A new insight into RecA filament regulation by RecX from the analysis of conformation-specific interactions

  1. Aleksandr Alekseev
  2. Georgii Pobegalov
  3. Natalia Morozova
  4. Alexey Vedyaykin
  5. Galina Cherevatenko
  6. Alexander Yakimov
  7. Dmitry Baitin
  8. Mikhail Khodorkovskii

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This paper is of interest to readers in the fields of DNA repair, DNA-protein interactions and those employing single-molecule techniques. Using single-molecule methods, the authors discovered that RecX exerts its regulatory effect on the RecA filament through two modes of action: i) by promoting RecA dissociation from ssDNA, and ii) by causing a reversible conformational change of the filament. The latter mode of RecX action is novel and of particular interest. The authors present a plausible model of the RecX-RecA-ATP-ssDNA system that could be further validated in future experiments.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

RecA protein mediates homologous recombination repair in bacteria through assembly of long helical filaments on ssDNA in an ATP-dependent manner. RecX, an important negative regulator of RecA, is known to inhibit RecA activity by stimulating the disassembly of RecA nucleoprotein filaments. Here we use a single-molecule approach to address the regulation of ( Escherichia coli ) RecA-ssDNA filaments by RecX ( E. coli ) within the framework of distinct conformational states of RecA-ssDNA filament. Our findings revealed that RecX effectively binds the inactive conformation of RecA-ssDNA filaments and slows down the transition to the active state. Results of this work provide new mechanistic insights into the RecX-RecA interactions and highlight the importance of conformational transitions of RecA filaments as an additional level of regulation of its biological activity.

Article activity feed

  1. Version published to 10.7554/elife.78409 on eLife
  2. eLife

    Author Response

    Reviewer #1 (Public Review):

    Weakness:

    I do not believe that the data support the statement given in lines 396-400. The authors state that RecX interaction with the apo form of the RecA-ssDNA filament inhibits the transition to the ATP-bound state, which I believe is supported by their data that transition into the ATP-bound state is delayed following incubation with RecX. However, they go on to say that this is in line with previous reports which show that RecX blocks ATP hydrolysis by RecA. I think rather that their data suggests that RecX binds to inactive RecA and slows down binding of ATP by RecA. This would be in line with their hypothesis that RecX binding between monomers inhibits the cooperativity between the RecA monomers and slows down the apo-ATP transition (lines 440-442).

    It is our understanding that the Reviewer does not agree that our data is in line with the inhibition of RecA ATPase activity by RecX.

    RecA is a DNA-dependent ATPase and performs ATP hydrolysis when it is bound to DNA. The inhibition of RecA ATPase by RecX was extensively reported before [see ref 39 for example]. According to the available reports, the underlying mechanism was that RecX stimulates net disassembly of RecA-DNA complexes and thus reduce the overall rate of ATP hydrolysis. In this way the inhibition of RecA ATPase is not direct, but it is mediated by RecX, which, as it was thought, stimulates the disassembly of the RecA-DNA filaments.

    Our data indicate that RecX not only stimulates the disassembly of RecA-DNA filaments but also reduces the fraction of active ATP-bound states within RecA-ssDNA filament. Since ATP hydrolysis event requires active ATP-bound RecA unit, the decrease in the amount of active states within the RecA-ssDNA filament should lead to the reduction of ATP turnover and, as consequence, the lower overall rate of ATP hydrolysis would be observed.

    We agree that our data do not support the statement that RecX directly blocks ATP hydrolysis by RecA. We believe that RecX acts rather indirectly by reducing the fraction of active ATP-bound states within the RecA-ssDNA filament. To rule out misunderstanding and to adequately address the comment of the Reviewer the following changes in the text have been introduced:

    Original version: “This is in line with the reported inhibition of the RecA ATPase activity by RecX [39] and directly shows that RecX can effectively block ATP hydrolysis by preventing the corresponding conformational transitions within the RecA-ssDNA filament without actual displacement of RecA monomers from ssDNA as proposed in [25].”

