA sequence with a nucleosome covering the HREs [7]. Though direct evidence for receptors bound to the HREs of M M T V in the intact cells has been elusive, it is generally accepted that such an interaction takes place after hormonal treatment and that it is essential for induction. Induction is accompanied by changes in chromatin structure which becomes hypersensitive to MPE and DNaseI over the HREs [7]. Concurrently, other transcription factors bind to the M M T V promoter downstream of the HREs [8].

Steroid receptors are ligand inducible eukaryotic transregulators that can activate or repress transcription of specific genes. In most cases hormone induction is mediated by binding of the hormone receptors to well-defined DNA sequences called hormone responsive elements (HREs) [ 1]. Though the molecular details of the interaction between receptors and HRE sequences have been studied extensively, the mechanisms by which binding of the receptors to the HRE brings about changes in the transcriptional efficiency of an adjacent promoter remain unclear. Probably cooperation with other transcription factors or with the basal transcriptional machinery is required. This cooperation could occur through protein-protein interactions, direct or indirect, *Present address: University of Newcastle upon Tyne, School of Neuroscience, Division of Clinical Neuroscience, The Medical School, Framlington Place, Newcastle upon Tyne, NE2 4HH,

U.K. tCorrespondence to M. Beato. Received 17 Mar. 1994; accepted 31 May 1994. 21

22

Christian C. M6ws et al.

Immediately adjacent to the HREs is a palindromic site with affinity for the transcription factor N F I [9]. T h o u g h it has not been formally proven that N F I is one of the factors bound to the M M T V promoter after hormonal induction of intact cells, the footprints observed in this region with the purified N F I are very similar to those found in vivo [8, 10] and N F I acts as a transcription factor on the M M T V promoter [11]. However, we have not found synergistic binding of hormone receptors and N F I to the M M T V promoter [11]. Moreover, in a cell-free transcription assay, purified progesterone receptor enhances transcription from the M M T V promoter in the presence of a competitor oligonucleotide containing the consensus sequence for N F I binding and is also able to activate transcription of a M M T V template lacking the N F I binding site [12]. These findings and the regular structure of M M T V chromatin led to the notion that hormonal induction could be mediated partly by facilitation of N F I access to the nucleosomaly organized M M T V promoter [1, 8, 13]. In addition to N F I , other transcription factors can also mediate hormonal induction of the M M T V promoter. Between the N F I binding site and the T A T A box there are two degenerate octamer motifs that are recognized by the transcription factor O T F 1/Oct 1 [14]. Mutation of these motifs drastically impairs the response to either glucocorticoids or progesterone in transfection assays [11, 14, 15]. U n d e r cell-free conditions binding of the purified hormone receptors to the H R E s facilitates the interaction of purified O T F 1 with the octamer motifs [14, 16]. This synergistic D N A binding is likely to be responsible for the enhancement of transcription from the M M T V promoter observed in the presence of purified progesterone receptor, since induction can be inhibited by addition of competitor oligonucleotide with the octamer consensus sequence or by mutations in the octamer motifs [14]. How O T F 1 recruits the general transcriptional machinery is not completely clear, although it has been shown to mediate the effects of cellular and viral transactivators [17-20].

T o further our understanding of the interplay of various transcription factors on the M M T V promoter we have used purified proteins and recombinant variants of N F I and octamer transcription factors in a cell-free system derived from H e L a cell nuclei [21], that efficiently mediates induction by purified progesterone [12] or glucocorticoid receptor (GR). By depleting the extract from N F I activity we show that the complete N F I , but not its D N A binding domain, acts as a basal transcription factor on the naked M M T V promoter. Induction by either hormone receptor in vitro is independent of N F I . Under cell-free conditions both purified O T F 1 and O T F 2 added to the depleted extract activate transcription independently of N F I . T h e r e is no indication for synergistic binding of O T F and N F I to the wild type template. T h e cell-free assay reported here should be useful for dissecting the functional interactions between the various factors on the M M T V promoter.

EXPERIMENTAL DNA

Template-plasmids for the cell-free transcription were derived from p(C2AT)19 [22] and contained stretches of the G R - M M T V or the Adenovirus Major Late ( A d M L ) - p r o m o t e r fused upstream to a region encoding a G-free transcript of defined length. T h e indicated mutations in the factor binding sites were derived from sequences described [14]. A schematic representation of the constructions is given in Fig. I(A). Oligonucleotides have the following sequences: N F I and " O c t c " as described [11]; ENFB, AATTCCTTTT TTTGGATTGA AGCC A A T C G G A T A A T G A G G ; (Octc)2, G A T C C T A T GA T T T G C A T A A G C A T T T G C A T A A G A C T A . T h e GNO-oligonucleotide contains the M M T V sequences from - 8 3 to - 3 2 with 9 additional nucleotides added to the 5'-end to generate a perfect H R E palindrome.

[Fig. 1(.4-C) Opposite] Fig. 1. P R a n d GR induce transcription f r o m the M M T V promoter. (A) DNA t e m p l a t e s . S c h e m a t i c representation of the M M T V t e m p l a t e p l a s m i d s . The n u m b e r s indicate the distance in nueleotides u p s t r e a m o f the correct start site ( + 1). The binding sites for k n o w n transcription factors, h o r m o n e receptor (HR), NFI, OTF a n d TFIID, are boxed and the transcript lengths are given for correct initiation events. A 20 nucleotides longer satellite b a n d originates f r o m r e a d - t h r o u g h transcripts starting u p s t r e a m o f the G-free cassette. (B) Analysis of cell-free transcripts. Cell-free transcription reactions were p e r f o r m e d w i t h 25 ng o f the h o r m o n e responsive p l a s m i d p M M T V 240/380 ("360nt") a n d 12.5 ng o f the control p l a s m i d pMMTV80/280 ("260nt") as t e m p l a t e s , according to the indicated schedule (top). P R (6 ng) or GR (72 ng) were added 15 rain prior to the start of transcription and incubation with the t e m p l a t e DNA was p e r f o r m e d in the absence (lanes 1 a n d 3) or in the presence o f 8% P E G (lanes 2, 4 a n d 5) for 15 rain at 250°C. The strong signal corresponds to correctly initiated transcripts and the weaker signal to r e a d - t h r o u g h transcripts. (C) D e p e n d e n c e of specific transcription on receptor concentration. Increasing a m o u n t s o f GR or P R were added to analogous transcription reactions as described in Fig. I(B) (data not shown). For quantitation, densitometric data of the specific h o r m o n e responsive transcript (360nt) were n o r m a l i z e d against the internal control (260 nt) to exclude variations during the purification of the samples. The obtained values of induction (ratio of n o r m a l i z e d level o f the 360 n t transcript to a control reaction w i t h o u t r e c e p t o r ) are d i s p l a y e d for P R and for GR against the amount of protein.

