Experimental snapshots of a protein-DNA binding landscape

Abstract

Protein recognition of DNA sites is a primary event for gene function. Its ultimate mechanistic understanding requires an integrated structural, dynamic, kinetic, and thermodynamic dissection that is currently limited considering the hundreds of structures of protein-DNA complexes available. We describe a protein-DNA-binding pathway in which an initial, diffuse, transition state ensemble with some nonnative contacts is followed by formation of extensive nonnative interactions that drive the system into a kinetic trap. Finally, nonnative contacts are slowly rearranged into native-like interactions with the DNA backbone. Dissimilar protein-DNA interfaces that populate along the DNA-binding route are explained by a temporary degeneracy of protein-DNA interactions, centered on "dual-role" residues. The nonnative species slow down the reaction allowing for extended functionality.

Fig. 1.

Effect of point mutations on the stability of the first and last transition state ensembles in the multistate E2C-DNA kinetic route. (A) Representation of the effects on on the surface of the DNA-bound conformation of the HPV16 E2C homodimer (23). Relevant residues are labeled in a single monomer for clarity. Residues in α1 contacting the DNA bases as deduced from homologous E2C-DNA complexes, mutagenesis, and biophysical measurements (, –26) (N294, K297, C298, Y301, and R302) are Underlined. Residues N294, K297, Y301, and R302 also form nonspecific contacts with the DNA backbone as well as two other residues in α1, T295 and R300 (, –26). K304 and K305 from the 3₁₀ helix; V324, K325, and K327 from the β2-β3 loop; and T316, H318, K349, and T353 outside of the major recognition elements also form nonspecific contacts with the DNA backbone (23, 24, 26). Residues are colored according to changes in free energy upon mutation to alanine, except for K297R, R300M, and R302M. Uncharacterized residues are colored gray. (B) Comparison of the effects on (Blue Bars) and on the final complex (Black Bars). (C) Representation of the effects on on the surface of the DNA-bound conformation of the HPV16 E2C homodimer (23). (D) Comparison of the effects on (Green Bars) and on the final complex (Black Bars).

Fig. 2.

Effect of point mutations on the fast and slow concentration-independent association phases in the multistate E2C-DNA kinetic route. (A) Bar plot of the effects on the fast (Empty Bars) and slow phases (Black Bars). Thin Lines are a guide for the eye and correspond to a 2-fold change in the rate constant. (B) Correlation between the mutational effects on the fast and slow phases. R is -0.75, p-value is 0.011.

Fig. 3.

Free energy correlations in E2C-DNA complex formation. Each ΔΔG-value (in kcal/mol) describes the effect of a mutation on a given state along the E2C-DNA kinetic routes, taking the unbound reagents as a reference. The p-value describes the probability of observing the correlation by chance. (A) Final complex versus TSE^2-state. (B) Final complex versus . (C) Final complex vs.. (D) TSE^2-state vs. . (E) TSE^2-state vs. . (F) vs. .

Fig. 4.

Schematic free energy landscape for DNA sequence recognition by E2C. (A) Illustration of free energy as a function of the similarity to the native complex (Q_n, X-Axis) and to I_late (Q_t, Y-Axis), at conditions where both the native complex and the unbound reagents are populated. Low free energy regions are in Blue, high free energy regions in Red. The free energy minimum for the unbound reagents (U) is at the lower left corner of the plot, the native complex (N) is at the lower right corner and I_late at the upper left corner. The two-state kinetic route takes place along the x-axis with the formation of native-like contacts. The multistate kinetic route takes place first along the y-axis with the formation of nonnative contacts in the encounter complex and I_late, and then along the diagonal with the simultaneous disruption of nonnative contacts and formation of native-like ones to reach the final complex. (B) Double funnel projection of the free energy landscape (36). The width of the plot represents the conformational entropy of the E2C:DNA system, and its depth represents the energy of interaction between the two molecules. The top of the picture corresponds to non-interacting E2C and DNA, and the two funnels group native-like and nonnative protein-DNA interactions. The two-state kinetic route takes place in the minimally frustrated native funnel (Right), centered on the free energy minimum for the native complex. The multistate kinetic route maps to a frustrated nonnative funnel (Left) that is centered on I_late. Conversion of this intermediate into the global minimum requires the E2C:DNA complex to switch to the native funnel over . Dissociation of the native complex may take place over TSE^2-state or .