Conformational Dynamics of Lysine Methyltransferase Smyd2. Insights

Nov 29, 2016 - ABSTRACT: Smyd2, the SET and MYND domain containing protein lysine methyltransferase, targets histone and nonhistone substrates...
0 downloads 0 Views 4MB Size
Article

Conformational Dynamics of Lysine Methyltransferase Smyd2. Insights into the Different Substrate Crevice Characteristics of Symd2 and Symd3. Balasubramanian Chandramouli, and Giovanni Chillemi J. Chem. Inf. Model., Just Accepted Manuscript • DOI: 10.1021/acs.jcim.6b00652 • Publication Date (Web): 29 Nov 2016 Downloaded from http://pubs.acs.org on December 2, 2016

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 free 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 accessible to all readers and 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.

Journal of Chemical Information and Modeling 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 27

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

Journal of Chemical Information and Modeling

Conformational Dynamics of Lysine Methyltransferase Smyd2. Insights into the Different Substrate Crevice Characteristics of Symd2 and Symd3. Balasubramanian Chandramouli1*, Giovanni Chillemi2* 1

Scuola Normale Superiore, Piazza dei Cavalieri 7, I-56126, Pisa, Italy

2

CINECA, SCAI - SuperComputing Applications and Innovation Department,Via dei Tizii 6, 00185

Rome, Italy Correspondence to: *Tel: 39-050-508406, Email: [email protected] *Tel: 39-06-44486706 Email: [email protected]

1 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

ABSTRACT Smyd2, the SET and MYND domain containing protein lysine methyltransferase, targets histone and non-histone substrates. Methylation of non-histone substrates has direct implications in cancer development and progression. Dynamic regulation of Smyd2 activity and the structural basis of broad substrate specificity still remains elusive. Herein, we report on extensive molecular dynamics simulations on a full length Smyd2 in the presence and absence of AdoMet cofactor (covering together 1.3 µs of sampling), and the accompanying conformational transitions. Additionally, dynamics of the C-terminal domain (CTD) and structural features of substrate crevices of Smyd2 and Smyd3 are compared. The CTD of Smyd2 exhibits conformational flexibility in both states. In the holo form, however, it undergoes larger hinge motions resulting in more opened configurations than the apo form, which is confined around the partially open starting X-ray configuration. AdoMet binding triggers increased elasticity of the CTD leading Smyd2 to adopt fully opened configurations, which completely exposes the substrate binding crevice. These long range concerted motions highlight Smyd2's ability to target substrates of varying sizes. Substrate crevices of Smyd2 and Smyd3 show distinct features in terms of spatial, hydration and electrostatic properties, that emphasize their characteristic modes of substrates interaction and entry pathways for inhibitor binding. On the whole, our study shows how the elasticity and hinge motion of the CTD regulate its functional role and underpin the basis of broad substrate specificity of Smyd2. We also highlight the specific structural principles that guide substrate and inhibitor binding to Smyd2 and Smyd3.

2 ACS Paragon Plus Environment

Page 2 of 27

Page 3 of 27

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

Journal of Chemical Information and Modeling

INTRODUCTION Smyds (SET and MYND domain containing proteins) constitute a five member subfamily of protein lysine methyltransferases (PKMT) that methylate histone and non-histone targets 1⁠ . Smyds recruit S-adenosylmethionine cofactor (AdoMet/SAM) as a methyl donor substrate, catalyze the transfer of the methyl group to specific lysine residues of target protein substrates, and, finally, release the cofactor product S-adenosylhomocysteine (AdoHcy/SAH). Histone targets of Smyds1-3 have so far been reported that include H3-Lys4 (Smyd1-3), H3-Lys36 (Smyd2), and H4-Lys5,Lys20 (Smyd3)

2–6

⁠ . Smyd3 methylates two protein kinases, VEGRF1 (Lys831) and MAP3K2 (Lys260),

leading to augmented VEGFR1 activity in cancer cells and activation of RAS-driven tumorigenesis 7,8

. In the Smyd family, Smyd2 is known to methylate a broad spectrum of substrates. Non-histone

targets of Smyd2 include tumor suppressor proteins p53 (Lys370) and Rb (Lys860), Estrogen receptor ERα (Lys266), PARP1 (Lys528) and HSP90 (Lys531, Lys574)

9–13

⁠ ⁠ . There are ongoing

efforts to identify new substrates of Smyd2 to unveil its associations in novel biological processes and human diseases. Recent proteomics and multistate CPD based studies have identified 14 novel cellular substrates whose methylation is catalyzed by Smyd2 14–16⁠ . Lately, Smyd2 has gained larger attention as its role in cancer biology, via methyltransferase activity on non-histone targets, has been unfolded

17

⁠ . Methylation of p53 impairs its binding

efficiency to promoter genes, thereby repressing p53 transcriptional activity, and down-regulation of Smyd2 by siRNA in cells promotes p53 mediated apoptosis 9⁠ . Methylation of a conserved lysine residue of Rb generates an epitope, selectively recognized by transcriptional repressor L3MBTL1, and thus regulates Rb activity in cell cycle progression and cellular differentiation

10

⁠ . Smyd2

mediated methylation has been shown to be important for dimerization of HSP90, which is essential for its activity to fold oncogenic kinases in cancer cells

13

⁠ . Overexpression of Smyd2 in

esophageal squamous cell carcinoma has been shown to inversely correlate with patients survival rate, leading to the identification of Smyd2 as a prognostic marker 18⁠ . Crystal structures of Smyd2 in complex with AdoMet cofactor revealed a bi-lobed 3D conformation formed by the N-terminal (NTD) and C-terminal (CTD) domains

19,20

⁠ . The deep groove between

the domains acts as the binding crevice for protein substrates (Figure 1A). The NTD (res. 1-271) comprises the catalytic SET, MYND and post-SET (pSET) domains. The SET domain is split by the MYND into two segments (NSET and SET). The MYND domain is a zinc-finger motif that binds to two zinc ions, coordinated by seven cysteines and a histidine (Figure S1). The pSET domain is

