5. Repressor protein binds to the operator and is released by inducer

10.5 Repressor protein binds to the operator and is released by inducer

The repressor was isolated originally by purifying the component able to bind the gratuitous inducer IPTG (Gilbert and Muller-Hill, 1966; see 100.1 Isolation of Repressor). (Because the amount of repressor in the cell is so small, in order to obtain enough material it was necessary to use a promoter up mutation to increase lacI transcription, and to place this lacI locus on a DNA molecule present in many copies per cell. This results in an overall overproduction of 100 V1000-fold.)

The repressor binds to double-stranded DNA containing the sequence of the wild-type lac operator. The repressor does not bind DNA from an O^c mutant. The addition of IPTG releases the repressor from operator DNA in vitro. The in vitro reaction between repressor protein and operator DNA therefore displays the characteristics of control inferred in vivo; so it can be used to establish the basis for repression (Gilbert and Muller-Hill, 1967).

Figure 10.11 The lac operator has a symmetrical sequence. The sequence is numbered relative to the startpoint for transcription at +1. The regions of dyad symmetry are indicated by the shaded blocks.

How does the repressor recognize the specific sequence of operator DNA? The operator has a feature common to many recognition sites for bacterial regulator proteins: it is palindromic. The inverted repeats are highlighted in Figure 10.11. Each repeat can be regarded as a half-site of the operator.

We can use the same approaches to define the points that the repressor contacts in the operator that we used for analyzing the polymerase-promoter interaction (see 9 Transcription). Deletions of material on either side define the end points of the region; constitutive point mutations identify individual base pairs that must be crucial. Experiments in which DNA bound to repressor is compared with unbound DNA for its susceptibility to methylation or UV crosslinking identify bases that are either protected or more susceptible when associated with the protein.

Figure 9.16 One face of the promoter contains the contact points for RNA.

The region of DNA protected from nucleases by bound repressor lies within the region of symmetry, comprising the 26 bp region from -5 to +21. The area identified by constitutive mutations is even smaller. Within a central region extending over the 13 bp from +5 to +17, there are eight sites at which single base-pair substitutions cause constitutivity. This emphasizes the same point made by the promoter mutations summarized earlier in Figure 9.16. A small number of essential specific contacts within a larger region can be responsible for sequence-specific association of DNA with protein.

The symmetry of the DNA sequence reflects a symmetry in the protein. Repressor is a tetramer of identical subunits, each of which must therefore have the same DNA-binding site. Each inverted repeat of the operator is contacted in the same way by a repressor monomer. This is shown by symmetry in the contacts that repressor makes with the operator (the pattern between +1 and +6 is identical with that between +21 and +16) and by matching constitutive mutations in each inverted repeat. (However, the operator is not perfectly symmetrical; the left side binds more strongly than the right side to the repressor. A stronger operator would be created by a perfect inverted duplication of the left side.)

Various inducers cause characteristic reductions in the affinity of the repressor for the operator in vitro. These changes correlate with the effectiveness of the inducers in vivo. This suggests that induction results from a reduction in the attraction between operator and repressor. So when inducer enters the cell, it reduces the affinity for the operator of any repressor to which it binds. But consider a repressor tetramer that is already bound tightly to the operator. How does inducer cause this repressor to be released?

Figure 10.12 Does the inducer bind to free repressor to upset an equilibrium (left) or directly to repressor bound at the operator (right)?

Two models for repressor action are illustrated in Figure 10.12:

• The equilibrium model (left) calls for repressor bound to DNA to be in rapid equilibrium with free repressor. Inducer would bind to the free form of repressor, and thus unbalance the equilibrium by preventing reassociation with DNA.
  
• But the rate of dissociation of the repressor from the operator is much too slow to be compatible with this model (the half-life in vitro in the absence of inducer is >15 min). This means that instead the inducer must bind directly to repressor protein complexed with the operator. As indicated in the model on the right, inducer binding must produce a change in the repressor that makes it release the operator. Indeed, addition of IPTG causes an immediate destabilization of the repressor-operator complex in vitro.

