Proteins are the workhorses of the cell. The shape that a protein molecule adopts enables it to carry out its role. However, a protein’s shape, or 'conformation', is not static. Instead, a protein can shift between different conformations. This is particularly true for enzymes – the proteins that catalyze chemical reactions. The region of an enzyme where the chemical reaction happens, known as the active site, often has to change its conformation to allow catalysis to proceed. Changes in temperature can also make a protein shift between alternative conformations. Understanding how a protein shifts between conformations gives insight into how it works.

A common method for studying protein conformation is X-ray crystallography. This technique uses a beam of X-rays to figure out where the atoms of the protein are inside a crystal made of millions of copies of that protein. At room temperature or biological temperature, X-rays can rapidly damage the protein. Because of this, most crystal structures are determined at very low temperatures to minimize damage. But cooling to low temperatures changes the conformations that the protein adopts, and usually causes fewer conformations to be present.

Keedy, Kenner, Warkentin, Woldeyes et al. have used X-ray crystallography from a very low temperature (-173°C or 100 K) to above room temperature (up to 27°C or 300 K) to explore the alternative conformations of an enzyme called cyclophilin A. These alternative conformations include those that have previously been linked to this enzyme’s activity. Starting at a low temperature, parts of the enzyme were seen to shift from having a single conformation to many conformations above a threshold temperature. Unexpectedly, different parts of the enzyme have different threshold temperatures, suggesting that there isn’t a single transition across the whole protein. Instead, it appears the way a protein’s conformation changes in response to temperature is more complex than was previously realized. This result suggests that conformations in different parts of a protein are coupled to each other in complex ways.

Keedy, Kenner, Warkentin, Woldeyes et al. then performed X-ray crystallography at room temperature using an X-ray free-electron laser (XFEL). This technique can capture the protein’s structure before radiation damage occurs, and confirmed that the alternative conformations observed were not affected by radiation damage.

The combination of X-ray crystallography at multiple temperatures, new analysis methods for identifying and measuring alternative conformations, and XFEL crystallography should help future studies to characterize conformational changes in other proteins.

Current structural biology methods provide only incomplete pictures of how proteins interconvert between distinct conformations (Motlagh et al., 2014; van den Bedem and Fraser, 2015). Although X-ray crystallography reveals atomic coordinates with relatively high accuracy and precision, the resulting electron density maps contain contributions from multiple alternative conformations reflecting the ensemble average of 10⁶–10¹⁵ copies of the protein in one crystal (Rejto and Freer, 1996; Smith et al., 1986; Woldeyes et al., 2014). At high resolution, it is often possible to detect and discretely model these alternative conformations (Burnley et al., 2012; Davis et al., 2006; Lang et al., 2010; van den Bedem et al., 2009). Structural characterization of alternative conformations by X-ray crystallography can complement NMR (Baldwin and Kay, 2009; Fenwick et al., 2014) and computational simulations (Dror et al., 2012; Ollikainen et al., 2013) in defining the structural basis of protein dynamics and ultimately in linking dynamics to function (Henzler-Wildman and Kern, 2007).

However, more than 95% of crystal structures are determined at cryogenic temperatures (∼100 K) to reduce radiation damage by minimizing diffusion of reactive intermediates and chemical-damage-induced structural relaxations (Garman, 2010; Holton, 2009; Warkentin et al., 2013). Unfortunately, cryocooling can modify main chain and side chain conformational distributions throughout the protein, including at active sites and distal regions important for allosteric function (Fraser et al., 2011; Halle, 2004; Keedy, et al., 2014). Recent studies have instead used room temperature data collection to reveal a multitude of previously ‘hidden’ alternative conformations that are not evident at cryogenic temperatures, many of which have important ramifications for determining molecular mechanisms (Deis et al., 2014; Fraser et al., 2009; Fukuda and Inoue, 2015; van den Bedem et al., 2013).

Between these temperature extremes, protein conformational heterogeneity changes in complex ways. Previous studies using a wide variety of biophysical probes including NMR, X-ray crystallography, and neutron scattering have revealed a change in the character of conformational heterogeneity and/or protein dynamics around 180–220 K (Doster, 2010; Frauenfelder et al., 2009; Lewandowski et al., 2015; Ringe and Petsko, 2003) however, the molecular origins of this ‘glass’ or ‘dynamical’ transition remain incompletely understood. Classic work has examined the temperature dependence of protein conformational heterogeneity across individual X-ray structures determined at temperatures from ∼80 to 320 K (Frauenfelder et al., 1979, 1987; Tilton et al., 1992). These studies used atomic B-factors as a proxy for conformational heterogeneity and identified a global inflection point around 180–220 K. This inflection point was interpreted in terms of a transition driven by dynamical arrest of the coupled hydration layer-protein system (Doster et al., 1989; Frauenfelder et al., 1979, 1987; Tilton et al., 1992). By contrast, solution NMR studies of picosecond–nanosecond (ps–ns) timescale methyl side chain order parameters showed heterogeneous changes in motional amplitudes at temperatures between 288 and 346 K. Thermal deactivation of these motions was suggested to predict a transition near 200 K without invoking solvent arrest (Lee and Wand, 2001). Recent solid-state NMR (ssNMR) experiments suggest that protein motions are coupled to solvent, and that three transitions at ∼195, 220, and 250 K mark the onset of distinct classes of motions as temperature increases (Lewandowski et al., 2015). Unfortunately, these studies used either globally averaged data (as with ssNMR or neutron scattering) or imprecise atomic-level models of conformational heterogeneity (as with B-factors in X-ray crystallography or NMR order parameters), thus preventing an all-atom understanding of the complex temperature response of protein crystals.

New crystallographic and computational techniques now enable a more detailed investigation of the temperature dependence of protein conformational heterogeneity at the atomic level. First, the program Ringer (Lang et al., 2014, 2010) evaluates low-level electron density traditionally considered noise to uncover statistically significant ‘hidden’ alternative conformations, which may become populated or depopulated as a function of temperature. Second, multiconformer models with explicit alternative conformations of both backbone and side chain atoms, as created by manual building or methods such as the program qFit (Keedy et al., 2015, van den Bedem et al., 2009), can account for non-harmonic motions across separate energy wells (encoded by discrete alternative conformations with distinct occupancies and coordinates) and harmonic motions within energy wells (encoded by B-factors). Third, crystallographic order parameters (S²) weigh these harmonic and non-harmonic contributions in a single metric that quantifies the disorder of each residue in a multiconformer model, allowing direct comparison with NMR-determined order parameters (Fenwick et al., 2014). Finally, methodological advances based on the physics of ice formation have enabled variable-temperature crystallographic data collection at temperatures between 300 and 100 K with modest or no use of potentially conformation-perturbing cryoprotectants (Warkentin et al., 2012; Warkentin and Thorne, 2009). Together, these methods overcome many of the limitations of previous X-ray-based approaches and will contribute to an integrated view of how protein conformational heterogeneity and dynamics evolve with temperature.

The human proline isomerase cyclophilin A (CypA) is an excellent model system for deploying these tools to study the structural basis of functional conformational dynamics and, in particular, to use temperature to understand the extent of correlated motions during an enzyme’s catalytic cycle. Previous NMR relaxation data for CypA (Eisenmesser et al., 2005, 2002) indicated a single common exchange process, both in the apo state and during catalysis, for a network of dynamic residues extending from the core to the active site. Room temperature crystallography later suggested the precise alternative conformations that collectively interconvert during catalysis (Fraser et al., 2009). However, subsequent NMR relaxation experiments of mutants designed to perturb the dynamics suggested that multiple exchange processes occur within this network (Schlegel et al., 2009). Here, we analyze multitemperature synchrotron experiments to examine the temperature-dependent conformational heterogeneity of CypA. Additionally, we report X-ray-free electron laser (XFEL) data, which are free of conventional radiation damage (Kern et al., 2014; Spence et al., 2012), to validate previous connections between alternative conformations determined by synchrotron crystallography and NMR experiments performed in solution (Eisenmesser et al., 2005; Fraser et al., 2009). Our analysis shows that the temperature dependence of alternative protein conformations is heterogeneous and that the character of this heterogeneity bridges previous models for protein dynamical transitions. Our results also suggest new ways to use variable temperature with both synchrotron and XFEL crystallography to probe the dynamic underpinnings of protein function.