3 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

organized around a zinc, coordinated by four conserved cysteines. The CTD (res. 272-433) contains seven α-helices that mainly form the tetratricopeptide repeat (TPR) domain (Figure S1). Smyd2 shares ~30% sequence identity with Smyd3 primarily at the level of the SET and MYND domains. An overlay of their X-ray structures shows that while both proteins adopt similar cradle shaped 3D topology with a well superimposed NTD, orientation differences exhibited by the CTD result in a closed (Smyd3) or partially open (Smyd2) conformation (Figure 1B). Despite this knowledge, understanding the specific structural feature that drives their substrate recognition demands detailed investigation of their conformational dynamics. In this regard, we recently reported an extensive MD simulation study on human Smyd3. We demonstrated that AdoMet binding restricts the flexibility of the CTD and locks Smyd3 in a closed clamshell like conformation, and highlighted the importance of the CTD in stabilizing the substrate crevice for target recognition 21⁠ . Herein, we report MD simulations performed on a full length human Smyd2 both in the absence (hereinafter, apo) and presence (hereinafter, holo) of the AdoMet cofactor, covering a total ~1.3 µsecond sampling. The present report paves the route for a better understanding of the structural and dynamic changes of Smyd2 upon cofactor binding, by comparing state-of-the-art simulations in the two states. The structural and dynamical features that underpin the broad substrate recognition have been examined by discriminating the CTD dynamics, spatial and electrostatic characteristics of the substrate binding crevices of Smyd2 and Smyd3.

4 ACS Paragon Plus Environment

Page 4 of 27

Page 5 of 27

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

Journal of Chemical Information and Modeling

METHODS Model Generation and Simulation Protocol Atomic coordinates for the simulation were extracted from the better resolved (at 2.0Å resolution) X-ray crystallographic structure of Smyd2-AdoMet binary complex (PDB ID:3TG4) available in PDB19⁠ . This structure has a similar 3D conformation as other X-ray structures of Smyd2 in complex with AdoMet cofactor (PDB ID: 3S7B,4YND), and well superimposed with a RMSD of 0.7Å over all Cα atoms

20,22

⁠ . Simulation representing the apo state was performed after removing

the cofactor from the X-ray structure. Starting structures of apo and holo systems were immersed in a cubic box of TIP3P water model

23

⁠ , which extended up to ~15 Å from solute surface and

additional counter ions were added to render electroneutrality. Histidine protonation states were assigned based on the consensus as predicted by H++ (v3.1), Protoss and Propka programs24–26⁠ . Simulations were performed with Amber package (v.14) using ff14SB forcefield for the protein 27⁠ . Parameters for AdoMet were adapted from Amber ff10 forcefield and a previous report

27–29

⁠.

Partial charges were derived by fitting the electrostatic potential, obtained after a geometry optimization using the HF/6-31G* level of theory using Gaussian software (Gaussian 09, Gaussian Inc.), following the restrained electrostatic potential (RESP) formalism using RED tools

30,31

⁠.

Dihedral restraints were applied during geometry optimization to retain the enzyme bound cofactor conformation. Amber ZAFF parameters were used to treat the zinc sites, involving a zinc ion coordinated to four protein residues 32,33⁠ . Starting systems were equilibrated following a multi-step protocol as follows; i) two rounds of minimization (6000 iterations) and dynamics (100 ps, ∆t = 1fs) of the solvent and counter ions in the bulk, keeping the protein Cα atoms and zinc sites restrained with decreasing force constants 10, 5, 3, and 2 Kcal/(mol Å2), ii) an unrestrained minimization of the whole system (10000 steps), iii) final heating to 303 K at constant volume (250 ps, ∆t = 1fs), followed by density equilibration at constant pressure (3 ns, ∆t=1fs). The production phase was initiated in NVT ensemble and extended for ~640 ns. MD snapshots were saved at 10 ps intervals. Simulation conditions included periodic boundary conditions, constraining covalent bonds involving hydrogens via SHAKE that permitted a 2 fs time step for the numerical integration

34

⁠ , 12 Å cut-off for calculating non-bonded

interactions, Particle Mesh Ewald (PME) method for evaluating long-range electrostatic interactions 35

⁠ , temperature regulation with Langevin coupling using a collision frequency of 2.0 ps-1 36⁠ .

5 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

Analysis of MD Trajectories Trajectories were analyzed using Ambertools (v1.5), Gromacs utilities (v.5.0.7) or in-house code written using MDAnalysis

37–39

⁠ . Backbone RMSD was calculated using either the X-ray or

average structure from simulations as the reference; the latter was obtained by extracting the frames over the last 500 ns at 100 ps interval. Per residue RMSF was calculated, fitting the frames against the X-ray structure using the backbone all residues except the terminal ones. Segmented RMSF for the individual domains was obtained by a restricted fitting using the backbone of residues constituting the domains. Throughout the text, reference to backbone means Cα,C,N atoms. Dynamic cross correlation matrix was computed considering the backbone coordinates as follows, C ij= ⁠∆ r i ∆r j ⁠ / [ √ ⁠∆r 2i ⁠√ ⁠∆r 2j ⁠]

where ∆ri is the displacement of ith atom from the mean position and < > represents the ensemble average over analyzed portion of the trajectory. Principal component analysis (PCA) was performed by diagonalizing the covariance matrix obtained from atomic fluctuations (using Gromacs utilities), after removing the rotations and translations by fitting onto the X-ray structure 40⁠ . Geometric angles were calculated using an in-house code after fitting the trajectory over the Nterminal lobe of the X-ray structure. Opening angle was defined between the vectors from res.32-37 (NSET) to res.104-118 (SET), and to res.312-314 (CTD), respectively, considering the Cα centroid as vector coordinates. Slide angle was defined between the vectors from res.291 to 310 (along the first alpha helix of CTD) and from res.291 to res.314 (pointing to the second α-helix of CTD), using the Cα positions as vector points. The first vector was fixed at the X-ray structure. Hydration of substrate binding crevice was examined by counting the number of water molecules within a distance threshold of 12 Å from the geometric center of all Cα atoms. Solvent contact surface area was calculated with Naccess program41⁠ (v2.1.1), using a solvent probe radius of 1.4 Å. Buried surface was obtained by subtracting the contact surface areas calculated separately for the N and Cterminal domains from that of the whole protein. Reported values indicate the average of the buried surface areas of N and C-terminal domains, respectively. Electrostatic potential was obtained using 2.0 and 78.5 as solute and solvent dielectric permittivity, at zero salt concentration, with APBS (v1.5) program 42⁠ . Figures and plots were generated using UCSF-Chimera and Matplotlib library 43,44

⁠.

