The catalytic mechanism of the Serine Hydroxymethyltransferase – a

2 days ago - Serine Hydroxymethyltransferase (SHMT) is an important drug target to fight malaria - one of the most devastating infectious diseases tha...
0 downloads 0 Views 138MB Size
Subscriber access provided by University of South Dakota

Article

The catalytic mechanism of the Serine Hydroxymethyltransferase – a computational ONIOM QM/MM study Henrique Silva Fernandes, Maria João Ramos, and Nuno M. F. Sousa A Cerqueira ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b02321 • Publication Date (Web): 18 Sep 2018 Downloaded from http://pubs.acs.org on September 18, 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 57 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

ACS Catalysis

The catalytic mechanism of the Serine Hydroxymethyltransferase – a computational ONIOM QM/MM study Henrique S. Fernandes§, Maria João Ramos, and Nuno M. F. S. A. Cerqueira§* UCIBIO@REQUIMTE, Departamento de Química e Bioquímica, Faculdade de Ciências da Universidade do Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal.

* Corresponding author: [email protected]

Abstract Serine Hydroxymethyltransferase (SHMT) is an important drug target to fight malaria - one of the most devastating infectious diseases that accounted in 2016 with 216 million cases and almost 450 thousand deaths. In this paper, computational studies were carried out to unveil the catalytic mechanism of SHMT using QM/MM methodologies. This enzyme is responsible for the extraordinary cyclisation of a tetrahydrofolate (THF) into 5,10-methylene-THF. This process is catalyzed by a pyridoxal-5’-phosphate (PLP) cofactor that binds L-serine and from which one molecule of L-glycine is produced. The results show that the catalytic process takes place in eight sequential steps that involve an α-elimination, the cyclization of the 5-hydroxymethyl-THF intermediate into 5,10-methylene-THF and the protonation of the quinonoid intermediate. According to the calculated energetic profile, the rate-limiting step of the full mechanism is the elimination of the hydroxymethyl group, from which results a formaldehyde intermediate that then becomes covalently bonded to the THF cofactor. The calculated barrier (DLPNO-CCSD(T)/CBS:ff99SB) for the ratelimiting step (18.0 kcal/mol) agrees very well with the experimental kinetic results (15.7-16.2 kcal/mol). The results also highlight the key role played by Glu57 during the full catalytic process and particularly in the first step of the mechanism that requires an anionic Glu57, contrasting with some proposals available in the literature for

ACS Paragon Plus Environment

1

ACS Catalysis 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 57

this step. It was also concluded that the cyclisation of THF must take place in the enzyme, rather than in solution as it has been proposed also in the past. All of these results together provide knowledge and insight on the catalytic mechanism of SHMT that now can be used to develop inhibitors targeting SHMT and therefore anti-malaria drugs.

Keywords Serine Hydroxymethyltransferase (SHMT), malaria, tetrahydrofolate (THF), pyridoxal-5’-phosphate (PLP), serine, catalytic mechanism, ONIOM, QM/MM

ACS Paragon Plus Environment

2

Page 3 of 57 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

ACS Catalysis

1. Introduction Malaria is one of the most devastating infectious diseases that is caused by a parasitic infection transmitted by mosquitoes. According to the last World Health Organization (WHO), 216 million new cases of Malaria and almost 450 thousand deaths were caused by this disease in 2016. Despite the resources and investments allocated to find a universal cure for this disease (2.7 billion dollars in 2016), the control and eventual elimination of malaria is seriously questioned due to the resistance of parasites to antimalarial drugs. This effect has become even more dramatic in five countries of the Greater Mekong Subregion and in the Cambodia-Thailand border, where Plasmodium falciparum - the species that causes the deadliest form of malaria 1 - has become resistant to the major antimalarial drugs used clinically. 1-2 All of these facts together demand an urgent need for new drug targets, which can be used to design new and more efficient therapeutics against the malaria pathogen. Particular attention is being given to those that provide a selective interference with the parasite metabolism without harming the human host.

1, 3-4

In this regard, some of the most promising targets are enzymes

involved in the de novo purine biosynthesis in the malaria parasites, such as Serine Hydroxymethyltransferase (SHMT). 5-8

Figure 1 – Structure of the external aldimine (pyridoxal-5’-phosphate + L-serine) (Left) and the tetrahydrofolate (THF) cofactor (Right).

The SHMT (EC 2.1.2.1) is a pyridoxal-5’-phosphate (PLP)-dependent enzyme that catalyzes the reversible transfer of a carbon-unit from the L-serine to a second cofactor, the tetrahydrofolate (THF) (Figure 1). At the end of the reaction, a glycine molecule is released as well as a 5,10-methylene-THF

ACS Paragon Plus Environment

3

ACS Catalysis 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 4 of 57

molecule, which is a precursor in the synthesis of purines and thymidylate.9 Since the purines biosynthesis is vital for the pathogen survival (synthesis of nucleic acids), the inhibition of SHMT has been pointed out as an important drug target for malaria.

5

Furthermore, the comparison between the

tridimensional structure of the human and P. falciparum forms of the SHMT revealed key differences that can help to rationalize the specific synthesis of new inhibitors for the P. falciparum form of the SHMT. 10 Despite the importance that SHMT has gathered in the last two years, this enzyme continues to be poorly understood. This happens due to the complex chemistry that it catalyzes, which requires the participation of two cofactors (PLP and THF) and the multi sequential steps required for the cyclization of the THF into the final product, 5,10-methylene-THF. This knowledge is very important since it will provide clues about the enzymatic activity and how it can be inhibited. This information can then be used to develop new inhibitors targeting this enzyme with potential application as new anti-malaria drugs. Since SHMT is a PLP dependent enzyme, the catalytic activity of this enzyme is composed of three stages. The first two stages are common to all PLP-dependent enzymes and were already studied.

11-12

The first stage

involves the activation of the enzyme, in which the PLP cofactor becomes covalently bonded to an active site Lys residue and forms what is commonly known as the internal aldimine. The second stage is triggered when the substrate is available, and it becomes covalently bonded to the PLP cofactor, forming the external aldimine. The third stage is what differentiates the PLPdependent enzymes, and it is specific to each class of enzymes. In the case of SHMT, the enzyme catalyzes an α-elimination of L-Ser that is subsequently inserted into the THF cofactor. (Figure 2) The experimental kinetic data available for the E. coli form of the SHMT indicates an activation barrier limit of 15.7 – 16.2 kcal/mol based on the available kcat values of 686.6 ± 21.7 min-1 (pH 7.2, 20 °C)

13

and 790 min-1 (pH

7.2, 30 °C) 14.

ACS Paragon Plus Environment

4

Page 5 of 57 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

ACS Catalysis

Figure 2 – Scheme representing the mechanism behind the internal and external aldimine formation for PLP-dependent enzymes and the specific α-elimination reaction.

Currently, there is no accepted mechanism for the α-elimination catalyzed by SHMT, and only two hypotheses are available in the literature. They diverge in the way through which the hydroxymethyl group is transferred from L-serine, which is bonded to the PLP cofactor and form an external aldimine, to THF. One of the hypothesis suggests that the reaction takes place in one step, where nitrogen N1 of THF makes a nucleophilic attack on the Cβ of the L-Ser promoting the α-elimination reaction. In the other hypothesis, the reaction requires two steps. The first step involves the formation of a formaldehyde molecule (through the dissociation of the hydroxymethyl group from the external aldimine), which then undergoes a nucleophilic attack from nitrogen N1 of the

ACS Paragon Plus Environment

5

ACS Catalysis 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

THF cofactor.

15

Page 6 of 57

These two hypotheses also diverge regarding the protonation

state of Glu57, which interacts very closely with the hydroxyl group of the external aldimine. In the first proposal, Glu57 is proposed to be protonated, whereas, in the second proposal, the mechanism requires an unprotonated Glu57. 15-17 The role played by Glu57 in the reaction is known to be fundamental for catalysis since several mutagenic studies have shown that its mutation by Gln, Leu, or Lys residues reduces the activity of the enzyme drastically (more than 300 times), but the role played by this residue in the mechanism continues to be poorly understood. 16, 18 Some authors also suggest that the reaction of the formaldehyde molecule with THF can occur in the solvent and not in the enzyme

15, 19

. However, so far,

no experimental result is currently available to provide insights into this hypothesis. In this paper, the catalytic mechanism of the SHMT was studied employing computational means and using QM/MM methodologies. Based on the mechanistic proposals available in the literature, and described above, two models of the enzyme were set up to assess the role of the protonation of the Glu57 residue in the catalytic mechanism and also to understand if the hydroxymethyl group is transferred directly to the THF cofactor or it involves the formation of a formaldehyde intermediate in an intermediate step. A third model was also built to study the possibility in which the cyclisation of the THF intermediate occurs in the solvent and does not require the presence of the enzyme.

ACS Paragon Plus Environment

6

Page 7 of 57 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

ACS Catalysis

2. Methodology 2.1.

Preparing the structure This study was conducted using the crystallographic structure of SHMT

from Escherichia coli deposited in the Protein Data Bank (PDB) with the reference 1DFO. 17 Although here, it was our aim to study the SHMT present in the Plasmodium falciparum, we have chosen to use the PDB from Escherichia coli because this structure contains the external aldimine with the product of the reaction (PLP + Glycine) and an analog of first THF intermediate. This sort of information is not available in the PDB file of SHMT from P. falciparum (PDB code: 4O6Z)20. Although the identity between both proteins is not very high (43%, Figure S1 A) (BLASTP 2.8.0+)21-22, the structure of both proteins (Figure S1 B) is very similar nearby, and the region of the active site can be almost overlapped (Figure S1 C). The model system that was used in the computational studies includes one of the homodimers of SHMT that is present in the crystallographic structure with the PDB code 1DFO. The external aldimine (PLP+Gly) present on the PDB was used to model the quinonoid intermediate of the reaction, adding only one proton to the Cα carbon of the substrate. The 5-formyl-THF inhibitor was used as a template to build the 5-hydroxymethyl-THF intermediate that results from the transfer of the hydroxymethyl group during the first step of the reaction. From this model, two additional models were built to accommodate different protonation states of Glu57. These models were called Model 1 and Model 2. An additional model was built to study if the formation of the 5,10methylene-THF could occur in a non-enzymatic process. This model (called Model 3) only contains the 5-hydroxymethyl-THF molecule that results from the reaction of the THF cofactor with the formaldehyde molecule (that takes place inside the enzyme). To model the solvent effect an implicit solvent model was employed (water, ε = 78.3553) using the IEFPCM formalism integrated with Gaussian09 23.