Cell-free Transcription from the M M T V Promoter

23

A) pMMTV 2 4 0 / 3 8 0

L--:

HR

.

.

.

NF1

.

.

O TF

.

TFIID I

t r a n s c r i p t length :

.

.

~

.

.

.

360nt MMTV

pMM-FV 80•280

G-free

-so

-20

+1 260 nt

MMTV

B)

G-free

..,S

rac ~nin

1251 ! -15

!

30 I 0

I 45

PR 1

4-

PEG

GR

q ='

-.

I

|

i

+

4-

360 nt

260 nt

1

2

3

4

5

c) 6O 50

30 ~ 2010-

0

I

i

i

i

I0

20

30

40

•

1

i

|

50

60

70

receptor/ng Fig. l(A--C)-legend opposite.

24

Christian C. M6ws et al.

Preparation of HeLa-cell nuclear extract (FINE) and the transcription factors HeLa cell nuclear extracts were prepared essentially as described [21]. The recombinant transcription factors vNF1; the DNA-binding domain of NF1, vNF1DBD (aa 4-240); vOTF1; the POU-domain, v O T F 1 - P O U (aa 1-23 and 269-440); and vOTF2, were prepared via a vaccinia virus expression system as described [23,24]. For some experiments n O T F I , purified from Namalwa cells was used [25]. T h e purification of v N F I - D B D and O T F 1 - P O U domain was as described [26]. Recombinant N F I was purified essentially as described for endogenous HeLa N F I [27]. The PR was purified from rabbit uteri [4] and the GR from rat liver [28]. T he DNA-binding domain of GR (aa 432-523) was obtained from a Staphylococcus aureus expression system as described [29].

DNA-affinity matrix preparation and H N E depletion Specific DNA-affinity resin was obtained by coupling of double-stranded oligonucleotides containing the relevant factor recognition sites to CNBr-activated sepharose 4B (Pharmacia) as described [30], with some modifications. Inactivation of surplus coupling sites was performed with 1M Tris buffer and the coupling efficiency was monitored by spectrophotometric determination of the DNA content in the various buffers. Typically, a resin contained approx. 13 fmol oligonucleotide per #1 of settled volume. For depletion, HeLa cell nuclear extracts were inverted with 1/4 to 1/2 of their volume of DNA-affinity resin at 4°C for 10-14 h. T h e supernatant was then aliquoted and refrozen. T he resin was inverted for 1 h with anologous buffer containing 5 0 0 mM NaC1 to elute the bound material. Depleted and complete H N E as well as the eluate were assayed for factor content by Gel Mobility Shift Analysis as given below.

Gel Mobility Shift Analysis H N E depletion and DNA-binding activity in the various factor preparations were determined by gel mobility shift experiments essentially as described [29] NF1- and OTF-binding was assayed in a final volume of 22 pl of reaction mixture based on H N E buffer C [21] supplemented with 100ng/#l poly-dI • dC, 3 mg/ml BSA, 0.05% NP 40 and 100 mM (Na/K)C1. All steps were performed at 4°C. T he protein was preincubated with the competitor for 10min (20/,1) followed by the addition of [32P]-end labeled specific oligonucleotide (10,000 cpm in 1/11) and further incubation for 15 rain. Complexes were resolved by gel electrophoresis, visualized by autoradiography and quantitated by densitometry. Th e binding assays with the GNO-oligonucleotide were performed under the same conditions except that NP40 was omitted.

Cell free transcription The assay follows the original description [22] with few modifications. A standard reaction mixture (final volume: 25 ~1) contained 16.3 mM HEPES, 8 mM Tris (pH 7.9 at 4°C), 6 0 - 8 0 m M (Na/K)C1, 7 . 3 m M MgC12, 0.44 mM EDTA, 0 . 8 m M ( D T T , DTE), 40ng//~l BSA, 7.4% (w/v) glycerol, 4% (w/v) polyethylene glycol 20000 (PEG), 0 . 4 r a m ATP, 0. 4m M U T P , 20/~M CTP, 11.4 kBq/#l c~-[32p]CTP (15 TBq/mmol), 0.16 mM 3'-O-methyl-GTP, 9.8 mM creatin phosphate, 0.4U//~l RNase T 1, 10ng//~l calf thymus DNA, approx. 1.8 ktg//~l H N E protein, plus the transcription factors and template plasmids as indicated in the figure legends. Transcription was then allowed to proceed for 45 rain at 30°C. Preincubation of the steroid hormone receptors with the template plasmids was carried out for 15 rain at room temperature (volume: 12.5 #1) in 5mM HEPES, 16 mM Tris (pH 7.9 at 4°C), 80-120 mM NaC1, 0.8 mM EDTA, 0.8 m M D T E , 8% (w/v) glycerol, 8% PEG, 80/~g/ml BSA and 4 ng/pl calf thymus DNA, before H N E and nucleotide mix was added. The reactions were finally stopped by addition of 200#1 stop buffer (0.1 M sodium acetate (pH 6), 10mM EDTA, 0.1% SDS, 125/~g/ml total yeast RNA). Samples were then phenol/CHC13 extracted twice before precipitation with ethanol. Transcripts were separated on a 6.5% denaturing polyacrylamide gel, visualized by autoradiography and quantitated by densitometry. Induction was determined with respect to the transcription of the control-template.