Binding of the repressor-IPTG complex to the operator can be studied by using greater concentrations of the protein in the methylation protection/enhancement assay. The large amount compensates for the low affinity of the repressor-IPTG complex for the operator. The complex makes exactly the same pattern of contacts with DNA as the free repressor. An analogous result is obtained with mutant repressors whose affinity for operator DNA is increased; they too make the same pattern of contacts.

Overall, a range of repressor variants whose affinities for the operator span seven orders of magnitude all make the same contacts with DNA. Changes in the affinity of the repressor for DNA must therefore occur by influencing the general conformation of the protein in binding DNA, not by making or breaking one or a few individual bonds.

Figure 10.13 The structure of a monomer of Lac repressor identifies several independent domains. Photograph kindly provided by Mitchell Lewis.

The repressor has several domains. The DNA-binding domain occupies residues 1-59. It is known as the headpiece. It can be cleaved from the remainder of the monomer, which is known as the core, by trypsin. The crystal structure illustrated in Figure 10.13 offers a more detailed account of these regions (Friedman et al., 1995; Lewis et al., 1996).

The N-terminus of the monomer consists of two a-helices separated by a turn. This is a common DNA-binding motif, known as the HTH (helix-turn-helix); the two a-helices fit into the wide groove of DNA, where they make contacts with specific bases (see 11 Phage strategies). This region is connected by a hinge region to the main body of the protein. In the DNA-binding form of repressor, the hinge forms a small a-helix (as shown in the figure); but when the repressor is not bound to DNA, this region is disordered. The HTH and hinge together correspond to the headpiece.

The bulk of the core consists of two regions with similar structures. Each has a six-stranded parallel β-sheet sandwiched between two a-helices on either side. The inducer binds in a cleft between the two regions .

At the C-terminus, there is an a-helix that contains two leucine heptad repeats. The helices of four monomers associate to maintain the tetrameric structure.

Figure 10.14 The crystal structure of the core region of Lac repressor identifies the interactions between monomers in the tetramer. Each monomer is identified by a different color. Photographs kindly provided by Alan Friedman.
Multiple figure

Figure 10.14 shows the structure of the tetrameric core (using a different modeling system from Figure 10.13). It consists in effect of two dimers. The body of the dimer contains a loose interface between the N-terminal regions of the core monomers, a cleft at which inducer binds, and a hydrophobic core (top). The C-terminal regions of each monomer protrude as parallel helices. (The headpiece would join on to the N-terminal regions at the top.) Together the dimers interact to form a tetramer (center) that is held together by a C-terminal bundle of 4 helices.

Sites of mutations are shown by beads on the structure at the bottom. lacI^s mutations map in two groups: yellow shows those that affect the dimer interface, and gray shows those in the inducer-binding cleft. lacI- mutations that affect oligomerization map in two groups; white shows those that prevent dimer formation, and purple shows those that prevent tetramer formation from dimers.

Early work suggested a model in which the headpiece is relatively independent of the core. It can bind to operator DNA by making the same pattern of contacts with a half-site as intact repressor. However, its affinity for DNA is many orders of magnitude less than that of intact repressor. The reason for the difference is that the dimeric form of intact repressor allows two headpieces to contact the operator simultaneously, each binding to one half-site. This enormously increases affinity for the operator.

Figure 10.15 Inducer changes the structure of the core so that the headpieces of a repressor dimer are no longer in an orientation that permits binding to DNA. Photographs kindly provided by Mitchell Lewis.

Binding of inducer causes an immediate conformational change in the repressor protein. Binding of two molecules of inducer to the repressor tetramer is adequate to release repression. Figure 10.15 shows that binding of inducer changes the orientation of the headpieces relative to the core, with the result that the two headpieces in a dimer can no longer bind DNA simultaneously. This eliminates the advantage of the multimeric repressor, and reduces the affinity for the operator.