ACS Paragon Plus Environment

7

ACS Catalysis 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

2.2.

Page 8 of 57

Structure Minimization The protein, and the external aldimine (L-Ser bounded to the PLP

cofactor), as well as the THF cofactor were solvated by using TIP3P

24

type

water molecules. The protein was placed inside a water box so that each atom of the protein distanced at least 12 Å of the from the faces. The SANDER software from the AMBER12 package was used to add the hydrogen atoms to the model.25 The protonation states for all amino acids were predicted by PROPKA 3.1

26-27

at pH 7.0, excepting the Glu57 residue whose protonation

state was tested in models 1 and 2. The parameterization of the 5-hydroxymethyl-THF and external aldimine was made optimizing their structures using the Hartree–Fock (HF/6-31G(d)) method to obtain the atomic charges and the antechamber software from AMBER12 package

25

to assign the atom types. The frcmod, prepc, and lib files

containing all the parameters and topology for the external aldimine and 5hydroxymethyl-THF molecules are available in supporting information (SI). The geometry of the full system was minimized through four sequential steps using the AMBER12 package

25

and the GAFF

28

and ff99SB

29

force

fields. Firstly, we executed a minimization of all water molecules, and afterwards all of the hydrogen atoms. Subsequently, all atoms from the system, except backbone atoms, were minimized, and finally, a minimization of all system with no constraint was conducted.

2.3.

QM/MM The QM/MM model was built loading the minimized structure from the

previous procedure, using the molUP plugin (VMD) software

31

30

for Visual Molecular Dynamics

. The QM/MM model included the full dimer, the two external

aldimines, the two 5-hydroxymethyl-THF molecules, and a 5.0 Å coat of water molecules. It was followed the general protocol that is widely accepted to study enzymatic mechanisms. 32 The subtractive QM/MM model methodology ONIOM was used in all the calculations.

33-35

This method allows splitting a model system into different

regions that are computed by different theoretical levels. In this work, the system was divided into two regions: the high-level (HL) layer was calculated

ACS Paragon Plus Environment

8

Page 9 of 57 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

ACS Catalysis

with density functional theory (DFT), and the low-level (LL) layer with molecular mechanics (MM).

Figure 3 – (Left) Structure of the SHMT dimer (subunit A: green; subunit B: blue) and licorice representation of the HL layer. (Right) The substrate, cofactors and the active site residues are represented in Licorice and colored according to the subunit that they belong to (subunit A: green; subunit B: blue). Each molecule is identified by a circular tag.

The atoms included in the HL layer were adjusted from step to step since different types of reactions take place during the full catalytic process. In the αelimination reactions, the HL includes the atoms from the external aldimine, THF cofactor, and some active site residues that in total account to 106 atoms (Arg3643, Asp200, His129, Asn102, Glu57, Ser35, and a water molecule). In the study of the conversion of the THF intermediate to 5,10-methylene-THF only the THF intermediate and the active site residue Glu57 were included in the HL layer, amounting to 85 atoms. In the last stage of the mechanism, the similar set of atoms used in the α-elimination was used (121 atoms in total). (Figure 3) The remaining atoms of the system were included in the LL layer. Hydrogen atoms were used as link atoms to complete the valence of the bonds crossing between the two layers of the ONIOM QM/MM scheme. The geometry optimization of the HL layer was carried out using the B3LYP functional (DFT)

36

. The B3LYP functional was selected due to its very

good results in the study of similar biological systems. basis set was employed as available in Gaussian 09

23

37-44

The 6-31G(d,p)

.

We explored the reactional space through linear transit scans along the reaction coordinates implicated in each step of the reaction. Afterward, the

ACS Paragon Plus Environment

9

ACS Catalysis 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 10 of 57

transition states (TS) were optimized, using the structures of higher energy obtained with each linear scans. The reactant and product of each step were determined using internal reaction coordinate (IRC)

45

calculations. The TSs

were confirmed by vibrational frequency calculations, resulting in a single imaginary frequency with the correct transition vectors assigned. The vibrational frequency calculations were also conducted for the minima, and no imaginary frequencies were observed. The ZPE, thermal, and entropic energy corrections were estimated at 310.15 K and 1.0 bar during the frequency calculation of TS and related minima structures. Finally, the energies of all TS, reactant, and product were determined using DLPNO-CCSD(T)/CBS, for the HL layer. The recent domain-based local pair natural orbital coupled cluster method with single, double, and perturbative triple excitations (DLPNO-CCSD(T))

46-47

was employed since it allows

calculating energies that closely fit those obtained with the canonical method (CCSD(T)) at a much lower computational cost

48-49

. Moreover, several studies

have already shown the application of this new method in several biochemistry areas

50-52

. For each minima and TS of each step, the HL layer was isolated to

perform DLPNO-CCSD(T) single-point energy calculations (SP) using the ORCA software (version 4.0.1.2) pVDZ

54

and cc-pVTZ

54

53

. The SPs were carried out using the cc-

basis set, and the cc-pVDZ/C

55-56

and cc-pVTZ/C

56

correlation fitting basis sets, respectively. The combination of the energies obtained with cc-pVDZ | cc-pVDZ/C and cc-pVTZ | cc-pVTZ/C basis sets for the HL layer were used to extrapolate to the complete basis set (CBS) limit according to the Truhlar’s extrapolation approach

57

, excepting for the α and β

exponents. The values 4.42 and 2.46 were considered as α and β exponents in the extrapolation according to the ORCA documentation. Concerning the interaction between the HL and LL layers, the mechanical embedding method was used to perform all the calculations because the electrostatic embedding scheme is computational very expensive considering the size of the system and the number of reactions studied in this work. The activation and reaction free energies presented here were determined by the difference between the Gibbs free energies of TS and reactant, or product and reactant, respectively.

ACS Paragon Plus Environment

10

Page 11 of 57 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

ACS Catalysis

All the preparation of the Gaussian09 results analysis were made using molUP plugin

23 30

input files and subsequent

for VMD software 31.

The structures (PDB files) for the reactant, TS, and product of each step studied in this work are available in SI.

ACS Paragon Plus Environment

11

ACS Catalysis 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 57

3. Results and Discussion The computational calculations regarding the catalytic mechanism of SHMT were divided into three main stages: i) the first stage involved the αelimination process; ii) the second stage, the conversion of the THF intermediate to 5,10-methylene-THF; and iii) the third stage, the protonation of the quinonoid intermediate (QI) and subsequent enzymatic turnover. Each of these stages is described and discussed in detail in the following sections of the manuscript. Each of these stages is described and discussed in detail in the following sections of the manuscript. The analysis of all key interactions for the minima and TS of each step is provided in SI (Figure S3).

3.1. First Stage: the α-Elimination The α-elimination of the hydroxymethyl group of the L-Ser is the main point of discussion considering the catalytic mechanism of the SHMT. In the literature, there are two proposals for this reaction. In the first one, it is proposed that Glu57 is protonated and involves the nucleophilic attack of nitrogen N1 from THF to carbon Cβ of the substrate and from which results the elimination of the hydroxymethyl group 15-17. (Figure 4 Top) The other hypothesis proposes that the Glu57 residue is not protonated and at the beginning of the reaction it abstracts a proton from the hydroxyl group of the substrate. From this reaction results the formation of formaldehyde that afterwards reacts with the nitrogen N1 from THF and becomes covalently bonded to it. (Figure 4 Bottom)

ACS Paragon Plus Environment

12

Page 13 of 57 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

ACS Catalysis

Figure 4 – Schematic representation of the two mechanistic proposals for the α-elimination of Lserine that is bonded to the PLP cofactor and forms the external aldimine, considering different protonation states of Glu57 (Top: protonated and Bottom: unprotonated). The spheres represent the phosphate group of the PLP cofactor.

Unprotonated Glu57 When Glu57 is not protonated, it interacts with the hydroxyl group of the external aldimine by a hydrogen bond, and this interaction endorses a good alignment of the latter group with the THF cofactor. This interaction will be very important during the two steps that are required for the α-elimination process. The first step is characterized by a proton transfer between the hydroxyl group of the external aldimine and the negatively charged Glu57. The computational calculations also show that this reaction triggers, simultaneously, the cleavage of the bond between carbons Cα and Cβ (the α-elimination reaction) and, the formation of a double bond between oxygen Oγ and carbon Cβ (R: 1.38 Å to P: 1.23 Å). At the end of this reaction, one molecule of formaldehyde is obtained, Glu57 becomes protonated, and the PLP cofactor adopts a quinonoid configuration that is negatively charged (Charge of PLP’s ring: R: +0.18 a.u.; P: -0.04 a.u.) (Figure 5 – Step 1). This reaction requires an activation Gibbs energy of 18.0 kcal/mol, and it is endergonic in 8.1 kcal/mol. The TS structure of this reaction is characterized by one imaginary frequency at 229.1i cm-1 that is associated with the vibration of the atoms involved in the reaction (Figure 5 – Step 1 Center). In the TS structure, the bond

ACS Paragon Plus Environment

13

ACS Catalysis 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 14 of 57

that is cleaved is perpendicular to the PLP cofactor ring (Cβ-Cα-C7-C2 angle: 98.6°), a condition that is important in the chemistry of the PLP cofactor to have the reaction kinetically favored. 58-59 The second step involves the nucleophilic attack of nitrogen N1 of THF to carbon Cβ of the recently formed formaldehyde, and the simultaneous proton transfer from Glu57 to oxygen Oγ of formaldehyde. At the end of this reaction, a hydroxymethyl group is obtained that becomes covalently bonded to the THF cofactor, which becomes positively charged (R: +1.74; P: -1.17). In this step, the quinonoid intermediate that resulted from the first step does not participate in the reaction. (Figure 5 – Step 2) The TS of this reaction is characterized by one imaginary frequency at 206.8i cm-1, in which carbon Cβ of formaldehyde adopts an sp3 hybridization and the bond between carbon Cβ and oxygen Oγ becomes elongated, preparing it for the nucleophilic attack endorsed by the THF and the proton transfer from Glu57, respectively. This reaction requires a very low Gibbs activation energy (2.0 kcal/mol), and it is thermoneutral.

