1. Biophysics and Structural Biology
Download icon

Mapping the conformational landscape of a dynamic enzyme by multitemperature and XFEL crystallography

  1. Daniel A Keedy
  2. Lillian R Kenner
  3. Matthew Warkentin
  4. Rahel A Woldeyes
  5. Jesse B Hopkins
  6. Michael C Thompson
  7. Aaron S Brewster
  8. Andrew H Van Benschoten
  9. Elizabeth L Baxter
  10. Monarin Uervirojnangkoorn
  11. Scott E McPhillips
  12. Jinhu Song
  13. Roberto Alonso-Mori
  14. James M Holton
  15. William I Weis
  16. Axel T Brunger
  17. S Michael Soltis
  18. Henrik Lemke
  19. Ana Gonzalez
  20. Nicholas K Sauter
  21. Aina E Cohen
  22. Henry van den Bedem Is a corresponding author
  23. Robert E Thorne Is a corresponding author
  24. James S Fraser Is a corresponding author
  1. University of California, San Francisco, United States
  2. Cornell University, United States
  3. Lawrence Berkeley National Laboratory, United States
  4. SLAC National Accelerator Laboratory, United States
  5. Stanford University, United States
  6. Howard Hughes Medical Institute, Stanford University, United States
Research Article
Cited
26
Views
2,881
Comments
0
Cite as: eLife 2015;4:e07574 doi: 10.7554/eLife.07574

Abstract

Determining the interconverting conformations of dynamic proteins in atomic detail is a major challenge for structural biology. Conformational heterogeneity in the active site of the dynamic enzyme cyclophilin A (CypA) has been previously linked to its catalytic function, but the extent to which the different conformations of these residues are correlated is unclear. Here we compare the conformational ensembles of CypA by multitemperature synchrotron crystallography and fixed-target X-ray free-electron laser (XFEL) crystallography. The diffraction-before-destruction nature of XFEL experiments provides a radiation-damage-free view of the functionally important alternative conformations of CypA, confirming earlier synchrotron-based results. We monitored the temperature dependences of these alternative conformations with eight synchrotron datasets spanning 100-310 K. Multiconformer models show that many alternative conformations in CypA are populated only at 240 K and above, yet others remain populated or become populated at 180 K and below. These results point to a complex evolution of conformational heterogeneity between 180-–240 K that involves both thermal deactivation and solvent-driven arrest of protein motions in the crystal. The lack of a single shared conformational response to temperature within the dynamic active-site network provides evidence for a conformation shuffling model, in which exchange between rotamer states of a large aromatic ring in the middle of the network shifts the conformational ensemble for the other residues in the network. Together, our multitemperature analyses and XFEL data motivate a new generation of temperature- and time-resolved experiments to structurally characterize the dynamic underpinnings of protein function.

https://doi.org/10.7554/eLife.07574.001

eLife digest

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.

https://doi.org/10.7554/eLife.07574.002

Introduction

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.

Results

Multitemperature X-ray datasets reveal modulated conformational ensembles of CypA

To probe the conformational landscape of CypA, we collected eight high-resolution (1.34 –1.58 Å) synchrotron crystallographic datasets across a wide range of temperatures from 100 to 310 K (Table 1) with no added cryoprotectants. For each dataset, we initially refined single-conformer models. Although the single-conformer models are very similar to each other, the accompanying electron density maps reveal differences throughout the protein. In the active-site network, the mFo-DFc difference electron density maps are relatively featureless below 200 K, suggesting that a single conformation is a valid fit below this temperature. By contrast, positive and negative mFo-DFc peaks become gradually more prevalent as temperature increases above 200 K, suggesting that multiple conformations are increasingly required to explain the data as temperature increases (Figure 2—figure supplement 1).

Table 1

Crystallographic statistics for multitemperature synchrotron datasets collected on a single crystal per dataset. Statistics for the highest resolution shell are shown in parentheses.