6 ACS Paragon Plus Environment

Page 6 of 27

Page 7 of 27

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

Journal of Chemical Information and Modeling

RESULTS The C-terminal Domain Undergoes Larger Structural Relaxation in Both Apo, Holo States Structural deviation of the simulated systems with respect to the starting X-ray configuration was examined in terms of the backbone RMSD (Figure 2A). In the apo system (gray line), the RMSD quickly reached a plateau and remained stable around an average value of 1.86 Å after 50 ns. While in the holo system, after an initial increase, the RMSD fluctuated around a higher average value of 2.69 Å, indicating a larger relaxation from the starting configuration as compared to the apo system. All the following analyses were restricted to the last 500 ns omitting the initial portion of the trajectory (Figure 2A, shaded). Structural deviation of the individual domains within each system was examined by calculating their RMSD with respect to the average structure from the respective simulations (Figure 2B). In the apo system, all domains have a similar average RMSD value. In the holo system, a slightly higher average is observed for the CTD and MYND, indicating a larger contribution to the overall deviation compared to the other domains. Further, to examine the extent of structural changes that each domain exhibits across the two states, RMSDs were also computed by swapping the apo and holo average structures. The CTD displayed more difference between the apo and holo systems, followed by the MYND domain (Figure 2C). The pSET has the lowest deviations in both systems, indicating its minor relaxation in the two states. Additionally, we monitored the conservation of system's temperature, potential and total energies. Time evolved variations of these parameters showed a stable profile throughout in both simulations (Figure S2). To identify and compare the mobile regions of the protein in the two states, per residue averaged backbone RMSF was computed. In both systems, the first anti-parallel α-helical region of the CTD displayed more flexibility which is higher in the holo case (Figure 2D, shaded). In the apo system, two additional peaks are observed around residues 14-17 and 97-102, respectively. These regions can be better visualized in the 3D representation of the RMSF (Figure S3A, A'). The first peak corresponds to the β-turn segment in the NSET, which forms a part of the cofactor binding cavity. Residues in this segment make direct contacts with AdoMet and stabilize its binding 45⁠ . Hence the increased flexibility is an expected effect of the cofactor removal (Figure S3A). The second peak corresponds to the solvent exposed loop that connects the MYND and SET domains. The figure also shows the localized flexibility observed around the CTD's first pair of anti-parallel α-helices. To verify if the flexibility of the mobile regions is influenced by its relative motion with respect to other parts of the protein, segmented RMSF was computed for each individual domain after fitting the snapshots based on the backbone of residues constituting the domains. Segmented RMSF 7 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

profile, depicted in 3D (Figure S3B, B'), confirms that intrinsic flexibility of the mobile regions is unaffected by the other regions. C-terminal Domain Adopts Common Modes of Opening Motion Among the Smyds. To examine inter-residue communications within the protein, we computed the dynamic cross correlation (DCC) matrix to get insights into the degree of correlation in protein motion 46–48⁠ . The cross correlation coefficient Cij between a pair of residues i and j indicates a correlated (in-phase, Cij > 0) or an anti-correlated (out-phase, Cij < 0) motion. The 2D-DCC matrix is reported in Figure 3A for apo (upper left triangle) and holo (lower right triangle) systems. It is apparent that in both systems no strong in-phase correlation is visible between the N and C-terminal domains, indicating that the two lobes act like independent clusters of the protein. Within the N-terminal domain some subtle correlation in motion, though not strong, is observed between the NSET, MYND and SET domains. The CTD exhibited anti-correlated motions largely against the MYND and SET (Figure 3A, encircled), which is slightly higher in the holo system. This feature can be better seen in the 3D depiction of negative correlations for apo (Figure 3B) and holo (Figure 3C) systems, where Cα pairs with negative correlation coefficient (-0.65 < Cij < -0.45) are connected by blue lines. This result primarily hints that the CTD undergoes a hinge motion with respect to the N-terminal counterpart. The conformational changes that characterize global protein motion can be inferred from PCA, which permits to decouple the complex motion in a subset of dimensions (i.e., eigen vectors) that largely describe the variance in motion

40,49,50⁠

⁠ . The cumulative contribution of the PCA eigen

vectors revealed the first three vectors to explain ~79% (apo) to 89% (holo) of the total variance in motion (Figure S4A). Projection of protein motion along the principal eigen vectors showed the CTD dynamics to involve two major modes of motion that include hinge (opening) and shear (sliding) movements in both systems (Figure S4B). Previously we reported a similar observation in the CTD motion of Smyd3 in the apo state 21⁠ . These results suggest that the bi-lobed structure of the Smyds directs the CTD to adopt similar modes of opening that leads to the exposure of the substrate binding crevice. Next, we compared the extent of the CTD opening in the two states by monitoring time evolved variations in two geometric descriptors, the open and slide angles (refer section 2.2 for definitions) that describe the relative movement of the CTD with respect to the N-terminal domain (Figure 4). In the apo system, open angle remained stable and showed a smaller spread around the average value which is closer to the X-ray structure (Figure 4A'). In the holo system, instant angles displayed higher fluctuations and broader variations about the average value that is well shifted 8 ACS Paragon Plus Environment