Figure 5 - Structures of reactant (R), transition state (TS), and product (P) of the two sequential steps associated with the α-elimination reaction catalyzed by SHMT. Wedge-Dash representation of the structures of reactant and product. Three-dimensional representation of the TS structure. The yellow

ACS Paragon Plus Environment

14

Page 15 of 57 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

ACS Catalysis

arrows correspond to the major vibrational frequency associated with the imaginary frequency of the TS. The electrostatic potential was calculated and mapped on the surface of the molecules involved in the reaction.

Protonated Glu57 When Glu57 is protonated, the α-elimination takes place in a single step. However, the computational results show that the nucleophilic attack of nitrogen N1 from THF to carbon Cβ of the external aldimine is impossible under biological conditions since the activation barrier is too high (∆E‡ = 53.9 kcal/mol) to occur at reasonable conditions of temperature and pressure. In addition, the reaction is very endothermic (∆ER = 46.7 kcal/mol) which results from the lack of stabilization endorsed by Glu57, contrasting to what was observed with the previous model. These results indicate therefore that Glu57 cannot be protonated during the α-elimination of the external aldimine by the SHMT. (Figure 6)

Figure 6 – Energy profile of the two evaluated mechanisms for the α-elimination process catalyzed by SHMT with Glu57 protonated (∆E) (gray) or unprotonated (∆G) (black).

ACS Paragon Plus Environment

15

ACS Catalysis 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 16 of 57

3.2. Second Stage: Conversion of THF intermediate into 5,10methylene-THF After the α-elimination stage, the THF is covalently bonded to a hydroxymethyl group and becomes positively charged. The next step involves the cyclization of this reaction intermediate (THF-I) to yield one of the final products of the reaction: the 5,10-methylene-THF. In the literature, there are two hypotheses for this reaction. Some authors propose that this reaction does not require that participation of the enzyme, and therefore, it could take place in the cytoplasm of the cell (water environment). Other authors suggest that the reaction needs to continue inside of the enzyme since the cyclization process is a rather complex task and the enzyme will be very important to catalyze this process.

15, 19

Each of these two hypotheses is

discussed below.

3.2.1. Enzymatic Catalysis Under enzymatic catalysis, the formation of 5,10-methylene-THF requires five steps that include: i) three proton transfers, ii) the formation and release of one water molecule, and iii) the cyclization process. In most of these reactions, Glu57 plays an active role, and it is directly involved in the catalysis. The quinonoid intermediate (PLP cofactor) is neither involved in any of these reactions nor does it affect the cyclization process. However, it was kept in all of the models used to study these reactions, because it stays in the quinonoid intermediate stage waiting for the final reaction where it is protonated.

Proton transfer reactions The first two steps consist of three proton transfers. The first step involves the intramolecular proton transfer from nitrogen N1 to nitrogen N4, resulting in the migration of the positive charge that was concentrated at nitrogen N1, in the reactant, to nitrogen N4. This reaction has a very small Gibbs activation energy (∆G‡ = 0.7 kcal/mol), and it is exergonic in 8.5 kcal/mol. The low activation energy results from the configuration of the TS structure (1230.6i cm-1) in which both nitrogen atoms are very close to each other favoring the proton transfer (N1-N4: 2.69 Å). The proximity between both

ACS Paragon Plus Environment

16

Page 17 of 57 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

ACS Catalysis

atoms is endorsed by Gly57 that establishes two hydrogen bonds with the THF intermediate: one with the hydroxyl group (1.47 Å) and another with the proton that is bonded to nitrogen N4 (2.57 Å). The stabilization of the product of the reaction relies on the better delocalization of the positive charge (mesomeric effect) that is provided by the benzyl group that is attached to nitrogen N4. (Figure 7 – Step 3)

The second step involves two coupled proton transfers: one between nitrogen N4 to oxygen Oγ and another between oxygen Oγ and Glu57. The full process is almost spontaneous (∆G‡ = 1.3 kcal/mol), and the reaction is very exergonic in 8.3 kcal/mol (Figure 7 – Step 4). The low activation energy of this reaction is only possible due to the proximity of the atoms involved in the chemical reaction at the TS structure (639.6i cm-1). The higher stabilization of the product (THF-III) of this reaction is due to the absence of point charges in the active site, which contrasts with the reactants where nitrogen N4 of the THF intermediate (THF-II) was positively charged, and Glu57 was negatively charged. This effect can be seen in the electrostatic maps represented in Figure 7. The key role played by Glu57 in these two steps reinforces, once again, its importance in the catalytic process as well as the importance of its protonation state at the beginning of the catalytic process.

ACS Paragon Plus Environment

17

ACS Catalysis 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 18 of 57

Figure 7 - Structures of reactant (R), transition state (TS), and product (P) of the two sequential steps associated with the three proton transfers catalyzed by SHMT. Wedge-Dash representation of the structures of reactant and product. Three-dimensional representation of the TS structure. The yellow arrows correspond to the major vibrational frequency associated with the imaginary frequency of the TS. The electrostatic potential was calculated and mapped on the surface of the molecules involved in the reaction.

Dehydration reaction The next step of the cyclization process involves the formation of a carbocation through the elimination of a water molecule. Our initial attempts to study this reaction involved the direct protonation of the hydroxyl group by Glu57 that, led to the formation of a water molecule, which should dissociate subsequently (Figure 7 – Step 4). Although this step is rather straightforward, the computational results indicate that this reaction is only possible when the hydroxyl group does not interact with nitrogen N4 by a hydrogen bond. Only in this conformation, it was possible to characterize a TS structure for the dehydration process. This new conformation is possible through the rotation of the N1-Cβ bond from the product of the previous step by -125.0º and requires an energy cost of 9.6 kcal/mol (Figure 8).

ACS Paragon Plus Environment

18

Page 19 of 57 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

ACS Catalysis

Figure 8 – Energy profile for the dihedral rotation across the N1-Cβ bond. Newman representation was used to show the relative configuration of the Cβ and N1 atoms. The energies were computed using the ONIOM(DLPNO-CCSD(T)/CBS:ff99SB) scheme. The full energy profile obtained from the scan for rotational increments of 5.0° can be found in SI (Figure S2).

In the new conformation of the THF intermediate (THF-III), the elimination of the water molecule requires a Gibbs activation energy of 5.4 kcal/mol, and the reaction is slightly endergonic in 3.0 kcal/mol. The small activation energy of this reaction results from the good alignment between Glu57 and the hydroxyl group of the THF intermediate in the TS structure (113.6i cm-1) (Figure 9). Although, in this new conformation, Glu57 does not interact directly with nitrogen N4, it establishes a hydrogen bond with a water molecule (1.83 Å) that interacts with nitrogen N4 by another hydrogen bond (2.00 Å). At the end of this step, the release of the water molecule restores the negative charge of Glu57 and generates a positive charge at nitrogen N1 that becomes bonded to carbon Cβ by a double bond (N1-Cβ R: 1.43 Å to P: 1.32 Å). The formation of the new point charges in the THF intermediate (THF-IV) are very important since they polarize the Cβ-N1 bond that will be crucial to increase the electrophilic character of the Cβ atom, which is required for the cyclisation of the THF intermediate (THF-IV) in the next step.

ACS Paragon Plus Environment

19

ACS Catalysis 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 20 of 57

Figure 9 - Structures of reactant (R), transition state (TS), and product (P) of the dehydration of the THF intermediate catalyzed by the SHMT. Wedge-Dash representation of the structures of reactant and product. Three-dimensional representation of the TS structure. The yellow arrows correspond to the major vibrational frequency associated with the imaginary frequency of the TS. The electrostatic potential was calculated and mapped at the surface of the molecules involved in the reaction.

Cyclization The cyclization of the THF intermediate (THF-IV) to yield the 5,10methylene-THF molecule requires two sequential steps: i) an intramolecular nucleophilic attack and ii) a deprotonation. The first step involves the nucleophilic attack of nitrogen N4 to carbon Cβ of the THF intermediate (THF-IV), from which results a 5-membered ring. This reaction is favored by the correct orientation of carbon Cβ in relation to nitrogen N4 that is endorsed by the hydrogen bond that Glu57 establishes with one water molecule present in the active site and the latter one with nitrogen N4. For this reason, the TS structure (205.0i cm-1) is very close to the reactant of this reaction, from which results a in a spontaneous reaction (Figure 10 – Step 6). The positive charge that was lodged at nitrogen N1 of the THF intermediate (THF-IV) migrates at the end of the reaction to nitrogen N4, where it is better delocalized due to the presence of the neighbor benzyl group that is bonded to the nitrogen N4. The Gibbs reaction energy of this reaction reflects this stabilization being exergonic in 16.9 kcal/mol. The next step of the cyclization process involves the proton transfer from nitrogen N4 of the THF intermediate (THF-V) to Glu57. This reaction proceeds with the participation of a water molecule that catalyzes the proton transfer between both groups. The TS of this reaction is characterized by one imaginary frequency at 670.7i cm-1. At the end of the reaction, Glu57 becomes protonated

ACS Paragon Plus Environment

20

Page 21 of 57 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

ACS Catalysis

and one of the final products of the enzymatic catalysis, 5,10-methylene-THF, is obtained. This step is almost spontaneous requiring a Gibbs activation energy of 1.1 kcal/mol, and the reaction is slightly endergonic in 0.5 kcal/mol (Figure 10 – Step 7). Overall, it can be concluded that the cyclization process requires a Gibbs activation energy of 1.1 kcal/mol and it is exergonic in 16.4 kcal/mol.

