university of groningen stereoselective synthesis of

University of Groningen

Stereoselective synthesis of glycerol-based lipidsFodran, Peter

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2015

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Fodran, P. (2015). Stereoselective synthesis of glycerol-based lipids. [S.n.].

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-amendment.

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 17-10-2021

https://research.rug.nl/en/publications/stereoselective-synthesis-of-glycerolbased-lipids(eb53c35c-a5c6-4d5a-bb42-df323c314cab).html

1

Stereoselective Synthesis of Glycerol-based

Lipids

Peter Fodran

The work described in this thesis was carried out at the Stratingh Institute for

Chemistry, University of Groningen, The Netherlands.

This work was financially supported by the Zernike Institute for Advanced Materials.

Printed by: Ipskamp Drukkers, Enschede

Cover design: Peter Fodran

3

Stereoselective Synthesis of

Glycerol-based Lipids

PhD thesis

to obtain the degree of PhD at the University of Groningen on the authority of the

Rector Magnificus Prof. E. Sterken and in accordance with

the decision by the College of Deans

This thesis will be defended in public on

Friday 13th March 2015 at 12:45

By

Peter Fodran

born on 12th June 1985

in Bratislava, Slovak Republic

Supervisor Prof. A. J. Minnaard Assessment committee Prof. M. H. Clausen Prof. F. J. Dekker Prof. J. G. Roelfes

5

How come it is not possible? People are flying to the moon and you want to tell me

that you cannot fix this (vacuum cleaner, stove, hairdryer…)?

(Iveta Fodranová)

7

CONTENTS

Chapter 1 An Introduction to Phospholipids 1 Introduction 2 Nomenclature 4 Biosynthesis of fatty acids, sphingolipids, triacylglycerols, and glycerophospholipids 6

Biosynthesis of fatty acids 6 Biosynthesis of sphingolipids 8 Biosynthesis of triacylglycerols and glycerophospholipids 10

Biosynthesis of non-archaeal ether based lipids 15 Outline of this thesis 16 References and footnotes 18

Chapter 2 Synthesis of Methyl-branched Fatty Acids 21 Introduction 22

Tuberculostearic acid 25 Results and discussion 27

(R)-Tuberculostearic acid 27 Caspofungin fatty acid 28

Conclusions 32 Experimental part 32 References and footnotes 42

Chapter 3 Catalytic Synthesis of Enantiopure Mixed Diacylglycerols 45 Introduction 46 Results and discussion 49

Synthesis of enantiopure phospholipids 49 Enantiopurity does not decrease during the ring opening 51 Synthesis of the platelet-activating factor (PAF) 52

Conclusions 53 Experimental section 53 References and footnotes 61

Chapter 4 Enantiopure Triacylglycerols in Three Steps 63 Introduction 64

Reported syntheses of triacylglycerols 64 Results and discussion 68

Co[(R,R)-salen] catalyzed ring opening of glycidyl esters 68 Towards the automated synthesis of triacylglycerols 70

Conclusions 75 Experimental part 75 References and footnotes 93

Chapter 5 A Methyl Matters 95 Introduction 96 Results and discussion 101

Chemical synthesis of the phospholipids 101 Formation of the liposomes and proteoliposomes 102 Molecular dynamics study of the bilayers 104 Calcein efflux assay 106

Conclusions and outlook 110 Experimental part 111 References and footnotes 121

Chapter 6 Synthesis of a Cyclooctyne–based Lipidation Probe 123 Introduction 124

Design of a lipophilic lipidation probe 127 Results and discussion 128 Conclusions and outlook 131 Experimental part 132 References 140

Chapter 7 A Missing Link in Archaeal Lipid Biosynthesis; a Contribution from Organic Synthesis 143

Introduction 144 Biosynthesis of archaeal membrane lipids 145 Archaeal lipids as taxanomic markes 149

Results and discussion 149 Synthesis of 2,3-bis-O-(geranylgeranyl)-sn-glycero-1-phosphate 149 Identification of CDP-archaeol synthase 153 Catalytic alcoholysis of benzylglycidol as a key step in the synthesis of cyclo-

archaeol and -glucosyl-cyclo-archaeol 154 Conclusion 157 Experimental part 157 References and footnotes 167

Summary 171 Samenvatting 175 Acknowledgements 180

1

Chapter 1 An Introduction to Phospholipids

Abstract: Phospholipids are compounds with enormous significance for Life. For a

long time, they were considered as passive building blocks of the membranes.

However with the discovery of the phospholipid signalling, understanding of their

roles changed. In the first part, this chapter introduces the reader to the

nomenclature of phospholipids. The second part of this chapter briefly summarizes

the biosynthetic pathways of various phospholipid classes and presents some of their

functions in living organisms. The last part of this chapter presents an outline of this

thesis.

Chapter 1

2

1

Introduction

“Lipids” is a loosely defined term for substances of biological origin, soluble

in non-polar solvents.1 Chemically, lipids can be divided into non-saponifiable and

saponifiable lipids. Steroids, prostaglandins and fat soluble vitamins comprise the

class of non-saponifiable lipids. Glycerolipids, phospholipids, sphingolipids and

waxes constitute the class of saponifiable lipids. An intriguing difference between

these classes is that while non-saponifiable lipids act mostly as single molecules,

saponifiable lipids mainly act as a collective. This can be illustrated by the following

examples. Retinal (vitamin A) is a non-saponifiable lipid. Its molecular properties

allow light-induced cis/trans isomerization which is essential for vision (Figure 1-I).

Figure 1. ( I ) Cis/trans isomerization as a principle of vision; ( II ) organization of a lipid

raft in a liquid ordered membrane; ( III ) examples of a saponifiable lipids which act as single

molecules.

An Introduction to Phospholipids

3

1

A mixture of saturated phospholipids, cholesterol and sphingolipids is a collective

(Figure 1-II) in the liquid ordered phase. In the membrane this collective forms an

organized lipid raft2 which is essential for signal transduction.3 As this classification

to saponifiable and non-saponifiable lipids is historical, it is easy to find exceptions.

For example, lipid 1 (Figure 1-III) is a typical membrane lipid surrounded by billions

of similar lipids (slightly) differing in the length of the fatty acids and the number of

the double bonds. However, in the entire collective of the membrane lipids, 1 is also

the lipid involved in inflammation processes.4 An example of a non-saponifiable lipid

fulfilling the role of a saponifiable lipid is the archaeal membrane lipid 2. The ether

bonds make 2 resistant to saponification, but the function which 2 fulfils is typical

for saponifiable lipids.

The main challenge in the studies of saponifiable lipids is their variability.

Terms like glycero-, phospho- and spingolipids frequently account for 10s to 1000s

related, but different molecular species. This variability is determined by the modular

structure of these lipids, described in Figure 2. In Nature, 40 common fatty acids

occur which differ in their chain length and the degree of unsaturation (Figure 2-I).

Triacylglycerols (Figure 2-II) are esters of glycerol and fatty acids. Given that the 3

hydroxyl groups can be esterified with any of the 40 fatty acids, the estimated number

of possible triacylglycerols approaches 64 000 (403). In the case of glycero-

phospholipids (Figure 2-II), one position of the glycerol is already occupied by any

of the 6 common phosphorus headgroups. The remaining 2 positions can again be

esterified by any of the 40 fatty acids resulting in up to 9 600 (6 x 402) different

species. Spingolipids can display even greater variability (Figure 2-III). The primary

hydroxyl group can carry either a phosphorous headgroup or a glycan core resulting

in more than 100 000 different species. The current knowledge of lipids is far away

from understanding the biological significance of this variability, but it has been

established that subtle deviations in the fatty acid composition of lipids can be linked

to heart5 and neurodegenerative diseases6 or metabolic syndrome.7

This chapter briefly introduces 3 topics that are important for this thesis. In

the first part, the reader is introduced to the nomenclature of the lipids. The second

part offers a brief overview of the biosynthesis of fatty acids, triacylglycerols and

phospholipids together with some of their biological properties. The last part of the

chapter presents the outline of this thesis.

Chapter 1

4

1

Figure 2. Variability of saponifiable lipids. ( I ) 40 common fatty acids; ( II ) variability in

glycerol based lipids; ( III ) variability in sphingolipids.

Nomenclature

Lipid research covers multiple fields of chemistry, biology and medicine.

Therefore it is not a surprise that a unified and universally applied nomenclature is

lacking. Despite the nomenclature for organic compounds is rigorously defined by

IUPAC8, this is happily ignored in lipid research. The following table (Table 1)


5

1

summarizes all the fatty acids that are mentioned in this thesis in all the common

nomenclatures.

Table 1. Names and symbols for fatty acids in this thesis.

Numerical

symbol

Structure

H3C-(hydrocarbon)-CO2H

Systematic name

(acid)

Trivial name

(acid)

4:0 -(CH2)2- butanoic butyric

6:0 -(CH2)4- hexanoic caproic

8:0 -(CH2)6- octanoic caprylic

10:0 -(CH2)8- decanoic capric

12:0 -(CH2)10- dodecanoic lauric

14:0 -(CH2)12- tetradecanoic myristic

16:0 -(CH2)14- hexadecanoic palmitic

18:0 -(CH2)16- octadecanoic stearic

18:1(11) -(CH2)7-CH=CH-(CH2)9- Z-9-octadecenoic oleic

18:2(9,12) -(CH2)5-(CH2CH=CH)2-(CH2)7- Z,Z-octadeca-9,12-

dienoic

linoleic

20:4(5,8,11,14) -(CH2)4-(CH2CH=CH)4-(CH2)3 Z,Z,Z,Z-eicosa-

5,8,11,14-tetraenoic

arachidonic

The description of the stereochemistry of the glycerol-based lipids might be

confusing. Given that glycerol is a prochiral compound, its substitution can lead to

a pair of enantiomers. Chemically, these are easily described using the Cahn-Ingold-

Prelog system (CIP). Although unambiguous, this nomenclature can obscure

biosynthetic relationships, for example in the case of triacylglycerols. Triacylglycerols

are biosynthesized by acylation of a diacylglycerol (Figure 3). An example below

(Figure 3-I) shows that depending on the length of the introduced fatty acid, the

corresponding triacylglycerols might have opposite configurational prefixes in CIP

system. In order to clearly present biological relationships, Hirschmann9 introduced

a stereospecific numbering (sn) system. This is based on the Fischer projection of the

substituted glycerol, placed in such a way that the secondary hydroxyl group points

to the left. The carbon on top is then designated as the sn-1 position and carbon on

the bottom sn-3. The advantage of the Hirschmann system is that a formal inversion

of the configuration is not possible. For comparison, the acylation of diacylglycerol

that was confusing in CIP (Figure 3-I) is now defined as an acylation on the sn-3

position (Figure 3-II).

Chapter 1

6

1

Figure 3. Comparison of 2 different systems for the stereochemical description of

glycerol-based lipids. ( I ) the CIP system commonly used in organic chemistry; ( II )

Hirschmanns system used in biology and biochemistry.

Biosynthesis of fatty acids, sphingolipids, triacylglycerols, and glycerophospholipids

Biosynthesis of fatty acids

Fatty acids are the building blocks of lipids. Their de novo synthesis is one of

the key metabolic pathways in living organisms. Chemically, this process is a

decarboxylative malonic ester synthesis of acyl coenzyme A with malonyl coenzyme

A (6) (Figure 4) followed by a deoxygenation. The synthesis of fatty acids starts with

a covalent attachment of acetyl coenzyme A 3 to the acyl carrying protein (ACP).

Malonyl coenzyme A (6) enters the cycle after covalent attachment to the acyl carrier

protein (ACP). The condensation of 4 and 6 results in thioaceto acetate 7 and

liberation of CO2.


7

1

Figure 4. De novo biosynthesis of fatty acids.

The ß-keto group is first reduced to alcohol 8 by NADPH/H+ (Figure 4), which is

subsequently dehydrated to α,ß-unsaturated 9. The conjugated double bond of 9 is

reduced by NADPH/H+ and the resulting 10 can enter the second cycle.

Alternatively, 10 (or its higher homologue) can either be hydrolysed to the

corresponding fatty acid 11 or transthioesterified to acyl coenzyme A 12. All steps in

the fatty acid synthesis are catalyzed by a fatty acid synthase (FAS). In Nature, there

are 2 types of FAS. FAS type 1 is present in animals and fungi and FAS type 2 is

found in bacteria and plants. The difference between the 2 types is that FAS type 1

is a single enzyme with 7 distinct domains and FAS type 2 is an assembly of 7

separable enzymes. A notable exception is the CMN group of bacterial species

(Corynebacterium, Mycobacterium, and Nocardia), which possesses both types of FAS.10

Desaturation of fatty acids

The fatty acid synthases tightly cooperate with desaturases that introduce

double bonds in the fatty acid chain. The most common desaturation is the

conversion of stearic acid to oleic acid by abstraction of 9-pro-R and 10-pro-R

Chapter 1

8

1

hydrogens. In human metabolism, this is catalyzed by 3 membrane-bound proteins

(Figure 5). The necessary electrons come from the electron transport chain, which

begins by reduction of reductase bound FAD (E-FAD) by NADH.

Figure 5. Δ9 desaturation of fatty acids.

The electrons are further transferred to cytochrome b5 and finally to the non-heme

Fe of the desaturase. Iron in its Fe2+ state can interact with O2 and oxidize 13 to 14.

The resulting oleoyl coenzyme A (14) can be further elongated or desaturated.11,12

Once the fatty acid has the desired length and unsaturation(s), it can enter

other metabolic pathways. This can be, for example, further modification of the fatty

acid (i.e. methylation as in Mycobacterium tuberculosis, see Chapter 2) or conversion into

sphingolipids, triacylglycerols and glycerophospholipids.

Biosynthesis of sphingolipids

The biosynthesis of fatty acids is tightly connected to the biosynthesis of

sphingolipids (derivatives of sphingosine (21)) via palmitoyl coenzyme A (15). The

biosynthesis of 21 (Figure 6) starts with a decarboxylative condensation of 15 and

serine (16).


9

1

Figure 6. Biosynthesis of sphingosine.

The resulting 17 is reduced to aminoalcohol 18. The nitrogen reacts with a fatty acid

coenzyme A and 19 is desaturated resulting in ceramide 20, which after hydrolysis of

the amide affords 21. Sphingosine (21) can be further modified (Figure 7) to

sphinholipids like cerebrosides 22, sphingomyelines 23 or gangliosides 24.

Figure 7. Examples of sphingolipids.

Chapter 1

10

1

Sphingolipids are responsible for diverse physiological functions. As the

membrane building blocks they are located at the outer leaflet of the phospholipid

bilayer.13 As signalling molecules14 sphingolipids are an important link between

overproduction of lipids and obesity.15

Biosynthesis of triacylglycerols and glycerophospholipids

The biosynthesis of triacylglycerols and glycerophospholipids is closely

related. Both pathways start with (R)-glycerol-1-phosphate (sn-glycerol-3-phosphate)

and share the same intermediates until the phosphatidic acid stage where the

pathways divide. First the biosynthesis of triacylglycerols is discussed.

Biosynthesis of triacylglycerols

The dominant route producing more than 90%16,17,18 of the triacylglycerols is

called the Kennedy pathway.19 In the endoplasmic reticulum, (R)-glycerol-1-

phosphate (sn-glycerol-3-phosphate) 25 (Figure 8) is esterified with a fatty acid

coenzyme A to form lysophosphatidic acid 26. In the next step, 27 is esterified with

a second fatty acid coenzyme A.

Figure 8. The Kennedy pathway.

The phosphate in the phosphatidic acid 27 is hydrolysed and the resulting

diacylglycerol 28 is esterified with a third fatty acid coenzyme A to afford the


11

1

triacylglycerol 29. Subsequently, triacylglycerols can be stored in a specialized

organelle (a lipid droplet20) where they serve as energy reserve and precursors of

other lipid products.

Figure 9. ( I ) accumulation of triacylglycerols and fatty acids between the membrane

leaflets; ( II ) budding of a lipid droplet; ( III ) mature lipid droplet.

The mechanism of formation of the lipid droplets is poorly understood, but

a generally accepted theory states that these are formed by budding of the

endoplasmic reticulum (Figure 9)21 as a response to an elevated triacylglycerol

synthesis.22 Initially, the synthetized triacylglycerols concentrate between the leaflets

of the membrane (Figure 9-I). With the increasing amount of triacylglycerols, the

bud grows collecting more and more triacylglycerols (Figure 9-II). Finally, the lipid

droplet forms (Figure 9-III) as an independent organelle that can move into the

cytosol and interact with other organelles. Alternative mechanisms for the formation

of lipid droplets have been proposed by Ploegh23 and Walter and Farase.24

The content of the lipid droplets can be utilized when needed. The main

mechanism of utilization of the stored triacylglycerols and sterol esters is lipolysis.

Adipose triglyceride lipase and hormone sensitive lipase are moved to the surface of

the lipid droplet. The first enzyme hydrolyses at the sn-2 position of the

triacylglycerol. Hormone sensitive lipase further hydrolyses the 1,3-diacylglycerol to

a monoacylglycerol. The hydrolysis of the final fatty acid occurs in the cytosol and is

catalyzed by a monoacylglycerol lipase.25 The products of the hydrolysis of

triacylglycerols from the lipid droplets might be utilized for example in the

biosynthesis of phospholipids.

Chapter 1

12

1

As it was already mentioned, the biosynthesis of phospholipids is closely

related to the biosynthesis of triacylglycerols. The phosphatidic acid can be converted

to any of the 6 common phospholipids by 2 mechanisms. The first mechanism

involves cytidine triphosphate (CTP) activation of the phosphatidic acid leading to

phosphatidylinositols (PI), phosphatidylglycerols (PG) and cardiolipins. The second

mechanism utilizes CTP activation of the headgroup precursors leading to

phosphatidylcholines (PC), phosphatidylethanolamines (PE) and

phosphatidylserines (PS). The biosyntheses are presented in this order.

The activation of phosphatidic acid 27 (Figure 10) by CTP results in the cytidine

diphosphate (CDP) activated diacylglycerol 30 and liberation of diphosphate (PPi)

CDP activated 30 can react with inositol (31) resulting in phosphatidylinositol 32 and

cytidine monophosphate (CMP).

Figure 10. Biosynthesis of phospholipids via CTP activation of 27.


13

1

Alternatively, the CDP activated 30 (Figure 10) can react with 25, which after

hydrolysis of phosphate results in phosphatidylglycerol 34 type lipid. 34 can react

with an additional molecule of CDP-diacylglycerol, affording cardiolipin 35.26

The phosphatidyl inositol 32, phosphatidyl glycerol 34 and cardiolipin 35

families of lipids fulfil various important functions in the entire cellular life. For

example the PI lipids anchor membrane proteins to the outer leaflet of the

membrane (via protein lipidation, see chapter 6). Another important role of the PI

lipids is in signal transduction in the plant and the animal kingdom via the action of

a specific phospholipase C.27 By this mechanism, the PI lipids influence the activity

of dozens of enzymes belonging to the protein kinase C family, thus controlling key

cellular functions like differentiation, proliferation, metabolism and apoptosis. The

PG lipids serve as precursors for the cardiolipins. Cardiolipins are mainly found in

the mitochondrial membrane, where they bind and regulate the activity of various

proteins.28 Abnormalities in the cardiolipin metabolism can be linked to a variety of

diseases, including Barth syndrome29, Parkinson, Alzheimer30 and Tengier disease.31

The mentioned functions of the families of lipids (PI, PG and cardiolipins) form

only a fraction of what has been reported.

The phosphatidic acid 27 can be transformed into PC, PE and PS

phospholipids via the second mechanism involving activation of the headgroup

precursor by CTP. This pathway starts with hydrolysis of 27 to diacylglycerol 28

(Figure 11-I). The CDP-phosphorylating agents 38 and 39 are synthetized in the

cytosol by the same mechanism (Figure 11-II). The corresponding alcohols 36 are

phosphorylated with ATP resulting in phosphates 37. These react with CTP leading

to the phosphorylating agents 38 and 39, which are further transported to the

endoplasmic reticulum, where they phosphorylate diacylglycerol 28. Phosphorylation

of 28 with CDP-ethanolamine 38 results in phosphatidylethanolamine type lipid 40

and phosphorylation of 27 with 39 results in phosphatidylcholine lipid type 41. Both

40 and 41 can be further converted to phosphatidylserine 42 type lipids. And finally,

42 can be converted back to 40 by decarboxylation.