Here, we have mapped the conformational landscape of the dynamic enzyme CypA by analyzing multiconformer models from multitemperature crystallography. Unlike previous temperature-dependent analyses of X-ray crystallography, here we consider both harmonic disorder (B-factors) and non-harmonic displacements (alternative conformations and occupancies), characterized using crystallographic bond order parameters. We have four primary findings:

1. The occupancy of alternative side chain conformations present at 310 K generally decreases with decreasing temperature.

2. However, some residues adopt alternative conformations only at 180 K and below.

3. The temperature response of residues is highly heterogeneous: the ‘intersection temperature’ below which side chain conformational heterogeneity remains nearly constant has a broad distribution across the molecule, with a peak near 250 K.

4. Residues composing a dynamic active-site network that is important for function do not have a consistent intersection temperature.

Our results provide new insight into the relationship between energy landscapes, the glass transition, and protein function. Glasses and other disordered systems have complex energy landscapes, in part due to the large number of microenvironments and the extensive frustration that disorder generates in the intermolecular interactions. Proteins at biological temperatures are ‘glassy’ in this sense—they have complex energy landscapes, due to their large size, heterogeneous amino acid composition, and many degrees of conformational freedom (Frauenfelder et al., 1991). The extensive heterogeneity in the temperature response of individual residues in CypA that we observe here provides additional direct evidence for this underlying energetic heterogeneity.

In addition to the inherent ‘glassiness’ of proteins at biological temperatures, dynamical transitions, some of which have been called ‘glass transitions’, have been reported at lower temperatures, including 180, 200, 220, 240, and 250 K, based on Mössbauer spectroscopy, X-ray crystallography, liquid and solid-state NMR, neutron scattering, and other techniques (Lewandowski et al., 2015; Schiro et al., 2015). These transitions typically manifest as a change in slope of some measurement in the vicinity of the suggested transition temperature. However, many of these measurements are sensitive to motions only within some timescale window (often ps-ns), monitor only a subset of amino acid types, and/or spatially average over all residues in the protein. By contrast, multitemperature crystallography with multiconformer models has many advantages by providing a time-independent and fully site-resolved measurement of ensemble-averaged atomic displacements, including both harmonic and discrete conformational heterogeneity, within the crystal.

This combined methodology lets us examine dynamical and glass transitions in protein crystals from a new perspective. Glass transitions are by definition non-equilibrium phenomena that arise when the kinetics of relaxation toward equilibrium slow so dramatically that equilibrium cannot be reached on experimental timescales. One signature of a true glass transition in proteins would be if occupancies of minority alternative conformations were arrested at non-zero values below some temperature. Indeed, here we see no appreciable temperature evolution of individual conformer occupancies or B-factors at 180 K and below, and the average disorder at these temperatures is far in excess of what is predicted from high-temperature extrapolations. These observations are consistent with the falling out of equilibrium expected in a glass transition, but not with a transition driven by the thermal freeze-out of alternative side chain conformations (Lee and Wand, 2001). Moreover, the persistence of multiple rotameric states at low temperatures is consistent with solvent arrest that impedes further changes in side chain disorder. Although the details of protein–solvent interactions may differ in crystals versus in solution, local variability at different protein–solvent interface microenvironments in the crystal (Teeter et al., 2001) likely contributes to the heterogeneity of temperature responses that we observe. The critical importance of site resolution is evident in the results of Figure 6. Averaging over side chain disorder in all residues yields an apparent transition near 250 K (Figure 6 and Figure 6—figure supplement 7). Perhaps coincidentally, this same temperature has been associated with a transition for protein side chains in site-averaged solid-state NMR measurements (Lewandowski et al., 2015). However, the ‘transition’ in our residue-averaged result obscures the highly heterogeneous temperature dependence of the individual side chains. We find no evidence of tight cooperativity or of a collective global response near 250 K that would be expected in the case of a ‘true’ dynamical transition. Instead, our data are consistent with local, non-cooperative freeze-out of conformational states defined by the energy landscape over a broad temperature range.

The heterogeneous response of side chain order parameters to temperature is driven largely by the changes to the populations of alternative conformations, which ‘flat-line’ at different temperatures across CypA. This diversity of ‘flat-lining’ temperatures is present even within the dynamic active-site network, even though the constituent residues have similar occupancies for their major and minor states at high temperatures (Figure 7). This result contrasts with previous NMR and X-ray experiments that hypothesized correlated motions of this network as rate-limiting for the catalytic cycle (Eisenmesser et al., 2005; Fraser et al., 2009) (Figure 8A).

Figure 8

Download asset Open asset

To bridge these views, we propose that the active-site network adopts two substates, which are primarily distinguished by Phe113 rotamer interconversion. Each of these substates adopts a differently weighted ensemble of conformations for other residues (Figure 8B). In this model, Met61 and Arg55 can switch rotamers more easily than Ser99 when Phe113 is in its χ1 gauche (p) rotamer pointed toward Ser99, whereas Ser99 can switch rotamers more easily than Met61 and Arg55 when Phe113 is in its χ1 gauche (m) rotamer pointed toward Met61. Additionally, thermal ‘breathing’ motions within rotameric wells may relieve minor steric overlaps within some of these macro- and microstates (Figure 8B , bottom right). This model is consistent with Phe113 having the lowest ‘flat-lining’ temperature of the network (Figure 7), and makes sense sterically because of the large size of the aromatic ring. These hypothesized motions are consistent with the timescales and temperature dependencies of motion assigned by solid-state NMR studies of crystalline protein GB1 (Lewandowski et al., 2015). Furthermore, the model helps explain the difficulty of fitting NMR relaxation data for perturbed versions of the active-site network as a single collective exchange process (Schlegel et al., 2009). The aromatic ring of Phe113 could play a dominant role in determining the chemical shift changes of the surrounding residues. Each of these residues could also populate multiple rotamers in the excited state measured by NMR. Our hierarchical perspective evokes the ‘population shuffling’ model of (Smith et al., 2015), in which a protein macrostate (in CypA, defined by the Phe113 rotamer) also determines the different relative populations of rotamers for a subset of other residues (in CypA, the other residues in the active-site network). In this model, the interconversion between macrostates, and not the collective motion of all residues between distinct rotamers, is correlated with the rate-limiting step of the CypA catalytic cycle.

Diversity in the temperature dependences of alternative conformations as we see here is inevitable given the limitations of the amino acid alphabet, yet its spatial pattern within a protein may provide insight into selective pressures. Evolutionary optimization must ensure that functionally important alternative conformations are robustly populated and interconvert appreciably at physiological conditions. However, the energy landscapes of individual residues are coupled to varying extents, such that some subsets of residues must be collectively optimized to preserve some, but not perfect, collectivity in functional motions. For proteins with large sequence alignments, evolutionary covariation has been used to predict ‘sectors’ of functionally cooperative residues, which are often dispersed in primary sequence but strikingly contiguous in tertiary structure (Halabi et al., 2009). By contrast, temperature-dependent crystallography has the potential to unveil couplings in atomic detail by identifying sets of residues whose conformational ensembles respond concertedly to temperature change. Based on our results with CypA, we expect this coupling to be weak, but measurable. Serial femtosecond XFEL crystallography combined with ultra-fast temperature jumps could enable a temporal view of these coupled conformational changes. Novel static and time-resolved multitemperature crystallographic approaches will provide powerful tools for resolving concerted motions to explore how proteins function and evolve.

1. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 100 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yug).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUG
2. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 150 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yuh).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUH
3. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 180 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yui).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUI
4. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 240 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yuj).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUJ
5. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 260 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yuk).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUK
6. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 280 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yul).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUL
7. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 300 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yum).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUM
8. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) Multiconformer synchrotron model of CypA at 310 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yun).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUN
9. 1. Keedy DA
   2. Kenner LR
   3. Warkentin M
   4. Woldeyes RA
   5. Thompson MC
   6. Brewster AS
   7. Van Benschoten AH
   8. Baxter EL
   9. Hopkins JB
   10. Uervirojnangkoorn M
   11. McPhillips SE
   12. Song J
   13. Mori RA
   14. Holton JM
   15. Weis WI
   16. Brunger AT
   17. Soltis M
   18. Lemke H
   19. Gonzalez A
   20. Sauter NK
   21. Cohen AE
   22. van den Bedem H
   23. Thorne RE
   24. Fraser JS

   (2015) High-resolution multiconformer synchrotron model of CypA at 273 K

   Publicly available at the RCSB Protein Data Bank (Accession no: 4yuo).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUO
10. 1. Keedy DA
    2. Kenner LR
    3. Warkentin M
    4. Woldeyes RA
    5. Thompson MC
    6. Brewster AS
    7. Van Benschoten AH
    8. Baxter EL
    9. Hopkins JB
    10. Uervirojnangkoorn M
    11. McPhillips SE
    12. Song J
    13. Mori RA
    14. Holton JM
    15. Weis WI
    16. Brunger AT
    17. Soltis M
    18. Lemke H
    19. Gonzalez A
    20. Sauter NK
    21. Cohen AE
    22. van den Bedem H
    23. Thorne RE
    24. Fraser JS

    (2015) Multiconformer fixed-target X-ray free electron (XFEL) model of CypA at 273 K

    Publicly available at the RCSB Protein Data Bank (Accession no: 4yup).

    http://www.rcsb.org/pdb/explore/explore.do?structureId=4YUP
11. 1. Fraser JS
    2. Fraser J

    (2015) X-Ray Diffraction data from Cyclophilin A, source of 4YUO structure

    Publicly available at the SRGrid Data Bank (Accession no: 10.15785/SBGRID/68).

    https://data.sbgrid.org/dataset/68/

1. 1. Ke H

   (1992) Similarities and differences between human cyclophilin a and other beta-barrel structures. Structural refinement at 1.63 angstroms resolution

   Publicly available at the RCSB Protein Data Bank (Accession no: 2CPL).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=2CPL

1. 1. Alcorn T
   2. Juers DH

   (2010) Progress in rational methods of cryoprotection in macromolecular crystallography

   Acta Crystallographica Section D Biological Crystallography 66:366–373.

   https://doi.org/10.1107/S090744490903995X

   • PubMed
   • Google Scholar
2. 1. Baldwin AJ
   2. Kay LE

   (2009) NMR spectroscopy brings invisible protein states into focus

   Nature Chemical Biology 5:808–814.

   https://doi.org/10.1038/nchembio.238

   • PubMed
   • Google Scholar
3. 1. Burnley BT
   2. Afonine PV
   3. Adams PD
   4. Gros P

   (2012) Modelling dynamics in protein crystal structures by ensemble refinement

   eLife 1:e00311.

   https://doi.org/10.7554/eLife.00311

   • PubMed
   • Google Scholar
4. 1. Butterfoss GL
   2. Richardson JS
   3. Hermans J

   (2005) Protein imperfections: separating intrinsic from extrinsic variation of torsion angles

   Acta Crystallographica Section D Biological Crystallography 61:88–98.

   https://doi.org/10.1107/S0907444904027325

   • PubMed
   • Google Scholar
5. 1. Chinte U
   2. Shah B
   3. Chen YS
   4. Pinkerton AA
   5. Schall CA
   6. Hanson BL

   (2007) Cryogenic (<20 K) helium cooling mitigates radiation damage to protein crystals

   Acta Crystallographica Section D, Biological Crystallography 63:486–492.

   https://doi.org/10.1107/S0907444907005264

   • PubMed
   • Google Scholar
6. 1. Cohen AE
   2. Ellis PJ
   3. Miller MD
   4. Deacon AM
   5. Phizackerley RP

   (2002) An automated system to mount cryo-cooled protein crystals on a synchrotron beam line, using compact sample cassettes and a small-scale robot

   Journal of Applied Crystallography 35:720–726.

   https://doi.org/10.1107/S0021889802016709

   • PubMed
   • Google Scholar
7. 1. Cohen AE
   2. Soltis SM
   3. González A
   4. Aguila L
   5. Alonso-Mori R
   6. Barnes CO
   7. Baxter EL
   8. Brehmer W
   9. Brewster AS
   10. Brunger AT
   11. Calero G
   12. Chang JF
   13. Chollet M
   14. Ehrensberger P
   15. Eriksson TL
   16. Feng Y
   17. Hattne J
   18. Hedman B
   19. Hollenbeck M
   20. Holton JM
   21. Keable S
   22. Kobilka BK
   23. Kovaleva EG
   24. Kruse AC
   25. Lemke HT
   26. Lin G
   27. Lyubimov AY
   28. Manglik A
   29. Mathews II
   30. Mcphillips SE
   31. Nelson S
   32. Peters JW
   33. Sauter NK
   34. Smith CA
   35. Song J
   36. Stevenson HP
   37. Tsai Y
   38. Uervirojnangkoorn M
   39. Vinetsky V
   40. Wakatsuki S
   41. Weis WI
   42. Zadvornyy OA
   43. Zeldin OB
   44. Zhu D
   45. Hodgson KO

   (2014) Goniometer-based femtosecond crystallography with x-ray free electron lasers

   Proceedings of the National Academy of Sciences of the United States of America 111:17122–17127.

   https://doi.org/10.1073/pnas.1418733111

   • PubMed
   • Google Scholar
8. 1. Davis IW
   2. Arendall WB
   3. Richardson DC
   4. Richardson JS

   (2006) The backrub motion: how protein backbone shrugs when a sidechain dances

   Structure 14:265–274.

   https://doi.org/10.1016/j.str.2005.10.007

   • PubMed
   • Google Scholar
9. 1. Deis LN
   2. Pemble CW
   3. Qi Y
   4. Hagarman A
   5. Richardson DC
   6. Richardson JS
   7. Oas TG

   (2014) Multiscale conformational heterogeneity in staphylococcal protein a: possible determinant of functional plasticity

   Structure  22:1467–1477.

   https://doi.org/10.1016/j.str.2014.08.014

   • PubMed
   • Google Scholar
10. 1. Doster W
    2. Cusack S
    3. Petry W

    (1989) Dynamical transition of myoglobin revealed by inelastic neutron scattering

    Nature 337:754–756.

    https://doi.org/10.1038/337754a0

    • PubMed
    • Google Scholar
11. 1. Doster W

    (2010) The protein-solvent glass transition

    Biochimica Et Biophysica Acta 1804:3–14.

    https://doi.org/10.1016/j.bbapap.2009.06.019

    • PubMed
    • Google Scholar
12. 1. Dror RO
    2. Dirks RM
    3. Grossman JP
    4. Xu H
    5. Shaw DE

    (2012) Biomolecular simulation: a computational microscope for molecular biology

    Annual Review of Biophysics 41:429–452.

    https://doi.org/10.1146/annurev-biophys-042910-155245

    • PubMed
    • Google Scholar
13. 1. Eisenmesser EZ
    2. Bosco DA
    3. Akke M
    4. Kern D

    (2002) Enzyme dynamics during catalysis

    Science 295:1520–1523.

    https://doi.org/10.1126/science.1066176

    • PubMed
    • Google Scholar
14. 1. Eisenmesser EZ
    2. Millet O
    3. Labeikovsky W
    4. Korzhnev DM
    5. Wolf-Watz M
    6. Bosco DA
    7. Skalicky JJ
    8. Kay LE
    9. Kern D

    (2005) Intrinsic dynamics of an enzyme underlies catalysis

    Nature 438:117–121.

    https://doi.org/10.1038/nature04105

    • PubMed
    • Google Scholar
15. 1. Emsley P
    2. Lohkamp B
    3. Scott WG
    4. Cowtan K

    (2010) Features and development of coot

    Acta Crystallographica Section D Biological Crystallography 66:486–501.

    https://doi.org/10.1107/S0907444910007493

    • PubMed
    • Google Scholar
16. 1. Fenwick RB
    2. van den Bedem H
    3. Fraser JS
    4. Wright PE

    (2014) Integrated description of protein dynamics from room-temperature x-ray crystallography and NMR

    Proceedings of the National Academy of Sciences of the United States of America 111:E445–454.

    https://doi.org/10.1073/pnas.1323440111

    • PubMed
    • Google Scholar
17. 1. Fraser JS
    2. Clarkson MW
    3. Degnan SC
    4. Erion R
    5. Kern D
    6. Alber T

    (2009) Hidden alternative structures of proline isomerase essential for catalysis