Figure 10 - Structures of reactant (R), transition state (TS), and product (P) of the steps involved in cyclization of the THF intermediate catalyzed by the SHMT. Wedge-Dash representation of the structures of reactant and product. Three-dimensional representation of the TS structure. The yellow arrows corresponds to the major vibrational frequency associated with the imaginary frequency of the TS. The electrostatic potential was calculated and mapped on the surface of the molecules involved in the reaction.

ACS Paragon Plus Environment

21

ACS Catalysis 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 22 of 57

3.2.2. Non-enzymatic catalysis In the last section, it was described the synthesis of the 5,10-methyleneTHF molecule under enzymatic catalysis. In this section, it is described the energy profile of the same reaction that resembles the cytoplasm, and it was modelled as a water environment. Under enzymatic catalysis, the full process requires five sequential steps, in which Glu57 plays an active role involving a conformational rearrangement (the rotation of the N1-Cβ bond) of the THF intermediate (THF-I(S)). In the cytoplasm, the cyclization process requires only three steps and does not need any conformational rearrangement of THF. The first step of the reaction in the cytoplasm is very similar to the one that takes place during the enzyme catalysis (Figure 11 – Step 1 and Figure 7– Step 3). It involves the proton transfer from nitrogen N1 to nitrogen N4 of the THF reaction intermediate (THF-I(S) versus THF-I). In the cytoplasm, the reaction is marginally slower requiring a Gibbs activation energy of 1.4 kcal/mol (versus 0.7 kcal/mol in the enzymatic catalysis). The most significant difference is observed in the reaction free energy. In the cytoplasm, this step is endergonic in 0.7 kcal/mol whereas under enzymatic catalysis it is exergonic in 8.5 kcal/mol. This difference comes from the stabilization of the product of the reaction that is provided by the enzyme, and mainly by Glu57, which in the cytoplasm is absent. The next step involves the dehydration process. This is where the differences between both mechanisms are observed. In the cytoplasm, the abstraction of the proton bonded to nitrogen N4 by oxygen Oγ occurs in a single step without requiring any conformational rearrangement of the reaction intermediate. At the end of this step, one water molecule is formed, and the same reactional product is obtained. Despite the simplicity of the reaction, the energetic cost for this reaction in the cytoplasm is much higher than under enzymatic catalysis. In the cytoplasm, this step requires a Gibbs activation energy of 10.2 kcal/mol (versus 5.4 kcal/mol), which means that the dehydration process is approximately 3,000 times slower than when it happens with enzymatic catalysis. As expected, the Gibbs reaction free energy is almost the same for the dehydration whether the reaction occurs under enzymatic catalysis

ACS Paragon Plus Environment

22

Page 23 of 57 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

ACS Catalysis

(∆GR = +3.0 kcal/mol) or in the cytoplasm (∆GR = +1.9 kcal/mol) (Figure 11 – Step 2; Figure 9 – Step 5). The last step required for the cyclization of the THF intermediate (THFIII(S) versus THF-IV) involves the nucleophilic attack of nitrogen N4 to carbon Cβ. This reaction is spontaneous in both the cytoplasm and the enzyme (∆G‡ = -3.5 kcal/mol versus ∆G‡ = -1.4 kcal/mol), which means that both processes are energetically similar (Figure 11– Step 3; Figure 10 – Step 6). However, it requires only one step instead of two. This energy difference is explained by the higher freedom of the structure in the cytoplasm that allows an easier cyclization of the 5-membered ring of THF without the geometric constraints imposed by the active site of the enzyme. However, once more, the enzyme is able to provide further stabilization of the reactional product (∆GR = -16.9 kcal/mol versus ∆GR = -6.8 kcal/mol), contributing to a more favorable reaction. All the TS structure of these reactions were checked using the vibrational frequency analysis and are characterized by a single imaginary frequency for each TS. Those frequencies correspond to the vibration of the atoms directly involved in each step and can be found in Figure 11.

ACS Paragon Plus Environment

23

ACS Catalysis 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 24 of 57

Figure 11 – Structures of reactant (R), transition state (TS), and product (P) of the steps involved in cyclization of the THF intermediate in the cytoplasm. Wedge-Dash representation of the structures of reactant and product. Three-dimensional representation of the TS structure. The yellow arrows correspond to the major vibrational frequency associated with the imaginary frequency of the TS.

Looking at the overall results for the cyclization of the THF intermediate in the cytoplasm, it can be concluded that although the reaction in the water requires fewer steps (three instead of four steps), the reaction is from the kinetic and thermodynamic point of view favored with enzymatic catalysis. Under ACS Paragon Plus Environment

24

Page 25 of 57 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

ACS Catalysis

enzymatic catalysis, the rate-limiting step of the cyclization process is the dihedral rotation across the N1-Cβ bond and requires 9.6 kcal/mol, whereas, in the cytoplasm, the rate-limiting step is the elimination of the water molecule with an activation barrier of 10.2 kcal/mol. (Figure 12). When the reaction takes place in water, the process is almost thermoneutral (-4.2 kcal/mol), whereas with enzymatic catalysis of the same process originates a product that is highly stabilized in 28.5 kcal/mol. (Figure 12) In sum, these results show that the production of the 5,10-methyleneTHF can occur in the cytoplasm and without enzymatic catalysis, but it would be much slower. However, taking into account that this process requires the migration of the THF intermediate to the cytoplasm, which implies an additional energetic cost (dissociation and desolvation effects), the full process will be even slower. For this reason, the cyclization of THF intermediate (THF-I) has to occur through an enzymatic process catalyzed by the SHMT.

Figure 12 - Energy profile of the two proposals for the conversion of the THF intermediate into 5,10methylene-THF: (black) the process is catalyzed by the SHMT enzyme or (gray) the reactions occur in the cytoplasm. The values in brackets correspond to the energy of each single step.

3.3. Protonation of the Quinonoid Intermediate Once the 5,10-methylene-THF is formed, the active site has Glu57 protonated and a quinonoid intermediate (Figure 10 – Step 7 that needs to be

ACS Paragon Plus Environment

25

ACS Catalysis 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 26 of 57

converted into an internal aldimine in order to enhance the enzymatic turnover and the release of one molecule of Gly. It is worth mentioning that nothing is known if the 5,10-methylene-THF is released from the SHMT active site to the cytoplasm before or after the protonation of the quinonoid intermediate. For this reason, we chose to maintain the 5,10-methylene-THF inside the active site during this step. We observed that it is neither involved in any chemical step nor is it involved in the catalytic process. This step involves a proton transfer from Glu57 to carbon Cα of the QI, through a water molecule that catalyzes the reaction. This water molecule is the one that was generated at step 5 (Figure 9), which plays, in this final step, an important role for the continuity of the catalytic process. This reaction has a Gibbs activation energy of 9.0 kcal/mol, and it is exergonic in 6.1 kcal/mol. The TS was validated through the vibrational frequencies analysis showing a single imaginary frequency assigned to the atoms that are intrinsically related to the reaction (775.0i cm-1). This reaction is favored because the QI is negatively charged and therefore it meets good conditions to act as a nucleophile and becomes protonated (Figure 13). At the end of this reaction, Glu57 becomes negatively charged, restoring its initial protonation state, and an external aldimine with a Gly bonded to the PLP cofactor is obtained. After the transimination reaction - a process that is common to all PLP-dependent enzymes - Gly is released from the active site, and the PLP cofactor becomes bonded to the active site Lys229, forming an internal aldimine, which makes the enzyme ready for a new catalytic cycle, once the substrate becomes available.

ACS Paragon Plus Environment

26

Page 27 of 57 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

ACS Catalysis

Figure 13 - Structures of reactant (R), transition state (TS), and product (P) of the protonation of the quinonoid intermediate (QI) catalyzed by the SHMT. Wedge-Dash representation of the structures of reactant and product. Three-dimensional representation of the TS structure. The yellow arrows correspond to the major vibrational frequencies associated with the imaginary frequency of the TS. The electrostatic potential was calculated and mapped on the surface of the molecules involved in the reaction.

ACS Paragon Plus Environment

27

ACS Catalysis 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 28 of 57

Conclusion In this work, the catalytic mechanism of the SHMT was studied by computational means employing QM/MM methodologies. This enzyme catalyzes the reversible conversion of L-serine and THF into glycine and 5,10methylene-THF, with the help of the PLP cofactor. The reaction requires eight sequential steps that can be divided into three main stages: the α-elimination, the cyclization of the 5-hydroxymethyl-THF intermediate into 5,10-methyleneTHF, and the protonation of the quinonoid intermediate (QI). The full mechanism is represented in Figure 14. In the first step, the hydroxymethyl group of the substrate is eliminated in the form of a formaldehyde molecule that undergoes, in the second step, a nucleophilic attack by the nitrogen N1 of the THF cofactor. The resulting THF intermediate (THF-I) undergoes an intramolecular reaction, that involves a proton transfer from the nitrogen N1 to nitrogen N4 (step 3) and finally to Glu57 (step 4). Afterwards, a conformational rearrangement of the THF intermediate (THF-III) takes place, required for the elimination of a water molecule that enhances the electrophilic nature of carbon Cβ, which in turn is required for the next step (step 5). Subsequently, carbon Cβ becomes suitable to undergo a nucleophilic attack by the nitrogen N4, producing a 5-membered ring THF intermediate (step 6). Immediately after, Glu57 abstracts the remaining proton from the nitrogen N4 and originates the final product 5,10-methylene-THF (step 7). Finally, the water molecule that was released in step 5 assists, in the final step, the proton transfer from the Glu57 to the carbon Cα of the QI producing an external aldimine with the glycine bonded to the PLP cofactor (step 8). At the end of this reaction, the release of 5,10-methylene-THF and L-Gly (after the transimination reaction) from the active site takes place, and the enzyme is ready for a new enzymatic cycle.

