Theoretical Studies on the Catalytic Mechanism and Substrate

Subscriber access provided by UNIV OF SCIENCES PHILADELPHIA

Theoretical studies on the Catalytic Mechanism and Substrate Diversity for Macrocyclization of Pikromycin Thioesterase Ting Shi, Lanxuan Liu, Wentao Tao, Shenggan Luo, Shuobing Fan, Xiao-Lei Wang, Linquan Bai, and Yi-Lei Zhao ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b01156 • Publication Date (Web): 10 Apr 2018 Downloaded from http://pubs.acs.org on April 10, 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 42 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

Theoretical studies on the Catalytic Mechanism and Substrate

2

Diversity for Macrocyclization of Pikromycin Thioesterase

3 4

Ting Shi,1 Lanxuan Liu,1 Wentao Tao,1 Shenggan Luo,1 Shuobing Fan,1

5

Xiao-Lei Wang,1 Linquan Bai,1 Yi-Lei Zhao1,*

6 7 8 9 10 11

1

State Key Laboratory of Microbial Metabolism, Joint International Research Laboratory of Metabolic and Developmental Sciences, School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, China

12 13 14 15 16 17 18 19 20 21 22

*To whom correspondence should be addressed:

23 24

Prof. Yi-Lei Zhao

25

School of Life Sciences and Biotechnology

26

Shanghai Jiao Tong University

27

800 Dongchuan Road, Shanghai 200240, China

28

Tel/Fax: +86-21-34207190;

29

Email: [email protected]

30 1

ACS Paragon Plus Environment

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

1

Page 2 of 42

ABSTRACT

2

Polyketide synthases (PKSs) share a subset of biosynthetic steps in construction

3

of a polyketide and the offload from PKS main module of specific product release is

4

most often catalyzed by a thioesterase (TE). In spite of various PKS systems were

5

discovered in polyketide biosynthsis, the molecular basis of TE-catalyzed

6

macrocyclization remains challenging. In this study, MD simulations and QM/MM

7

methods were combined to investigate the catalytic mechanism and substrate diversity

8

of pikromycin (PIK) TE with two systems (PIK-TE-1 and PIK-TE-2), where

9

substrates

1

and

2

are

corresponding

to

TE-catalyzed

precursors

of

10

10-deoxymethynolide and narbonolide, respectively. The results showed that in

11

comparison with PIK-TE-2, system PIK-TE-1 exhibited more tendency to form stable

12

pre-reaction state, which is critical to macrocyclization. Besides, the structural

13

characteristics of pre-reaction states were uncovered through analyses of

14

hydrogen-bonding and hydrophobic interactions, which were found to play a key role

15

in substrate recognition and product release. Furthermore, potential energy surfaces

16

were calculated to study the molecular mechanism of macrocyclization including the

17

formation of tetrahedral intermediates from si- and re-face nucleophilic attacks and

18

the release of products. The energy barrier of macrocyclization from si-face attack

19

was calculated to be 16.3 kcal/mol in PIK-TE-1, 3.6 kcal/mol lower than from re-face

20

attack and 4.1 kcal/mol lower than si-face attack in PIK-TE-2. These results are in

21

agreement with experimental observations that the yield of 10-deoxymethynolide is

22

superior to narbonolide in PIK TE-catalyzed macrocyclization. Our findings elucidate

23

the catalytic mechanism of PIK TE and provide a better understanding of type I PKS

24

TEs in protein engineer.

25 26

KEYWORDS

27

thioesterase, macrocyclization, catalytic mechanism, substrate diversity, pre-reaction

28

state 2


Page 3 of 42 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

2

Polyketides and analogs have been developed to macrolide antibiotics, antifungal

3

agents, immunosuppressants and anticancer agents1-5, owing to their significant

4

medicinal properties. They are efficiently assembled by type I polyketide synthase

5

(PKS) 6, which are comprised of multiple requisite domains (i.e. “modules”). Each

6

module is responsible for a particular function in the construction of polyketide

7

scaffolds including initiation, elongation, reduction and extension. Finally, when the

8

polyketide chain reaches the desired size, the product can be offloaded from the PKSs

9

with the assistance of a thioesterase (TE) via hydrolysis or macrocyclization7-10.

10

Pikromycin, a naturally occurring ketolide antibiotic, is biosynthesized by a type

11

I PKS in Streptomyces venezuelae ATCC 1543911. Unlike other type I PKS systems,

12

the pikromycin PKS demonstrates the unique ability to efficiently generate two types

13

of macrolactone products: methymycin and pikromycin. As we all know pikromycin

14

thioesterase (PIK TE) could serve to cyclize both linear hexaketide and heptaketide

15

chain elongation intermediates towards the 12- and 14-membered macrolactones

16

10-deoxymethynolide and narbonolide, respectively (Figure 1). Further processing of

17

10-deoxymethynolide and narbonolide by the post-PKS tailoring enzymes, including

18

a glycosyltransferase and a cytochrome P450 hydroxylase, completes the biosynthesis

19

of methymycin and pikromycin.

20

PIK TE adopts an α,β-hydrolase fold and it contains a central seven-stranded

21

β-sheet connected by α-helices, with β2 antiparallel to the remaining strands. Similar

22

to substrate-binding region of 6-deoxyerythronolide B synthase (DEBS) TE, the lid

23

region is composed of two helices (L1 and L2) and a lid loop (Figure 1). Crystal

24

structures of the two homologous enzymes show many similarities, including

25

hydrophobic interface for dimerization, an open substrate channel and characteristic

26

Ser-His-Asp catalytic triad (Ser148, His268 and Asp176 in PIK TE and Ser142,

27

His259 and Asp169 in DEBS TE)12. The triad is located at the center of a long open

28

substrate channel13. In vivo, the polyketide substrate is initially bound to the 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

1

phosphopantetheinyl arm of an acyl carrier protein (ACP) and undergoes

2

transesterification to the active site Ser148 of PIK TE. Intramolecular attack by the

3

distal hydroxyl group of the substrate on the acyl-enzyme intermediate affords the

4

macrolactone, with simultaneous releasing of the product from the enzyme. Although

5

the biochemical function of these PKSs has been unveiled14,15, a detailed description

6

of how protein dynamics determines their biological role at the atomic and molecular

7

level remains a scientific challenge.

8 9 10

Figure 1. Stucture of PIK TE and macrocyclization catalyzed by PIK TE with substrate 1 and 2.

11 12

As we known, PKS TEs have been shown to be highly specific in catalytic

13

processes. Both acyl-enzyme intermediate formation and macrocyclization are

14

substrate-specific. For example, the DEBS TE appears to require a carbonyl function

15

group connected to the carbon (C7) of the substrate for the nucleophilic cyclization.