    Nature 462:669–673.

    https://doi.org/10.1038/nature08615

    • PubMed
    • Google Scholar
18. 1. Fraser JS
    2. van den Bedem H
    3. Samelson AJ
    4. Lang PT
    5. Holton JM
    6. Echols N
    7. Alber T

    (2011) Accessing protein conformational ensembles using room-temperature x-ray crystallography

    Proceedings of the National Academy of Sciences of the United States of America 108:16247–16252.

    https://doi.org/10.1073/pnas.1111325108

    • PubMed
    • Google Scholar
19. 1. Frauenfelder H
    2. Chen G
    3. Berendzen J
    4. Fenimore PW
    5. Jansson H
    6. Mcmahon BH
    7. Stroe IR
    8. Swenson J
    9. Young RD

    (2009) A unified model of protein dynamics

    Proceedings of the National Academy of Sciences of the United States of America 106:5129–5134.

    https://doi.org/10.1073/pnas.0900336106

    • PubMed
    • Google Scholar
20. 1. Frauenfelder H
    2. Hartmann H
    3. Karplus M
    4. Kuntz ID
    5. Kuriyan J
    6. Parak F
    7. Petsko GA
    8. Ringe D
    9. Tilton RF
    10. Connolly ML

    (1987) Thermal expansion of a protein

    Biochemistry 26:254–261.

    https://doi.org/10.1021/bi00375a035

    • PubMed
    • Google Scholar
21. 1. Frauenfelder H
    2. Petsko GA
    3. Tsernoglou D

    (1979) Temperature-dependent x-ray diffraction as a probe of protein structural dynamics

    Nature 280:558–563.

    https://doi.org/10.1038/280558a0

    • PubMed
    • Google Scholar
22. 1. Frauenfelder H
    2. Sligar S
    3. Wolynes P

    (1991) The energy landscapes and motions of proteins

    Science 254:1598–1603.

    https://doi.org/10.1126/science.1749933

    • PubMed
    • Google Scholar
23. 1. Fukuda Y
    2. Inoue T

    (2015) High-temperature and high-resolution crystallography of thermostable copper nitrite reductase

    Chemical Communications  51:.

    https://doi.org/10.1039/c4cc09553g

    • PubMed
    • Google Scholar
24. 1. Garman EF

    (2010) Radiation damage in macromolecular crystallography: what is it and why should we care?

    Acta Crystallographica Section D Biological Crystallography 66:339–351.

    https://doi.org/10.1107/S0907444910008656

    • PubMed
    • Google Scholar
25. 1. Halabi N
    2. Rivoire O
    3. Leibler S
    4. Ranganathan R

    (2009) Protein sectors: evolutionary units of three-dimensional structure

    Cell 138:774–786.

    https://doi.org/10.1016/j.cell.2009.07.038

    • PubMed
    • Google Scholar
26. 1. Halle B

    (2004) Biomolecular cryocrystallography: structural changes during flash-cooling

    Proceedings of the National Academy of Sciences of the United States of America 101:4793–4798.

    https://doi.org/10.1073/pnas.0308315101

    • PubMed
    • Google Scholar
27. 1. Hattne J
    2. Echols N
    3. Tran R
    4. Kern J
    5. Gildea RJ
    6. Brewster AS
    7. Alonso-Mori R
    8. Glöckner C
    9. Hellmich J
    10. Laksmono H
    11. Sierra RG
    12. Lassalle-Kaiser B
    13. Lampe A
    14. Han G
    15. Gul S
    16. Difiore D
    17. Milathianaki D
    18. Fry AR
    19. Miahnahri A
    20. White WE
    21. Schafer DW
    22. Seibert MM
    23. Koglin JE
    24. Sokaras D
    25. Weng TC
    26. Sellberg J
    27. Latimer MJ
    28. Glatzel P
    29. Zwart PH
    30. Grosse-Kunstleve RW
    31. Bogan MJ
    32. Messerschmidt M
    33. Williams GJ
    34. Boutet S
    35. Messinger J
    36. Zouni A
    37. Yano J
    38. Bergmann U
    39. Yachandra VK
    40. Adams PD
    41. Sauter NK

    (2014) Accurate macromolecular structures using minimal measurements from x-ray free-electron lasers

    Nature Methods 11:545–548.

    https://doi.org/10.1038/nmeth.2887

    • PubMed
    • Google Scholar
28. 1. Henzler-Wildman K
    2. Kern D

    (2007) Dynamic personalities of proteins

    Nature 450:964–972.

    https://doi.org/10.1038/nature06522

    • PubMed
    • Google Scholar
29. 1. Hirata K
    2. Shinzawa-Itoh K
    3. Yano N
    4. Takemura S
    5. Kato K
    6. Hatanaka M
    7. Muramoto K
    8. Kawahara T
    9. Tsukihara T
    10. Yamashita E
    11. Tono K
    12. Ueno G
    13. Hikima T
    14. Murakami H
    15. Inubushi Y
    16. Yabashi M
    17. Ishikawa T
    18. Yamamoto M
    19. Ogura T
    20. Sugimoto H
    21. Shen JR
    22. Yoshikawa S
    23. Ago H

    (2014) Determination of damage-free crystal structure of an x-ray-sensitive protein using an XFEL

    Nature Methods 11:734–736.

    https://doi.org/10.1038/nmeth.2962

    • PubMed
    • Google Scholar
30. 1. Holton JM

    (2009) A beginner's guide to radiation damage

    Journal of Synchrotron Radiation 16:133–142.

    https://doi.org/10.1107/S0909049509004361

    • PubMed
    • Google Scholar
31. 1. Keedy DA
    2. Fraser JS
    3. van den Bedem H
    4. Shehu A

    (2015) Exposing hidden alternative backbone conformations in x-ray crystallography using qFit

    PLOS Computational Biology 11:e1004507.

    https://doi.org/10.1371/journal.pcbi.1004507

    • PubMed
    • Google Scholar
32. 1. Keedy DA
    2. van den Bedem H
    3. Sivak DA
    4. Petsko GA
    5. Ringe D
    6. Wilson MA
    7. Fraser JS

    (2014) Crystal cryocooling distorts conformational heterogeneity in a model michaelis complex of DHFR

    Structure (London, England : 1993) 22:899–910.

    https://doi.org/10.1016/j.str.2014.04.016

    • PubMed
    • Google Scholar
33. 1. Kern J
    2. Tran R
    3. Alonso-Mori R
    4. Koroidov S
    5. Echols N
    6. Hattne J
    7. Ibrahim M
    8. Gul S
    9. Laksmono H
    10. Sierra RG
    11. Gildea RJ
    12. Han G
    13. Hellmich J
    14. Lassalle-Kaiser B
    15. Chatterjee R
    16. Brewster AS
    17. Stan CA
    18. Glöckner C
    19. Lampe A
    20. Difiore D
    21. Milathianaki D
    22. Fry AR
    23. Seibert MM
    24. Koglin JE
    25. Gallo E
    26. Uhlig J
    27. Sokaras D
    28. Weng TC
    29. Zwart PH
    30. Skinner DE
    31. Bogan MJ
    32. Messerschmidt M
    33. Glatzel P
    34. Williams GJ
    35. Boutet S
    36. Adams PD
    37. Zouni A
    38. Messinger J
    39. Sauter NK
    40. Bergmann U
    41. Yano J
    42. Yachandra VK

    (2014) Taking snapshots of photosynthetic water oxidation using femtosecond x-ray diffraction and spectroscopy

    Nature Communications 5:4371.

    https://doi.org/10.1038/ncomms5371

    • PubMed
    • Google Scholar
34. 1. Lang PT
    2. Holton JM
    3. Fraser JS
    4. Alber T

    (2014) Protein structural ensembles are revealed by redefining x-ray electron density noise

    Proceedings of the National Academy of Sciences of the United States of America 111:237–242.

    https://doi.org/10.1073/pnas.1302823110

    • PubMed
    • Google Scholar
35. 1. Lang PT
    2. Ng HL
    3. Fraser JS
    4. Corn JE
    5. Echols N
    6. Sales M
    7. Holton JM
    8. Alber T

    (2010) Automated electron-density sampling reveals widespread conformational polymorphism in proteins

    Protein Science  19:1420–1431.

