Lac Repressor at Last Robert T Sauer

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Minireview 219

Lac repressor at last Robert T Sauer

Crystal and cocrystal structures of the LacI and PurR Figure 1 repressors reveal a novel use of hinge ␣ helices which bind in the minor groove of the operator and mediate transmission of the allosteric signals that modulate DNA-binding activity.

Address: Department of Biology, MIT, Cambridge, MA 02139, USA.

Structure 15 March 1996, 4:219–222

The central role of Lac repressor (LacI) in the history of molecular biology and transcriptional regulation has ensured its inclusion in virtually every modern text book of molecular biology, genetics, and biochemistry. As every student of these disciplines learns, LacI negatively regu- lates transcription of ␤-galactosidase and other genes in the lac operon of E. coli. It does so by binding to a partially symmetric operator of approximately 20 bp that is located near the start point of transcription. Secondary operator sites located roughly 90 bp and 400 bp to each side of the primary operator aid in the repression process. The activity of LacI is modulated by the binding of small inducer molecules such as allolactose or isopropyl-␤-D-thiogalacto- side (IPTG), which decrease its operator affinity, resulting in dissociation and relief of repression. Like most transcription factors, the LacI protein is oligomeric (a tetramer of 360-residue subunits) and modular. An isolated N-terminal headpiece of 50–60 residues binds weakly to operator half-sites as a monomer [1] and a C-terminal core fragment of about 300 residues binds inducer and mediates tetramer formation [2]. Two views of the LacI dimer–operator interaction, related by a 90° rotation [2]. The repressor dimer is shown as a ribbon trace with one Although micro-crystals of LacI were first reported in 1974 red monomer and one yellow monomer. Residue positions involved in [3], obtaining diffraction-quality crystals and solving the binding the inducer IPTG are shown as purple spheres. The double structure proved to be extremely difficult. Fortunately, helical backbone of operator DNA is shown as a black ribbon trace. Coordinates were generously provided by M Kercher and M Lewis. two groups persevered and have now reported the crystal structures of intact LacI [4], intact LacI bound to IPTG [4], the LacI core bound to IPTG [5], and a complex of looping, and provide a simple molecular understanding of LacI with two operator DNA fragments [4]. The wait was the allosteric regulation of DNA binding by ligands. worthwhile. These structures can now be compared with the NMR structures of the headpiece and headpiece–half- Operator recognition and bending operator complex [6,7]. Moreover, the crystal structure of a The LacI tetramer forms a V-shaped molecule in which related repressor, the Purine repressor (PurR), complexed two dimers are attached tail-to-tail with their dyad axes with corepressor and operator DNA has also been solved arranged roughly parallel but at a small angle to each other [8], in addition to the structure of the unliganded PurR [4,5]. In the cocrystal [4], each dimer contacts a single core [9]. Together these structures reveal the mechanism 21 bp operator in an equivalent manner. Figure 1 shows of DNA recognition for the LacI family of repressors. They the structure of the LacI dimer bound to an operator frag- also show an interesting use of ␣ helices in minor-groove ment. The dimeric PurR–operator complex is almost iden- binding and bending of operator DNA, exhibit unexpected tical [8]. In both complexes there are two discrete types of quaternary asymmetry, suggest models for long-range DNA DNA contact. First, each N-terminal headpiece contacts 220 Structure 1996, Vol 4 No 3

