Molecular Dynamics Simulations Reveal ... - ACS Publications

Subscriber access provided by Kaohsiung Medical University

B: Biophysics; Physical Chemistry of Biological Systems and Biomolecules

Molecular Dynamics Simulations Reveal Differentiated Context– Dependent Conformational Dynamics of Two Proteins of the Same Family Feng Tao, and Haiyan Liu J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.8b08468 • Publication Date (Web): 08 Nov 2018 Downloaded from http://pubs.acs.org on November 8, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Molecular Dynamics Simulations Reveal Differentiated Context –Dependent Conformational Dynamics of Two Proteins of the Same Family Feng Taoa and Haiyan Liu*, a, b aSchool bHefei

of Life Sciences, University of Science and Technology of China, 230027, Hefei, Anhui, China;

National Laboratory for Physical Sciences at the Microscales, 230027, Hefei, Anhui, China;

*Corresponding author. E-mail: [email protected]

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 45

Abstract

The Arabidopsis pyrabactin resistant 1 (PYR1)-like family of proteins (PYLs) are receptors of abscisic acid (ABA), an essential small signaling molecule in plants. Here we report a comparative molecular dynamics (MD) study on two PYL members, PYR1 and PYL10, which, despite their highly similar sequences and structures, have been suggested to belong to two different subclasses of PYLs, one being dimeric and relying on binding to ABA to inhibit downstream type 2C protein phosphatases (PP2Cs), the other being monomeric and able to constitutively inhibit downstream PP2Cs without ABA. MD simulations have been carried out on these proteins in various monomeric or complexation states. Analyses of the simulations unambiguously confirm that ABA has large effects on the conformational dynamics of PYR1 but not PYL10, while a downstream PP2C has much larger effects on PYL10 than on PYR1. The differentiated effects are consistent with the functional differences between the two proteins. Potential of mean forces (PMF) calculated by umbrella sampling showed that binding to ABA strengthens the PYR1-PP2C complex, increasing the PMF change for dissociation from 7.5 to 12.0 kcal mol-1. On the other hand, the same PMF change for an apo-PYL10-PP2C complex was computed to be 9.5 kcal mol-1, suggesting stronger binding in apo-PYL10-PP2C than in apo-PYR1-PP2C. Several specific sequence features that may contribute to the functional differentiation between PYR1 and PYL10 are suggested based on the inter-subunit residue-residue contacts occurred in the simulations.

2


Page 3 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1.


INTRODUCTION It is common for proteins of the same sequence family and with high structural similarity to display

subtle but notable differences in important functional expects. To understand possible sequence-structuredynamics relationships that underlie such variations is an important goal for molecular dynamics studies of proteins. One interesting case is the Arabidopsis pyrabactin resistant 1 (PYR1)-like family of proteins, or PYLs, which have been identified as the molecular receptors of abscisic acid or ABA,1-2 an essential molecular signal in plant growth, development and responses to environmental stresses.3-6 The PYL family of proteins belong to the steroidogenic-acute-regulatory-related-lipid-transfer (START) domain superfamily, whose members adopt a helix-grip fold and often host a large cavity to bind hydrophobic ligands. The PYL family of Arabidopsis contains 14 members that have high sequence similarity to each other (sequence identities between 40% to 80%, see the sequence alignments in Figure 1A and Figure S1), most of them sharing a conserved ABA binding pocket in their START domain ligandbinding cavity. Structural and biochemical analyses of the PYL family proteins have provided a general picture of how they may sense and propagate the ABA signal. A number of structural studies have revealed that ABA induces notable conformational changes in some PYLs.7-10 Melcher et al. reported the structures of apo- and ABA-bound PYL2 and of apo-PYL1.9 They suggested that ABA-binding leads to the closing up of a “gate” and a “latch” region (Figure 1B) in the protein. Similar ABA-induced open-to-closed transformations of PYL2 was reported by Yin et al.10 and of PYR1 by Santiago et al.7 and Nishimura et al..8 Nishimura et al. also pointed out a “recoil” motif (Figure 1B) in PYR1 that existed as a part of a loop in the apo-structure but merged into a helix in the ABA-bound or holo-structure. The sequence locations of these regions in PYR1 are indicated in Figure 1A. 3