https://doi.org/10.7554/eLife.07574.003
100 K150 K180 K240 K260 K280 K300 K310 K
PDB ID4YUG4YUH4YUI4YUJ4YUK4YUL4YUM4YUN
Wavelength (Å)0.97670.97670.97670.97670.97670.97670.97670.9767
Resolution range (Å)33.58–1.48 (1.53–1.48)16.95–1.34 (1.39–1.34)16.12–1.38 (1.43–1.38)34.05–1.42 (1.47–1.42)33.98–1.48 (1.53–1.48)25.23–1.42 (1.47–1.42)22.67–1.5 (1.55–1.50)22.66–1.58 (1.64–1.58)
Space groupP212121P212121P212121P212121P212121P212121P212121P212121
Unit cell (a, b, c)42.24, 51.91, 88.0642.45, 51.82, 88.0142.42, 51.96, 88.2143.04, 53.22, 88.6343.09, 52.79, 88.8143.00, 52.61, 89.1243.01, 52.61, 89.3242.85, 52.58, 89.41
Total reflections160,129 (15,842)160,780 (7,437)154,202 (11,295)152,578 (13,600)134,699 (13,381)168,932 (15,019)144,734 (14,433)125,225 (12,326)
Unique reflections32,657 (3,240)42,288 (3,471)39,548 (3,820)38,881 (3,710)34,411 (3,391)38,763 (3,794)32,999 (3,254)28,291 (2,760)
Multiplicity4.9 (4.9)3.8 (2.1)3.9 (3.0)3.9 (3.7)3.9 (3.9)4.4 (4.0)4.4 (4.4)4.4 (4.5)
Completeness (%)99 (100)95 (80)97 (95)99 (96)100 (100)100 (100)99 (100)100 (100)
Mean I/sigma (I)14.07 (1.57)25.95 (3.24)16.47 (1.64)12.86 (1.66)10.09 (1.46)15.51 (1.52)16.90 (1.63)13.26 (1.45)
Wilson B-factor (Å2)16.0713.1216.9515.5516.0617.6219.7521.44
R-merge (%)6.8 (99.4)3.0 (29.4)4.2 (71.8)6.2 (99.2)8.1 (104.3)4.9 (100.0)4.7 (101.7)6.7 (127.3)
R-measurement (%)7.6 (111.0)3.4 (36.9)4.8 (85.6)7.2 (116.8)9.4 (120.8)5.6 (115.3)5.4 (115.9)7.6 (144.5)
CC1/21.00 (0.62)1.00 (0.90)1.00 (0.60)1.00 (0.50)1.00 (0.52)1.00 (0.52)1.00 (0.59)1.00 (0.56)
CC*1.00 (0.88)1.00 (0.97)1.00 (0.87)1.00 (0.82)1.00 (0.83)1.00 (0.83)1.00 (0.86)1.00 (0.85)
Refinement resolution range (Å)33.085–1.48 (1.558–1.48)19.117–1.34 (1.394–1.34)16.995–1.38 (1.435–1.38)34.055–1.42 (1.477–1.42)33.98–1.48 (1.547–1.48)25.23–1.42 (1.477–1.42)22.67–1.5 (1.579–1.5)25.2221.58 (1.679 –1.58)
Reflections used in refinement32,627 (4,654)42,278 (3,932)39,545 (4,265)38,879 (4,161)34,411 (4,237)38,762 (4,256)32,999 (4,643)28,287 (4,632)
Reflections used for R-free1,028 (147)1,325 (125)1,238 (133)1,218 (130)1,080 (133)1,217 (133)1,036 (145)889 (146)
R-work (%)13.3 (20.4)12.4 (16.4)13.3 (25.4)12.6 (26.3)13.1 (26.0)11.1 (22.6)10.8 (20.0)11.7 (21.8)
R-free (%)18.3 (26.8)15.6 (21.3)17.5 (33.0)15.6 (30.4)16.8 (31.2)14.3 (25.5)14.4 (24.8)15.0 (28.8)
Number of non-hydrogen atoms2,2792,4331,9691,9932,0352,1202,0962,172
Macromolecule atoms1,9332,1321,7451,7501,8371,9241,9522,061
Protein residues165164164163163163163163
RMS (bonds) (Å)0.0090.0080.0080.0090.0090.0080.0090.009
RMS (angles) (°)1.161.201.231.201.161.161.141.14
Ramachandran favored (%)9794979697969796
Ramachandran allowed (%)3.35.72.74.134.23.33.9
Ramachandran outliers (%)00000000
Rotamer outliers (%)2.41.30.531.11.51.91.40.88
Clashscore0.571.080.001.240.270.780.520.00
Average B-factor (Å2)21.7417.2521.8520.1420.0021.4824.0925.77
Macromolecule average B-factor (Å2)18.4814.6719.9917.9518.1719.6122.8224.94
Solvent average B-factor (Å2)39.9935.5436.3435.8937.0139.8941.2341.30
  1. PDB: Protein Data Bank. CC: correlation coefficient.

We monitored the shift from single-conformation to multiple conformations both visually (Figure 1A,B) and using the automated electron density scanning program Ringer (Figure 1C,D). Briefly, Ringer identifies alternative conformations at low levels of electron density by evaluating the density value for the γ atom at each possible position about the χ1 dihedral angle, given a fixed main chain conformation (Lang et al., 2014, 2010). We focused on two residues, Ser99 and Leu98, which are key markers of the conformational exchange by NMR (Eisenmesser et al., 2002, 2005) and were implicated in our previous room-temperature X-ray and mutagenesis experiments (Fraser et al., 2009). For both Ser99 (Figure 1A) and Leu98 (Figure 1B), a dominant peak is evident at all temperatures. The reduced height of this peak as temperature increases is accompanied by the increase in a secondary peak corresponding to the electron density of the minor conformation. To quantify this trend, we computed correlation coefficients between the electron density versus dihedral angle curves for each residue (Figure 1C,D). Pairs of curves for similar temperatures have higher correlations than those for different temperatures. In particular, pairs of curves for temperatures that span the low-temperature (100–180 K) and high-temperature (240–310 K) regimes are more poorly correlated than are curves from the same temperature regime. The dynamical transitions observed in previous studies (Doster, 2010; Lee and Wand, 2001; Lewandowski et al., 2015; Ringe and Petsko, 2003; Schiro et al., 2015) generally occur between these two temperature regimes.

Figure 1 with 1 supplement see all
Automated electron density sampling reveals increased conformational redistribution.

Ringer curves of 2mFo-DFc electron density versus χ1 dihedral angle for (A) Ser99 and (B) Leu98 show large peaks for modeled major conformations and smaller peaks for additional minor conformations (dashed vertical lines). These secondary peaks become more evident as temperature increases (color gradient from blue to purple to red). A backrub motion was used for Ser99. For (C) Ser99 and (D) Leu98, a Pearson correlation coefficient was calculated between each pair of Ringer curves from the corresponding panel in (A) or (B). Circles in diagonal elements are colored as in (A) or (B); circles in off-diagonal elements are all gray but scaled by pairwise correlation coefficient (see legend). Pairs of curves from similar temperatures are generally more correlated to each other (larger circles) than are pairs of curves from more different temperatures (smaller circles).

https://doi.org/10.7554/eLife.07574.004

To ground this conformational redistribution in all-atom detail, we built a multiconformer model with qFit (Keedy et al., 2015; van den Bedem et al., 2009) for each multitemperature dataset. We then finalized the model by manually editing alternative conformations and refining to convergence, resulting in models that were improved relative to the single-conformer models (Table 2, Video 1).

Table 2

Improvements in validation statistics from finalizing raw qFit models. Statistics calculated with phenix.molprobity.

https://doi.org/10.7554/eLife.07574.006
RT synchrotronXFEL100 K150 K180 K240 K260 K280 K300 K310 K
Rfree (%)Raw qFit16.725.219.016.918.517.517.915.716.316.1
Final14.624.918.315.617.515.616.814.314.415.0
Δ–2.1–0.3–0.7–1.3–1.0–1.9–1.1–1.4–1.9–1.1
MolProbity scoreRaw qFit1.471.801.791.311.211.181.451.280.951.19
Final1.081.391.191.290.631.140.911.250.990.76
Δ–0.39–0.41–0.80–0.02–0.58–0.04–0.54–0.030.04–0.43
  1. RT: Room temperature; XFEL: X-ray-free electron laser.