PC, PE and PS lipids are the main building blocks of biological membranes.

PC is the most common lipid in animals and plants where it constitutes up to 50%

of all phospholipids. In bacteria, PC lipids are scarcer. Due to their molecular shape,

PC, PE and PS lipids have their preferred location in the membranes. PC lipids are

mainly located in the outer leaflet while PE and PS are located in the inner leaflet.

Distribution of the lipids between the leaflets is tightly regulated by enzymes –

flippases. However, in some events the distribution of the membrane lipids is altered.

Chapter 1

14

1

For example during apoptosis, PS lipids are moved to the outer leaflet, where they

are recognized by macrophages. By this mechanism the apoptic cell is removed

without triggering an inflamation.32

Figure 11. ( I ) Biosynthesis of phospholipids via CTP activation of the headgroup

precursors; ( II ) biosynthesis of the CDP activated headgroup precursors 38 and 39.


15

1

Biosynthesis of non-archaeal ether based lipids

Plasmologens are ether analogues of the PE and PC lipids. Despite being

structurally related, their biosynthesis requires a specific pathway (Figure 12-I).

Figure 12. ( I ) Biosynthesis of plasmalogens; ( II ) mechanism of the substitution of acyl

for a long chain alcohol as the key step in the biosynthesis of plasmalogens.

Chapter 1

16

1

The biosynthesis starts in the peroxisome, by acylation of dihydroxyacetone

phosphate (43). In the second step the carboxylate is substituted by a long-chain

alcohol resulting in 45. The mechanism of this step was elucidated by Brown and

Snyder (Figure 12-II)33. In the active site of the alkylglycerone phosphate synthase,

44 tautomerizes to 46, which after protonation leads to 47. The resulting carbocation

is attacked by a nucleophilic centre Nu of the protein (probably an amino group in

the active site) resulting in departure of the carboxylate. In a subsequent step 49

reacts with a long-chain alcohol. 50 undergoes an E1 type elimination leading to 51

which finally tautomerizes to ketone 45. Reduction of 45 (Figure 12-I) results in 53

which is acylated in the endoplasmic reticulum. From 57 on, the biosynthesis is

similar to the synthesis of PC or PE lipids. First the phosphate is hydrolyzed and

resulting 55 is phosphorylated by CDP activated choline or ethanolamine. In case of

the choline headgroup, the biosynthesis stops at this point. The lipids with

ethanolamine headgroup can be further desaturated to 58.

The biological functions of plasmalogens are still not fully understood.

Structurally, they help to maintain physical properties of the membranes.34 Broniec

et al.35 reported that the analogues of 56 act as scavengers of reactive oxygen

suggesting that they play a role in oxidative stress. An important plasmalogen is the

platelet activating factor (Figure 12-I), which is an extremely potent signalling

molecule triggering the platelet aggregation and immunological responses at pM

concentrations (10-11 M).36 An efficient synthesis of PAF is described in Chapter 3.

Outline of this thesis

Lipids play vital roles in many processes essential for life. This is illustrated

by a lipid membrane, which is a complex mixture of (phospho)lipids with various

chain lengths and degree of unsaturation. In this complex mixture, every single

component has an irreplaceable role. Of course, lipids are in principle accessible from

their natural sources, but their isolation and purification (from other lipids) is tedious

and often virtually impossible. A convincing example in this connection starts with

the impressive contribution of R. J. Anderson in 1927,37 who was the first to isolate

and describe tuberculostearic acid 59. For his studies, he needed 2 200 culture flasks

with a volume of 200 cm3. Only in 2010, 83 years later, it was established beyond

reasonable doubt that 59 is part of phospholipid 60 in M. tuberculosis.38 From 1 g of a

total lipid extract of M. tuberculosis, the authors isolated 50 μg of pure lipid 60, and

determined its structure by independent synthesis. For further illustration, 1 g of the

total lipid extract corresponds roughly to 20 g of bacteria.


17

1

Biology has a lot to gain from the availability of pure, well-defined, natural

and unnatural lipids in sufficient amounts, and organic chemistry can fulfil this need.

This is realized and illustrated in this thesis. In 7 chapters, novel, efficient, and

stereoselective approaches are described for the synthesis of ester-based and ether-

based phospholipids and triacylglycerols.

Chapter 2 describes the catalytic asymmetric synthesis of methyl-branched

fatty acids (59 in Figure 13). The approach is based on conjugate addition of

methylmagnesium bromide to α,ß-unsaturated thioesters and subsequent chain

elongation to the desired length by Wittig reaction with a functionalized ylide. This

modular approach is applied in the synthesis of the fatty acid chain of caspofungin,

which allowed a study in the group of Prof. R. M. J. Liskamp (Molecular Medicinal

Chemistry, University of Utrecht) on the influence of the stereochemistry of this

fatty acid on its antifungal properties.

Figure 13. Examples of a fatty acid and lipid isolated from natural sources.

The theme of Chapter 3 is the transformation of fatty acids into

phospholipids. Here, the Jacobsen Co(II) salen complexes play an important role,

granting the regiospecific opening of protected glycidol with fatty acids. The chapter

further describes a migration-free deprotection of the resulting silylated

diacylglycerols, solving a long-standing problem in this field. It allows the synthesis

of various glycerophospholipids. A small modification of the catalyst opens a

convenient access to mixed ether/ester lipids represented by platelet activating

factor.

Chapter 4 is an extension of this methodology to the synthesis of enantiopure

triacylglycerols in just 3 synthetic steps. This allows the preparation of a small (>15)

Chapter 1

18

1

library of triacylglycerols, as a prelude to the determination of the composition of

(cow) milk fat, a piece de resistance in diary research.

Chapter 5 describes the influence of phospholipids on the function of

mechanosensitive channels of large conductance (MscL). In particular, the role of

methyl-branched lipid 60 on the MscL from the same species is studied and related

to their non-branched analogues.

Chapter 6 describes the synthesis of a fatty acid equipped with a strained

cyclooctyne. This “clickable fatty acid” is a promising tool for further studies in

chemical biology.

Chapter 7, composed of 2 parts, is dedicated to the synthesis of ether-based

Archaea lipids. In this chapter, the introduction is dedicated to the metabolism of

the unique Archaea lipids. Part one describes the synthesis of an intermediate in

Archaea lipid biosynthesis. This lipid has been used in the department of Molecular

Microbiology (GBB, Prof. A. J. M. Driessen) for the identification of CDP-archaeol

synthase, the missing link in this biosynthesis. The second part describes the

application of the aforementioned Co(II) salen complexes in a total synthesis of

cyclo-archaeol.

References and footnotes

(1) Moss, G. P.; Smith, P. A. S.; Tavernier, D. Pure Appl. Chem. 1995, 67, 1307.

(2) Pike, L. J. J. Lipid Res. 2003, 44, 655.

(3) Simons, K.; Toomre, D. Nat. Rev. Mol. Cell Biol. 2000, 1, 31.

(4) Fernandis, A. Z.; Wenk, M. R. Curr. Opin. Lipidol. 2007, 18, 121.

(5) Beilin, L. J.; Burke, V.; Puddey, I. B.; Mori, T. A.; Hodgson, J. M. Clin. Exp.

Pharmacol. Physiol. 2001, 28, 1078.

(6) Han, X. Front. Biosci. 2007, 12, 2601.

(7) Carpentier, Y. A.; Portois, L.; Malaisse, W. J. Am. J. Clin. Nutr. 2006, 83, S1499.

(8) Eur. J. Biochem. 1977, 79, 11.

(9) Hirschmann, H. J. Biol. Chem. 1960, 235, 2762.

(10) Gebhardt, H.; Meniche, X.; Tropis, M.; Krämer, R.; Daffé, M.; Morbach, S.

Microbiology 2007, 153, 1424.

(11) Nakamura, M. T.; Nara, T. Y. Annu. Rev. Nutr. 2004, 24, 345.

(12) Qiu, X. Prostaglandins Leukot. Essent. Fatty Acids 2003, 68, 181.

(13) Berridge, M. J. Nature 1993, 361, 315.

(14) Spiegel, S.; Milstien, S. J. Biol. Chem. 2002, 277, 25851.

(15) Summers, S. A. Prog. Lipid Res. 2006, 45, 42.

(16) Lehner, R.; Kuksis, A. J. Biol. Chem. 1993, 268, 8781.


19

1

(17) For other biosynthetic pathways see ref. 18 and 19.

(18) (a) Cagliari, A.; Margis, R.; Dos, S. M. F.; Turchetto-Zolet, A. C.; Loss, G.; Margis-

Pinheiro, M. Int. J. Plant Biol. 2011, 2, 40 (b) Coleman, R. A.; Lee, D. P. Prog. Lipid Res. 2004,

43, 134 (c) Karantonis, H. C.; Nomikos, T.; Demopoulos, C. A. Curr. Drug Targets 2009, 10,

302 (d) Lehner, R.; Kuksis, A. Prog. Lipid Res. 1996, 35, 169 (e) Lehner, R.; Kuksis, A. Prog.

Lipid Res. 1996, 35, 169 (f) Sorger, D.; Daum, G. Appl. Microbiol. Biotechnol. 2003, 61, 289 (g)

Sorger, D.; Daum, G. Appl. Microbiol. Biotechnol. 2003, 61, 289 (h) Yen, C.-L. E.; Stone, S. J.;

Koliwad, S.; Harris, C.; Farese, R. V., Jr. J. Lipid Res. 2008, 49, 2283 (i) Coleman, R. A.;

Mashek, D. G. Chem. Rev. 2011, 111, 6359.

(19) Weiss, S. B.; Kennedy, E. P. J. Am. Chem. Soc. 1956, 78, 3550.

(20) the alternative terms describing same organelle are: lipid bodies, oil bodies and

adiposomes

(21) Martin, S.; Parton, R. G. Nat. Rev. Mol. Cell Biol. 2006, 7, 373.

(22) Pol, A.; Martin, S.; Fernandez, M. A.; Ferguson, C.; Carozzi, A.; Luetterforst, R.;

Enrich, C.; Parton, R. G. Mol. Biol. Cell 2004, 15, 99.

(23) Ploegh, H. L. Nature 2007, 448, 435.

(24) Walther, T. C.; Farese Jr, R. V. Biochim. Biophys. Acta. - Mol. Cell Biol. L. 2009, 1791,

459.

(25) Guo, Y.; Cordes, K. R.; Farese, R. V.; Walther, T. C. J. Cell Sci. 2009, 122, 749.

(26) the biosynthesis of cardiolipins differs in prokaryotic and eucaryotic cells. The

depicted sequence corresponds to the prokaryotic cells.

(27) Irvine, R. F. Curr. Opin. Cell Biol. 1992, 4, 212.

(28) Haines, T. H. Biochim. Biophys Acta - Biomembranes 2009, 1788, 1997.

(29) Xu, Y.; Malhotra, A.; Ren, M.; Schlame, M. J. Biol. Chem. 2006, 281, 39217.

(30) Ruggiero, F. M.; Cafagna, F.; Petruzzella, V.; Gadaleta, M. N.; Quagliariello, E. J.

Neurochem. 1992, 59, 487.

(31) Oram, J. F. Biochim. Biophys. Acta - Mol. Cell Biol. L. 2000, 1529, 321.

(32) Verhoven, B.; Schlegel, R. A.; Williamson, P. J. Exp. Med. 1995, 182, 1597.

(33) Brown, A. J.; Snyder, F. J. Biol. Chem. 1983, 258, 4184.

(34) Farooqui, A. A.; Horrocks, L. A.; Farooqui, T. Chem. Phys. Lipids 2000, 106, 1.

(35) Broniec, A.; Klosinski, R.; Pawlak, A.; Wrona-Krol, M.; Thompson, D.; Sarna, T.

Free Radic. Biol. Med. 2011, 50, 892.

(36) Prescott, S. M.; Zimmerman, G. A.; Stafforini, D. M.; McIntyre, T. M. Annu. Rev.

Biochem. 2000, 69, 419.

(37) Anderson, R. J. J. Biol. Chem. 1927, 74, 525.

(38) ter Horst, B.; Seshadri, C.; Sweet, L.; Young, D. C.; Feringa, B. L.; Moody, D. B.;

Minnaard, A. J. J. Lipid Res. 2010, 51, 1017.

Chapter 1

20

1

21

Chapter 2 Synthesis of Methyl-branched Fatty Acids

Abstract: Branched-chain fatty acids are common in yeast and bacteria, where they

fulfil diverse functions. Their isolation from natural sources is lengthy and tedious.

This chapter presents an efficient and modular synthesis of methyl branched fatty

acids.

Parts of this chapter have been published:

Mulder, M. P. C.; Fodran P.; Kemmink, J.; Breukink, E. J.; Kruijtzer, J. A. W.;

Minnaard, A. J.; Liskamp, R. M. J. Org. Biomol. Chem. 2012, 10, 7491.

Fodran, P.; Minnaard, A. J. Org. Biomol. Chem. 2013, 11, 6919.

Chapter 2

22

2

Introduction

Methyl-branched fatty acids (a subset of branched-chain fatty acids) are

found at many places in Nature,1 although they are much less abundant than their

straight-chain congeners. They can be divided into several categories, but for the

purpose of this chapter a straightforward division based on the position of the

methyl-branch in the linear chain is sufficient. The most common pattern is a mono-

methyl branch, but also poly-methyl branched fatty acids occur.1b

Fatty acids can branch next to the carboxylic acid group - at the beginning of

the chain. Mycocerosic acids (Figure 1) 1 which are components of the Mycobacterium

tuberculosis cell wall are a prominent example. The biosynthesis2 of 1 is analogous to

the biosynthesis of fatty acids (Chapter 1). The long-chain acyl coenzyme A (2) is

initially extended by methylmalonyl coenzyme A, then decarboxylated and finally

deoxygenated. Additional cycle(s) lead to a multi-methyl branched 1. Remarkably,

the same M. tuberculosis produces the structurally very similar phthioceranic acid (3)

which has the opposite configuration of the methyl substituents.

Figure 1. Biosynthesis of fatty acids branched at the beginning of the chain.

Based on the abovementioned division, fatty acids can also branch in the

middle of the chain. Probably the best known example from this group is (R)-

tuberculostearic acid (4) (Figure 2-I). 4 is an important cytoplasmic membrane

component of M. tuberculosis. The biosynthesis3 (Figure 2-II) involves methylation of

the 10th carbon of oleolate 5 by S-adenosylmethionine (SAM). A subsequent Wagner-

Meerwein rearrangement of 6 results in a thermodynamically more stable tertiary

carbocation 7, which after deprotonation affords non-Zaitsev olefin 8. The final step

is the reduction of the methylene in 8 to the (R)-tuberculostearic acid carrying

phospholipid 9.

Synthesis of Methyl-branched Fatty Acids

23

2

Figure 2. ( I ) Structure of (R)-tuberculostearic acid (9); ( II ) biosynthesis of 9.

A third class of branched fatty acids carries the methyl-branch at the terminus of the

chain4 (Figure 3). They are biosynthetized in the same manner as linear fatty acids.

Figure 3. Biosynthesis of fatty acids branched at the terminus of the chain.

Chapter 2

24

2

The only difference is the primer, which determines the position of the

branch and the odd/even number of carbons in the chain. These primers, isobutyryl-

(10) isovarelyl- (11) and (S)-2-methylbutyryl-coenzyme A (12) are products of the

catabolism of L-valine, L-leucine and L-isoleucine respectively. Important to note is

that while iso-branched fatty acids 13 and 14 are achiral, the anteiso-branched acids 15

are chiral and enantiopure.

A particular example of a fatty acid branched at the end of the chain is 16

(Figure 4-I). 16 forms a lipophilic portion of lipopeptide pneumocandin B0 (17),

which is found in the fungus Zalerion arboricola. Chemical modification of 17

(Figure 4-II), leads to the better known caspofungin (18).5 18 together with

anidulafungin and micafungin6 are approved drugs of the enchinocandin class of

antifungals.

Figure 4. ( I ) An example of a fatty acid branched at the terminus of the chain; ( II ) antifugal

agents 17 and 18 originating from Zalerion arboricola.

Enchinocandins are modern non-competitive 1,3-ß-glucan synthase inhibitors, used

for the treatment of candidiasis and aspergillosis (also in immunocompromised

patients). Their advantages are low toxicity and a high antifungal activity. Their

pharmacological properties are dependent on the fatty acid component.6

A common denominator of all above-mentioned branched acids is their

relevance in medicine. For example, O’Sullivan et al. 7 reported that detection of

mycocerosic acid (1) in sputum can be used for the rapid detection of tuberculosis.


25

2

Similarly, French et al.8 developed a method for diagnostics of tuberculous

meningitis based on the presence of tuberculostearic acid (4) in the cerebrospinal

fluid. The importance of branched fatty acids for medicine is not limited to

tuberculosis research, but extents also to cardiac9 and pediatric10 studies. A

procedure allowing an efficient synthesis of methyl-branched fatty acids, with the

methyl branch at any desired position in the chain would be a great aid in further

research on the role of methyl branching in fatty acids. To develop such a method,

(R)-tuberculostearic acid (4) served as prototype branched fatty acid.

Tuberculostearic acid

Since its first description in 1927, tuberculostearic acid has been synthetized

by multiple researches either in racemic or enantiopure form.11 Furthermore, the

natural enantiomer has been prepared starting from the chiral pool, using a chiral

auxiliary, and recently, also by enantioselective catalysis.11f The approach using

enantioselective catalysis was reported by Ter Horst et al.11f and utilizes an

enantioselective conjugate addition to an α,ß-unsaturated thioester as a key-step for

the introduction of the methyl branch (Scheme 1). The authors prepared thioester

22 by cross-metathesis of 19 (prepared from commercially available 10-undecenoic

acid) and 20 (prepared in 3 steps).

Scheme 1. Synthesis of 4 according to Ter Horst et al.11f

Chapter 2

26

2

Conjugate addition to 22 afforded 23 in an excellent 91% yield and a good 95:5 e.r.

(Scheme 1). A chemoselective reduction of the thioester in the presence of an oxo-

ester afforded the aldehyde and subsequent Wittig reaction afforded 25 in 92% yield

over 2 steps. Finally, hydrogenation using organocatalytically12 generated diimide and

subsequent alkaline hydrolysis of the isopropyl ester afforded (R)-tuberculostearic

acid 4 in 91% yield (over 2 steps). Although at that time the highest yielding synthesis,

it takes 7 linear steps starting from commercially available compounds.

A retrosynthetic analysis (Figure 5) revealed that the synthesis of 4 can in

principle be carried out in 5 linear steps. A disconnection between the 7th and 8th

carbon of 4 gives 27 and 28. Phosphonium salt 27 can be prepared in one step from

commercially available 7-bromoheptanoic acid.13

Figure 5. Retrosynthetic analysis of 4.

Aldehyde 28 can be obtained by enantioselective conjugate addition to α,ß-

unsaturated undecenoate14 followed by a reduction. From a broad spectrum of

suitable enoates,15 α,ß-unsaturated thioesters remain the best option as these afford

products of the conjugate addition in high yields and optical purity. Furthermore,

thioesters offer a better control of the reduction step compared to oxo-esters,16

therefore 29 can be envisioned as a suitable substrate. Finally, 29 is available via a

Wittig or Horner-Wadsworth-Emmons reaction of commercially available nonanal

(30). Alternatively, in analogy to Ter Horst, 29 can be prepared via cross-metathesis

of 20 and decene (31).17 On first sight, the cross-metathesis might seem more


27

2

appealing especially from an atom economy point of view. This is however just

appearance as 20 is prepared by a Wittig reaction with paraformaldehyde.

The versatility of this approach will be illustrated with the synthesis of

tuberculostearic acid (4) and the caspofungin side chain 16. Its flexibility becomes

clear in particular because 16 (Figure 6) bears the methyl groups at different positions

of the chain compared to 4 and their absolute stereochemistry is opposite, this

underscores the need for a catalytic enantioselective approach.


Another aspect of this chapter is a study of the influence of 16 on the antifungal

activity of caspofungin and its derivatives.