Page 8 of 27

Page 9 of 27

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

Journal of Chemical Information and Modeling

above the X-ray structure. Time evolved variations of the slide angle followed a similar pattern as described above (Figure 4B'); the apo system has lesser fluctuations and the average value being closer to the X-ray structure. On the contrary, the holo system has higher fluctuations with a larger average value compared to the X-ray. These results confirm that the CTD exhibits an opening in both systems. Distinct Features of the CTD Dynamics and Substrate Crevices of Smyd2 & 3 To gain additional insights concerning the CTD dynamics and associated conformational transitions in Smyd2 and Smyd3 21⁠ , we compared the CTD motion, spatial and electrostatic properties of the interaction surface of their substrate crevices (Figure 5). Distributions of open and slide angles for Smyd2 showed that in the apo system the angle values display shorter spread and the average values stayed closer to the X-ray (Figure 5A). In contrast, in the holo system, the angle values showed broader distributions and the average values are shifted above the X-ray structure. Due to larger degrees of freedom of CTD motion, Symd2 samples more opened conformations in its holo state in contrast to the apo form. Interestingly, the angle distributions showed a reverse trend for Smyd3 (Figure 5A, inset); the CTD in the holo state exhibits a more confined motion than the apo state. Indeed, we observed the holo enzyme to maintain a closed clamshell like conformation similar to the respective X-ray structure. The contrasting difference in CTD motion drives Smyd2 and Smyd3 to present, in their holo states, an expanded deep canyon (Smyd2) or narrow pocket (Smyd3) for the substrates. Spatial features of the substrate crevices of Smyd2 and Smyd3 were compared by counting the accommodated water molecules and the amount of buried surface between the N and C-terminal domains (Figure 5B,C). The number of water molecules occupying the substrate crevice is higher in Smyd2 than Smyd3. The larger pocket of Smyd2 permits the housing of ~20 more water molecules as indicated by the population averages (Figure 5B). Plot of buried solvent contact surface area showed a distinct profile for the two Smyds (Figure 5C). In Smyd2, the amount of buried surface at the domains interface is lower than Smyd3. Further, in Smyd2 the distribution of the buried surface is shifted to lower values compared to the average buried area in the apo state (Figure 5C, Inset; compare solid and dashed blue lines). In contrast, distribution of the buried surface in Smyd3 is shifted to higher values compared to the average buried area in the apo state (Figure 5C, Inset; compare solid and dashed green lines). This result demonstrates the propensity of spatial expansion (Smyd2) or contraction (Smyd3) of the substrate crevices moving from respective apo states to the holo ones.

9 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

Electrostatics of the substrate crevices of the Smyds was examined by obtaining the electrostatic potential of a representative holo configuration (Figure S5A), using the APBS program 42⁠ . Surface depiction of the structures shows the closed conformation of Smyd3 to well maintain the interface contacts between the CTD and MYND, consequently presenting a small substrate pocket (Figure S5B). This feature may provide a restricted entry pathway for target substrates. The narrow entrance of the pocket is shaped by a cluster of acidic residues, as a result of which a high negative potential is concentrated at the vestibule (Figure S5C). In Smyd2, the fully opened conformation presents a larger entrance that can permit the entry and accommodation of different sized substrates (Figure S5B). Further, the electrostatics shows a dispersed negative potential at the substrate crevice (Figure S5C). Larger degrees of the CTD motion may alter the spread of negative potential to exert specific, substrate dependent dragging force for recruitment via long range electrostatic interactions. Another notable feature apparent from the electrostatics concerns the large positive iso-surface around the Smyd3's MYND domain which correlates with the previously reported role of the domain in protein-DNA interactions. Mutation of basic residues in the MYND, in fact, has been shown to disrupt the DNA binding to Smyd3 3,51⁠ .

10 ACS Paragon Plus Environment

Page 10 of 27

Page 11 of 27

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

Journal of Chemical Information and Modeling

DISCUSSION Smyd2 methylates a spectrum of substrates and has a defined role in the dysregulation of different tumour associated signaling pathways 17⁠ . Overexpression of Smyd2 correlates with alterations of gene expressions involved in cell cycle, chromatin remodeling and transcriptional regulation

52

⁠.

Smyd2 gained a large therapeutic interest after its oncogenic potential was established, following which the first Smyd2-inhibitor (AZ505) has been reported

18,20

⁠ . Hence, examining how Smyd2

dynamically regulates its activity serves not only to understand the factor that underpins broad substrate recognition but also adds to the knowledge of designing potent inhibitors. Herein, taking advantage of the available structure data, we performed extensive MD simulations on Smyd2 to delineate the conformational changes in the presence and absence of the AdoMet cofactor. Further, we compared differences in Smyd2 and Smyd3 dynamics to infer about their characteristic substrate crevices. AdoMet Drives Enzyme Specific Conformational Changes Simulations of Smyd2 apo and holo systems resulted in stable trajectories without any significant structural distortions. This is expected, as overall, the protein is largely composed of closely packed α-helices and β-sheets. The CTD remained flexible in both systems but undergoes larger relaxation in the holo system with respect to the X-ray structure (Figure 2B). The pSET domain stayed rigid without significant relaxation across the two systems (Figure 2C). The cysteine rich pSET plays a critical role in maintaining the structural integrity of the enzyme. Mutation of conserved cysteine or complete deletion of the pSET has been shown to reduce or abolish the methyltransferase activity 45

⁠ . The CTD exhibited strong anti-correlation motions with respect to the NTD in both states

resulting in a hinge like opening of the domain (Figure 3). Analysis of the motion along the principal vectors showed common modes of opening transitions with respect to the NTD counterpart (Figure S4). However the extent of the CTD opening is different in the two states, as inferred from the time evolved variations of geometric angles (Figure 4). Distribution of angle values showed that the CTD motion in the apo state is confined around the X-ray (Figure 5A). In contrast, in the holo state, appreciable shift with broad variance is observed. Intriguingly, a reverse trend observed in for Smyd3 (Figure 5A, Inset). Taken together, these results indicate that AdoMet binding drives enzyme specific conformational transitions in the CTD. In Smyd2, the presence of the cofactor enhances elasticity of the CTD which results in larger hinge motions and fully opened conformations. In Smyd3, the AdoMet's presence

11 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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 12 of 27

restricts the CTD's hinge motion and, consequently, in the holo state a closed conformation is retained

21

⁠ . In the Smyd3 apo state, the CTD motion is significantly enhanced upon the cofactor

removal, leading to the sampling of fully opened conformations. It is worth noting that starting configurations of all simulations represent the holo form of the enzymes. Hence, the observed difference in the CTD dynamics can be attributed to the binding or removal of the cofactor. We posit that AdoMet cofactor acts as a single key with dual role, driving either the opening (Smyd2) or closing (Smyd3) of the hosting enzyme conformation. The conformational preference of the holo enzymes may impact substrate recognition capability and specificity. To date, Smyd2 has been shown to exhibit broadest substrate specificities among the Smyds, and so far 19 non-histone proteins involved in multitude of biological processes have been identified as cellular substrates of Smyd2 14–16⁠ . Enhanced CTD flexibility, in concert with a larger interaction surface at the substrate crevice, could facilitate the recruitment of a broad span of different sized substrates. Similarly, reduced CTD flexibility and corresponding closed Smyd3 conformation in the holo state support its experimentally shown preference to harbor the PKMT activity on smaller substrates 5,8⁠ . The CTDs of the Smyd1-3 are structurally conserved and form similar TPR-repeat domain irrespective of low sequence identity. In Smyd1, the CTD deletion resulted in increased binding and methylation of H3 suggesting that the CTD inhibits substrate binding through steric effects 53⁠ . In Smyd3, the CTD has been shown to be essential for basal histone H4 methylation. Deletion of the complete CTD or the last three helices eliminated H4 methylation both in vitro and vivo