the major groove of each operator half-site. Second, hinge HMG domain [13]. In each of these cases, the minor ␣ helices immediately adjacent to each headpiece bind in groove is wider and shallower and hydrophobic side chains the minor groove near the center of the site. The hinge from the proteins act to wedge open the bases. helices pack against one another, providing communica- tion with the other half-site, and also connect the head- Dimerization and inducer/corepressor binding pieces to the protein core. As discussed below, the hinge In LacI, dimer formation and inducer binding are medi- helices seem to mediate the allosteric response to inducer ated by two subdomains of the core, which have similar or corepressor binding. folds [4,5]. In each subdomain, a parallel ␤ sheet is packed between ␣ helices, forming a three-layered structure. The The DNA-binding headpieces of LacI and PurR consist of dimer interface is formed by interactions between the three ␣ helices [4,6–8]. The first two comprise a helix- DNA-proximal subdomains in adjacent subunits as well turn-helix motif which makes base-specific contacts in the as between the DNA-distal subdomains (Fig. 1). The major groove as well as DNA-backbone contacts. These inducer IPTG binds in a cleft at the interface between the overall interactions are similar in the LacI–operator proximal and distal subdomains of the same monomer. complex [4], the PurR–operator complex [8], and the LacI The C-terminal domain of PurR has the same overall headpiece–half-operator complex [7]. One notable differ- architecture and the corepressor hypoxanthine also binds ence between the whole- and half-operator LacI com- at the equivalent position at the interface of the two sub- plexes is the use of the hinge helix. NMR studies of a LacI domains [8,9]. Because the inducer/corepressor-binding fragment containing residues 1–56 do not show the hinge pockets are distant from the protein–DNA binding helix or interactions with the minor groove [6,7], suggest- surface, in both LacI and PurR, the binding of these small ing that this region is disordered. It is somewhat difficult to molecules must affect DNA binding indirectly. interpret this result because several residues from the C terminus of the intact hinge helix are missing in the 1–56 Tetramers, asymmetry and looping headpiece and might be needed to stabilize the helix or to Tetramer formation in LacI is mediated by a C-terminal bind the minor groove. Alternatively, the interaction of ␣ helix which contains two heptad repeats and is con- each hinge helix with the minor groove may only be stable nected to the rest of the core by a linker of approximately when it can pack against the symmetric hinge helix in 10 residues [4,5]. PurR lacks this helix and only forms a dimeric complex and/or when the C-terminal core is dimers. In the LacI tetramer, the four C-terminal helices attached. LacI headpieces bind operator DNA with a Kd of form an antiparallel four-helix bundle with D2 symmetry approximately 1 ␮M, whereas the intact protein binds with (three mutually perpendicular dyad axes). This is the sym- a Kd of almost 1 pM [1]. Thus, the DNA-binding affinity metry observed in nearly all known tetramer structures. of intact LacI is enhanced by approximately six orders of However, the linker regions have different conformations magnitude by some combination of the hinge-helix–DNA in each monomer, breaking the symmetry, and giving rise interactions and the chelate effect which reduces the to the asymmetric V-shaped tetramer. As a consequence, entropy of binding of the second headpiece. the three twofold axes in intact LacI are roughly parallel, placing all four DNA-binding domains on the same side of The overall arrangement and use of the hinge helices the molecule. appear to be similar for LacI and PurR [4,8], although the details of the interactions can be seen at higher resolution Asymmetric linkage of DNA-recognition domains to sym- in the latter cocrystal. By inserting into the minor groove, metric oligomerization domains may be a recurring feature the hinge helices induce a set of DNA distortions: a minor in DNA-binding oligomers that are larger than dimers. For groove that is significantly wider and shallower than that example, LacI-type models in which D2-symmetric of B-DNA; a 45° kink caused by intercalation of leucine tetramerization domains are asymmetrically linked to side chains between the minor-groove edges of the central dimeric DNA-binding domains have been proposed to base pairs; an approximate doubling of the helical rise explain how tetramers of both the tumor-suppressor between the central bases; and unwinding of the central protein p53 [14,15] and the Mnt repressor [16] bind base pairs by approximately 40–50°. As a consequence, the to their DNA recognition sites. An even more dramatic hinge helices and minor groove form complementary case of mismatched symmetry occurs for the trimeric binding surfaces allowing protein side chains and func- heat-shock transcription factor which must also utilize tional groups to contact the minor-groove edges of certain asymmetric linkers to connect its trimerization and DNA- base-pairs and the phosphate backbone. The set of minor- binding domains [17,18]. Such schemes permit great adapt- groove deformations observed in the LacI and PurR com- ability in the recognition of binding sites spaced at varying plexes causes the DNA to bend away from the protein distances and orientations along a DNA molecule because (Fig. 1). Interactions with the minor groove also lead to the symmetry of DNA recognition need not be related to similar modes of bending in protein–DNA complexes of the symmetry of oligomerization. In general, the closed TBP [10,11], the LEF-1 HMG domain [12], and the SRY symmetry of the oligomerization domain will function to Minireview Lac repressor Sauer 221

prevent indefinite aggregation of the protein (this may be Figure 2 especially critical in non-specific DNA binding), whereas the linkers and intersubunit contacts must allow sufficient conformational freedom to adapt to the necessary DNA- binding orientations without introducing so much flexibility that high-affinity binding would be precluded. In the LacI structures, the linkage between the tetramerization domain allows enough flexibility to permit the angle between the two dimers to change by about 10° [4].