Video 1
Animated interpolation between electron density maps in temperature trajectory.

For each pair of adjacent temperatures (e.g. 100 and 150 K), the temperature regime between them was bisected and an average 2mFo-DFc electron density map was calculated in reciprocal space using CCP4 utilities, until temperature points were spaced by <1 K. A new multiconformer model is shown when the animation reaches the corresponding temperature.

https://doi.org/10.7554/eLife.07574.007

At 180 K and below, the active-site network is best modeled as a single state, with electron density corresponding to ordered water molecules clearly evident adjacent to Phe113 (Figure 2, top row). At 240 K and above, by contrast, multiple conformations provide a better explanation of the data. Interestingly, some partial-occupancy water molecules are still present and likely co-occur with the major conformations (Figure 2, middle and bottom rows). Met61 appears to populate additional conformations above 180 K, although it is difficult to precisely define changes in its conformational ensemble as temperature increases. This residue bridges Phe113 and the catalytic residue Arg55 via steric contacts between alternative conformations in both directions, emphasizing the importance of modeling multiple conformations in all-atom detail for understanding inter-residue coupling.

Figure 2 with 1 supplement see all
Multiconformer modeling across temperatures captures increasing conformational heterogeneity.

Residues extending from the core to the active site of cyclophilin A (CypA) adopt a single conformer at low temperatures, but gradually transition to increasing occupancy of secondary conformations as temperature increases. These conformations are well supported by 2mFo-DFc electron density contoured at 0.6 σ (cyan mesh) and 3.0 σ (dark blue mesh). This is corroborated by the room-temperature X-ray free-electron laser (XFEL) model (gray), which is free from conventional radiation damage and features the same secondary conformations. Water molecules (red spheres) are more fully ordered at low temperatures, but become only partially occupied at higher temperatures because they are mutually exclusive with the secondary Phe113 conformation.

https://doi.org/10.7554/eLife.07574.008

XFEL data confirm conformational heterogeneity in synchrotron data is not due to radiation damage

Quantifying radiation damage versus exposure dose (Figure 1—figure supplement 1) and limiting exposure dose per dataset ensured that the conformational heterogeneity observed in multitemperature synchrotron datasets was not dominated by radiation damage. However, XFELs can generate data that are entirely free from conventional radiation damage by diffraction-before-destruction data collection (Kern et al., 2014; Spence et al., 2012). To compare the distribution of alternative conformations between synchrotron and XFEL data, we collected two ambient-temperature datasets: a 1.75 Å resolution radiation-damage-free dataset using serial femtosecond rotation crystallography (Table 3) (Hirata et al., 2014; Schlichting, 2015; Suga et al., 2015) and an additional 1.2 Å resolution synchrotron dataset (Table 4). For the XFEL experiment, we collected 1,239 individual diffraction images, translating to unique unexposed regions of 71 crystals between each shot (Video 2), and processed them using cctbx.xfel (Hattne et al., 2014) with post-refinement in PRIME (Uervirojnangkoorn et al., 2015). Automated molecular replacement yielded interpretable electron density maps that allowed us to refine a single-conformer structural model with reasonable quality statistics. Electron density sampling analysis using Ringer and multiconformer refinement using qFit were performed as for the multitemperature synchrotron data.

Table 3

Crystallographic statistics for room-temperature XFEL dataset collected across 71 crystals. Statistics for the highest resolution shell are shown in parentheses.

https://doi.org/10.7554/eLife.07574.010
XFEL
PDB ID4YUP
Resolution range (Å)43.98 –1.75 (1.81 –1.75)
Space groupP212121
Unit cell (a, b, c)42.42, 51.82, 87.96
Unique reflections19,942 (1894)
Completeness (%)99 (96)
Wilson B-factor (Å2)21.12
Refinement resolution range (Å)43.98 –1.75 (1.93 –1.75)
Reflections used in refinement19,936 (4,811)
Reflections used for R-free625 (151)
R-work (%)20.0 (34.3)
R-free (%)24.9 (36.1)
Number of non-hydrogen atoms1,762
Macromolecular atoms1,559
Protein residues164
RMS (bonds) (Å)0.017
RMS (angles) (°)1.44
Ramachandran favored (%)96
Ramachandran allowed (%)3.6
Ramachandran outliers (%)0
Rotamer outliers (%)1.8
Clashscore1.92
Average B-factor (Å2)29.03
Macromolecule average B-factor (Å2)26.52
Solvent average B-factor (Å2)48.25
Number of TLS groups3
  1. PDB: Protein Data Bank; TLS: translation libration screw; XFEL: X-ray-free electron laser.

Table 4

Crystallographic statistics for room-temperature synchrotron dataset collected on a single crystal. Statistics for the highest resolution shell are shown in parentheses.

https://doi.org/10.7554/eLife.07574.011
1.2 Å Synchrotron
PDB ID4YUO
Wavelength (Å)0.9795
Resolution range (Å)44.60 –1.20 (1.24 –1.20)
Space groupP212121
Unit cell (a, b, c)42.9, 52.43, 89.11
Total reflections307,722 (18,999)
Unique reflections58,118 (5,122)
Multiplicity5.3 (3.7)
Completeness (%)91 (82)
Mean I/sigma (I)10.99 (5.93)
Wilson B-factor (Å2)15.22
R-merge (%)11.2 (20.4)
R-measurement (%)12.2 (23.4)
CC1/20.99 (0.96)
CC*1.00 (0.99)
Refinement resolution range (Å)45.19 –1.20 (1.23 –1.20)
Reflections used in refinement58,108 (3,657)
Reflections used for R-free2,000 (126)
R-work (%)12.7 (31.3)
R-free (%)14.6 (33.5)
Number of non-hydrogen atoms2327
Macromolecular atoms2143
Protein residues163
RMS (bonds) (Å)0.009
RMS (angles) (°)1.16
Ramachandran favored (%)96
Ramachandran allowed (%)4.1
Ramachandran outliers (%)0
Rotamer outliers (%)0.84
Clashscore0.98
Average B-factor (Å2)19.62
Macromolecule average B-factor (Å2)18.40
Solvent average B-factor (Å2)33.86
  1. PDB: Protein Data Bank.

