Study of SHMT2 inhibitors and their Binding Mechanism by

Subscriber access provided by Nottingham Trent University

Computational Biochemistry

Study of SHMT2 inhibitors and their Binding Mechanism by Computational Alanine Scanning LIping He, Jinxiao Bao, Yunpeng Yang, Suzhen Dong, Lujia Zhang, Yifei Qi, and John Z.H. Zhang J. Chem. Inf. Model., Just Accepted Manuscript • DOI: 10.1021/acs.jcim.9b00370 • Publication Date (Web): 23 Aug 2019 Downloaded from pubs.acs.org on August 27, 2019

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 19 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

Study of SHMT2 inhibitors and their Binding Mechanism by Computational Alanine Scanning Liping He1, Jingxiao Bao1, Yunpeng Yang1, Suzhen Dong1, Lujia Zhang1,2, Yifei Qi1,2*, and John Z.H. Zhang1,2,3* 1Shanghai

Engineering Research Center of Molecular Therapeutics & New Drug Development, Shanghai Key Laboratory of Green Chemistry & Chemical Process, School of Chemistry and Molecular Engineering, East China Normal University, Shanghai 200062, China 2NYU-ECNU

Center for Computational Chemistry at NYU Shanghai, Shanghai 200062, China

3Department

of Chemistry, New York University, NY, NY 10003, USA

ABSTRACT Mitochondrial serine hydroxymethyl transferase isoform 2 (SHMT2) has attracted increasing attention as a pivotal catalyzing regulator of the serine/glycine pathway in the onecarbon metabolism of cancer cells. However, few inhibitors that target this potential anticancer target have been discovered. Quantitative characterization of the interactions between SHMT2 and its known inhibitors should benefit future discovery of novel inhibitors. In this study, we employed a recently developed alanine-scanning-interaction-entropy method to quantitatively calculate the residue-specific binding free energy of 28 different SHMT2 inhibitors that originate from a same skeleton. Major contributing residues from SHMT2 and chemical groups from the inhibitors were identified and the binding energy of each residue was quantitatively determined, revealing essential features of the protein-inhibitor interaction. The most important contributing residue is Y105 of the B chain followed by L166 of the A chain. The calculated protein-ligand binding free energies are in good agreement with the experimental results, and showed better correlation and smaller errors compared with those obtained using the conventional MM/GBSA with normal mode method. These results may aid the rational design of more effective SHMT2 inhibitors.

*Correspondence

to: [email protected], [email protected]

Key words: SHMT2, interaction entropy, binding free energy, alanine scanning, GBSA, dielectric constant

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

Page 2 of 19

INTRODUCTION Abnormal cellular metabolism such as highly aerobic glycolysis and folate-dependent onecarbon metabolism is one of the most important features of infinite proliferation of cancer cells.1-3 The serine catabolism represents a major one-carbon source, which not only supplies the essential components for the synthesis of purines and pyrimidine nucleotides, but also contributes to the maintenance of redox homeostasis under hypoxic conditions.4-7 Many studies have revealed that the one-carbon metabolism in mitochondria has deep significance in human tumors, such as B-cell lymphomas, non-small-cell lung cancer, breast cancer, and melanoma.810

Various antifolate agents such as pemetrexed, 5-fluorouracil, and methotrexate are

routinely used as clinical chemotherapeutics targeting the one-carbon metabolism.11-13 The serine hydroxy methyltransferase enzyme (SHMT, E.C. 2.1.2.1) has two members, SHMT1 and SHMT2, that catalyze the conversion of serine and tetrahydrofolate into glycine and 5,10-methylene-tetrahydrofolate in cytoplasm and mitochondrion, respectively.9-10, 14 The cytosolic SHMT isoform (SHMT1) and mitochondrial isoform (SHMT2) share 66% sequence identity but have distinct cellular roles.15 Results from recent research suggest that SHMT2 is of vital importance in the growth and reprogramming of cancer cells and is an important anticancer target as a main catalyzing enzyme in the one-carbon metabolism.16-17 Therefore, targeting SHMT2 is potentially an effective way for the treatment of various cancers.6,