    https://doi.org/10.1002/pro.423

    • PubMed
    • Google Scholar
36. 1. Lee AL
    2. Wand AJ

    (2001) Microscopic origins of entropy, heat capacity and the glass transition in proteins

    Nature 411:501–504.

    https://doi.org/10.1038/35078119

    • PubMed
    • Google Scholar
37. 1. Lewandowski JR
    2. Halse ME
    3. Blackledge M
    4. Emsley L

    (2015) Protein dynamics. direct observation of hierarchical protein dynamics

    Science  348:578–581.

    https://doi.org/10.1126/science.aaa6111

    • PubMed
    • Google Scholar
38. 1. Motlagh HN
    2. Wrabl JO
    3. Li J
    4. Hilser VJ

    (2014) The ensemble nature of allostery

    Nature 508:331–339.

    https://doi.org/10.1038/nature13001

    • PubMed
    • Google Scholar
39. 1. Ollikainen N
    2. Smith CA
    3. Fraser JS
    4. Kortemme T

    (2013) Flexible backbone sampling methods to model and design protein alternative conformations

    Methods in Enzymology 523:61–85.

    https://doi.org/10.1016/B978-0-12-394292-0.00004-7

    • PubMed
    • Google Scholar
40. 1. Rejto PA
    2. Freer ST

    (1996) Protein conformational substates from x-ray crystallography

    Progress in Biophysics and Molecular Biology 66:167–196.

    https://doi.org/10.1016/S0079-6107(97)85629-5

    • PubMed
    • Google Scholar
41. 1. Ringe D
    2. Petsko GA

    (2003) The ‘glass transition’ in protein dynamics: what it is, why it occurs, and how to exploit it

    Biophysical Chemistry 105:667–680.

    https://doi.org/10.1016/S0301-4622(03)00096-6

    • PubMed
    • Google Scholar
42. 1. Schirò G
    2. Fichou Y
    3. Gallat F-X
    4. Wood K
    5. Gabel F
    6. Moulin M
    7. Härtlein M
    8. Heyden M
    9. Colletier J-P
    10. Orecchini A
    11. Paciaroni A
    12. Wuttke J
    13. Tobias DJ
    14. Weik M

    (2015) Translational diffusion of hydration water correlates with functional motions in folded and intrinsically disordered proteins

    Nature Communications 6:6490.

    https://doi.org/10.1038/ncomms7490

    • Google Scholar
43. 1. Schlegel J
    2. Armstrong GS
    3. Redzic JS
    4. Zhang F
    5. Eisenmesser EZ

    (2009) Characterizing and controlling the inherent dynamics of cyclophilin-a

    Protein Science  18:811–824.

    https://doi.org/10.1002/pro.89

    • PubMed
    • Google Scholar
44. 1. Schlichting I

    (2015) Structural biology: photosynthetic complex in close-up

    Nature 517:26–27.

    https://doi.org/10.1038/nature14072

    • PubMed
    • Google Scholar
45. 1. Smith CA
    2. Ban D
    3. Pratihar S
    4. Giller K
    5. Schwiegk C
    6. de Groot BL
    7. Becker S
    8. Griesinger C
    9. Lee D

    (2015) Population shuffling of protein conformations

    Angewandte Chemie  54:207–210.

    https://doi.org/10.1002/anie.201408890

    • PubMed
    • Google Scholar
46. 1. Smith JL
    2. Hendrickson WA
    3. Honzatko RB
    4. Sheriff S

    (1986) Structural heterogeneity in protein crystals

    Biochemistry 25:5018–5027.

    https://doi.org/10.1021/bi00366a008

    • PubMed
    • Google Scholar
47. 1. Spence JCH
    2. Weierstall U
    3. Chapman HN

    (2012) X-ray lasers for structural and dynamic biology

    Reports on Progress in Physics 75:102601.

    https://doi.org/10.1088/0034-4885/75/10/102601

    • Google Scholar
48. 1. Suga M
    2. Akita F
    3. Hirata K
    4. Ueno G
    5. Murakami H
    6. Nakajima Y
    7. Shimizu T
    8. Yamashita K
    9. Yamamoto M
    10. Ago H
    11. Shen JR

    (2015) Native structure of photosystem II at 1.95 å resolution viewed by femtosecond x-ray pulses

    Nature 517:99–103.

    https://doi.org/10.1038/nature13991

    • PubMed
    • Google Scholar
49. Book

    1. Team R

    (2014)

    R: A Language and Environment for Statistical Computing

    Vienna Austria: R Foundation for Statistical Computing.

    • Google Scholar
50. 1. Teeter MM
    2. Yamano A
    3. Stec B
    4. Mohanty U

    (2001) On the nature of a glassy state of matter in a hydrated protein: relation to protein function

    Proceedings of the National Academy of Sciences of the United States of America 98:11242–11247.

    https://doi.org/10.1073/pnas.201404398

    • PubMed
    • Google Scholar
51. 1. Tilton RF
    2. Dewan JC
    3. Petsko GA

    (1992) Effects of temperature on protein structure and dynamics: x-ray crystallographic studies of the protein ribonuclease-a at nine different temperatures from 98 to 320K

    Biochemistry 31:2469–2481.

    https://doi.org/10.1021/bi00124a006

    • PubMed
    • Google Scholar
52. 1. Uervirojnangkoorn M
    2. Zeldin OB
    3. Lyubimov AY
    4. Hattne J
    5. Brewster AS
    6. Sauter NK
    7. Brunger AT
    8. Weis WI

    (2015) Enabling x-ray free electron laser crystallography for challenging biological systems from a limited number of crystals

    eLife 4:e05421.

    https://doi.org/10.7554/eLife.05421

    • Google Scholar
53. 1. van den Bedem H
    2. Bhabha G
    3. Yang K
    4. Wright PE
    5. Fraser JS

    (2013) Automated identification of functional dynamic contact networks from x-ray crystallography

    Nature Methods 10:896–902.

    https://doi.org/10.1038/nmeth.2592

    • PubMed
    • Google Scholar
54. 1. van den Bedem H
    2. Dhanik A
    3. Latombe J-C
    4. Deacon AM

    (2009) Modeling discrete heterogeneity in x-ray diffraction data by fitting multi-conformers

    Acta Crystallographica Section D Biological Crystallography 65:1107–1117.

    https://doi.org/10.1107/S0907444909030613

    • PubMed
    • Google Scholar
55. 1. van den Bedem H
    2. Fraser JS

    (2015) Integrative, dynamic structural biology at atomic resolution--it's about time

    Nature Methods 12:307–318.

    https://doi.org/10.1038/nmeth.3324

    • PubMed
    • Google Scholar
56. 1. Warkentin M
    2. Badeau R
    3. Hopkins JB
    4. Thorne RE

    (2012) Spatial distribution of radiation damage to crystalline proteins at 25-300 k

    Acta Crystallographica. Section D, Biological Crystallography 68:1108–1117.

    https://doi.org/10.1107/S0907444912021361

    • PubMed
    • Google Scholar
57. 1. Warkentin M
    2. Hopkins JB
    3. Badeau R
    4. Mulichak AM
    5. Keefe LJ
    6. Thorne RE

    (2013) Global radiation damage: temperature dependence, time dependence and how to outrun it

    Journal of Synchrotron Radiation 20:7–13.

    https://doi.org/10.1107/S0909049512048303

    • PubMed
    • Google Scholar
58. 1. Warkentin M
    2. Thorne RE

    (2009) Slow cooling of protein crystals

    Journal of Applied Crystallography 42:944–952.

    https://doi.org/10.1107/S0021889809023553

    • PubMed
    • Google Scholar
59. 1. Woldeyes RA
    2. Sivak DA
    3. Fraser JS

    (2014) E pluribus unum, no more: from one crystal, many conformations

    Current Opinion in Structural Biology 28:56–62.

    https://doi.org/10.1016/j.sbi.2014.07.005

    • PubMed
    • Google Scholar

Thank you for sending your work entitled "Mapping the Conformational Landscape of a Dynamic Enzyme by Multitemperature and XFEL Crystallography" for consideration at eLife. Your article has been favorably evaluated by Michael Marletta (Senior Editor) and two reviewers, Stephen C Harrison and A Joshua Wand. The former is a member of our Board of Reviewing Editors.

