THEORETICAL INVESTIGATION OF

ENANTIOSELECTIVE LIGAND-HOST BINDING

INTERACTIONS

BY

SURIYAWUT KULATEE

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE

REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

(ENGINEERING AND TECHNOLOGY)

SIRINDHORN INTERNATIONAL INSTITUTE OF TECHNOLOGY

THAMMASAT UNIVERSITY

ACADEMIC YEAR 2017

Ref. code: 25605922040059HVU

THEORETICAL INVESTIGATION OF

ENANTIOSELECTIVE LIGAND-HOST BINDING

INTERACTIONS

BY

SURIYAWUT KULATEE

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE

REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

(ENGINEERING AND TECHNOLOGY)

SIRINDHORN INTERNATIONAL INSTITUTE OF TECHNOLOGY

THAMMASAT UNIVERSITY

ACADEMIC YEAR 2017

Ref. code: 25605922040059HVU

ii

Acknowledgements

I would like to thank Assoc. Prof. Dr. Luckhana Lawtrakul for designing,

conceiving, and analyzing the experiments. Special thanks to Assoc. Prof. Dr. Pisanu

Toochinda and Col. Asst. Prof. Dr. Anotai Suksangpanomrung for analyzing and co-

writing the publication.

Special acknowledgements to my beloved advisor, Assoc. Prof. Dr. Luckhana

Lawtrakul for her invaluable encouragement and guidance throughout the study. I

would like to convey my utmost gratitude to her understandings in both the academic

and personal activities of mine. Without her special care and support, this study would

not be a success.

I would love to dedicate my integrity towards my parents, Mr. Yuwachart

Kulatee and Mrs. Pornthip Sukchuen, for their endless encouragements and supports

throughout the study.

Suriyawut Kulatee was financially supported by the Excellent Thai Student

(ETS) scholarship program of Sirindhorn International Institute of Technology (SIIT).

The author gratefully acknowledges the Center of Nanotechnology, Kasetsart

University for the Gaussian 09 program package.

Mr. Suriyawut Kulatee

Ref. code: 25605922040059HVU

iii

Abstract

THEORETICAL INVESTIGATION OF ENANTIOSELECTIVE

LIGAND-HOST BINDING INTERACTIONS

by

SURIYAWUT KULATEE

Bachelor of Engineering (Chemical Engineering), Sirindhorn International Institute of

Technology, Thammasat University, 2016.

Master of Science (Engineering and Technology), Sirindhorn International Institute of

Technology, Thammasat University, 2018.

Enantiomerically pure compounds are highly demanded by the pharmaceutical,

agrochemical, food additives, fragrances, and catalysts industries. Their separations

require appropriate technique and the problem arise with the selection of appropriate

systems of chiral selector-selectand combinations. The identification of suitable

selectors for separating enantiomers requires considerable experimentation which is

highly demanding with respect to time, material and labor. With the help of molecular

modeling, the researchers can visualize three-dimensional structures of any chemical

compounds in any kind of simulated environment, allowing them to study their physical

and chemical properties. In this study, two chiral systems are studied; the chemical

system consisting of beta-cyclodextrin (chiral host) and nicotine enantiomers (chiral

guests), and the biological system consisting of wild-type and mutant pfDHFR (chiral

hosts) and cycloguanil derivatives enantiomers (chiral ligands). The research aims to

study the effect of stereochemistry in chiral hosts on the chiral guests and to predict

their enantioselectivity as the potential chiral selector. The findings indicate the

potential use of both beta-cyclodextrin and pfDHFR as the chiral selectors in separating

nicotine and cycloguanil derivatives enantiomers, respectively.

Keywords: Enantioselectivity, enantiomer, chiral resolution, and molecular modeling

Ref. code: 25605922040059HVU

iv

Table of Contents

Chapter Title Page

Signature Page i

Acknowledgements ii

Abstract iii

Table of Contents iv

List of Figures vi

List of Tables vii

1 Introduction 1

1.1 Background 1

1.2 Scope of the study 4

1.3 Objectives of the study 5

2 Literature Review 6

2.1 Chemical system (beta-cyclodextrin and nicotine enantiomers) 6

2.1.1 Beta-cyclodextrin 9

2.1.2 Nicotine enantiomers 10

2.2 Biological system (Plasmodium falciparum Dihydrofolate 11

Reductase and cycloguanil derivatives enantiomers)

2.2.1 Plasmodium falciparum Dihydrofolate Reductase (pfDHFR) 13

2.2.2 Cycloguanil (Cyc) enantiomers 14

3 Methodology 16

Ref. code: 25605922040059HVU

v

3.1 Guest/ligand structures preparation 16

3.2 Host structures preparation 16

3.3 Molecular docking calculation setup 16

3.4 Complex optimization setup 17

3.5 ∆E calculation 17

4 Result and discussion 20

4.1 Enantioselectivity of nicotine by beta-cyclodextrin 20

4.2 Enantioselectivity of Cycloguanil derivatives by pfDHFR 25

5 Conclusions and Recommendations 29

References 30

Appendices 35

Appendix A 36

Ref. code: 25605922040059HVU

vi

List of Tables

Tables Page

2.1 Properties of alpha-, beta-, and gamma-cyclodextrin [16] 6

2.2 Chiral resolution studies using BCD or its derivatives as the chiral

selector

8

2.3 General properties of nicotine enantiomers 11

2.4 Cycloguanil derivatives substituents (R1, R2) dataset. Compound

names and substituent details are taken from [38]

15

4.1 Binding energy (kcal mol-1) calculation of (R)-nicotine from

molecular docking calculations and ∆E (kcal mol-1) calculation of

nicotine/BCD inclusion complex from PM6 calculations

20

4.2 Binding energy (kcal mol-1) calculation of (S)-nicotine from

molecular docking calculations and ∆E (kcal mol-1) calculation of

nicotine/BCD inclusion complex from PM6 calculations

20

4.3 Binding energy (kcal mol-1) comparison of enantiomeric Cyc

derivatives between molecular docking calculation and experimental

data [38]

25

Ref. code: 25605922040059HVU

vii

List of Figures

Figures Page

1.1 Diagrammatic representation of two non-superimposable mirror

images of (R)- and (S)-enantiomer

1

1.2 Optical activities of (R)- and (S)-enantiomer, respectively 1

1.3 Examples of racemic drugs that exert different biological responses

[1]

2

1.4 Chiral resolution techniques in both the analytic and preparative

scale [3]

3

1.5 Lock and key mechanism of binding interaction. (R)- and (S)-

enantiomer are shown as left and right compound, respectively. (R)-

enantiomer binds to the host while (S)-enantiomer cannot

5

2.1 Chemical structure of beta-cyclodextrin. Image source: Sci.

Pharm. 2018,86(2), 20; doi:10.3390/scipharm86020020

7

2.2 Stereocenters of a single glucose unit in beta-cyclodextrin (BCD)

structure. A total of five chiral centers are located at C-1 to C-5, each

depicted by R and S symbols. R and S stands for chiral center in that

configuration, respectively

9

2.3 Chemical structures of nicotine enantiomers (a) (R)-nicotine; (b) (S)-

nicotine, respectively. Chiral center is shown as black asterisk. (R)-

nicotine has hydrogen atom pointed inside the plane of paper (hollow

wedge), while (S)-nicotine has hydrogen atom pointing outside the

paper (bold wedge)

10

2.4 Conversion of dihydrofolate (DHF) to tetrahydrofolate (THF) by

Dihydrofolate Reductase (DHFR) [32]

11

2.5 Chemical structures of (a) cycloguanil; (b) the general structure of

its derivatives. Cyc consists of a chlorophenyl ring and a 1,3,5-

dihydrotriazine ring. Chiral center is shown as black asterisk. X and

X’ are meta-positions, while Y is para-position

12

Ref. code: 25605922040059HVU

viii

2.6 Three-dimensional structures of wild-type pfDHFR crystal, obtained

from Protein Data Bank (PDB ID: 3UM8 [33])

13

2.7 Amino acid comparisons within the binding pockets of (a) wild-type;

(b) mutant pfDHFR (right), respectively. Cycloguanil and amino

acids are shown as stick and line model, respectively. Dark grey,

blue, red, green, and yellow represents carbon, nitrogen, oxygen,

chlorine, and sulfur atoms, respectively. Hydrogen atoms were

removed for clarity. Asterisk represents chiral centers

14

3.1 Summary of methodological flowchart for chemical system 18

3.2 Summary of methodological flowchart for biological system 19

4.1 ∆E calculation for determining the favorability of nicotine/BCD

inclusion complex formation

21

4.2 Structure of nicotine/BCD inclusion complex from molecular

docking (left side) and PM6 calculations (right side). (a), (b), and (c)

represents (R)-nicotine/BCD inclusion complex of ranking 1,2, and

3, respectively. (d) and (e) represents (S)-nicotine/BCD inclusion

complex of ranking 1 and 2, respectively

23

4.3 The minimized structure of nicotine/BCD inclusion complex’s

binding interactions of (a) (R)-nicotine; (b) (S)-nicotine

24

4.4 Simplified view of the binding interactions of Cyc derivatives inside

the wild-type and mutant pfDHFR binding pockets

26

4.5 Superposition image of Cyc derivatives (p- and m-chlorophenyl

substituent) with the reference structure in the wild-type pfDHFR

binding pocket. Cyc24, 25, 26, 27, 28, 29, 30, 31, and 42 (R2 is alkyl

chain) in: (a) R configuration; (b) S configuration. Cyc32, 33, 34, 35,

36, 37, 38, 39, 40, 41, 43, 44, 45, and 46 (R2 is phenol chain) in: (c)

R configuration; (d) S configuration. Cyc derivatives and the

reference structure are shown as line model and stick model,

respectively. Black, blue and green indicates carbon, nitrogen, and

chlorine atom, respectively

27

Ref. code: 25605922040059HVU

ix

4.6 Superposition image of Cyc derivatives (p- and m-chlorophenyl

substituent) with the reference structure in the mutant pfDHFR

binding pocket. Cyc24, 25, 26, 27, 28, 29, 30, 31, and 42 (R2 is alkyl

chain) in: (a) R configuration; (b) S configuration. Cyc32, 33, 34, 35,

36, 37, 38, 39, 40, 41, 43, 44, 45, and 46 (R2 is phenol chain) in: (c)

R configuration; (d) S configuration. Cyc derivatives and the

reference structure are shown as line model and stick model,

respectively. Black, blue and green indicates carbon, nitrogen, and

chlorine atom, respectively

28

Ref. code: 25605922040059HVU

1

Chapter 1

Introduction

1.1 Background

Enantiomers are isomeric compounds containing at least one asymmetric chiral

center (C, N, F, etc.) that have two non-superimposable images. These compounds exist

in two enantiopure forms; (R)- and (S)-enantiomer, with optical activity. Enantiomers

have identical physical and chemical properties, except they rotate plane polarized light

in opposite directions (Figure 1.1 and 1.2).

Figure 1.1 Diagrammatic representation of two non-superimposable mirror images of

(R)- and (S)-enantiomer.

Figure 1.2 Optical activities of (R)- and (S)-enantiomer, respectively.

4

1

2

3

1

4

3

2

(+) or D = Clockwise

(-) or L = Anti-clockwise

R configuration S configuration

Absolute configuration Optical Activity

(+) or D = Clockwise

(-) or L = Anti-clockwise

Optical Activity

S

Ref. code: 25605922040059HVU

2

Enantiomeric drugs (or chiral drugs) are gaining immeasurable importance in

the field of pharmaceutical industry, as well as in the field of therapeutic applications.

On a molecular level, chirality represents an intrinsic property of the building blocks of

life [1]. Most of the functional enzymes in living organisms are chiral compounds, e.g.

amino acids, sugars, proteins and nucleic acids [2-3]. For some therapeutics, single-

enantiomer formulations can provide greater desirable effects than a mixture of

racemates (equal mixture of two pure enantiomers) [2-4]. This would, in theory, only

require half of the effective dose of 50:50 racemic mixture [2-3]. In fact, very often one

of them represents the more active isomer (eutomer), while the other one might be

active (distomer) in a different way, contributing to side-effects, displaying toxicity, or

acting as antagonist [1]. Taking a mixture of a eutomer and a distomer (in the form of

racemates) may lead to different biological responses like: (i) distomer is inactive when

compared to eutomer, (ii) distomer has the same biological activity as eutomer, (iii)

distomer is less potent than eutomer, (iv) distomer acts as an antagonist to eutomer, (v)

distomer exerts an adverse effect on eutomer, and (vi) distomer exerts different

therapeutic effects than eutomer [1,5]. A few examples of racemic drugs that exert

different therapeutic effects are shown in Figure 1.3 [1].

Figure 1.3 Examples of racemic drugs that exert different biological responses [1].

* *

****

Ref. code: 25605922040059HVU

3

The impact of chirality on almost any pharmacological and biological process

is well recognized and is finding strong repercussion on many fields of economic

interests, such as the development of drugs, agrochemicals, food additives, fragrances,

new materials, and catalysts [3]. The increasing demands in enantiomerically pure

compounds, techniques that are robust, cost-effective, and capable of parallelism are

required for both the analytical and preparative scales [3]. The process of separating

pure enantiomers from racemic mixture is called chiral resolution. Until date, chiral

resolution is achieved via three main techniques; (i) chiral resolution by crystallization,

(ii) chiral resolution by chiral derivatizing agents, and (iii) chiral resolution by chiral

column chromatography. New emerging methods like capillary electrophoresis, liquid-

liquid extraction, sensors, membranes, and biotransformation asymmetric catalysis are

still developing and have been studied extensively [3, 6-9]. A summarized information

about chiral resolution techniques is shown in Figure 1.4. Problems, however, arise

with the selection of appropriate systems chiral selector-selectand combinations from

the constantly growing repertoire of chiral selectors [10]. The identification of suitable

selectors for a specific pair of enantiomers requires considerable experimentation and

might be, therefore, highly demanding with respect to time, material and labor [3].

Clearly, there is a need for empirical and/or rational strategies to facilitate this tedious

selection procedure and to avoid to some extent the “trial-and-error” approach [3].

Figure 1.4 Chiral resolution techniques in both the analytic and preparative scale [3].

ENANTIOSEPERATIONTECHNIQUES

BiotransformationAsymmetric catalysis

Liquid-liquidextraction

Sensors

Capillary electrophoresis Chromatography

Crystallization

Membranes

Analytical Analytical/Preparative Frequently used Emerging techniques

Ref. code: 25605922040059HVU

4

In order to reduce such inconveniences, chiral resolution of racemic compounds

can be investigated by computational chemistry; molecular modeling approach.

Molecular modeling is a computer-based means of representing, visualizing and

investigating the three-dimensional structures and related properties of molecules.

Modern biochemical texts are resplendent with marvelous computer-generated pictures

representing chemical and biological molecules [11]. This technique aids the

researchers to visualize three-dimensional structures of any chemical and biological

molecules in any kind of simulated environment, allowing them to study their physical

and chemical properties, for example shape, size and charge; to simulate the dynamic

behavior of atoms and molecules, such as their vibrational, twisting and rotational

movements; to explore their interactions with other molecules; to design rationally

molecules of biological and clinical interest; and, perhaps most importantly, to greatly

improve scientific communication and the teaching of all aspects of biomolecular

sciences [11]. With the help of molecular modeling, enantioselectivity of the chiral

hosts towards chiral guests can be investigated. Such measures can help reduce the

experimental work load, by generating hypotheses of the subsequent experiment

testing, which is quicker and cost-effective.

1.2 Scope of the study

In this study, two systems (chemical and biological) of chiral molecules are

investigated by molecular modeling. Chemical system of compounds consists of beta-

cyclodextrin (BCD) and nicotine enantiomers; (R)- and (S)-nicotine. While biological

system of compounds consists of Plasmodium falciparum Dihydrofolate Reductase

(pfDHFR) and cycloguanil (Cyc) derivatives enantiomers; (R)- and (S)-cycloguanil

derivatives. The hosts and guests in both the systems are chiral. Based on the lock and

key mechanism of binding interaction, compounds that can fit in perfectly inside the

host’s binding cavity, will results in binding interaction and thus, the formation of

chemical/or biological complexes (Figure 1.5).

Three-dimensional structures of nicotine and Cyc derivatives enantiomers were

constructed using GaussView 5 [12]. The resulting structures were further optimized

for the lowest energy state via Gaussian 09 [13]. While the three-dimensional structures

of BCD and pfDHFR (wild and mutant type) can be obtained from central database.

Ref. code: 25605922040059HVU

5

The downloaded structures were optimized via Discovery Studio Visualizer 4.0 [14].

Then, nicotine enantiomers were docked into the BCD’s binding cavity and Cyc

derivatives enantiomers were docked into the pfDHFR’s binding site via AutoDock 4.2,

using default settings at a hundred docking frequencies [15]. In this calculation, ligands

were freely flexible while hosts were kept rigid. The results obtained from AutoDock

calculation were reported in terms of binding energies (BE). A hundred docking

calculations were classified into many clusters depending on the structural

conformations of ligands within the hosts’ molecules. The best BE were selected from

the cluster with highest frequencies and were further analyzed. From the analysis,

binding interactions of chemical and biological complexes can be studied and predicted

using hypotheses, as whether the chiral hosts can be used to separate pure enantiomers

from racemic mixture or not.

Figure 1.5 Lock and key mechanism of binding interaction. (R)- and (S)-enantiomer are

shown as left and right compound, respectively. (R)-enantiomer binds to the host while

(S)-enantiomer cannot.

1.3 Objectives of the study

This study aims at understanding the relationship of stereochemistry of chiral

hosts on chiral guests. The scope is to investigate the enantioselective properties of

chiral hosts towards chiral guests, using molecular modeling approach. The obtained

results then will be used for predicting the potential chiral hosts which will be beneficial

for other researchers and experimentalists.

Host

SR

Ref. code: 25605922040059HVU

6

Chapter 2

Literature Review

2.1 Chemical system (beta-cyclodextrin and nicotine enantiomers)

Cyclodextrins (CDs) are membered rings of glucose units connected to each

other by (alpha-1,4)-glycosidic bondage. Three primary types of CDs are alpha-, beta-

, and gamma-cyclodextrin (ACD, BCD, and GCD, respectively), each containing six,

seven, and eight glucose units. Summarized details of CDs are given in Table 2.1.

Table 2.1 Properties of alpha-, beta-, and gamma-cyclodextrin [16].

Properties ACD BCD GCD

No. of glucopyranose units 6 7 8

Chemical formula C36H60O30 C42H70O35 C48H80O40

Molar mass in Dalton units 972.8 1134.9 1297.1

Water molecules in cavity 6 11 17

Water solubility (w/v at 25℃) 14.5 1.8 23.2

Inner diameter in Å 5 ± 0.3 6.3 ± 0.3 7.9 ± 0.4

Outer diameter in Å 14.6 ± 0.5 15.4 ± 0.1 17.5 ± 0.6

Height of structure in Å 7.9 ± 0.1 7.9 ± 0.1 7.9 ± 0.1

Melting range in ˚C 255-260 255-265 240-245

Water of crystallization 10.2 13-15 8-18

Cavity volume (ml/mol) 174 262 472

Price 1.0 0.025 0.8

(USD/g of pharmaceutical grade)

What sets CDs apart from other compounds is that, CDs exhibit intrinsic

property of both hydrophobicity and hydrophilicity. The interior of CDs is joined by

glycosidic bonds (or ethereal linkages), resulting in the internal cavity of all CDs to

have hydrophobic property. Whereas the outer edges of CDs are occupied with primary

and secondary hydroxyl (-OH) functional groups. Primary -OH groups are located at

C-2 and C-3 of the wider rim while secondary -OH group is located at C-6 of the narrow

Ref. code: 25605922040059HVU

7

rim. The presence of -OH groups on the outer edges of CDs resulted in the hydrophilic

property. Three-dimensional structure of BCD showing seven membered-ring,

glycosidic linkage, and -OH functional groups is shown in Figure 2.1.

Figure 2.1 Chemical structure of beta-cyclodextrin. Image source: Sci. Pharm. 2018,

86(2), 20; doi:10.3390/scipharm86020020.

Cyclodextrins (CDs) are the most popular of the many chiral selectors used in

capillary electrophoresis (CE); a chiral resolution technique used for analytical purpose.

CE is an analytical technique used for separation of charges/ionic analytes based on their

electrophoretic mobility, under the influence of applied voltage. During the process,

positively charged analytes are reduced (gain e-) at cathode and negatively charged

analytes are oxidized (loss e-) at anode, via the capillary tube of the size 20 to 100 μm.

Due to their good enantioselective property towards a wide range of analytes, their good

water solubility, and their transparency towards UV light down to low wavelengths, CDs

closely resemble the ideal chiral selector and they are used in about two-thirds of the

literature applications (refer to Table 2.2) [17-18].

Ref. code: 25605922040059HVU

8

Table 2.2 Chiral resolution studies using BCD or its derivatives as the chiral selector.

Method Selector Selectant Year Ref.

CE Sulfobutyl-BCD Ephedrine, and pseudo- 1994 [19]

ephedrine enantiomers

CE Carboxymethyl-BCD Neurotransmitters 2001 [20]

Membrane BCD Tryptophan enantiomers 2007 [21]

CE BCD derivatives Higenamine enantiomers 2017 [22]

MM BCD Isopulegol enantiomers 2013 [23]

MM BCD Propranolol enantiomers 2014 [24]

MM BCD Valine enantiomers 2015 [25]

MM, NMR BCD Asenapine enantiomers 2016 [26]

MM BCD Leucine enantiomers 2017 [27]

Sensors BCD Tyrosine enantiomers 2017 [28]

MM BCD Mansonone enantiomers 2018 [29]

CE = capillary electrophoresis, Membrane = membrane technology,

Sensors = enantioselectivity sensing, MM = molecular modeling,

NMR = nuclear magnetic resonance spectroscopy.

CDs are available in many sizes. They give fast kinetics for the formation and

breakdown of complexes with enantiomers and are relatively cheap [17]. Successful

application of CDs in CE, has followed their use as chiral stationary phases in gas

chromatography (GC), thin-layered chromatography (TLC), and high-performance liquid

chromatography (HPLC), and as mobile phase additives in TLC and HPLC. Most early

workers used the parent ACD, BCD, and GCD but most interest has now shifted to the

substituted cyclodextrin derivatives, particularly those of BCD and its derivatives

because of its low cost and accessibility [17-18]. However, these techniques require

considerable experimentation and might be, therefore, highly demanding with respect to

time, material and labor [3]. To reduce such inconveniences, theoretical investigation

using molecular modeling can be quite handful, as seen in later studies shown in Table

2.2 [23-29]. Therefore, preliminary analysis via molecular modeling can be useful in

Ref. code: 25605922040059HVU

9

predicting possible hypotheses before conducting real experiments and thus, saves time

and cost.

2.1.1 Beta-cyclodextrin (BCD)

Figure 2.2 Stereocenters of a single glucose unit in beta-cyclodextrin (BCD) structure.

A total of five chiral centers are located at C-1 to C-5, each depicted by R and S

symbols. R and S stands for chiral center in that configuration, respectively.

Beta-cyclodextrin is a seven-membered glucose ring, joined by (alpha-1,4)-

glycosidic bonds. BCD is known for its cheap price and high productivity in the field of

chiral resolution [19-29]. Each glucose unit has five chiral centers as shown in Figure 2.2.

Therefore, BCD is a chiral host. As mentioned earlier, BCD and its derivatives are

gaining popularity among researchers due to its lower cost and higher efficiency in its

fast kinetic for inclusion complex formation with the chiral molecules [23-29]. The use

of BCD in forming inclusion complex with guest enantiomers will provide greater

insights towards its enantioselectivity towards guest enantiomers, as well as providing us

with the in-depth molecular interactions that are responsible for higher affinity of BCD

towards one of the guest enantiomers. From these findings, the possibility of using BCD

in separating nicotine enantiomers can be concluded. Please be noted that such

predictions are derived from preliminary analysis. In order to reduce biases from fixed

host flexible guest calculation, molecular dynamics calculation where both the host and

guest are flexible, should be further performed.

Ref. code: 25605922040059HVU

10

2.1.2 Nicotine enantiomers

Nicotine is a volatile alkaloid that can be extracted from the tobacco leaves.

Nicotine forms a very stable complex with hemoglobin, allowing them to be transported

from the lungs to the brain. Nicotine indirectly stimulates the release of neurotransmitters

in various parts of the brain, which collectively resulted in the feeling of relaxation,

sustained attention, stimulation, alertness, and calmness [30]. Nicotine is an optically

active stereoisomer. Due to its structural arrangement, the molecule has one chiral center

at C-1. Nicotine enantiomer compositions in naturally occurring tobacco leaves is

dominated by ~99.0-99.9% of (S)-nicotine and ~0.1-1.2% of (R)-nicotine [31]. The

structure of nicotine is made up of the covalent linkage between pyrimidine ring and

methylpyrrolidine ring. The chiral center exists at the position of carbon atom of

methylpyrrolidine ring that is linked to the pyrimethamine ring. Molecular structures of

nicotine enantiomers and its properties are shown in Figure 2.3 and Table 2.3,

respectively. In this study, the enantioselectivity of BCD towards nicotine enantiomers

were investigated.

Figure 2.3 Chemical structures of nicotine enantiomers (a) (R)-nicotine; (b) (S)-nicotine,

respectively. Chiral center is shown as black asterisk. (R)-nicotine has hydrogen atom

pointed inside the plane of paper (hollow wedge), while (S)-nicotine has hydrogen atom

pointing outside the paper (bold wedge).

Pyrimethamine ring

(R)-nicotine (S)-nicotine

Methylpyrrolidine ring

Ref. code: 25605922040059HVU

11

Table 2.3 General properties of nicotine enantiomers.

Properties (R)-nicotine (S)-nicotine

Chemical formula C10H14N2 C10H14N2

Molar mass in g/mol 162.236 162.236

Polarization effect Rotate clockwise (+) Rotate anti-clockwise (-)

Melting point in ˚C -79 -79

Boiling point in ˚C 247 247

No. of H bond donor 0 0

No. of H bond acceptor 2 2

No. of chiral center 1 1

Rotatable bond count 1 1

2.2 Biological system (Plasmodium falciparum Dihydrofolate Reductase and

cycloguanil derivatives enantiomers)

Figure 2.4 Conversion of dihydrofolate (DHF) to tetrahydrofolate (THF) by

Dihydrofolate Reductase (DHFR) [32].

Ref. code: 25605922040059HVU

12

Plasmodium falciparum Dihydrofolate Reductase (pfDHFR) is a key enzyme

responsible for the livelihood of Plasmodium falciparum parasites [5]. This parasite

contributes up to 87% of all the Plasmodium population in Thailand (based on reports

from Thailand’s Center of Disease Control). The enzyme Dihydrofolate Reductase

(DHFR) is responsible for the reproductive cycle of the Plasmodium parasite, which is

responsible for the malaria infection in human. DHFR converts dihydrofolate to

tetrahydrofolate, which is the building blocks of nucleic acids (refer to Figure 2.4).

Cycloguanil (Cyc), a kind of the anti-folate, is used to tackle DHFR enzymatic

activity by inhibiting DHFR’s activation and thus, no conversion of dihydrofolate.

However, prolong use of same line of drug resulted in pfDHFR resistance towards anti-

folate. Among various types of mutated pfDHFR, double-point mutations at residue 16

(alanine mutates to valine) and at residue 108 (serine mutates to threonine) confer

cycloguanil (Cyc) resistance in the double mutant variant pfDHFR (A16V + S108T) [5,

34-35]. Cyc derivatives (a new line of anti-folate) are designed and experimentally tested

against both the wild-type and mutant pfDHFR (A16V + S108T) [34-35]. The structure

of Cyc derivatives contain one chiral center at the C-2 position of chlorophenyl ring,

resulting them to be enantiomers (Figure 2.5).

Figure 2.5 Chemical structures of (a) cycloguanil; (b) the general structure of its

derivatives. Cyc consists of a chlorophenyl ring and a 1,3,5-dihydrotriazine ring. Chiral

center is shown as black asterisk. X and X’ are meta-positions, while Y is para-position.

Ref. code: 25605922040059HVU

13

So far, there is no reports about which type of Cyc derivatives enantiomers; the

(R)- or (S)-enantiomer is more potent towards mutant pfDHFR. This is important because

the pure enantiomeric form of a chiral drug can exert desirable or non-desirable responses

on the body or both [1-2,4-5]. Important living enzymes are mostly made up of chiral

molecules. Chances are that, they will exhibit “lock and key” mechanism as illustrated in

Figure 1.2. Since both pfDHFR and cycloguanil derivatives are chiral molecules, we can

investigate the enantioselectivity of chiral host (wild-type and mutant pfDHFR) towards

chiral ligands (Cyc derivatives enantiomers) via molecular modeling.

2.2.1 Plasmodium falciparum Dihydrofolate Reductase (pfDHFR)

Figure 2.6 Three-dimensional structures of wild-type pfDHFR crystal, obtained from

Protein Data Bank (PDB ID: 3UM8 [33]).

In this study, two types of pfDHFR are being studies; the wild-type pfDHFR

(PDB ID: 3UM8 [33]) and double mutant variant (A16V+S108T) pfDHFR (PDB ID:

3UM6 [34]). The crystal structure of pfDHFR is shown in Figure 2.6. The three-

dimensional protein network is made up of two domains; the dihydrofolate (DHFR)

DHFR domain

(N-terminal)

TS domain

(C-terminal)

pfDHFR

binding site

-helix -helix -loop

Ref. code: 25605922040059HVU

14

domain and the thymidylate synthase (TS) domain. A total of 608 amino acid residues

with 231 residues belonged to DHFR domain and 288 residues belonged to the TS

domain. The remaining 89 residues belonged to the junction region that serves as the

bridge joining the DHFR to TS domain [35]. The enantioselectivity of both the wild-type

and mutant pfDHFR will be investigated in the region of interests i.e., the DHFR’s

binding site. From exterior view of both types of pfDHFR, it is not possible to notice their

differences. For better perspectives, the comparison of binding pockets of both the wild-

type and mutant pfDHFR is shown in Figure 2.7.

Figure 2.7 Amino acid comparisons within the binding pockets of (a) wild-type; (b)

mutant pfDHFR (right), respectively. Cycloguanil and amino acids are shown as stick

and line model, respectively. Dark grey, blue, red, green, and yellow represents carbon,

nitrogen, oxygen, chlorine, and sulfur atoms, respectively. Hydrogen atoms were

removed for clarity. Asterisk represents chiral centers.

2.2.2 Cycloguanil (Cyc) derivatives enantiomers

Cyc is well known for its inhibiting potential against DHFR enzyme. Lately,

pfDHFR mutated and became resistant to Cyc. One of the type i.e., the double mutant

variant (A16V+S108T) was found to be directly associated with the Cyc resistance

[5,36-37]. Chemical structures of Cyc and Cyc derivatives are shown in Figure 2.5. Cyc

consists of a 1,3,5-dihydrotriazine ring with a 2,2-dimethyl substitution at the C-2

position and p-chlorophenyl substitution at the N-1 position. In this study, there are two

Ref. code: 25605922040059HVU

15

types of R1, R2 substituents; the non-bulky alkyl chains and the bulky phenolic chains.

The substitution of R1, R2 substituents into the general structure of Cyc derivatives

gives the R-enantiomer. On the other hand, switching the position of R1, R2 substituents

constitute the S-enantiomer. The substitution of flexible substituents at C-2 gives rise

to asymmetric carbon chiral center. As a result, Cyc derivatives can exist as (R)- or (S)-

enantiomers. The dataset of Cyc derivatives are presented in Table 2.4.

Table 2.4 Cycloguanil derivatives substituents (R1, R2) dataset. Compound names and

substituent details are taken from [38].

Comp. X Y R1 R2

Cyc H Cl Me Me

23 Cl H Me Me

24 H Cl Me nPr

25 Cl H Me iPr

26 H Cl Me iPr

27 Cl H Me nPr

28 H Cl Me nHex

29 Cl H Me nHex

30 H Cl H Me

31 Cl H H Me

32 H Cl H C6H5

33 Cl H H C6H5

34 H Cl H 4-C6H5OC6H4

35 Cl H H 4-C6H5OC6H4

36 H Cl H 3-C6H5OC6H4

37 Cl H H 3-C6H5OC6H4

38 H Cl H 3-C6H5CH2OC6H4

39 Cl H H 3-C6H5CH2OC6H4

40 H Cl H 3-(4-ClC6H4O)C6H4

41 Cl H H 3-(4-ClC6H4O)C6H4

42 Cl H H nC7H15

43 Cl H H 4-PrOC6H4

44 Cl H H 3-(3,5-Cl2C6H3O)C6H4

45 Cl H H 3-[2,4,5-Cl3C6H2O(CH2)3O]C6H4

46 Cl H H 3-(3-CF3C6H4O)C6H4

X and X’: m-positions; Y: p-position; green: chlorine atom.

Ref. code: 25605922040059HVU

16

Chapter 3

Methodology

In this study, there are two types of molecular systems; the chemical and

biological system. In chemical system, the host is beta-cyclodextrin (BCD) and the

guests are nicotine enantiomers. In biological system, the host is wild-type and double

mutant variant (A16V+S108T) Plasmodium falciparum Dihydrofolate Reductase

(pfDHFR) and the guests are cycloguanil (Cyc) derivatives enantiomers.

3.1 Guest/Ligand structures preparation

Three dimensional structures of nicotine (guests) and Cyc derivatives

enantiomers (ligands) are constructed using GaussView 0 5 [ 12]. Chemical structures

of nicotine enantiomers are submitted to geometry optimization with the basis set PM6,

gaseous state, by Gaussian 09 [13]. Similarly, Cyc derivatives underwent geometry

optimization but with the different basis set i.e., Hatree-Fock/6-31G (d,p), gaseous

environment by Gaussian 09 [13].

3.2 Host structures preparation

The crystal structures of BCD is downloaded from Cambridge Crystallographic

Data Center (CCDC). The reference code for BCD is BCDEXD03 [39]. pfDHFR’s

crystal structure is download from RCSB Protein Data Bank; Wild-type pfDHFR (PDB

ID: 3UM8) [33] and double-mutant variant (A16V+S108T) pfDHFR (PDB ID: 3UM6)

[34]. The downloaded crystal structures are then optimized in Discovery Studio

Visualizer by removing water of crystallization molecules and add polar hydrogen

atoms [14]. The final structures of both the optimized ligands and hosts are docked in

together using AutoDock 4.2.6 tools [15].

3.3 Molecular docking calculation setup

The optimized structures of nicotine enantiomers are docked into the crystal

structure of BCD while Cyc derivatives enantiomers are docked into the crystal

structure of wild-type and mutant pfDHFR, using AutoDock 4.2.6 [15]. The small

Ref. code: 25605922040059HVU

17

molecules (guests or ligands) are kept flexible, while the hosts are kept rigid. Gasteiger

charges are assigned to the system before performing molecular docking simulation. A

grid size of 40 × 40 × 20 (chemical system) and 60 x 60 x 60 (biological system) with

0.375 Å spacing was assigned. The dimensions and coordinates of grid boxes are kept

constant throughout the docking process. One hundred docking calculations are

performed on each guest-host complex using the Lamarckian genetic algorithm with

remaining parameters run at default settings [40]. The results obtained are classified

into different clusters with different binding energies which are used for further

analysis.

3.4 Complex optimization setup

The results obtained from molecular docking calculations predict the

configuration of guests/ligands within the host’s binding site. Each configuration is

saved from the results and then embedded into the host’s structure to form binding

complex. The inclusion complex of biological system (Cyc derivatives/pfDHFR

complex) is rather big and cannot be optimize due to computational limitations.

Therefore, smaller complex like nicotine/BCD is furthered optimized using PM6,

gaseous state basis set, by Gaussian 09 [13].

3.5 ∆E calculation

After performing nicotine/BCD complex optimization, ∆E is calculated. The

negative and positive value of ∆E indicates that inclusion complex formation is

favorable and non-favorable, respectively.

The equilibrium equation is described by the equation:

Guest + Host ↔ Guest/Host inclusion complex)

Guest = (R)- and (S)-nicotine

Host = BCD

∆E calculation: ∆E (kcal mol-1) = EINCLUSION COMPLEX – (EHOST + EGUEST)

Ref. code: 25605922040059HVU

18

Figure 3.1 Summary of methodological flowchart for chemical system

Ref. code: 25605922040059HVU

19

Figure 3.2 Summary of methodological flowchart for biological system

Ref. code: 25605922040059HVU

20

Chapter 4

Result and Discussion

4.1 Enantioselectivity of nicotine by beta-cyclodextrin

Table 4.1 Binding energy (kcal mol-1) calculation comparison of (R)-nicotine from

molecular docking calculations and ∆E (kcal mol-1) calculation of nicotine/BCD

inclusion complex from PM6 calculations.

Docking Calculation ∆E Calculation

Rank Freq. BE S.D. EBCD E(R)-NICOTINE ECOMPLEX ∆E

1 76 -4.16 ± 0.02 -1572.31 27.76 -1557.33 -12.78

2 16 -4.10 ± 0.00 -1572.31 27.76 -1561.91 -17.37

3 8 -4.08 ± 0.00 -1572.31 27.76 -1556.06 -11.51

Freq. = frequency; BE = binding energy; S.D. = standard deviation

Table 4.2 Binding energy (kcal mol-1) calculation comparison of (S)-nicotine from

molecular docking calculations and ∆E (kcal mol-1) calculation of nicotine/BCD

inclusion complex from PM6 calculations.

Docking Calculation ∆E Calculation

Rank Freq. BE S.D. EBCD E(S)-NICOTINE ECOMPLEX ∆E

1 60 -4.15 ± 0.01 -1572.31 27.72 -1554.31 -9.72

2 40 -4.13 ± 0.00 -1572.31 27.72 -1553.58 -8.98

Freq. = frequency; BE = binding energy; S.D. = standard deviation

The results obtained from molecular docking calculation of (R)- and (S)-

nicotine enantiomers into the beta-cyclodextrin (BCD) binding site is shown in Table

4.1 and 4.2, respectively. From one hundred docking frequencies, rankings of the

clusters are classified as rank 1, 2, and 3. One rank classification represents one

structural conformation of the nicotine enantiomer within the BCD’s binding site. It is

observed that (R)- and (S)-nicotine has three and two rankings, respectively. To obtain

∆E, the inclusion complexes were re-optimized by Gaussian 09 using PM6 calculation

by Gaussian 09 [13].

Ref. code: 25605922040059HVU

21

Figure 4.1 ∆E calculation for determining the favorability of nicotine/BCD inclusion

complex formation.

The results in Table 4.1 and 4.2 indicate the enantioselectivity of BCD towards

(R)-nicotine is more than the (S)-nicotine. In terms of molecular docking calculations,

clear difference could not be established to predict the enantioselectivity of BCD

towards one of the nicotine enantiomers. In order to observe the movement of nicotine

inside BCD’s binding site, nicotine/BCD inclusion complex optimization was

performed using PM6 gas phase calculation. From PM6 calculation, we obtain ∆E

values which are calculated by subtracting the energy of reactants i.e., nicotine and

BCD from the energy of nicotine/BCD inclusion complex, as shown in Figure 4.1. In

equilibrium perspective, negative and positive ∆E means that the formation of inclusion

complex is favorable and non-favorable, respectively.

From the results, BCD is more enantioselective towards the (R)-nicotine than

the (S)-nicotine. To support the claim, ∆E values of each nicotine/BCD inclusion

complex were compared and the comparison shows that (R)-nicotine of ranking 2 has

the lowest ∆E value of -17.37 kcal mol-1 while (S)-nicotine of ranking 2 has the highest

∆E value of -8.98 kcal mol-1. This value shows that the formation of nicotine/BCD

inclusion complex is more favorable by the (R)-nicotine than the (S)-nicotine. The

lower ∆E values corresponds not only to the favorability of inclusion complex, they

also indicate the distribution of retention time of each nicotine enantiomer within the

BCD Nicotine Nicotine/BCD inclusion complex

∆E= EINCLUSION COMPLEX - (EBCD + ENICOTINE)

- ∆E means formation of inclusion complex is favorable

+∆E means formation of inclusion complex is not favorable

Ref. code: 25605922040059HVU

22

chromatography column. The lower the ∆E value, the longer that inclusion complex

retentate within the column.

In order to understand the phenomena of higher and lower ∆E values, binding

interactions of nicotine enantiomers with BCD were investigated. As previously stated,

one ranking obtained from molecular docking calculation represents one structural

conformation of nicotine within the BCD’s binding pocket. The comparison of

nicotine/BCD inclusion complex conformation obtained from molecular docking

calculation and PM6 calculation is shown in Figure 4.2. In the figure, the structure of

nicotine is shown as stick model while BCD is shown as line model. The gray opaque

surface surrounding the BCD’s structure is the Van der Waals surface. From the figure,

it is observed that the structure of nicotine experiences minor shift from its original

position after performing nicotine/BCD inclusion optimization via PM6 gaseous phase

calculation. The minor shift in the structure of nicotine was the response to avoid steric

hindrance between the nicotine and the BCD molecule.

Nicotine molecule consists of methylpyrrolidine and pyrimethamine ring (refer

to Figure 2.3). The pyrimethamine ring is aromatic because of the presence of

delocalized electrons. One torsional point is present at the chiral carbon, allowing the

methylpyrrolidine to rotate freely. From Figure 4.2, the methylpyrrolidine ring of both

the (R)- and (S)-nicotine rotated itself to avoid steric hindrance, while the

pyrimethamine ring remained rigid, except in Figure 4.2c and 4.2d where the

pyrimethamine did actually move. From Figure 4.2, nicotine/BCD inclusion complex

of 4.2b (lowest ∆E; (R)-nicotine/BCD) and 4.2d (lowest ∆E; (S)-nicotine/BCD) were

further evaluated for the effect of BCD’s stereoselectivity on nicotine enantiomers.

Their binding interactions are shown in Figure 4.3a and 4.3b, respectively.

Ref. code: 25605922040059HVU

23

Figure 4.2 Structure of nicotine/BCD inclusion complex from molecular docking (left

side) and PM6 calculations (right side). (a), (b), and (c) represents (R)-nicotine/BCD

inclusion complex of ranking 1,2, and 3, respectively. (d) and (e) represents (S)-

nicotine/BCD inclusion complex of ranking 1 and 2, respectively.

Ref. code: 25605922040059HVU

24

Figure 4.3 The minimized structure of nicotine/BCD inclusion complex’s binding

interactions of (a) (R)-nicotine; (b) (S)-nicotine.

From binding interaction analysis of Figure 4.3, (R)- and (S)-nicotine were

primarily driven into BCD’s binding pocket by van der Waals forces and hydrophobic

interactions. In (R)-nicotine/BCD inclusion complex, one hydrogen bonding of the

distance 2.53 Å was present. This hydrogen bonding is formed between nitrogen atom

of methylpyrrolidine ring of nicotine and a secondary -OH group of BCD. The bond

length of 2.53 Å represents medium strong hydrogen bonding with the bond energy of

the range 4-14 kcal mol-1 [41]. The formation of hydrogen bond in (R)-nicotine/BCD

inclusion complex, results it to have more negative ∆E value than the (S)-nicotine/BCD

inclusion complex by (-17.37) - (-9.72) = -7.65 kcal mol-1. Hydrogen bonding formed

also prevented (R)-nicotine from being rejected from the nicotine/BCD inclusion

complex. As a result, the folding of secondary -OH groups of the wider rim, furthermore

strengthened the inclusion complex formed between (R)-nicotine of ranking 2 and

BCD.

The stereochemistry of BCD is dominated by 80% (R)-chiral carbon centers and

20% (S)-chiral carbon centers. The stereochemistry of both the BCD and nicotine

enantiomers were well preserved after molecular docking and PM6 calculations. No

special relationship between BCD’s stereochemistry and nicotine enantiomers were

observed. Despite the presence of high majority of (R)-chiral centers in BCD, BCD

still bind to both nicotine enantiomers but BCD binds more preferably with the (R)-

nicotine due to the formation of hydrogen bonding.

Ref. code: 25605922040059HVU

25

4.2 Enantioselectivity of Cycloguanil derivatives by pfDHFR

The results in Table 4.3 are selected according to the guideline of essential

binding characteristics of a good pfDHFR inhibitor (Figures 4.4) [42]. The selected

results are compared with the experimental data by superimposing the structural

conformations of the selected results with the Cyc structure present in the parent x-ray

crystal structure. The good superposition indicates that Cyc derivatives have similar

binding interaction characteristics to the guideline (Figure 4.4) and the bad

superposition indicates some internal steric hindrance.

Table 4.3 Binding energy (kcal mol-1) comparison of enantiomeric Cyc derivatives

between molecular docking calculation and experimental data [38].

a m-Cl at X flips to X’ position; b Poor conformation; X: m-position; Y: p-position; R1,

R2: substituents; R: Cyc derivatives in R configuration; S: Cyc derivatives in S

configuration; Exp.: Experimental data; Comp.: Compound; Bold: Enantiomer

configuration with the lowest BE.

3UM8 3UM6

Comp X Y R1 R2 R S Exp R S Exp

Cyc H Cl Me Me -8.12

-7.98

-12.04 -7.70

-7.70

-8.02

23 Cl H Me Me -11.63 -8.88

24 H Cl Me nPr -8.07 -6.85b -11.54 -8.08 -7.09b -6.87

25 Cl H Me iPr -8.59 -7.30b -10.36 -8.20a -7.41b -7.72

26 H Cl Me iPr -8.72 -7.12b -10.15 -7.70 -7.37b -5.93

27 Cl H Me nPr -8.14 -8.01 -11.37 -8.07 -7.57a -8.63

28 H Cl Me nHex -7.75b -8.26 -12.58 -7.81 -7.79 -8.21

29 Cl H Me nHex -7.85 -8.17 -11.76 -7.65a -7.62a -9.54

30 H Cl H Me -8.34 -7.76 -11.44 -7.98 -7.53 -9.41

31 Cl H H Me -8.26a -7.83 -10.90 -8.40a -7.48 -10.11

32 H Cl H C6H5 -8.97 -8.39 -11.39 -9.18 -7.21b -9.97

33 Cl H H C6H5 -8.80a -8.67 -10.82 -9.34a -7.19 -10.88

34 H Cl H 4-C6H5OC6H4 -8.57b -9.47 -12.82 -9.19 -9.04b -11.49

35 Cl H H 4-C6H5OC6H4 -8.47b -9.97 -12.49 -9.01a -7.22b -11.69

36 H Cl H 3-C6H5OC6H4 -8.60 -9.49 -12.69 -8.91 -6.81 -11.69

37 Cl H H 3-C6H5OC6H4 -8.73b -9.85 -12.22 -8.50 -8.55b -11.73

38 H Cl H 3-C6H5CH2OC6H4 -8.12 -8.82 -12.49 -8.40 -6.74 -11.20

39 Cl H H 3-C6H5CH2OC6H4 -9.31b -9.82 -11.78 -8.14 -7.01 -11.59

40 H Cl H 3-(4-ClC6H4O)C6H4 -8.82 -10.04 -12.08 -8.96 -7.27 -11.08

41 Cl H H 3-(4-ClC6H4O)C6H4 -9.25b -10.40 -12.12 -8.79 -7.39 -11.54

42 Cl H H nC7H15 -8.39a -8.03 -11.69 -8.33a -7.59 -11.49

43 Cl H H 4-PrOC6H4 -7.43b -8.87 -11.57 -9.43a -8.05b -10.96

44 Cl H H 3-(3,5-Cl2C6H3O)C6H4 -8.90b -10.09 -11.93 -8.62 -6.93b -11.36

45 Cl H H 3-[2,4,5-Cl3C6H2O(CH2)3O]C6H4 -9.18b -10.20 -11.46 -7.68 -3.49b -11.71

46 Cl H H 3-(3-CF3C6H4O)C6H4 -8.20b -9.98 -11.69 -8.62 -6.74 -11.17

Ref. code: 25605922040059HVU

26

Figure 4.4 Simplified view of the binding interactions of Cyc derivatives inside the

wild-type and mutant pfDHFR binding pockets.

The presence of poor conformations in the results were analyzed. The outcomes

indicated that chiral binding pocket of both the wild-type and mutant pfDHFR have

preferential binding towards one form of enantiomer. The non-preferred enantiomer

cannot bind property due to its failure to comply with the “lock and key” mechanism,

thereby, making them experiencing steric hindrance with one or more of the pfDHFR

side chains. For better insights, the structures of enantiomeric Cyc derivatives were

superimposed inside the wild and mutant pfDHFR binding pocket as shown in Figure

4.5 (wild-type pfDHFR) and 4.6 (mutant pfDHFR).

In wild-type pfDHFR, Cyc derivatives with alkyl chains (except Cyc28 and 29)

are preferred for the (R)-enantiomer and Cyc derivatives with phenol chains (except

Cyc32 and 33) are preferred for the (S)-enantiomer. (R)-Cyc derivatives with alkyl

chains have better binding activity than (S)-Cyc derivatives because they can avoid

steric hindrance with the Phe58 side chains. (S)-Cyc derivatives with phenol chains

have better binding activity than (R)-Cyc derivatives because they can avoid steric

hindrance with the Leu46 and Met55 side chains. Cyc28, 29, 32, and 33 are exceptions

because the size of their substituents is the transition between non-bulky alkyl chains

and bulky phenol chains.

Ref. code: 25605922040059HVU

27

Figure 4.5 Superposition image of Cyc derivatives (p- and m-chlorophenyl substituent)

with the reference structure in the wild-type pfDHFR binding pocket. Cyc24, 25, 26,

27, 28, 29, 30, 31, and 42 (R2 is alkyl chain) in: (a) R configuration; (b) S configuration.

Cyc32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 43, 44, 45, and 46 (R2 is phenol chain) in: (c)

R configuration; (d) S configuration. Cyc derivatives and the reference structure are

shown as line model and stick model, respectively. Black, blue and green indicates

carbon, nitrogen, and chlorine atom, respectively.

In mutant pfDHFR, the BE values have similar trend to the experimental data.

Cyc derivatives, irrespective of the substituent type, are preferred for the (R)-

enantiomer. Mutant pfDHFR is made up of chiral centers. The chirality within mutant

pfDHFR are similar to that of wild-type, except for Thr108 that contains two chiral

centers (R and S). The highest available enantiomers are in S configuration (refer to

Figure 2.6). The increase in the bulkiness of Val16 and Thr108 results in the reduction

of binding pocket volume around them. Mutation at residue 108 results in Cyc

derivatives with p-Cl (except Cyc28, 34, 36, 38 and 40) to experience steric hindrance

Ref. code: 25605922040059HVU

28

with Thr108 side chains. Val16 is situated in front Phe58 (refer to Figure 2.6). Because

Val16 is bulkier than Ala16, the pocket volume Val16 and Phe58 is reduced, resulting

in the (R)-Cyc derivatives to have better binding activity than the (S)-Cyc derivatives.

The superposition images show that poor conformation or the non-preferred enantiomer

configurations have binding interactions that are different from the reference structure.

This is because they have steric hindrance with one or more of the amino acid side

chains within the binding pocket, resulting them to have higher energy state and

therefore higher BE as shown in Table 4.3.

Figure 4.6 Superposition image of Cyc derivatives (p- and m-chlorophenyl substituent)

with the reference structure in the mutant pfDHFR binding pocket. Cyc24, 25, 26, 27,

28, 29, 30, 31, and 42 (R2 is alkyl chain) in: (a) R configuration; (b) S configuration.

Cyc32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 43, 44, 45, and 46 (R2 is phenol chain) in: (c)

R configuration; (d) S configuration. Cyc derivatives and the reference structure are

shown as line model and stick model, respectively. Black, blue and green indicates

carbon, nitrogen, and chlorine atom, respectively.

Ref. code: 25605922040059HVU

29

Chapter 5

Conclusions and Recommendations

The theoretical investigation of enantioselective ligand-host binding

interactions in chemical and biological systems reveals that both the chiral hosts of

these two systems were capable of separating pure enantiomers from the racemic

mixture. In the chemical system, BCD binds more preferably towards (R)-nicotine than

the (S)-nicotine, due to the formation of hydrogen bonding. In the biological system,

both the wild-type and mutant pfDHFR are enantioselective towards Cyc derivatives

enantiomers because they cannot satisfy the lock and key mechanism, together with the

steric hindrance that one enantiomeric form of Cyc derivative experienced within the

pfDHFR binding pocket. The enantioselectivity of chiral hosts i.e. BCD and pfDHFR

from both systems demonstrated good activity in selective binding with one form of

enantiomer over another form. Both hosts are the potential chiral selectors to be used

in real environment for enantiomer separations. Therefore, theoretical investigation of

enantioselective ligand-host binding interaction by molecular modeling approach

proves to be a very powerful tool towards understanding the three-dimensional aspects

of chiral ligands-hosts binding interactions, which is very beneficial to the

experimentalist, not only in saving time and cost, but also provide them with the general

idea of selecting a good chiral selector.

For in-depth understanding of the molecular interactions of chiral hosts and

guests (or ligands) and to obtain the docking-based binding energies that are sufficiently

accurate to discriminate the preferred ligand stereochemistry, more accurate methods

for binding energy prediction as well as incorporating protein flexibility may be

required to improve the quality of the predicted binding energies. These could be done

by molecular dynamics simulations (MD) in the real aqueous environment.

Ref. code: 25605922040059HVU

30

References

1. Sekhon, B.S. (2013). Exploiting the power of stereochemistry in drugs: an

overview of racemic and enantiopure drugs. Journal of Modern Medicinal Chemistry,

1, 10–36.

2. Nguyen, L. A., He, H., & Pham-Huy, C. (2006). Chiral drugs: an overview.

International Journal of Biomedical Science: IJBS, 2(2), 85–100.

3. Maier, N. M., Franco, P., & Lindner, W. (2001). Separation of enantiomers:

needs, challenges, perspectives. Journal of Chromatography A, 906(1), 3-33.

4. McConathy, J., & Owens, M. J. (2003). Stereochemistry in drug action. Primary

Care Companion to The Journal of Clinical Psychiatry, 5(2), 70–73.

5. Kulatee, S., Toochinda, P., Suksangpanomrung, A., & Lawtrakul, L. (2017).

Theoretical investigation of the enantioselective complexations between pfDHFR and

cycloguanil derivatives. Scientia Pharmaceutica, 85(4), 37.

6. Chankvetadze, B., Lindner, W., & Scriba, G. K. E. (2004). Enantiomer

separations in capillary electrophoresis in the case of equal binding constants of the

enantiomers with a chiral selector: commentary on the feasibility of the concept.

Analytical Chemistry, 76(14), 4256-4260.

7. Keurentjes, J. T. F., Nabuurs, L. J. W. M., & Vegter, E. A. (1996). Liquid

membrane technology for the separation of racemic mixtures. Journal of Membrane

Science, 113(2), 351-360.

8. Buser, H. R., & Mueller, M. D. (1992). Enantiomer separation of chlordane

components and metabolites using chiral high-resolution gas chromatography and

detection by mass spectrometric techniques. Analytical Chemistry, 64(24), 3168-3175.

9. Collet, A., Brienne, M. J., & Jacques, J. (1980). Optical resolution by direct

crystallization of enantiomer mixtures. Chemical Reviews, 80(3), 215-230.

10. Pirkle, W. H., & Pochapsky, T. C. (1989). Considerations of chiral recognition

relevant to the liquid chromatography separation of enantiomers. Chemical Reviews,

89(2), 347-362.

Ref. code: 25605922040059HVU

31

11. Sansom, C. E., & Smith, C. A. (1998). Computer applications in the

biomolecular sciences. Part 1: Molecular modelling. Biochemical Education,

GaussView, version 5; Semichem Inc.: Shawnee Mission, KS, USA.

13. Frisch, M.J., Trucks, G.W., Schlegel, H.B., Scuseria, G.E., Robb, M.A.,

Cheeseman, J.R., Scalmani, G., Barone, V., Petersson, G.A., Nakatsuji, H., & et al.

(2016). Gaussian 09; Revision A.02; Gaussian, Inc.: Wallingford, CT, USA.

14. Accelrys (2016). Discovery Studio Modeling Environment Release 4.0;

Accelrys Inc.: San Diego, CA, USA.

15. Morris, G.M., Huey, R., Lindstrom, W., Sanner, M.F., Belew, R.K., Goodsell,

D.S., & Olson, A.J (2009). Autodock4 and AutoDockTools4: Automated docking with

selective receptor flexibility, Journal of Computational Chemistry. 16, 2785–2791.

16. Kurkov, S. V., & Loftsson, T. (2013). Cyclodextrins. International Journal of

Pharmaceutics, 453(1), 167-180.

17. Wren S., & et al. (2001) The Use of Cyclodextrins as Chiral Selectors, An

International Journal for Rapid Communication in Chromatography, Electrophoresis,

and Associated Techniques, 6, 59-77.

18. Gübitz, G., & Schmid, M.G (1997). Chiral separation principles in capillary

electrophoresis, Journal of Chromatography A, 792(1), 179–225.

19. Tait, R. J., Thompson, D. O., Stella, V. J., & Stobaugh, J. F. (1994). Sulfobutyl

ether beta-cyclodextrin as a chiral discriminator for use with capillary electrophoresis.

Analytical Chemistry, 66(22), 4013-4018.

20. Maruszak, W., Trojanowicz, M., Margasinska, M., & Engelhardt, H. (2001).

Application of carboxymethyl-beta-cyclodextrin as a chiral selector in capillary

electrophoresis for enantiomer separation of selected neurotransmitters. Journal of

Chromatography A, 926(2), 327-336.

21. Xiao, Y., & Chung, T.-S. (2007). Functionalization of cellulose dialysis

membranes for chiral separation using beta-cyclodextrin immobilization. Journal of

Membrane Science, 290(1), 78-85.

22. Liu, Y., Yu, H., Zhang, H., Yu, L., & Xu, W. (2017). Use of various beta-

cyclodextrin derivatives as chiral selectors for the enantiomeric separation of

higenamine by capillary electrophoresis. Microchemical Journal, 134, 289-294.

Ref. code: 25605922040059HVU

32

23. Ceborska, M., Szwed, K., & Suwinska, K. (2013). Beta-Cyclodextrin as the

suitable molecular container for isopulegol enantiomers. Carbohydrate Polymers,

97(2), 546-550.

24. Ghatee, M. H., & Sedghamiz, T. (2014). Chiral recognition of Propranolol

enantiomers by beta-Cyclodextrin: Quantum chemical calculation and molecular

dynamics simulation studies. Chemical Physics, 445, 5-13.

25. Alvira, E. (2015). Theoretical study of the separation of valine enantiomers by

beta-cyclodextrin with different solvents: a molecular mechanics and dynamics

simulation. Tetrahedron: Asymmetry, 26(15), 853-860.

26. Szabó, Z.-I., Tóth, G., Völgyi, G., Komjáti, B., Hancu, G., Szente, L., & et al.

(2016). Chiral separation of asenapine enantiomers by capillary electrophoresis and

characterization of cyclodextrin complexes by NMR spectroscopy, mass spectrometry

and molecular modeling. Journal of Pharmaceutical and Biomedical Analysis, 117,

398-404.

27. Alvira, E. (2017). Influence of solvent polarity on the separation of leucine

enantiomers by beta-cyclodextrin: a molecular mechanics and dynamics simulation.

Tetrahedron: Asymmetry, 28(10), 1414-1422.

28. Xuanping, T., Qin, L., Yizhong, S., Huan, W., Yanmei, Z., & Jidong, Y. (2015).

Chiral recognition of tyrosine enantiomers based on decreased resonance scattering

signals with silver nanoparticles as optical sensor. Chirality, 27(3), 194-198.

29. Mahalapbutr, P., Nutho, B., Wolschann, P., Chavasiri, W., Kungwan, N., &

Rungrotmongkol, T. (2018). Molecular insights into inclusion complexes of

mansonone E and H enantiomers with various beta-cyclodextrins. Journal of Molecular

Graphics and Modelling, 79, 72-80.

30. W., A. D., Xiande, W., & Nuran, E. (1998). Enantiomeric composition of

nicotine in smokeless tobacco, medicinal products, and commercial reagents. Chirality,

10(7), 587-591.

31. Ernst, M., Matochik, J. A., Heishman, S. J., Van Horn, J. D., Jons, P. H.,

Henningfield, J. E., & London, E. D. (2001). Effect of nicotine on brain activation

during performance of a working memory task. Proceedings of the National Academy

of Sciences of the United States of America, 98(8), 4728–4733.

Ref. code: 25605922040059HVU

33

32. Schnell, J. R., Dyson, H. J., & Wright, P. E. (2004). Structure, dynamics, and

catalytic function of Dihydrofolate Reductase. Annual Review of Biophysics and

Biomolecular Structure, 33(1), 119-140.

33. RCSB PDB-3UM8: Wild-Type Plasmodium falciparum DHFR-TS Complexed

with Cycloguanil and NADPH Structure Summary Page. Available online:

https://www.rcsb.org/pdb/explore.do?structureId=3UM8 (accessed on 30 May 2017).

34. RCSB PDB-3UM6: Double Mutant (A16V + S108T) Plasmodium falciparum

DHFR-TS (T9/94) Complexed with Cycloguanil, NADPH and dUMP Structure

Summary Page. Available online: https://www.rcsb.org/pdb/explore.do?structureId=

3UM6 (accessed on 30 May 2017).

35. Yuvaniyama, J., Chitnumsub, P., Kamchonwongpaisan, S., Vanichtanankul, J.,

Sirawaraporn, W., Taylor, P., & et al. (2003). Insights into antifolate resistance from

malarial DHFR-TS structures. Nature Structural Biology, 10, 357.

36. Yuthavong, Y., Vilaivan, T., Chareonsethakul, N., Kamchonwongpaisan, S.,

Sirawaraporn, W., Quarrell, R., & Lowe, G. (2000). Development of a lead inhibitor

for the A16V+S108T mutant of Dihydrofolate Reductase from the cycloguanil resis-

tant strain (T9/94) of Plasmodium falciparum†. Journal of Medicinal Chemistry,

43(14), 2738-2744.

37. Vanichtanankul, J., Taweechai, S., Uttamapinant, C., Chitnumsub, P., Vilaivan,

T., Yuthavong, Y., & Kamchonwongpaisan, S. (2012). Combined Spatial Limitation

around Residues 16 and 108 of Plasmodium falciparum Dihydrofolate Reductase

Explains Resistance to Cycloguanil. Antimicrobial Agents and Chemotherapy, 56(7),

3928–3935.

38. Kamchonwongpaisan, S., Quarrell, R., Charoensetakul, N., Ponsinet, R.,

Vilaivan, T., Vanichtanankul, J., & et al. (2004). Inhibitors of multiple mutants of

Plasmodium falciparum Dihydrofolate Reductase and their antimalarial activities.

Journal of Medicinal Chemistry, 47(3), 673-680.

39. CSD Entry-BCDEXD03: Beta-Cyclodextrin hydrate clathrate. Available

online:https://www.ccdc.cam.ac.uk/structures/search?sid=iucr&id=doi:10.1021%2Fja

00091a014 (accessed on 22 November, 2017).

Ref. code: 25605922040059HVU

34

40. Morris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, W. E., Belew,

R. K., & Olson, A. J. (1998). Automated docking using a Lamarckian genetic algorithm

and an empirical binding free energy function. Journal of Computational Chemistry,

19(14), 1639-1662.

41. Grabowski, S.J. (2006). Theoretical studies of strong hydrogen bonds. Annual

Reports Section “C” (Physical Chemistry), 102, 131–165.

42. Yuthavong, Y. (2002). Basis for antifolate action and resistance in malaria.

Microbes and Infection, 4(2), 175-182.

Ref. code: 25605922040059HVU

35

Appendices

Ref. code: 25605922040059HVU

36

Appendix A

List of abbreviations

Abbreviation Meaning

ACD

BCD

BE

CCDC

CDs

CE

Cyc

DHFR

GC

Alpha-cyclodextrin

Beta-cyclodextrin

Binding energy

Cambridge Crystallographic Data Center

Cyclodextrins

Capillary electrophoresis

Cycloguanil

Dihydrofolate reductase

Gas chromatography

GCD

HPLC

Gamma-cyclodextrin

High-performance liquid chromatography

Membrane Membrane technology

MM Molecular modeling

NMR

PDB

pfDHFR

Sensors

Nuclear magnetic resonance spectroscopy

Protein Data Bank

Plasmodium falciparum Dihydrofolate Reductase

Enantioselectivity sensing

TLC Thin-layered chromatography

TS Thymidylate synthase

Ref. code: 25605922040059HVU