Results and discussion

(R)-Tuberculostearic acid

According to the retrosynthetic analysis in Figure 5, (R)-tuberculostearic acid

(4) was prepared in 5 linear steps. Unsaturated thioester 29 (Scheme 2) was prepared

by treatment of nonanal (30) with an excess of stabilized Wittig reagent 3318 in the

presence of LiCl (20 mol%) using “on water“ conditions. The olefination proceeds

smoothly under these conditions resulting in an excellent E : Z ratio (>>95 : 5).

Claridge et al.19 reported similar selectivity in a related Horner-Wadsworth-Emmons

reaction using MeMgBr as a base. The on water procedure is advantageous as it does

not require inert atmosphere or exclusion of water. After the olefination, the trace

amount of undesired (Z)-29 (Scheme 2) was removed by flash chromatography and

(E)-29 was isolated in 92% yield. Next, 29 was subjected to a CuBr.(34)14,20 catalyzed

conjugate addition of MeMgBr. This afforded the branched product 35 in 94% yield

and 95 : 5 e.r. Thioester 35 was reduced with DIBAL to afford 28, which could be

subjected to Wittig reaction without any purification. Although multiple authors have

applied acid-functionalized Wittig reagents in their syntheses,21 their reported

Chapter 2

28

2

conditions resulted in low yields of 36. After optimization, a combination of an

excess of 27 in combination with LiHMDS, were the best conditions affording 36 in

79% yield over 2 steps.

Reagents and conditions: a) 33 (1.4 equiv), LiCl (20 mol%), water, RT, 18 h; b) CuBr.SMe2 (1.5 mol%), 34 (1.65 mol%), MeMgBr (1.2 equiv), tBuOMe, –78 °C, 3 h addition of 29 followed by stirring for 16 h; c) DIBAL (1.3 equiv), CH2Cl2, -78 °C, 2 h; d) 27 (1.4 equiv), LiHMDS (2.8 equiv), THF, 0 °C – 21 °C, 3 h; e) 26 (5 mol%), NH2NH2.H2O (21.0 equiv), EtOH, O2 (balloon), RT, 24 h.

Scheme 2. Second generation synthesis of tuberculostearic acid.

To avoid racemization of the homoallylic methyl-branched stereocenter in 36, the

double bond was reduced with diimide, produced by controlled oxidation of

hydrazine by O2 in the presence of 26.12 In this way, (R)-tuberculostearic acid (4) was

prepared in 63% overall yield, the shortest and highest yielding route to date.

Caspofungin fatty acid

The relative and absolute configuration of 16 (Scheme 3) was reported by

Leonard et al.22 The authors initially applied the Enders auxiliary method23 to prepare

ent-16 in 8 steps as a 90:10 (syn : anti) ratio of diastereomers. Later, using prolinol as

a chiral auxiliary,24 they synthetized the correct enantiomer as an 80 : 20 (syn : anti)

mixture of diastereomers also in an 8 step sequence.

The same strategy as in (R)-tuberculostearic acid (Scheme 2) could be applied

to the synthesis of caspofungin fatty acid 16 and its monomethyl analogue 46


29

2

(Scheme 3). Commercially available (E)-pent-2-enoic acid 37 was converted into

thioester 32 using Steglich conditions.

Reagents and conditions: a) EtSH (2.0 equiv), DCC (1.1 equiv), DMAP (10 mol%), CH2Cl2, 0 - 21 °C, 3 h; b) MeMgBr (1.2 equiv), CuBr.Me2S (1.2 mol%), 25 (1.3 mol%), tBuOMe, –78 °C, 3 h addition of 32 followed by stirring for 16 h; c) DIBAL (1.2 equiv), CH2Cl2, –50 °C, 1 h; d) 40 (1.5 equiv), nBuLi (1.1 equiv), THF, 0-21 °C, 16 h; e) MeMgBr (1.3 equiv), CuBr.25 (1.3 mol%), tBuOMe, –78 °C, 3 h addition of 41 followed by stirring for 16 h; f) DIBAL, (1.2 equiv), CH2Cl2, -50 °C, 1 h; g) 27 (1.75 equiv), LiHMDS (2.0 equiv), THF, 0 °C – 21 °C, 3 h; h) NH2NH2.H2O (25.0 equiv), 26 (10 mol%), EtOH, O2 (balloon), 21 °C; i) 44 (1.7 equiv), LiHMDS (2.0 equiv), THF, 0 °C – 21 °C, 3 h; j) NH2NH2.H2O (30.0 equiv), 26 (10 mol%), O2, (balloon), EtOH, 21 °C.

Scheme 3. Synthesis of 15 and monomethyl analogue 46.

Distillation of the crude mixture under reduced pressure afforded pure 32 in

95%. Thioester 32 underwent CuBr.(S,Rp)-Josiphos (CuBr.25) catalyzed

enantioselective conjugate addition of MeMgBr smoothly, affording the desired 38

in 92% yield with a 98.5 : 1.5 e.r. This branched thioester was a key precursor in the

Chapter 2

30

2

synthesis of both 16 and 46. To achieve the caspofungin fatty acid 16, thioester 38

was reduced with DIBAL. After work-up, the aldehyde 39 was treated with Horner-

Wadworth-Emmons reagent 40 in basic conditions to afford 41 in 70% yield over 2

steps. After separation of the minor Z stereoisomer 41 was subjected to the second

conjugate addition resulting in a 42 in a 20 : 1 diastereomeric ratio. Flash

chromatography afforded diastereomerically pure 42 in 78% yield. This was reduced

with DIBAL and the corresponding aldehyde underwent Wittig reaction with 27 to

extend the chain to the desired length. Acid 43 was isolated in 47% over 2 steps. The

double bond in 43 was reduced with diimide, in situ formed by catalytic oxidation of

hydrazine. The caspofungin fatty acid 16 was isolated in 81% yield.

The synthesis of the mono methyl analogue 46 (Scheme 3) was performed in

a similar manner. The key thioester 38 was reduced by DIBAL and extended using

44. These 2 steps afforded 45 in 60% yield. The reduction of the resulting double

bond by in situ generated diimide resulted in monomethyl branched 46 in 80% yield.

The lower yields compared to the synthesis of tuberculostearic acid (Scheme

2) are caused by the volatility of the intermediates. Compared to the previously

published approach22 the present synthesis of the caspofungin side chain is higher

yielding (20% vs. 9% by Leonard et al.). The biggest advantage arises from the

iterative protocol, which yields 42 as 20:1 ratio of epimers, which are fully separable.

In the synthesis published by Leonard et al. the mismatching pair of chiral enolate

and chiral alkylating agent results in only a 4 : 1 ratio of epimers. The authors needed

2 recrystallizations of the fatty acid (as the corresponding cinchonidine salt) to

improve the diastereometic ratio to 97:3.

Influence of different branched fatty acids on the antifungal activity of caspofungine analogues

The following part of this research has been conducted by Dr. Mulder, from

the group of Prof. Liskamp, University of Utrecht.

The core of caspofungin (18) (Figure 7) consists of 6 amino acids. Given that

some of these amino acids are not commercially available, the effect of the branched

fatty acids on the antifungal activity of caspofungin was studied on the simplified

hexapeptide. In the simplified analogue (red in Figure 7) 3,4-dihydroxy

homotyrosine, 3-hydroxy ornithine and its diaminoethyl analogue were substituted

for homotyrosine and ornithine.


31

2

Figure 7. Caspofungine and its analogues used in the study.

Caspofungine 18 and the analogues 47, 48 and 49 were then tested against a

panel of common Candida species in a broth microdilution assay (Table 1). The

results are expressed as minimum inhibitory concentration (MIC) values – the

minimum concentration of a compound which completely inhibits visible growth.

Table 1. Antifungal activity of caspofungin and its analogues.

species

derivative

C. albicans

CBS9975

(μg.ml-1)

C. dubliniensis

CBS7987

(μg.ml-1)

C. tropicalis

CBS94

(μg.ml-1)

C. glabrata

CBS138

(μg.ml-1)

C. krusei

CBS573

(μg.ml-1)

C. parapsilosis

CBS604

(μg.ml-1)

16 0.023 0.014 0.006 0.027 0.006 0.281

47 0.117 0.07 0.094 0.469 2.25 >4.25

48 0.188 0.047 0.188 0.625 1.875 >4.25

49 0.203 0.047 0.063 0.438 0.813 >4.25

The simplification of the core (derivative 16 vs. 47), leads to a decreased activity. In

the case of C. albicans and C. dubliniensis the decrease is only 5-fold. The simplification

has a larger impact on the activity against C. tropicalis and C. glabrata where the

decrease is in the order of a magnitude (15- and 19-fold). More pronounced effects

were observed in the case of C. krusei where 47 was less active in the order of 2

magnitudes. Analogue 47 was inactive against C. parapsilosis. To conclude on the

influence of the methyl groups, the dimethyl-47, monomethyl-48 and desmethyl-49

Chapter 2

32

2

demonstrated roughly the same activity against all the studied Candida species. This

result is partially in agreement with the study reported by Fujie6 on micafungin

derivatives. The substitution of the fatty acid did not have a major influence on the

antifungal activity, but was used to improve hemolysis, which is the most common

side effect of treatment by enchinocandins.

Conclusions

This chapter presents a novel synthetic approach to methyl-branched fatty

acids. The combination of an enantioselective conjugate addition of MeMgBr to

linear aliphatic unsaturated thioesters, in combination with a Wittig reaction with an

acid functionalized phosphonium salt, places the methyl group at the desired position

in the chain. These reactions were applied in the synthesis of (R)-tuberculostearic

acid, the caspofungin fatty acid, and its analogues. (R)-Tuberculostearic acid was

prepared in only 5 steps with an overall yield of 63%. The fatty acid residue of

caspofungin 16 was prepared in 8 steps with an overall yield of 20%. Its monomethyl

analogue 46 was prepared in 5 steps with 42% overall yield. Given that the methyl

substituents of these 2 fatty acids have the opposite configuration, enantioselective

catalysis is the approach of choice. In the case of the caspofungin fatty acid residue,

also the influence of methyl branching on antifungal activity was studied where no

obvious relationship was observed.

Experimental part

Solvents for chemical reactions were dried according to the standard procedures.

Solvents for flash chromatography were used without further purification. All

reagents were used without further purification unless noted otherwise. All the

reactions were performed using Schlenk techniques unless noted otherwise.

Glassware was dried by heating (150 °C) for at least 2 h and subsequent cooling

under vacuum before use. Reactions were monitored by GC/MS (GC, HP6890: MS

HP5973) equipped with an HP1 column (Agilent Technologies, Palo Alto, CA) or

by TLC on silica coated aluminium foils (60 Å, 0.25 mm coating thickness). TLCs

were visualized by the following stains: iodine stain, Seebach’s stain (2.5 g

phosphomolybdic acid 1.0 g Ce(SO4)2 and 6.0 ml conc H2SO4 sequentially added to

94 ml H2O, bromcresol green stain (40.0 mg of bromocresol green dissolved in 100

ml EtOH, 0.1M NaOH solution added until blue) or Dittmer stain (phospholipid

stain, Preparation: solution A: 4.0 g MoO3 in 100 ml of hot concentrated H2SO4;

solution B: dissolve 180 mg Mo (metallic molybdenum) in 50 ml of hot solution A,

stock solution: after cooling, mix 50 ml solution B with 50 ml solution A for


33

2

phospholipids). Flash column chromatography was performed on 230-430 mesh

silica gel.1H-, 13C-, and 31P-NMR spectroscopy was performed on Varian VXR300 or

AMX400 spectrometers. Chemical shifts were determined relative to the residual

solvent peaks (CHCl3, δ = 7.26 ppm for 1H NMR, δ = 77.16 ppm for 13C NMR).

The reported shifts are in ppm. The ( - ) sign in the 13C NMR reports stands for

negative phase in APT (Attached Proton test). Optical rotations were measured on

a Schmidt+Haensch polarimeter (Polartronic MH8) with a 10 cm cell. The mass

spectra were recorded on an Thermoscientific LTQ OrbitrapXL spectrometer.

Synthesis of (R)-tuberculostearic acid (Scheme 2)

(E)-S-ethyl undec-2-enethioate (29)

To a suspension of 33 (5.10 g, 14.0 mmol, 1.4 equiv.) in water (10 ml),

lithium chloride (84.8 mg, 2.00 mmol, 20 mol%) and nonanal (1.70 ml, 10.0 mmol)

were added. The mixture was subsequently stirred for 18 h in an opened flask.

Subsequently, the water was evaporated, the residue was dissolved in CH2Cl2 (30 ml)

and adsorbed on silica. After evaporation, the residue was placed on top of the

column (dry loading) and chromatographed using 5% toluene in pentane.

The desired (E)-S-ethyl undec-2-enethioate25 (2.11 g, 92%) was obtained as a

colourless thick liquid.

1H NMR (400 MHz, CDCl3, δ): 6.88 (dt, J = 15.5, 7.0 Hz, 1H), 6.08 (d, J = 15.5 Hz,

1H), 2.93 (q, J = 7.4 Hz, 2H), 2.17 (ddd, J = 14.7, 7.3, 1.5 Hz, 2H), 1.43 (m, 3H),

1.26 (m, 12H), 0.87 (t, J = 6.9 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 190.11, 145.41 ( - ), 128.63 ( - ), 32.14, 31.80, 29.3,

29.14, 27.97, 22.99, 22.62, 14.80 ( - ), 14.05 ( - ) (1 signal overlapping)

HRMS-ESI+ (m/z): [M + H]+ calculated for C13H25OS, 229.162; found 229.162.

(R)-S-ethyl 3-methylundecanethioate (35)

(R,SFe)-Josiphos.EtOH adduct (49.0 mg, 82.5 μmol, 1.65 mol%) and

CuBr.Me2S (15.0 mg, 75.0 μmol, 1.50 mol%) were stirred in freshly distilled tBuOMe

(45 ml) until homogeneous (approx. 20 min). The mixture was cooled to –78 °C

(cryostat) and a solution of MeMgBr in Et2O (Acros Organics, 3 M, 2.0 ml,

Chapter 2

34

2

7.50 mmol, 1.2 equiv) was added dropwise. After stirring for 10 min, a solution of

26 (1.14 g, 5.0 mmol) in tBuOMe (5 ml) was added over 3 h by syringe pump. After

complete addition, the mixture was stirred for an additional 16 h. Then, EtOH

(5.0 ml) was added and the flask was removed from the cooling bath. An aqueous

solution of NH4Cl (1 M, 20 ml) was added and the mixture was stirred for 20 min at

rt. The resulting solution was transferred to a separatory funnel and the aqueous layer

was diluted with water (30 ml). Layers were separated, the aqueous layer was

extracted with Et2O (3x15 ml), and the combined organic layers were washed with

brine (50 ml), dried over MgSO4 and evaporated. The crude residue (a yellow-orange

liquid) was purified by column chromatography (SiO2, 2% Et2O in pentane) to afford

35 (1.15 g, 94%) as a thick colourless liquid.

1H NMR (400 MHz, CDCl3, δ): 2.87 (q, J = 7.4 Hz, 3H), 2.52 (dd, J = 14.4, 6.0 Hz,

1H), 2.33 (dd, J = 14.4, 8.1 Hz, 1H), 2.00 (m, 1H), 1.25 (m, 17H), 0.92 (d, J = 6.7

Hz, 3H), 0.87 (t, J = 6.8 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 199.32, 51.40, 36.61, 31.86, 31.06, 29.69 ( - ), 29.55,

29.27, 26.82, 23.23 , 22.65, 19.51 ( - ), 14.78 ( - ), 14.08 ( - ).

HRMS-ESI+ (m/z): [M + H]+ calculated for C14H29OS, 245.194; found, 245.195.

[α]D +3.3° (c = 1.02, CHCl3)

The enantiomeric excess was determined on the

corresponding carbamate, obtained by LiAlH4 reduction, treatment with phosgene

in toluene (5 equiv) and treatment with (S)-(−)-1-(1-naphthyl)ethylamine (1 equiv).

The retention times were compared to a racemic sample.

Chiracel OD-H, flow=1 ml/min, tminor=11.4 min, tmajor=12.4 min, e.r=95:5.

(R)-3-methylundecanal (28)

This compound was used without purification. One time, isolation was

performed in order to confirm its absolute configuration.

To a solution of (R)-S-ethyl 3-methyl undecathioate 35 (978 mg, 4.00 mmol) in

anhydrous CH2Cl2 (23 ml), cooled to –78 ºC (EtOH/dry ice), a solution of DIBAL


35

2

in CH2Cl2 (1 M, 5.2 ml, 5.2 mmol, 1.3 equiv) was added. The mixture was stirred

until complete consumption of starting material (approx. 2 h, TLC). The solution

was poured into saturated Rochelle salt and stirred until phases separated (overnight).

The aqueous layer was extracted with CH2Cl2 (2x30 ml), the combined organic layers

were washed with brine, dried over MgSO4 and carefully evaporated. The residual

colourless liquid was purified by column chromatography (2% Et2O in pentane) to

afford 737 mg (94%) of a 28 as pleasantly smelling colourless liquid.

1H NMR (400 MHz, CDCl3, δ): 9.89 – 9.64 (m, 1H), 2.39 (ddd, J = 16.0, 5.7, 2.1 Hz,

1H), 2.22 (ddd, J = 16.0, 7.8, 2.6 Hz, 1H), 2.04 (d, J = 6.7 Hz, 1H), 1.28 (m, 14H),

0.96 (d, J = 6.7 Hz, 3H), 0.88 (t, J = 6.9 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 180.15, 41.58, 36.64, 31.87, 30.12 ( - ), 29.69, 29.55,

29.27, 26.86, 22.65, 19.66 ( - ), 14.07 ( - ).

HRMS-ESI+ (m/z): [M + H]+ calculated for C12H25O, 185.191; found, 185.191.

[α]D + 10.3 (c = 0.25, hexanes).

Value of the optical rotation matches with a value previously reported.26

(R)-10-methyloctadec-7-enoic acid (33).

To a vigorously stirred suspension of 7-

(bromotriphenylphosphoranyl)heptanoic acid 27 (2.51 g, 5.3 mmol, 1.4 equiv) in

THF (4.0 ml) at 21 °C, a solution of LiHMDS (Sigma-Aldrich, 1 M, 10.6 ml, 10.6

mmol, 2.8 equiv) was added dropwise. After 30 min of stirring, (R)-3-

methylundecanal 28 (700 mg, 3.8 mmol) in a small amount of THF (1.00 ml) was

added dropwise over 5 min. The resulting reaction mixture was stirred until the

solution remained pale yellow (3 h), then HCl (1 M, aqueous) was added until the

pH reached 1. The mixture was transferred to a separatory funnel, the organic layer

was separated, and the aqueous layer was extracted with Et2O (2 x 25 ml). The

combined organic layers were washed with brine, dried over MgSO4 and evaporated.

The residual thick colourless liquid was purified by column chromatography (20%

Et2O in pentane 1% formic acid) to afford 890 mg (79%) of E and Z isomers of 36

as colourless liquid.

Chapter 2

36

2

1H NMR (400 MHz, CDCl3, δ): 5.38 (m, 2H), 2.35 (t, J = 7.5 Hz, 2H), 2.02 (m, 3H),

1.84 (m, 3H), 1.65 (dd, J = 9.9, 5.0 Hz, 1H), 1.27 (m, 2H), 0.86 (dt, J = 14.7,

7.3 Hz, 17H), 0.86 (dt, J = 14.7, 7.3 Hz, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.49, 130.11, 128.81, 36.70, 34.52, 33.40, 31.91,

29.96 ( - ), 29.66, 29.34, 29.30, 28.72, 27.17, 27.07, 24.58, 22.67, 19.58 ( - ), 14.10 ( - ).