54

⁠ . In

support, we reported that restricted CTD motion and closed conformation could indeed stabilize the substrate crevice for binding

21

⁠ . In Smyd2, deletion of the CTD did not impair the methylation

activity on p53 peptide but led to 5-fold reduction in the methylation on p53 protein 19⁠ . Further, the CTD deletion while resulted in 5-fold increase in the methylation of H3 peptide, did not affect the activity on H3 protein. The exact mechanism of how the CTD plays multiple roles is hard to explain due to the lack of Smyd2 structures in complex with complete substrate proteins. However, the observed large conformational flexibility of the CTD implies on its ability to impact the PKMT activity of Smyd2 in a substrate-dependent manner. Different Modes of Substrate and Inhibitor Binding to the Smyds Different spatial and electrostatics features of the substrate binding pockets of Smyd2 and Smyd3 may provide an additional level of specificity in target recognition (Figure 5). Residues shaping the

12 ACS Paragon Plus Environment

Page 13 of 27

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

Journal of Chemical Information and Modeling

wall of substrate pockets are involved in substrate recruitment and binding. Different electrostatic potential, particularly at the vestibule of substrate pockets suggests that the enzymes may provide discrete interaction sites that determine the modes of substrate interaction and binding. Indeed, inspection of the X-ray structures of Smyd2 and Smyd3 resolved in complex with substrate peptides reveals distinct binding modes of their substrate peptides (Figure S6). In Smyd2, the peptides are bound forming an U-shaped conformation, while in Smyd3 the substrate peptides adopt an extended conformation. Intriguingly, notable difference is also seen in the binding of inhibitor molecules to Smyd2 and Smyd3 (Figure S7). To date, compounds of two chemical series have been identified as potent inhibitors of Smyd2 20,22,55

⁠ , which include AZ505, A-893 (benzoxazinone) and LLY-507 (pyrrolidine). These inhibitors

showed exceptional selectivity to Smyd2 and their binding was competitive with peptide substrates. X-ray structure of Smyd2-inhibitor complexes reveal the compounds to bind in a similar fashion (Figure S7); clasping the protein surface and well inserted into the target lysine access channel (deep hydrophobic pocket with conserved aromatic residues that shelter the target lysine of protein substrates, Figure S8). In contrast, the Smyd3 selective inhibitor (EPZO30456) binds in extended mode in the substrate crevice resembling the bound conformation of the substrate peptides (Figure S7)56⁠ . Lysine access channels of Smyd2 and Smyd3 are structurally similar with well superimposed backbone and composed of conserved aromatic residues. We performed a superimposition of Smyd3 X-ray structure against the Smyd2-inhibitor complexes (Figure S9). Several Smyd3 residues (V195,R140,Q256), that shape the entrance of lysine access channel, extend a physical hindrance for the Smyd2 selective inhibitors to access the channel. We envisage that distinct dynamics of Smyd2 and Smyd3 and associated conformational transitions provide different entry pathways which may impact both the selectivity and interaction of inhibitors.

13 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

CONCLUSIONS AdoMet cofactor acts as a single key with dual role upon binding to Smyd2 and Smyd3. The cofactor binding enhances the opening of Smyd2 and closing of Smyd3 conformation, via influencing the elasticity of the CTD. Differential CTD flexibility and concerted changes in substrate binding crevice have crucial implications in substrate recognition. The increased elasticity, together with large substrate pocket highlights the basis of the broad substrate specificity of Smyd2. Further, the different dynamics of the CTD's implies on its multiple roles in the methylation activity of the Smyds. Distinct properties of the substrate binding crevices indicate an additional layer of specificity in the target recognition. In support, X-ray structures of Smyd2 and Smyd3 in complex with substrate peptides reveal distinct modes of binding and conformations. Our results also highlight plausible reasons for the selective binding of inhibitors to the Smyds. We caution that the presented results describe the dynamical characteristics of the Smyds in sub µsecond regime. Hence, further MD studies at longer time scale, especially in the presence of substrates, are needed to shed light on the conformational transitions at higher time regime and the functional mechanism that drives the recognition of specific targets. On the whole, the present study provides a road map for future investigations focused on mechanistic bases of substrates recognition and small molecule mediated inhibition of Smyd's activity. These details are crucial in the realm of better inhibitors, designed for hindering Smyd2 activity on substrates whose methylation is related to diseased states.

14 ACS Paragon Plus Environment

Page 14 of 27

Page 15 of 27

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

Journal of Chemical Information and Modeling

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: Figures S1-S9 depicting Smyd2 structure, System stability vs time, RMSF, PCA projections, Electrostatic surface, Smyd-peptide complex, Smyd-inhibitor complex, Lysine access channel, and Overlay of Smyd3 X-ray structure against Smyd2-inhibitor complexes. ACKNOWLEDGEMENT We acknowledge Department of Chemistry, University of Rome “La Sapienza” and Cineca for computing resources. The authors acknowledge the anonymous reviewers for their comments which helped in improving the manuscript. ABBREVIATIONS SET, Suppressor of variegation, Enhancer of Zeste, Trithorax; MYND, Myeloid-Nervy-DEAF1; siRNA, Small interfering RNA; VEGFR1, Vascular endothelial growth factor receptor-1; Rb, Retinoblastoma; CPD, Computational protein design; MD, Molecular dynamics; HKMT, Histone lysine methyltranferase; PDB, Protein data bank; RMSD, root-mean square deviation; RMSF, rootmean-square fluctuation; PCA, principal component analysis

15 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

References (1)

Spellmon, N.; Holcomb, J.; Trescott, L.; Sirinupong, N.; Yang, Z. Structure and Function of SET and MYND Domain-Containing Proteins. Int. J. Mol. Sci. 2015, 16, 1406–1428.

(2)

Gottlieb, P. D.; Pierce, S. A.; Sims, R. J.; Yamagishi, H.; Weihe, E. K.; Harriss, J. V; Maika, S. D.; Kuziel, W. A.; King, H. L.; Olson, E. N.; Nakagawa, O.; Srivastava, D. Bop Encodes a Muscle-Restricted Protein Containing MYND and SET Domains and Is Essential for Cardiac Differentiation and Morphogenesis. Nat. Genet. 2002, 31, 25–32.