Page 4 of 45

Not all reported structures of PYLs show large ABA-induced conformational changes.11-12 Sun et al. reported two closed-conformation structures of PYL10 (similar to the PYL10 conformation shown in Figure 1C, which has a Cα root mean square deviation (RMSD) of 0.78 Å compared with the closed conformation of PYR1) in the presence and in the absence of ABA, respectively.11 More recently, Li et al. reported a structure for ABA-free PYL13, which is also in the closed conformation despite the absence of ABA.12 Further, PYL13 is ABA-irresponsive and may not be able to bind ABA.12 Besides possible conformational changes induced by ABA, a number of PYLs have been found to form homo-dimers in solution, the stability of the dimer probably subjected to regulation by ABA.7-10 ABA-free homodimer structures have been reported for PYL1, PYL2 and PYL3.9-10, 13 These structures displayed two-folded symmetry, with both subunits in the open conformation. However, in the presence of ABA, the structures of PYR1 have been reported as asymmetric homodimers,7-8 in which only one of the two subunits contained a bound ABA (Figure 1B). In these symmetric and asymmetric dimers, the sequence locations of the dimer interfaces (Figure 1A) as well as the relative positioning of the two subunits in the three-dimensional space are highly conserved. In the asymmetric dimer, the aforementioned “gate” and “recoil” regions of both subunits also take significant parts in the proteinprotein dimerization interactions. In addition, the “latch” region of the ABA-bound or “closed” subunit is at the dimer interface. In the symmetric dimer (not shown), both subunits are in the open conformation, which causes decreased inter-subunit contacts of the gate regions. Simple substitution of the structure of the “open” subunit by the structure of the “closed” subunit in the asymmetric dimer can lead to serious steric clashes between the gate and the recoil regions of the two subunits (Figure S2). These simple analyses seem to suggest that saturated binding to ABA, which makes the open-conformation monomer unavailable, may lead to significant destabilization of the homodimer, which is in agreement with the 4


Page 5 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


experimental results on PYR114 and PYL3,13 and provides a possible mechanism for the dimeric PYLs to respond to ABA concentration. We would like to mention that two studies have reported homodimer structures of PYLs in which both subunits contain bound ABA.10, 13 However, with the relative positioning of the two subunits being significantly different from those in the ABA-free or one-ABA-bound homodimer structures, these structures have been suggested as metastable.13 While the regulation of the stability of the homodimer by ABA may play a part in ABA-sensing by the dimeric PYLs, which include PYR1 and PYLs 1 to 3, this may not apply to all ABA-responsive PYLs. A number of PYLs, including PYL4, 5, 6, 8, 9 and 10, have been reported to take a monomeric form, irrespective of the presence or absence of ABA.11,

15

Dupeux et al. and Hao et al. assigned these

monomeric PYLs into a new subclass as the subclass 2, with the dimeric PYLs assigned to the subclass 1.11,

14-15

Besides being monomeric, the subclass 2 PYLs have been reported to have higher intrinsic

affinities to ABA than their subclass 1 counterparts.14 Similar to the subclass 1 PYLs, the subclass 2 PYLs should also adopt the closed conformation when binding to ABA, as indicated by the ABA-bound structures of PYL9 and PYL10.15-17 Interestingly, structural evidence, including structures of apo-PYL10 in the closed conformation as well as those in a half-open conformation,11, 15 suggested that even in the absence of ABA, the subclass 2 PYLs may still largely prefer the closed conformation, which distinguishes them from the subclass 1 PYLs. In other words, the open-to-closed transition of subclass 2 PYLs may be autonomous or feasibly-induced independent of ABA, providing a thermodynamic explanation for their higher ABA affinities.14 Further studies suggested that the two subclasses of PYLs may have distinct ABA-dependences in their inhibiting interactions with downstream type 2C protein phosphatases, or PP2Cs, through which the ABA signals sensed by PYLs are further propagated.1-2, 18-19 The inhibition of PP2Cs is achieved through 5



Page 6 of 45

the formation of PYL-PP2C complexes.10, 18, 20 Structural evidence as well as biochemistry experiments suggested that while the dimeric subclass 1 PYLs depend on ABA for PP2C inhibition, the monomeric subclass 2 PYLs may have ABA-independent constitutive activity to inhibit PP2Cs.9,