9

Although the importance of SHMT2 in cancer growth and proliferation is gaining many attentions, only a few SHMT2 inhibitors have been discovered.18-19 Most of these inhibitors are pyrazolopyran derivatives sharing a 1,4-dihydropyrano [2,3-c] pyrazole ring scaffold, which were initially reported as plant SHMT inhibitors, and later found to inhibit the Plasmodium falciparum SHMT and human SHMT as well.18 Recently, Ducker et al. optimized the structure of this scaffold and found two compounds that bind human SHMT1/2 with nanomolar affinity.10 In a subsequent patent, they also synthesized and evaluated a series of derivatives of these two compounds with IC50 ranges from ~10 nM to 5 μM.20 These compounds provide an excellent opportunity to characterize the binding pocket and interactions between human SHMT2 and the inhibitors using computational methods. The quantitative understanding of protein-ligand interaction and the accurate prediction of ligand binding affinity by computational methods are essential in rational drug design.21-23 In protein-ligand interactions, small molecules typically bind to a few hot-spots residues that contribute most of the binding energy.24-25 Many computational approaches have been developed to predict protein–ligand binding free energy and hot-spot residues. In general, they can be classified into knowledge-based,26-27 empirical formula-based,28-29 and physics-based methods.30-32 Among them, the MM/PB(GB)SA approach is a widely used method with a moderate computational cost.33 Recently, we developed a new method that uses alanine scanning (AS) together with the interaction entropy (IE) approach to calculate protein-ligand binding free energy, identify hot spots and explain the structure–activity relationship (SAR) by 2


Page 3 of 19 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


quantitatively analyzing the contribution of each pocket residue.34-37 In this study, to gain quantitative understanding of the binding between SHMT2 and its inhibitors, we used the AS-IE method to analyze the interactions of 28 SHMT2 inhibitors that share the same skeleton. Key contributing residues from SHMT2 were identified and quantitative structure-activity relationship from the inhibitors were calculated. The calculated total binding free energies are in good correlation with the experimental results and comparisons with those from the conventional MM/GBSA method were made.

METHODS Molecular dynamics simulations Of the 28 compounds studied, only compound 16 has a complex structure with SHMT2 (PDB ID 5V7I10). This structure was used to manually build the complex structures of the other 27 compounds by directly modifying the chemical groups of compound 16. The force field parameters for the 28 compounds were obtained using the gaff force field.38 The charges for the 28 ligands were calculated using antechamber in antechamber of Amber16 with AM1-BCC method.39 To keep conditions identical with experiment, protonation of the protein-ligand complex was done at pH 7.4. The parameters of non-standard protein residue LLP were prepared according to the amber tutorial (http://ambermd.org/tutorials/basic/tutorial5/). Molecular dynamics (MD) simulations were carried out using Amber1640-41 and the ff14SB force field.42 The protein-ligand complex was solvated in a cubic box using TIP3P water, extending 10 Å away from the protein. Cut-off of non-bonded interactions was set to 10 Å. Counterions were added to neutralize the systems. Each system was minimized for 4000 steps and heated from 0 to 300 K in 1 ns with Langevin dynamics, followed by a 10-ns MD simulation in NPT ensemble with a time step of 2 fs. The SHAKE algorithm43 was used to constrain the bonds involving hydrogen atoms. Binding free energy calculation In the computational alanine-scanning approach, it is assumed that the mutated alanine contributes negligibly to the binding free energy and the free energy difference before and after alanine mutation gives a quantitative measure of the contribution of the specific residue to the total binding free energy. Therefore, the difference of binding free energy for mutating residue X to A (alanine) is defined by xa a x Gbind  Gbind  Gbind xa xa  Ggas  Gsol

(1)

where the binding free energy for the gas-phase and solvation components were evaluated by the following scheme: 3



x a a x Ggas  Ggas  Ggas

x a a x Gsol  Gsol  Gsol

Page 4 of 19

(2)

(3)

where Ggas and Ggas is the gas-phase component of the binding free energy between the a

x

ligand L and residue A and X, respectively. This alanine scanning approach was applied to each pocket residue that is within 6 Å of any atoms of the ligand to obtain residue-specific contributions to the total binding energy. The gas-phase free energy between a residue and the ligand was evaluated by summation of the enthalpy and entropy, which was calculated using the standard molecular mechanics and the IE method, respectively:34-35, 44-46 x Ggas  Eintx  T Sintx x

 Eintx  KT ln e Eint

(4)

and similarly, for the mutant a

a a Ggas  Eint  KT ln e Eint

(5)

Here Eintx and Einta denote the interaction energies (van der Waals and electrostatic energies) between the ligand and residue X and A, respectively. The exponential argument β is

1/ KT ,

and Eintx/a is the deviation from the average Eintx/a . The exponential average was evaluated by discrete time averaging,

e

x Eint

1  N

N

 e

x Eint  ti 

i 1

(6)

where N is the number of MD snapshots. Finally, Eq. (2) becomes x a x a x a Ggas  E gas  T S gas a x a  Eint  Eintx  KT ln e Eint  ln e Eint   