RESULTS AND DISCUSSION

The transcription assay: effect of progesterone (PR) and GR We have shown previously that PR purified from rabbit uterus stimulates transcription from the M M T V promoter in HeLa nuclear extracts and that this effect depends on the integrity of the octamer motifs [12, 14]. Using a similar HeLa nuclear extract and templates containing G-free cassettes [22] of different lengths [Fig. I(A)] we can now obtain a better stimulation of transcription by the progesterone receptor and detect a clear induction by GR purified from rat liver [Fig. I(B)]. Optimal transcription is observed when "macromolecular crowding" is generated by the addition of polyethylene glycol (PEG; [31 ]). Under these conditions, the extent of induction is dependent on the concentration of added receptor and maximal induction is 50-fold for the PR and 12-fold for the GR [Fig, I(C)]. The difference in the level of induction observed with PR and GR is probably due to the quality of the two receptor preparations. Quantitation of receptor content was based on bound hormone ligand, assuming one molecule of steroid bound per receptor monomer [12]. However, when PR and GR were compared in terms of

Cell-free Transcription from the MMTV Promoter their ability to shift a G R E / P R E oligonucleotide in the gel retardation assay, the P R was considerably more efficient (data not shown), correlating with its higher transcriptional activity. T h e reasons for the differences in D N A binding and transcriptional activity of both receptor preparations are unknown and could reflect posttranslational modification, interaction with other proteins, or protein stability.

25

T h e fact that other groups have reported high transcriptional induction with recombinant G R [32] suggests that the differences are not intrinsic to the receptor molecules. T o make sure that the effect of hormone receptors was due to binding of the receptor proteins to the H R E we included in the reaction mixture a control template containing the sequences of the M M T V promoter up

A)

B) ERE

Protein A

GRE

~---"~

"~--""~

GR-DBD

.4 :~ ~ i i ~ ! :

~¸z¸¸¸i

GR-DBD 1

2

3

4

5

6

c) GR-DBD i

PR

v

i

GR

PR

GR

360 nt

260 nt

l

2

3

o

29

13

4

5

6

26

3

-fold induction

Fig. 2. T h e G R - D B D is n o t able to i n d u c e M M T V t r a n s c r i p t i o n . (A) S D S - P A G E of r e c o m b i n a n t G R - D B D . T h e p u r i f i e d r e c o m b i n a n t p r o t e i n G R - D B D m i g r a t e s w i t h a n a p p r o x i m a t e m o l e c u l a r w e i g h t of 15 k D a On 10% S D S - P A G E . T h e m o l e c u l a r w e i g h t m a r k e r s are: o v a l b u m i n , l a c t a t e d e h y d r o g e n a s e , c a r b o n i c a n h y d r a s e a n d l a c t o g l o b u l i n . (B) B a n d s h i f t assay of D N A b i n d i n g specifieity of G R - D B D . T h e p u r i f i e d G R - D B D b i n d s specifically to a n oligonucleotide c o m p r i s i n g t h e G R E c o n s e n s u s s e q u e n c e (lanes 4-6---1, 3 a n d 7 p l of t h e p r o t e i n p r e p a r a t i o n ) b u t n o t to t h e E R E c o n s e n s u s (lanes 1-3---corresponding a m o u n t s ) in gel m o b i l i t y shift e x p e r i m e n t s . (C) Influence of G R - D B D on b a s a l a n d r e c e p t o r - d e p e n d e n t t r a n s c r i p t i o n . In t r a n s c r i p t i o n r e a c t i o n s as d e s c r i b e d in Fig. I(B) a n a d d i t i o n a l p r e i n c u b a t i o n step w i t h 2.5 ~1 (50 ng) of t h e G R - D B D h a s b e e n i n c l u d e d (15 m i n a t 25°C), followed by i n c u b a t i o n w i t h o u t a d d e d r e c e p t o r (lane 4), w i t h 42 n g P R (lane 5), or w i t h 55 ng G R (lane 6). C o n t r o l lanes w i t h o u t a d d i t i o n of t h e G R - D B D a r e d i s p l a y e d on t h e left (lanes 1-3). T h e c o r r e s p o n d i n g values of i n d u c t i o n a r e s h o w n u n d e r n e a t h t h e lanes.

26

Christian C. M6ws et al.

to - 8 0 and thus lacking the H R E region [Fig. I(A)]. Transcription of this template was not influenced by addition of PRs or GRs, supporting the specificity of the receptor-mediated induction [Fig. I(B)].

G R - D B D or other factors absent from our nuclear extract. Of course, it is also possible that bacterially expressed G R - D B D is not modified posttranslationally in the appropriate way.

The D N A binding domain of G R is not sufficient for transactivation in vitro

Involvement of N F I in M M T V

Using this sensitive assay we asked whether the D N A binding domain of the GR, that has been reported to be sufficient for transactivation in gene transfer experiments [33], was able to activate transcription from the M M T V promoter in vitro. T o this end we expressed the D N A binding domain of the glucocorticold receptor ( G R - D B D ) in Staphylococcus aureus as a fusion protein with protein-A [34]. After affinity chromatography on immunoglobulin-sepharose and cleavage of the protein-A moiety, the recombinant protein was purified by D N A cellulose chromatography. T h e purified preparation yielded a dominant band of molecular weight 15 k D A in S D S - P A G E along with some residual protein A [Fig. 2(A)]. This protein binds as a homodimer selectively and with high affinity to a synthetic G R E / P R E [20] [Fig. 2(B)], but is not able to activate transcription from the M M T V promoter in vitro [Fig. 2(C)]. W h e n added together with the intact hormone receptors, the D N A binding domain inhibits induction by GR, but not by PR, suggesting that either it forms inactive heterodimers with native G R [Fig. 2(c)] or competes for the D N A binding site, or both. We conclude that a homodimer of the intact receptor, but not of its D N A binding domain, is able to activate transcription in vitro. T h e apparent contradiction with previous gene transfer experiments performed with canonical G R E s [33] could be due to the use of slightly shorter G R fragments in these studies or to the lower D N A binding affinity of G R - D B D for the M M T V - H R E . Alternatively, a synergistic interaction between G R molecules bound to the four sites of the H R E , that is important for optimal induction in vivo [4], could require domains of G R not present in

