Gene, 1i0 (1992) 225-228 © 1992 Elsevier Science Publishers B.V. All fights reserved. 0378-1119/92/$05.00

GENE 06228

225

Interactions between proteins bound to the duck/IA-globin gene promoter and enhancer detected by the DNasel footprinting

(Avian globin-encoding genes; recc\mbinant DNA; transcription factors; DNA looping)

Alei Cvekl" and V~clav Pa~es b

a Institute of Organic Chemistry and Biochemistry and b Insgitute of Molecular Genetics of the Czechoslovak Academy of Sciences, Prague (Czechoslovakia) Tel. (42-2)311-9568: Fax (42-2) 311-9017

Received by W.A. Fenton: 3 September 1991 Revised/Accepted: 18 September/20 September 1991 Received at publishers: 28 October 1991

SUMMARY

The duck/~A-globin (pAGLB) enhancer was closely linked to the duck ~A-GLB promoter, and the construct was used to study binding of nuclear proteins to specific sites of these regulatory elements. DNaseI-footprint analysis showed that the presence of the enhancer induced binding of proteins to additional sites on the promoter. The results are consistent with the looping-out model, based on specific interactions of enhancer-bound and promoter bound-proteins.

INTRODUCTION

The transcription of eukaryotic genes is controlled by DNA elements located either close to the start point of transcription (proximal elements, e.g., promoters) or fur- ther upstream or downstream (distal elements, e.g., en- hancers (Dynan, 1989). Proximal elements are thought to bind general transcription factors such as TFIID, whereas distal elements bind factors required for efficient initiation of transcription and its regulation (Lewin, 1990). The pre- cise mechanism(s) by which distally bound factors operate is not clear; they may increase (or decrease) the binding of general transcription factors to the promoter or recruit sep-

Correspondence to: Dr. A. Cvekl at his present address, National Institutes of Health, Bldg. 6, Rm. 214, Bethesda, MD 20892 (U.S.A.) Tel. (301)496- 7490; Fax (301)402-0781; and Dr. V. Pa~es, Institute of Molecular Ge- netics of the (~SAV, CS-166 37 Prague (Czechoslovakia) Tel. (42-2) 311-9568; Fax (42-2) 311-9017.

Abbreviations: bp, base pair(s); GLB, globin-encoding gene(s); kb, kilo- base(s) or 1000 bp; nt, nucleotide(s).

arate factors which are essential for efficient transcription. In either case, it is likely that distal factors modify the binding of proximal factors which themselves interact witl" other, non DNA-binding, proteins (Lewin, 1990; Lin and Green, 1991).

In vitro assembly of initiation complexes may be de- tected by various techniques: (i) kinetic studies using com- pletely or partially purified individual transcription factors (Reinberg et al., 1987; Maldonado et al., 1990); (ii) electro- phoretic mobility shift assays using the same sources of protein factors (Buratowski et al., 1989); and (iii) compar- ison of the DNasel protection patterns both in the presence and absence of distally bound factors (Hal et al., 1988; Horikoshi et al., 1988a,b; Li et al., 1991). The last method offers two alternative approaches, using either purified pro- teins or different templates. The former approach was used for studying the interaction of TFIID protein with two activators, GAIA (Horikoshi et al., 1988a) and ATF (Hal et al., 1988; Horikoshi et al., 1988b), and, more recently, for delineating the interaction between the Spl factor and the E2 protein (Li et al., 1991).

Here we report a study carried out with the ¢ ~ r ~.;~,-

EXPERIMENTAL AND DISCUSSION

proach, namely a comparison of DNasel protection pat- te-~ns using two different DNA probes with the same pro- tein factors. The footprint formed on the promoter region ot the duck ~A-GLB gene has been compared with the footprint on the same region when the promoter is joined closely to its enhancer, normally located more than 1.8 kb downstream from the cap site (Kretsovali et al., 1987). The differences found can be interpreted in terms of strong ef- fect of distally bound proteins on binding of proximal fac- tors, possibly mediated by the DNA loop between both elements.

Lanes

(a) Analysis of proximal elements Three DNA probes were used: P (promoter only)

(298 bp) is an EcoRI-HindIII fragment ofplasmid pDG/~gl, which was formed by cloning the PstI fragment (nt -201 to + 54, see Fig. 1) ofpDG~29 (Kretsovali et al., 1987) into pUCI8; PE (promoter plus enhancer) (698 bp) is the EcoRI.HindIIl fragment ofpDG1PE, which was prepared by inserting the same PstI fragme~tt into pDGEnh (Cvekl et al., 1991) in the wild-type orientation (Kretsovali et al., 1987); and E (enhancer only) (433 bp) is an EcoRI-HindIII fragment ofpDGEnh (Cvekl et al., 1991). The probes were 5'-end labelled as indicated (Fig. 1), and sites of specific protein-DNA interactions were detected by DNasel foot- printing using fractionated nuclear extracts of 13-day duck embryonic erythrocytes (Cvekl et al., 1991).