16

Replacement the carbonyl group with a hydroxyl group results in the exclusive

17

occurrence of hydrolysis16-17. On the other hand, DEBS TE demonstrates broad

18

tolerance for the ring size of the polyketide generated. In more detail, although the

19

natural product of DEBS TE is a 14-membered ring, this enzyme has been shown to

20

support the formation of alternatively functionalized 6-, 8-, 12-, 14-, and

21

16-memberedring systems. This study was expected to provide understanding of the

22

mechanism and substrate selectivity of these type I PKS TEs, which could present 4


Page 4 of 42

Page 5 of 42 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

new opportunities for chemoenzymatic synthesis of polyketides and analogs, and

2

thereby accelerate the development of new medicines in furture.

3

In this study, PIK TE is employed to develop a high-resolution model detailing

4

enzyme-substrate-interactions to rationalize macrocyclization of 10-deoxymethonilide

5

and narbonylide. We combined molecular dynamic (MD) simulations and quantum

6

mechanics/molecular mechanics (QM/MM) calculations on systems of PIK-TE-1 and

7

PIK-TE-2 in an aqueous environment. In addition, we report the formation of a

8

substrate-enzyme pre-reaction state, which is critial to mutual recognition and

9

catalytic preparation and description of its structural characteristics with

10

hydrogen-bonding and hydrophobic interactions. Furthermore, the two-layer

11

ONIOM-based QM/MM calculations were employed to investigate the catalytic

12

mechanism in detail. Transition states and key intermediates were located in the

13

calculated potential energy surface. A possible molecular mechanism of the

14

macrocyclization process was proposed. These results are in good agreement with

15

biochemical and structural studies. Our study elucidates the catalytic mechanism of

16

PIK-TE with its substrates, provides insight into the macrocyclization process of

17

PIK-TE, and increases a better understanding of type I PKS TEs in engineered PKS

18

pathway.

19 20

MATERIALS AND METHODS

21

System preparation: The crystal structure of PIK TE (PDB entry: 2HFK)15 was used as the

22

starting structure in the preparation of MD simulations. Both substrates 1 and 2 were covalently

23

bonded to residue Ser148 of PIK TE. Conformation of the substrates was determined through the

24

following procedure: 1) We blocked the terminal of the substrates by adding an N-terminal cap

25

(-CO-CH3) and a C-terminal cap (-NH-CH3); 2) We conducted a classical conformational search

26

with a systematic search method (CAESAR)18 encoded in Discovery Studio 3.5, and 300 ligand

27

conformations were generated; 3) The top 20 low-energy conformations were picked and further

28

optimized with PM3 method19–21 in Gaussian0922; 4) Conformation with the lowest energy was

29

then placed into PIK TE’s catalytic pocket and connected with the protein through covalent bond, 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

1

and the cap was removed from the substrate. Following the steps above, two complex systems,

2

PIK-TE-1 and PIK-TE-2, were set up. Files for complexes were prepared with the tleap module of

3

AMBER 12.23 After preparation of the complex structures, parameter for substrates 1 and 2 was

4

generated. We firstly performed an conformational optimization with Gaussian 09 program at the

5

level of HF/6-31G(d)24-26 and then computed its electrostatic surface potential (ESP) charge. A

6

two-step restrained electrostatic potential (RESP)27 charge fitting procedure was then carried out on

7

the substrates. Finally, missing parameters as bond and dihedral information were generated by the

8

Antechamber package.

9

MD simulation and trajectory analysis: MD simulations were performed on PIK-TE-1 and

10

PIK-TE-2 complexes using AMBER ff03.r1 force field. The proteins were solvated in a cubic box

11

of TIP3P water molecules, with the thickness of external water layer exceeding 10Å. Sodium ions

12

were then added to achieve charge neutralization in the system.

13

Both solvated systems were subjected to a minimization and heating cycle, gradually raising

14

the temperature to 298 K. After equilibrated for 50 ps, a 50-ns molecular dynamics (MD)

15

simulation was conducted on the complexes without any restraints under NPT conditions.

16

Afterwards, simulation for each complex was repeated 6 times with a different random number,

17

and one of them was extended to 300 ns. The Particle Mesh Ewald (PME) method28 was

18

employed to account for long-range electrostatic interactions, and the SHAKE algorithm in its

19

matrix form was used to fix bonds and angles involving hydrogen atoms29. Cutoff for van der

20

Waals interactions was set to 10.0 Å. This protocol was applied to all MD simulations for both

21

systems using the parallel version of PMEMD.cuda in AMBER12 suite.

22

QM and QM/MM calculation: QM/MM calculations were started from one snapshot closest to

23

the ensemble average of the most dominant cluster from MD trajectories. All the QM calculations

24

were performed with M062x30/6-31G(d) method. QM/MM calculations at the ONIOM

25

(M062x/6-31G(d):Amber) level were performed using a two-layered ONIOM method31,32 encoded

26

in the Gaussian09 program. We selected substrates 1 or 2, which contains the group (-Oser-CH2-)

27

of residue Ser148, the anionic carboxymethyl group (–CH2–COO-) of Asp176 and the side chain

28

of His268 as our QM region. A total of 74 atoms were included in the QM system of substrate 1

29

and 82 atoms in system of substrate 2. This caused the QM level to bear one negative charge while

30

the whole system had 17 negative charges, due to the removal of Na+ ions from the structure. 6


Page 6 of 42

Page 7 of 42 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

The QM region was described with density functional theory and geometries were optimized

2

with M062x exchange-correlation functional and 6-31G(d) basis set. Single-point energy

3

calculations were performed on the basis of the optimized structures using larger basis sets,

4

including 6-311+G(d), 6-311+G(d,p), and 6-311+G(2df,2p). The remainder of the system (MM

5

region) was treated with the AMBER Parm99 force field. QM/MM calculations of PIK-TE-1 and

6

PIK-TE-2 consisted of 8649 atoms and 8657 atoms, respectively. The electrostatic interactions

7

between the QM and MM regions were calculated using the electronic embedding method, which

8

provides a better description of the electrostatic interaction between the QM and MM regions and

9

allows the QM wavefunction to be polarized.

10 11 12

RESULTS AND DISCUSSION To

understand

the

catalytic

mechanism

and

substrate

diversity

of

13

macrocyclization process catalyzed by PIK TE, each of two complex systems

14

(PIK-TE-1 and PIK-TE-2) were constructed six times in 50 ns molecular dynamic

15

(MD) simulations, where substrates 1 and 2 covalently binded with the residue Ser148

16

in the active site of PIK TE. After that, two additional 300 ns simulations of PIK-TE-1

17

and PIK-TE-2 were performed to investigate the structural characteristics of

18

pre-reaction states. Furthermore, potential energy surfaces with QM/MM method

19

were calculated to study the molecular mechanism of macrocyclization including (1)

20

forming of a tetrahedral intermediate from si- or re-face nucleophilic attack of C1

21

carbonyl group, and (2) the releasing of a product 10-deoxymethynolide or

22

narbonolide. Pivotal transition states were located and the energy barriers were

23

obtained. Finally, a detailed mechanism of PIK TE catalyzed macrocyclization was

24

proposed.