This manuscript, a major technical achievement, gives an initial glimpse from a new perspective on several long-standing issues and controversies in protein biophysics. There are two significant observations – that local (mostly side chain) conformations are heterogeneous and in many cases show strong temperature dependence and that the temperature dependence is also heterogeneous.

Essential revisions:

A vast amount of work has gone into this report, and the reviewers agreed that it is a tour de force of experimental and computational effort. In its current form however, the manuscript makes it difficult to tease out where it leads. There seem to be two problems, both of which stem from efforts to fit the work into a framework that does not do it justice. Firstly, the authors interpret the results in terms of a transition at 200^o even when the data seem to show that there is no such sharp distinction. More importantly, the authors fail to draw obvious conclusions from the heterogeneity of the apparent temperature dependence and try to force it into an old model that seems to be disproven. Secondly, the inclusion of the XFEL experiments is unjustified.

Problem 1:

There are no data between 180° and 240°, the largest jump in the temperature range. This jump exaggerates the apparent break in the curves in Figure 5B-C, and the relatively small (and imprecisely determined) occupancies of minor states at 180° below makes including those points in the "pseudo Van't Hoff" plots awkward. But within the large error bars that would obtain, many of the plots could probably include the lower temperature data and remain essentially linear (certainly true for my own hand replotting of Figure 5C as lnK vs. 1/T). In any case, it would seem that the heterogeneous response does not support the slaving model of Fraunfelder et al., since the natural thermal dependence of motion at individual sites will, as shown by Lee and Wand, give an average response typical of the scattering studies of old that started the dynamical transition and slaving models.

Problem 2:

The XFEL data are in many ways "scooped" by the 1.2 Å synchrotron data. In several figures, the XFEL points are lonely dots that simply fall on the expected curve. The one justification for inclusion of the XFEL data is the assertion that the dataset is damage free. But Figure 1–figure supplement 1 asserts that damage is minimal across data collection temperatures. There is some increase in R as exposure progresses, and the data were collected to avoid damage differences at different temperatures. Nonetheless, over the whole range of temperatures, the resolution is always better than that of the XFEL data, and had the synchrotron data been cut off at 1.75 Å spacing, it is possible that there would have been no evidence for any damage at any temperature.

Both these problems can most likely be addressed with some thorough reorganization and considerable rewriting of the manuscript, along the lines outlined below.

1) The Introduction should be about the real point of the paper (and the reason that it would be interesting to eLife readers): the internal dynamics of a protein. It should be addressed to someone interested in protein structure, but who is not acquainted with the earlier Frauenfelder and Petkso work, who has not read Lee and Wand (but needs to know about it), and who would like to understand whether the authors agree or disagree with the interpretations given by Kern to her studies of CypA dynamics. In its current form, the Introduction provides an unclear description of the inadequacies of current methods and the promise of XFEL data collection, and the paper then leaves a careful reader believing that the paper has in no way demonstrated that promise and if anything shown that current synchrotrons do better. Only around the fourth paragraph does the Introduction use the possibility of temperature jump (not used for any of the experiments here) as a reason to segue into what are (for the uninitiated) obscure references to Frauenfelder, Lee and Wand, and Eisenmesser et al.

2) In the presentation of Results, the same forced effort to justify the XFEL experiment shows through. For example, in the third paragraph, the failure of the XFEL data alone clearly to show the Leu98 alternative conformation gives rise to the argument "oh well, in the future, with less hardy crystals, we'll need it". We suggest at the end of this set of points how to include the XFEL data more modestly.

3) The Results are written for an "insider". What is "Ringer analysis", for example? If "Ringer" is the name of a program, it should be in all uppercase (like HKL2000, RELION, CHARMM, etc.). Surely there is a two or three word phrase that can summarize sampling the density for evidence of variation of torsion around chi1. Likewise, sentences like "our multiconformer models separate harmonic from non-harmonic contributions to flexibility" will be clear to MD savvy folks, but not to a sufficient fraction of the eLife readership to justify the journal as a good venue for a paper written this way.

4) As already mentioned, the interpretation of a sharp divide above and below 200° K seems forced. Sure, above 200° multiple conformations provide a better explanation than does a single one, but if two conformations are evident at 240, and the occupancy of the minor is not evident at 180°, that does not convince at least one reviewer that there is anything special about 200° or that there is something going on that is different from a smooth decrease in the minor state occupancy that simply falls below the detection threshold somewhere between 240° and 180°, but not in any way that suggests an abrupt transition (at least for the whole protein). This is the key point, both for presentation and analysis of the data and for the conclusions.

5) The paragraph about CypA in HIV is not particularly relevant, and the throw-away remark about an "arms race" should be rethought. Some evolutionary relationship between loop states and Gag binding might hold, but there are no lineage data cited even to hint at such a possibility.

6) A diagram is needed to explain what is meant by "tiered energy landscapes" – or at least what Frauenfelder meant by it. It is important to note that the tiers of Fraunfelder are related in timescale yet the heterogeneity of rotamer distributions seen here have been shown by many NMR studies to be on the same sub-nanosecond timescale.

7) To what extent are these data more or less consistent with the notions put forth by Lee and Wand, by Tilton et al, by Kern and coworkers, and ideally also with the (not mentioned) "sectors" analyzed by Ranganathan?

8) The reviewers are comfortable with inclusion of the XFEL data, but simply as another dataset. Some modest remark is of course justified, about how valuable the new postrefinement methods are for a good dataset, but Figure 8 should be removed.

Minor points:

Figure 6 fails to get across that the conformational heterogeneity and its temperature dependence are in direct opposition to the global "glass transition" popularized by the physics community. It is implied but not made explicit that this heterogeneous temperature dependence predicts the 200 K inflection using data obtained 100 degrees above it. In other words it is a direct consequence of the thermal behavior of internal motion of proteins in solution and has nothing really to do with a global concerted glass transition driven by the freezing of solvent.

The authors talk a lot about "dynamics", but this is not what they observe. Rather they observe presumably (near) equilibrium distributions. What is missing is a strong acknowledgement of timescale. For CypA, they somewhat casually link distributions to slower time scales (i.e. us-ms) rather than the ps-ns motions that actually govern side chain rotamer interconversion. The work of Fraser and Wright (Fenwick et al. 2014) addresses this point somewhat, but that work is only summarized in the present manuscript. Further to this point, the "hierarchical" model (beyond the tautological) is not necessarily consistent, since it requires higher "tiers" of motion to occur on increasingly slower timescales, which is something that this work cannot really address (although the NMR studies say it is not consistent, since the various classes of side chain motion observed occur on the same timescale).

The idea of frustration is overworked. In ubiquitin, one of only two proteins where an extensive temperature dependence of NMR-detected fast motion has been done, it has been shown by pressure perturbation that the coupling (effectively the source of frustration) between side chains is very limited.

Essential revisions:

A vast amount of work has gone into this report, and the reviewers agreed that it is a tour de force of experimental and computational effort. In its current form however, the manuscript makes it difficult to tease out where it leads. There seem to be two problems, both of which stem from efforts to fit the work into a framework that does not do it justice. Firstly, the authors interpret the results in terms of a transition at 200^o even when the data seem to show that there is no such sharp distinction. More importantly, the authors fail to draw obvious conclusions from the heterogeneity of the apparent temperature dependence and try to force it into an old model that seems to be disproven. Secondly, the inclusion of the XFEL experiments is unjustified.

We thank Professors Harrison and Wand for their thoughtful and considerate review. We have extensively reworked the Introduction and Discussion of the manuscript to focus on the divergent models of the relationship between the glass transition and functional dynamics. The results are now tighter and more linear in driving towards these points, reducing the emphasis of the XFEL dataset and accurately positioning it as an important positive control. Our detailed responses to these issues are contained below. Because of the extensive revisions to the manuscript we only include outlines of the revisions.

Problem 1:

There are no data between 180° and 240°, the largest jump in the temperature range.

Technical issues made it difficult to collect data in this temperature range.

