Global conformational dynamics of a Y-family DNA polymerase during catalysis

Abstract

Replicative DNA polymerases are stalled by damaged DNA while the newly discovered Y-family DNA polymerases are recruited to rescue these stalled replication forks, thereby enhancing cell survival. The Y-family DNA polymerases, characterized by low fidelity and processivity, are able to bypass different classes of DNA lesions. A variety of kinetic and structural studies have established a minimal reaction pathway common to all DNA polymerases, although the conformational intermediates are not well defined. Furthermore, the identification of the rate-limiting step of nucleotide incorporation catalyzed by any DNA polymerase has been a matter of long debate. By monitoring time-dependent fluorescence resonance energy transfer (FRET) signal changes at multiple sites in each domain and DNA during catalysis, we present here a real-time picture of the global conformational transitions of a model Y-family enzyme: DNA polymerase IV (Dpo4) from Sulfolobus solfataricus. Our results provide evidence for a hypothetical DNA translocation event followed by a rapid protein conformational change prior to catalysis and a subsequent slow, post-chemistry protein conformational change. Surprisingly, the DNA translocation step was induced by the binding of a correct nucleotide. Moreover, we have determined the directions, rates, and activation energy barriers of the protein conformational transitions, which indicated that the four domains of Dpo4 moved in a synchronized manner. These results showed conclusively that a pre-chemistry conformational change associated with domain movements was too fast to be the rate-limiting step. Rather, the rearrangement of active site residues limited the rate of correct nucleotide incorporation. Collectively, the conformational dynamics of Dpo4 offer insights into how the inter-domain movements are related to enzymatic function and their concerted interactions with other proteins at the replication fork.

Figure 1. Front and back views of domain motions during a single, correct nucleotide incorporation.

The domains of Dpo4 are shown in blue (finger), red (palm), green (thumb), and purple (LF); the DNA is in gold; all of the nine mutant residues (Table S1) are in yellow and the Alexa488-labeled DNA base is in cyan. The arrows represent the direction of residue movement based on the FRET signals for phase P₁ (black) and phase P₂ (white).

Figure 2. Steady-state fluorescence spectra of finger domain mutant (N70C) at 20°C.

Alexa488-labeled DNA (100 nM, black trace) was excited at a wavelength of 493 nm. The sequential addition of Alexa594-labeled Dpo4 (600 nM) and dTTP (1 mM) produced the red and green traces, respectively. Spectra were normalized to 1 by using the donor as a reference. Emission spectra are shown for both (A) S-1 and (B) S-2 DNA substrates.

Figure 3. Stopped-flow kinetics of dTTP incorporation into S-1 DNA at 20°C.

Dpo4 mutant•S-1 DNA complexes were reacted with dTTP and fluorescence was monitored using a stopped-flow apparatus. Donor (green) and acceptor (red) traces are shown for (A) the finger (N70C), (B) palm (S112C), (C) thumb (S207C), and (D) LF (K329C) domains. Data for finger (E49C), palm (S96C and N130C), thumb (K172C), and LF (R267C) residues are shown in Figure S3. Each Dpo4 mutant (Table S1) and S-1 were labeled with Alexa594 and Alexa488, respectively. Notably, some changes in fluorescence upon dTTP binding occurred during the instrument's dead time and the donor and acceptor fluorescence signals at time zero or close to time zero were not recorded.

Figure 4. Stopped-flow kinetics of dTTP incorporation into S-2 DNA at 20°C.

Dpo4 mutant•S-2 DNA complexes were reacted with dTTP and the fluorescence was monitored using a stopped-flow apparatus. Donor (green) and acceptor (red) traces are shown for the (A) finger (N70C), (B) palm (S112C), (C) thumb (S207C), and (D) LF (K329C) domains. Data for finger (E49C), palm (S96C and N130C), thumb (K172C), and LF (R267C) residues are shown in Figure S4. Each Dpo4 mutant (Table S1) and S-2 were labeled with Alexa594 and Alexa488, respectively. Notably, some changes in fluorescence upon dTTP binding occurred during the instrument's dead time and the donor and acceptor fluorescence signals at time zero or close to time zero were not recorded.

Figure 5. Activation energy barrier for dTTP incorporation into S-1 DNA catalyzed by the S112C Dpo4 mutant.

The extracted rates of the P₁ (circle) and P₂ (square) phases were plotted as a function of reaction temperature to yield the activation energy barriers of 16.4±0.6 and 22±1 kcal/mol for P₁ and P₂, respectively.

Figure 6. Stopped-flow kinetics of domain-domain motions at 20°C.

A pre-incubated mixture of Dpo4 finger mutants (A) Y274W-N70C^CPM or (B) Y274W-K26C^CPM (200 nM) with either S-3 (black trace) or S-4 (red trace) DNA substrates (300 nM) was reacted with dTTP (1 mM). Only the fluorescence signals of acceptor CPM were recorded. Notably, some changes in fluorescence upon dTTP binding occurred during the instrument's dead time and the acceptor fluorescence signals at time zero or close to time zero were not recorded.

Figure 7. Mechanism of a single, correct nucleotide incorporation catalyzed by Dpo4.

P₀, P₁, and P₂ are the three phases observed in Figure 3. DNA* and DNA respectively represent the location of DNA in the active site of Dpo4 before and after DNA sliding by one base pair. E^apo, E, E′, and E″ represent four different conformations of Dpo4. PP_i denotes pyrophosphate.