(3)

Hamamoto, R.; Furukawa, Y.; Morita, M.; Iimura, Y.; Silva, F. P.; Li, M.; Yagyu, R.; Nakamura, Y. SMYD3 Encodes a Histone Methyltransferase Involved in the Proliferation of Cancer Cells. Nat. Cell Biol. 2004, 6, 731–740.

(4)

Brown, M. A.; Sims, R. J.; Gottlieb, P. D.; Tucker, P. W. Identification and Characterization of Smyd2: A Split SET/MYND Domain-Containing Histone H3 Lysine 36-Specific Methyltransferase That Interacts with the Sin3 Histone Deacetylase Complex. Mol. Cancer 2006, 5, 26.

(5)

Van Aller, G. S.; Reynoird, N.; Barbash, O.; Huddleston, M.; Liu, S.; Zmoos, A.-F.; McDevitt, P.; Sinnamon, R.; Le, B.; Mas, G.; Annan, R.; Sage, J.; Garcia, B. A.; Tummino, P. J.; Gozani, O.; Kruger, R. G. Smyd3 Regulates Cancer Cell Phenotypes and Catalyzes Histone H4 Lysine 5 Methylation. Epigenetics 2012, 7, 340–343.

(6)

Foreman, K. W.; Brown, M.; Park, F.; Emtage, S.; Harriss, J.; Das, C.; Zhu, L.; Crew, A.; Arnold, L.; Shaaban, S.; Tucker, P. Structural and Functional Profiling of the Human Histone Methyltransferase SMYD3. PLoS One 2011, 6, e22290.

(7)

Kunizaki, M.; Hamamoto, R.; Silva, F. P.; Yamaguchi, K.; Nagayasu, T.; Shibuya, M.; Nakamura, Y.; Furukawa, Y. The Lysine 831 of Vascular Endothelial Growth Factor Receptor 1 Is a Novel Target of Methylation by SMYD3. Cancer Res. 2007, 67, 10759–10765.

(8)

Mazur, P. K.; Reynoird, N.; Khatri, P.; Jansen, P. W. T. C.; Wilkinson, A. W.; Liu, S.; Barbash, O.; Van Aller, G. S.; Huddleston, M.; Dhanak, D.; Tummino, P. J.; Kruger, R. G.; Garcia, B. A.; Butte, A. J.; Vermeulen, M.; Sage, J.; Gozani, O. SMYD3 Links Lysine Methylation of MAP3K2 to Ras-Driven Cancer. Nature 2014, 510, 283–287.

(9)

Huang, J.; Perez-Burgos, L.; Placek, B. J.; Sengupta, R.; Richter, M.; Dorsey, J. A.; Kubicek, S.; Opravil, S.; Jenuwein, T.; Berger, S. L. Repression of p53 Activity by Smyd2-Mediated Methylation. Nature 2006, 444, 629–632.

(10)

Saddic, L. A.; West, L. E.; Aslanian, A.; Yates, J. R.; Rubin, S. M.; Gozani, O.; Sage, J. Methylation of the Retinoblastoma Tumor Suppressor by SMYD2. J. Biol. Chem. 2010, 285, 37733–37740.

16 ACS Paragon Plus Environment

Page 16 of 27

Page 17 of 27

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

Journal of Chemical Information and Modeling

(11)

Zhang, X.; Tanaka, K.; Yan, J.; Li, J.; Peng, D.; Jiang, Y.; Yang, Z.; Barton, M. C.; Wen, H.; Shi, X. Regulation of Estrogen Receptor α by Histone Methyltransferase SMYD2-Mediated Protein Methylation. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 17284–17289.

(12)

Piao, L.; Kang, D.; Suzuki, T.; Masuda, A.; Dohmae, N.; Nakamura, Y.; Hamamoto, R. The Histone Methyltransferase SMYD2 Methylates PARP1 and Promotes poly(ADPRibosyl)ation Activity in Cancer Cells. Neoplasia 2014, 16, 257–264.

(13)

Hamamoto, R.; Toyokawa, G.; Nakakido, M.; Ueda, K.; Nakamura, Y. SMYD2-Dependent HSP90 Methylation Promotes Cancer Cell Proliferation by Regulating the Chaperone Complex Formation. Cancer Lett. 2014, 351, 126–133.

(14)

Olsen, J. B.; Cao, X.-J.; Han, B.; Chen, L. H.; Horvath, A.; Richardson, T. I.; Campbell, R. M.; Garcia, B. A.; Nguyen, H. Quantitative Profiling of the Activity of Protein Lysine Methyltransferase SMYD2 Using SILAC-Based Proteomics. Mol. Cell. Proteomics 2016, 15, 892–905.

(15)

Ahmed, H.; Duan, S.; Arrowsmith, C. H.; Barsyte-Lovejoy, D.; Schapira, M. An Integrative Proteomic Approach Identifies Novel Cellular SMYD2 Substrates. J. Proteome Res. 2016, 15, 2052–2059.

(16)

Lanouette, S.; Davey, J. A.; Elisma, F.; Ning, Z.; Figeys, D.; Chica, R. A.; Couture, J.-F. Discovery of Substrates for a SET Domain Lysine Methyltransferase Predicted by Multistate Computational Protein Design. Structure 2015, 23, 206–215.

(17)

Hamamoto, R.; Saloura, V.; Nakamura, Y. Critical Roles of Non-Histone Protein Lysine Methylation in Human Tumorigenesis. Nat. Rev. Cancer 2015, 15, 110–124.

(18)

Komatsu, S.; Imoto, I.; Tsuda, H.; Kozaki, K.; Muramatsu, T.; Shimada, Y.; Aiko, S.; Yoshizumi, Y.; Ichikawa, D.; Otsuji, E.; Inazawa, J. Overexpression of SMYD2 Relates to Tumor Cell Proliferation and Malignant Outcome of Esophageal Squamous Cell Carcinoma. Carcinogenesis 2009, 30, 1139–1146.

(19)

Wang, L.; Li, L.; Zhang, H.; Luo, X.; Dai, J.; Zhou, S.; Gu, J.; Zhu, J.; Atadja, P.; Lu, C.; Li, E.; Zhao, K. Structure of Human SMYD2 Protein Reveals the Basis of p53 Tumor Suppressor Methylation. J. Biol. Chem. 2011, 286, 38725–38737.