This jump exaggerates the apparent break in the curves in Figure 5B-C, and the relatively small (and imprecisely determined) occupancies of minor states at 180° below makes including those points in the "pseudo Van't Hoff" plots awkward. But within the large error bars that would obtain, many of the plots could probably include the lower temperature data and remain essentially linear (certainly true for my own hand replotting of Figure 5C as lnK vs. 1/T). In any case, it would seem that the heterogeneous response does not support the slaving model of Fraunfelder et al., since the natural thermal dependence of motion at individual sites will, as shown by Lee and Wand, give an average response typical of the scattering studies of old that started the dynamical transition and slaving models.

We have deleted the Van’t Hoff plots because we agree that the precision of the occupancy estimate is problematic and can potentially be obscured with increasing B-factors. This uncertainty is a major advantage of the crystallographic order parameter approach, which allows the interaction between the B-factor and occupancy estimates to be accounted in one metric.

In any case, it would seem that the heterogeneous response does not on the face of it support the slaving model of Fraunfelder et al since the natural thermal dependence of motion at individual sites will, as shown by Lee and Wand, give an average response typical of the scattering studies of old that started the dynamical transition and slaving models.

We will argue below for an interpretation that reconciles many aspects of the two models. The crystal solvent does indeed undergo some transition (or more likely multiple transitions) that arrests the progression of disorder in the protein sidechains. The heterogeneity in the temperature this occurs at is reflected in our original analysis of the intersection temperatures. In addition, our data confirm the central ideal of Lee and Wand: that a transition can be observed by monitoring the average disorder after extrapolating the displacements across all protein sites (as in Figure 4 of Lee and Wand, 2001, with individual order parameters bounded between 0 and 1). In our data, however, a flattening transition of the average behavior is only seen when the high temperature data are used (red line) and not when all data are used (purple line). This transition occurs at a lower temperature than in Lee and Wand. In contrast, plotting the average behavior of our observed data (data points) also reveals a flattening transition, but with a higher degree of ordering (and between the 180-240 K data points) than the extrapolation of what is projected from Lee and Wand.

We interpret the origin of this flattening as the solvent glassy state preventing the evolution of further motion in the crystal. The barrier for the protein to transition between different conformations is therefore raised above the projection we would expect based on the simple temperature extrapolation (red line). However, we note that many distinct transitions likely occur across this temperature range as different parts of the bulk solvent and different ordered solvent regions become much more strongly ordered.

These observations thus reconcile the two models. The bulk solvent likely does undergo a complex transition that arrests protein motions, and the average protein sidechain likely does undergo a transition that is driven by thermal depopulation of alternative conformations and unrelated to a global bulk solvent transition. It is also important to note that the bulk solvent transition is not global; even Teeter, Yamano, Stec, and Mohanty, PNAS, 2001, indicate distinct local transition temperatures for waters based on X-ray data. In summary, the discordance in the observed vs. projected average order strongly suggests that the sidechain motions reaching maximum order and the sidechain motions arresting arise from different underlying causes.

A plot has now been added as a panel of Figure 6 (Figure 6C), which now focuses on the non-global nature of the averaged transition measured by this comparison and the intersection analysis presented in the previous version. This topic also forms a major part of the revised Discussion.

Problem 2:

The XFEL data are in many ways "scooped" by the 1.2 Å synchrotron data. In several figures, the XFEL points are lonely dots that simply fall on the expected curve. The one justification for inclusion of the XFEL data is the assertion that the dataset is damage free. But Figure 1–figure supplement 1 asserts that damage is minimal across data collection temperatures. There is some increase in R as exposure progresses, and the data were collected to avoid damage differences at different temperatures. Nonetheless, over the whole range of temperatures, the resolution is always better than that of the XFEL data, and had the synchrotron data been cut off at 1.75 Å spacing, it is possible that there would have been no evidence for any damage at any temperature.

We think this positive control is important because we have experienced skepticism about the influence of radiation damage on the functionally important alternative conformations we observe only at higher temperatures. In the revised manuscript, we have emphasized the control aspect of the XFEL work and de-emphasized its role in general.

Both these problems can most likely be addressed with some thorough reorganization and considerable rewriting of the manuscript, along the lines outlined below. 1) The Introduction should be about the real point of the paper (and the reason that it would be interesting to eLife readers): the internal dynamics of a protein. It should be addressed to someone interested in protein structure, but who is not acquainted with the earlier Frauenfelder and Petkso work, who has not read Lee and Wand (but needs to know about it), and who would like to understand whether the authors agree or disagree with the interpretations given by Kern to her studies of CypA dynamics. In its current form, the Introduction provides an unclear wishy-description of the inadequacies of current methods and the promise of of XFEL data collection, and the paper then leaves a careful reader believing that the paper has in no way demonstrated that promise and if anything shown that current synchrotrons do better. Only around the fourth paragraph does the Introduction use the possibility of temperature jump (not used for any of the experiments here) as a reason to segue into what are (for the uninitiated) obscure references to Frauenfelder, Lee and Wand, and Eisenmesser et al.

We have re-written the Introduction extensively. The major themes of the paragraphs in this shortened introduction are:

a) General problems in protein dynamics and links to function;

b) Problems with cryocrystallography;

c) The two classic models of the glass transition and the recent ssNMR results;

d) How Ringer, multiconformer X-ray models, and crystallographic order parameters overcome the limitations of previous X-ray studies based on B-factors;

e) How the functional dynamics of CypA provide a new model system for testing the links between functional dynamics and the glass transition.

2) In the presentation of Results, the same forced effort to justify the XFEL experiment shows through. For example, in the third paragraph, the failure of the XFEL data alone clearly to show the Leu98 alternative conformation gives rise to the argument "oh well, in the future, with less hardy crystals, we'll need it". We suggest at the end of this set of points how to include the XFEL data more modestly.

We have simplified the presentation of the XFEL experiments, properly positioning the data as a positive control.

3) The Results are written for an "insider". What is "Ringer analysis", for example? If "Ringer" is the name of a program, it should be in all uppercase (like HKL2000, RELION, CHARMM, etc.). Surely there is a two or three word phrase that can summarize sampling the density for evidence of variation of torsion around chi1. Likewise, sentences like "our multiconformer models separate harmonic from non-harmonic contributions to flexibility" will be clear to MD savvy folks, but not to a sufficient fraction of the eLife readership to justify the journal as a good venue for a paper written this way.

We have included simpler explanations of the Ringer and multiconformer modeling procedures and included a more simple definition of harmonic and non-harmonic contributions. In addition, we made other textual changes to increase the clarity of the manuscript. The uppercase naming convention derives from acronyms (e.g. REgularised LIkelihood OptimisatioN). Since Ringer doesn’t stand for anything other than the name Ringer, we have not included it in uppercase.

4) As already mentioned, the interpretation of a sharp divide above and below 200° K seems forced. Sure, above 200° multiple conformations provide a better explanation than does a single one, but if two conformations are evident at 240, and the occupancy of the minor is not evident at 180°, that does not convince at least one reviewer that there is anything special about 200° or that there is something going on that is different from a smooth decrease in the minor state occupancy that simply falls below the detection threshold somewhere between 240° and 180°, but not in any way that suggests an abrupt transition (at least for the whole protein). This is the key point, both for presentation and analysis of the data and for the conclusions.

Indeed, our data do not favor a single transition temperature (that is what is shown in Figure 6). However, our data do suggest that the temperature dependent ordering is strongly restricted at 180 K and below. The slopes are significantly different on either side of this (relatively broad) transition range (the average slope is 2.5x10^-4 (1-S²)/K for 180K and below, and 2.6x10^-3 (1-S²)/K for 240K and above; p=1x10^-62 by a paired T-test).

5) The paragraph about CypA in HIV is not particularly relevant, and the throw-away remark about an "arms race" should be rethought. Some evolutionary relationship between loop states and Gag binding might hold, but there are no lineage data cited even to hint at such a possibility.

We have deleted this section. For the reviewers’ interest, we point to the excellent papers by Leo James, cited in the original manuscript, on the relationship between loop states, CypA/TRIMCypA lineage, and binding.

6) A diagram is needed to explain what is meant by "tiered energy landscapes" – or at least what Frauenfelder meant by it. It is important to note that the tiers of Fraunfelder are related in timescale yet the heterogeneity of rotamer distributions seen here have been shown by many NMR studies to be on the same sub-nanosecond timescale.

We explain more about the population shuffling model and tiers below. Briefly, the exchange of individual sidechain conformations may be fast on the ns timescale, but the dominant conformational exchange that shuffles these conformations is on the ms timescale. We suggest below that the rotameric switch of Phe113 is a dominant contributor to the NMR dynamics.

