Natural Polymorphisms D60E and I62V Stabilize a Closed Conformation in HIV-1 Protease in the Absence of an Inhibitor or Substrate

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors provide an interesting structural analysis of HIV PR that draws on consistent DEER EPR (experimental) and MD studies in order to illustrate that isolated D60E or I62V single mutations of B subtype HIV PR both lead to dramatic changes in the conformational distribution that correspond to a higher prevalence of more closed conformations that more closely resemble inhibitor-bound conformations.  Meanwhile the double-mutant exhibits a compensatory mechanism leading to a conformational distribution that overall more closely resembles the conformational landscape of the unmodified B subtype, but that still retains a not insignificant population of the closed conformation.  This explains earlier results where the F and H subtypes, which possess these mutations, demonstrate a significant population of more closed conformations.

More broadly, I believe this study demonstrates the power of combining experimental DEER studies with actual molecular dynamics simulations (rather than relying on MMM or other more rudimentary tools.)

Bearing in mind that I think that the core of this paper represents an important contribution that should be showcased, I would highlight several areas where I think the authors could quite quickly improve the paper:

For those not intimate familiarity with HIV PR, some very brief background in the intro that explains how the enzymatic activity of F and H subtypes and their response to inhibitors differs from the B subtype would prove useful.  Line 420-422 of the conclusion are the very first point in the manuscript where it is very clearly stated that this work is solving a problem set up in the previous work from this lab --> namely, that the F and H subtypes exhibit a significant population of closed conformations that the B subtype does not exhibit, and that we want to know which mutations are responsible for this difference.  This is actually stated in lines 67-73, but in a way that is very indirect/roundabout.  In a related issue, lines 77-81 give important information, but are not at all integrated into the broader narrative of the various subtypes and which conformations they adapt.  Along similar lines, lines 44-48 could more directly and simply state that one can attempt to target resistant strands of the virus by better understanding and then targeting the various open conformations of the HIV PR.

It seems that the authors have missed an opportunity to more directly connect the named conformations to the numbered conformations that are discovered by the MD results.  Specifically, Figure 8 shows 5 numbered conformations, and the authors don't commit to identifying these with the "curled/tucked," "closed," "semi-open," etc conformations that they talk about with respect to the DEER data.  The authors should at least clearly propose an assignment of the names vs. numbers for the different conformations.  In particular, note that it's implied that conformation 1 in figure 8 is alternately identified as "wide open" in 7C and "curled tucked" in 7B (assuming that 7B/C are essentially projections of 8A along the two axes.)

In Fig 6, there is some confusion generated by the fact that structures are shown from the crystal structures that do not match the corresponding MD data.  (Specifically -- note that the 4.0 Ang salt bridge in 6C (wt) does not show up in 6D).  In fact this highlights the potential drawbacks of looking exclusively at crystal structures, but a direct comment in the text would be very useful.

Line 108: please give the concentration of the MTSL in the reaction mixture.  In the literature, this number is typically given in terms of molar equivalents, as here, but for kinetics as well as for equilibrium populations, it's the concentration of MTSL that matters, not the number of molar equivalents.  Another confusing detail is why the pH is so high (line 111) -- this seems more like a solution NMR prep, and not clear why this is needed for DEER.

line 129-132: It would be good to know the tool that was used to incorporate the AA sidechain substitutions -- even if given in previous publications that are cited here. Also (142-147) it's not clear if, in the restrained minimizations that you discuss, the releasing of constraints are inclusive: e.g. after minimizing the position of all water molecules and counterions, do you freeze the position of the water molecules to minimize the hydrogen atoms and sidechains, or is this a progressive releasing of constraints.  (I would assume it's a progressive releasing of constraints, as that would be typical, but that's not clear from the way it's written).

In Fig 3A, the distance distribution for I62V is actually somewhat different in nature relative to both the other mutants as well as the MD, because it's not actually adapting a closed conformation (like D60E) or a distribution that includes both closed and semi-open (like the double mutant) -- rather, it's adapting a single roughly gaussian distribution between closed and semi-open.  Some commentary in the text would be appreciated.

line 172-180 -- the Gaussian distribution figures in the SI are quite nice, and I think you want to point to them here.

Figure 8B: These plots seem to use a relatively coarse rectangular histogramming followed by a linear grid interpolation that leads to cross-shaped artefacts and a couple dramatic hotspots that throw off your color scale.  Please consider switching to a plotting method that uses hexagonal binning, which are available for both matlab and python, and which would undoubtedly do a better job of displaying this data.

Comments on the Quality of English Language

There are various low-level grammatical, etc, issues that could use a close pass by a copy editor, or a careful pass by the PI.

Here is a somewhat random selection:

line 10: "more closed conformational landscape" should be "landscape of more closed conformations"

line 23: remove "is"

line 24-25: tense agreement

line 36: β not b

line 93: change to "cloned between the"

line 143,146: missing "be" in "allowed to be"

line 386: "minimal"

398, 408, 413: missing articles (a/the) -->  I think in other places, as well 

***the SI includes some MSWord comments***

Author Response

Please see attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Evaluating the manuscript entitled "Natural Polymorphisms D60E and I62V Stabilize a Closed Conformation in HIV-1 Protease in the Absence of Inhibitor or Substrate", the authors examined the effect of natural polymorphisms of the HIV-1 protease on viral activity. HIV infection is a worldwide health problem characterized by medication resistance and virological failure. When compared to Subtype B, natural polymorphisms (NPs) in African and Brazilian protease variations create a more closed conformational environment. When two common NPs, D60E and I62V, are injected, they can induce a closed conformation in HIV-1PR Subtype B. D60E modifies the conformation by establishing a salt bridge with K43, whereas I62V modifies the hydrophobic clusters of the cantilever and fulcrum, resulting in a more closed conformation.

The methodology applied by the authors is fully integrated and in my opinion their conclusions are quite safe. I propose to publish this paper after solving some minor problems:

1) Figure 1 should be replaced with another one, in which the amino acids can be seen better. You may need a new image next to the zoomed one.

 

2) In figure 9, the amino acids should also be added to show the differences of each form with the wild form.

Author Response

Please see attachment

Author Response File: Author Response.pdf