    Revised version: “This is in line with the reported inhibition of the RecA ATPase activity by RecX measured in bulk [39] since depletion of the ATP-bound conformation within the filament should effectively reduce the overall rate of ATP hydrolysis. Interestingly, the possibility of RecX to inhibit RecA ATPase without actual displacement of RecA monomers from ssDNA was proposed previously in [25].”

    We consider the suggestion of the Reviewer, “I think rather that their data suggests that RecX binds to inactive RecA and slows down binding of ATP by RecA.” as one possible mechanism of how RecX retards the apo-ATP transition of the RecA-ssDNA filament. However, our results do not allow elucidating whether RecX slows down binding of ATP by RecA filament or RecX binding directly prevents the conformational change without affecting ATP binding by RecA. Interestingly, a rather elegant mechanism in which RecX prevents conformational changes of RecA filament directly (without affecting ATP binding) was proposed previously based on the results of electron microscopy studies [ref 25 in the manuscript]. Corresponding discussion concerning this mechanism was added in the Discussion:

    “It is noteworthy that previous electron microscopy studies provide a possible explanation of how RecX binding hampers the apo-ATP transition of the RecA filament. It was shown that the conformational change of the filament is accompanied by a large movement of RecA’s C-terminal domain, which is supposed to be allosterically coupled to the ATPase site [7]. According to low resolution electron microscopy studies, RecX binds from the C-terminal domain of one RecA subunit to the core domain of another [25]. Thus it was proposed that RecX inhibits RecA ATPase activity by preventing conformational transition through clamping RecA’s C-terminal domain. Although the proposed mechanism is in line with the results of the current study, we believe that additional research is required to elucidate the mechanistic basis of the RecX effect on the conformational transitions of the RecA-ssDNA filament.”

    Reviewer #2 (Public Review):

    In the last paragraph of the introduction, the authors describe their previous work, in which 3 mechanically distinct states of RecA-ssDNA filaments are identified. Yet in this paper the authors only refer to two mechanically distinct states: active (ATP-bound) and ap (ATP hydrolyzed). Is there a role for a third state in the model to describe these experiments? This should be addressed.

    Thank you for this comment. Indeed, we reported previously that apo and ADP states are distinct by mechanical properties and stability. However binding of RecX to apo and ADP RecA-ssDNA filaments resulted in the similar slowdown of the following transition to the active state. Thus the interaction of RecX with ADP RecA-ssDNA filaments was only briefly addressed in the original version of the manuscript (Lines 287-289 in the original version). In order to address this point more extensively, we provided figure S4 in the Supplementary and made following corrections in the manuscript:

    Original version: “Similar results were obtained for the interaction of the RecX with the ADP-bound state of RecA-ssDNA filament (data not shown).”

    Revised version: “We also examined the interaction of RecX with ADP state of the RecA-ssDNA filament. Recently, it was shown that ADP and apo conformations represent two distinct inactive states of RecA-ssDNA filament [18]. Incubation of ADP-bound form of the RecA-ssDNA filament with RecX also resulted in the slowdown of the following decompression (Figure S4) similarly to the RecX interaction with apo RecA-ssDNA filament. Interestingly, ADP-bound RecA-ssDNA filaments exhibited greater stability when supplemented with RecX”

    Original version: “Unexpectedly, we discovered that RecX interacts with the compressed apo form of the RecA-ssDNA filament and inhibits its transition into the ATP-bound state.”

    Revised version: “Unexpectedly, we discovered that RecX interacts with the ADP and apo forms of the RecA-ssDNA filament and inhibits their transition into the ATP-bound state.”