(7)

The solvation energy in equation (3) was calculated using the OBC GBSA model (igb =2 in Amber),33, 47- 48 with a dielectric constant of 1, 3, and 5 for nonpolar, polar, and charged residues, respectively.49-50

4


Page 5 of 19 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


Combining the alanine-scanning and the IE method described above, the total proteinligand binding free energy can be approximated by the summation,36-37 x a Gbind    Gbind x

(9)

where the summation is over all the contributing residues (6 Å around the ligand). Equation (9) is not theoretically rigorous, but provides a good estimate of the total binding free energy, especially when relative binding free energies are important for a series of compounds targeting the same receptor. In this calculation, a total of 1000 snapshots from the last 5 ns of the MD trajectory were used to calculate the binding energy using the AS-IE method. For comparison, we also calculated the binding free energy using the conventional MM/GBSA method using the MMPBSA.py script51 with igb parameter also set to 2. We set the dielectric constant to one in the MM/GBSA calculation as this is typical in standard MM/GBSA calculation and a total of 100 snapshots were used to calculate the MM/GBSA energy. Considering the computational cost, we used 10 snapshots to perform the normal mode calculation for each complex.

RESULTS AND DISCUSSION Protein-ligand complex structures We have collected 28 compounds that bind human SHMT2 with different binding affinities20 (Table S1). Five representative compounds with different magnitude of binding affinities are shown in Table 1. These compounds share a skeleton structure that consists of a 1,4-dihydropyrano [2,3-c] pyrazole ring and a benzene ring (Fig. 1). Among them, compound 16 has a co-crystal structure with SHMT2 (PDB code: 5V7I10) and was used as a template to build all other 27 complex structures for MD simulations (Fig. 2).

Figure 1. 2D skeleton structure of the 28 SHMT2 inhibitors

5



Page 6 of 19

Figure 2. Co-crystal structure of SHMT2 and compound 16. The receptor SHMT2 consists of two chains, the A (orange) and B (green). The residues in the A chain is labeled directly by residue number, while residues in the B chain has an extra apostrophe after the number, and this convention is used throughout this paper. The red, yellow and blue lines represent the π-π, π-alkyl, and van der Waals interactions. The picture was prepared with Pymol52 using the coordinates from PDB ID 5V7I.10 Table 1 The structures of the five representative compounds and their experimental binding energies. SHMT2

ΔGexp*

IC50 (nM)

(kcal/mol)

1

0.17

13.3

8

4.0

11.5

13

35

10.2

17

100

9.6

Compound

Representative 2D-structure

6


Page 7 of 19 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


27

*Experimental

5000

7.2

binding free energy was estimated using ΔG = RT ln IC50 with T = 300 K.

Identification of hot-spots in SHMT2-ligand binding The binding free energy of 28 complexes were calculated by the AS-IE method and the detailed decomposition of complex 1, the strongest inhibitor, is shown in Table 2. The total binding free energy of the complex consists of enthalpy and entropy from the interactions with each pocket residue, and the enthalpy is further decomposed into four components, van der Waals, electrostatic energy, GB and non-polar solvation energy. Over all, the van der Waals interaction contributes most of the binding energy. Of the binding residues, Tyr105’ is the dominant contributing residue followed by Leu166 (Fig. 3A and 3B). The Arg425 and Leu172 are also important residues contributing to the binding. As shown in Table 2 for compound 1, these four are hot spots whose binding interactions with the compound are dominated by hydrophobic interactions. Table 2. Residue-specific binding free energies of SHMT2 and compound 1 from AS-IE. Only residues with ΔΔG > 0.6 kcal/mol are listed and all values are in kcal/mol. Residue