7) To what extent are these data more or less consistent with the notions put forth by Lee and Wand, by Tilton et al, by Kern and coworkers, and ideally also with the (not mentioned) "sectors" analyzed by Ranganathan?

The discussion is now focused on the following points:

a) A general summary of our major findings on CypA;

b) New insights into the relationship between energy landscapes, the glass transition, and protein function;

c) Glass transition models;

d) Relationship of CypA motion to catalysis and the population shuffling models (Kern and our prior work with her group);

e) The longer term potential for multitemperature and XFEL data to reveal coupled motions and evolutionary selection in proteins (Ranganathan and others).

8) The reviewers are comfortable with inclusion of the XFEL data, but simply as another dataset. Some modest remark is of course justified, about how valuable the new postrefinement methods are for a good dataset, but Figure 8 should be removed.

We have deleted the figure and refocused the discussion of the XFEL data.

Minor points:Figure 6 fails to get across that the conformational heterogeneity and its temperature dependence are in direct opposition to the global "glass transition" popularized by the physics community. It is implied but not made explicit that this heterogeneous temperature dependence predicts the 200 K inflection using data obtained 100 degrees above it. In other words it is a direct consequence of the thermal behavior of internal motion of proteins in solution and has nothing really to do with a global concerted glass transition driven by the freezing of solvent.

We have replaced the original Figure 6. Figure 6A now plots disorder vs. temperature for one representative residue. Figure 6B now plots the distribution of the intersection temperature between the two lines across the protein. Figure 6C now plots the new analysis bridging different models of the dynamical transition. To address the reviewer’s criticism, we have moved the original “small multiples” Figure 6A to figure supplements, to support the new Figure 6B.

The authors talk a lot about "dynamics", but this is not what they observe. Rather they observe presumably (near) equilibrium distributions.

The reviewers are correct that the observations we make are of “conformational heterogeneity” near equilibrium. We are careful not refer to dynamics at all in the Discussion of the X-ray results directly.

What is missing is a strong acknowledgement of timescale. For CypA, they somewhat casually link distributions to slower time scales (i.e. us-ms) rather than the ps-ns motions that actually govern side chain rotamer interconversion. The work of Fraser and Wright (Fenwick et al. 2014) addresses this point somewhat, but that work is only summarized in the present manuscript. Further to this point, the "hierarchical" model (beyond the tautological) is not necessarily consistent, since it requires higher "tiers" of motion to occur on increasingly slower timescales, which is something that this work cannot really address (although the NMR studies say it is not consistent, since the various classes of side chain motion observed occur on the same timescale).

Indeed, our earlier analysis (Fraser et al, 2009), performed in collaboration with the Kern lab, focused on the correspondence between the two states observed by NMR relaxation studies (with an exchange rate of ~1000s-1) and the two dominant alternative conformations observed from room-temperature X-ray studies, which obviously do not specify the timescale. The heterogeneity in the temperature response indicates that the sidechain conformational heterogeneity is not as strongly correlated as we had previously suspected. In the revised manuscript, we propose that many CypA sidechains are in relatively fast exchange, but that a slower process (most probably the coupled sidechain and backbone “backrub” motion of Phe113) shuffles these conformations. This phenomenon of “population shuffling” has been suggested based on ubiquitin (Smith et al, 2015) and is also consistent with the observation that simplistic model of clashes in CypA cannot classify two distinct states (van den Bedem et al, 2013).

The idea of frustration is overworked. In ubiquitin, one of only two proteins where an extensive temperature dependence of NMR-detected fast motion has been done, it has been shown by pressure perturbation that the coupling (effectively the source of frustration) between side chains is very limited.

We have removed the original sentence that contains the reference to frustration, and now mention it only with respect to glassy systems in general.

Author details

1. Daniel A Keedy

   Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States

   Contribution

   DAK, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   Lillian R Kenner, Matthew Warkentin and Rahel A Woldeyes

   Competing interests

   No competing interests declared.

   "This ORCID iD identifies the author of this article:" 0000-0002-9184-7586

2. Lillian R Kenner

   Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States

   Contribution

   LRK, Conception and design, Acquisition of data, Drafting or revising the article

   Contributed equally with

   Daniel A Keedy, Matthew Warkentin and Rahel A Woldeyes

   Competing interests

   No competing interests declared.

3. Matthew Warkentin

   Department of Physics, Cornell University, Ithaca, United States

   Contribution

   MW, Conception and design, Acquisition of data, Drafting or revising the article

   Contributed equally with

   Daniel A Keedy, Lillian R Kenner and Rahel A Woldeyes

   Competing interests

   No competing interests declared.

4. Rahel A Woldeyes

   Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States

   Contribution

   RAW, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   Daniel A Keedy, Lillian R Kenner and Matthew Warkentin

   Competing interests

   No competing interests declared.

   "This ORCID iD identifies the author of this article:" 0000-0003-0737-8383

5. Jesse B Hopkins

   Department of Physics, Cornell University, Ithaca, United States

   Contribution

   JBH, Acquisition of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

6. Michael C Thompson

   Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States

   Contribution

   MCT, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

7. Aaron S Brewster

   Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, United States

   Contribution

   ASB, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

8. Andrew H Van Benschoten

   Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States

   Contribution

   AHVB, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

9. Elizabeth L Baxter

   Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States

   Contribution

   ELB, Acquisition of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

10. Monarin Uervirojnangkoorn

    1. Department of Molecular and Cellular Physiology, Stanford University, Stanford, United States
    2. Howard Hughes Medical Institute, Stanford University, Stanford, United States

    Contribution

    MU, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents

    Competing interests

    No competing interests declared.

11. Scott E McPhillips

    Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    SEM, Acquisition of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

12. Jinhu Song

    Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    JS, Acquisition of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

13. Roberto Alonso-Mori

    Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    RA-M, Acquisition of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

14. James M Holton

    1. Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, United States
    2. Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States
    3. Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, United States

    Contribution

    JMH, Analysis and interpretation of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

15. William I Weis

    1. Department of Molecular and Cellular Physiology, Stanford University, Stanford, United States
    2. Department of Structural Biology, Stanford University, Stanford, United States
    3. Department of Photon Science, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    WIW, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents

    Competing interests

    No competing interests declared.

    "This ORCID iD identifies the author of this article:" 0000-0002-5583-6150

16. Axel T Brunger

    1. Department of Molecular and Cellular Physiology, Stanford University, Stanford, United States
    2. Howard Hughes Medical Institute, Stanford University, Stanford, United States
    3. Department of Structural Biology, Stanford University, Stanford, United States
    4. Department of Photon Science, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    ATB, Acquisition of data, Drafting or revising the article, Contributed unpublished essential data or reagents

    Competing interests

    ATB: Reviewing editor, eLife.

    "This ORCID iD identifies the author of this article:" 0000-0001-5121-2036

17. S Michael Soltis

    Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    SMS, Acquisition of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

    "This ORCID iD identifies the author of this article:" 0000-0003-4678-2995

18. Henrik Lemke

    Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    HL, Acquisition of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

19. Ana Gonzalez

    Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    AG, Acquisition of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

20. Nicholas K Sauter

    Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, United States

    Contribution

    NKS, Analysis and interpretation of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

21. Aina E Cohen

    Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    AEC, Acquisition of data, Drafting or revising the article

    Competing interests

    No competing interests declared.

22. Henry van den Bedem

    Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, United States

    Contribution

    HvdB, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

    For correspondence

    vdbedem@slac.stanford.edu

    Competing interests

    No competing interests declared.

    "This ORCID iD identifies the author of this article:" 0000-0003-2358-841X

23. Robert E Thorne

    Department of Physics, Cornell University, Ithaca, United States

    Contribution

    RET, Conception and design, Analysis and interpretation of data, Drafting or revising the article

    For correspondence

    ret6@cornell.edu

    Competing interests

    No competing interests declared.

24. James S Fraser

    Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States

    Contribution

    JSF, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

    For correspondence

    jfraser@fraserlab.com

    Competing interests

    No competing interests declared.

    "This ORCID iD identifies the author of this article:" 0000-0002-5080-2859

• 4,345

  Page views

• 1,036

  Downloads

• 114

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.