    The presented model shows RecX binding specifically to inactive (ATP hydrolyzed) RecA proteins, reasoning that even in the presence of free ATP, patches of inactive RecA will be available for RecX binding. Thus, the model should be sensitive to the fraction of RecA units in the inactive state at equilibrium, which is not altered systematically in the described experiments. This inactive fraction would be determined by the balance of the rate of ATP hydrolysis and ATP binding. The latter could be altered by adjusting the concentration of free ATP in buffer before the introduction of RecX, with the model predicting shortening should be faster at lower ATP concentrations (RecX binding enhanced). Alternatively, the use of ATP analog ATP-gamma-S, which resists hydrolysis and stabilizes RecA filaments, should inhibit RecX binding and compaction according to the model. At least one of these experiments would help to validate the proposed model.

    We carried out the experiments with ATP-gamma-S, suggested by Reviewers. Figure 4 was updated, and the text covering these results was added to the manuscript:

    “We also assesed RecXmNG binding to the active form of RecA-ssDNA using non-hydrolyzable ATP analog, ATPγS. The RecA-ssDNA filament was formed in the presence of 0.5 mM ATPγS, followed by incubation in the channel containing 1 μM RecXmNG and 0.5 mM ATPγS for 30 seconds and then was visualized in the channel containing 0.5 mM ATPγS and no proteins. As a result, average intensity of the tether was close to the background level (Figure 4D) indicating that RecXmNG did not remain bound to the active RecA-ssDNA filament. Thus we suppose that RecX interaction with the active form of the RecA-ssDNA filament is much weaker compared to the binding of RecX to the apo state. Interestingly, in the presence of ATPγS RecX did not induce any shortening of the RecA-ssDNA filaments (Figure S5) indicating the essential role of ATP hydrolysis in the RecX induced destabilization of RecA-ssDNA filaments.”

  3. eLife

    Evaluation Summary:

    This paper is of interest to readers in the fields of DNA repair, DNA-protein interactions and those employing single-molecule techniques. Using single-molecule methods, the authors discovered that RecX exerts its regulatory effect on the RecA filament through two modes of action: i) by promoting RecA dissociation from ssDNA, and ii) by causing a reversible conformational change of the filament. The latter mode of RecX action is novel and of particular interest. The authors present a plausible model of the RecX-RecA-ATP-ssDNA system that could be further validated in future experiments.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

  4. eLife

    Reviewer #1 (Public Review):

    In this paper, Pobegalov et al. use single-molecule C-trap technology to examine the dynamics of RecX interaction with distinct conformational states of RecA-ssDNA filaments. Using a multi-channel microfluidic flow cell, the authors demonstrate the biphasic disassembly of E. coli RecA-ssDNA nucleoprotein filaments by E. coli RecX. The authors trap ssDNA between two beads, assemble RecA onto this filament in a buffer containing ATP and RecA, and then while in a force clamp mode can move this filament into buffer containing RecX and observe change in bead distance as a read-out for nucleoprotein filament changes. With this method the authors demonstrate that RecX, in a concentration-dependent manner, leads to an initial steep decrease in filament length, and then a gradual reduction in bead distance (filament disassembly) over about 200s. This is in line with previous single-molecule studies done with RecA and RecX homologues in Mt and Bs.

    The authors then go on to show that the RecX-mediated disassembly is reversible to a degree. Incubation of RecA-ssDNA with RecX, followed by incubation with ATP leads to an increase in filament length, albeit not a full restoration of length, indicating that the alteration of the RecA filament is comprised of both reversible (compression) and irreversible (disassembly) mechanisms. They also show greater incubation times with RecX leads to a slower decompression rate of the filament, which indicated a prevention of RecA switching between active (ATP-bound) and inactive (ADP-bound or apo) forms. Through creation of a fluorescently tagged RecX they also demonstrate further evidence that RecX binds the apo form of RecA-ssDNA nucleoprotein filament and that upon association with ATP (or switch to the active form) RecX gradually dissociates from the filament. And finally, the authors demonstrate that RecA-dsDNA filaments are rapidly and irreversibly disassembled by RecX, with no restoration of filament length upon incubation in buffer lacking RecX.