(20)

Ferguson, A. D.; Larsen, N. A.; Howard, T.; Pollard, H.; Green, I.; Grande, C.; Cheung, T.; Garcia-Arenas, R.; Cowen, S.; Wu, J.; Godin, R.; Chen, H.; Keen, N. Structural Basis of Substrate Methylation and Inhibition of SMYD2. Structure 2011, 19, 1262–1273.

(21)

Chandramouli, B.; Silvestri, V.; Scarno, M.; Ottini, L.; Chillemi, G. Smyd3 Open & Closed Lock Mechanism for Substrate Recruitment: The Hinge Motion of C-Terminal Domain Inferred from µ-Second Molecular Dynamics Simulations. Biochim. Biophys. Acta 2016, 1860, 1466–1474.

17 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

(22)

Sweis, R. F.; Wang, Z.; Algire, M.; Arrowsmith, C. H.; Brown, P. J.; Chiang, G. G.; Guo, J.; Jakob, C. G.; Kennedy, S.; Li, F.; Maag, D.; Shaw, B.; Soni, N. B.; Vedadi, M.; Pappano, W. N. Discovery of A-893, A New Cell-Active Benzoxazinone Inhibitor of Lysine Methyltransferase SMYD2. ACS Med. Chem. Lett. 2015, 6, 695–700.

(23)

Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. Comparison of Simple Potential Functions for Simulating Liquid Water. J. Chem. Phys. 1983, 79, 926.

(24)

Anandakrishnan, R.; Aguilar, B.; Onufriev, A. V. H++ 3.0: Automating pK Prediction and the Preparation of Biomolecular Structures for Atomistic Molecular Modeling and Simulations. Nucleic Acids Res. 2012, 40, 537–541.

(25)

Bietz, S.; Urbaczek, S.; Schulz, B.; Rarey, M. Protoss: A Holistic Approach to Predict Tautomers and Protonation States in Protein-Ligand Complexes. J. Cheminform. 2014, 6, 12.

(26)

Li, H.; Robertson, A. D.; Jensen, J. H. Very Fast Empirical Prediction and Rationalization of Protein pKa Values. Proteins 2005, 61, 704–721.

(27)

Hornak, V.; Abel, R.; Okur, A.; Strockbine, B.; Roitberg, A.; Simmerling, C. Comparison of Multiple Amber Force Fields and Development of Improved Protein Backbone Parameters. Proteins 2006, 65, 712–725.

(28)

Wang, J.; Cieplak, P.; Kollman, P. A. How Well Does a Restrained Electrostatic Potential (RESP) Model Perform in Calculating Conformational Energies of Organic and Biological Molecules? J. Comput. Chem. 2000, 21, 1049–1074.

(29)

Markham, G. D.; Norrby, P.-O.; Bock, C. W. S-Adenosylmethionine Conformations in Solution and in Protein Complexes: Conformational Influences of the Sulfonium Group. Biochemistry 2002, 41, 7636–7646.

(30)

Vanquelef, E.; Simon, S.; Marquant, G.; Garcia, E.; Klimerak, G.; Delepine, J. C.; Cieplak, P.; Dupradeau, F.-Y. R.E.D. Server: A Web Service for Deriving RESP and ESP Charges and Building Force Field Libraries for New Molecules and Molecular Fragments. Nucleic Acids Res. 2011, 39, 511-517.

(31)

Bayly, C. I.; Cieplak, P.; Cornell, W.; Kollman, P. A. A Well-Behaved Electrostatic Potential Based Method Using Charge Restraints for Deriving Atomic Charges: The RESP Model. J. Phys. Chem. 1993, 97, 10269–10280.

(32)

Peters, M. B.; Yang, Y.; Wang, B.; Füsti-Molnár, L.; Weaver, M. N.; Merz, K. M. Structural Survey of Zinc Containing Proteins and the Development of the Zinc AMBER Force Field (ZAFF). J. Chem. Theory Comput. 2010, 6, 2935–2947.

(33)

Li, P.; Roberts, B. P.; Chakravorty, D. K.; Merz, K. M. Rational Design of Particle Mesh Ewald Compatible Lennard-Jones Parameters for +2 Metal Cations in Explicit Solvent. J. Chem. Theory Comput. 2013, 9, 2733–2748.

18 ACS Paragon Plus Environment

Page 18 of 27

Page 19 of 27

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

Journal of Chemical Information and Modeling

(34)

Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. . Numerical Integration of the Cartesian Equations of Motion of a System with Constraints: Molecular Dynamics of N-Alkanes. J. Comput. Phys. 1977, 23, 327–341.

(35)

Darden, T.; York, D.; Pedersen, L. Particle Mesh Ewald: An N⋅log(N) Method for Ewald Sums in Large Systems. J. Chem. Phys. 1993, 98, 10089-10092.

(36)

Izaguirre, J. A.; Catarello, D. P.; Wozniak, J. M.; Skeel, R. D. Langevin Stabilization of Molecular Dynamics. J. Chem. Phys. 2001, 114, 2090-2098.

(37)

Roe, D. R.; Cheatham, T. E. PTRAJ and CPPTRAJ: Software for Processing and Analysis of Molecular Dynamics Trajectory Data. J. Chem. Theory Comput. 2013, 9, 3084–3095.

(38)

Michaud-Agrawal, N.; Denning, E. J.; Woolf, T. B.; Beckstein, O. MDAnalysis: A Toolkit for the Analysis of Molecular Dynamics Simulations. J. Comput. Chem. 2011, 32, 2319–2327.

(39)

Van Der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark, A. E.; Berendsen, H. J. C. GROMACS: Fast, Flexible, and Free. J. Comput. Chem. 2005, 26, 1701–1718.

(40)

Amadei, A.; Linssen, A. B.; Berendsen, H. J. Essential Dynamics of Proteins. Proteins 1993, 17, 412–425.

(41)

Hubbard, S., Thornton, J. “NACCESS”, Computer Program. 1993.

(42)

Baker, N. A.; Sept, D.; Joseph, S.; Holst, M. J.; McCammon, J. A. Electrostatics of Nanosystems: Application to Microtubules and the Ribosome. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 10037–10041.