25

Compared to PIK-TE-2, more “active” states were observed in PIK-TE-1

26

In our previous studies of DEBS TE33, we observed a pre-reaction state during

27

MD simulations, which is not only critical to the precise recognition of

28

enzyme-substrate, but also significant to the macrocyclization process. In this special

29

conformation, His259 and Asp169 in the catalytic triad constitute a proton transfer 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

1

chain to facilitate the deprotonation of distal hydroxyl group of the substrate and

2

further promote the intramolecular nuclear attack to obtain a tetrahedral intermediate.

3

According to our study, whether the enzyme-substrate could develop to the

4

pre-reaction state seemed to be decisive of whether macrocyclization or hydrolysis

5

occurred.

6 7

Figure 2. Pre-reaction states in PIK-TE-1 and PIK-TE-2.

8

Substrate 1 is in salmon and 2 is in magenta.

9 10

To identify the existence of pre-reaction states in PIK TE systems, six times 50

11

ns MD simulations were performed in PIK-TE-1 and PIK-TE-2, respectively (Figure

12

S1). Clustering algorithm was utilized to collect all typical substrate conformations

13

based on twelve simulation trajectories. Five representative conformations were

14

observed to be pre-reaction states including three in PIK-TE-1 system and two in

15

PIK-TE-2 system (Figure 2). Furthermore, we used two distances, which were

16

highlighted in DEBS TE pre-reaction state, to evaluate the reliability of the

17

pre-reaction state in PIK TE. One is defined by the distance between the Nε atom of

18

His268 in catalytic triad and the O atom of distal hydroxyl group of the substrate,

19

designated as d(Nε-O11) in PIK-TE-1 and d(Nε-O13) in PIK-TE-2, indicating the

20

formation of hydrogen bonding and further deprotonation of substrate hydroxyl group.

21

The other is defined between the C1 and the O atom of distal hydroxyl group,

22

designated as d(C1-O11) in PIK-TE-1 and d(C1-O13) in PIK-TE-2, demonstrating the

23

nucleophilic attack on the acyl-enzyme intermediate. 8


Page 8 of 42

Page 9 of 42 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 2

Figure 3. Conformer populations with distances of d(Nε-O11)/d(Nε-O13) and

3

d(C1-O11)/d(C1-O13) obtained from 6×50 ns MD simulations in PIK-TE-1 and

4

PIK-TE-2. Distances d(Nε-O11)/d(Nε-O13) ≤ 3.0 Å and d(C1-O11)/d(C1-O13) ≤ 4.5

5

Å are highlighted in cyan and pink.

6 7

It has been shown by QM calculations33,34 that pre-reaction state adopts a

8

reactive conformation, where the former distances is around 3.0 Å and the later is

9

close to 4.5 Å. Therefore, conformations with restraints of both d(Nε-O11)/d(Nε-O13)

10

≤ 3.0 Å and d(C1-O11)/d(C1-O13) ≤ 4.5 Å will be an indication of pre-reaction state.

11

The two distances were calculated from all conformations of MD simulations and

12

their frequency distributions were analyzed to evaluate the formation of pre-reaction

13

state in PIK-TE-1 and PIK-TE-2. As shown in the Figure 3, the average distance of

14

d(Nε-O11) in all 6×50 ns simulations is 4.32 Å, less than 4.73 Å of d(Nε-O13) in

15

PIK-TE-2. Compared d(Nε-O11) with d(Nε-O13), it is found that the population of

16

d(Nε-O11) ≤ 3.0 Å in PIK-TE-1 is about 24.3%, larger than 14.2% of d(Nε-O13) in

17

PIK-TE-2. These results suggest that the formation of a hydrogen bond between the

18

Nε atom of His268 and O atom of distal hydroxyl group in PIK-TE-1 might be

19

preferable to that in PIK-TE-2. On the other hand, although the average distance of

20

d(C1-O11) is 4.94 Å, similar to 5.03 Å of d(C1-O13), the population of distance

21

d(C1-O11) ≤ 4.5 Å in PIK-TE-1 increased by 10.3% compared with PIK-TE-2, 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

the

preference

in

formation

Page 10 of 42

1

suggesting

of

favorable

conformations

for

2

macrocyclization in PIK-TE-1. Taken together, these results demonstrated that

3

compared to PIK-TE-2, PIK-TE-1 complex could generate more “active”

4

conformations, which might be easily transform to pre-reaction states.

5 6

Figure 4. Distances of d(Nε-O11)/d(Nε-O13) and d(C1-O11)/d(C1-O13) in 300 ns

7

MD simulations in PIK-TE-1 and PIK-TE-2. The different distance proportions are

8

indicated in cycles.

9 10

To further uncover the stability of pre-reaction states, additional 300 ns MD

11

simulations were performed separately on the basis of two 50 ns simulations of

12

PIK-TE-1 and PIK-TE-2. Both d(Nε-O11)/d(Nε-O13) and d(C1-O11)/d(C1-O13) were

13

monitored (Figure 4). According to the 300 ns root-mean-square deviation (RMSD)

14

values, the systems of PIK-TE-1 and PIK-TE-2 had all reached their respective

15

equilibriums after 10 ns simulations (Figure S2). To ensure the accuracy of our results,

16

the first 10 ns simulation were not considered in our analysis. The MD trajectories

17

revealed that the most prevalent conformations adopted pre-reaction states in either

18

PIK-TE-1 or PIK-TE-2 system. To be precise, distances d(Nε-O11) in almost half of

19

conformers (46.0%) are less than 3.0 Å in PIK-TE-1, while the proportion was 10


Page 11 of 42 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

reduced to 27.5% in PIK-TE-2, indicating more appropriate orientations of distal

2

hydroxyl group towards His268 of PIK TE. What's more, once the hydrogen bond had

3

formed in PIK-TE-1, it maintained the bond for more than 90 ns simulations,

4

reflecting the stability of the hydrogen bond. On the other hand, although the average

5

distances of d(C1-O11) and d(C1-O13) were essentially in accord with each other,

6

(4.63 Å in PIK-TE-1 and 4.59 Å in PIK-TE-2), the proportion of d(C1-O11) ≤ 4.5 Å

7

in PIK-TE-1 was 24.3%, larger than the 14.2% observed in PIK-TE-2. Remarkably,

8

the “active” conformations with both d(Nε-O11) ≤ 3.0 Å and d(C1-O11) ≤ 4.5 Å in

9

PIK-TE-1 were 11.2%, i.e. approximately twice that observed in PIK-TE-2 (5.7%). In

10

conclusion, it was observed again that in comparison with PIK-TE-2, PIK-TE-1

11

indicated more favorable in formation of pre-reaction state, and once this emerged,

12

the pre-reaction state would maintain for a relatively long time in PIK-TE-1.