ACS Paragon Plus Environment

28

Page 29 of 57 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

ACS Catalysis

Figure 14 – Overview of the catalytic mechanism of the PLP-dependent enzyme SHMT obtained in this work.

According to the calculated energy profile (Figure 15), the rate-limiting step of the entire mechanism is where the elimination of the hydroxymethyl group occurs, forming a formaldehyde intermediate that becomes immediately bonded to the THF cofactor. The intramolecular passage of the proton between nitrogen N1 and nitrogen N4 of the THF-I intermediate also contributes to assign this step as the rate-limiting one. These two reactions require an activation energy boost of 18.0 kcal/mol. These results closely fit the experimental kinetic data that predicted an activation barrier of 15.7-16.2 kcal/mol 13-14, 60.

ACS Paragon Plus Environment

29

ACS Catalysis 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 30 of 57

It is worth mentioning that SHMT is also capable of catalyzing the reverse reaction, in other words, the conversion of glycine and 5,10-methylene-THF into L-serine and THF. Analyzing the energy profile in the reverse order (Figure 15), all the steps are feasible under biological conditions, agreeing with the experimental evidence of reversible catalysis. However, there is no kinetic result about the kinetics of the reverse reaction catalyzed by the SHMT from E.coli. There is only a kinetic study using Hydrogenobacter thermophilus from which it was concluded that the reaction is almost 4 times slower in the reverse direction 61

. Based on this evidence, it might be expected a slower conversion of glycine

into L-serine. This also agrees very well with the mechanism proposed in this work, since the rate-limiting step, for the reverse reaction, would be 29.7 kcal/mol, which compared with the rate-limiting step of the direct catalytic would lead to a slower reaction.

Figure 15 – Complete energy profile (DLPNO-CCSD(T)/CBS::ff99SB) for the catalysis of L-serine and THF into glycine and 5,10-methylene-THF by the SHMT.

The results obtained with this work also provide new insights into the catalytic process and on the α-elimination and the production of the 5,10methylene-THF molecule. Regarding the two proposals that are available in the literature for the αelimination reaction, the results presented in this paper indicate that the reaction proceeds with the formation of the formaldehyde molecule as an intermediate, and only then, it becomes covalently bonded to nitrogen N1 of the THF cofactor. In this process, Glu57 cannot be protonated since it plays a key role in holding the proton of the hydroxymethyl group. (Figure 14). Although

ACS Paragon Plus Environment

30

Page 31 of 57 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

ACS Catalysis

structural studies showed a close proximity between Glu57 and the hydroxyl group of the substrate, from which was suggested that the Glu57 was protonated

16-17

, , the calculations showed that this configuration of the active

site does not allow the α-elimination. Such reaction was only possible to model when we considered the Glu57 in an anionic form at the beginning of the reaction. When Glu57 is anionic, the hydroxyl group from the substrate establishes a hydrogen bond with the Glu57, according to the experimental evidence that shows a close proximity between the Glu57 residue and the Oγ atom of the substrate. Glu57 was found to play a key role in several reactions of the mechanism, namely steps 1, 2, 4, 5, 7, and 8. This is in line with the mutagenic studies that show that any mutations to Glu57 caused a drastic inhibition of the enzyme. 18

16,

This is the first time where the role played by this residue is explained.

Furthermore, Glu57 is a key residue to handle most of the proton transfers that are required for the reaction to occur, holding a negative charge during some steps (step 3 and 6) to provide conditions for intermediate reactions to happen. Then, Glu57 becomes protonated during the remaining steps of the mechanism. In sum, Glu57 works as an proton pump that handles the charge transfers across the mechanism and establishes a link between the external aldimine and the THF cofactor. (Figure 16)

Figure 16 – Schematic representation of the behavior of the main groups and charges across the entire catalytic mechanism.

ACS Paragon Plus Environment

31

ACS Catalysis 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 32 of 57

In the literature, there are also two different proposals suggesting that the conversion of the THF intermediate to 5,10-methylene-THF can alternatively occur outside the enzyme through a non-enzymatic mechanism (in the solvent) 15, 19

. In this work, both proposals were studied, and the results showed that the

reaction is thermodynamically and kinetically favored under enzymatic catalysis. The results presented in the manuscript are also in line with the experimental evidence

15

observed, confirming that the conversion of the THF into 5,10-

methylene-THF is very fast and almost spontaneous. Overall, this stage of the mechanism (step 3 to 7) is very exergonic and with low free activation energies agreeing with the experimental observation of an almost spontaneous process. Additionally, the steps involved in the cyclization of the THF are important to prepare the enzyme for the final step where a proton is transferred to the QI to produce the glycine. This evidence agrees with the fact that the THF cyclization must occur inside the active site of the SHMT. Moreover, our results agree with the proposal of Dunathan

59, 62

, which

postulates that the bound complex that is formed or cleaved during the reaction involving the PLP cofactor must be placed perpendicularly to the conjugated system of the PLP. The hydroxymethyl group is released almost perpendicularly to the PLP cofactor (Cβ-Cα-C7-C2 angle: 98.6°), and the protonation of the QI occurs in the same way (H-Cα-C7-C2 angle: 115.7°). Together with the previous stages that are common to all PLP-dependent enzymes, this work provides for the first time and with an atomistic level of detail, a portrait of the entire mechanism catalyzed by the SHMT. We believe that this knowledge provides valuable information for the rational molecular design of new drugs against malaria.

ACS Paragon Plus Environment

32

Page 33 of 57 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

ACS Catalysis

Supporting Information This information is available free of charge on the ACS Publications website. Alignment of the protein sequences of the E.coli and P. falciparum forms of the SHMT enzyme and tridimensional image of the overlapping of the structures of all dimer or active site residues (Figure S1); Energy profile of the dihedral rotation across the N1-Cβ bond of the THF intermediate (Figure S2); Analysis of all key interactions for each step of the catalytic mechanism (Figure S3) (PDF). Molecular Mechanics parameters for the external aldimine (PLP + glycine) and 5-hydroxymethyl-THF (ZIP). PDB coordinate files for all stationary states of the entire catalytic mechanism of SHMT (ZIP).

Author Information Present Address: §

UCIBIO@REQUIMTE, BioSIM, Departamento de Biome- dicina, Faculdade

de Medicina da Universidade do Porto, Alameda Professor Hernâni Monteiro, 4200-319, Porto, Portugal

Acknowledgments European Union (FEDER funds POCI/01/0145/FEDER/007728) and National Funds (FCT/MEC, Fundação para a Ciência e Tecnologia and Ministério da Educação

e

Ciência)

under

the

Partnership

UID/MULTI/04378/2013.UID/MULTI/04378/2013;

Agreement

PT2020

NORTE-01-0145-FEDER-

000024, supported by Norte Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF); Fundação para a Ciência e a Tecnologia (FCT) through projects IF/01310/2013. Henrique Fernandes is grateful to the Fundação para a Ciência e Tecnologia (FCT), Portugal for a PhD Grant (SFRH/BD/115396/2016).

ACS Paragon Plus Environment

33

ACS Catalysis 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 34 of 57

References 1. Haldar, K.; Bhattacharjee, S.; Safeukui, I., Drug resistance in Plasmodium. Nat. Rev. Microbiol. 2018, 16, 156. 2. Organization, W. H., World malaria report 2017. Geneva, 2017. 3. Baragaña, B.; Hallyburton, I.; Lee, M. C. S.; Norcross, N. R.; Grimaldi, R.; Otto, T. D.; Proto, W. R.; Blagborough, A. M.; Meister, S.; Wirjanata, G.; Ruecker, A.; Upton, L. M.; Abraham, T. S.; Almeida, M. J.; Pradhan, A.; Porzelle, A.; Martínez, M. S.; Bolscher, J. M.; Woodland, A.; Luksch, T.; Norval, S.; Zuccotto, F.; Thomas, J.; Simeons, F.; Stojanovski, L.; Osuna-Cabello, M.; Brock, P. M.; Churcher, T. S.; Sala, K. A.; Zakutansky, S. E.; Jiménez-Díaz, M. B.; Sanz, L. M.; Riley, J.; Basak, R.; Campbell, M.; Avery, V. M.; Sauerwein, R. W.; Dechering, K. J.; Noviyanti, R.; Campo, B.; Frearson, J. A.; AnguloBarturen, I.; Ferrer-Bazaga, S.; Gamo, F. J.; Wyatt, P. G.; Leroy, D.; Siegl, P.; Delves, M. J.; Kyle, D. E.; Wittlin, S.; Marfurt, J.; Price, R. N.; Sinden, R. E.; Winzeler, E. A.; Charman, S. A.; Bebrevska, L.; Gray, D. W.; Campbell, S.; Fairlamb, A. H.; Willis, P. A.; Rayner, J. C.; Fidock, D. A.; Read, K. D.; Gilbert, I. H., A novel multiple-stage antimalarial agent that inhibits protein synthesis. Nature 2015, 522, 315. 4. McCarthy, J. S.; Lotharius, J.; Rückle, T.; Chalon, S.; Phillips, M. A.; Elliott, S.; Sekuloski, S.; Griffin, P.; Ng, C. L.; Fidock, D. A.; Marquart, L.; Williams, N. S.; Gobeau, N.; Bebrevska, L.; Rosario, M.; Marsh, K.; Möhrle, J. J., Safety, tolerability, pharmacokinetics, and activity of the novel long-acting antimalarial DSM265: a two-part first-in-human phase 1a/1b randomised study. Lancet Infect. Dis. 2017, 17, 626-635. 5. Kronenberger, T.; Lindner, J.; Meissner, K. A.; Zimbres, F. M.; Coronado, M. A.; Sauer, F. M.; Schettert, I.; Wrenger, C., Vitamin B6-dependent enzymes in the human malaria parasite Plasmodium falciparum: a druggable target? BioMed Res. Int. 2014, 2014, 108516. 6. Schwertz, G.; Witschel, M. C.; Rottmann, M.; Bonnert, R.; Leartsakulpanich, U.; Chitnumsub, P.; Jaruwat, A.; Ittarat, W.; Schafer, A.; Aponte, R. A.; Charman, S. A.; White, K. L.; Kundu, A.; Sadhukhan, S.; Lloyd, M.; Freiberg, G. M.; Srikumaran, M.; Siggel, M.; Zwyssig, A.; Chaiyen, P.; Diederich, F., Antimalarial Inhibitors Targeting Serine Hydroxymethyltransferase (SHMT) with in Vivo Efficacy and Analysis of their Binding Mode Based on Xray Cocrystal Structures. J. Med. Chem. 2017, 60, 4840-4860. 7. Schwertz, G.; Frei, M. S.; Witschel, M. C.; Rottmann, M.; Leartsakulpanich, U.; Chitnumsub, P.; Jaruwat, A.; Ittarat, W.; Schafer, A.; Aponte, R. A.; Trapp, N.; Mark, K.; Chaiyen, P.; Diederich, F., Conformational Aspects in the Design of Inhibitors for Serine Hydroxymethyltransferase (SHMT): Biphenyl, Aryl Sulfonamide, and Aryl Sulfone Motifs. Chemistry (Easton) 2017, 23, 14345-14357.