ΔΔEvdw

ΔΔEele

ΔΔGB

Y105’A

5.6 ± 0.2

0.1 ± 0.2

-0.7 ± 0.1

L166A

3.4 ± 0.1

-0.2 ± 0.1

R425A

2.0 ± 0.3

0.1 ± 0.0 -0.3 ± 0.1

L172A

2.0 ± 0.2

0.0 ± 0.0

-0.1 ± 0.0

H171A

1.2 ± 0.1

-0.1 ± 0.1

0.1 ± 0.0

Y176A

1.5 ± 0.1

-0.1 ± 0.0

-0.1 ± 0.0

K409A

1.1 ± 0.3

2.3 ± 0.0

-2.3 ± 0.0

N410A

1.3 ± 0.2

0.6 ± 0.2

-0.6 ± 0.1

E98’A

1.3 ± 0.2

0.3 ± 0.1

-0.7 ± 0.1

N408A

1.1 ± 0.1

0.1 ± 0.1

-0.4 ± 0.0

0.2 ± 0.1

ΔΔNP 0.5 ± 0.0 0.2 ± 0.0 0.1 ± 0.0 0.1 ± 0.0 0.0 ± 0.0 0.1 ± 0.0 0.3 ± 0.0 0.1 ± 0.0 0.2 ± 0.0 0.1 ± 0.0

ΔΔH 5.5 ± 0.1 3.4 ± 0.1 2.0 ± 0.2 2.0 ± 0.2 1.2 ± 0.1 1.4 ± 0.1 1.4 ± 0.3 1.4 ± 0.3 1.0 ± 0.1 0.9 ± 0.2

7


IE -0.6 ± 0.1 -0.5 ± 0.3 -0.1 ± 0.1 -0.2 ± 0.1 0.1 ± 0.0 -0.2 ± 0.1 -0.2 ± 0.1 -0.3 ± 0.2 -0.1 ± 0.0 -0.1 ± 0.1

x a Gbind

4.9 ± 0.1 2.9 ± 0.4 1.9 ± 0.2 1.8 ± 0.2 1.3 ± 0.1 1.3 ± 0.2 1.2 ± 0.4 1.1 ± 0.1 1.0 ± 0.2 0.8 ± 0.2


I183A

1.1 ± 0.1

-0.1 ± 0.0

-0.3 ± 0.1

S226A

0.5 ± 0.0 22.1 ± 0.4

0.3 ± 0.1

-0.1 ± 0.0

3.1 ± 0.1

-5.2 ± 0.2

Total

0.1 ± 0.0 0.0 ± 0.0 1.9 ± 0.0

0.9 ± 0.0 0.7 ± 0.1 21.7 ± 0.5

Page 8 of 19

-0.3 ± 0.1 -0.1 ± 0.1 -2.4 ± 0.2

0.6 ± 0.1 0.6 ± 0.2 19.4 ± 0.6

Figure 3. Residue-specific interactions between pocket residues and compound 1. (A) Average contributions of the pocket residues in SHMT2 to compound 1. Only residues with ΔΔG > 0.6 kcal/mol are plotted. (B) Interaction diagram of pocket residues with compound 1. Residues that mainly provide van der Waals and electrostatic interactions are colored in blue and brown. The radius of the sticks is proportional to the binding free energy of each residue. Figure B was prepared with Pymol52 using the coordinates from PDB ID 5V7I.10 We further analyzed the average residue-specific free energy in the 28 complexes. On average, a few pocket residues show lager contributions to the binding energy compared to others. The residues that contribute more than 1 kcal/mol are identified as hot spots, including Tyr105’, Arg425, Leu166, Asn410, Leu172, Tyr176 and Lys409 with an average contribution 8


Page 9 of 19 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


of 3.6, 1.8, 1.7, 1.4, 1.2, 1.2 and 1.0 kcal/mol (Fig. 4A). The interaction diagram of the crystal structure of 5V7I explains the quantitative residue-specific binding free energies (Fig. 4B). Tyr105’ forms strong π-π and π-alkyl interaction with the benzene and methyl group of the inhibitors with a distance of 4.7 Å and 4.6 Å, respectively (Fig. 2). As the main contributor, Tyr105’ is predicted as the strongest hot spot in SHMT2-inhibitor interactions. Arg425, the second highest contributing residue, forms a strong van der Waals interaction with the isopropyl group of inhibitors with a distance of 4.0 Å (Fig. 2). There is a π-sigma interaction between Leu166 and imidazole ring of ligand. These three residues are most important in stabilizing the protein-inhibitor complex based on their contribution to the binding energy.

Figure 4. Residue-specific interactions between pocket residues and the inhibitors. (A) Average contributions of the pocket residues in SHMT2. The hot spots are colored with deeper blue. The black line of each residue presents the range of their contribution. (B) Interaction diagram of pocket residues with 28 inhibitors. The radius of the sticks is proportional to the average binding free energy of each residue. Figure B was prepared with Pymol52 using the coordinates from PDB ID 5V7I.10 Analysis of binding energy of different ligands 9



The above hot-spot analysis calculated the average interaction of each pocket residue to all inhibitors. To characterize the interaction between a pocket residue and different ligands and identify the key residues that determine the binding strength, we classified the 28 complexes into five different groups according to their magnitude of binding affinities, group 1-5 consist of compound 23-28 (IC50 ≥ 10000 nM), 17-22 (100 nM ≤ IC50

Study of SHMT2 inhibitors and their Binding Mechanism by

Recommend Documents