Video 2
Fixed-target X-ray-free electron laser (XFEL) data collection from cyclophilin A (CypA) crystals at the LCLS-XPP end station. 

Screen capture image of the Blu-Ice GUI showing a video display of a CypA crystal. After each shot, a new damage line appears and the crystal is translated.

https://doi.org/10.7554/eLife.07574.012

In agreement with our previous room-temperature studies (Fraser et al., 2009), the XFEL and synchrotron mFo-DFc difference maps reveal evidence for the rate-limiting alternative conformations extending from the active site into the core of the protein (Figure 3A,B). For example, the backrub-coupled (Davis et al., 2006) rotamer jump of Phe113 is apparent from a large positive mFo-DFc peak in both maps. Alternative conformations for core residue Ser99 are also evident from mFo-DFc peaks (Figure 3A,B) and Ringer electron density sampling curves (Figure 3E). We did not conclusively observe a secondary peak in the electron density sampling curve corresponding to a discrete alternative conformation of Leu98 (Figure 3F), but that is likely due to the lower resolution of the XFEL dataset. Multiconformer models for both datasets (Figure 3C,D) again feature alternative conformations across the active-site network and are strongly supported by 2mFo-DFc electron density. These results provide an important positive control on the observation of conformational heterogeneity in our synchrotron studies by establishing that electron density corresponding to the alternative conformations of CypA is not an artifact of radiation damage. The ability of XFEL crystallography to reveal native and functionally important alternative conformations at high resolution may be especially useful for other systems that are presently intractable for room- or variable-temperature synchrotron crystallography due to the small size of available crystals.

Figure 3 with 1 supplement see all
The active-site conformational ensemble of cyclophilin A (CypA) determined without radiation damage at room temperature.

(A) Electron density maps for room-temperature synchrotron (red) and (B) X-ray-free electron laser (XFEL) (silver) single-conformer models reveal conformational heterogeneity extending from the protein core (Leu98 and Ser99) to the active site (Arg55) of CypA. The primary conformation is well supported by 2mFo-DFc electron density contoured at 0.6 σ (cyan mesh) and 3.0 σ (dark blue mesh). mFo-DFc difference electron density contoured at 3.0 σ (green mesh) and − 3.0 σ (red mesh) suggests unmodeled alternative conformations. (C, D) Finalized multiconformer models explicitly model these alternative conformations, which are well-supported by 2mFo-DFc electron density. (E, F) Ringer electron density sampling for the single-conformer models shows peaks representing alternative conformations for (E) Ser99 and (F) Leu98. The primary conformations of both residues are obvious as peaks for both models, but the minor conformations (dashed vertical line; as modeled in 3k0n) are also evident, with 2mFo-DFc values well above the 0.3σ (darker gray horizontal line) threshold, except for the Leu98 in the XFEL model (due to the lower resolution). A backrub motion of −10° positions the backbone properly for Ringer to best detect the minor conformation for Ser99, but not for Leu98.

https://doi.org/10.7554/eLife.07574.013

Some regions feature conformational heterogeneity only at low temperatures

Although more conformational heterogeneity is expected with our higher temperature synchrotron datasets, and is evident in the active site of CypA, cooling can also stabilize new conformations (Halle, 2004). For example, the loop containing residues 79–83 (Figure 4, Video 3) exhibits conformational heterogeneity only at cryogenic temperatures. This region is well fit by a single conformation at 240 K and above, but a secondary loop conformation is necessary to explain the electron density at 100, 150, and 180 K. Additionally, the loop is clearly single-state in the highest resolution (1.2 Å) dataset (Figure 4—figure supplement 1), demonstrating that the slightly lower resolution of the elevated-temperature datasets does not obscure a secondary conformation.

Figure 4 with 1 supplement see all
Alternative loop conformations can appear at lower temperatures.

The surface loop containing residues 79–83 adopts alternative conformations at low temperatures (top row) but not at high temperatures (bottom two rows). The secondary loop conformation is separated from the body of the protein by an ordered water molecule (red sphere); the van der Waals interactions between the loop and the water may reflect an enthalpic stabilization that is more dominant at low temperatures. The electron density peak to the right of the water corresponds to the backbone carbonyl oxygen of Glu81. 2mFo-DFc electron density contoured at 0.6σ (cyan mesh) and 2.0σ (dark blue mesh). XFEL: X-ray-free electron laser.

https://doi.org/10.7554/eLife.07574.015
Video 3
Animated rotation around the 100 K (blue) and 310 K (red) models and electron density maps from Figure 4.
https://doi.org/10.7554/eLife.07574.017

In the primary conformation, the 79–83 loop is not involved in any main chain–main chain hydrogen bonds to the rest of CypA, suggesting that the barrier to forming the secondary conformation does not involve breakage of cooperative secondary-structure-like interactions. The observation of a secondary state for residues 79–83 at 100–180 K, but not at 240–310 K, suggests that it is enthalpically stabilized at lower temperatures (Halle, 2004; Keedy et al., 2014). Consistent with this mechanism, the secondary conformation of the 79–83 loop is accompanied by an ordered, partial-occupancy water molecule (Figure 4, top row). This water molecule, which is clearly distinct from the carbonyl oxygen of the primary conformation of Glu81, wedges between the loop and the rest of the protein. The surprising appearance of specific solvent-linked protein conformational heterogeneity exclusively below 240 K emphasizes the complex and heterogeneous changes in protein–solvent energetics that can occur at cryogenic temperatures.