ACS Paragon Plus Environment

34

Page 35 of 57 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

ACS Catalysis

8. Pornthanakasem, W.; Kongkasuriyachai, D.; Uthaipibull, C.; Yuthavong, Y.; Leartsakulpanich, U., Plasmodium serine hydroxymethyltransferase: indispensability and display of distinct localization. Malar. J. 2012, 11, 387. 9. Newman, A. C.; Maddocks, O. D. K., One-carbon metabolism in cancer. Br. J. Cancer 2017, 116, 1499. 10. Franca, T. C.; Pascutti, P. G.; Ramalho, T. C.; Figueroa-Villar, J. D., A three-dimensional structure of Plasmodium falciparum serine hydroxymethyltransferase in complex with glycine and 5-formyl-tetrahydrofolate. Homology modeling and molecular dynamics. Biophys. Chem. 2005, 115, 1-10. 11. Cerqueira, N. M.; Fernandes, P. A.; Ramos, M. J., Computational Mechanistic Studies Addressed to the Transimination Reaction Present in All Pyridoxal 5'-Phosphate-Requiring Enzymes. J. Chem. Theory. Comput. 2011, 7, 1356-68. 12. Oliveira, E. F.; Cerqueira, N. M.; Fernandes, P. A.; Ramos, M. J., Mechanism of formation of the internal aldimine in pyridoxal 5'-phosphatedependent enzymes. J. Am. Chem. Soc. 2011, 133, 15496-505. 13. Florio, R.; Chiaraluce, R.; Consalvi, V.; Paiardini, A.; Catacchio, B.; Bossa, F.; Contestabile, R., The role of evolutionarily conserved hydrophobic contacts in the quaternary structure stability of Escherichia coli serine hydroxymethyltransferase. FEBS J. 2009, 276, 132-43. 14. Angelaccio, S.; di Salvo, M. L.; Parroni, A.; Di Bello, A.; Contestabile, R.; Pascarella, S., Structural stability of cold-adapted serine hydroxymethyltransferase, a tool for β-hydroxy-α-amino acid biosynthesis. J. Mol. Catal. B: Enzym. 2014, 110, 171-177. 15. Schirch, V.; Szebenyi, D. M., Serine hydroxymethyltransferase revisited. Curr. Opin. Chem. Biol. 2005, 9, 482-7. 16. Szebenyi, D. M.; Musayev, F. N.; di Salvo, M. L.; Safo, M. K.; Schirch, V., Serine hydroxymethyltransferase: role of glu75 and evidence that serine is cleaved by a retroaldol mechanism. Biochemistry 2004, 43, 6865-76. 17. Scarsdale, J. N.; Radaev, S.; Kazanina, G.; Schirch, V.; Wright, H. T., Crystal structure at 2.4 Å resolution of E. coli serine hydroxymethyltransferase in complex with glycine substrate and 5-formyl tetrahydrofolate11Edited by I. A. Wilson. J. Mol. Biol. 2000, 296, 155-168. 18. Rao, J. V.; Prakash, V.; Rao, N. A.; Savithri, H. S., The role of Glu74 and Tyr82 in the reaction catalyzed by sheep liver cytosolic serine hydroxymethyltransferase. Eur. J. Biochem. 2000, 267, 5967-76. 19. Kallen, R. G.; Jencks, W. P., The mechanism of the condensation of formaldehyde with tetrahydrofolic acid. J. Biol. Chem. 1966, 241, 5851-63. 20. Chitnumsub, P.; Ittarat, W.; Jaruwat, A.; Noytanom, K.; Amornwatcharapong, W.; Pornthanakasem, W.; Chaiyen, P.; Yuthavong, Y.;

ACS Paragon Plus Environment

35

ACS Catalysis 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 36 of 57

Leartsakulpanich, U., The structure of Plasmodium falciparum serine hydroxymethyltransferase reveals a novel redox switch that regulates its activities. Acta Crystallogr. Sect. D. Biol. Crystallogr. 2014, 70, 1517-1527. 21. Altschul, S. F.; Madden, T. L.; Schaffer, A. A.; Zhang, J.; Zhang, Z.; Miller, W.; Lipman, D. J., Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997, 25, 3389-402. 22. Altschul, S. F.; Wootton, J. C.; Gertz, E. M.; Agarwala, R.; Morgulis, A.; Schaffer, A. A.; Yu, Y. K., Protein database searches using compositionally adjusted substitution matrices. FEBS J. 2005, 272, 5101-9. 23. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J.; Brothers, E. N.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Gaussian, Inc.: Wallingford, CT, USA, 2009. 24. Mahoney, M. W.; Jorgensen, W. L., A five-site model for liquid water and the reproduction of the density anomaly by rigid, nonpolarizable potential functions. The Journal of Chemical Physics 2000, 112, 8910-8922. 25. Case, D. A.; Darden, T. A.; Cheatham, T. E.; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.; Walker, R. C.; Zhang, W.; Merz, K. M.; Roberts, B.; Hayik, S.; Roitberg, A.; Seabra, G.; Swails, J.; Goetz, A. W.; Kolossváry, I.; Wong, K. F.; Paesani, F.; Vanicek, J.; Wolf, R. M.; Liu, J.; Wu, X.; Brozell, S. R.; Steinbrecher, T.; Gohlke, H.; Cai, Q.; Ye, X.; Wang, J.; Hsieh, M. J.; Cui, G.; Roe, D. R.; Mathews, D. H.; Seetin, M. G.; Salomon-Ferrer, R.; Sagui, C.; Babin, V.; Luchko, T.; Gusarov, S.; Kovalenko, A.; Kollman, P. A., AMBER 12. University of California, San Francisco: 2012. 26. Sondergaard, C. R.; Olsson, M. H.; Rostkowski, M.; Jensen, J. H., Improved Treatment of Ligands and Coupling Effects in Empirical Calculation and Rationalization of pKa Values. J. Chem. Theory. Comput. 2011, 7, 228495. 27. Olsson, M. H.; Sondergaard, C. R.; Rostkowski, M.; Jensen, J. H., PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pKa Predictions. J. Chem. Theory. Comput. 2011, 7, 525-37.

ACS Paragon Plus Environment

36

Page 37 of 57 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

ACS Catalysis

28. Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A., Development and testing of a general amber force field. J. Comput. Chem. 2004, 25, 1157-74. 29. 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, 712-25. 30. H, S. F.; Ramos, M. J.; N, M. F. S. A. C., molUP: A VMD plugin to handle QM and ONIOM calculations using the gaussian software. J. Comput. Chem. 2018, 39, 1344-1353. 31. Humphrey, W.; Dalke, A.; Schulten, K., VMD: visual molecular dynamics. J. Mol. Graph. 1996, 14, 33-8, 27-8. 32. Cerqueira, N.; Fernandes, P. A.; Ramos, M. J., Protocol for Computational Enzymatic Reactivity Based on Geometry Optimisation. Chemphyschem 2018, 19, 669-689. 33. Dapprich, S.; Komaromi, I.; Byun, K. S.; Morokuma, K.; Frisch, M. J., A new ONIOM implementation in Gaussian98. Part I. The calculation of energies, gradients, vibrational frequencies and electric field derivatives. J. Mol. Struct.: THEOCHEM 1999, 461, 1-21. 34. Svensson, M.; Humbel, S.; Froese, R. D. J.; Matsubara, T.; Sieber, S.; Morokuma, K., ONIOM:  A Multilayered Integrated MO + MM Method for Geometry Optimizations and Single Point Energy Predictions. A Test for Diels−Alder Reactions and Pt(P(t-Bu)3)2 + H2 Oxidative Addition. J. Phys. Chem. 1996, 100, 19357-19363. 35. Maseras, F.; Morokuma, K., IMOMM: A new integrated ab initio + molecular mechanics geometry optimization scheme of equilibrium structures and transition states. J. Comput. Chem. 1995, 16, 1170-1179. 36. Yanai, T.; Tew, D. P.; Handy, N. C., A new hybrid exchange–correlation functional using the Coulomb-attenuating method (CAM-B3LYP). Chem. Phys. Lett. 2004, 393, 51-57. 37. Fernandes, H. S.; Ramos, M. J.; Cerqueira, N., The Catalytic Mechanism of the Pyridoxal-5'-phosphate-Dependent Enzyme, Histidine Decarboxylase: A Computational Study. Chemistry (Easton) 2017, 23, 9162-9173. 38. Ribeiro, A. J. M.; Santos-Martins, D.; Russo, N.; Ramos, M. J.; Fernandes, P. A., Enzymatic Flexibility and Reaction Rate: A QM/MM Study of HIV-1 Protease. ACS Catal. 2015, 5, 5617-5626. 39. Oliveira, E. F.; Cerqueira, N. M. F. S. A.; Ramos, M. J.; Fernandes, P. A., QM/MM study of the mechanism of reduction of 3-hydroxy-3-methylglutaryl coenzyme A catalyzed by human HMG-CoA reductase. Catal. Sci. Technol. 2016, 6, 7172-7185.