transcription

T h e requirement of N F I for transcription of the M M T V promoter has been previously reported [11, 12]. Using the new transcription assay we confirm the deleterious effect of mutations on the NFI binding site on transcription from the M M T V promoter [Fig. 3(A)]. Mutation of both halves of the N F I binding site reduces transcription to about 1/3 of the level observed with the wild type M M T V promoter [Fig. 3(A)]. T h e relatively small effect of these mutations compared to previously published results [12] is not due to residual N F I binding, but probably reflects the macromolecular crowding effect of PEG. However, none of these mutations reduced the extent of induction observed after addition of either GRs or PRs [Fig. 3(A), G R data not shown], confirming previous results [11]. These data show that induction of M M T V transcription in vitro by either PR or G R is independent of N F I and are in apparent contradiction to those reported for the ovalbumin promoter [35]. This discrepancy may result from the different array of H R E s and N F I binding sites in the two promoters. It is known that when the distance between these two sites is appropriate one observes a functional synergism between them in transient transfection assays [36]. We conclude that although N F I and the hormone receptors have the potential to synergize under certain conditions, they do not synergize on the M M T V promoter. T h e H e L a cell nuclear extract can be depleted of more than 90% of its N F I binding activity by incubation with a N F I consensus oligonucleotide linked to sepharose. In the depleted extract transcription from the M M T V promoter decreases drastically, and can be

[Fig. 3(A-C)--Opposite] Fig. 3. Role of NF1 in MMTV transcription. (A) Influence of mutations in the NFI binding site on MMTV-transcription. T r a n s c r i p t i o n r e a c t i o n s h a v e been c a r r i e d out as d e s c r i b e d in Fig. 1, but in addition to the wild type (wt) MMTV promoter, we used templates containing mutations in the NF1 binding site. Either the promoter distal half (NFd), the promoter proximal half (NFp) or both halves (NFt) of the NFI binding site w e r e mutated [11] in the pMMTV 240/380 template as schematically indicated. Reactions w e r e p e r f o r m e d after preincubation with PR or control buffer. (B) Complementation of a NFI-depleted extract with r e c o m b i n a n t NFI fractions. HNE was depleted of NFI by incubation with the NFI-oligonucleotide resin (HNE-NF1, see e x p e r i m e n t a l section) and used for cell-free t r a n s c r i p t i o n e x p e r i m e n t s . R e a c t i o n s c o n t a i n e d 40 ng of pMMTV 80/280 and, as an internal control, 10 ng of pML(CzAT)19 that yields a 380 nt transcript. The d e p l e t e d e x t r a c t w a s tested without added factors (lane 4) or after c o m p l e m e n t a t i o n with equivalent DNA-binding activities of either recombinant intact NFI (21ng of vNFI, lane 1), recombinant NFI DNA binding domain (100 ng of vNFI-DBD, lane 3) or the corresponding buffer (c, lane 2). A control r e a c t i o n with non-depleted extract is included (lane 5). The densitometrie evaluation of the data is shown at the bottom. (C) Influence of r e c o m b i n a n t NFI on basal and receptor-dependent transcription. Cell-free transcription (as d e s c r i b e d in Fig. 1) w e r e p e r f o r m e d with the NFl-depleted HNE (HNE-NFI) without NFI (lane 2, 4 and 6) or after c o m p l e m e n t a t i o n with 10.5 ng vNFI (lanes 1, 3, 5). The reactions were preincubated with either buffer (lanes 1 and 2), 45 ng of PR (lanes 3 and 4) or 58 ng of GR (lanes 5 and 6). A p p r o p r i a t e control r e a c t i o n s with non-depleted extract have been included (lanes 7-9).

Cell-free Transcription

from the MMTV

Promoter

27

A) HRE

NF I

OTF

PR

260 nt

361) nl

÷

[

L L B)

C) HNE - NFI

transcript length :

HNE

PR

.4O4

t--:: .... AdML

÷'[~ .... / G-free"

380 nt

vNFI

+

÷

GR "'PR GR

-

+

360 nt

HNE - NFI

HNE

/o/, nt 380 nt

2

260 nt

!

2

3

4

5

rel. t r a n s c r i p t i o n (260 n t ) Fig. 3(A-C)--legend opposite.

3

5

6

7

8

9

Christian C. M6ws et al.

28

Participation of octamer transcription factors in basal transcription

nOTF1

vOTF2

vOTF1.POU

stimulation of transcription

Fig. 4. Role of o c t a m e r transcription factors in intact and N F l - d e p l e t e d HeLa nuclear extract. Influence of o c t a m e r transcription factors on r e c e p t 0 r - i n d e p e n d e n t transcription. Either NFI depleted extract or intact nuclear extract, were used to test the influence of purified or r e c o m b i n a n t o c t a m e r transcription factors (OTF1, OTF2 or the POU domain of OTFt, equivalent amounts in terms of D N A binding) on transcription of the pMMTV 80/280 template (260nt transcript). As internal control, 10 ng of pML(C2AT)19 that yields a 380nt transcript, were included in each reaction. The a u t o r a d i o g r a m s o f the gel retardation assay were quantitated by d e n s i o m e t r y . The figure represents the average of four separate e x p e r i m e n t s corrected for variations by reference to the internal adenovirus major late p r o m o t e r control.

partly restored by adding back purified N F I [Fig. 3(B)]. T h e r e a d - t h r o u g h signals as well as transcription f r o m the Adenovirus M a j o r Late p r o m o t e r are independent of N F I . Purified wild type N F I obtained from recombinant vaccinia virus infected H e L a cells [23] restores the transcriptional activity of the M M T V promoter, without affecting transcription from the Adenovirus Major Late p r o m o t e r [Fig. 3(B), lanes 1 and 2]. T h i s effect is not observed when the recombinant D N A binding domain of N F I , containing amino acids 4-240 [23, 27], is added to the depleted extract [Fig. 3(B), lane 3]. T h e amount of truncated protein added was equivalent to the a m o u n t of N F I as judged in band shift experiments with a N F I consensus oligonucleotide (data not shown). However, the D N A binding domain of N F I has been reported to be sufficient for adenovirus D N A replication [23, 38, 39], suggesting that different regions of N F I are required for transcriptional activation and D N A replication. We conclude that the carboxy-terminal half of N F I , containing the potential transactivation domains [38] is required for the observed transcriptional activity on the M M T V promoter. T h e N F I - d e p l e t e d extract was used to confirm the lack of effect of N F I upon h o r m o n e receptor dependent transcription. In the depleted extract the effect of adding purified PR or G R was indistinguishable from that observed in the intact H e L a nuclear extract or after complementation of the depleted extract with recombinant N F I [Fig. 1(C)]. T h u s , neither P R nor G R require N F I for in vitro activation of M M T V transcription.