Three protein fractions, eluted frown a heparin-Ultrogel column with 200 raM, 400 mM, and 600 mM KCI (und designated H200E, H400E, and H600E, respee~ivel~), were used as sources of protein factors.

pDglPE 73 +2050. -201 ÷54 +1~1,1 ENHANCER " , PROMOTER ],H R

- I I ~'/////////////////////~ ,r,'///////////////////////~ r

226

L -13o~ ~

IE

t. PE"

Fig. 1. Schematic illustration of the DNA probes (DNA*). The general location of protein-binding regions (hatched boxes) on plasmid pDGIPE is indicated. For the experiment in Fig. 2, plasmids pDG~gl and pDRIPE were digested with HfndIII (H), S'-end labelled (asterisks) at the HindIff sites, and redigested with EcoRI (R) to yield the probes P and PE, re- spectively. For the experiment in Fig, 3, plasmids pDGEnh and pDglPE were digested with EcoRI, 5'-end labelled (dots) at the EcoRI sites, and redigested with HindIII to yield probes E and PE, respectively. The numbers (nt) indicate the positions in the intact gene (Kretsovali et al., 1987) of the various parts of the construct.

The results using probes P and PE are shown in Fig. 2. Two major conclusions can be drawn from this data: (i) the proximal element alone (probe P) interacts with nuclear proteins similarly to the analogous chicken promoter (lane 3, cf. Emerson et al., 1989; Evans et al., 1990); and (ii) the protection pattern over this element changes in at least six regions (I-VI) when the distal element is also present (probe PE) (compare lane 3 with lane 4, and lane 5 with 6, etc.). The differences all occur within or immediately adjacent to

Sp~ .~

PAL

DNA* P PE P PEP PE P PE P PE H400E 10 101~ 1~1 1010 10!05}/u H600E . . . . 2.5 I H200E " " 5 5 2.5

~9 1 2 3 4 5 6 7 8 9 10

Protein Site

spl.~

I I vt

NF-Et~

BS3 I CACCC box J~

GC box E

CCAATbox [~

TFIID

i V

i IV

B tit

i II

I i

Fig. 2. DNasc! ~'.otprintin~ analysis with probes P and PE. The indicated volumes of the hepann-Ukr0gel fractions, H200E, H400E, and H600E, were'first in~'.~bated on ice with 500 ng of poly(dI-dC), then with indicated probe (P or PE) at 30°C for 15 min. Digestions with DNaseI (1.25#g/ ml) were at 21°C for 60 sec. Protein binding sites (BS3, CACCC-box, CCAA T-box, and GC-box) are indicated by brackets, and the transcrip- tion factors involved (NF-E4, PAL, Spl, and TFIID) are also identified; Spl? indicates uncertainty that Spl actually binds to these sites. The differences in protection patterns seen between the two probes are indi- cated on the right by solid boxes and Roman numerals (I-VI). G + A, Maxam and Gilbert (1977) G + A reaction. A 6% polyacrylamide/8 M urea sequencing gel was dried before autoradiography.

sequence elements in the proximal region that show protein binding, suggesting that the presence of the distal element does indeed alter the mode or extent of protein binding to the proximal element.

(b) Analysis of distal elements A similar experiment to investigate changes in binding

the distal element when the proximal element was present did not reveal such profound alterations (Fig. 3). Never- theless, subtle but reproducible changes are visible in the border between PAL and AP-I sites, as well as in the AP-2

site. In both experiments, all changes were detected in the

presence of the H400E fraction, where typically most of DNA binding proteins elute; addition of the other fractions did not alter the pattern.

Lanes

Protein Site

DNA* E lie E PE E lie E lie E PE Hdl00E 10 101O1O 1 0 1 0 1 0 @ ] , H600E - - 5 S - - 2.525.~1 HI00E . . . . 5 S 2.$2.SJ

< ~ 1 2 3 4 5 6 7 8 9 10

BS2 [~ GATA'I "

. , . , ,

GATA'I

CACCC box

AP2~p 1

PAL

Bs, I

Fig. 3. DNasel footprintiag analysis of E and PE. Conditions of the experiment were identical with those described in Fig. 2, except that the probes PE and E, labelled at the EcoRl site (see Fig. 1), were used. Protein-binding sites (API, AP2, BSI, BS2, and CACCC-box) are indi- cated by brackets, and the transcription factors involved (GATA-1 and PAL) are also identified.

227

(c) Interpretation of ~'ootprinting patterns These effects of the distal element on protein binding to

the proximal element probably reflect interactions between proteins bound to these elements, with 'looping out' of the intervening DNA (Li et al., 1991; Mastrangelo et al., 1991; Ptashne, 1986; Su ~: ~., 1991). Such interactions have been proposed to occur in the case of the chicken flA-GLB pro- moter, and it has been suggested that they contribute to the stage-specific regulations of this gene (Gallarda et al., 1989). Our data are consistent with this suggestion. Re- gions I and II (Fig. 2) could result from interaction between NF-E4 and factors bound to the AP-1 and AP-2 motifs of the distal (enhancer) element, whereas region III might be produced by distal CACCC-box binding protein and region IV by erythroeyte specific factor GATA-1 (Eryfl). The greater degree of change detected at the proximal element in comparison with the distal element could reflect either the relative apparent binding constants of different factors or the directional specificity of protein-protein contacts be- tween individual factors (Takahashi et al., 1986).