    Strength:

    Overall I believe that the data presented here accurately supports the authors interpretations and that they have presented novel findings regarding regulation of RecA-DNA nucleoprotein filaments via RecX.

    Weakness:

    I do not believe that the data support the statement given in lines 396-400. The authors state that RecX interaction with the apo form of the RecA-ssDNA filament inhibits the transition to the ATP-bound state, which I believe is supported by their data that transition into the ATP-bound state is delayed following incubation with RecX. However, they go on to say that this is in line with previous reports which show that RecX blocks ATP hydrolysis by RecA. I think rather that their data suggests that RecX binds to inactive RecA and slows down binding of ATP by RecA. This would be in line with their hypothesis that RecX binding between monomers inhibits the cooperativity between the RecA monomers and slows down the apo-ATP transition (lines 440-442).

  5. eLife

    Reviewer #2 (Public Review):

    In this study, optical tweezers are used to isolate a single ssDNA molecule filamented by RecA-ATP. Shortening of the substrate is observed in the presence of RecX. This shortening occurs on two distinct timescales and is partially reversible, suggesting RecX both reversibly compacts the RecA filament upon binding and irreversibly promotes depolymerization of the RecA filament. In contrast, while RecX did not shorten the substrate in the absence of free ATP (apo state), the bound RecX inhibited subsequent elongation when RecA and ATP were reintroduced. Taken together, the different experimental combinations presented help test the validity and completeness of previously proposed models, resulting in a proposed model in which RecX regulates filamentation by binding to RecA in an inactive (ATP hydrolyzed) state. Overall, the presented experiments are of great value to further understand the system, though the final presented model makes one critical an assumption that could likely be tested and validated in an additional experiment using the same experimental apparatus and procedures described.

    In the last paragraph of the introduction, the authors describe their previous work, in which 3 mechanically distinct states of RecA-ssDNA filaments are identified. Yet in this paper the authors only refer to two mechanically distinct states: active (ATP-bound) and ap (ATP hydrolyzed). Is there a role for a third state in the model to describe these experiments? This should be addressed.

    The presented model shows RecX binding specifically to inactive (ATP hydrolyzed) RecA proteins, reasoning that even in the presence of free ATP, patches of inactive RecA will be available for RecX binding. Thus, the model should be sensitive to the fraction of RecA units in the inactive state at equilibrium, which is not altered systematically in the described experiments. This inactive fraction would be determined by the balance of the rate of ATP hydrolysis and ATP binding. The latter could be altered by adjusting the concentration of free ATP in buffer before the introduction of RecX, with the model predicting shortening should be faster at lower ATP concentrations (RecX binding enhanced). Alternatively, the use of ATP analog ATP-gamma-S, which resists hydrolysis and stabilizes RecA filaments, should inhibit RecX binding and compaction according to the model. At least one of these experiments would help to validate the proposed model.

  6. eLife

    Reviewer #3 (Public Review):

    This study by Alekseev et all focuses on the RecA nucleoprotein filament dynamics and its regulation by RecX protein. RecA and its eukaryotic homolog RAD51 is the key protein of recombination and DNA repair in all organisms including humans. Understanding the mechanism of RecA action and regulation is an important task. In spite of extended studies, the dynamics of the RecA nucleoprotein filament and its transition from active to inactive form is still poorly understood. The authors use a single-molecule approach to investigate the effect of RecX on the conformational status of the RecA nucleoprotein filament. This effect was measured by a change in the RecA filament length as a readout of a transition between the active and inactive filament conformation. The authors found that RecX exerts its regulatory effect through two modes of action: i) by promoting RecA dissociation from ssDNA and ii) causing a reversible conformational change of the filament. The later mode of RecX action is novel and is of particular interest. The authors suggest that RecX binds to and stabilizes the apo form of RecA causing RecA filament inactivation and delaying its reactivation in the presence of ATP. They used fluorescently labeled RecX to confirm their model. Overall, this is an interesting study.

  7. Version published to 10.1101/2022.03.14.484239v1 on bioRxiv