ACS Paragon Plus Environment

37

ACS Catalysis 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 38 of 57

40. Neves, R. P. P.; Fernandes, P. A.; Ramos, M. J., Unveiling the Catalytic Mechanism of NADP+-Dependent Isocitrate Dehydrogenase with QM/MM Calculations. ACS Catal. 2016, 6, 357-368. 41. Moreira, C.; Ramos, M. J.; Fernandes, P. A., Reaction Mechanism of Mycobacterium Tuberculosis Glutamine Synthetase Using Quantum Mechanics/Molecular Mechanics Calculations. Chem. - Eur. J. 2016, 22, 92189225. 42. Gesto, D. S.; Cerqueira, N. M. F. S. A.; Fernandes, P. A.; Ramos, M. J., Unraveling the enigmatic mechanism of l-asparaginase II with QM/QM calculations. J. Am. Chem. Soc. 2013, 135, 7146-7158. 43. Cerqueira, N. M. F. S. A.; Gonzalez, P. J.; Fernandes, P. A.; Moura, J. J. G.; Ramos, M. J., Periplasmic Nitrate Reductase and Formate Dehydrogenase: Similar Molecular Architectures with Very Different Enzymatic Activities. Acc. Chem. Res. 2015, 48, 2875-2884. 44. Cassimjee, K. E.; Manta, B.; Himo, F., A quantum chemical study of the omega-transaminase reaction mechanism. Org. Biomol. Chem. 2015, 13, 845364. 45. Fukui, K., The path of chemical reactions - the IRC approach. Acc. Chem. Res. 1981, 14, 363-368. 46. Riplinger, C.; Neese, F., An efficient and near linear scaling pair natural orbital based local coupled cluster method. J. Chem. Phys. 2013, 138, 034106. 47. Riplinger, C.; Sandhoefer, B.; Hansen, A.; Neese, F., Natural triple excitations in local coupled cluster calculations with pair natural orbitals. J. Chem. Phys. 2013, 139, 134101. 48. Liakos, D. G.; Sparta, M.; Kesharwani, M. K.; Martin, J. M.; Neese, F., Exploring the Accuracy Limits of Local Pair Natural Orbital Coupled-Cluster Theory. J. Chem. Theory Comput. 2015, 11, 1525-39. 49. Liakos, D. G.; Neese, F., Is It Possible To Obtain Coupled Cluster Quality Energies at near Density Functional Theory Cost? Domain-Based Local Pair Natural Orbital Coupled Cluster vs Modern Density Functional Theory. J. Chem. Theory Comput. 2015, 11, 4054-63. 50. Sparta, M.; Neese, F., Chemical applications carried out by local pair natural orbital based coupled-cluster methods. Chem. Soc. Rev. 2014, 43, 5032-41. 51. Paiva, P.; Sousa, S. F.; Ramos, M. J.; Fernandes, P. A., Understanding the Catalytic Machinery and the Reaction Pathway of the Malonyl-Acetyl Transferase Domain of Human Fatty Acid Synthase. ACS Catal. 2018, 48604872. 52.Bistoni, G.; Polyak, I.; Sparta, M.; Thiel, W.; Neese, F., Towards accurate QM/MM reaction barriers with large QM regions using Domain Based Pair

ACS Paragon Plus Environment

38

Page 39 of 57 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

ACS Catalysis

Natural Orbital Coupled Cluster Theory. J. Chem. Theory Comput. 2018, 14, 3524-3531. 53. Frank, N., Software update: the ORCA program system, version 4.0. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2018, 8, e1327. 54. Dunning Jr, T. H., Gaussian basis sets for use in correlated molecular calculations. I. The atoms boron through neon and hydrogen. J. Chem. Phys. 1989, 90, 1007-1023. 55. Hättig, C., Optimization of auxiliary basis sets for RI-MP2 and RI-CC2 calculations: Core–valence and quintuple-ζ basis sets for H to Ar and QZVPP basis sets for Li to Kr. PCCP 2005, 7, 59-66. 56. Weigend, F.; Köhn, A.; Hättig, C., Efficient use of the correlation consistent basis sets in resolution of the identity MP2 calculations. J. Chem. Phys. 2002, 116, 3175-3183. 57. Truhlar, D. G., Basis-set extrapolation. Chem. Phys. Lett. 1998, 294, 4548. 58. Ngo, H. P.; Cerqueira, N. M.; Kim, J. K.; Hong, M. K.; Fernandes, P. A.; Ramos, M. J.; Kang, L. W., PLP undergoes conformational changes during the course of an enzymatic reaction. Acta Crystallogr. D Biol. Crystallogr. 2014, 70, 596-606. 59. Dunathan, H. C., Stereochemical aspects of pyridoxal phosphate catalysis. Adv. Enzymol. Relat. Areas Mol. Biol. 1971, 35, 79-134. 60. Di Salvo, M. L.; Scarsdale, J. N.; Kazanina, G.; Contestabile, R.; Schirch, V.; Wright, H. T., Structure-based mechanism for early PLP-mediated steps of rabbit cytosolic serine hydroxymethyltransferase reaction. BioMed Res. Int. 2013, 2013, 458571. 61. Jain, M.; Nilsson, R.; Sharma, S.; Madhusudhan, N.; Kitami, T.; Souza, A. L.; Kafri, R.; Kirschner, M. W.; Clish, C. B.; Mootha, V. K., Metabolite Profiling Identifies a Key Role for Glycine in Rapid Cancer Cell Proliferation. Science 2012, 336, 1040-1044. 62. Dunathan, H. C., Conformation and reaction specificity in pyridoxal phosphate enzymes. Proc. Natl. Acad. Sci. U. S. A. 1966, 55, 712-6.

ACS Paragon Plus Environment

39

ACS Catalysis 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 40 of 57

For Table of Contents Only

ACS Paragon Plus Environment

40

Page 41 of 57 Oγ Cβ H O

H

C

COO



N N C8 H O 1 C7 P O 2O O1 O C2 3 O 4 NN4 5 H 6 Pyridoxal-5’-Phosphate 7 8 PLP

ACS Catalysis O H H N1N C2 N H 2N

N H

N H

C3

H N

O

ACS Paragon Plus Environment Tetrahydrofolate

THF

COO

H N

N4

COO

Lys ACS Catalysis H

N H

Page 42 of 57

O 1 H O 2 P O O O 3 O 4 N 5 H 6 7 8 9 10 Lys 11 H 2O 12 H 2O N 13 H O 14 P O O O 15 O 16 N 17 H 18 H Internal Aldimine 19 H NH2 NH2 20 HO 21 COO22 COOH H Glycine 23 HO H COO COO L-serine 24 OOC 25 N N H O H O 26 COO O P O P O 27 NH O O O O 28 O O 29 N N OOC External Aldimine External Aldimine 30 NH H (PLP + Gly) OOC (PLP + L-Ser) H 31 O HN O NH 32 NH TetraHydroFolate N 33 (THF) H 2O + NH 34 H 2N 35 N O OR 36 N ? Glu57 Protonated ? Glu57 Deprotonated N 37 N 38 α-Elimination N 5,10-mehtylene-THF H2 N H H 39 40 Glu Glu O O 41 O O 42 H H H H H NH NH OOC OOC H 43 H H H H O O 44 NH N NH N H NH NH 45 O O O O 46 NH NH N N 47 N NH2 N NH2 Conversion of THF to 48 H H 5,10-methylene-THF 49 ? Non-Enzymatic Catalysis ? Enzymatic Catalysis 50 = -CH2PO42(cytoplasm) (SHMT) 51 OR 52 ACS Paragon Plus Environment 53 54

Activation

Transimination

Page 43 of 57 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

ACS Catalysis

Tyr 65

Glu 57

THF

Arg 363

WAT

EA

Ser 175

His 126

Ser 99

Subunit B

ACS Paragon Plus Environment WAT

Subunit A

His 129 Asp 200

Protonated Glu57

Glu

Unprotonated Glu57

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

O H H H OOC H O NH H

ACS Catalysis O NH H N

O

O

Glu

N N H

OOC H NH

NH

O

NH

OOC H

H H

NH O

O

O H

H N O N

N H

Glu

NH

O H O

OOC H NH H O

NH NH2

NH H N

N H

O NH

O H O

O

N

NH2 Glu

O H

Page 44 of 57

H H N

O

Glu

N

NH

ACS Paragon Plus Environment N NH2 H

O

OOC H NH O

NH

H O

O NH H N

O N

N H

NH NH2

O NH

NH

NH NH NH2

Page 45 of 57

R

TS

P

STEP

1

THF

-0.15

-0.15

+0.05

&[

O

O

&[

H

H

O

N

O P O O O

H

Glu

2\

N

O H

O

&[

2\

H

H

N

COO

&Z

2\

H

H

H O

COO &Z

N

NH NH

O N

N H

2

&Z

H

O

THF

NH2

R

N

O P O O O

PLP + Ser

¨*‡ 18.5 ¨*R +12.1

+0.05

Glu

Glu

H N

H O

NH NH

O N

N H

215.3i cm-1

TS

STEP

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

ACS Catalysis

THF

NH2

P

Glu

-0.08

O

H

Glu O H

-0.08

+0.15

O 2\

H O P O O O

Glu

H

H

N1

H O

H N1 N

N N H

THF

N THF

NH NH

O

NH2

¨*‡ 1.7 ¨*R -0.3

H

O

&[

N

COO N

O

2\

&[

+0.15

ACS Paragon Plus Environment

H N

O P O O O

301.7i cm-1

COO H O

2\

H N1 N

O &[ H O

NH NH

N N H

THF-I

NH2

G or ∆E (kcal/mol)