15, 18

So far,

structures of the subclass 1 PYLs in complex with PP2Cs have all been solved in their ABA-bound form.9, 10, 13, 18, 20-22

On the other hand, an ABA-free complex between a subclass 2 PYL, PYL10, and a

downstream PP2C, HAB1 or hypersensitive-to-ABA-1, has been reported.15 Irrespective of the presence or absence of ABA in these complexes, the different subclasses of PYLs use residues located similarly in sequence (Figure 1A) and in tertiary structure (Figure 1C) to interact with PP2C. The different ABA-dependence of the two subclasses of PYLs in their conformational preferences and their protein-protein interactions provides a highly intriguing case for the investigation of how sequence variations cause changes in structures and conformational dynamics, which in turn regulate functionally important molecular interactions. As the sequence and function diversifications of the PYL family members exist in other plant species besides Arabidopsis,23-24 understanding the differentiated biophysical mechanism of ABA response in different Arabidopsis PYLs may help the understanding of the sequence-function relationships of PYLs from other species as well. A number of experimental studies have already been carried out along this direction.14-15, 25-27 For example, to investigate how the oligomerization state may be related to PYLs’ functions, site-directed mutations have been introduced into dimeric PYLs to obtain monomeric mutants.14-15 Specifically, Dupeux et al. designed the H60P mutation of PYR1 to obtain a partially monomeric mutant.14 Highly interestingly, the affinity of this mutant to ABA increases significantly, in consistence with the stronger intrinsic ABA affinity of the wild type monomeric PYLs. In another study, Hao et al. designed the V83L/I84K double mutant of the subclass 1 protein PYL2, so that its gate region has the same sequence 6


Page 7 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


as the subclass 2 protein PYL10.15 The resulting mutant is monomeric and gains some constitutive activity to inhibit PP2C, although not as strong as PYL10. On the other hand, the Leu83 of PYL10 does not seem to be representative of other subclass 2 PYLs including PYLs 4,5,6, 8 and 9, for which the residue type at this position is valine, the same as in PYL2 but not in PYL10 (Figure S1). This suggests context dependence of the effects of amino acid substitutions. In a study using experimentally-screened rather than rationally-chosen mutations, Mosquna et al. scanned 39 conserved positions in PYR1, a subclass 1 PYL, to identify mutants with ABA-independent PP2C inhibition activity.27 Combining mutations at different sites led to the triple (H60P/V83F/F159V) and quadruple mutants (H60P/V83F/M158I/F159V) that have full PP2C inhibition activity in the absence of ABA (it was not reported whether the ABAbinding affinities of these mutants have been affected).27 Another potentially interesting sequence difference between the two subclasses of PYLs concerns a pair of cysteine residues (Cys30 and Cys157) that are conserved in the subclass 2 proteins but absent from the subclass 1 proteins.8,16 In different crystal structures of PYL10, the pair have been observed in both disulfide-bonded11 and non-disulfide-bonded15 configurations. So far, no function implication has been suggested for this cysteine pair. Besides mutational experiments, molecular dynamics (MD) simulations can also be used to obtain insights into this interesting problem about sequence, structure, dynamics and function relationships.28-32 A few MD studies on PYLs have been reported. In one study, Dorosh et al. have applied MD to compare the essential collective dynamics of the PYR1 subunits in the PYR1 homodimer and in a PYR1-PP2C complex.33 Their results suggested that even for the apo form of a subclass 1 PYL, complex formation with PP2C may have stronger stabilization effects on its intra-molecular structure than homodimer formation, which implicates some intrinsic propensity of ABA-free PYLs to interact with PP2C. In

