Allosteric communication in DNA polymerase clamp loaders relies on a critical hydrogen-bonded junction

  1. Subu Subramanian
  2. Kent Gorday
  3. Kendra Marcus
  4. Matthew R Orellana
  5. Peter Ren
  6. Xiao Ran Luo
  7. Michael E O'Donnell
  8. John Kuriyan

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This paper will be of interest not only to scientists working in the primary field of DNA replication, but also to molecular biologists employing deep mutagenesis as well as structural biologists interested in the functions of the broader class of AAA+ ATPase molecular machines. The work examines relationships between protein sequence, structure and function in the bacteriophage T4 clamp-clamp loader complex, a highly studied AAA+ ATPase that deposits ring-shaped proteins onto DNA to support DNA polymerase processivity and DNA replication. The clamp loader system is revealed to have a high tolerance to amino acid substitution, with little correlation between permitted substitutions and phylogenetic variation. A hitherto unrecognized residue in the clamp loader, which appears to be shared among certain AAA+ ATPase members, is identified as critical for the maintenance of a functional structure and for allosteric coupling. The key claims of the paper are well supported by the data presented, and the employed methodology has undergone rigorous validation. Although a few control studies are still needed, this is a novel and significant paper overall.

    (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 See related articles

Listed in

Log in to save this article

Abstract

Clamp loaders are AAA+ ATPases that load sliding clamps onto DNA. We mapped the mutational sensitivity of the T4 bacteriophage sliding clamp and clamp loader by deep mutagenesis, and found that residues not involved in catalysis or binding display remarkable tolerance to mutation. An exception is a glutamine residue in the AAA+ module (Gln 118) that is not located at a catalytic or interfacial site. Gln 118 forms a hydrogen-bonded junction in a helical unit that we term the central coupler, because it connects the catalytic centers to DNA and the sliding clamp. A suppressor mutation indicates that hydrogen bonding in the junction is important, and molecular dynamics simulations reveal that it maintains rigidity in the central coupler. The glutamine-mediated junction is preserved in diverse AAA+ ATPases, suggesting that a connected network of hydrogen bonds that links ATP molecules is an essential aspect of allosteric communication in these proteins.

Article activity feed

  1. Version published to 10.7554/elife.66181 on eLife
  2. eLife

    Author Response:

    We thank the editors and the reviewers for their positive assessment of our work.

    Reviewer #1 (Public Review):

    [...] One major concern is that the levels of protein expression and folding are not verified. This is concerning for the Gln118 mutation because lack of fitness could result trivially from misfolding or accelerated degradation that might result from increased flexibility and conformational stability. Moreover, the authors' finding that it was not possible to purify Gln118 mutant proteins for biochemical studies is consistent with this sort of trivial explanation for apparent lack of biological function.

    As described in the manuscript, the sidechain of Gln 118 makes hydrogen bonds with the backbone segment leading into an adjacent helix. We had omitted to point out in the original manuscript that Gln 118 is completely buried in thestructure (we now do so in the revised manuscript, on page 29). As shown by Worth and Blundell, buried polar sidechains that form backbone hydrogen bonds (as Gln 118 does) are highly conserved in proteins, and these polar sidechains are important for the stabilization of the protein architecture (Worth and Blundell BMC Evolutionary Biology 2010, 10:161). Thus, we do expect the mutation of Gln118 to destabilize the clamp loader structure. However, we do not find the identification of the importance of Gln118 to be a trivial finding, because the role of polar residues in maintaining structure is quite commonly linked to their functional role, making it difficult to separate the two effects. For example, the proximal histidine that links the F helix in hemoglobin to the iron atom is perhaps the most important residue for allosteric communication in hemoglobin. Mutation of the proximal histidine severely destabilizes hemoglobin, due to loss of heme binding and conversion to a molten globule state (see, for example, Brennan and Matthews, Hemoglobin, 21:393-403, 1997).

    It was an oversight for us to have not analyzed the effects of Q118 mutations on stability and function, and we have now rectified this. We now include the results of the following four experiments, in which we compare the expression of the mutant and wild-type forms of the clamp loader, their behavior on gel filtration analysis, and their activities in ATPase assays and DNA replication assays. These experiments demonstrate that the mutation most likely destabilizes the protein, and affects the nature of the assembled complex. These results further emphasize the crucial nature of the hydrogen-bonding interactions made by the Gln 118 sidechain.