(R)-Tuberculostearic acid (4)

To a solution of (E)- and (Z)-(R)-10-

methyloctadec-7-enoic acid 33 (741 mg, 2.5 mmol) and hydrazine hydrate (2.5 ml, 53

mmol, 21 equiv) in an O2 atmosphere, riboflavin catalyst (47.5 mg, 0.13 mmol,

5.0 mol%) was added in one portion. The mixture turned from red to yellow and the

reaction was stirred for 24 h at ambient temperature (21 °C). After complete

conversion of the starting material, the solution was acidified with concentrated HCl

to pH = 1 and extracted with Et2O (3x50 ml). The combined organic layers were

dried over MgSO4 and evaporated. The residual yellow thick liquid was purified by

column chromatography (20% Et2O in pentane) to afford 724 mg of (R)-

tuberculostearic acid 23 (97%) as colourless liquid.

1H NMR (400 MHz, CDCl3, δ): 2.35 (t, J = 7.5 Hz, 2H), 1.63 (dt, J = 15.0, 7.5 Hz,

2H), 1.42 – 1.00 (m, 27H), 0.88 (t, J = 6.9 Hz, 3H), 0.83 (d, J = 6.5 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 180.82, 37.08, 37.05, 34.06, 32.73 ( - ), 31.91, 30.02,

29.92, 29.68, 29.45, 29.35, 29.23, 29.05, 27.07, 27.02, 24.65, 22.67, 19.69 ( - ), 14.09

( - ).

HRMS-ESI+ (m/z): [M + H]+ calculated for C19H38O, 299.295; found, 299.295.

[α]D –0.2 (c = 3.1, CHCl3).

Caspofungin fatty acids (Scheme 3)

(E)-S-ethyl pent-2-enethioate (32)

To a cooled solution (0 °C) of (E)-pentenoic acid 37 (5.0 g, 50.0 mmol),

DCC (11.4 g, 55.0 mmol, 1.1 equiv) and DMAP (611 mg, 5.0 mmol, 10 mol%) in

pentane (500 ml), neat EtSH (7.2 ml, 0.1 mol, 2.0 equiv) was added dropwise. The


37

2

mixture was allowed to slowly reach room temperature and stirred for 16 h. After

filtration through a short silica pad, volatiles were evaporated. The crude residue was

distilled on Kugelrohr and afforded product 32 as a colourless liquid (6.85 g; 95%).

1H NMR (400 MHz, CDCl3, δ): 6.90 (m, 1H), 6.06 (dd, J = 15.5, 1.5 Hz, 1H), 2.90

(m, 2H), 2.18 (qd, J = 7.2, 3.7 Hz, 2H), 1.15 (m, 3H), 0.99 (s, 3H).

13C NMR (101 MHz, CDCl3, δ): 190.03, 146.46 ( - ), 127.79 ( - ), 25.18, 22.95,

14.78 ( - ), 12.05 ( - ).

Spectral data correspond to literature.[27a]

(S)-S-ethyl 3-methylpentanethioate (38)

(S,RFe)-Josiphos.EtOH adduct (43.2 mg, 67 µmol, 1.3 mol%) and

CuBr.Me2S (12.8 mg, 62 µmol, 1.2 mol%) were stirred in freshly distilled tBuOMe

(56 ml) until the mixture remained homogeneous (typically 10-30 min). Then the

mixture was cooled to –78 °C and after 10 min a solution of MeMgBr in Et2O

(2.7 ml, 8.1 mmol, 1.6 equiv) was added dropwise during 10 min. After 15 min of

stirring a solution of thioester 10 (721 mg, 5.0 mmol) in tBuOMe (6.0 ml) was added

over 3 h by a syringe pump. The reaction mixture was stirred for an additional 16 h

at –78 °C, quenched with EtOH (5.0 ml) and allowed to reach ambient temperature.

Then a solution of NH4Cl (1 M, 50 ml) was added. The organic layer was separated

and the aqueous layer extracted with Et2O (3 20.0 ml). The combined organic layers

were dried over MgSO4 and carefully evaporated (the product is volatile). The

residual yellow liquid was purified by flash chromatography (20% Et2O in pentane)

to afford (S)-S-ethyl 3-methylpentanethioate (38) (800 mg, 80%) as a colourless

liquid.


1H), 2.34 (dd, J = 14.4, 8.1 Hz, 1H), 1.95 (m, 1H), 1.37 (m, 1H), 1.24 (t, J = 7.4 Hz,

4H), 0.9 (m, 6H).

13C NMR (101 MHz, CDCl3, δ): 199.36, 51.01, 32.63 ( - ), 29.23, 23.24, 19.02 ( - ),

14.78 ( - ), 11.21 ( - ).NMR spectra contain traces of solvents (≈5%). Spectral data

correspond to literature.20]

HRMS-ESI+ (m/z): [M + H]+ calculated for C8H16OS, 161.099; found 161.099.

Chapter 2

38

2

The enantiomeric ratio was determined on the corresponding methyl ester by chiral

stationary phase gas chromatography on a Chiraldex G-TA column (30 m x

0.25 mm), 60 ºC, retention times: 5.94 (R) / 6.05 (S) min: 98.5:1.5 (e.r.), 97% ee (as

reported in literature).20

[α]D +8.4 (c = 1.0 in CHCl3).

The absolute configuration was determined on the alcohol obtained by reduction

with LiAlH4.

[α]D +7.4 (c = 0.95 in CHCl3).

Literature[44] reports the opposite enantiomer with [α]D –8.5 (c = 1 in CHCl3).

(S,E)-S-ethyl 5-methylhept-2-enethioate (41)

A solution of (S)-S-ethyl 3-methylpentanethioate 38 (801 mg;

5.0 mmol) in CH2Cl2 (50 ml) was cooled to –55 °C and then a solution of DIBAL

(1 M in CH2Cl2, 6.0 ml, 6.0 mmol, 1.2 equiv) was added. The mixture was stirred

until complete conversion of starting material (ca 1.5 h). Subsequently, the mixture

was poured into a saturated Rochelles salt (potassium sodium tartrate) solution and

stirred until the phases separated (mostly within 2 h). Layers were separated and the

aqueous layer was extracted with CH2Cl2 (315 ml). The combined layers were dried

and carefully evaporated (the product is volatile) until the weight of the residue

corresponded to quantitative yield (501 mg).

To a cooled solution (0 °C) of 39 (1.8 g, 7.5 mmol, 1.5 equiv) in THF (25 ml) a

solution of n-BuLi (1.6 M in hexanes, 3.4 ml, 5.5 mmol, 1.1 equiv) was added

dropwise. The reaction mixture was stirred for 20 min at 0 °C. Then (S)-3-

methylpentanal from the previous step (ca 5.0 mmol) in a small amount of THF

(0.3 ml) was added. The reaction mixture was stirred overnight (16 h). The reaction

was quenched with water (10 ml). Layers were separated and the aqueous layer

extracted with Et2O (3 15 ml). The combined organic layers were dried over

MgSO4 and carefully evaporated (the product is volatile). The residue was purified

by flash chromatography on SiO2 (0.4% tBuOMe in pentane) and affored (S,E)-S-

ethyl 5-methylhept-2-enethioate 41 (651 mg, 70 %) as a colourless liquid.

1H NMR (400 MHz, CDCl3, δ): δ 6.85 (ddd, J = 8.6, 7.6, 1.2 Hz,), 6.08 (dd, J = 15.5,

1.3 Hz, 1H), 2.94 (m, 2H), 2.17 (ddd, J = 7.2, 6.5, 3.5 Hz, 1H), 2.00 (m, 1H), 1.53


39

2

(dq, J = 13.3, 6.9 Hz, 1H), 1.36 (m, 2H), 1.26 (ddd, J = 7.4, 4.9, 1.2 Hz, 3H), 0.86 (m,

6H).

13C NMR (101 MHz, CDCl3, δ): 189.95, 144.27 ( - ), 129.69 ( - ), 39.21, 34.19 ( - ),

29.18, 22.99, 19.14 ( - ), 14.79 ( - ), 11.34 ( - ) + 2 peaks at 65.81 and 15.24 as Et2O

residues.


[α]D +8.2 (c = 1.8 in CHCl3);

(3S,5S)-S-ethyl 3,5-dimethylheptanethioate (42)

(S,RFe)-Josiphos-CuBr complex (29.1 mg, 1.30 mol%) was dissolved

in freshly distilled tBuOMe (24.0 ml) until the mixture remained homogeneous

(typically 10-30 minutes). Then the mixture was cooled to –78°C and after 10 min

a solution of MeMgBr in Et2O (3 M in Et2O, 1.20 ml, 1.30 equiv) was added

dropwise. After 15 min of stirring, a solution of thioester 41 (490 mg, 2.63 mmol) in tBuOMe (2.60 ml) was added over 3 h by a syringe pump. The mixture was stirred

for an additional 16 h at –78°C. The reaction was quenched by addition of EtOH

(2.00 ml) and the mixture was allowed to reach ambient temperature. Then an

aqueous solution of NH4Cl (1 M, 30 ml) was added, the organic layer separated and

the aqueous layer extracted with Et2O (3 20 ml). The combined organic layers were

dried over MgSO4 and carefully evaporated (the product is volatile). The residual

yellow liquid was purified by flash chromatography (0.4% tBuOMe in pentane) to

afford (3S,5S)-S-ethyl 3,5-dimethylheptanethioate 41 (418 mg, 78%) as a colourless

liquid.


1H), 2.28 (dd, J = 14.4, 8.5 Hz, 1H), 1.43 – 1.27 (m, 2H), 1.24 (t, J = 7.4 Hz, 4H),

1.04 (m, 3H), 0.92 (d, J = 6.6 Hz, 3H), 0.85 (m, 6H).

13C NMR (101 MHz, CDCl3, δ): 199.33, 51.29, 44.07, 31.53 ( - ), 29.00 ( - ), 28.66,

23.23, 20.15 ( - ), 19.48 ( - ), 14.79 ( - ), 11.07 ( - ).


[α]D +4.2 (c = 1.4 in CHCl3).

Chapter 2

40

2

(10S,12S)-10,12-dimethyltetradec-7-enoic acid (43)

A solution of (S)-S-ethyl 3-methylpentanethioate 41

(417 mg, 2.1 mmol) in CH2Cl2 (20 ml) was cooled to –55 °C and a solution of DIBAL

(1 M in CH2Cl2, 2.5 ml, 2.5 mmol, 1.2 equiv) was added. The mixture was stirred

until complete conversion of starting material (ca 1.5 h), then poured into saturated

aqueous Rochelles salt and stirred until phases separated (mostly within 2 h). Layers

were separated and the aqueous layer was extracted with CH2Cl2 (315 ml). The

combined layers were dried and carefully evaporated (the product is volatile) to a

weight corresponding to quantitative yield (294 mg).

To a stirred suspension of 7-(bromotriphenylphosphoranyl) heptanoic acid 27

(1.65 g, 3.8 mmol, 1.8 equiv) in THF (2.0 ml) at ambient temperature a solution of

LiHMDS (1 M in THF, 4.0 ml, 4.0 mmol, 2.0 equiv) was added dropwise. The

mixture was stirred until the suspension turned into a deep red solution. Then a

solution of (3S,5S)-3,5-dimethylheptanal (ca 2.1 mmol) in a small amount of THF

(300 µl) was added and the reaction mixture was stirred until complete consumption

of starting material (2 h). The mixture was acidified to pH=1 by dilute aq. HCl and

extracted with Et2O (3 20 ml). The combined organic layers were dried and

evaporated. The resulting thick liquid was purified by column chromatography (20%

Et2O in pentane) and afforded 43 (331 mg, 47%) as a colourless liquid (a mixture of

E and Z isomers).

1H NMR (400 MHz, CDCl3, δ): 5.37 (m, 2H), 2.35 (t, J = 7.5 Hz, 2H), 1.34 (m, 26H).

13C NMR (101MHz, CDCl3, δ): 131.22, 130.17, 129.03, 128.67, 44.23, 44.00, 39.77,

34.26, 34.03, 32.35, 31.61, 31.55, 30.70, 30.42, 29.29, 29.16, 28.72, 28.49, 27.08, 24.57,

24.51, 20.16, 20.06, 19.68, 11.14. – additional peaks observed due to the inseparable

E/Z mixture.

HRMS-ESI+ (m/z): [M + H]+ calculated for C16H32O2, 255.232; found, 255.235.

[α]D +12.2 (c = 1.8 in CHCl3).

(10R,12S)-10,12-dimethyltetradecanoic acid (16)

To a vigorously stirred solution of (10S,12S)-10,12-

dimethyltetradec-7-enoic acid 14 (310 mg, 1.2 mmol) and flavine catalyst


41

2

(49.6 mg, 10 mol%) in EtOH (1.0 ml) under oxygen atmosphere, hydrazine hydrate

(1.6 ml, 31.7 mmol, 26 equiv) was added in one portion. Vigorous stirring continued

for 16 h. Then the reaction mixture was acidified to pH=1 by dilute aq. HCl and


evaporated. The residual red liquid was purified by column chromatography

(20% Et2O in pentane) and afforded 16 (524 mg, 81%) as a colourless thick liquid.

1H NMR (400 MHz, CDCl3, δ): 11.20 (bs, 1H), 2.34 (t, J = 7.5 Hz, 2H), 1.13 (m,

27H).

13C NMR (101MHz, CDCl3, δ): 180.15, 44.70, 36.85, 34.21, 31.56 ( - ), 30.00 ( - ),

29.93, 29.46, 29.24, 29.20, 29.07, 26.86, 24.73, 20.26 ( - ), 19.72 ( - ), 11.17 ( - ).

HRMS-ESI- (m/z): [M - H]+ calculated for C16H32O2, 255.232; found, 255.216.

[α]D +14.1 (c = 1.3 in CHCl3).

(S)-12-methyltetradec-9-enoic acid (45)

To a stirred solution of (S)-S-ethyl 3-

methylpentanethioate 38 (73.5 mg, 0.5 mmol) in CH2Cl2 (0.8 ml) at –50 °C, DIBAL

was added. After complete conversion of the thioester (2 h), the reaction mixture

was poured into a saturated solution of Rochelle salt. After clear layer separation, the

organic layer was separated and the aqueous layer extracted with Et2O (3 10 ml).

The combined organic layers were dried and carefully evaporated until the weight

corresponded to a quantitative yield (43.0 mg).

Then, to a stirred suspension of 9-(bromotriphenyl phosphoranyl)nonanoic acid 44

(425 mg, 0.9 mmol, 1.7 equiv.) in THF (0.8 ml) a solution of LiHMDS (1 M in THF;

1.0 ml, 1.0 mmol, 2.0 equiv) was added until the solution remained deep red. To this

solution, 3-methyl-pentanal in a small amount of THF (300 µL) was added. The

reaction mixture was stirred until complete conversion of the starting material (2 h)

and acidified to pH=1 by dilute aq. HCl. The resulting solution was extracted with

Et2O (3 20 ml) and the combined organic layers were dried and evaporated. The

resulting thick liquid was purified by column chromatography (20% Et2O in pentane)

and afforded 45 (65.9 mg, 60%) as a colourless liquid (as a mixture of E and Z

isomers).

Chapter 2

42

2

1H NMR (400 MHz, CDCl3, δ): 5.37 (dt, J = 6.0, 4.6 Hz, 2H), 2.35 (t, J = 7.5 Hz,

2H), 2.01 (d, J = 5.4 Hz, 2H), 1.63 (m, 2H), 1.30 (m, 14H), 0.88 (m, 6H).

13C NMR (101 MHz, CDCl3, δ): 177.89, 130.32 ( - ), 128.66 ( - ), 35.11, 34.13, 33.63,

29.69, 29.46, 29.19, 28.94, 28.88, 27.19, 24.64, 19.11 ( - ), 11.55 ( - ) (1 signal

overlapping)

HRMS-ESI- (m/z): [M - H]+ calculated for C15H29O2, 241.216; found, 241.217.

[α]D= +0.3 (c = 0.5 in CHCl3).

(S)-13- methylpentadecanoic acid (46)

To a vigorously stirred solution of (13S)-13-

methyltetradec-7-enoic acid 15 (60 mg, 250 μmol) and flavine catalyst 26 (10 mg,

25 µmol, 10 mol%) in EtOH (1.0 ml) under oxygen atmosphere, hydrazine hydrate

(375 µl, 7.5 mmol, 30 equiv.) was added in one portion. Vigorous stirring was

continued for 16 h and the mixture was acidified to pH=1 by dilute aq. HCl and


evaporated. The residual red liquid was purified by column chromatography (20%

Et2O in pentane) and afforded 46 (50.0 mg, 83%) as a colourless thick liquid.

1H NMR (400 MHz, CDCl3, δ): 2.34 (t, J = 7.5 Hz, 2H), 1.62 (m, 2H), 1.28 (m, 20H),

0.84(m, 6H).

13C NMR (101 MHz, CDCl3, δ): 220.66, 36.62, 34.38, 34.07, 29.99, 29.66, 29.58,

29.48, 29.42, 29.23, 29.05, 27.09, 24.66, 19.20 ( − ), 11.39 ( − ).


[α]D= +0.3 (c = 0.5 in CHCl3).


(1) (a) Ohagan, D. Nat. Prod. Rep. 1993, 10, 593(b) Minnikin, D. E.; Kremer, L.; Dover,

L. G.; Besra, G. S. Chem. Biol. 2002, 9, 545(c) Rezanka, T.; Sigler, K. Prog. Lip. Res.

2009, 48, 206.

(2) Rainwater, D. L.; Kolattukudy, P. E. J. Biol. Chem. 1985, 260, 616.


43

2

(3) Jauregui.G; Lenfant, M.; Toubiana, R.; Azerad, R.; Lederer, E. Chem. Commun. 1966,

855.

(4) Kaneda, T. Microbiol. Rev. 1991, 55, 288.

(5) Denning, D. W. The Lancet 2003, 362, 1142.

(6) Fujie, A. Pure Appl. Chem. 2007, 79, 603.

(7) O'Sullivan, D. M.; Nicoara, S. C.; Mutetwa, R.; Mungofa, S.; Lee, O. Y. C.; Minnikin,

D. E.; Bardwell, M. W.; Corbett, E. L.; McNerney, R.; Morgan, G. H. PLoS One

2012, 7.

(8) French, G. L.; Chan, C. Y.; Cheung, S. W.; Teoh, R.; Humphries, M. J.; O'Mahony,

G. The Lancet 1987, 330, 117.

(9) Knapp, F. F., Jr.; Ambrose, K. R.; Goodman, M. M. Eur. J. Nucl. Med. 1986, 12, S39.

(10) Ran-Ressler, R. R.; Devapatla, S.; Lawrence, P.; Brenna, J. T. Pediatr. Res. 2008, 64,

605.

(11) (a) Prout, F. S.; Cason, J.; Ingersoll, A. W. J. Am. Chem. Soc. 1947, 69, 1233 (b) Prout,

F. S.; Cason, J.; Ingersoll, A. W. J. Am. Chem. Soc. 1948, 70, 298 (c) Liu, X.; Stocker,

B. L.; Seeberger, P. H. J. Am. Chem. Soc. 2006, 128, 3638 (d) Dyer, B. S.; Jones, J. D.;

Ainge, G. D.; Denis, M.; Larsen, D. S.; Painter, G. F. J. Org. Chem. 2007, 72, 3282

(e) Roberts, I. O.; Baird, M. S. Chem. Phys. Lipids 2006, 142, 111 (f) ter Horst, B.;

Seshadri, C.; Sweet, L.; Young, D. C.; Feringa, B. L.; Moody, D. B.; Minnaard, A. J.

J. Lipid Res. 2010, 51, 1017.

(12) (a) Smit, C.; Fraaije, M. W.; Minnaard, A. J. J. Org. Chem. 2008, 73, 9482 (b) Teichert,

J. F.; den Hartog, T.; Hanstein, M.; Smit, C.; ter Horst, B.; Hernandez-Olmos, V.;

Feringa, B. L.; Minnaard, A. J. ACS Catal. 2011, 1, 309.

(13) Carballeira, N. M.; Cruz, H.; Hill, C. A.; De Voss, J. J.; Garson, M. J. Nat. Prod. 2001,

64, 1426.

(14) López, F.; Minnaard, A. J.; Feringa, B. L. Acc. Chem. Res. 2006, 40, 179.

(15) Harutyunyan, S. R.; den Hartog, T.; Geurts, K.; Minnaard, A. J.; Feringa, B. L. Chem.

Rev. 2008, 108, 2824.

(16) Fukuyama, T.; Tokuyama, H. Aldrichimica Acta 2004, 37, 87.