7



Page 8 of 45

another recent work, Timucin et al. have investigated the effects of magnesium on the thermostability of the PYL10-PP2C complex.34 In the current work, we apply MD simulations to compare two PYL family members belonging to subclass 1 and subclass 2, respectively. The sequence identity between the two members, PYR1 and PYL10, is 45.9%, and the RMSD of Cα positions between their most similar X-ray structures is only 0.78 Å. Our objectives are to investigate, first, whether there exist significant differences between their conformational dynamics; second, how these differences may be related to their differentiated dimeric propensities and ABA-responding interactions with PP2Cs; and third, can any particular sequence features contributing to these differences be identified. These issues have not been the focus of previous MD simulations. To address them, we carry out a series of molecular dynamics simulations of more than 4 µs in total length on holo- and apo- PYR1 and PYL10 in monomer, homodimer (only for PYR1) and HAB1complexed forms. Structural stability, dynamic fluctuations and inter-residue correlations are comparatively analysed. In addition, inter-subunit binding strengths of different complexes are estimated by umbrella sampling along inter-subunit distances. We focus our discussions on those context-dependent properties that are statistically different between the two PYLs, on the functional relevance of these differentiated properties, and on how these differentiations may correlate with some specific sequence variations between the two proteins.

2.

METHODS

2.1 Molecular dynamics simulations MD simulations have been carried out on PYR1 and PYL10 in variously states (Table 1). We note that based on the experimental structures used, all the PYL10 systems have been started from the closed 8


Page 9 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conformation. In addition, the various PYL10 systems are also differentiated by the presence (the oxidized or “-o” systems) or absence (the reduced or the “-r” systems) of the disulphide bond between Cys30 and Cys157 (Figure 1C). Initial structures have been prepared based on available PDB structures (PDB IDs given in Table 1), incomplete or missing residues reconstructed with the Modeller program.35 The initial structures have been solvated in cubic boxes filled with TIP3P water extending at least 12 Å in all directions from the solutes. The CHARMM27 force filed36 has been used. All Asp, Glu, Lys and Arg residues have been assumed to be charged. The protonation states of His residues have been manually assigned according to their predicted pKa values using the ProPKA program37 and possible hydrogenbonding partners in the respective initial structures. When needed, an appropriate number of sodium ions have been added to neutralize the net negative charges of a system. Periodic boundary conditions have been applied with the long-range electrostatic interactions treated using the Particle-Mesh-Ewald38 approach. Force field parameter files have been generated using the Psfgen plugin of the VMD program,39 with parameters for ABA estimated by the CGenFF program.40 The addition of solvent and counter ions have been performed using the VMD plugins Solvate and Ionize. The simulations have been performed with the ACEMD program,41 with a time step of 4 fs enabled by using a hydrogen mass repartitioning scheme.42 Each of the solvated initial structures has been subjected to 10,000 steps of conjugate-gradient energy minimization, followed by 12 consecutive constant volume (NVT) simulations, each lasted 250 ps. In these successive simulations, harmonic restraints have been applied to the Cα positions with progressively decreasing force constants from 1 kcal mol-1 Å-2 to 0, the temperatures gradually increased from 0 to 300 K. To further equilibrate the respective system, another 2 ns constant temperature (300 K) and constant pressure (1 atm.) (NPT) simulations were carried out. Finally, the equilibrated configurations were 9



Page 10 of 45

subjected to 100 ns NVT sampling simulations. For statistically meaningful analysis, for most of the systems, the sampling simulations have been repeated three times with different initial velocities (Table 1). For most of the simulations, the RMSD from respective initial structures reached plateau values within far less than the 100 ns total simulation time (see Results). In addition, analyses carried out using different replicated simulations yielded consistent results. Thus we did not attempt to extend the simulations longer. In the analysis of the MD trajectories, RMSDs refer to those of Cα positions computed in reference to respective starting x-ray crystal structures. The RMSFs refer to root mean square fluctuations of these atomic positions around the structures averaged over the respective 100 ns simulations. To facilitate comparative analysis, PYL10 residue numbers have been replaced by their aligned residue numbers in PYR1. For each of the simulated PYL-PP2C complexes, two sets (A and B) of inter-subunit contacting residue pairs have been identified separately using the criteria that the closest distance between any two non-hydrogen atoms should be smaller than 3.2 Å (set A) or 4 Å (set B). Inter-residue cross correlations of the fluctuations of C α positions have been calculated and plotted using the python libraries mdtraj43 and matplotlib. The protein community detecting tool used in this study was developed by Anurag Sethi,44 which is based on the Givan-Newman algorithm.45 The PYMOL46 and VMD programs have been used for protein visualization. Sequence alignments have been plotted using ESPript webserver47 and sequence identities have been calculated by SIM.48