Quantifying temperature-dependent changes in conformational heterogeneity

Despite counter examples such as the 79–83 loop, most residues in CypA, especially in the active site, exhibit increases in discrete conformational heterogeneity above 180 K. To quantify these changes in regions implicated by NMR relaxation experiments, we measured the 2mFo-DFc electron density in the volumes occupied by the alternative conformations of Ser99 and Phe113. By contrast, B-factors, which can model the harmonic motions near any single conformation, are poor proxies for the non-harmonic change between discretely separated conformations. To quantify the change in minor state occupancy as a function of temperature, we summed the electron density in the volume that is occupied exclusively by the minor conformation and avoided any voxels that overlap with the van der Waals radii of atoms of the major conformation (Figure 5A). The resulting curves of minor-state electron density versus temperature have a shallow slope at 180 K and below, but a much steeper slope at 240 K and above (Figure 5B,C). Additionally, the electron density for the XFEL data is consistent with the data collection temperature (273 K) and the overall trends.

Quantifying temperature titration of conformational heterogeneity in multiconformer models.

(A) 2mFo-DFc electron density was summed over the volume occupied by the minor conformation but not the major conformation (blue grid points) for Ser99 and Phe113. (B,C) Minor-state 2mFo-DFc electron density increases with temperature. Electron density sums were normalized for each residue. Multitemperature points from synchrotron data are shown in colors corresponding to temperature. The X-ray-free electron laser point is shown as a gray triangle. Best-fit lines are shown for 180 K and below (blue) versus 240 K and above (red).

https://doi.org/10.7554/eLife.07574.018

However, most residues that populate alternative conformations do not have such easily characterized and separable regions of electron density. To quantify how conformational heterogeneity throughout CypA varies as a function of temperature, we used B-factor-dependent crystallographic order parameters (S2) (Fenwick et al., 2014). These order parameters include both harmonic contributions, which reflect conformational heterogeneity near one conformation (encoded by B-factors), and non-harmonic contributions, which reflect conformational heterogeneity between multiple discretely separated conformations (encoded by occupancies and displacements in coordinates). Importantly, these order parameters account for both conformational heterogeneity within energy wells, whether it is modeled by B-factors or by subtly different alternative conformations, as well as discretely separated alternative conformations that occupy distinct rotamers. Similar to the 2mFo-DFc electron density integration results for Phe113, we observed a large change in χ1 bond order parameters at 240 K and above (Figure 6A).

Figure 6 with 7 supplements see all
Diversity in temperature dependences of side chain disorder across cyclophilin A (CypA) does not predict the observed average arrest of disorder.

(A) The complement of B-factor-influenced side chain order parameter for the bond most closely associated with the χ1 dihedral angle for Phe113. Lines reflect least-squares fits to synchrotron models at 180 K and below (blue) versus 240 K and above (red). Multitemperature synchrotron points in colors; X-ray-free electron laser (XFEL) point (not included in fits) as gray triangle. (B) Distribution of the intersection temperature between the <200 and >200 K lines fitted with kernel density function. The peak is near 250 K, although there is a tail toward lower temperatures. Intersection temperatures were <170 K for four residues and >330 K for five residues. (C) Predicted and observed values for the complement of side chain order parameter, averaged over all residues in CypA. The predicted values were obtained by extrapolating each residue’s fit line for 240 K and above (red curve) or for the full 100–300 K (purple curve), flooring the result to 0, then averaging across all residues in CypA. Observed values, similarly floored and averaged, are shown as points.

https://doi.org/10.7554/eLife.07574.019

Mapping the transitions in CypA conformational heterogeneity

Next, we applied the order parameter analysis to all side chain χ1 angles in CypA. Although conformational heterogeneity generally increases with temperature throughout the enzyme, we observed a diverse set of conformational responses (Figure 6—figure supplement 1). The trends for the majority of residues suggested a transition somewhere between our data points at 180 and 240 K, below which the change in conformational heterogeneity with temperature is reduced. To quantify this trend, we performed separate fits for the low-temperature (≤180 K) and high-temperature (≥240 K) data points for all residues. The slopes of conformational heterogeneity (1 – S2) versus temperature were significantly different (p=1 × 10–62, paired T-test) on either side of this transition range: the average slope for the low-temperature fit lines (2.5 × 10–4 K–1) was an order of magnitude smaller than for the high-temperature fit lines (2.6 × 10–3 K–1). This is consistent with the idea that heterogeneity is much less dependent on temperature below the 180–240 K ‘transition’ range.

However, some residues behaved differently from the rest of the protein. Val2 retains its conformational heterogeneity at all temperatures, which is expected based on its weakly constrained position at the N terminus. Many of the remaining outlier residues (Glu15, Glu81, Pro105, Ala117, Glu120, Lys125, Met142, Ser147, Lys151) appear to be involved in a spatially contiguous set of crystal contacts across symmetry mates in the context of the crystal lattice (Figure 6—figure supplement 2). This cluster includes Glu81, which adopts alternative backbone conformations only at low temperatures (Figure 4). The variability of these residues can likely be explained by distinct sets of conformations across crystal contacts that are differentially, but somewhat stochastically, favored during the cooling process (Alcorn and Juers, 2010).

Our data suggest that CypA does not undergo a single global transition from having strongly temperature-dependent changes in side chain conformational heterogeneity to relatively temperature-independent behavior. An ‘intersection’ or ‘transition’ temperature for each bond angle can be estimated from the intersection of the low-temperature and high-temperature fit lines of 1 – S2 versus temperature. The distribution of these intersection temperatures is broad and asymmetrical, with an elongated tail from the peak near 250 K toward 200 K (Figure 6B). Furthermore, the distribution of intersection temperatures is more complex for order parameters reporting on the terminal heavy-atom bond of the side chain than for χ1 (Figure 6—figure supplement 3,4). This increase likely occurs because side chain end orientations are subject to more degrees of freedom and therefore temperature changes may redistribute them in a greater variety of ways.

Distinguishing between models of protein heterogeneity as a function of temperature

Our data provide insight into models for the origin of the temperature dependence of protein conformational heterogeneity and into proposed dynamical transitions. In one model, deactivation of different internal protein motions at high temperatures (near 300 K) is sufficient to predict a dynamical transition near 200 K (Lee and Wand, 2001). In a second model, solvent-coupled arrest of protein motions produces a transition in a similar temperature range (Ringe and Petsko, 2003). To distinguish between these two models, we analyzed the average side chain disorder across all residues in CypA, focusing on the bond most closely associated with the χ1 dihedral angle, at each of the eight temperatures we studied (Figure 6C). These averaged disorder values drop as temperature is decreased from 310 K, then flatten out somewhere between 240 and 180 K, with some scatter due to the variability in the cryocooling process (data points in Figure 6C and Figure 6—figure supplement 7).