Figure 10.16 When a repressor tetramer binds to two operators, the stretch of DNA between them is forced into a tight loop. (The blue structure in the center of the looped DNA represents CAP, another regulator protein that binds in this region). Photograph kindly provided by Mitchell Lewis.

The allosteric transition that results from binding of inducer occurs in the repressor dimer. So why is a tetramer required to establish full repression? Each dimer can bind an operator sequence. This enables the intact repressor to bind to two operator sites simultaneously. In fact, there are two further operator sites in the initial region of the lac operon. The original operator, O1, is located just at the start of the lacZ gene. It has the strongest affinity for repressor. Weaker operator sequences (sometimes called pseudo-operators) are located on either side; O2 is 410 bp downstream of the startpoint, and O3 is 83 bp upstream of it. When Lac repressor binds simultaneously to O1 and to one of the other operators, it causes the DNA between them to form a loop. Figure 10.16 shows a model for binding of tetrameric repressor to two operators (Oehler, 1990).

Binding at the additional operators affects the level of repression. Elimination of either the downstream operator (O2) or the upstream operator (O3) reduces the efficiency of repression by 2-4 . However, if both O2 and O3 are eliminated, repression is reduced 100 . This suggests that the ability of the repressor to bind to one of the two other operators as well as to O1 is important for establishing repression. We do not know how and why this simultaneous binding increases repression.

We know most about the direct effects of binding of repressor to the operator (O1). It was originally thought that repressor binding would occlude RNA polymerase from binding to the promoter. However, we now know that the two proteins may be bound to DNA simultaneously, and the binding of repressor actually enhances the binding of RNA polymerase! But the bound enzyme is prevented from initiating transcription.

Figure 9.10 RNA polymerase initially contacts the region from -55 to +20. When sigma dissociates,the core enzyme contracts to -30; when the enzyme moves a few base pairs, it becomes more compactly organized into the general elongation complex.

The equilibrium constant for RNA polymerase binding alone to the lac promoter is 1.9 10⁷ M^-1. The presence of repressor increases this constant by two orders of magnitude to 2.5 10⁹ M^-1. In terms of the range of values for the equilibrium constant K_B given in Figure 9.10, repressor protein effectively converts the formation of closed complex by RNA polymerase at the lac promoter from a weak to a strong interaction.

What does this mean for induction of the operon? The higher value for K_B means that, when occupied by repressor, the promoter is 100 times more likely to be bound by an RNA polymerase. And by allowing RNA polymerase to be bound at the same time as repressor, it becomes possible for transcription to begin immediately upon induction, instead of waiting for an RNA polymerase to be captured.

The repressor in effect causes RNA polymerase to be stored at the promoter. The complex of RNA polymerase Prepressor PDNA is blocked at the closed stage. When inducer is added, the repressor is released, and the closed complex is converted to an open complex that initiates transcription. The overall effect of repressor has been to speed up the induction process.

Figure 10.20 Operators may lie at various positions relative to the promoter.

Does this model apply to other systems? The interaction between RNA polymerase, repressor, and the promoter/operator region is distinct in each system, because the operator does not always overlap with the same region of the promoter (see Figure 10.20). For example, in phage lambda, the operator lies in the upstream region of the promoter, and binding of repressor occludes the binding of RNA polymerase (see 11 Phage strategies). So a bound repressor does not interact with RNA polymerase in the same way in all systems.

Research

Friedman, A. M., Fischmann, T. O., and Steitz, T. A. (1995). Crystal structure of lac repressor core tetramer and its implications for DNA looping. Science 268, 1721-1727.

Gilbert, W. and Muller-Hill, B. (1966). Isolation of the lac repressor. Proc. Nat. Acad. Sci. USA 56, 1891-1898.

Gilbert, W. and Muller-Hill, B. (1967). The lac operator is DNA. Proc. Nat. Acad. Sci. USA 58, 2415-2421.

Lewis, M. et al. (1996). Crystal structure of the lactose operon repressor and its complexes with DNA and inducer. Science 271, 1247-1254.

Oehler, S. (1990). The three operators of the lac operon cooperate in repression. EMBO J. 9, 973-979.