2.2 Umbrella sampling Umbrella sampling has been applied to obtain potential of mean forces (PMF) for the dissociation of the PYR1 homodimers, the apo-PYR1-HAB1 complex, the holo-PYR1-HAB1complex, and the apoPYL10-HAB1 complex. For the PYR1 homodimer, both the ABA-free and the one-ABA-bound asymmetric forms have been investigated. The reaction coordinate for umbrella sampling is the distance 10


Page 11 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


between the geometric centres of the two subunits. The umbrella potentials are harmonic restraints centred at a set of prechosen distance points, beginning at the respective starting values, ending at a distance 12~15 Å longer, where the PMFs become flat. The separations between the centres of neighbouring windows (1 Å) as well as the restraining force constant (1 kcal mol-1 Å-2) have been chosen based on short test simulations on the PYR1 homodimers to make sure that there is sufficient overlapping between the sampled distance distributions of neighbouring windows.48 The simulation length of each window is 20 ns, with the first 4 ns used for equilibration. Longer sampling simulations have been tested for the PYR1homodimers, and they gave similar results as the 20 ns simulations. Varying the equilibration time between 2 and 6 ns had only small effects (20% in set A, Table 2). (A) apo-PYR1-HAB1; (B) ABA-PYR1-HAB1; (C) apo-PYL10-r-HAB1. The PYR1 and PYL10 are shown as cartoons and HAB1 as surfaces. Contacting residues are shown as sticks, color coded according to occupancies (cyan corresponds to the lowest and red the highest). For each structure, two views rotated by 90 degrees from each other are shown.

40


Page 41 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry A

B

C

D

Figure 8. Residues of PYL10 that form extra polar-polar contacts with HAB1, compared to PYR1. For each illustrated contact pair, a snapshot with the corresponding contact formed was randomly chosen from the simulated trajectory. Contacting residue pairs between (A) Arg64-Glu201; (B) Lys84-Gly246; (C) Glu154-Lys381; and (D) Cys166-Tyr404 are shown

41



Page 42 of 45

Tables Table1. Systems investigated by MD simulations. proteins

PDB ID

systems

repeated times e

apo-PYR1-opena

3

ABA-PYR1-closeda

3

apo-PYR1-closeda

3

ABA-PYR1-asymb

1

apo-PYR1-asymb

1

apo-PYR1-HAB1c

3

ABA-PYR1-HAB1

1

apo-PYL10-od

3

ABA-PYL10-od

3

monomer

apo-PYL10-rd

3

heterodimer

apo-PYL10-r-HAB1d

3

complex states monomer

3K3K PYR1

PYL10

homodimer 3QN1

heterodimers

3R6P

monomer

3RT0 3RT0

a The

ABA-free subunit in the open conformation has been extracted from the asymmetric dimer in 3K3K to obtain the initial structure for apo-PYR1-open, while the other ABA-bound subunit in the closed conformation has been extracted from the same dimer to initiate the trajectories for both ABA-PYR1closed and apo-PYR1-closed. b Only the second one of the two protomers binds to ABA in ABA-PYR1-asym. The initial structure for apo-PYR1-asym has been obtained by removing the ABA from the asymmetric dimer. c Initial structure has been obtained by removing the ABA from the ternary ABA-PYR1-HAB1 complex. d PYL10-o refers to the oxidized form, in which a disulfide bond is formed between Cys30 and Cys157. The corresponding initial structures have been taken from 3R6P, which has been determined in complex with ABA. PYL10-r refers to the reduced form, which does not contain the disulfide bond. The corresponding initial structures have been extracted from 3RT0, which is a complex of PYL10 with HAB1 without ABA. e Times of repeated 100 ns simulations for each system.