Next, we used two different linear fits to extrapolate 1 – S2 across all temperatures for each residue, floored the result at maximum order (1 – S2 = 0), and then averaged across all residues to obtain predictions for the residue-averaged disorder versus temperature. In the first fit, the linear function was fit to data for each residue at all temperatures. Consistent with the necessity of using separate fit lines for the low-temperature and high-temperature data (Figure 5 and Figure 6—figure supplement 1), the resulting prediction gives a poor account of the averaged experimental data and does not indicate a transition (purple line in Figure 6C). In the second fit, only the high-temperature (240 K and above) data for each residue were fit. The resulting prediction is more consistent with the averaged high-temperature experimental data and does indicate a transition (red line in Figure 6C). The flattening of this predicted curve at low temperatures occurs as more individual residues achieve maximal predicted order (S2→1) (Figure 6—figure supplement 5). This latter prediction, which is extrapolated from high-temperature crystallographic data, is reminiscent of predictions based on NMR relaxation experiments conducted at 288–346 K (Lee and Wand, 2001). Our observations are consistent with the idea that thermal depopulation of protein alternative conformations is sufficient to predict the existence of an average inflection without invoking a transition of the solvent. However, the low temperature of the predicted inflection (∼100 K), as well as the large separation in low-temperature disorder between our experimental data (data points in Figure 6C) and predictions from high temperature (red line in Figure 6C), suggest that thermal depopulation of protein alternative conformations cannot by itself account for the observed ∼200 K transition. This large separation also indicates that data collected at high temperatures (>260 K) cannot be reliably extrapolated to predict conformational heterogeneity at low temperatures. It also follows that data from low temperatures (≤180 K) cannot be simply extrapolated to predict the features of the energy landscape that may be important above ∼200 K.

Many effects may contribute to the discrepancy between the observed data and the behavior projected from the high-temperature fits. To gain additional insight into this discrepancy, we decomposed the order parameters into their B-factor versus discrete-conformers components and examined their temperature dependences (Figure 6—figure supplement 6). Roughly 67% of residues (e.g. Phe8) remain within one rotamer well across all temperatures. Approximately 13% of residues (e.g. Ser99 and Phe113) populate clearly separable multiple rotameric states at high temperatures, and then show complete depopulation of minority states on cooling so that only a single rotameric state remains at 180 K and below. However, 6% of residues (e.g. Thr5) continue to populate multiple rotameric states at or below 180 K. An additional 6% of residues (e.g. Lys91) populate new rotameric states only at or below 180 K. These results help explain the excess residual disorder in our experimental structures below 240 K compared to projections based on high-temperature fits. Although slopes of crystallographic B-factors with temperature remain nearly flat below 240 K (Figure 6—figure supplement 6B), we expect that true harmonic thermal disorder does subtly decrease from 180 to 100 K; these thermal effects could be more detectable if even higher resolution data were collected at even lower temperatures, perhaps by using liquid helium as a cryogen to cool to ∼15 K or below (Chinte et al., 2007).

Imperfect coupling between active- site residues in CypA

While the results above show a variety of thermal and non-thermal conformational responses, it remains unclear whether these responses involve coupled conformational shifts of multiple residues. In particular, the network of alternative side chain conformations spreading from the core of the protein (Ser99) into the active site (Arg55) across multiple β-strands exhibits qualitatively similar behavior of increasing occupancy above 240 K. In previous work (Fraser et al., 2009), the collective presence of these alternative conformations at room temperature, but not at cryogenic temperatures, and the close contacts between these residues had suggested a concordance with the single exchange process fit by NMR relaxation dispersion for this dynamic active-site network. However, using our new multitemperature data, this network appears subdivided based on the apparent intersection or transition temperatures of the constituent residues, with Ser99 and Phe113 behaving most similarly to each other (Figure 7, Video 4).

The temperature dependence of side chain disorder is non-homogenously spatially distributed in CypA.

Intersection temperatures from (A) χ1 order parameters as in Figure 6B or (B) side chain terminus order parameters as in Figure 6—figure supplement 3 B are mapped to the 1.2 Å room-temperature synchrotron model. Each residue is marked with a sphere colored based on its apparent transition temperature, from low (blue) to high (red). The active-site network is subdivided: Ser99 and Phe113 (left of boxed region) both transition at a low temperature regardless of order parameter bond vector, but Met61 and Arg55 transition at higher, different temperatures.

https://doi.org/10.7554/eLife.07574.027
Video 4
360° rotation around Figure 7B.
https://doi.org/10.7554/eLife.07574.028

Discussion

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).

The dynamic active-site network of cyclophilin A (CypA) has a complex energy landscape.

(A) The previous simple model in which Ser99, Phe113, and Arg55 (Met61 omitted for clarity) interconvert from one macrostate (blue) to the other (red) completely collectively. NMR data suggest this process occurs on a millisecond timescale. (B) A more nuanced model in which network microstates are populated differently depending on the network macrostate, defined by the Phe113 rotameric state. In the left macrostate, Ser99 rotamer changes are disfavored because of steric overlaps with Phe113, but Arg55 rotamer changes are accommodated; the reverse is true (perhaps to a lesser extent) in the right macrostate. Within each microstate, rapid thermal motions occur (bottom right), and may alleviate some minor steric overlaps. Timescales are estimates consistent with NMR observables for CypA and other systems. ms: millisecond; ns: nanosecond; ps: picosecond.

https://doi.org/10.7554/eLife.07574.029

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.

Materials and methods

Protein expression, purification, and crystallization

Wild-type CypA was produced and crystallized as previously reported (Fraser et al., 2009). Briefly, crystals were grown by mixing equal volumes of well solution (100 mM (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) HEPES pH 7.5, 23% PEG 3350, 5 mM Tris (2-carboxymethyl) phosphine [TCEP]) and protein (60 mg mL –1 in 20 mM HEPES pH 7.5, 100 mM NaCl, 0.5 mM TCEP) in the hanging-drop format.