(17) van Zijl, A. W.; Minnaard, A. J.; Feringa, B. L. J. Org. Chem. 2008, 73, 5651.

(18) Keck, G. E.; Boden, E. P.; Mabury, S. A. J. Org. Chem. 1985, 50, 709.

(19) Claridge, T. D. W.; Davies, S. G.; Lee, J. A.; Nicholson, R. L.; Roberts, P. M.; Russell,

A. J.; Smith, A. D.; Toms, S. M. Org. Lett. 2008, 10, 5437.

(20) Des Mazery, R.; Pullez, M.; López, F.; Harutyunyan, S. R.; Minnaard, A. J.; Feringa,

B. L. J. Am. Chem. Soc. 2005, 127, 9966.

(21) (a) Corey, E. J.; Mascitti, V. J. Am. Chem. Soc. 2004, 126, 15664 (b) Corey, E. J.;

Mascitti, V. J. Am. Chem. Soc. 2006, 128, 3118 (c) Jakubowski, J. A.; Kohn, T. J.;

Mais, D. E.; Takeuchi, K.; True, T. A.; Wyss, V. L.; Mais, D. E.; True, T. A. Bioorg.

Med. Chem. Let. 1998, 8, 1943

(22) Leonard, W. R.; Belyk, K. M.; Bender, D. R.; Conlon, D. A.; Hughes, D. L.; Reider,

P. J. Org. Lett. 2002, 4, 4201.

(23) Enders, D.; Eichenauer, H. Chem. Ber. 1979, 112, 2933.

Chapter 2

44

2

(24) White, J. D.; Johnson, A. T. J. Org. Chem. 1994, 59, 3347.

(25) The Z-isomer elutes first.

(26) Bowen, E. G.; Wardrop, D. J. J. Am. Chem. Soc. 2009, 131, 6062.

45

Chapter 3 Catalytic Synthesis of Enantiopure Mixed Diacylglycerols

Abstract: Protected diacylglycerols are valuable precursors of phospholipids. A

catalytic one-pot synthesis of TBDMS-protected diacylglycerols has been developed,

starting from enantiopure glycidol. Subsequent migration-free deprotection leads to

stereo- and regiochemically pure diacylglycerols, which can be converted into the

desired phospholipids. Application of a more electrophilic catalyst allows synthesis

of mixed ether/ester lipids.

Parts of this chapter have been published: Fodran, P.; Minnaard, A. J. Org. Biomol.

Chem. 2013, 11, 6919.

Chapter 3

46

3

Introduction

Diacylglycerols and diacyl glycerophospholipids are compounds with an

enormous significance for living organisms.1 Besides being the major components of

cell membranes, phospholipids in particular are involved in numerous physiological

processes, and some of them even found therapeutic application.2 Natural

glycerophospholipids contain a phosphate head group and usually two different acyl

chains (Figure 1). A different chain length results in a different hydrophobic

thickness of the membrane formed from these lipids, which has a direct effect on

the functioning of membrane embedded proteins.3 The variability in head group and

acyl chains results in many different glycerophospholipid species (see Chapter 1).

While separation of phospholipids based on the head group is relatively

straightforward, separation based on the chain length is practically impossible.

Figure 1. Mixed diacyl glycerolphospholipids.

Despite their apparent structural simplicity, 1,2-diacylglycerols bearing two

different acyl residues are challenging synthetic targets. A commonly encountered

difficulty is the migration of the acyl group from the secondary to the primary

hydroxyl (Figure 2), which is accompanied by the release of steric strain.4

Figure 2. 1,2-to-1,3 acyl shift in diacylglycerols.

This acyl shift is catalyzed by traces of acids and bases. Furthermore, it also occurs

on chromatography stationary phases (silica gel, aluminium oxide) leading to

decreased yields and tedious purification.

Catalytic Synthesis of Enantiopure Mixed Diacylglycerols

47

3

Over the years, the limited access to diacylglycerols and phospholipids

bearing 2 different chains attracted the attention of organic chemists. Several

synthetic routes circumvent the acyl migration, usually by application of multiple

protection/deprotection steps. Starting from enantiopure 4-methoxybenzyl

protected glycerol, Martin et al.5 obtained mixed diacyl glycerolphospholipids in a

sequence of 8 steps and an overall 52% yield. Gras and Bonfanti6 used 5-hydroxy-

1,3-dioxan (formaldehyde protected glycerol) to synthetize selectively mixed 1,2-

diacylglycerols. A drawback of this approach is that the products are racemic.

Massing and Eibl7 applied protected D-mannitol in the synthesis of ether-based lipids

represented by platelet activating factor (PAF). Guanti et al.8 synthetized 1,2-

diacylglycerols by chemo-enzymatic methods. Starting from racemic solketal

(glycerol acetonide), the authors obtained 1,2-dipalmitoyl-sn-glycerol in 55% yield

using the Amano P protease.

Regioselective glycidol opening can be an attractive and atom economic

option for the synthesis of phospholipids. This strategy was pursued in several ways

(Scheme 1).

Scheme 1. Phospholipid syntheses with epoxide ring opening as the key step.

Chapter 3

48

3

Burgos et al.9 obtained monoacylglycerol 4 in 25% yield starting from glycidol 3 by

applying excess of stearic acid and Ti(OiPr)4. Besides the low yield (25%), the authors

reported a decreased enantiopurity of 4 compared to the starting material. Ali and

Bittman10 used BF3.OEt2 for ring-opening of tosyl glycidol 5 by a fatty acid

anhydride. Although the reaction afforded 6 in a good 76% yield, it is limited to 1,2-

diacylglycerols containing two identical acyl chains. Lindberg et al.11 achieved

regioselective acylation of epoxide 7 in conditions promoting SN2 reaction (good

nucleophile, polar aprotic solvent). Reaction with an excess (3 equiv) of palmitic acid

and cesium palmitate resulted in 8 in a good 65% yield. Stawinski and Stamatov12

described a 3 step procedure starting from 9. The epoxide 9 was first treated with an

excess of nBu4NI and the resulting iodohydrine was treated with an excess of acid

chloride to afford 10. The iodide, which prevented acyl migration, was substituted in

the last step with an excess of nBu4N salt of fatty acid affording 11 in an excellent

92% over 3 steps.

In the current era, in which sustainability is a frequently used term, a general,

efficient, and preferably catalytic approach to enantiopure mixed diacyl glycerols

would be highly desirable. A drawback of the development of such an approach is

the lack of catalysts that allow regioselective ring opening of terminal epoxides with

acids. Jacobsen et al.13 reported the application of a cobalt catalyst 13 in a

desymmetrization of meso-epoxide 12 (Scheme 2).

Scheme 2. Desymmetrization and kinetic resolution of epoxides using 13.


49

3

Although application of 13 to the kinetic resolution of terminal epoxides by fatty

acids was unsuccessful, in the case of hydrolytic kinetic resolutions of terminal

epoxides 15 the catalyst gives excellent results (Scheme 2).

The following part of this chapter describes the development of a catalytic

one-pot synthesis of protected diacylglycerols, and their further conversion into the

desired phospholipids bearing two different acyl groups. Furthermore, a slight

modification of the conditions allowed the synthesis of mixed ether/ester products,

which can be further converted into, for example, platelet activating factor.


Synthesis of enantiopure phospholipids

It turned out that in solvent-free conditions, enantiopure TBDMS protected

glycidol (17) reacted smoothly with stearic acid in the presence of catalytic 13 and

stoichiometric iPr2NEt. These conditions led to monoacyl glycerol 18 as a single

regioisomer in quantitative yield as confirmed by 1H-NMR, 13C-NMR and GC-MS.

As 18 was the only product of the ring opening reaction it could be esterified in the

same pot using a second fatty acid, DCC and DMAP. After full conversion of 18,

the crude reaction mixture was transferred on a silica gel column without any workup

and chromatographed to afford 19 in 92% yield over two steps (Scheme 3). The

reaction could be readily scaled-up.

Reagents and conditions: (a) C17H35CO2H (1.0 equiv), iPr2NEt (1.0 equiv), 13 (1 mol%), 21 °C, 16 h; (b) C15H31CO2H (1.2 equiv), DCC (1.1 equiv), DMAP (10 mol%), heptane, 21 °C, 16 h.

Scheme 3 A two-step-one-pot synthesis of protected diacylglycerol 18.

Chapter 3

50

3

At 5 mmol scale, the protected diacylglycerol was isolated in 91% yield,

corresponding to 3.2 g of the desired product. The two-step-one-pot protocol also

allows synthesis of unsaturated protected diacylglycerols. When oleic acid was

applied for the ring opening and palmitic acid for the final esterification, the product

was isolated in 82% yield over 2 steps.

Acyl migration during deprotection of protected 1,2-diacylglycerols is a

common problem, which is enhanced by a difficult separation of the undesired

rearranged 1,3-diacylglycerol. Although conditions for migration-free deprotection

of benzylated and tritylated14 diacylglycerols have been reported,

Reagents and conditions: (a) BF3.CH3CN (1.1 equiv), CH2Cl2, 21 °C, 5 min; (b) 21 (1.2 equiv), 4,5-dicyanoimidazole (1.0 equiv), CH2Cl2, 15 min; (c) tBuOOH (3.0 equiv), 10 min; (d) Pd/C (5 mol%), MeOH / HCO2H (96 / 4), 21 °C.

Scheme 4. Conversion of a protected diacylglycerol into the corresponding phospholipid.

migration-free desilylation of 19 (Scheme 4) is a challenge for nearly 40 years.14-15

Examination of a wide variety of conventional deprotection conditions including

TBAF and TFA,16 led to substantial migration prior to full conversion. Conditions

that applied Lewis or Brønsted acids in only catalytic amounts17 gave similar results

at best. Finally, utilization of a small excess of BF3 (as its etherate or as its acetonitrile

complex) in CH2Cl2 at 0 °C, cleaved the TBDMS group with no detectable migration

(TLC, NMR18) within 5 min. The reaction was closely monitored (by TLC) and was

quenched immediately upon full conversion. To exclude any migration during the

work-up, the reaction was quenched with chilled phosphate buffer (1 M, pH = 7).


51

3

The 1,2-diacylglycerol 20 was isolated in quantitative yield and used without further

purification.

The target 1,2-diacyl phosphatidylethanolamine 23 was synthetized using the

phosphoramidite coupling/oxidation methodology19 which is widely used in solid-

phase DNA synthesis. The mild reaction conditions and short reaction times

minimize the possibility of acyl migration. Full conversion of 20 was achieved within

15 min using 4,5-dicyanoimidazole as the activator, 21 as the head group precursor

and tBuOOH as the oxidant. The reaction was very clean and 22 was obtained in

85% yield. Finally, the protecting groups were removed by hydrogenolysis using

catalytic Pd/C and HCO2H as hydrogen donor. This led, after purification over

silica-gel, to the desired phospholipid 23 in 72% yield starting from 19.

Enantiopurity does not decrease during the ring opening

There is a frequently overlooked pathway that might decrease the

enantiopurity of the diacylglycerol product. In an ideal case, the epoxide opening

step displays high regioselectivity on the C1 position of 17 (Figure 3), therefore the

optical purity of 24 does not erode. But, if the regioselectivity is insufficient, then an

SN220 attack on the C2 position of 17 results in 25, which after acyl migration gives

26. As a result, the optical purity of the ring-opened product is lower.

Figure 3. Decrease of enantiopurity as a result of insufficient regioselectivity in the ring opening.

Routinely, in the field of phospholipids, the optical purity is studied via analysis of

the corresponding Mosher’s esters.21 However, the reported 1H- and 19F-NMR shifts

are in some cases contradictory. 9-11,22 Furthermore, incomplete conversion of the

starting material in the esterification with Mosher’s acid chloride (leading to a kinetic

Chapter 3

52

3

resolution), together with an attempt to purify the diastereomeric products, might

lead to an undesired diastereomeric enrichment, thus leading to erroneous results.

To circumvent these issues, the optical purity in the cobalt catalyzed reactions was

studied by HPLC with a chiral stationary phase. To this end, the ring-opening

reaction was performed on the closely related TBDPS-glycidyl ether.23

Application of the identical reaction conditions as shown above (Scheme 3) led to

smooth ring opening, and the resulting protected monoacylglycerol was analyzed

without any further sample manipulation. HPLC analyses on a chiral stationary phase

showed no decrease of the optical purity during this step. It is therefore safe to

conclude that the reaction occurs regio-specifically on the terminal position of the

epoxide.

Synthesis of the platelet-activating factor (PAF)

The developed methodology was applied in the synthesis of platelet-

activating factor (33) (Scheme 5). This mixed ether/ester type phosphatidylcholine

lipid affects the aggregation of platelets, and has a function in processes like glycogen

degradation, reproduction, brain function and blood circulation. Up to date, several

syntheses of platelet-activating factor have been reported.7,24

Initially, the reaction of glycidol 17 with hexadecanol in the same conditions

as with carboxylic acids (Scheme 3) led only to a marginal conversion. However,

application of the more electrophilic 2725 led to full conversion (Scheme 5) of

hexadecanol in the presence of epoxide 17 in 3 days. The reaction proceeded in the

same clean manner as above, affording ring opened product 28 as a single

regioisomer. This was further converted to 29 with excess Ac2O in the presence of

DMAP and Et3N. 29 was further converted to 33 in 3 steps (Scheme 5). Desilylation

in the same conditions as in the case of diacylglycerols (CH3CN.BF3 complex)

resulted in a migration-free deprotection. The phosphatidylcholine head group was

installed by a known 2 step procedure.26 In the first step 30 was phosphorylated with

an excess of 31 in the presence of Et3N. After full conversion and removal of the

excess of the reagents, 31 was treated with Me3N in the presence of TMSOTf. The

crude reaction mixture was loaded on a low-surface silica gel column (see

experimental section) and chromatographed. This afforded 33 in 71% yield over 3

steps (46% starting from 17).


53

3

Reagents and conditions: (a) hexadecanol (0.55 equiv), 27 (2 mol%), THF, 21 °C, 3 d (b) Ac2O (2.5 equiv), Et3N (2.5 equiv), DMAP (10 mol%), CH2Cl2, 21 °C, 16 h; (c) BF3.CH3CN (1.1 equiv), CH2Cl2, 0 °C, 20 min; (d) 31 (4.0 equiv), iPr2NEt (4.0 equiv), CH2Cl2, 0 °C, 16 h; (e) Me3N (1.5 equiv), TMSOTf (2 equiv), CH2Cl2.

Scheme 5. Synthesis of platelet activating factor (33).

Conclusions

A one-pot synthesis of enantiopure mixed diacylglycerols was developed,

starting from TBDMS-protected glycidol. The longstanding problem of acyl

migration upon deprotection of silyl-protected diacylglycerols was solved with the

use of a TBDMS protecting group and a BF3-complex for the desilylation. The

protocol is experimentally straightforward and can be readily scaled-up to a multi-

gram scale. The overall synthesis of phospholipids was optimized and scaled-up. The

possibility of erosion of the optical purity during the ring-opening reaction was ruled

out. Application of the more electrophilic catalyst 27 allowed synthesis of mixed

ether/ester type phospholipids represented by platelet activating factor.

Experimental section

A two-step-one-pot synthesis of protected diacylglycerol (Scheme 3)

(R)-3-((tert-butyldimethylsilyl)oxy)-2-(palmitoyloxy)propyl stearate (19)

Chapter 3

54

3

Stearic acid (1.42 g, 5.0 mmol, 1.0 equiv) and 1313 (30 mg,

50 µmol, l mol%) were suspended in a small amount of ether (ca 1 ml) and stirred

in an oxygen atmosphere for 15 min at 21 °C. The solvent was evaporated, and

Hünigs base (873 µl, 5.0 mmol, 1.0 equiv) was added. After 5 min of stirring, (R)-

TBDMS-glycidyl ether (1.0 ml, 5 mmol) was added, and the mixture was stirred for

16 h.

After 1H-NMR showed (attenuation of the signals at 2.63 ppm and 2.77 ppm)

complete conversion of the glycidyl ether, all volatiles were evaporated in high

vacuum. To a solution of this intermediate (5.0 mmol) in heptane (10 ml), palmitic

acid (1.54 g, 6.0 mmol, 1.2 equiv) and DMAP (61.0 mg, 0.5 mmol, 5 mol%) were

added, the mixture was chilled to 0 °C, and DCC (1.24 g, 6.0 mmol, 1.2 equiv) was

added in one portion. The reaction mixture was stirred for 16 h and subsequently

directly placed on a SiO2 column and chromatographed using 9% Et2O in pentanes

to afford the desired product (3.24 g, 91%) as a white solid.


(dd, J = 11.9, 6.3 Hz, 1H), 3.70 (m, 2H), 2.29 (td, J = 7.6, 2.1 Hz, 4H), 1.60 (m, 4H),

1.26 (broad s, 61H), 0.87 (m, 15H), 0.04 (s, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.36, 173.01, 71.64( - ), 62.41, 61.43, 34.31, 34.13,

31.90, 29.68, 29.64, 29.61, 29.46, 29.34, 29.27, 29.11, 29.08, 25.72 ( - ), 24.92, 24.89,

22.66, 18.16, 14.07, -5.52 ( - ), -5.56 ( - ).

[α]D +7.1 (c = 2.3, CHCl3).

Melting point 45 °C

HRMS: (ESI+) calculated for C43H87O5Si [M+H] +: 710.624 found: 710.635.

Conversion of a protected diacylglycerol into the corresponding phospholipid (Scheme 4).

(S)-3-hydroxy-2-(palmitoyloxy)propyl stearate (20)


55

3

A solution of 19 (1.5 g, 2.1 mmol) in CH2Cl2 (20 ml) was immersed

in an ice bath (ice/water). Then CH3CN.BF3 (2.0 ml, 2.3 mmol, 1.1 equiv) was added,

and the resulting light yellow mixture was stirred for 5 min, while carefully monitored

by TLC. After full conversion, the reaction mixture was diluted with Et2O (100 ml)

and poured onto cooled phosphate buffer (pH = 7, 1 M, 25 ml). The organic layer

was separated and washed with saturated brine (50 ml), dried and evaporated to

dryness to afford 1-stearoyl-2-palmitoyl glycerol (1.25 g, 99%) as a white solid. The

compound was used directly without delay and further purification.


(dd, J = 11.9, 5.7 Hz, 1H), 3.72 (d, J = 4.9 Hz, 2H), 2.31 (m, 4H), 1.61 (dd, J = 12.9,

6.8 Hz, 4H), 1.25 (s, 54H), 0.87 (t, J = 6.8 Hz, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.75, 173.40, 72.09 ( - ), 62.00, 61.51, 34.27, 34.08,

31.90, 29.68, 29.64, 29.60, 29.46, 29.34, 29.25, 29.10, 29.07, 24.92, 24.87, 22.67, 14.09

( - ).

(2R)-3-(((benzyloxy)(2-(((benzyloxy)carbonyl)amino)ethoxy)phosphoryl)oxy)-2-(palmitoyloxy)propyl stearate (22)

To a stirred solution of (S)-3-hydroxy-2-

(palmitoyloxy)propyl stearate (1.25 g, 2.1 mmol) in CH2Cl2 (10 ml), phosphoramidite

21 (1.13 g, 2.5 mmol, 1.2 equiv) was added. The mixture was cooled to 0 °C and 1H-

imidazole-4,5-dicarbonitrile (323 mg, 2.7 mmol, 1.3 equiv) was added in one portion.

The reaction was stirred until complete conversion of the starting diacylglycerol

(monitored by TLC - typically 30 min). Subsequently, the mixture was cooled to –20

°C, and tBuOOH (ca 5 M in decane, 800 µl, 4.4 mmol, 2.1 equiv.) was added,

followed by stirring for 30 min. The reaction was then diluted with 10 ml of CH2Cl2

and poured into aqueous NaHCO3 (1 M, 200 ml). The organic layer was washed with

aqueous HCl (1 M, 200 ml), brine, dried and evaporated. The resulting crude yellow

oil was purified by column chromatography on SiO2 using 10% pentane in CHCl3 to

afford 22 (1.78 g, 85%) as a colorless thick liquid, together with an co-eluting

impurity.