    1. We created a clamp loader variant in which the ATPase subunit is C-terminally tagged with the fluorescent protein mCherry, allowing the expression levels of the proteins to be monitored by flow cytometry of E. coli cells. This experiment shows that introduction of the Q118N mutation leads to a very substantial reduction in protein expression (Figure 6 supplement 2 in the revised manuscript). An important point is that the proteins are expressed using a strong promoter (T7 RNA polymerase promoter), which was done so as to purify proteins for biochemical experimentation and also enable ready detection of the mCherry fluorescence. The natural T4 promoter that is used in the phage assay results in very low levels of protein expression (no detectable fluorescence signal when mCherry is fused to the ATPase subunit), and we do not know whether the expression defect that we see is also manifested under conditions where the protein expression is low. Nevertheless, the data do indicate that the Q118N mutation destabilizes the clamp loader complex.

    2. We purified mCherry tagged variants of the wild-type clamp-loader complex, the Q118N mutant complex, and the Q118N/I141L double mutant that has partial recovery of fitness in the phage propagation assay. SDS-PAGE analysis (not shown) confirms that all complexes have the ATPase and clasp subunits of the clamp loader in the proper 4:1 ratio. Gel filtration analysis shows that the wild-type complex corresponds to a single peak eluting at ~70 ml, which we assume corresponds to correctly assembled clamp loader (see Figure 9 supplement 1 in the revised manuscript). For both mutants, there is a peak at ~70 ml, corresponding to the properly assembled clamp loader, but also an additional peak that is close to the void volume of the column (~45 ml). For the Q118N mutant, the fraction of the protein corresponding to the properly assembled clamp loader is small. This fraction is substantially larger for the double mutant that has increased fitness (Q118N/I141L), indicating that one effect of the second mutation is to recover the ability of the clamp loader to assemble properly.

    3. We measured the rates of DNA-stimulated ATP hydrolysis for purified and mCherry-tagged wild-type clamp loader and the Q118N mutant, as we had described in the original manuscript for several other mutants (Figure 9 supplement 2 in the revised manuscript). Addition of the mCherry tag to the wild-type clamp loader results in a slight reduction of the ATPase activity. The Q118N mutation has a very low rate of DNA-stimulated ATPase activity (less than 10% of activity of the wild-type mCherry-tagged clamp loader). These data indicate that even in the fraction of Q118N mutant that can be purified as part of an intact clamp loader complex, the mutation compromises the ability to hydrolyze ATP. This is likely to be due to the failure to assemble into a competent conformation.

    4. We measured the extent of plasmid DNA replication by the T4 replisome, using wild-type and mutant clamp loaders, as described in the original manuscript (Figure 9 supplement 3 in the revised manuscript). As for the ATPase assay, addition of the mCherry tag to the wild-type clamp loader results in a slight reduction of replication efficiency. Introduction of the Q118N mutation leads to a near-total loss of replication efficiency, to a level comparable to that seen in the absence of the clamp loader.

    The main text of the manuscript now includes a description of these new results, and the new data are included as supplementary figures.

    Reviewer #2 (Public Review):

    [...] One potential weakness is that among all the questions posed at the beginning of the study not all received a definitive answer. In particular, the question "To what extent does the mutational sensitivity of the system in a particular organism, carrying out the essential function of DNA replication, reflect the sequence diversity seen across the spread of life?" is only partially addressed.

    We agree that we have not fully probed the issue of evolutionary sequence diversity versus mutational sensitivity. We find it exciting that the two sets of data show clear divergence in certain positions, pointing to epistasis. This will be an exciting direction for future exploration.

    The second question, "The clamp loader subunits respond cooperatively to the clamp, ATP and DNA. How do the mechanisms underlying this cooperativity impose constraints on the sequence?", has not been answered and goes beyond the scope of this study.

    Answering this question requires the analysis of second-order mutations. We have demonstrated the feasibility of using the phage propagation assay for such analysis, but we agree with the reviewer that this is beyond the scope of the present study.

  3. eLife

    Reviewer #2 (Public Review):

    This is paper constitutes an experimental tour de force in understanding bacteriophage T4 replication. The T4 replication system has served as a model for elucidating universal DNA replication mechanisms. Specifically, in this study a new platform for deep mutagenesis was developed, validated and successfully applied to yield a complete profile of mutationally sensitive sites in the DNA polymerase clamp loader gp62 and the DNA sliding clamp gp45. The platform supports high-throughput testing of mutations in replication genes for functional fitness and could be adapted to enable future in-vitro evolution studies of the replication proteins. The mutational profile, along with sequence conservation analysis, demonstrates that clamp loader residues in the AAA+ modules exhibit high tolerance to mutation. Mutationally sensitive residues appear to be directly involved in either ATP hydrolysis or DNA binding. The residue Gln118 was the one notable exception, being distal from both the active site and the DNA. Subsequent detailed molecular modeling and structural analyses establish a structural basis for the observed Gln118 sensitivity. Notably, Gln 118 participates in a critically important hydrogen bond network linking the ATP active sites around the circumference of the clamp loader and likely plays a role in allosteric communication during the clamp loading cycle. Mutation of Gln118 disrupts this network and affect the structural rigidity of an element of the clamp loader termed the central coupler. Function restoration by a second-site suppressor mutation clearly establishes the functional importance of this previously unanticipated mechanism.