Crystallographic data collection

For the multitemperature synchrotron datasets at 100, 150, 180, 240, 260, 280, 300, and 310 K, we collected data at the Cornell High Energy Synchrotron Source (CHESS) at beamline A1 with a 100 µm collimator using a wavelength of 0.9767 Å. Crystals were looped by hand, stripped of excess mother liquor (100 mM HEPES pH 7.5, 23% PEG 3350, 5 mM TCEP) using NVH oil (Warkentin and Thorne, 2009), and placed directly into the nitrogen-gas cryostream pre-set to the desired temperature at the beamline. Water inside protein crystals is nanoconfined so that ice nucleation is dramatically suppressed, but water outside crystallizes readily and rapidly. Careful removal of all excess solvent from the crystal surface is essential to obtaining ice-free diffraction between 260 K and 180 K without using large cryoprotectant concentrations.

For the XFEL experiment, we collected multiple diffraction images per crystal using a 10-µm X-ray beam with each irradiation point separated by at least 25–40 µm to avoid collateral radiation damage. A total of 1,239 still diffraction images were collected from 71 CypA crystals over the course of two experiments using a goniometer setup and a Rayonix MX325HE detector at LCLS-XPP (Cohen et al., 2014) (Video 2). All data were collected at ambient temperature (approximately 273 K). To prevent dehydration, crystals were coated with paratone oil immediately after looping and mounted on the goniometer at the XPP endstation of LCLS using the SAM sample exchange robot (Cohen et al., 2002).

For the new 1.2 Å room-temperature synchrotron dataset, paratone oil was applied to cover a 2 μL hanging drop containing a single large crystal of CypA. The crystal was harvested through the paratone and excess mother liquor was removed using a fine paper wick. Attenuated data were collected at SSRL beamline 11-1 at 273 K controlled by the cryojet on the PILATUS 6M PAD detector.

Crystallographic data processing

The synchrotron datasets were indexed, integrated, and scaled using XDS and XSCALE, and intensities were subsequently converted to structure factor amplitudes using XDSCONV. All datasets were from single crystals. Data reduction statistics for the highest resolution room-temperature dataset and the multitemperature datasets can be found in Tables 1,4 respectively.

The XFEL data were processed using cctbx.xfel  (Hattne et al., 2014). Of the 1,239 images collected, 772 were indexed and their intensities were integrated. Post-refinement, as implemented by PRIME  (post-refinement and merging, version date: November 11, 20:22:51 2014) (Uervirojnangkoorn et al., 2015), was used to correct the intensity measurements and merge the data. We optimized over the uc_tolerance, n-postref_cycle, sigma_min, partiality_min, and gamma_e values to obtain the final structure factor amplitudes. Data reduction statistics for the XFEL data are provided in Table 3 .

To promote consistency between models derived from different datasets, Rfree flags were generated using PHENIX for the highest resolution ‘reference’ (1.2 Å, 273 K) dataset first and were subsequently copied to all other multitemperature and XFEL datasets for the automated molecular replacement and refinement pipeline.

Model building

For each dataset, we calculated initial phases by performing molecular replacement with phenix.auto_mr using PDB ID 2cpl as a search model. We next refined XYZs and ADPs of the initial model with phenix.refine for 4 macrocycles with XYZ and ADP weight optimization turned on; identified translation libration screw (TLS) groups with phenix.find_tls_groups; and refined optimized XYZs, ADPs, and TLS parameters for six more macrocycles. These single-conformer models and associated electron density maps were used as input for two subsequent steps.

First, the single-conformer models were analyzed with Ringer (Lang et al., 2010) via mmtbx.ringer using default settings. A coupled side chain–backbone ‘backrub’ motion (Davis et al., 2006) of −10° for Ser99 (see Figure 5A) was necessary to match the Cα and Cβ positions of the minor conformation as modeled in PDB ID 3k0n; using this modified backbone indeed yielded maximal minor-conformation Ringer peaks for our multitemperature datasets. No backrub motion was necessary for Leu98 due to the different type of backbone displacement (Fraser et al., 2009). Correlation coefficients between pairs of Ringer electron density versus dihedral angle curves were calculated using the cor function in R (Team, 2014).

Second, the single-conformer models were used as input to qFit (Keedy et al., 2015; van den Bedem et al., 2009). Subsequent to the automated model building, we manually deleted ill-fitting waters and edited alternative protein side chain conformations based on fit to the electron density in Coot  (Emsley et al., 2010) and refinement with phenix.refine. For example, at 240 K, qFit automatically modeled Phe113 as single-state, but significant mFo-DFc peaks remained, so we decided on a two-state model. Met61 was particularly difficult to model accurately across temperatures due to convolved issues of χ3 non-rotamericity for Met in general (Butterfoss et al., 2005), the relatively high electron count for sulfur, and likely temperature-modulated Met-specific radiation damage. For these reasons, visual inspection of the maps and manual building is currently essential for alternative backbone conformations with moderate displacements, as observed in residues 79–83 (Figure 4). We are currently developing new methods to automatically detect and model such backbone excursions in multiscale multiconformer models. These efforts improved Rfree and MolProbity scores across datasets (Table 2). Because of the lower resolution, the XFEL model was refined with three TLS groups and with optimization of X-ray versus geometry and ADP weights.

Model and electron density analysis

For minor-state electron density sums, 2mFo-DFc (Fc filled) map values were summed across a grid of points defined by superimposing each model onto PDB ID 3k0n using all Cα atoms, summing the 2mFo-DFc value at each point with 0.25 Å of a target minor-state heavy atom (Oγ for Ser99; Cδ1, Cε1, Cε2, or Cζ for Phe113), and normalizing to unity across datasets for each residue being analyzed. This procedure allowed a strictly common reference set of map evaluation points. Results were very similar when using unfilled maps (data not shown).