In the cell, the binding of a LacI tetramer to both the primary operator and a secondary operator, with looping of the intervening DNA, appears to be important for achiev- ing maximum repression [19]. Beginning with the cocrystal structure of the repressor tetramer bound to two operator fragments, a loop corresponding to the spacing between the primary operator and the closest secondary operator could be modeled ([4]; Fig. 2). One interesting feature of this model is the prediction that the binding sites for the catabolite gene activator protein (CAP) and RNA polymerase should be exposed on the inner surface of the loop [4]. Clearly, this has implications for the ways in which RNA polymerase, LacI and CAP may interact in controlling transcription of the lac operon.

Allostery LacI and PurR respond differently to ligand binding. IPTG binding reduces the affinity of LacI for operator, whereas purine binding is needed for operator binding and repression in PurR. Although this difference is critical Plausible structure of a DNA loop stabilized by binding of one dimer in a functional sense, it is relatively minor in terms of of the LacI tetramer to the primary operator and binding of the other allosteric mechanism as the effect of a ligand will simply dimer of the tetramer to a secondary operator 93 base pairs away. The V-shaped LacI tetramer (yellow) is shown as a ribbon trace. The depend on whether it binds more strongly to a quaternary DNA is shown in white. Details of the model building and regulatory structure compatible with strong DNA binding or one that consequences for lac operon control are discussed in [2]. Coordinates binds DNA weakly [20]. One of the exciting aspects of the were generously provided by M Kercher and M Lewis. LacI and PurR structures now available is that we are afforded molecular views of both proteins with and without bound ligand [4,5,8,9]. would pull the hinge helices apart; in a dynamic alterna- tive, these changes could result in unfolding of the hinge Interestingly, there is no electron density for the head- helices. In either case, both the minor-groove interactions pieces or the hinge helices in the crystal structures of free of the hinge helices and the positioning of the headpieces LacI or the LacI–IPTG complex, indicating that these for DNA binding would be compromised. By binding pref- regions must be statically or dynamically disordered [4]. erentially to the T-state, IPTG increases the free energy However, the C-terminal regions of these proteins are very required to convert to the R-state. Thus, the overall DNA- similar to each other and to the crystal structure of the LacI binding affinity is reduced because more of the energy core–IPTG complex [4,5]. In the jargon of allostery [20], resulting from protein–DNA interactions must be used to each of these structures represents the T-state. The drive the T→R conformational change. LacI–operator complex represents the R-state. Comparing the core regions of the T- and R-structures shows that the In PurR, similar changes in the relative positions of the positions of the DNA-distal subdomains remain essentially two DNA-proximal subdomains cause the residues closest unchanged. Hinge-bending motions between subdomains, to the hinge helices to move farther apart in the T-state however, cause significant changes in the conformations (unliganded) than the R-state (PurR–hypoxanthine–DNA) and relative orientations of the two DNA-proximal core [8,9]. In this case, however, binding of the purine co- subdomains, which rotate by about 10° causing the repressor between the DNA-proximal and DNA-distal residues closest to the hinge helices to move 3–4 Å farther subdomains stabilizes the conformation in which the apart in the T-state. In a static model, this movement hinge helices can pack against each other and interact with 222 Structure 1996, Vol 4 No 3