13 14

Structural characteristics of pre-reaction states

15

To further probe into the structural characteristics of pre-reaction states, we

16

analyzed the trajectories of PIK-TE-1 and PIK-TE-2 systems. Hydrogen-bonding and

17

hydrophobic interactions between PIK TE and substrates 1 and 2 were carefully

18

investigated. More specifically, the critical hydrogen-bonding interactions between Nε

19

atom of His268 and the active distal hydroxyl group were observed in both PIK-TE-1

20

and PIK-TE-2 (Figure 5). Additionally, a hydrogen bond between the C7 carbonyl and

21

the side chain of Thr77, which is located in the loop connecting strand β3 and helix α3,

22

was found in PIK-TE-2. By contrast, a water molecular bridged between the substrate

23

and side chain of Thr77 were captured in PIK-TE-1 (Figure S3). These were in accord

24

with previous reports that Thr77 plays a key role in stabilization of the oxygenion

25

intermediate13 and the cycle reactive conformation34. Furthermore, two dominant

26

hydrophobic interactions between binding cavity residue (Ala78 and Gly150) and

27

substrate 1 were found in PIK-TE-1 and three hydrophobic interactions between

28

Ala78, Gly150, Tyr178 and substrate 2 existed in PIK-TE-2. The mutation of Y171F

29

in DEBS TE, corresponding to Tyr178 in PIK TE, were found catalytically inactive35,

30

indicating the key role of Tyr171 in DEBS TE catalysis. However, theoretical studies 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

1

on the role of Y178 in PIK TE are quite rare. According to our simulations, Y178 has

2

hydrophobic interactions with substrate 2, suggesting sort of correlation with substrate

3

recognition. On the other hand, Y178 in PIK-TE-1 was observed to leave away from

4

active site, expanding the exit gate and consequently promoting product release

5

(Figure S4). Therefore, it was proposed that Y178 of PIK-TE might play a different

6

role in substrate recognition and product release. In short, these results demonstrated

7

that hydrogen-bonding interaction with His268 is essential to the formation of the

8

pre-reaction state and hydrophobic interaction may play a significant role in substrate

9

recognition and product release.

10 11

Figure 5. The key interactions between substrates and PIK TE.

12 13

For further study, we focused on the active site pockets in PIK-TE-1 and

14

PIK-TE-2 in 5 ns MD simulations (obtained from the dominant clusters of MD

15

simulations) utilizing POVME 2.036,37, a computational tool that characterizes the

16

shape and size of protein pockets. Although the ring sizes of substrates 1 and 2 are

17

slightly different, their binding pockets have comparable volumes of 335 Å3 in 12


Page 12 of 42

Page 13 of 42 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

PIK-TE-1 and 355 Å3 in PIK-TE-2, indicating the tolerance of PIK TE (Figure 6). We

2

speculated that PIK TE has sufficient space to accommodate different substrates in its

3

pockets and whether the substrate could transform to its pre-reaction state seemed to

4

be decisive of whether macrocyclization occurred.

5 6

Figure 6. The active binding pocket and exit of PIK TE. Substrate 1 is in salmon and

7

2 is in magenta.

8 9

Next, we carefully compared the active sites of pre-reaction states in PIK-TE-1

10

and PIK-TE-2. It observed that the major interactions between PIK TE and substrates

11

were similar in both systems, especially at the end of substrates, where the Nε atom of

12

His268 grabed the distal hydroxyl group (O11 and O13) by a hydrogen bond and

13

further stablized the substrates. However, the major difference lay in the reactive

14

center, where the orientation of C1 carbonyl group looked different, if not reverse, in

15

PIK-TE-1 and PIK-TE-2 (Figure 7). As a result, the nucleophilic attack, accompanied

16

by deprotonated hydroxyl group, would target the re- or si-face of the C1 carbonyl

17

group. The different conformations were supported by the optimization calculated by

18

the QM/MM method, where the C1 carbonyl group was found either in the same

19

direction with distal hydroxyl group, accompanied by re-face nucleophilic attack, or

20

in reverse orientation, followed by si-face attack. Besides, the conformer

21

corresponding to si-face attack was found to be 7.1 kcal/mol more stable than that of

22

re-face, demonstrating the preponderance of conformer taken by si-face attack. 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

we

optimized

10-deoxymethynolide

Page 14 of 42

1

Meanwhile,

and

narbonolide

at

the

2

M062x/6-31g(d) level and compared with the stability of them. Products obtained

3

from re-face nucleophilic attack were less stable than those from si-face by 2.7

4

kcal/mol (10-deoxymethynolide) and 3.0 kcal/mol (narbonolide). These results

5

suggested that macrocyclization from si-face nucleophilic attack may be more

6

favorable than that from re-face in thermodynamics.

7 8

Figure 7. Scheme of nucleophilic attacks of substrate 1 from re- or si-face of C1

9

carbonyl group to form tetrahedral intermediates and the optimized products of

10

10-deoxymethynolide and narbonolide.

11 12

Macrocyclization in PIK-TE-1

13

To investigate the energetic distinction between macrocyclization reacted from

14

re- or si-face nucleophilic attack, system PIK-TE-1 was carefully studied and four

15

potential energy surfaces demonstrating the formation of a tetrahedral intermediate

16

and the releaseof product 10-deoxymethynolide were plotted. These computational

17

results demonstrated that the activated energy barrier via si-face nucleophilic attack is

18

16.3 kcal/mol, 3.6 kcal/mol lower than that via re-face nucleophilic attack.

19

Macrocyclization from si-face

20

Two two-dimensional potential energy surfaces were calculated with QM/MM

21

method to aid in understanding the macrocylization catalyzed by PIK TE from si-face 14


Page 15 of 42 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

nucleophilic attack. First, a potential energy surface corresponding to formation of

2

tetrahedral intermediate was calculated by defining the distances d2(O11-H) and

3

d1(O11-C1) as the reaction coordinates, which represents the deprotonation of

4

hydroxyl group and the si-face nucleophilic attack, respectively. The potential energy

5

surface with the key structures along the reaction pathway, including 1_R, 1_TS1,

6

1_IM1, 1_TS2 and 1_P are shown in Figure 8. The distance of d1 was found to be

7

2.67 Å in the optimized structure 1_R and 1.60 Å in the optimized intermediate

8

1_IM1, and the distance of d2 is 1.01 Å in 1_R and 1.65 Å in 1_IM1. The transition

9

state 1_TS1 was located at d1=1.60 Å and d2=1.60 Å. The calculated potential energy

10

barrier of formation of the tetrahedral intermediate was 10.2 kcal/mol, indicating that

11

the formation of tetrahedral intermediate could proceed spontaneously.

12 13

Figure 8. Two-dimensional potential energy surfaces in PIK-TE-1 system including

14

the formation of tetrahedral intermediates from si-face nucleophilic attack and the

15

release of product 10-deoxymethynolide are shown with optimized structures of

16

transition state.

17 18

Next, another potential energy surface was calculated to uncover the energy

19

barrier of product release step by defining the distance d3(Oser-C1) and d4(Oser-H) as

20

the reaction coordinates, which represented the releasing product and the refreshing

21

Ser148 hydroxyl group. The distance of d3 was found to be 1.46 Å in the optimized

22

intermediate 1_IM1 and 2.57 Å in the optimized product 1_P, and the distance of d4

23

was 2.36 Å in 1_IM1 and 1.00 Å in 1_P. The transition state 1_TS2 was located at 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

1

d3=1.80 Å and d4=1.30 Å. The calculated potential energy barrier of formation of the

2

tetrahedral intermediate was 6.9 kcal/mol. Taken together, the energy barrier of the

3

macrocylization catalyzed by PIK TE from si-face nucleophilic attack was 16.3

4

kcal/mol.

5

Macrocyclization from re-face

6 7


8

the formation of tetrahedral intermediates from re-face nucleophilic attack and the

9

release of product 10-deoxymethynolide are shown with optimized structures of the

10

transition state.

11 12

We also studied the macrocyclization reacted from re-face nucleophilic attack.

13

Similar to that from si-face, the potential energy surfaces were calculated by defining

14

the distance d2(O11-H) and d1(O11-C1) as the reaction coordinates, representing the

15

deprotonation of hydroxyl group and the re-face nucleophilic attack. The key

16

structures along the reaction pathway consisting of 1_TS1' and 1_TS2' were located

17

(Figure 9). The distance of d1 was found to be 2.77 Å in the optimized structure 1_R'

18

and 1.58 Å in the optimized intermediate 1_IM1', and the distance of d2 was 1.03 Å in

19

1_R' and 1.86 Å in 1_IM1'. The transition state 1_TS1' was located at d1=1.90 Å and

20

d2=1.60 Å. The calculated potential energy barrier of formation of the tetrahedral

21

intermediate from re-face was 12.3 kcal/mol, which was 2.1 kcal/mol higher than that 16


Page 16 of 42

Page 17 of 42 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

from si-face nucleophilic attack, suggesting the macrocyclization procceding from

2

si-face nucleophilic attack is more favorable than that from re-face in energy.

3

Next, we carefully examined the structure of 1_IM1', whose energy was found to

4

be slightly lower than that of the transtion state 1_TS1' by 1.0 kcal/mol, suggesting

5

that dissociation of the intermediate state could occur easily38,39. We proposed a

6

conformational transition through the movement of imidazole in His268 to transfer its

7

proton to Ser148, since the distance between Nε of His268 and O of Ser148 is larger

8

than 4.61 Å in 1_IM1'. The conformational transition was supported through further

9

optimizations. 1_IM2' and 1_IM3' were located along the reaction pathway (Figure

10

10), which was 1.4 and 2.4 kcal/mol higher than 1_IM1' in energy, suggesting the

11

conformational transition could proceed spontaneously. This structural reorganization

12

made the O atom of Ser148 more inclined to accept the proton of His268 and promote

13

the release of products.

14 15

Figure 10. The conformational transformations in PIK-TE-1 system.

16 17

Then the energy barrier of the product releasing step was calculated with a

18

potential energy surface by defining the distance d3(Oser-C1) and d4(Oser-H) to

19

represent the departure of the product and the refreshing Ser148 hydroxyl group. The

20

distance of d3 was found to be 1.50 Å in the optimized intermediate 1_IM3' and 2.79

21

Å in the optimized product 1_P', and the distance of d4 was 2.49 Å in 1_IM3' and 1.04

22

Å in 1_P'. The transition state 1_TS2' was located at d3=1.90 Å and d4=1.70 Å. The 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

1

calculated potential energy barrier of product release was 6.2 kcal/mol, suggesting

2

that the releasing step should proceed readily.

3 4

Figure 11. The energy profile of PIK-TE-1 system.

5 6

Taken together, our computational results revealed that although the formation of

7

tetrahedral intermediate is advantageous in energy (10.2 and 12.3 kcal/mol), the

8

dissociation of the intermediate could occur easily, and the macrocyclization

9

proceeding from si-face nucleophilic attack was more favorable than that from re-face

10

attack (Figure 11). Therefore, we will focus on the macrocyclization arising from

11

si-face nucleophilic attack pathway in the following study.

12

Macrocyclization from si-face in PIK-TE-2

13

Similar to PIK-TE-1 system, two potential energy surfaces were calculated to

14

understand the formation of the tetrahedral intermediate and release of product

15

narbonolide in PIK-TE-2. According to calculations presented above, only the si-face

16

nucleophilic attack pathway was involved in further study. Distances d1(O13-C1) and

17

d2(O13-H) were used to represent the nucleophilic attack and the deprotonation of

18

hydroxyl group, while d3(Oser-C1) and d4(Oser-H) represented the releasing products

19

and the refreshing Ser148 hydroxyl group (Figure 12). The distance of d1 was found

20

to be 3.33 Å in optimized structure 2_R and 1.50 Å in optimized intermediate 2_IM,

21

and the distance of d2 was 1.00 Å in 2_R and 1.82 Å in 2_IM. The transition state

22

2_TS1 was located at d1=1.50 Å and d2=1.70 Å. The potential energy barrier for the

23

formation of the tetrahedral intermediate was calculated to be 15.4 kcal/mol, 3.1

24

kcal/mol higher than that in PIK-TE-1. Furthermore, the distance of d3 was found to

25

be 1.46 Å in optimized intermediate 2_IM and 2.66 Å in optimized product 2_P, and 18


Page 18 of 42

Page 19 of 42 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

the distance of d4 was 2.10 Å in 2_IM and 0.99 Å in 2_P. The transition state 2_TS2

2

was located at d3=1.70 Å and d4=1.30 Å. The calculated potential energy barrier of

3

product release was 5.1 kcal/mol, 1.1 kcal/mol lower than that in PIK-TE-1. To sum

4

up, the energy barrier for the PIK-TE-catalyzed macrocylization from si-face

5

nucleophilic attack in PIK-TE-2 was 20.4 kcal/mol (Figure S5).

6

7 8

Figure 12. Two-dimensional potential energy surfaces and key structures in the

9

PIK-TE-2 system.

10 11

Catalytic Mechanism

12

These data together suggest the nucleophilic attack, when accompanied by the

13

deprotonation of pre-reaction state, can proceed from either si-face or re-face to the

14

formation of a charged tetrahedral intermediate. In system PIK-TE-1, the energy

15

barrier was calculated to be of 10.2 kcal/mol in the former and 12.3 kcal/mol in the

16

later. Unlike in the si-face nucleophilic attack, the conformational reorganization was

17

found to be indispensable in re-face attack to refreshing Ser148. Finally, following the

18

collapse of the transient intermediate, the final macrocyclic product was released with

19

a relative low barrier of 6.9(6.2) kcal/mol in PIK-TE-1 and 5.1 kcal/mol in PIK-TE-2.

20

The energy barrier of the whole pathway was 16.3 kcal/mol in PIK-TE-1 and 20.5

21

kcal/mol in PIK-TE-2. Moreover, energetic corrections of key structures at different

22

level were performed and the results consistently suggest the favored preference of 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 42

1

product 10-deoxymethynolide (Table 1). As we all known, the distribution of

2

polyketide products is governed by a number of programmed events that are

3

controlled by polyketide synthases. Herein, our calculated results were coincidently in

4

agreement with the experimental observation40, where the catalysis of hexaketide

5

N-acetylcysteamine thioester (SNAC)-connected substrate under PIK module 6+TE

6

(in the presence of cosubstrate methylmalonyl CoA) outputed a 4:1 mixture of

7


8 9

Table 1. The energetic corrections at different level with M06-2x method. M062x/ 6-31g(d) 1 1_TS1 1_IM 1_TS2 EB 2 2_TS1 2_IM 2_TS2 EB

10

M062x/ 6-311+g(d)

M062x/ 6-311+g(d,p)

M062x/ 6-311+g(2df,2p)

Hartree

Hartree

Hartree

Hartree

-1618.101628 -1618.085385 10.2 kcal/mol -1618.086696 -1618.075711 6.9 kcal/mol 16.3 kcal/mol -1770.724558 -1770.700096 15.4 kcal/mol -1770.691987 -1770.728631 5.1 kcal/mol 20.4 kcal/mol




EB: energy barrier

11 12

CONCLUSIONS

13

TE-catalyzed macrocyclization of a linear polyketide acyl chain is known to be

14

the offload step in polyketide synthase (PKS)-mediated biosynthesis of macrocyclic

15

polyketides. Although substantial effort has been undertaken to elucidate the

16

structural characteristics of TEs, the molecular mechanism remains unclear. Here we

17

combined MD simulations with QM/MM calculations on the complexes of PIK-TE-1

18

and

19

enzyme-substrate-interactions to rationalize macrocyclization of 10-deoxymethonilide

20

and narbonylide by the pikromycin (PIK) TE. Compared with PIK-TE-2, "active"

PIK-TE-2

to

develop

a

high-resolution

20


model

detailing

Page 21 of 42 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

conformers were more frequently observed in PIK-TE-1 with potential to easily

2

transform into the pre-reaction state. Once the pre-reaction state had formed, it would

3

maintain for a relatively long time, which appeared to be critical to macrocyclization.

4

Besides, hydrogen-bonding interaction with His268 was pivotal to the formation of

5

the pre-reaction state and the hydrophobic interactions had potentially an essential

6

role in substrate recognition and product release. Particularly, Tyr178 was proposed to

7

widen the exit and promote product release in PIK-TE-1. Calculations on volumes of

8

binding pockets indicated that PIK TE had sufficient space to accommodate both

9

substrates 1 and 2, exhibiting the tolerance of PIK TE. Moreover, potential energy

10

surfaces were calculated with QM/MM method to obtain the transition states and the

11

energy barriers. Our computaional results indicated that the macrocylization catalyzed

12

by PIK TE from si-face nucleophilic attack is more favorable than that from re-face

13

attack in both thermodynamic and kinetic aspects. Although the formation of

14

tetrahedral intermediate was advantageous in energy, the reverse dissociation of the

15

intermediate occured easily. The energy barrier of the whole pathway was calculated

16

to be of 16.3 kcal/mol in PIK-TE-1 and 20.4 kcal/mol in PIK-TE-2. This is in good

17

agreement with biochemical and structural studies. Our study provides insight into

18

TE-catalyzed macrocyclization and increases a better understanding of type I PKS

19

TEs in the engineered PKS pathway.

20 21 22

ASSOCIATED CONTENT

23

Supporting Information

24

The Supporting Information is available free of charge on the ACS Publications

25

website at http://pubs.acs.org.

26

The root-mean-square deviation (RMSD) and root-mean-square fluctuation (RMSF)

27

values of MD simulations of PIK-TE-1 and PIK-TE-2 systems, the hydrogen-bonding

28

interactions between Thr77 and substrates, the Tyr178 in PIK-TE-1, and the energy

29

profile in PIK-TE-2 system. 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

1 2

AUTHOR INFORMATION

3

Corresponding Author

4

*Email: [email protected]. Tel/Fax: +86-21-34207190.

5

Notes

6

The authors declare no competing financial interest.

7 8

ACKNOWLEDGMENTS

9

TS, LB and YLZ conceived and designed the investigation. TS, LL, WT, SL and XLW

10

performed calculations and analyses. TS, SF and YLZ wrote up the paper.

11

The authors thank the National Basic Research Program of China “973”

12

(2012CB721005) and the National High-tech R&D Program of China “863”

13

(2012AA020403), National Science Foundation of China (21377085 and 31770070),

14

SJTU-YG2016MS42, and the SJTU-HPC computing facility award for financial

15

supports and computational hours. The authors acknowledge Dr. Timo Törmäkangas

16

for his helpful revision on language.

17

This work is dedicated to Professors Zhenyi Wen, Kendall N. Houk, and Yundong Wu

18

on the occasion of their 80th, 75th and 60th birthdays, respectively.

19 20

REFERENCES

21

(1)

Sieber, S. A.; Marahiel, M. A. Molecular mechanisms underlying nonribosomal

22

peptide synthesis: approaches to new antibiotics. Chem. Rev. 2005, 105,

23

715-738.

24

(2)

enzymes. Chem. Soc. Rev. 2009, 38, 2012-2045.

25 26

(3)

Williams, G. J. Engineering polyketide synthases and nonribosomal peptide synthetases. Curr. Opin. Struct. Biol. 2013, 23, 603-612.

27 28

Meier, J. L.; Burkart, M. D. The chemical biology of modular biosynthetic

(4)

Newman, D. J.; Cragg, G. M. Natural products as sources of new drugs over the 22


Page 22 of 42

Page 23 of 42 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

30 years from 1981 to 2010. J. Nat. Prod. 2012, 75, 311–335.

1 2

(5)

of polyketide natural products. Nat. Prod. Rep. 2015, 32, 1183-1206.

3 4

(6)

Khosla, C. Structures and mechanisms of polyketide synthases. J. Org. Chem. 2009, 74, 6416-6420.

5 6

Larsen, E. M.; Wilson, M. R.; Taylor, R. E. Conformation-activity relationships

(7)

Wang, M.; Zhou, H., Wirz, M.; Tang, Y.; Boddy, C. N. A thioesterase from an

7

iterative fungal polyketide synthase shows macrocyclization and cross coupling

8

activity and may play a role in controlling iterative cycling through product

9

offloading. Biochemistry. 2009, 48, 6288-6290.

10

(8)

255-278.

11 12

Du, L.; Lou, L. PKS and NRPS release mechanisms. Nat. Prod. Rep. 2010, 27,

(9)

Pinto, M.; Wang, M.; Horsman, M.; Boddy, C. N. 6-Deoxyerythronolide B

13

synthase thioesterase-catalyzed macrocyclization is highly stereoselective. Org.

14

Let. 2012, 14, 2278-2281.

15

(10) Xu, Y.; Zhou, T.; Zhang, S.; Xuan, L-J.; Zhan, J.; Molnar, I. Thioesterase

16

domains of fungal nonreducing polyketide synthases act as decision gates

17

during combinatorial biosynthesis. J. Am. Chem. Soc. 2013, 135, 10783-10791.

18

(11) Xue, Y.; Zhao, L.; Liu, H-W. Sherman, D. H. A gene cluster for macrolide

19

antibiotic biosynthesis in Streptomyces venezuelae: architecture of metabolic

20

diversity. Proc. Nat. Acad. Sci. USA 1998, 95, 12111-12116.

21

(12) Lu, H.; Tsai, S-C.; Khosla, C.; Cane, D. E. Expression, site-directed

22

mutagenesis, and steady state kinetic analysis of the terminal thioesterase

23

domain of the methymycin/picromycin polyketide synthase. Biochemistry. 2002,

24

41, 12590-12597.

25

(13) Tsai, S.-C.; Lu, H.; Cane, D. E.; Khosla, C.; Stroud, R. M. Insights into channel

26

architecture and substrate specificity from crystal structures of two

27

macrocycle-forming

28

Biochemistry. 2002, 41, 12598-12606.

29 30

thioesterases

of

modular

polyketide

synthases.

(14) Giraldes, J. W.; Akey, D. L.; Kittendorf, J. D.; Sherman, D. H.; Smith, J. L.; Fecik,

R.

A.

Structural

and

mechanistic 23


insights

into

polyketide

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 42

1

macrolactonization from polyketide-based affinity labels. Nat. Chem. Biol. 2006,

2

2, 531-536.

3

(15) Akey, D. L.; Kittendorf, J. D.; Giraldes, J. W.; Fecik, R. A.; Sherman, D. H.;

4

Smith, J. L. Structural basis for macrolactonization by the pikromycin

5

thioesterase. Nat. Chem. Biol. 2006, 2, 537-542.

6

(16) Aldrich, C. C.; Venkatraman, L.; Sherman, D. H.; Fecik, R. A.

7

Chemoenzymatic

synthesis

of

the

polyketide

8

10-deoxymethynolide. J. Am. Chem. Soc. 2005, 127, 8910-8911.

macrolactone

9

(17) He, W.; Wu, J.; Khosla, C.; Cane, D. E. Macrolactonization to

10

10-deoxymethynolide catalyzed by the recombinant thioesterase of the

11

picromycin/methymycin polyketide synthase. Bioorg. Med. Chem. Lett. 2006,

12

16, 391-394.

13

(18) Li, J.; Ehlers, T.; Sutter, J.; Varma-O’Brien, S.; Kirchmair, J. CAESAR: a new

14

conformer generation algorithm based on recursive buildup and local rotational

15

symmetry consideration. J. Chem. Inf. Model. 2007, 47, 1923-1932.

16 17 18 19 20 21

(19) Stewart, J. J. P. Optimization of parameters for semiempirical methods I: Method. J. Comput. Chem. 1989, 10, 209-220. (20) Stewart, J. J. P. Optimization of parameters for semiempirical methods II: Applications. J. Comput. Chem. 1989, 10, 221-264. (21) Anders, E.; Koch, R.; Freunscht, P. Optimization and application of lithium parameters for PM3. J. Comput. Chem. 1993, 14, 1301-1312.

22

(22) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;

23

Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.;

24

Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.;

25

Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;

26

Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.;

27

Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.;

28

Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.;

29

Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.;

30

Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. 24


Page 25 of 42 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

B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.;

2

Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R.

3

L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J.

4

J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.;

5

Cioslowski, J.; Fox, D. J. Gaussian 09; Gaussian, Inc. Wallingford, CT, 2013.

6

(23) Case, D. A.; Darden, T. A.; Cheatham, III, T. E.; Simmerling, C. L.; Wang, J.;

7

Duke, R. E.; Luo, R.; Walker, R. C.; Zhang, W.; Merz, K. M.; Roberts, B.;

8

Hayik, S.; Roitberg, A.; Seabra, G.; Swails, J.; Götz, A. W.; Kolossváry, I.;

9

Wong, K. F.; Paesani, F.; Vanicek, J.; Wolf, R. M.; Liu, J.; Wu, X., Brozell, S.

10

R.; Steinbrecher, T.; Gohlke, H.; Cai, Q.; Ye, X.; Wang, J.; Hsieh, M.-J.; Cui,

11

G.; Roe, D. R.; Mathews, D. H.; Seetin, M. G.; Salomon-Ferrer, R.; Sagui, C.;

12

Babin, V.; Luchko, T.; Gusarov, S.; Kovalenko, A.; Kollman, P. A. AMBER 12,

13

University of California, San Francisco, 2012.

14

(24) Ditchfield, R.; Hehre, W. J.; Pople, J. A. Self-consistent molecular orbital

15

methods. IX. An extended Gaussian-type basis for molecular orbital studies of

16

organic molecules. J. Chem. Phys. 1971, 54, 724-728.

17

(25) Hehre, W. J.; Ditchfield, R.; Pople, J. A. Self-consistent molecular orbital

18

methods. XII. Further extensions of Gaussian-type basis sets for use in

19

molecular orbital studies of organic molecules. J. Chem. Phys. 1972, 56,

20

2257-2261.

21 22

(26) Payne, P. W. The Hartree-Fock theory of local regions in molecules. J. Am. Chem. Soc. 1977, 99, 3787-3794.

23

(27) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson,

24

D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. A Second

25

Generation Force Field for the Simulation of Proteins, Nucleic Acids, and

26

Organic Molecules. J. Am. Chem. Soc. 1995, 117, 5179-5197.

27

(28) Darden, T.; York, D.; Pedersen, L. Particle Mesh Ewald: An N•log(N) Method

28

for Ewald Sums in Large Systems. J. Chem. Phys.1993, 98, 10089-10092.

29

(29) Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. Numerical integration of the

30

cartesian equations of motion of a system with constraints: molecular dynamics 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

1 2

Page 26 of 42

of n-alkanes. J. Comput. Phys. 1977, 23, 327-341. (30) Zhao, Y.; Truhlar, D. G. The M06 Suite of Density Functionals for Main Group

3

Thermochemistry,

Thermochemical

Kinetics,

Noncovalent

Interactions,

4

Excited States, and Transition Elements: Two New Functionals and Systematic

5

Testing of Four M06-Class Functionals and 12 Other Functionals. Theor. Chem.

6

Acc. 2008, 120, 215-241.

7

(31) Dapprich, S.; Komáromi, I.; Byun, K. S.; Morokuma, K.; Frisch, M. J. A new

8

ONIOM implementation in Gaussian 98. Part 1. The calculation of energies,

9

gradients and vibrational frequencies and electric field derivatives. J. Mol.

10

Struct. (Theochem) 1999, 462, 1-21.

11

(32) Vreven, T.; Byun, K. S.; Komáromi, I.; Dapprich, S.; Montgomery, J. A.;

12

Morokuma, K.; Frisch, M. J. Combining Quantum Mechanics Methods with

13

Molecular Mechanics Methods in ONIOM. J. Chem. Theory Comput. 2006, 2,

14

815-826.

15

(33) Chen, X-P.; Shi, T.; Wang, X-L.; Wang, J.; Chen, Q.; Bai, L.; Zhao, Y-L.

16

Theoretical

Studies

on

the

Mechanism

of

Thioesterase-Catalyzed

17

Macrocyclization in Erythromycin Biosynthesis. ACS Catal. 2016, 6,

18

4369-4378.

19

(34) Koch, A. A.; Hansen, D. A.; Shende, V. V.; Furan, L. R.; Houk, K. N.; Jimé

20

nez-Osés, G.; Sherman, D. H. A Single Active Site Mutation in the Pikromycin

21

Thioesterase Generates a More Effective Macrocyclization Catalyst. J. Am.