We calculated B-factor-influenced order parameters (S2) as previously reported (Fenwick et al., 2014) except that we monitored one of two different types of bond vector. For the χ1 order parameter, we used Cβ-Xβ (where X = C or O) for most amino acids, Cα-Cβ for Ala, and Cα-Hα for Gly. For the side chain-end order parameter, we used the heavy-atom to heavy-atom bond vector for each amino acid that was closest to the side chain terminus, with ties broken by the number in the atom name (e.g. Cγ-Cδ1 instead of Cγ-Cδ2 for Leu). All negative order parameters (caused by high B-factors) were floored to 0. χ1 order parameters were floored for 7 residues, and side chain-end order parameters were floored for 23 residues. Per-residue ‘apparent dynamic transition temperatures’ were then calculated as the intersection between the <200 K and >200 K fit lines in order parameter versus temperature plots and floored to 0 K if necessary. The kernel density curve was fit with the density function in R (Team, 2014).

For extrapolation of fit lines in Figure 6 , we used a fit to all data points or to just the high-temperature data points (≥240 K) for each residue, and extrapolated to the temperature at which order would be maximized (1 – S2 = 0). To predict global behavior, at each temperature we averaged across all residues the predicted 1 – S2 values from the fit, making sure to floor non-physical predicted values of 1 – S2 < 0 to 0, as in (Lee and Wand, 2001).

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
    Features and development of coot
    1. P Emsley
    2. B Lohkamp
    3. WG Scott
    4. K Cowtan
    (2010)
    Acta Crystallographica Section D Biological Crystallography 66:486–501.
    https://doi.org/10.1107/S0907444910007493
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
    High-temperature and high-resolution crystallography of thermostable copper nitrite reductase
    1. Y Fukuda
    2. T Inoue
    (2015)
    Chemical Communications , 51, 10.1039/c4cc09553g, 25563214.
  24. 24
  25. 25
  26. 26
    Biomolecular cryocrystallography: structural changes during flash-cooling
    1. B Halle
    (2004)
    Proceedings of the National Academy of Sciences of the United States of America 101:4793–4798.
    https://doi.org/10.1073/pnas.0308315101
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
    R: A Language and Environment for Statistical Computing
    1. R Team
    (2014)
    Vienna Austria: R Foundation for Statistical Computing.
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59

Decision letter

  1. Stephen C Harrison
    Reviewing Editor; Harvard Medical School, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

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 200o 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.

https://doi.org/10.7554/eLife.07574.054

Author response

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 200o 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.

https://doi.org/10.7554/eLife.07574.055

Article and author information

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.
    ORCID icon 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.
    ORCID icon 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.
    ORCID icon 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.
    ORCID icon 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.
    ORCID icon 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.
    ORCID icon 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.
    ORCID icon 0000-0002-5080-2859

Funding

National Science Foundation (Graduate Research Fellowship)

  • Rahel A Woldeyes

National Institutes of Health (GM095887)

  • Nicholas K Sauter

Kinship Foundation (Searle Scholar)

  • James S Fraser

Pew Charitable Trusts (Pew Scholar)

  • James S Fraser

David and Lucile Packard Foundation (Packard Fellow)

  • James S Fraser

Howard Hughes Medical Institute (Collaborative Innovation Award (HCIA))

  • William I Weis
  • Axel T Brunger

U.S. Department of Energy (DE-AC02-76SF00515)

  • Aina E Cohen

National Science Foundation (BioXFEL Postdoctoral Fellowship)

  • Michael C Thompson

National Institutes of Health (GM102520)

  • Nicholas K Sauter

National Institutes of Health (GM094586)

  • Henry van den Bedem

National Science Foundation (MCB-1330685)

  • Robert E Thorne

SLAC National Laboratory (SLAC-LDRD-0014-13-2)

  • Henry van den Bedem

National Institutes of Health (OD009180)

  • James S Fraser

National Institutes of Health (GM110580)

  • James S Fraser

National Science Foundation (STC-1231306)

  • James S Fraser

National Science Foundation (DMR-1332208)

  • Robert E Thorne

National Institutes of Health (GM103485)

  • Robert E Thorne

National Institutes of Health (GM103393)

  • Aina E Cohen

A.P. Giannini Foundation (Postdoctoral Research Fellowship)

  • Daniel A Keedy

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Justin Biel, Bryn Fenwick, Robert Stroud, Ian Wilson, and Peter Wright for helpful conversations. DAK is supported by an A. P. Giannini Postdoctoral Research Fellowship. RAW is supported by a NSF Graduate Research Fellowship. MCT is supported by a BioXFEL Postdoctoral Fellowship. NKS and ASB are supported by NIH GM095887 and NIH GM102520. ATB and WIW acknowledge a Howard Hughes Medical Institute Collaborative Innovation Award (HCIA) that also provided funds for the purchase of the microdiffractometer for the goniometer setup. HvdB is supported by the NIH Protein Structure Initiative U54GM094586 at the Joint Center for Structural Genomics and SLAC National Accelerator Laboratory grant SLAC-LDRD-0014-13-2. RET is supported by NSF MCB-1330685. JSF is a Searle Scholar, Pew Scholar, and Packard Fellow, and is supported by NIH OD009180, NIH GM110580, and NSF STC-1231306. Use of the Linac Coherent Light Source (LCLS), SLAC National Accelerator Laboratory, is supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Contract No. DE-AC02-76SF00515. The use of the Cornell High Energy Synchrotron Source (CHESS) is supported by NSF DMR-1332208; the Macromolecular Diffraction at CHESS (MacCHESS) facility is supported by NIH GM103485; and the Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, is supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Contract No. DE-AC02-76SF00515. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research, and NIH GM103393.

Reviewing Editor

  1. Stephen C Harrison, Reviewing Editor, Harvard Medical School, United States

Publication history

  1. Received: March 19, 2015
  2. Accepted: September 29, 2015
  3. Accepted Manuscript published: September 30, 2015 (version 1)
  4. Version of Record published: December 23, 2015 (version 2)

Copyright

© 2015, Keedy et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,881
    Page views
  • 755
    Downloads
  • 26
    Citations

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

Comments

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading