Conformational Change of Transcription Factors from Search to Specific Binding: A lac Repressor Case Study

Abstract

In a process known as facilitated diffusion, DNA-binding proteins find their target sites by combining three-dimensional diffusion and one-dimensional scanning of the DNA. Following the trade-off between speed and stability, agile exploration of DNA requires loose binding, whereas, at the DNA target site, the searching protein needs to establish tight interactions with the DNA. To enable both efficient search and stable binding, DNA-binding proteins and DNA often switch conformations upon recognition. Here, we study the one-dimensional diffusion and DNA binding of the dimeric lac repressor (LacI), which was reported to adopt two different conformations when binding different conformations of DNA. Using coarse-grained molecular dynamic simulations, we studied the diffusion and the sequence-specific binding of these conformations of LacI, as well as their truncated or monomeric variants, with two DNA conformations: straight and bent. The simulations were compared to experimental observables. This study supports that linear diffusion along DNA combines tight rotation-coupled groove tracking and rotation-decoupled hopping, where the protein briefly dissociates and reassociates just a few base pairs away. Tight groove tracking is crucial for target-site recognition, while hopping speeds up the overall search process. We investigated the diffusion of different LacI conformations on DNA and show how the flexibility of LacI's hinge regions ensures agility on DNA as well as faithful groove tracking. If the hinge regions instead form α-helices at the protein-DNA interface, tight groove tracking is not possible. On the contrary, the helical hinge region is essential for tight binding to bent, specific DNA, for the formation of the specific complex. Based on our study of different encounter complexes, we argue that the conformational change in LacI and DNA bending are somewhat coupled. Our findings underline the importance of two distinct protein conformations for facilitated diffusion and specific binding, respectively.

Figure 1

Schematic representations of the facilitated diffusion performed by LacI and the protein conformations involved. (A) Facilitated diffusion as a combination of 3D and 1D search mechanisms. The zoom panels demonstrate the structural properties of the specific and nonspecific complexes of LacI and the interacting DNA. The specific complex of LacI contains a bent DNA target as well as helices at the protein–DNA interface. The conformation of LacI in the nonspecific complex during the search process is still elusive, and usage of both conformations of LacI (or others) are possible. (B) Cartoon representation of the recognition conformation with cyan DBD and DNA, and violet, helical hinge region (PDB ID 1EFA). (C) Cartoon representation of the repressor search conformation with DBD and DNA colored in orange (PDB ID 1OSL) and the hinge region in pink. In both conformations, the core domain is shown in gray. The full-length search conformation was modeled by incorporating the core domain adopted from the full-length structure of the recognition conformation.

Figure 2

Linear diffusion of variants of LacI along linear DNA studied using coarse-grained simulations. Diffusion of LacI along DNA was studied for the recognition (A, C, and E) and search (B, D, and F) conformations. Each conformation was studied in three variants: monomeric DBD (A and B), dimeric DBD (C and D), and full-length LacI (E and F). The coarse-grained simulations follow the linear diffusion of the lac repressor on a 100 bp long coarse-grained DNA. Each panel depicts the translational movement along Z and the angular movement around the DNA (φ) collected for each variant of LacI in eight independent simulations, each comprising 10⁸ steps. The plots highlight the areas of DNA that were sampled by the recognition region of the corresponding variant of LacI. The dotted lines in the scatter plots mark the DNA backbones based on the definition in Figure S2. The figure shows results from simulations with salt concentration of 0.02 M. Plots showing the distributions for four different salt concentrations (0.01, 0.02, 0.03, and 0.04 M) are shown in Figures S3 and S4.

Figure 3

Molecular characteristics of linear diffusion of LacI along DNA: experiments versus computation. (A) Schematic representation of the diffusion dynamics and its molecular characterization. The diffusion is quantified by the diffusion coefficient for linear translocation, D (B), the pitch (C), and the rotational diffusion coefficient, D_r (D). Each plot shows the mean value of the corresponding parameter; the standard deviation is indicated by the shaded region. The pitch is obtained from the two diffusion coefficients via eq 9. The diffusion parameters were measured for the recognition (green) and search (orange) conformations of LacI, in monomeric DBD, dimeric DBD, and dimeric full-length variants. The experimentally measured values of the rotational and translational diffusion coefficients and the pitch at a salt concentration of 0.016 M are shown in gray.

Figure 4

Structural and dynamic properties of LacI diffusion along DNA. (A) Typical trajectory of the full-length search conformation of LacI diffusing along DNA. The locations of the two recognition regions are shown by light and dark orange curves as a function of time. The panel under the plot demonstrates when the analysis algorithm detects groove tracking (line), hops (stars), and flips (arrows). The distances covered during groove tracking and hopping at different salt concentrations are shown in (B) and (D), for the full-length dimeric LacI in the search (orange) and recognition (green) conformations. The respective durations of these events are shown in (C) and (E). (F) Frequencies of hops that are longer than five base pairs. The frequency is compared to the experimental value that is shown in gray.

Figure 5

Interplay between conformational changes in LacI and in DNA. (A) Mean distance, d, of Cα atoms in the LacI DBD measured from the DNA axis for recognition and search conformations, respectively. The data represents an average over frames recorded during groove tracking performed by full-length LacI. (B) Representative specific interaction between arginine 22 and base G5 in the DNA based on the crystal structure of the specific complex (PDB ID 1EFA). (C) Electrostatic interaction energies of the search and recognition conformations of LacI with either straight or bent specific DNA. (D) Occupancies of specific interactions at the straight or bent target site. The energetics and occupancies of the four complexes between LacI and DNA were obtained from five independent CG simulations with a salt concentration of 0.03 M. The frames for analysis were further selected based on alignment with the target site.