(43)

Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. UCSF Chimera--a Visualization System for Exploratory Research and Analysis. J. Comput. Chem. 2004, 25, 1605–1612.

(44)

Hunter, J. D. Matplotlib: A 2D Graphics Environment. Comput. Sci. Eng. 2007, 9, 90–95.

(45)

Wu, J.; Cheung, T.; Grande, C.; Ferguson, A. D.; Zhu, X.; Theriault, K.; Code, E.; Birr, C.; Keen, N.; Chen, H. Biochemical Characterization of Human SET and MYND DomainContaining Protein 2 Methyltransferase. Biochemistry 2011, 50, 6488–6497.

(46)

Chandramouli, B.; Chillemi, G.; Giombini, E.; Capobianchi, M. R.; Rozera, G.; Desideri, A.; Kusters, J. Structural Dynamics of V3 Loop with Different Electrostatics: Implications on Co-Receptor Recognition: A Molecular Dynamics Study of HIV gp120. J Biomol Struct Dyn 2013, 31, 403–413.

(47)

D’Abramo, M.; Bešker, N.; Desideri, A.; Levine, A. J.; Melino, G.; Chillemi, G. The p53 Tetramer Shows an Induced-Fit Interaction of the C-Terminal Domain with the DNABinding Domain. Oncogene 2016, 35, 3272-3281.

19 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling

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

(48)

Tesauro, C.; Morozzo della Rocca, B.; Ottaviani, A.; Coletta, A.; Zuccaro, L.; Arnò, B.; D’Annessa, I.; Fiorani, P.; Desideri, A. Molecular Mechanism of the Camptothecin Resistance of Glu710Gly Topoisomerase IB Mutant Analyzed in Vitro and in Silico. Mol. Cancer 2013, 12, 100.

(49)

Mancini, G.; D’Annessa, I.; Coletta, A.; Sanna, N.; Chillemi, G.; Desideri, A. Structural and Dynamical Effects Induced by the Anticancer Drug Topotecan on the Human Topoisomerase I – DNA Complex. PLoS One 2010, 5, e10934.

(50)

Mancini, G.; Zazza, C. F429 Regulation of Tunnels in Cytochrome P450 2B4: A Top Down Study of Multiple Molecular Dynamics Simulations. PLoS One 2015, 10, e0137075.

(51)

Xu, S.; Wu, J.; Sun, B.; Zhong, C.; Ding, J. Structural and Biochemical Studies of Human Lysine Methyltransferase Smyd3 Reveal the Important Functional Roles of Its Post-SET and TPR Domains and the Regulation of Its Activity by DNA Binding. Nucleic Acids Res. 2011, 39, 4438–4449.

(52)

Abu-Farha, M.; Lambert, J.-P.; Al-Madhoun, A. S.; Elisma, F.; Skerjanc, I. S.; Figeys, D. The Tale of Two Domains: Proteomics and Genomics Analysis of SMYD2, a New Histone Methyltransferase. Mol. Cell. Proteomics 2008, 7, 560–572.

(53)

Sirinupong, N.; Brunzelle, J.; Ye, J.; Pirzada, A.; Nico, L.; Yang, Z. Crystal Structure of Cardiac-Specific Histone Methyltransferase SmyD1 Reveals Unusual Active Site Architecture. J. Biol. Chem. 2010, 285, 40635–40644.

(54)

Brown, M. A.; Foreman, K.; Harriss, J.; Das, C.; Zhu, L.; Edwards, M.; Shaaban, S.; Tucker, H. C-Terminal Domain of SMYD3 Serves as a Unique HSP90-Regulated Motif in Oncogenesis. Oncotarget 2015, 6, 4005–4019.

(55)

Nguyen, H.; Allali-Hassani, A.; Antonysamy, S.; Chang, S.; Chen, L. H.; Curtis, C.; Emtage, S.; Fan, L.; Gheyi, T.; Li, F.; Liu, S.; Martin, J. R.; Mendel, D.; Olsen, J. B.; Pelletier, L.; Shatseva, T.; Wu, S.; Zhang, F. F.; Arrowsmith, C. H.; Brown, P. J.; Campbell, R. M.; Garcia, B. A.; Barsyte-Lovejoy, D.; Mader, M.; Vedadi, M. LLY-507, a Cell-Active, Potent, and Selective Inhibitor of Protein-Lysine Methyltransferase SMYD2. J. Biol. Chem. 2015, 290, 13641–13653.

(56) Mitchell, L. H.; Boriack-Sjodin, P. A.; Smith, S.; Thomenius, M.; Rioux, N.; Munchhof, M.; Mills, J. E.; Klaus, C.; Totman, J.; Riera, T. V; Raimondi, A.; Jacques, S. L.; West, K.; Foley, M.; Waters, N. J.; Kuntz, K. W.; Wigle, T. J.; Scott, M. P.; Copeland, R. A.; Smith, J. J.; Chesworth, R. Novel Oxindole Sulfonamides and Sulfamides: EPZ031686, the First Orally Bioavailable Small Molecule SMYD3 Inhibitor. ACS Med. Chem. Lett. 2016, 7, 134–138.

20 ACS Paragon Plus Environment

Page 20 of 27

Page 21 of 27

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

Journal of Chemical Information and Modeling

Figure Legends Figure 1 A) X-ray structure of Smyd2-AdoMet binary complex. Smyd2's individual domains are colored for clarity. AdoMet is represented in ball-stick mode. B) Overlay of Smyd2 (PDB ID:3TG4) and Smyd3 (PDB ID:3RU0) structures. The C-terminal domains of Smyd2 (red) and Smyd3 (blue) are differentiated. The N-terminal domains are shown in dark cyan. Figure 2 Structural deviation and flexibility. A) RMSD from the X-ray structure, as a function of simulation time, for apo (gray) and holo (blue) systems. B) Average RMSD of individual domains with respect to average structure from the respective simulations. C) Average RMSD calculated by swapping apo and holo average structures. Error bar indicates the std. deviation. D) RMSF as a function of residue index. Anti-parallel α-helices of CTD are shaded in green. Asterisks (in red) over the peaks denote the mobile regions. Figure 3 Inter-residue dynamic cross correlation analysis. A) 2D-DCC matrix for apo (upper triangle) and holo (lower triangle) systems. Bottom panels: 3D mapping of negative correlations for apo (B) and holo (C) systems, where Cα pairs with correlations in range -0.65< Cij