    Overall, the manuscript makes progress on a topic clearly important to the DNA replication field. The findings are novel and well supported by the data. In particular, the molecular dynamics simulations and analysis appear to have been done using appropriate simulation protocols. Both the experimental and computational methods are described in sufficient detail. Approaching T4 replication from multiple angles, using multiple experimental and computational techniques is a notable strength of this manuscript.

    One potential weakness is that among all the questions posed at the beginning of the study not all received a definitive answer. In particular, the question "To what extent does the mutational sensitivity of the system in a particular organism, carrying out the essential function of DNA replication, reflect the sequence diversity seen across the spread of life?" is only partially addressed. The second question, "The clamp loader subunits respond cooperatively to the clamp, ATP and DNA. How do the mechanisms underlying this cooperativity impose constraints on the sequence?", has not been answered and goes beyond the scope of this study.

    On the technical side, more rigorous analysis of the molecular simulations performed as part of the study would be welcome. In particular, quantifying the effects Gln118 on dynamics and on the rigidity of the central coupler could have used additional analysis.

  4. eLife

    Reviewer #1 (Public Review):

    The authors sought to understand the relationship between sequence conservation and biological function for a protein complex that undergoes conformational changes during its functional cycle. This included understanding the extent to which phylogenetic comparisons can guide identification of functionally important residues for a specific family member. The particular focus was on the bacteriophage clamp-clamp loader complex, with the long-term goal of understanding structure-function principles that might facilitate the design of novel AAA+ ATPase proteins.

    A systematic mutagenesis screen was used to determine the relative fitness for single site mutations throughout the bacteriophage T4 clamp-loader and clamp proteins. A feature of the screen is selection for fitness in an (almost) authentic biological context. The vast majority of residues are highly permissive to substitution, which is notable because the considerable conformational changes required during the loader-clamp reaction cycle might have been expected to place more constraints throughout the structure. It is demonstrated that tolerance to mutation within the T4 proteins does not correlated well with conservation across 1,000 other bacteriophage sequences, thereby illustrating the importance of specific context and limits of inference from phylogenetic comparisons. The only critical residue distant from a catalytic active site or binding surface is a glutamine, whose importance was not previously noted, but imparts rigidity to the structure by coupling functionally-important clusters through a hydrogen bonded network. Inspection of distantly related AAA ATPases indicates that this residue is important for many, although not all members of this large and diverse family of molecular machines.

    One major concern is that the levels of protein expression and folding are not verified. This is concerning for the Gln118 mutation because lack of fitness could result trivially from misfolding or accelerated degradation that might result from increased flexibility and conformational stability. Moreover, the authors' finding that it was not possible to purify Gln118 mutant proteins for biochemical studies is consistent with this sort of trivial explanation for apparent lack of biological function.

  5. eLife

    Evaluation Summary:

    This paper will be of interest not only to scientists working in the primary field of DNA replication, but also to molecular biologists employing deep mutagenesis as well as structural biologists interested in the functions of the broader class of AAA+ ATPase molecular machines. The work examines relationships between protein sequence, structure and function in the bacteriophage T4 clamp-clamp loader complex, a highly studied AAA+ ATPase that deposits ring-shaped proteins onto DNA to support DNA polymerase processivity and DNA replication. The clamp loader system is revealed to have a high tolerance to amino acid substitution, with little correlation between permitted substitutions and phylogenetic variation. A hitherto unrecognized residue in the clamp loader, which appears to be shared among certain AAA+ ATPase members, is identified as critical for the maintenance of a functional structure and for allosteric coupling. The key claims of the paper are well supported by the data presented, and the employed methodology has undergone rigorous validation. Although a few control studies are still needed, this is a novel and significant paper overall.

    (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.)

  6. Version published to 10.1101/2020.12.31.424921v2 on bioRxiv
  7. Version published to 10.1101/2020.12.31.424921v1 on bioRxiv