22

Chem. Soc. 2017, 139, 13456-13465.

23

(35) Wang, M. Boddy, C.N. Examining the role of hydrogen bonding interactions in

24

the substrate specificity for the loading step of polyketide synthase thioesterase

25

domains. Biochemistry. 2008, 47, 11793-11803.

26

(36) Durrant, J. D.; Votapka, L.; Sørensen, J.; Amaro, R. E. POVME 2.0: An

27

Enhanced Tool for Determining Pocket Shape and Volume Characteristics. J.

28

Chem. Theory Comput. 2014, 10, 5047-5056.

29

(37) Wagner, J. R.; Sørensen, J.; Hensley, N.; Wong, C.; Zhu, C.; Perison, T.; Amaro,

30

R. E. POVME 3.0: Software for Mapping Binding Pocket Flexibility. J. Chem. 26


Page 27 of 42 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

Theory Comput. 2017, 13, 4584-4592.

2

(38) Świderek, K.; Martí, S.; Moliner, V. Theoretical Study of Primary Reaction of

3

Pseudozyma antarctica Lipase B as the Starting Point To Understand Its

4

Promiscuity. ACS Catal. 2014, 4, 426-434.

5

(39) Martínez-González, J. Á.; González, M.; Masgrau, L.; Martínez, R. Theoretical

6

Study of the Free Energy Surface and Kinetics of the Hepatitis C Virus

7

NS3/NS4A Serine Protease Reaction with the NS5A/5B Substrate. Does the

8

Generally Accepted Tetrahedral Intermediate Really Exist? ACS Catal. 2015, 5,

9

246-255.

10

(40) Wu, J.; He, W.; Khosla, C.; Cane, D. C. Chain elongation, macrolactonization,

11

and hydrolysis of natural and reduced hexaketide substrates by the

12

picromycin/methymycin polyketide synthase. Angew. Chem. Int. Ed. Engl. 2005,

13

44, 7557-7560.

14 15

FIGURE LEGEND

16

Figure 1. Stucture of PIK TE and macrocyclization catalyzed by PIK TE with

17

substrate 1 and 2.

18

Figure 2. Pre-reaction states in PIK-TE-1 and PIK-TE-2. Substrate 1 is in salmon and

19

2 is in magenta.

20

Figure 3. Conformer populations with distances of d(Nε-O11)/d(Nε-O13) and

21

d(C1-O13)/d(C1-O13) obtained from 6×50 ns MD simulations in PIK-TE-1 and

22

PIK-TE-2. Distances d(Nε-O11)/d(Nε-O13) ≤ 3.0 Å and d(C1-O11)/d(C1-O13) ≤ 4.5

23

Å are highlighted in cyans and pink.

24

Figure 4. Distances of d(Nε-O11)/d(Nε-O13) and d(C1-O11)/d(C1-O13) in 300 ns

25

MD simulations in PIK-TE-1 and PIK-TE-2. The different distance proportions are

26

indicated in cycles.

27

Figure 5. The key interactions between substrates and PIK TE.

28

Figure 6. The active binding pocket and exit of PIK TE. Substrate 1 is in salmon and

29

2 is in magenta. 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

1

Figure 7. Scheme of nucleophilic attacks of substrate 1 from re- or si-face of C1

2

carbonyl group to form tetrahedral intermediates and the optimized products of

3


4


5

the formation of tetrahedral intermediates from si-face nucleophilic attack and the

6

release of product 10-deoxymethynolide are shown with optimized structures of

7

transition state.

8


9

the formation of tetrahedral intermediates from re-face nucleophilic attack and the

10

release of product 10-deoxymethynolide are shown with optimized structures of the

11

transition state.

12

Figure 10. The conformational transformations in PIK-TE-1 system.

13

Figure 11. The energy profile of PIK-TE-1 system.

14

Figure 12. Two-dimensional potential energy surfaces and key structures in PIK-TE-2

15

system.

16 17 18 19 20 21 22 23 24 25 26 27 28 29 28


Page 28 of 42

Page 29 of 42 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

TABLE OF CONTENTS (TOC) GRAPHIC

2

3 4 5 6 7 8 9 10

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

Figure 1. Stucture of PIK TE and macrocyclization catalyzed by PIK TE with substrate 1 and 2. 1179x643mm (120 x 120 DPI)


Page 30 of 42

Page 31 of 42 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. Pre-reaction states in PIK-TE-1 and PIK-TE-2. Substrate 1 is in salmon and 2 is in magenta. 834x337mm (120 x 120 DPI)


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

Figure 3. Conformer populations with distances of d(Nε-O11)/d(Nε-O13) and d(C1-O11)/d(C1-O13) obtained from 6×50 ns MD simulations in PIK-TE-1 and PIK-TE-2. Distances d(Nε-O11)/d(Nε-O13) ≤ 3.0 Å and d(C1O11)/d(C1-O13) ≤ 4.5 Å are highlighted in cyans and pink. 1061x531mm (120 x 120 DPI)


Page 32 of 42

Page 33 of 42 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. Distances of d(Nε-O11)/d(Nε-O13) and d(C1-O11)/d(C1-O13) in 300 ns MD simulations in PIK-TE1 and PIK-TE-2. The different distance proportions are indicated in cycles. 903x518mm (120 x 120 DPI)


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

Figure 5. The key interactions between substrates and PIK TE. 855x789mm (120 x 120 DPI)


Page 34 of 42

Page 35 of 42 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 6. The active binding pocket and exit of PIK TE. Substrate 1 is in salmon and 2 is in magenta. 1141x836mm (120 x 120 DPI)


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

Figure 7. Scheme of nucleophilic attacks of substrate 1 from re- or si-face of C1 carbonyl group to form tetrahedral intermediates and the optimized products of 10-deoxymethynolide and narbonolide. 1113x873mm (120 x 120 DPI)


Page 36 of 42

Page 37 of 42 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. Two-dimensional potential energy surfaces in PIK-TE-1 system including the formation of tetrahedral intermediates from si-face nucleophilic attack and the release of product 10-deoxymethynolide are shown with optimized structures of transition state. 1159x876mm (120 x 120 DPI)


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

Figure 9. Two-dimensional potential energy surfaces in PIK-TE-1 system including the formation of tetrahedral intermediates from re-face nucleophilic attack and the release of product 10-deoxymethynolide are shown with optimized structures of the transition state. 1164x922mm (120 x 120 DPI)


Page 38 of 42

Page 39 of 42 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 10. The conformational transformations in PIK-TE-1 system. 1213x1038mm (120 x 120 DPI)


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

Figure 11. The energy profile of PIK-TE-1 system. 1791x609mm (120 x 120 DPI)


Page 40 of 42

Page 41 of 42 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 12. Two-dimensional potential energy surfaces and key structures in PIK-TE-2 system. 1037x606mm (120 x 120 DPI)


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

TOC


Page 42 of 42

Theoretical Studies on the Catalytic Mechanism and Substrate

Recommend Documents