Chapter 3

56

3

1H NMR (400 MHz, CDCl3, δ): 7.32 (m, 1H), 5.34 (broad s, 1H), 5.18 (dd, J = 9.6,

5.3 Hz, 1H), 5.07 (m, 4H), 4.27 (m, 1H), 4.09 (m, 4H), 3.42 (m, 2H), 2.28 (m, 4H),

1.57 (d, J = 7.0 Hz, 4H), 1.23 (m, 52H), 0.88 (t, J = 6.8 Hz, 6H), 0.83 (d, J = 6.5 Hz,

9H).

13C NMR (101 MHz, CDCl3, δ): 173.19, 172.81, 128.80( - ), 128.67 ( - ), 128.47 ( - ),

128.09 ( - ), 128.03 ( - ), 69.80, 66.78, 65.49, 61.56, 41.33, 37.09, 34.10, 33.97, 32.75

( - ), 31.91, 30.02, 29.97, 29.68, 29.64, 29.52, 29.47, 29.35, 29.28, 29.11, 29.06, 27.08,

24.81, 22.67, 19.69 ( - ), 14.10 ( - ).

31P NMR (162 MHz, CDCl3, δ): -0.81, -0.83.

HRMS (ESI+): calculated for C55H93O10NP [M+H]+: 958.653, found 958.653.

[α]D +7.1 (c = 2.3, CHCl3).

(R)-2-ammonioethyl (2-(palmitoyloxy)-3-(stearoyloxy)propyl) phosphate (23)

To a stirred solution of the lipid precursor 22 (1.78 g, 1.8

mmol) in MeOH/formic acid (96 / 4, 50 ml), Pd/C (Degussa Type E101 NE/W,

95.4 mg, 90 µmol, 5.0 mol%) was added. The mixture was stirred under hydrogen

atmosphere (balloon) until complete conversion of the starting material (typically 2 h,

according to TLC). Subsequently, the solution was diluted with CH2Cl2 (200 ml), and

SiO2 (10 g) was added, followed by evaporation of the volatiles. The SiO2 with

adsorbed phospholipid was transferred onto a short (20 g) SiO2 column, impurities

were eluted with Et2O (100 ml), followed by elution of the phospholipid with CHCl3

/MeOH / H2O (65 / 35 / 7), to afford 23 (1.09 g, 85%) as a white solid.

1H NMR (400 MHz, CDCl3/CD3OD/D2O 95/35/2, δ): 5.21 (d, J = 4.6 Hz, 1H),

4.38 (dd, J = 12.1, 2.9 Hz, 3H), 4.23 (d, J = 5.3 Hz, 1H), 4.15 (dd, J = 12.1, 7.3 Hz,

1H), 4.02 (t, J = 8.7 Hz, 2H), 3.95 (t, J = 5.9 Hz, 2H), 3.30 (dt, J = 3.2, 1.6 Hz, 1H),

3.19 – 3.10 (m, 2H), 2.29 (m, 4H), 1.58 (d, J = 6.6 Hz, 4H), 1.22 (broad s, J = 15.4

Hz, 54H), 0.86 (t, J = 6.8 Hz, 6H).

13C NMR (101 MHz, CDCl3/CD3OD/D2O, δ): 174.12, 173.77, 70.38 ( - ), 63.63,

62.68, 61.56, 34.15, 34.02, 31.83, 29.60, 29.47, 29.45, 29.25, 29.07, 29.03, 24.84, 24.77,

22.56, 13.84 ( - ).


57

3

31P NMR (162 MHz, CDCl3/CD3OD/D2O, δ): 0.10.

HRMS (ESI+): calculated for C39H79NO8P [M+H]+:720.554 found 720.559.

[α]D +6.5 (c = 1.0, toluene).

Melting point: 125 °C

benzyl (2-(((benzyloxy)(diisopropylamino)phosphanyl)oxy)ethyl)carbamate (21)

In a dry 50 ml round bottom flask (benzyloxy)bis(N,N-

diisopropylamino)-phosphine (1.51 g, 4.5 mmol, 1.5 equiv) and CBz-amino ethanol

(585 mg, 3.0 mmol) were dissolved in dry CH2Cl2 (6 ml). The solution was cooled to

0 °C in an ice/brine bath. To this solution, solid tetrazole (210 mg, 1 equiv) was

added in one portion. Stirring was continued at 0 °C until full conversion of the CBz-

amino ethanol (TLC, 2 h). The mixture was diluted with CH2Cl2 (to a final volume

of ca 30 ml), the organic layer was washed with saturated Na2CO3 (2 x 30 ml), dried

over MgSO4 and evaporated. The resulting thick residue was purified on silica using

pentane / ethyl acetate / Et3N in ration 95 / 5 / 5.

The reaction afforded 656 mg of the desired compound as a colorless liquid (51%).

1H NMR (400 MHz, CDCl3, δ): 7.30 (m, 10H), 5.18 (bs, 1H), 5.10 (s, 2H), 4.70 (m,

2H), 3.73 (dd, J = 12.0, 7.4 Hz, 1H), 3.63 (m, 2 H), 1.19 (d, J = 6.8 Hz, 12 H).

13C NMR (101 MHz, CDCl3, δ): 156.51( - ), 139.27( - ), 136.78( - ), 128.60, 128.42,

128.15, 127.51, 127.23, 66.73 ( - ), 65.66 ( - ), 65.48 ( - ), 62.84 ( - ), 62.68 ( - ), 43.14

(d C-P coupling), 24.76 (t).

31P NMR (162 MHz, CDCl3, δ): 149.1.

NMR data correspond to those published previously.19b

Studies on the regioselectivity of the ring-opening and the optical purity of the ring-opened products.

(rac)-tert-butyl(oxiran-2-ylmethoxy)diphenylsilane

Chapter 3

58

3

In an oven dried Schlenk flask, imidazole (1.58 g, 23.2 mmol, 2.2 equiv)

was dissolved in CH2Cl2 (5.0 ml). Neat TBDPSCl (3.4 ml, 13.2 mmol, 1.1 equiv) was

added whereupon the mixture turned into a thick suspension which was cooled to

0 °C. To this suspension (rac)-glycidol (0.8 ml, 12.0 mmol) was added. The mixture

was stirred for 17 h allowing to reach gradually to RT. Solids were filtered and washed

with CH2Cl2 (3 x 20 ml). The combined organic layers were dried and concentrated.

The crude residue was further purified on silica using 20% Et2O in pentane to yield

the desired product (3.75 g) in quantitative yield as a colourless oil.

1H NMR (400 MHz, CDCl3, δ): 7.68 (d, J = 7.8 Hz, 4H), 7.41 (m, J = 13.7, 6.8 Hz,

6H), 3.85 (dd, J = 11.9, 3.1 Hz, 1H), 3.71 (dd, J = 11.8, 4.8 Hz, 1H), 3.13 (m, J = 7.5,

3.8 Hz, 1H), 2.75 (t, J = 4.6 Hz, 1H), 2.61 (dd, J = 5.1, 2.7 Hz, 1H), 1.06 (s, 9H).

13C NMR (101 MHz, CDCl3, δ): 135.94, 135.88 ( − ), 133.58, 133.57, 130.08, 128.05,

128.04, 64.61 ( - ), 52.61, 44.77 ( - ), 27.08, 19.57( - ). 27

(R)-tert-butyl(oxiran-2-ylmethoxy)diphenylsilane

In an oven dried Schlenk flask, imidazole (295 mg, 4.4 mmol, 2.2 equiv)

was dissolved in CH2Cl2 (20.0 ml). Neat TBDPSCl (620 µl, 2.2 mmol, 1.1 equiv) was

added whereupon the mixture turned into a thick suspension which was cooled to

0 °C. To this suspension, (R)-glycidol (98% ee, 130 µl, 2.0 mmol) was added. The

mixture was stirred for 17 h allowing to reach gradually to rt. Solids were filtered and

washed with CH2Cl2 (3 x 20 ml), the combined organic layers were dried and

concentrated. The crude residue was further purified on silica using 20% Et2O in

pentane to yield the desired product in 86% yield (538 mg, colorless liquid).

1H NMR (400 MHz, CDCl3, δ): 7.68 (d, J = 7.8 Hz, 4H), 7.41 (m, J = 13.7, 6.8 Hz,

6H), 3.85 (dd, J = 11.9, 3.1 Hz, 1H), 3.71 (dd, J = 11.8, 4.8 Hz, 1H), 3.13 (m, J = 7.5,

3.8 Hz, 1H), 2.75 (t, J = 4.6 Hz, 1H), 2.61 (dd, J = 5.1, 2.7 Hz, 1H), 1.06 (s, 9H).

13C NMR (101 MHz, CDCl3, δ): 135.94, 135.88 ( - ), 133.58, 133.57, 130.08, 128.05,

128.04, 64.61 ( - ), 52.61, 44.77 ( - ), 27.08, 19.57( - ).

GC/MS: calculated for C15H15O2Si [M-tBu]: 255, found 255.

Spectral data in agreement with those previously published27.

[α]D =+2.5 (c = 2.0, CHCl3).


59

3

The epoxide ring opening:

Reaction for racemic TBDPS-glycidyl ether was performed at 1 mmol scale using

TBDPS-(rac)-glycidyl ether (312 mg, 1.0 mmol), butyric acid (91.8 mg, 1.0 mmol,

1.0 equiv), Hünigs base (175 µl, 1.0 mmol, 1.0 equiv) and Co[salen] catalyst (60.1

mg, 0.1 mmol, 10 mol%).

Reaction of enantiopure glycidyl ether was performed on 0.5 mmol scale using

glycidyl ether (156 mg, 0.5 mmol), butyric acid (46 µl, 0.5 mmol, 1.0 equiv), Hünigs

base (87 µl, 0.5 mmol, 1.0 equiv) and catalyst (6.0 mg, 10 μmol, 1 mol%). Reactions

were performed using following procedure:

A solution of Co[salen] (1.0 mol%) and butyric acid (1.0 equiv) in Et2O (1 ml) was

stirred under oxygen atmosphere (balloon) for 15 min. A change in color from bright

red to red-brown was observed. The solvent was evaporated, and to the resulting

brown mixture, Hünigs base (1.0 equiv) was added, and after 5 min of stirring (R)-

TBDPS-glycidyl ether (1.0 equiv) was added. The resulting mixture was stirred for

16 h after which 1H-NMR showed complete conversion of the glycidyl ether

(attenuation of the signals at 2.63 ppm and 2.77 ppm). Subsequently, volatiles were

evaporated using high vacuum. Crude residue was analyzed on HPLC.

1H NMR (400 MHz, CDCl3, δ): 7.65 (d, J = 7.7 Hz, 4H), 7.42 (m, 6H), 4.19 (m, 1H),

3.94 (m, 1H), 3.83 (m, 1H), 3.69 (m, 1H), 1.62 (dq, J = 15.0, 7.6 Hz, 2H), 1.06 (s, 9H),

0.93 (t, J = 7.3 Hz, 3H)

13C NMR (101 MHz, CDCl3, δ): 179.15, 135.51, 129.90, 127.81, 70.04, 64.90, 64.40,

35.99, 26.80, 26.72, 19.23, 18.37, 13.65.

HPLC (racemate, c = 1 mg/ml): CHIRACEL® OD-H 98:2 flow: 1 ml.min-1 t1=9.87

min t2=11.24 min.

HPLC (enantioenriched, c = 1 mg/ml): tmajor=10.03 min tminor= absent.

MS: calculated for C19H23O4Si [M-tBu]:343, found 343.

Synthesis of the platelet-activating factor (scheme 5).

(R)-1-((tert-butyldimethylsilyl)oxy)-3-(hexadecyloxy)propan-2-yl acetate (29)

Chapter 3

60

3

In a dry Schlenk flask, hexadecanol (242 mg, 1.0 mmol) was

dissolved in THF (0.4 ml). To this solution, (R)-TBDMS-glycidyl ether (390 µl, 1.9

mmol, 1.9 equiv) and Co[salen]OTs catalyst (20 mg, 25μmol, 2.5 mol%) were added.

The mixture was stirred for 3 days at rt (progress monitored by GC) until full

conversion of hexadecanol. The reaction mixture was diluted with anhydrous Et2O

(1.0 ml), cooled with an ice bath and subsequently DMAP (12.2 mg, 0.1 mmol,

10 mol%), Ac2O (240 µl, 2.0 mmol, 2.0 equiv) and Et3N (350 µl, 2.0 mmol, 2.0 equiv)

were added. The mixture was stirred for 16 h. Et2O was evaporated, the crude was

suspended in pentane, transferred onto a silica column and chromatographed with

5% Et2O in pentane to afford 321 mg of 29 (68% over 2 steps) as colorless liquid.

1H NMR (400 MHz, CDCl3, δ): 5.00 (m, 1H), 3.74 (m, 2H), 3.57 (m, 2H), 3.42 (m,

2H), 2.07 (s, 3H), 1.55 (s, 4H), 1.25 (s, 27H), 0.88 (s, 12H), 0.05 (s, 6H).

Shift at 1.55 ppm overlaps with water from the CDCl3.

13C NMR (101 MHz, CDCl3, δ): 170.62, 73.38 ( - ), 71.72, 68.96, 61.75, 32.06, 29.84,

29.82, 29.80, 29.76, 29.70, 29.60, 29.50, 26.20, 25.91 ( - ), 25.88, 22.83, 21.29 ( - ),

18.35, 14.25 ( - ), -5.31 ( - ).

HRMS (ESI+): calculated for C27H57O4Si [M+H]+: 473.401, found 473.402.

[α]D = +4.0 (c = 1.0, CHCl3).

Platelet-activating factor (33)

In a dry Schlenk flask, PAF precursor 29 (200 mg, 0.42

mmol) was dissolved in dry CH2Cl2 (4.2 ml). This solution was cooled to 0 °C in an

ice/water bath and treated with BF3.CH3CN (240 µl, 0.46 mmol, 1.1 equiv). The

reaction was closely monitored in the conversion of 29 (TLC), and after 20 min full

conversion was observed. The reaction was quenched by adding cooled phosphate

buffer (pH = 7, 1 M). The mixture was diluted with Et2O (20 ml), the organic layer

was washed with water (2 x 10 ml), and brine, dried and evaporated. The crude

residue was dried under high vacuum for 30 min and used without further

purification (1H NMR spectrum indicated no acyl migration).


61

3


2H), 2.11 (t, 3H), 1.55 (m, 30H), 0.88 (t, J = 6.7 Hz, 3H).

The residue was subsequently dissolved in THF (4.2 ml) and 2-chloro-1,3,2-

dioxaphospholan-2-oxide (154 µl, 1.7 mmol, 4.0 equiv) and iPr2NEt (300 µl,

1.7 mmol, 4.0 equiv) were added. The mixture was stirred overnight (ca 16 h). When

the reaction reached full conversion, the mixture was diluted with Et2O (25 ml). The

organic phase was washed with water, and brine, dried over MgSO4 and evaporated

to dryness. The crude was subsequently dissolved in CH2Cl2 (1.5 ml), and cooled in

an ice/water bath. The solution was treated with TMSOTf (140 µl, 0.84 mmol, 2.0

equiv). The addition was accompanied with a color change, first to brown then to

red). To this solution Me3N was added (55 µl, 0.63 mmol, 1.5 equiv). Given that

Me3N is a gas at RT, a syringe wrapped in cotton previously dipped in acetone/liquid

N2 was used. The reaction was monitored by TLC (disappearance of the spot with

Rf = 0.21 in Et2O). Upon full conversion all volatiles were evaporated. The crude

residue was transferred onto a silica column and carefully chromatographed on 150

Å Davisil silica gel using a gradient of 1% to 20% MeOH in CHCl3.

149 mg of PAF was obtained as a white waxy solid (71% over 3 steps). NMR spectra

of PAF were not indicative due to extensive peak-broadening.

HRMS (ESI+): calculated for C26H54NO7P [M+H]+: 524.362, found 524.366.

[α]D = +3.5 (c = 1.0, CHCl3).

Melting point: 240 °C (decomposition).


(1) (a) Simons, K.; Vaz, W. L. C. Annu. Rev. Biophys. Biomol. Struct. 2004, 33, 269 (b)

Simons, K.; Toomre, D. Nat. Rev. Mol. Cell Biol. 2000, 1, 31 (c) Cevc, G. Phospholipids

handbook; Marcel Dekker, Inc.: New York, 1993.

(2) Mintzer, M. A.; Simanek, E. E. Chem. Rev. 2008, 109, 259.

(3) Perozo, E.; Kloda, A.; Cortes, D. M.; Martinac, B. Nat. Struct. Mol. Biol. 2002, 9, 696.

(4) (a) Kodali, D. R.; Tercyak, A.; Fahey, D. A.; Small, D. M. Chem. Phys. Lipids 1990,

52, 163 (b) Crossley, A.; Freeman, I. P.; Hudson, B. J. F.; Pierce, J. H. J. Chem. Soc.,

Perkin Trans. 1959, 760.

(5) Martin, S. F.; Josey, J. A.; Wong, Y.-L.; Dean, D. W. J. Org. Chem. 1994, 59, 4805.

(6) Gras, J.-L.; Bonfanti, J.-F. Synlett 2000, 248.

(7) Massing, U.; Eibl, H. Chem. Phys. Lipids 1995, 76, 211.

Chapter 3

62

3

(8) Guanti, G.; Banfi, L.; Basso, A.; Bevilacqua, E.; Bondanza, L.; Riva, R. Tetrahedron:

Asymmetry 2004, 15, 2889.

(9) Burgos, C. E.; Ayer, D. E.; Johnson, R. A. J. Org. Chem. 1987, 52, 4973.

(10) Ali, S.; Bittman, R. J. Org. Chem. 1988, 53, 5547.

(11) Lindberg, J.; Ekeroth, J.; Konradsson, P. J. Org. Chem. 2001, 67, 194.

(12) (a) Stamatov, S. D.; Stawinski, J. Org. Biomol. Chem. 2007, 5, 3787 (b) Stamatov, S.

D.; Kullberg, M.; Stawinski, J. Tetrahedron Lett. 2005, 46, 6855.

(13) Jacobsen, E. N.; Kakiuchi, F.; Konsler, R. G.; Larrow, J. F.; Tokunaga, M.

Tetrahedron Lett. 1997, 38, 773.

(14) Kodali, D. R.; Duclos Jr, R. I. Chem. Phys. Lipids 1992, 61, 169.

(15) (a) Dodd, G. H.; Golding, B. T.; Ioannou, P. V. J. Chem. Soc., Chem. Commun. 1975,

249 (b) De Medeiros, E. F.; Herbert, J. M.; Taylor, R. J. K. J. Chem. Soc., Perkin Trans.

1991, 2725 (c) Buchnea, D. Lipids 1974, 9, 55.

(16) Wuts, P. G. M.; Greene, T. W. Greene's Protective Groups in Organic Synthesis, 4th Edition;

John Wiley & Sons: New Jersey, 2007.

(17) Pedersen, P. J.; Adolph, S. K.; Subramanian, A. K.; Arouri, A.; Andresen, T. L.;

Mouritsen, O. G.; Madsen, R.; Madsen, M. W.; Peters, G. N. H.; Clausen, M. H. J.

Med. Chem. 2010, 53, 3782.

(18) No 1,3-diacylglycerol is visible in the 1H-NMR spectrum. For a smooth

deprotection dry conditions are required.

(19) (a) Hayakawa, Y.; Kawai, R.; Hirata, A.; Sugimoto, J.-i.; Kataoka, M.; Sakakura, A.;

Hirose, M.; Noyori, R. J. Am. Chem. Soc. 2001, 123, 8165 (b) Rzepecki, P. W.;

Prestwich, G. D. J. Org. Chem. 2002, 67, 5454.

(20) Alternatively, an SN1 reaction might be considered. However, this also leads to lower

optical purity of the product.

(21) (a) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512 (b) Dale, J. A.; Dull, D.

L.; Mosher, H. S. J. Org. Chem. 1969, 34, 2543.

(22) (a) Guivisdalsky, P. N.; Bittman, R. J. Org. Chem. 1989, 54, 4637 (b) Guivisdalsky, P.

N.; Bittman, R. J. Org. Chem. 1989, 54, 4643.