We have previously shown that cell-free transcription from the M M T V p r o m o t e r is not drastically influenced by mutations in the octamer motifs [14]. T h e intact nuclear extract can be used to assay recombinant octarner binding factors in terms of their activity as transcription factors. T h e results of such experiments show that both O T F 1 and O T F 2 can enhance transcription from the M M T V p r o m o t e r (Fig. 4). T h e amount of factor added was calculated in band shift experiments with an octamer consensus oligonucleotide (data not shown). Based on this estimate the transcriptional activity of O T F 2 was slightly higher than that of O T F 1 (Fig. 4). Contrary to the intact octamer transcription factors, a recombinant protein containing the P O U - d o m a i n of O T F 1 (amino acids 269-440), albeit able to efficiently bind to D N A , does not exhibit transcriptional activity (Fig. 4), indicating that regions of the protein outside of the D N A binding domain are needed to generate a functional protein. T h e O T F 1 concentration in the intact H e L a extract seems to be limiting for transcription, as addition of purified native or recombinant O T F 1 enhances transcription from the M M T V promoter. T h e fact that not only O T F 1 but also O T F 2 is able to mediate basal M M T V transcription is interesting, as a p r o m o t e r selective activity of both octamer transcription factors has been reported for other genes [40]. Moreover, the P O U domain of O T F 1 , that is sufficient to support adenovirus D N A replication [27], is not effective in supporting transcription from the M M T V promoter. Again, these results suggest that different domains of O T F 1 are used for stimulation of D N A replication and M M T V transactivation. T h e effect of O T F 1 or O T F 2 on M M T V transcription is of similar magnitude in N F I - d e p l e t e d extract and in complete extract (Fig. 4). T h e r e might be two explanations for this finding: either the two classes of transcription factors bind to the same template but do not synergize, or each factor binds to a separate template molecule. Our new studies confirm the basic observation that the effect of PR on cell-free transcription of M M T V is at least partly mediated by binding of O T F 1 to the octamer motifs in the M M T V p r o m o t e r [14]. Mutation of these sites reduces the receptor-mediated enhancem e n t of transcription [14]. A similar behaviour is observed with G R (data not shown), suggesting that the PR and G R use similar pathways for in vitro transactivation. We have previously found that GR, as well as PR, stabilizes binding of O T F 1 to the M M T V p r o m o t e r [14] and this could explain the role of the octamer motifs in G R - i n d u c e d M M T V transcription. However, mutation of both octarner motifs and the

Cell-free T r a n s c r i p t i o n f r o m t h e M M T V NFI

binding

sites does not abolish induction

ing basic transcription

by PR

contribution

completely (data not shown), suggesting that on these mutant promoters the receptor can also act by recruit(A)

type promoter

~ ~

NF1

o

OTF1 L~ '

NF1 O~~ + r3 m ~ OTF-I ~' ~

NF1

+

+ ~

z

oz

29 factors to the TATA

of this pathway

to induction

box. The of the wild

remains to be established.

(B)

O~ ~ ÷ ~ ~~

OTF-I + o~

Promoter

,,~ !~ o

~ o ~

OTF-I

~

L~

OTF-1 ~. + B NF1

~

NF1 +

o

OTF-I

-

zz " ,x OTFI |FI + OTFI rFl FI

- 2 x OTFI - OTF1

"] NF1

- NFI-DBD - OTFI-POU 1 2 3 4 5 6 7 8 9101112131415161718192021

] Free 2 3 4 5 67

8-9 10~iii2~1314~i516F718192021

(c) OTFI

NFI

NF!

OTFI

.-:r~

"r

4- NFI+OTFI

NF!

~

~1~ O T F I

~,

NF!

2

3

4

$

6

7

$

9

Fig. 5. B i n d i n g of NF! a n d OTF1 to t h e M M T V p r o m o t e r . (A) DNA b i n d i n g e x p e r i m e n t w i t h of a n excess of p r o m o t e r f r a g m e n t . T h e G N O oligonucleotide (70,000 c p m ) was e n d - l a b e l e d a n d used for b a n d s h i f t assays w i t h r e c o m b i n a n t p r e p a r a t i o n s of NFI or OTF1. T h e a m o u n t s of NFI used were: 8 ng (lanes 1 a n d 9), 16 ng (lanes, 2, 10, 13-16 a n d 19), 24 ng (lanes 3, 4, 11 a n d 12). T h e a m o u n t s of OTF1 used were: 6 ng (lanes 5 a n d 13), 12 ng (lanes 6, 9-12, 14 a n d 20) a n d 24 n g (lanes 7, 8, 15 a n d 16). T h e a m o u n t of t h e D N A b i n d i n g d o m a i n of NFI ( N F I - D B D ) was 15 n g (lanes 17, 20 a n d 21) a n d the a m o u n t of t h e P O U d o m a i n o f OTF1 ( P O U ) was 4 ng (lanes 18, 19 a n d 21). W h e n i n d i c a t e d 100 ng of specific c o m p e t i t o r DNA c o n t a i n i n g t h e c o n s e n s u s s e q u e n c e of NFI ( N F I - C o m p . , lanes 4 a n d 12) or the o c t a m e r c o n s e n s u s m o t i f ( O T F 1 - C o m p . , l a n e s 8 a n d 16) were a d d e d to t h e r e a c t i o n s . (B) D N A b i n d i n g e x p e r i m e n t w i t h l i m i t i n g a m o u n t s of p r o m o t e r f r a g m e n t . T h e e x p e r i m e n t a l p r o t o c o l was s i m i l a r to t h a t d e s c r i b e d u n d e r (A), b u t using only one fifth t h e a m o u n t of t h e r a d i o a c t i v e GNO oligonucleotide. T h e a m o u n t s of NFI used were: 8 ng (lane 11), 16 ng (lanes 12 a n d 16-21) or 2 4 n g (lane 1, 13 a n d 14). T h e a m o u n t s of OTF1 used were: 6 n g (lanes 4 a n d 16), 12 ng (lanes 5, 11-14 a n d 17), 24 n g (lanes 7 a n d 18), 36 n g (lanes 8 a n d 19) a n d 48 n g (lanes 10, 20 a n d 21). Specific c o m p e t i t o r D N A was a d d e d as in (A). (C) D N A b i n d i n g e x p e r i m e n t u n d e r t h e s a m e c o n d i t i o n s as in (B), b u t u s i n g GNO o l i g o n u c l e o t i d e s c a r r y i n g m u t a t i o n s in t h e p r o m o t e r distal (top) or in t h e p r o m o t e r p r o x i m a l ( b o t t o m ) o c t a m e r motifs. T h e a m o u n t s of NFI used were: 8 ng (lane 1), 16 ng (lanes 2 a n d 4-9) a n d 24 n g (lane 3). T h e a m o u n t of OTF1 u s e d were: 6 ng (lane 4), 12 ng (lanes 1-3 a n d 5), 24 n g (lane 6) a n d 36 ng (lanes 7-9). W h e n i n d i c a t e d , specific c o m p e t i t o r DNA was a d d e d as in (A).

30

Christian C. M6ws et al.

Independent binding of N F I and O T F 1 to the M M T V promoter In view of the functional independence of N F I binding site and octamer motifs, we asked whether N F I and O T F 1 bind synergistically or independently to the M M T V promoter. We used the recombinant proteins and an excess of a p r o m o t e r fragment, in band shift assays which reflect the conditions in the in vitrotranscription experiments. We observe specific complexes with recombinant N F I or O T F 1 , that are selectively competed by the appropriate oligonucleotide [Fig. 5(A)]. At low concentrations of O T F 1 we see a single retarded complex, the majority of which reflects occupancy of the p r o m o t e r distal octamer m o t i f that is known to bind O T F 1 with high affinity [14]. At higher concentration of O T F 1 we detect an additional slower migrating complex than most likely represents the binding of two O T F 1 molecules to the two octamer motifs of the M M T V p r o m o t e r [Fig. 5(A), lanes 6 and 7]. W h e n N F I and O T F 1 are added together at low concentrations relative to the D N A fragment, we find no indication for the formation of a ternary complex containing N F I and O T F 1 on the same p r o m o t e r [Fig. 5(A)]. On the contrary both proteins seem to compete for binding to the M M T V promoter. A similar behavior is observed when truncated versions of the two factors are used. T h e N F I - D B D actually reduces binding of O T F 1 [Fig. 5(A), compare lane 6 and 21] or its P O U domain [Fig. 5(A), compare lanes 19 and 22], whereas the P O U domain of O T F 1 reduces binding of N F I [Fig. 5(A), compare lanes 2 and 20] or its D N A binding domain, although N F I - D B D appears to be dominant in our assay [Fig. 5(A), compare lanes 17 and 21]. T h u s , u n d e r optimal conditions for detecting synergism, there is no D N A binding cooperativity between N F I and O T F 1 . We next wanted to know whether N F I and O T F I can bind to the same template molecule. T o answer this question we p e r f o r m e d similar band shift experiments but using limiting amounts of the p r o m o t e r fragment. U n d e r these conditions, we see a larger fraction of the D N A molecules participating in complex formation with either N F I or O T F 1 and a very clear complex with two O T F 1 molecules [Fig. 5(B), lanes 11-13 and 17-20]. T h i s complex seems to contain N F I and O T F 1 as it is specifically competed by an N F I oligonucleotide as well as by an oligonucleotide containing the octamer m o t i f [Fig. 5(B), lanes 14 and 21]. T h e observed intensity of this complex is weak and does not indicate any cooperativity in D N A binding between N F I and OTF1. T h e ternary complex with N F I and O T F 1 could be formed with an O T F 1 molecule b o u n d to the p r o m o t e r proximal or to the p r o m o t e r distal octamer motifs. T o address this question we analyzed the binding behavior of p r o m o t e r fragments carrying mutations in one of the two octamer motifs [Fig. 5(C)]. Mutation of the strong

p r o m o t e r distal octamer m o t i f reduces binding of O T F 1 at low protein concentration but does not interfere with the appearance of the ternary complex at higher protein concentrations [Fig. 5(C) top]. On the contrary, mutation of the p r o m o t e r proximal octamer motif, while having little effect on binding of O T F 1 alone, drastically reduces the intensity of the band corresponding to the ternary complex with N F I and O T F 1 [Fig. 5(C) bottom]. We conclude that the M M T V p r o m o t e r can accommodate a N F I h o m o d i m e r and an O T F 1 m o n o m e r bound to the p r o m o t e r proximal octamer motif. Quantitative analysis of the retarded complex demonstrate that binding of O T F 1 and N F 1 to the p r o m o t e r carrying the octamer distal m u tation is synergistic. T h i s is demonstrated by the increment in intensity of the NF1 containing complex observed with increasing concentrations of added O T F 1 [Fig. 5(C) top panel, compare lanes 5, 6 and 7] and by the decreased intensity of the NF1 complex when an octamer oligonucleotide is added as competitor (compare lanes 7 and 9). Similarly, the intensity of the O T F 1 containing complex decreases when a NF1 oligonucleotide is used as competitor (compare lanes 7 and 8). T a k e n together the binding data suggest that N F I and O T F 1 have the potential to synergize in a properly oriented array of binding sites but compete for binding to their respective sites in the wild type M M T V promoter. W h e n the amount of D N A fragment was reduced and its concentration was limiting, we detected a ternary complex containing a dimer of N F I and a m o n o m e r of O T F 1 probably bound to the p r o m o t e r proximal octamer motif. However, the level of this ternary complex is low compared to the levels of complexes containing only N F 1 or O T F 1 . Therefore, though N F I and O T F 1 can physically bind to a single M M T V promoter, this interaction is not synergistic and is not observed under conditions of cell-free transcription on the natural configuration of sites of the M M T V promoter. T h i s suggests that basal transcription of the M M T V p r o m o t e r in vitro can occur via two independent pathways, one mediated by N F I and the other by octamer transcription factors. Only the second pathway is enhanced in the presence of hormone receptors in vitro. T h i s is reminiscent of the previously reported independent functions of N F I and O T F 1 in mediating adenovirus D N A replication [26]. It remains to be established whether the two pathways for transcriptional activation are independent, or even m u t u ally exclusive, in the intact cells. CONCLUSIONS In principle, hormone receptors could activate transcription either by modulating the efficiency of initiation at individual promoters or by recruiting inactive promoters into an active state. T h e first possibility envisages the existence of several states of activity

Cell-free T r a n s c r i p t i o n f r o m the M M T V P r o m o t e r for each promoter depending on the number and/or the array of transcription factors. The second possibility postulates the existence of only two functional promoter states, active or inactive. In the second model the f u n c t i o n o f i n d u c e r s is t o r e c r u i t p r o m o t e r s i n t o t h e a c t i v e state. T h i s m o d e l a c c o u n t s f o r m o s t e x p e r i m e n tal o b s e r v a t i o n s in i n d u c i b l e s y s t e m s a n d in p a r t i c u l a r for the hormonal induction of the MMTV promoter [41]. T h u s , t h e h o r m o n e r e c e p t o r s w o u l d a c t i v a t e t h e MMTV promoter by recruiting other transcription factors and the general transcriptional machinery. Our results suggest that they could accomplish this function b y at l e a s t t w o i n d e p e n d e n t p a t h w a y s . O n e p a t h w a y , mediated by octamer transcription factors, can be rep r o d u c e d in v i t r o u s i n g f r e e D N A as t e m p l a t e . A s t h e r e c e p t o r s f a c i l i t a t e t h e b i n d i n g o f p u r i f i e d O T F 1 to t h e MMTV p r o m o t e r [14], w e a s s u m e t h a t t h i s p a t h w a y involves a direct interaction between hormone receptors and octamer transcription factors. The second transactivation pathway, mediated by NFI, cannot be r e p r o d u c e d in v i t r o w i t h f r e e D N A . W e p o s t u l a t e t h a t this reflects the requirement for a nucleosomally organ i z e d c h r o m a t i n t e m p l a t e a n d h y p o t h e s i z e t h a t in v i v o h o r m o n a l i n d u c t i o n o f t h e M M T V p r o m o t e r is p a r t l y m e d i a t e d b y a n a l t e r a t i o n in c h r o m a t i n s t r u c t u r e t h a t e n a b l e s b i n d i n g o f N F I to t h e p r o m o t e r [13]. Acknowledgements--We thank H. Stunnenberg (EMBL) and W. Herr (Cold Spring Habor) for constructs, Claus Scheidereit (MPI Berlin) for n O T F 1, Bernhard Gross for purification of glucocorticoid and progesterone receptors and W. van Driel for purification of recombinant proteins. Work in Utrecht was supported in part by the Netherlands Foundation for Chemical Research (SON) with financial support from the Netherlands Organization for Scientific Research (NWO). Work in Marburg was supported by the Deutsche Forschungsgemeinschaft (DFG), the European Community BIOT E C H program and the Fonds der Chemischen Industrie (FCI).

REFERENCES

1. Beato M.: Gene regulation by steroid hormones. Cell 56 (1989) 335-344. 2. Beato M.: Transcriptional regulation of mouse mammary tumor virus by steroid hormones. Crit. Rev. Oneogen. 2 (1991) 195-210. 3. Cato A. C. B., Miksicek R., Schiitz G., Arnemann J. and Beato M.: The hormone regulatory element of mouse mammary tumour virus mediates progesterone induction. EMBO`7. 5 (1986) 2237-2240. 4. Chalepakis G., Arnemann J., Slater E. P., BrfiUer H., Gross B. and Beato M.: Differential gene activation by glucocorticoids and progestins through them hormone regulatory element of mouse mammary tumor virus. Cell 53 (1988) 371-382. 5. Chandler V. L., Maler B. A. and Yamamoto K. R.: DNA sequences bound specifically by glucocorticoid receptor in vitro render a heterologous promoter hormone response in vivo. Cell 33 (1983) 489--499. 6. Scheidereit C., Geisse S., Westphal H. M. and Beato M.: The glucocorticoid receptor binds to defined nucleotide sequences near the promoter of mouse mammary tumour. Nature 304 (1983) 749-752. 7. Richard-Foy H. and Hager G. L.: Sequence-specific positioning of nucleosomes over the steroid-inducible M M T V promoter. E M B O ,7. 6 (1987) 2321-2328. 8. Cordingley M. G., Riegel A. T. and Hager G. L.: Steroiddependent interaction of transcription factors with the inducible promoter of mouse mammary tumor virus in vivo. Cell 48 (1987) 261-270.

31

9, Nowock J., Borgrneyer A., P~schel A., Rupp A. W. and Sippel A. E.: The T G G C A protein binds to the M M T V - L T R , the adenovirus origin of replication, and BK virus. Nucl. Acid Res. 13 (1985) 2045-2062. 10. Miksicek R., Borgmeyer U. and Nowock J.: Interaction of the TGGCA-binding protein with upstream sequences is required for efficient transcription of mouse mammary tumor virus. E M B O ft. 6 (1987) 1355-1360. 11. Br/iggemeier U., Rogge L., Winnacker E. L. and Beato M.: Nuclear factor I acts as a transcription factor on the M M T V promoter but competes with steroid hormone receptors for DNA binding. E M B O `7. 9 (1990) 2233-2239. 12. Kalff M., Gross B. and Beato M.: Progesterone receptor stimulates transcription of mouse mammary tumor virus in a cell-free system. Nature 344 (1990) 360-362. 13. Pifia B., Briiggemeier U. and Beato M.: Nucleosome positioning modulates accessibility of regulatory proteins to the mouse mammary tumor virus promoter. Cell 60 (1990) 719-731. 14. Brfiggemeier U., Kalff M., Franke S., Scheidereit C. and Beato M.: Ubiquitous transcription factor OTF-1 mediates induction of the mouse mammary tumor virus promoter through synergistic interaction with hormone receptors. Cell 64 (1991) 565-572. 15. Toohey M. G., Lee J. W., Huang M. and Peterson D. O.: Functional elements of the steroid hormone-responsive promoter of mouse mammary tumor virus. J. Virol. 64 (1990) 4477--4488. 16. Meulia T. and Diggelmann H.: Tissue-specific factors and glucocorticoid receptors present in nuclear extracts binds next to each other in the promoter region of mouse mammary tumor virus. `7. Molec. Biol. 216 (1990) 859-872. 17. Stern S., Tanaka M. and Herr W.: The Oct-1 homeodoamin directs formation of a multiprotein-DNA complex with the HSV transactivator VP16. Nature 341 (1989) 624--630. 18. Janson L. and Pettersson U.: Cooperative interaction between transcription factor Spl and OTF-1. Proc. Natn. Acad. Sci. U.S.A. 87 (1990) 4732-4736. 19. Katan M., Haigh A., Verrijzer C. P., van der Vliet P. C. and O'Hare P.: Characterization of a cellular factor which interacts functionally with Oct-1 in the assembly of a multicomponent transcription complex. Nucl. Acid Res. 18 (1990) 6871-6880. 20. Kristie T. M. and Sharp P. A.: Interactions of the Oct-1 POU subdomains with specific DNA sequences and with the HSV alpha-tram-activator protein. Genes Devl. 4 (1990) 2283-2396. 21. Dignam J. D., Lebovitz R. M. and Roeder R. G.: Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucl. Acids Res. 11 (1983) 1475-1489. 22. Sawadogo M. and Roeder R. G.: Factors involved in specific transcription by human RNA polymerase II: analysis of a rapid and quantitative in vitro assay. Proc. Nam. Acad. Sci. U.S.A. 82 (1985) 4394--4398. 23. Gounari F., DeFrancesco R., Schmitt J., van Cortese R. and Stunnenberg H.: Amino-terminal domain of NF1 binds to DNA as a dimer and activates adenovirus DNA replication. E M B O `7. 9 (1990) 559-566. 24. Verrijzer C. P., Kal A. J. and van der Vliet P. C.: The Oct-1 homeo doamin contacts only part of the octamer sequence and full Oct-1 DNA-binding activity requires the POU-specific domain. Genes Devl. 4 (1990) 1964-1974. 25. Scheidereit C., Heguy A. and Roeder R. G.: Identification and purification of a human lymphoid-specific octamerbinding protein (OTF-2) that activates transcription of an immunoglobulin promoter in vitro. Cell 51 (1987) 783-793. 26. Mul Y. M., Verrijzer C. P. and van der Vliet P. C.: Transcription factors NFI and NFIII/Oct-1 function independently, employing different mechanisms to enhance adenovirus DNA replication. `7. Virol. 64 (1990) 5510-5518. 27. Verrijzer C. P., Kal A. J. and van der Vliet P.: The DNA binding domain (POU domain) of transcription factor Oct-1 suffices for stimulation of DNA replication. E M B O `7. 9 (1990) 1883-1888. 28. Geisse S., Scheidereit C., Westphal H. M., Hynes N. E., Groner B. and Beato M.: Glucocorticoid receptors recognize DNA sequences in and around murine mammary turnout virus DNA. E M B O `7. I (1982) 1613-1619. 29. Chalepakis G., Schauer M., Cao X. and Beato M.: Efficient binding of glucocorticoid receptor to its responsive element requires a dimer and DNA flanking sequences. D N A Cell Biol. 9 (1990) 355-368.

32

C h r i s t i a n C. M 6 w s et al.

30. Arndt-Jovin D. J., Jovin T. M., B~ihn W., Frischauf A.-M. and Marquardt M.: Covalent attachment of DNA to agarose. Improved synthesis and use in affinity chromatography. Eur. J. Biochem. 54 (1975) 411--418. 31. Zimmerman S. B. and Harrison B.: Macromolecular crowding increases binding of DNA polymerase to DNA: an adaptive effect. Proc. Natn. Acad. Sci. U.S.A. 84 (1987) 1871-1875. 32. Tsai S. Y., Srinivasan G., Allan G. F., Thompson E. B., O'Malley B. W. and Tsai M-J.: Recombinant human glucocorticold receptor induces transcription of hormone response genes in vitro. J. Biol. Chem. 265 (1990) 17,055-17,061. 33. Godowski P. J., Rusconi S., Miesfeld R. and Yarnamoto K. R.: Glucocorticoid receptor mutants that are constitutive activators of transcriptional enhancement. Nature 325 (1987) 365-368. 34. Cao X., Preiss T., Slater E. P., Westphal H. M, and Beato M.: Expression and functional analysis of steroid receptor fragments secreted from Staphylococcus aureus. J. Steroid Biochem. Molec. Biol. 44 (1993) 1-11. 35. Allan G. F., Ing N. H., Tsai S. Y., Srinivasan G., Weigel N. L., Thompson E. B., Tsai M-J. and O'Malley B. W.: Synergism between steroid response and promoter elements during cell-free transcription. J. Biol. Chem. 266 (1991) 5905-5910.

36. Schfile R., Muller M., Otsuka-Markami H. and Renkawitz R.: Cooperativity of the glucocorticoid receptor and the CACCCbox binding factor. Nature 332 (1988) 87-90. 37. Mull Y. M. and Van der Vliet P. C.: Nuclear factor I enhances adenovirus DNA replication by increasing the stability of a preinitiation complex. E M B O J. 11 (1992) 751-760. 38. Mermod N., O'Neill E. A., Kelly T. J. and Tjian R.: The proline-rich transcriptional activator of CTF-NF-1 is distinct from the replication and DNA binding domain. Cell 58 (1989) 741-753. 39. Bosher J., Leith I. R., Temperley S. M., Wells M. and Hay R. T.: The DNA-binding domain of nuclear factor-1 is sufficient to cooperate with the adenovirus type-2 DNA-binding protein in viral DNA replication. J. Gen~ Virol. 72 (1991) 2975-2980. 40. Tanaka M., Lai J-S. and Herr W.: Promoter-selective activation doamins in Oct-1 and Oct-2 direct differential activation of an snRNA and mRNA promoter. Cell 68 (1992) 755-767. 41. Ko M. S. H., Nakauchi H. and Takahashi N.: The dose dependence of glucocorticoid-inducible gene expression results from changes in the number of transcriptionally active templates. E M B O J. 9 (1990) 2835-2842.