1 2 3 4 5 6 7 8 9 10 11 12

ACS Catalysis

60.0

Page 46 of 57

+53.9

+46.7

50.0 40.0 30.0 20.0 10.0 0.0

+18.5

+13.8 +12.1

ACS Paragon Plus Environment

Reaction Coordinate

+11.8

Page 47 of R 57

TS

3

Glu

STEP

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

P

ACS Catalysis

N4 Glu

-0.20

O H

O H

O P O O O

N

H O

N H

N N4

H

N O Cα N1 H O N Oγ

THF-I

Glu

N1

-0.20

O

THF

NH NH NH2

R

∆G‡ 0.7 ∆GR -8.9

O P O O O

1298.0i cm-1

TS

H N4 N H

O H

COO N

H O

N H

+0.00

N O Cα N1 H O N Oγ

THF-II

NH NH NH2

P

Glu

STEP

4

COO

+0.00

Glu

COO N

N H

H O



+0.00

H N4 N H

O

H O P O O O

-0.20

O

H O Oγ



O

N

N1

N

THF-II

-0.20

Glu

N4 THF

NH NH NH2

∆G‡ ∆GR

1.5 ACS Paragon Plus Environment -7.5 1004.7i cm-1

COO N

N H

H N4 N

O H

H O P O O O

+0.00

O

H O

H

O





O

N

N1

N

THF-III

NH NH NH2

20.0

E (kcal/mol)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

ACS Catalysis

+18.6

Page 48 of 57

17.5 15.0 12.5 10.0

+9.6

Glu

H C

7.5 5.0 2.5 0.0 -180.0

Glu

HO C

H

Glu

H C2

Glu

N4

C HO

+1.7 -125.0

-90.0

H

H C C2 H N4

H OH

C2

H

OH C2

N4

N4

Plus Environment -60.0ACS Paragon -50.0 0.0

Dihedral Angle (°)

110.0

130.0

170.0

180.0

Page 49 of 57 R

TS

5

P

ACS Catalysis Tyr

STEP

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Glu

Wat

Glu

N H

N

COO N

H O



N

O Cβ N1 H O

+0.00

Wat

H

O H

H O P O O O

-0.20

O

N

THF-III

Glu

Oγ Cβ

THF

NH NH NH2

∆G‡ 3.3 ACS Paragon Plus Environment ∆GR +1.7

446.9i cm

O P O O O -1

COO N

N H

H

O

H

N1

-0.20

O

H O

+0.00

N

H N1 O C N H Oγ Cβ O N

THF-IV

NH NH NH2

6

TS

P

ACS Catalysis Tyr

Wat -0.20

H O P O O O

N H

H O H

O

Glu

O

H

COO

-0.20

+0.00

HOH H N1 N Cβ C N H H O O N

THF-IV

N1

Wat

N N4

N4

Glu



NH NH2

R

O P O O O

THF

∆G‡ 2.9 ∆GR -13.6 TS

O

H

NH

197.0i cm

-1

H H O

O

H N4N

COO N

H O



HOH

+0.00

NN1

O

NH

NH

N

NH2

N H

THF-V

P

Glu

STEP

7

Page 50 of 57

Glu

STEP

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

R

-0.15

O

H O P O O O

COO N

N H

H O H

O

Glu

H O

+0.00

Tyr

N

NH NH

N

NH2

THF-V

-0.15

O

Glu

H N N4

HOH O

Wat

Wat

∆G‡ 1.2 ACS Paragon Plus Environment ∆GR +1.0

N4

5,10-methyl

H

THF

O P O O O

744.2i cm

O H

-1

H

N

N4

COO

N

N H

H O

+0.00

H O H H O

N

O N

NH NH

NH2 5,10-methylene-THF

Page 51 of 57

1

N4

H N N4

H H

NH H N

NH

N1

O H

O

N NH2

N

THF-I(S) R

O H

¨*‡ 1.4 ¨*R +0.7

H

N1

N4 H

N4

N

2\

NH

O2\ O H

NH

H H O

N1

NH

N

2\

N

NH2

NH

N1

C

&[

&[

N

O

THF-II(S) R

¨*‡ 10.2 ¨*R +1.9

NH2

THF-III(S)

276.1i cm-1

TS

P

N1

H

NH2

N

THF-II(S)

N

N

NH

P

&[

H

O

1216.4i cm-1

TS

N4

N NH

STEP

3

P

N1

STEP

2

ACS Catalysis

TS

STEP

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

R

&[

N4

N N4 NH

NH H

N C&[ N1 H O

H N4N

NH N

N

N1

NH

&[

NH2

THF-III(S)

O

¨*‡ -3.5 ACS Paragon Plus Environment ¨*R -6.8

237.2i cm-1

N

NH2

THF-IV(S)

G (kcal/mol)

20.0 15.0 10.0 1 5.0 2 0.0 3 4 -5.0 5 -10.0 6 -15.0 7 8 -20.0 9 -25.0 10 -30.0 11 12

ACS Catalysis

Page 52 of 57 +10.9

Step1

Step2

+1.4

+2.6

+0.7

+0.7 -7.4

+0.7

Dihedral Rotation

-6.8

-8.9 -16.4

ACS Paragon Plus Environment

Reaction Coordinate

-10.1

-11.4 -14.7

-0.9

-4.2

-13.0 -25.4 -26.6

-25.6

Page 53 of R 57

TS

8

ACS Catalysis

P

STEP

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Tyr

5,10-methyl

-0.30

O Glu H

O H Cα

H O

N

N H

-0.30

O Glu



O

N

COO H O

THF

H

H O H



N



O P O O O

+0.00

Wat

Glu

H

NH

O N

NH NH2

QI ∆G‡ 13.5 ACS Paragon Plus Environment ∆GR -5.6

O P O O O

957.4i cm

-1

N H

H O

+0.00

H N

COO Oγ H N H O NH N O H H NH N O



NH2

Transimination ACS Catalysis

1 H Lys H COO2 H N 3 N H 4 H O 5 P O 6 O O O 7 N 8 H H COO9 H 10 Lys-NH2 N 11 H O 12 P O 13 O O O 14 N 15 H O 16 H Glu O H 17 External Aldimine O H N 18 (PLP + Product) 19 H COO N 20 H O NH N O 21 H O H 22 NH N P O 23 O O O 24 NH2 N 25 26 H 27 28 29 30 31 32 H O 33 Glu H O 34 O H N 35 H COO 36 N 37 HOH NH N O 38 O H 39 NH N P O 40 O O 41 O NH2 N 42 H 43 44 45 46 47 48 49 50 51 52 H O 53 Glu H O H 54 O 55 N 56 H COO 57 HOH H N 58 NH C N H O 59 H P O 60 O O NH O O N N NH2 H

5,10-methylene-THF

Lys O H P O O O H

HO O P O O O

N

Page 54 of 57

N

COOOH H Gem-diamine O

N H

COOH O

H N

Lys-NH2

THF

N H

Glu

External Aldimine (PLP + Substrate)

O H

O H

H O

O P O O O

tion a n oto r P I Q

N

COO

H

H O

N

O

n tio ina lim -E

Glu

O H

on N

e thy le n e- T H

F

N

NH NH2

Glu H

O

O P O O O

O O H COO

H

N

H N N H O O P O O O O O ACS Paragon Plus Environment N N H

NH NH NH2

N

N H

H O

H O

NH2

H

N

COO N

N H

N

N O

NH

O

H

COO

N

NH

O

H

O

N

N

N H

O

Glu H

NH

H

H O

O P O O O

si O H O

COO N

Glu

er

H O

,1 0- m

H

H

H

N

COO N

o5

H

O

O P O O O

nv

H

N H

H

O H

Ft

NH2

O

Co O

Glu

O P O O O

TH

NH

H

Serine HydroxyMethylTransferase

of

NH

N

N H

α

N

NH NH NH2

H O

H O H O

N

N N

NH NH NH2

Page 55 of 57

11 12 13

THF intermediate ACS Catalysis

5,10-methylene-THF

Protonation of the QI

18.0

kcat 10.1

G (kcal/mol)

25.0 20.0 15.0 1 10.0 2 5.0 3 0.0 4 5 -5.0 6 -10.0 7 -15.0 8 -20.0 9 -25.0 10 -30.0

α-Elimination

9.0 8.1

8.1

0.7 -7.2

-7.2

-9.7

-8.5

1

2

3

-12.1

-15.1

-16.8

ACS Paragon PlusDihedral Environment 4

Rotation

5

Reaction Coordinate

-6.1

-13.4

-29.0 -27.9 -28.5

6

7

8

H H

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

H H

THF

H H

THF

THF CH2 O

ACS Catalysis

H H

THF

H

CH2 O

H

THF

H

CH2 O

H

THF

H

H

Glu57

Glu57

Glu57

Glu57

H

Glu57H

H

CH2 O

CH2

Glu57 H

Glu57

CH2 O CH2 O

5,10 mTHF

THF

H

O

Page 56 of 57

H

5,10 mTHF

CH2

H

O

H

H

O

H

H

O

H

H

H

AA

AA

AA

AA

AA

AA

AA

AA

AA

AA

PLP

PLP

PLP

PLP

PLP

PLP

PLP

PLP

PLP

PLP

R

I1

I2

I3

I4

I5

I6

I7

I8

P

1

2

α-Elimination CH2 O

AA

3

4

5

6

7

Conversion of THF to 5,10-methylene-THF

H

= L-Ser (Substrate)

Dihedral Rotation

AA

H

ACS Paragon Plus Environment

= Gly (Product)

CH2

Glu57 H Glu57

Glu57

H

CH2

5,10 mTHF

8

QI Protonation

Page 57 of 57 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

ACS Catalysis

ACS Paragon Plus Environment