(23) The TBDPS group was chosen because it is chromophoric.

(24) (a) Nakamura, N.; Miyazaki, H.; Ohkawa, N.; Oshima, T.; Koike, H. Tetrahedron Lett.

1990, 31, 699 (b) Guivisdalsky, P. N.; Bittman, R. J. Org. Chem. 1989, 54, 4643 (c)

Kertscher, H. P.; Ostermann, G. Pharmazie 1986, 41, 596 (d) Tsuri, T.; Kamata, S.

Tetrahedron Lett. 1985, 26, 5195 (e) Marx, M. H.; Wiley, R. A. Tetrahedron Lett. 1985,

26, 1379 (f) Ohno, M.; Fujita, K.; Nakai, H.; Kobayashi, S.; Inoue, K.; Nojima, S.

Chem. Pharm. Bull. 1985, 33, 572.

(25) (a) Ferrer, C.; Fodran, P.; Barroso, S.; Gibson, R.; Hopmans, E. C.; Sinninghe

Damsté, J.; Schouten, S.; Minnaard, A. J. Org. Biomol. Chem. 2013, 11, 2482 (b)

Venkatasubbaiah, K.; Zhu, X.; Kays, E.; Hardcastle, K. I.; Jones, C. W. ACS Catal.

2011, 1, 489.

(26) Gadek, T. R. Tetrahedron Lett. 1989, 30, 915.

(27) Pospíšil, J.; Markó, I. E. Tetrahedron Lett. 2006, 47, 5933.

63

Chapter 4 Enantiopure Triacylglycerols in Three Steps

Abstract: With the development of chromatography and spectrometry, the analysis

of complex mixtures of closely related triacylglycerols has come within reach in a

number of laboratories. Next to the sophisticated instrumentation, this “lipidomics”

heavily relies on the analytical standards of triacylglycerols. However, these are

limited and expensive and their synthesis requires multiple steps, protecting groups,

and the use of an excess of reagents. This chapter presents an efficient, 3 step

synthesis of enantiopure triacylglycerols, which can be used as analytical standards in

the analysis of complex triacylglycerol mixtures like milk fat.

Fodran, P.; Das, N.; Eisink, N.; Welleman, I.; Kloek, W.; Minnaard, A. J. manuscript in preparation.

Chapter 4

64

4

Introduction

If a di- or triacylglycerol bears 2 different acyl chains on the sn-1 and the sn-3

position, the compound is chiral, which means it has a non-superimposable mirror

image (enantiomer). This fact is frequently overlooked in research connected to di-

and triacylglycerols, although it was noted already in 1939.1,2 In fact, their

stereochemistry is rarely considered. On the contrary, racemic versus enantiopure

triacylglycerols show for example different crystallization behavior – and, translated

to daily life, this means that butter, margarine or chocolate leave different sensations

depending on their stereochemical composition.3 Another area where the

stereochemistry of triacylglycerols is frequently overlooked is the analysis of fats and

oils. During the past 70 years, a lot of effort has been invested into the analysis of

milk fat. These efforts invariably involved chromatographic separation of the

triacylglycerols and their subsequent identification, and has met with moderate

success. Milk fat is typically composed of triacylglycerols containing 9 different fatty

acids, which results in 93=729 possible isomers. From this number, 92=81

triacylglycerols are achiral, the remaining 648 are chiral. Because the fatty acids are

similar in structure, that is, only differ in their chain length and possible unsaturation,

most triacylglycerols are extremely difficult to separate by chromatography.

Furthermore, they show only negligible optical rotations (are cryptochiral).4 This

severely complicates the analysis of oils and fats and in particular milk fat. An

unexplored approach in this connection is the unambiguous synthesis of

triacylglycerols and their use as reference compounds in HPLC-MS or GC-MS.

This chapter describes the synthesis of some of the most abundant

triacylglycerols of milk fat in a collaboration with the Dutch diary company

FrieslandCampina.

Reported syntheses of triacylglycerols

Chemical modification of triacylglycerols to alter the properties of edible fats

was first described in the 1920s and established as an industrial process in the 1940s

in Germany.5 To improve the spreadability and baking properties, lard (pig fat) was

treated with sodium methoxide. On the molecular level (Figure 1), this results in a

random intra- and intermolecular redistribution of the fatty acid residues on glycerol.

Today, this process is known as fat randomization 6 and plays an important role in the

food industry.7

Enantiopure Triacylglycerols in Three Steps

65

4

Figure 1. Triacylglycerols formed during fat randomization.6a

A more defined triacylglycerol composition can be achieved by lipases. Betapol, which is a human breast milk mimic, contains 47% of fat mainly as 1,3-dipalmitoyl-2-oleyl-sn-glycerol (3). This triacylglycerol has beneficial effects on infants like softer stool and better absorption of fatty acids and calcium.

Scheme 1. Industrial production of Betapol.

The industrial production of the fat component of Betapol is based on enzymatic

interesterification of tripalmitoyl glycerol (1) (Scheme 1)8 with oleic acid. (2).

Although enzymes are powerful tools in modifying triacylglycerols, they give

enantioenriched products only in a very limited number of cases.9 Another limitation

of the enzymatic synthesis of triacylglycerols is that different fatty acid chain lengths

often require different lipases.

Chemical synthesis can provide enantiopure triglycerides. Kristinsson and

Haraldsson10,11 reported a 6-step approach (Scheme 2). (S)-Solketal 4 (from mannitol)

Chapter 4

66

4

was first benzylated and then hydrolyzed in acidic aqueous ethanol affording 5 in

87% yield. An sn-3 selective acylation of 5 by immobilized Candida antarctica lipase

(CAL) and vinyl stearate, followed by hydrogenolysis with Pd/C gave

monoacylglycerol 6 in 85% yield over 2 steps. A second esterification with the same

enzyme and vinyl capriate afforded 7 in 85% yield. Finally, esterification of 7 with

eicosapentaenoic acid (EPA) led to triacylglycerol 8 in 91% yield. A careful choice of

protecting groups and the use of a selective lipase circumvents acyl migration

(Chapter 3).

Scheme 2. Synthesis of enantiopure triacylglycerols according to Kristinsson and Haraldsson.

Stamatov and Stawinski12 presented a different strategy. Heating glycidyl ether 9

(Scheme 3-I) with an excess of trifluoroacetic anhydride and Bu4NI afforded the

iodohydrin, which after esterification in the same pot yielded 11 in 90%. Substitution

of iodide 10 by acetate yielded 11 in 92% yield. The TBDMSO- group in 11 was

directly converted into the ester, yielding triacylglycerol 12 in 90% yield. Based on

their NMR studies, the authors proposed a mechanism for the direct conversion of

11 to 12 (Figure 3-II). Treatment of palmitic anhydride with TMSBr yields 13 and

acid bromide 14. The silyl substituted oxygen of 11 attacks 14 resulting in 15.

Subsequently, bromide can attack 15 at silicon, thus affording 12 and TBDMSBr. In

case the authors applied 16 as starting material (Scheme 3-III), the silyl substitution

could be omitted, thus affording 17 in only 3 steps. In this approach, the application

of iodohydrines and direct substitution of TBDMSO- group for an ester circumvents

the acyl migration.


67

4

Scheme 3. ( I ) Synthesis of triacylglycerols according to Stamatov and Stawinski; ( II ) proposed mechanism of direct substitution of 11 to 12; ( III ) in case glycidyl esters are applied, the synthesis is one step shorter.

Despite both approaches avoid acyl migration and yield the triacylglycerols in high

yields, they have some drawbacks. The shorter approach presented by Stamatov and

Stawinski uses significant excess of the reagents (3 to 5 equiv), what is not

economical in case of expensive (i.e. polyunsaturated) fatty acids. The approach

presented by Kristinsson and Haraldsson uses nearly stoichiometric (1.1 – 1.25

equiv) conditions, but requires 2 protecting groups and 6 steps.

The previous chapter (chapter 3) showed that 1 mol% of (R,R)-N,N′-bis(3,5-

di-tert-butylsalicylidene)-1,2-cyclohexanediaminocobalt(II) catalyzes the ring opening

of glycidyl ethers with fatty acids under basic conditions. If glycidyl esters would

exhibit the same reactivity, they can afford triacylglycerols by the same procedure in

3 steps and without utilization of a large excess of reagents or protecting groups.

Chapter 4

68

4


Co[(R,R)-salen] catalyzed ring opening of glycidyl esters

Glycidyl esters 16a-g were prepared under mild Steglich conditions (Scheme

4). In the presence of DCC and catalytic amounts of DMAP, (S)-glycidol 15 was

conveniently esterified with a small excess of fatty acid (1.2 equiv).

Reagents and conditions: a) RCO2H (1.2 equiv), DCC (1.1 equiv), DMAP (10 mol%), pentane, 0 °C, 4 h.

Scheme 4. Esterification of (S)-glycidol.

Homologous glycidyl esters 16a-f (Table 1, entries 1-6) were obtained in high yields

(82-92%). In addition to the saturated derivatives, unsaturated glycidyl oleate 16g

was prepared in 81% yield (Table 1, entry 7).

Table 1. Prepared glycidyl esters.

Entry Fatty acid R: Yield

1 butyric acid C3H7-16a 86%

2 caproic acid C5H11-16b 92%

3 capric acid C9H19-16c 86%

4 myristic acid C13H27-16d 82%

5 palmitic acid C15H31-16e 92%

6 stearic acid C17H35-16f 88%

7 oleic acid C17H33 [18:1 cis-9]-16g 81%

In an initial experiment (Scheme 5), neat glycidyl myristate 16d was treated

with a stoichiometric amount of stearic acid in the presence of 17 (1 mol%) and

Hünigs base. These conditions resulted in quantitative formation of the 1,3-

diacylglycerol. Progress of this reaction (Scheme 5) was monitored by 1H NMR by

following the attenuation of the signals corresponding to the protons of the primary

epoxide carbon (δ = 2.63 and 2.83 ppm). As the reaction afforded ring-opened

product 18 exclusively, and after dilution with heptane the crude residue was

esterified in the same flask with stearic acid, DCC and DMAP. After full conversion

of 18, the reaction mixture was transferred directly onto a silica column and

chromatographed to afford analytically pure triacylglycerol 19 in 83% yield.


69

4

Reagents and conditions: a) stearic acid (1.0 equiv), Hünigs base (1.0 equiv), 17 (1.0 mol%), neat, 20 h, RT; b) stearic acid (1.2 equiv), DCC (1.1 equiv), DMAP (5.0 mol%) heptane, 20 h, RT.

Scheme 5. Triacylglycerols, prepared according to the described method.

The scope of this two-step-one-pot procedure was further investigated with

various acids for ring opening and final esterification (Table 2). The synthetized

triacylglycerols were obtained in yields in the range of 79-92% over two steps. The

obtained yields are similar or higher than those described by Kristinsson and

Haraldsson, and by Stamatov and Stawinski.

Table 2. Synthetized triacylglycerols.

entry

-label

starting glycidyl ester

(sn-3 position)

Fatty acid used in ring opening

(sn-1 position)

Fatty acid in final esterification

(sn-2 position)

Yield

1-19a glycidyl butyrate 16a myristic acid oleic acid 86%

2-19b glycidyl butyrate 16a stearic acid palmitic acid 79%

3-19c glycidyl butyrate 16a oleic acid stearic acid 86%

4-19d glycidyl butyrate 16a oleic acid oleic acid 83%

5-19e glycidyl butyrate 16a linoleic acid myristic acid 85%

6-19f glycidyl caproate 16b oleic acid palmitic acid 82%

7-19g glycidyl caprate 16c oleic acid palmitic acid 88%

8-19h glycidyl palmitate 16e myristic acid stearic acid 80%

9-19i glycidyl palmitate 16e stearic acid myristic acid 79%

10-19j glycidyl palmitate 16e oleic acid oleic acid 92%

11-19k glycidyl stearate 16f butyric acid palmitic acid 82%

12-19l glycidyl stearate 16f myristic acid palmitic acid 80%

13-19m glycidyl stearate 16f oleic acid oleic acid 80%

14-19n glycidyl oleate 16g palmitic acid stearic acid 80%

Chapter 4

70

4

Furthermore, the fatty acids are used in stoichiometric or nearly stoichiometric

amounts. Compared to the enzymatic synthesis of triacylglycerols, the developed

reaction conditions tolerate different chain lengths in all three positions (Table 2). A

pair of isomeric triacylglycerols (entry 8 vs. 9) was prepared in very similar yields. The

unsaturated oleic acid is tolerated on all 3 (sn-1, sn-2, sn-3) positions (entries 1, 3, 4,

6, 7, 10, 13, 14). The double bonds of the polyunsaturated linoleic acid (18:2 cis,cis-

9,12, entry 5) remained intact. And finally, these conditions allow the synthesis of a

pair of enantiomers (entry 2 and 11) both from (S)-glycidol 16, if the fatty acids are

introduced in reversed order.

17 was initially developed for the kinetic resolution of various terminal

epoxides including glycidyl esters and ethers. This raised the question whether 17

could also catalyze the kinetic resolution of glycidyl esters with fatty acids.

Reagents and conditions: a) ent-17 (5.0 mol%), Hünigs base (0.5 equiv), palmitic acid (0.5 equiv), THF, 0 °C, 18 h.

Scheme 6. Kinetic resolution of 20.

When racemic 20 (Scheme 6) was reacted with palmitic acid in the presence of ent-

17 (5.0 mol%) and Hünigs base, the reaction went to full conversion (in palmitic

acid). After tedious work-up and purification by flash chromatography, recovered

stearate 22 showed only marginal enantioenrichment (55 : 45 e.r.).

Towards the automated synthesis of triacylglycerols

In case triacylglycerols are required as analytical standard in the analysis of

(milk) fat, the reported 15 triacylglycerols (Table 2) correspond to only 2% of the

required standards.13 The time required for the synthesis of the remaining 98% of

the triacylglycerols may be reduced by the development of an automated protocol.

Commercially available liquid-handling platforms for parallel synthesis can perform

up to 96 reactions in the same batch, but the options for work-up and purification

are limited. In order to utilize this equipment to maximum extent, the two-step-one-

pot protocol has to be modified in such a way that pure product is obtained by

filtration. The two-step-one-pot protocol consists of ring-opening and esterification.

Various esterification methods leave the second step with multiple options. On the


71

4

contrary, the first step is limited to 17 and Hünigs base. Hünigs base can be simply

evaporated. Then the removal of 15, which is soluble in a variety of organic solvents,

is the main challenge.

The wide application and excellent results in the numerous transformations

catalyzed by salen complexes made these catalysts attractive candidates for

immobilization.14 After covalent attachment to a support, the catalyst can be readily

removed from the reaction and recycled. Jacobsen et al.14a reported the synthesis and

application of modified salen ligand 27, which was immobilized on polystyrene beads

and converted into the corresponding cobalt catalyst.

Scheme 7. ( I ) Synthesis of modified salen ligand 27; ( II ) synthesis of aldehyde 26 as reported by Jacobsen et al.

The immobilization had no influence on the rate of the hydrolytic kinetic resolution

of terminal epoxides and the diol was obtained with similar enantiopurity after 4

times recycling of the catalyst. The authors prepared ligand 27 by condensation of

the aldehydes 25 and 26 with diamine 24 (Scheme 7). While diamine 24 and aldehyde

25 are commercially available, aldehyde 26 is not, but the same authors14a reported

its synthesis. First, diol 28 was monosilylated with TIPS-Cl on the sterically less

hindered 4-hydroxyl group. Then, the TIPS protected substrate was formylated with

paraformaldehyde in the presence of SnCl4. The final deprotection with TBAF

affords aldehyde 26. Even though Jacobsen et al.14a described the synthesis of

aldehyde 26, alternative routes avoiding toxic SnCl4 were explored. Formylation

according to Vilsmeier and Haack, Duff, and Hofsløkken and Skattebøl15,16 were

considered as suitable options (Scheme 8). The latter had already been reported17 for

the synthesis of 25 in a one-pot synthesis of the salen ligand, though not for the

Chapter 4

72

4

formylation of 28. These reactions were studied on benzyl protected substrate 29

(Scheme 8), which was obtained by benzylation of 28 followed by separation from

its bis-benzylated analogue.

Reagents and conditions: a) POCl3, DMF, CH2Cl2 b) hexamine (2.0 equiv), AcOH, reflux 4 h; c) MgCl2 (2.0 equiv), paraformaldehyde (3.0 equiv), Et3N (2.0 equiv), MeCN, reflux, 20 h; d) 1,4-cyclohexadiene (10 equiv), Pd/C (10 mol%), EtOH, 21 °C, 24 h.

Scheme 8. ( I ) Exploring the formylation of 29, ( II ) the mechanism of the MgCl2 mediated

formylation and formation of the side-product 32.

When 29 reacted with POCl3 and DMF (Scheme 8-I), only undesired O-formylated

30 was formed. An attempt to convert 30 into the desired C-formylated 31 by Fries


73

4

rearrangement18 in the presence of BCl3 resulted only in decomposition of the

starting material. On the other hand, Duff reaction of 30 gave the desired aldehyde

31 in a moderate 52% yield. Finally, the formylation according to Hofsløkken and

Skattebøl15 using an excess of paraformaldehyde and anhydrous MgCl2 in basic

conditions was the best alternative affording 31 in a good 72% yield. Occasionally,

traces of methoxymethylated product 32 were formed as a side product. The

mechanism of this formylation is depicted in scheme (Scheme 8-II). Under basic

conditions, phenol 29 is deprotonated affording phenoxymagnesium chloride (33).

33 reacts with formaldehyde (from paraformaldehyde) resulting in intermediate 34.

This can subsequently undergo 2 different pathways.This can subsequently undergo

2 different pathways. Oxidation by another formaldehyde molecule (in blue) results

in the formylated 35, which after hydrolysis, affords the desired phenol 31.

Alternatively, 34 might form quinone methide intermediate 36 (depicted in red). 36

can undergo a Michael addition with methanol(ate), affording the undesired 32.

Despite 31 and 32 were inseparable by column chromatography, 32 could be readily

removed by trituration with cold pentanes. Finally, 31 was debenzylated (Scheme 8-

I). From the explored hydrogenolysis conditions a combination of Pd/C and 1,4-

cyclohexadiene19 was the best, affording the desired aldehyde 26 in 95% yield.

With all the building blocks in hand, the modified Jacobsen ligand 27 was

prepared according to the previously published procedure.14a Condensation of

aldehydes (Scheme 7-I, see above) 25 and 26 with (1R,2R)-cyclohexane-1,2-diamine

Figure 2. Mixture of the three ligands used for immobilization to poly-4-hydroxystyrene.

(24) resulted in a statistical mixture of ligands 27, 37, and 38 in a ratio of 6 : 1 : 920

(Figure 2). Given that 37 was present in a small amount (6%) and 38 does not react

in the immobilization step14a, the mixture was used without any purification.

Finally, 27 was immobilized on polystyrene beads. Poly-4-hydroxystyrene

beads (Scheme 9) were treated with 4-nitrophenyl chloroformate. After washing, the

incorporation of the 4-nitrophenyl carbonate was confirmed by IR spectroscopy,

Chapter 4

74

4

which showed 2 new bands corresponding to a carbonate stretch vibration (υ = 1765

cm-1) and an asymmetric stretch21 vibration of the nitro group (υ= 1528 cm-1).

Reagents and conditions: a) 4-nitrophenyl chloroformate (2.0 equiv), DMAP (50 mol%), CH2Cl2, b) 27, 37, 38, Hünigs base (1.0 equiv), DMAP (50 mol%), c) Co(OAc)2.4H2O, MeOH/toluene, air.

Scheme 9. Immobilization of the ligand on poly-4-hydroxy styrene support.

The functionalized polymer 39 was treated with the mixture of the ligands 27, 37,

38. After washing, the band corresponding to the asymmetric stretch vibration of

the nitro group (υ = 1528 cm-1) had disappeared and the band corresponding to the

carbonate stretch vibration had shifted to lower frequency (υ = 1745 cm-1).