42


Page 43 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Frequently occurred inter-subunit residue-residue contacts in the MD simulations of the PYR1/PYL10-HAB1 heterodimers. Occupancies (%) in set A (set B) b ABA-PYR1-HAB1 PYR1-HAB1 PYL10-HAB1 Lys(Ser)63-Side Glu201 92.20(93.9) / 7.90(12.9) His(Pro)60-Side Glu323-Side 21.70(70.9) / / His(Pro)60-Side Tyr404-Main / 38.10(56.2) / His(Pro)60-Main Thr324-Side (72.3) / 8.70(66.1) Ser(Arg)64-Side Glu201-Side / / 55.30(56) Ile( Lys) 84-Side Glu203-Side / / 13.50(13.8) Ile( Lys) 84-Side Gly246-Main / / 24.60(76.9) Ser85-Side Glu203-Side 82.10(85.6) / 78(84.7) Ser85-Side Gly246-Main 34.20(65.7) / 63.30(79) Gly86-Main Arg389-Side 19.60(44.2) 42.20(63.2) 20.70(49.6) Leu87-Side Val393-Main 5.40(63.4) (54.8) (24.5) Pro88-Main Gln386-Side / (36.4) 10.40(46) Arg116-Side Gln386-Side 5.60(76) / 9.3(66.3) Asn151-Side Gln384-Main 81.00(99.3) 10.20(15.4) 62(87.4) Asn151 Trp385-Side 5.20(86.2) 14.60(47.4) 6.10(90.3) Asp(Glu)154-Side Lys381-Side / / 13(14.1) Phe159-Side Gly392-Main 6.00(90.8) (26.1) 9(91.8) Phe159-Side Trp385-Side 8.40(97.2) (70.7) 8.60(95.6) Asp(Glu)155-Side Gln384-Main / 22.00(39.9) 27.20(87.4) Phe159-Side Phe391-Main / 6.10(68.9) (46.8) Thr(Ala)162-Side Phe391-Main 5.40(79.5) (50.5) 7.40(52.5) Thr (Ala)162-Main Tyr404-Side / / 6.6(8.9) Leu(Cys)166-Side Tyr404-Main 6.70(69.3) / (64.1) Gln(Asn)169-Side Tyr408-Side / 7.10(25.2) / Lys170-Side Glu323-Side 82.50(84.2) 27.70(28.7) (5.1) a “Side” refers to the side chain. “Main” refer to the main chain. bSet A has been determined with an inter-non-hydrigen-atom-distance cutoff of 3.2 Å, and set B with a cut off 4 Å. Contacts with occupancies above 5% are listed. PYR1(PYL10) residues a

HAB1 residues a

43



Page 44 of 45

Table 3. Frequently occurred inter-subunit residue-residue contacts in the MD simulations of PYR1 asymmetric homodimers with or without ABA. Occupancies (%) in set A (set B) b ABA-PYR1-asym Apo-PYR1A-asym Met158-Main His60-Side 16.70(29.3) (43.8) Phe159-Main His60-Side 17.50(34.5) / Thr162-Side His60-Side (55.1) 64.20(89.6) Gln169-Side His60-Side 5.60(10.4) / Ser152-Main Lys63-Side / 23.90(27.9) Asp154-Side Lys63-Side 44.20(45.3) 91.60(93) Glu153-Side Lys63-Side 37.40(40.2) (9.6) Asp154-Side Ser85-Side 19.80(44.8) 47.90(65.7) Asp155-Main Ser85-Side 9.00(21.5) / Gly86-Main Gly86-Main (19.5) 22.70(53.2) Asp154-Side Gly86-Main 14.60(17.2) / Ser85-Main Pro88-Side 11.20(75.2) (11.4) Leu87-Main Arg116-Side 20.00(56.1) 15.60(33.5) Glu153-Side Arg116-Side 6.50(6.6) / Lys63-Side Asp154-Side 16.50(22.5) / Lys63-Side Asp155-Side 97.60(98.6) 99.50(99.9) His60-Side Thr162 23.50(53.9) 77(94.6) Ala89-Main Asp155-Side 23.40(30.4) 8.5(18.6) His60-Side Gln169-Side 41.60(51.1) (30.8) Leu166-Main Gln169-Side 34.70(60.3) (9) Gln169-Side Gln169-Side 41.20(78.9) 30.80(53.6) Thr173-Side Gln169-Side 11.40(24) / Gln169-Side Thr173-Side 55.90(74.4) / a “Side” refers to the side chain. “Main” refer to the main chain. b Set A has been determined with an inter-non-hydrogen-atom-distance cutoff of 3.2 Å, and set B with a cut off 4 Å. Contacts with occupancies above 5% are listed. PYR1-open a

PYR1-closed a

44


Page 45 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC Graphic

45


Molecular Dynamics Simulations Reveal ... - ACS Publications

Recommend Documents