(d) Conclusions The approach described here allows further development

of this and similar systems to study interactions of isolated motifs without using highly purified factors, such as are required for direct demonstratioi: of DNA looping using electron microscopy (Li e)al., 1991; Mastrangelo et al., 1991; Ptashne, 1986). Both functional and physical exper- iments will be required, however, for ultimate verification of this model system.

ACKNOWLEDGEMENT

We are grateful to G. Tebb for stimulating discussion and comment on the manuscript. The nt sequences have been deposited with GenBank under accession Nos.

X55351 and X55352.

REFERENCES

Buratowski, S., Hahn, S., Guarente, L. and Sharp, P.A.: Five interme- diate complexes in transcription initiation by RNA polymerase II. Cell 56 (1989) 549-561.

Cvekl, A., Horskfi, K., Vl~ek, C. and Pa~es, V.: Protein-binding A + T- rich motifs flank the duck flA-globin enhancer. Gene 103 (1991) 253- 257.

Dynan, W.S.: Modularity in promoters and enhoa~cers. Cell 58 (1989) 1-4.

Emerson, B.M., Nickol, J.M. and Fong, T.C.: Erythroid-specific activa- tion and derepression of the chick fl-globin promotel @n vitro. Cell 57 (1989) 1189-1200.

Evans, T., Felsenfeld, G. and Reitman, M.: Control of globin gene tran- scription. Annu. Rev. Cell Biol. 6 (1990) 95-124.

Gallarda, J.L., Foley, K.P., Yang, Z. and Engel, J.D.: The/]-globin stage

228

selector element factor is erythroid-specific promoter/enhancer bind- ing protein NF-E4. Genes Dev. 3 (1989) 1845-1859.

Hai, T., Horikoshi, M., Roeder, R.G. and Green, M.R.: Analysis of the role of the transcription factor ATF in the assembly of a functional preinitiation complex. Cell 54 (1988) 1043-1051.

Horikoshi, M., Carey, M.F., Kakidani, H. and Roeder, R.G.: Mechanism of action of a yeast activator: direct effect of GAL4 derivatives on mammalian TFllD-promoter interactions. Cell 54 (1988a) 665-669.

Horikoshi, M., Hal, T., Lin, Y.-S., Green, M.R. and Roeder, R.G.: Tran- scription factor ATF interacts with the TATA factor to facilitate es- tablishment of a preinitiation complex. Cell 54 (1988b) 1033-1042.

Kretsovali, A., Muller, M.M., Weber, F., Marcaud, L., Farache, G., Schreiber, E., Schaffner, W. and Scherrer, K.: A transcriptional en- hancer located between adult/~.globin genes in chicken and duck. Gene 58 (1987) 167-175.

Lewin, B.: Commitment and activation at polll promoters: a tail of protein-protein interaction s. Cell 61 (1990) 1161-1164.

Li, R., Knight, J.D., Jackson, S.P., Tjian, R. and Botchan, M.R.: Direct interaction between SPI and the BPV enhancer E2 protein mediates synergistic activation of transcription. Cell 65 (1991) 493-506.

Lin, Y.S. and Green, M.R.: Mechanism of action of an acidic transcrip- tional activator in vitro. Cell 64 (1991) 971-978.

Maldonado, E., Ha, I., Cortes, P., Weis, L. and Reinberg, D.: Factors

involved in specific transcription by mammalian RNA polymerase II: role of transcription factors IIA, IID, and IIB during formation of a transcription competent complex. Mol. Ceil. Biol. l0 (1990) 6335- 6347.

Mastrangelo, I.A., Cowney, A.J., Wall, J.S., Jackson, S.P. and Hough, P.V.C.: DNA looping and Spl multimer links: a mechanism for tran- scription synergism and enhancement. Proc. Natl. Acad. Sci. USA 88 (1991) 5670-5674.

Maxam, A. and Gilbert, W.: A new method for sequencing DNA. Proc. Natl. Acad. Sci. USA 74 (1977) 560-563.

Ptashne, M.: Gone regulation by proteins acting near by and at a distance. Nature 322 (1986) 697-701.

Reinberg, D., Horikoshi, M. and Roeder, R.G.: Factors involved in spe- cific transcription by mammalian RNA polymerase II. Functional analysis of initiation factors IIA and l i d and identification of a new factor operating at sequences downstream ofthe initiation site. J. Biol. Chem. 262 (1987) 3322-3330.

Su, W., Jackson, S., Tjian, R. and Echols, H.: DNA looping between sites for transcription activation: self-association of DNA bound SPI. Genes Dev. 5 (1991) 820-826.

Takahashi, K., Vigneron, M., Matthes, H., Wildeman, A., Zenke, M. and Chambon, P.: Requirement of stereospeeific alignments for initiation from the simian virus 40 early promoter. Nature 319 (1986) 121-126.