the minor groove of the operator. Hence, in both LacI and 14. Cho, Y., Gorina, S., Jeffrey, P.D. & Pavletich, N.P. (1994). Crystal structure of a p53 tumor suppressor–DNA complex; understanding PurR, the allosteric signals are transmitted via structural tumorigenic mutations. Science 265, 346–355. changes in the hinge helices. 15. Lee, W., Harvey, T.S., Yin, Y., Yau, P. Litchfield, D. & Arrowsmith, C.H. (1994). Solution structure of the tetrameric minimum transforming domain of p53. Nat. Struct. Biol. 1, 877–890. Future directions 16. Waldburger, C.D. & Sauer, R.T. (1995). Domains of Mnt repressor: Although the LacI and PurR structures answer many roles in tetramer formation, protein stability, and operator DNA binding. Biochemistry 34, 13109–13116. important questions, new ones arise. How, for example, 17. Sorger, P.K. & Nelson, H.C. (1989). Trimerization of a yeast are the protein–DNA complexes assembled? Are transient transcriptional activator via a coiled-coil motif. Cell 59, 807–813. structural deformations in the DNA trapped by repressor 18. Flick, K.E., Gonzalez, L., Jr., Harrison, C.J. & Nelson, H.C. (1994). Yeast heat shock transcription factor contains a flexible linker between binding or does a stepwise process of binding induce the the DNA-binding and trimerization domains. Implications for DNA observed changes? Are looped DNA structures formed by binding by trimeric proteins. J. Biol. Chem. 269, 12475–12481. dimerization of dimers bound at the primary and sec- 19. Oehler, S., Eismann, E.R., Kramer, H. & Muller-Hill, B. (1990). The three operators of the lac operon cooperate in repression. EMBO J. 9, ondary operators or by binding of a tetramer to one opera- 973–979. tor followed by binding to the second operator? Do the 20. Monod, J., Wyman, J., & Changeux, J.-P. (1965). On the nature of allosteric transitions: a plausible model. J. Mol. Biol. 12, 88–118. hinge-helix residues remain in an ␣-helical conformation 21. Markiewicz, P., Kleina, L., Cruz, C., Ehret, S. & Miller, J.H. (1994). in the T-state? Which interactions determine the relative Genetic studies of the lac repressor: analysis of 4000 altered stabilities of the different allosteric states? Can LacI or Escherichia coli lac repressors resulting from suppression of nonsense mutations at 328 positions in the lacI gene. J. Mol. Biol. PurR variants that recognize novel ligands be engineered? 240, 421–433. The remarkably complete genetic analysis of LacI by Miller and colleagues [21], together with the LacI and PurR structures, should now permit numerous hypotheses about DNA recognition and allosteric regulation to be framed in mechanistic terms and tested. The LacI family of repressors will clearly remain a paradigm for understanding transcriptional regulation.

References 1. Ogata, R.T. & Gilbert, W. (1979). DNA-binding site of lac repressor probed by dimethylsulfate methylation of lac operator J. Mol. Biol. 132, 709–728. 2. Platt, T., Files, J & Weber, K. (1973). Specific proteolytic destruction of the NH2-terminal region and loss of the deoxyribonucleic acid- binding activity. J. Biol. Chem. 248, 110–121. 3. Steitz, T.A., Richmond, T.J., Wise, D. & Engelman, D. (1974). The lac repressor protein: molecular shape, subunit structure, and proposed model for operator interaction based on structural studies of microcrystals. Proc. Natl. Acad. Sci. USA 71, 593–597. 4. Lewis, M., et al., & Lu, P. (1995). Structure of the Escherichia coli lactose operon and its complexes with DNA and inducer. Science, in press. 5. Friedman, A.M., Fischman, T.O. & Steitz, T.A. (1995). Crystal structure of lac repressor core tetramer and its implications for DNA looping. Science 268, 1721–1727. 6. Kaptein, R., Zuiderweg, E.R.P., Scheek, R.M., Boelens, R., & von Gunsteren, W.F. (1985). A protein structure from nuclear magnetic resonance data: lac repressor headpiece. J. Mol. Biol. 182, 179–182. 7. Chuprina, V.P, Rullmann, J.A.C., Lamerichs, R.M.J.N., van Boom, J.H., Boelens, R. & Kaptein, R. (1993). Structure of the complex of lac repressor headpiece and an 11 base-pair half-operator determined by nuclear magnetic resonance spectroscopy and restrained molecular dynamics. J. Mol. Biol. 234, 446–462. 8. Schumacher, M.A., Choi, K.Y., Zalkin, H. & Brennan, R.G. (1994). Crystal structure of LacI member, PurR, bound to DNA: minor groove binding by ␣ helices. Science 266, 763–770. 9. Schumacher, M.A., Choi, K.Y., Lu, F., Zalkin, H. & Brennan, R.G. (1995). Mechanisms of corepressor-mediated specific DNA binding by purine repressor. Cell 83, 147–155. 10. Kim, Y., Geiger, J.H., Hahn, S. & Sigler, P.B. (1993). Crystal structure of a yeast TBP/TATA-box complex. Nature 365, 512–520. 11. Kim, J.L., Nikolov, D.B. & Burley, S.K. (1993). Co-crystal structure of TBP recognizing the minor groove of a TATA element. Nature 365, 520-527. 12. Love, J.J., Li, X., Case, D.A., Giese, K., Grosschedl, R. & Wright, P. (1995). Structural basis for DNA bending by the architectural transcription factor LEF-1. Nature 376, 791–795. 13. Werner, M.H., Huth, J.R., Gronenborn, A.M. & Clore, G.M. (1995). Molecular basis of human 46X, Y Sex reversal revealed from the three- dimensional solution structure of the human SRY–DNA complex. Cell 81, 705–714.