Furthermore, a new band corresponding to the imine stretch vibration of the ligand

(υ = 1630 cm-1) appeared in the IR spectrum. Immobilized ligand 40 was treated with

Co(OAc)2.4H2O and exposed to air. In the IR spectrum, the band corresponding to

the imine (υ = 1630 cm-1) had disappeared. The elemental analysis indicated

incorporation of 0.83 mmol of Co3+[salen]OAc per gram of resin, based on the

nitrogen content.

The immobilized cobalt catalyst 41 was applied in the triacylglycerol synthesis

(Scheme 10). Glycidyl butyrate 16a afforded the corresponding diacylglycerol 36 in

the presence of oleic acid, Hünigs base and the immobilized catalyst (10 mol%) after

6 days. As envisioned, the salen catalyst was removed by filtration and the

diacylglycerol was esterified with palmitic acid under Steglich conditions to afford

19c in 80% yield. The yield of the reaction with the immobilized catalyst 41 is

comparable to the reaction with free catalyst 17. In the latter case, the same


75

4

triacylglycerol was obtained in 86% yield. While Jacobsen et al. 14a reported no

influence of immobilization on the rate of hydrolytic kinetic resolution, in the case

of ring opening with carboxylic acids the effect was significant. Typically, the glycidyl

ester was converted to product in 16 h with non-immobilized catalyst but needed 6

days to be fully converted with the immobilized catalyst.

Reagents and conditions: a) oleic acid (1 equiv), Hünigs base (1 equiv), 35 (10 mol%) 6 days, then filtration b) palmitic acid (1.2 equiv), DCC (1.2 equiv), DMAP (5 mol%), heptane.

Scheme 10. Synthesis of triacylglycerol 19c with immobilized Co[salen] catalyst 41.

Conclusions

This chapter presents a novel access to triacylglycerols. The products are

obtained in a straightforward two-step-one-pot procedure in 79-92% yields. Double

bonds tolerate the mild conditions and no side products are formed. Furthermore,

this chapter presents a step towards the robot-operated synthesis of triacylglycerols.

For this purpose, an immobilized catalyst 35 was synthetized according to a literature

procedure. A synthesis of one of the building blocks for this catalyst was improved.

The immobilized catalyst was also applied in a synthesis of a triacylglycerol. Despite

a significantly longer reaction time, the desired triacylglycerol was isolated in 80%

yield. With these tools now in hand, the entire identification of the triacylglycerols in

milk fat has come a significant step closer

Experimental part

The configuration of triacylglycerols is presented using the Hirschmann’s nomenclature22 (see introduction).

General procedure for the synthesis of glycidyl esters 17a-17g (Scheme 4; Table 1)

A dried round bottom flask was charged with the fatty acid (12.0 mmol, 1.2 equiv),

and DMAP (122 mg, 10 mol%). These were dissolved in heptane (100 ml) and cooled

to 0 °C (ice bath). After stirring for 10 min, (S)-glycidol (670 μl, 10.0 mmol) and

DCC (2.24 g, 11.0 mmol, 1.1 equiv) were added, and the resulting mixture was stirred

for 4 h. A precipitate (dicyclohexyl urea) formed within 5 min. After full conversion

Chapter 4

76

4

of (S)-glycidol (TLC, anisaldehyde stain), the white precipitate was filtered off and

washed with pentane. The combined filtrates were concentrated in vacuo and the

resulting crude residue was purified by column chromatography on silica gel using

20% Et2O in pentanes to afford the desired glycidyl ester.

Glycidyl butyrate (17a)

Following the general procedure for the synthesis of glycidyl esters

with butyric acid. The reaction afforded 1.61 g of glycidyl butyrate as

pale yellow liquid (86% yield)

1H NMR (400 MHz, CDCl3, δ): 4.30 (dd, J = 12.3, 3.0 Hz, 1H), 3.81 (dd, J = 12.3,

6.3 Hz, 1H), 3.10 (m, 1H), 2.74 (dd, J = 11.6, 7.2 Hz, 1H), 2.54 (dd, J = 4.9, 2.6 Hz,

1H), 2.23 (t, J = 7.6 Hz, 2H), 1.65 – 1.49 (m, 2H), 0.85 (t, J = 7.4 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 173.07, 64.59, 49.20 ( - ), 44.42, 35.73, 18.21, 13.45

( - ).

HRMS-ESI+ (m/z): [M + H]+ calculated for C7H13O3 , 145.085; found: 145.087.

Anal. Calcd for C7H12O3: C, 58.32; H, 8.39. Found: C 58.45; H 8.53.

[α]D –24.5° (c = 1.0, CHCl3).

Glycidyl caproate (17b)

Following the general procedure for the synthesis of glycidyl esters

with caproic acid. The reaction afforded 1.58 g of glycidyl caproate as pale yellow

liquid (92% yield).


3.2 Hz, 1H), 3.19 – 3.04 (m, 1H), 2.76 (dd, J = 8.2, 4.3 Hz, 1H), 2.57 (dd, J = 7.8, 2.9

Hz, 1H), 2.28 (t, J = 7.6 Hz, 2H), 1.43 (m, 6H), 0.83 (t, J = 6.9 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 173.31, 64.64, 49.25 ( - ), 44.46, 33.89, 31.16, 24.45,

22.19, 13.76 ( - ).

HRMS-ESI+ (m/z): [M + H]+ calculated for C9H17O3 , 173.117; found, 173.117.


77

4

Anal. Calcd for C9H16O3: C, 62.77; H, 9.36. Found: C, 63.04; H 9.51.

[α]D –27.0° (c = 1.0, CHCl3).

Glycidyl caprate (17c)

Following the general procedure for synthesis of glycidyl esters with

capric acid. Reaction afforded 2.17 g of glycidyl caprate as colourless liquid (86%

yield).


6.3 Hz, 1H), 3.15 (m, 1H), 2.80 (dd, J = 5.8, 3.3 Hz, 1H), 2.60 (dd, J = 4.9, 2.6 Hz,

1H), 2.31 (t, J = 7.6 Hz, 2H), 1.43 (m, 14H), 0.84 (t, J = 6.8 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 173.42, 64.68, 49.31 ( - ), 44.56, 34.00, 31.80, 29.34,

29.24, 29.19, 29.05 24.82, 22.60, 14.03 ( - ).


Anal. Calcd for C13H24O3: C, 68.38; H, 10.59. Found: C, 68.58; H, 10.70.

[α]D –19.8° (c = 1.0, CHCl3).

Glycidyl myristate (17d)

Following the general procedure for synthesis of the glycidyl esters

with myristic acid. The reaction afforded 2.40 g of glycidyl myristate as white solid

(82% yield).


6.3 Hz, 1H), 3.20 (ddd, J = 6.4, 4.1, 3.0 Hz, 1H), 2.84 (dd, J = 4.9, 4.2 Hz, 1H), 2.64

(dd, J = 4.9, 2.6 Hz, 1H), 2.35 (t, J = 7.6 Hz, 2H), 1.45 (m, 22H), 0.88 (t, J = 6.9 Hz,

3H).

13C NMR (101 MHz, CDCl3, δ): 173.68, 64.87, 49.53 ( - ), 44.80, 34.21, 32.05, 29.81,

29.78, 29.73, 29.58, 29.49, 29.38, 29.26, 25.01, 22.82, 14.25 ( - ). (2 signals are

overlapping)

Chapter 4

78

4



[α]D –18.6° (c = 1.0, CHCl3).


Glycidyl palmitate (17e)


with palmitic acid. The reaction afforded 3.03 g of glycidyl palmitate as white solid

(92% yield)

1H NMR (400 MHz, CDCl3). δ 4.41 (dd, J = 12.3, 3.1 Hz, 1H), 3.91 (dd, J = 12.3,

6.3 Hz, 1H), 3.19 (m, J = 6.3, 4.1, 3.0 Hz, 1H), 2.84 (dd, J = 4.8, 4.2 Hz, 1H), 2.64

(dd, J = 4.9, 2.6 Hz, 2H), 2.35 (t, J = 7.6 Hz, 2H), 1.41 (m, 26H), 0.88 (t, J = 6.8 Hz,

3H).

13C NMR (101 MHz, CDCl3, δ): 173.70, 64.89, 49.55 ( - ), 44.82, 34.24, 32.08, 29.85,

29.84, 29.83, 29.81, 29.75, 29.60, 29.51, 29.40, 29.28, 25.04, 22.85, 14.28 ( - ). (2

signals are overlapping)


Anal. Calcd for C19H36O3: C, 73.03; H 11.61. Found: C, 73.21; H, 11.65.

[α]D –9.4° (c = 1.0, CHCl3).

Melting point: 48 °C

Glycidyl stearate (17f)


with stearic acid. The reaction afforded 3.41 g of glycidyl stearate as white solid (88%

yield).


79

4


6.3 Hz, 1H), 3.20 (ddd, J = 6.4, 4.1, 3.0 Hz, 1H), 2.84 (dd, J = 4.9, 4.2 Hz, 1H), 2.64

(dd, J = 4.9, 2.6 Hz, 1H), 2.35 (t, J = 7.6 Hz, 2H), 1.70 – 1.20 (m, 30H), 0.88 (t,

J = 6.9 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 173.71, 64.89, 49.55 ( - ), 44.82, 34.22, 32.08, 29.85,

29.82, 29.81, 29.80, 29.75, 29.60, 29.52, 29.40, 29.27, 25.02, 22.85, 14.28 ( - ).

HRMS-APCI+ (m/z): [M + H]+ calculated for C21H41O3, 341.305; found, 341.305.


[α]D –11.2° (c = 1.0, CHCl3)


Glycidyl oleate (17g)

Following the general procedure for synthesis of glycidyl esters with

stearic acid. Reaction afforded 3.41 g of glycidyl oleate as white

solid (81% yield)


(dd, J = 12.3, 6.3 Hz, 1H), 3.19 (ddd, J = 6.4, 4.1, 2.9 Hz, 1H), 2.82 (m, 1H), 2.63

(dd, J = 4.9, 2.6 Hz, 1H), 2.34 (t, J = 7.6 Hz, 2H), 1.99 (m, 4H), 1.61 (dd, J = 14.7,

7.3 Hz, 2H), 1.27 (m, 16H), 0.87 (t, J = 6.8 Hz, 3H).

13C NMR (101 MHz, CDCl3, δ): 173.54, 130.05 ( - ), 129.78 ( - ), 64.83, 49.44 ( - ),

44.70, 34.11, 31.99, 29.84, 29.76, 29.61, 29.40, 29.23, 29.16, 27.29, 27.23, 24.93, 22.77,

14.20 ( - ).



α = +2.0° (c = 1.0, CHCl3).

Chapter 4

80

4

Synthesis of triacylglycerols (Scheme 5, Table 2):

A roundbottom flask was charged with cobalt catalyst 15 catalyst (6.2 mg, 1.0 mol%)

and fatty acid (1.0 mmol, 1.0 equiv). A small amount of Et2O was added to dissolve

the compounds (ca 2 ml). This solution was stirred for ca 15 min exposed to air.

After the solution turned from red to brown (due to oxidation of Co2+ to Co3+), Et2O

was evaporated. To this residue, DIPEA (170 µl, 1.0 mmol, 1.0 eq) and glycidyl ester

(1.0 mmol) were added in this order and the reaction was stirred overnight (ca 16 h)

at RT (21 °C). Conversion was monitored by 1H NMR spectroscopy of the crude

reaction mixture aliquots (attenuation of the signals at 2.64 ppm and 2.84 ppm). After

full conversion (reaction mixture turns back to red), the crude residue was dissolved

in n-heptane (10 ml), the second fatty acid (1.2 equiv), DMAP (12.2 mg, 10 mol%)

and DCC (227 mg, 1.1 mmol, 1.1 equiv) were added in this order. The reaction

mixture was stirred overnight (ca 20 h) at RT (21 °C). After full conversion of the

intermediate 1,2-diacylglycerol (TLC), the crude mixture was transferred directly

onto a silica column and chromatographed with 9% Et2O in pentane to afford the

desired triacylglycerol.

NOTE: The measured optical rotations for triacylglycerols were below the precision

of the instrument (<0.005°)

1-myristoyl-2-oleoyl-3-butyryl- sn-glycerol (19a)

Following the general procedure for synthesis of triacylglycerols,

Myristic acid was used for ring opening of glycidyl butyrate and

oleic acid was used for the final esterification. Reaction afforded

544 mg of desired product as pale yellow liquid (86%)

1H NMR (400 MHz, CDCl3, δ): 5.35 (m, 2H), 5.25 (m, 1H), 4.23 (dd, J = 11.9, 4.3

Hz, 2H), 4.08 (dd, J = 11.9, 6.0 Hz, 2H), 2.30 (m, 6H), 2.00 (m, 4H), 1.61 (m, 6H),

1.26 (m, 40H), 0.94 (t, J = 7.4, 3H), 0.88 (t, J = 6.8 Hz, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.30, 173.12, 172.92, 130.08 ( - ), 129.79 ( - ), 68.98

( - ), 62.18, 36.01, 34.31, 34.14, 34.13, 32.04, 32.02, 29.88, 29.81, 29.78, 29.74, 29.64,

29.60, 29.48, 29.43, 29.40, 29.28, 29.22, 29.19, 29.18, 27.33, 27.27, 25.01, 24.97, 24.95,

22.80, 18.45, 14.21 ( - ), 13.71 ( - ). (4 signals are overlapping)




81

4

1-stearoyl-2-palmitoyl-3-butyryl-sn-glycerol (19b)

Following the general procedure for synthesis of

triacylglycerols, stearic acid was used for ring opening of

glycidyl butyrate and palmitic acid was used for the final

esterification. The reaction afforded 523 mg of desired product as white solid (79%)


(dd, J = 11.9, 4.4 Hz, 2H), 2.31 (m, 6H), 1.65 (m, 6H), 1.26 (m, 52H), 0.94 (t,

J = 7.4 Hz, 3H), 0.88 (t, J = 6.8 Hz, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.28, 173.07, 172.86, 68.85 ( - ), 62.07, 36.04,

35.89, 34.20, 34.03, 31.91, 29.68, 29.66, 29.64, 29.62, 29.61, 29.59, 29.53, 29.52, 29.47,

29.45, 29.34, 29.27, 29.25, 29.24, 29.10, 29.09, 29.06, 24.88, 24.84, 22.67, 18.35, 18.32,

14.10 ( - ), 13.59 ( - ), 13.53( - ) (6 signals are overlapping).



Melting point 52 – 56 °C

1-oleoyl-2-palmitoyl-3-butyryl-sn-glycerol (19c)


triacylglycerols, oleic acid was used for ring opening of

glycidyl butyrate and palmitic acid was used for the final

esterification. Reaction afforded 575 mg of desired product as

white solid (86%)


Hz, 2H), 4.13 (dd, J = 11.9, 6.0 Hz, 2H), 2.28 (m, 6H), 2.01 (m, 4H), 1.62 (td, J =

11.4, 5.7 Hz, 6H), 1.26 (m, 44H), 0.93 (t, J = 7.4 Hz, 3H), 0.86 (t, J = 6.8 Hz, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.15, 172.97, 172.78, 129.94 ( - ), 129.64 ( - ), 68.84

( - ), 62.03, 36.02, 35.86, 34.16, 33.98, 31.90, 29.73, 29.67, 29.50, 29.46, 29.33, 29.29,

29.25, 29.14, 29.08, 29.05, 27.18, 27.13, 24.86, 24.81, 22.66, 18.30, 14.07 ( - ),

13.57 ( - ) ppm. (11 signals are overlapping)


Chapter 4

82

4


1-oleoyl-2-oleoyl-3-butyryl-sn-glycerol (19d)

Following the general procedure for synthesis of triacylglycerols,

oleic acid was used for ring opening of glycidyl butyrate and oleic

acid was used for the final esterification. Reaction afforded

575 mg of desired product as pale yellow liquid (83%).



1.27 (m, 40H), 0.93 (t, J = 7.4 Hz, 3H), 0.87 (t, J = 6.7 Hz, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.18, 172.99, 172.77, 129.97 ( - ), 129.95 ( - ),

129.68 ( - ), 129.66 ( - ), 68.85 ( - ), 62.05, 35.84, 34.19, 33.99, 31.81, 29.74, 29.63,

29.61, 29.58, 29.50, 29.41, 29.36, 29.25, 29.21, 29.16, 29.05, 29.03, 27.19, 27.14, 27.11,

27.09, 24.80, 22.65, 18.31, 14.07 ( - ), 13.58 ( - ) (12 signals are overlapping).

HRMS-ESI+ (m/z): [M + H]+ calculated for C43H79O6 , 691.587; found, 691.587.


1-oleoyl-2-oleoyl-3-butyryl-sn-glycerol (19e)


triacylglycerols, linoleic acid was used for ring opening of

glycidyl butyrate and myristic acid was used for the final

esterification. Reaction afforded 541 mg of desired product

as pale yellow liquid (85%)


Hz, 2H), 4.14 (dd, J = 11.9, 6.0 Hz, 2H), 2.76 (t, J = 6.4 Hz, 2H), 2.30 (m, 6H), 2.04

(q, J = 6.7 Hz, 4H), 1.65 (m, 6H), 1.27 (m, 34H), 0.94 (t, J = 7.4, 3H), 0.87 (t,

J = 3.4 Hz, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.18, 173.01, 172.81, 130.15, 129.95 ( - ), 128.02

( - ), 127.86 ( - ), 68.84 ( - ), 62.04, 36.03, 35.88, 34.18, 34.01, 33.89, 31.50, 29.65,

29.62, 29.60, 29.58, 29.45, 29.33, 29.25, 29.23, 29.16, 29.09, 29.05, 27.17, 25.60, 24.87,

24.83, 22.66, 22.54, 22.31, 18.35, 14.08 ( - ), 13.58 ( - ) (3 signals are overlapping).


83

4



1-oleoyl-2-palmitoyl-3-caproyl-sn-glycerol (19f)


triacylglycerols, oleic acid was used for ring opening of glycidyl

caproate and palmitic acid was used for the final esterification.

Reaction afforded 565 mg of desired product as pale yellow

liquid (82%).



(m, 6H), 1.27 (m, 48H), 0.85 (t, J = 6.9 Hz, 9H).

13C NMR (101 MHz, CDCl3, δ): 173.19, 173.17, 172.79, 129.95 ( - ), 129.65 ( - ), 68.84

( - ), 62.06, 34.18, 33.99, 33.97, 31.90, 31.21, 29.74, 29.67, 29.63, 29.50, 29.47, 29.34,

29.29, 29.26, 29.15, 29.08, 29.06, 29.00, 27.19, 27.14, 24.87, 24.81, 24.50, 22.66, 22.26,

14.08 ( - ), 13.85 ( - ) (10 signals are overlapping).



1-oleoyl-2-palmitoyl-3-capryl-sn-glycerol (19g)



caprate and palmitic acid was used for the final esterification.


liquid (88%)



(m, 6H), 1.27 (m, 56H), 0.87 (t, J = 6.8 Hz, 9H).

13C NMR (101 MHz, CDCl3, δ): 173.09, 173.07, 172.70, 129.90 ( - ), 129.61 ( - ),

68.83, 62.02, 34.14, 33.98, 33.96, 31.89, 31.87, 31.83, 29.73, 29.67, 29.63, 29.60, 29.58,

Chapter 4

84

4

29.49, 29.44, 29.33, 29.29, 29.26, 29.23, 29.13, 29.07, 29.04, 27.17, 27.12, 24.86, 24.82,

24.80, 22.64, 22.62, 14.05 ( - ) (12 signals are overlapping).



1-myristoyl-2-stearoyl-3-palmitoyl-sn-glycerol (19h)


triacylglycerols, myristic acid was used for ring opening of

glycidyl palmitate and stearic acid was used for the final

esterification. Reaction afforded 730 mg of desired product as white solid (80%).

NOTE: small amount of THF (ca 0.5 ml) is necessary to dissolve the starting

materials

1H NMR (400 MHz, CDCl3, δ): 5.25 (m, 1H), 4.29 (dd, J = 11.9, 4.3, 2H), 4.14 (dd,

J = 11.9, 6.0, 2H), 2.29 (m, 6H), 1.63 (m, 6H), 1.27 (m, 72H), 0.88 (t, J = 6.8, 9H).

13C NMR (101 MHz, CDCl3, δ): 173.28, 172.86, 68.84 ( - ), 62.08, 34.21, 34.04, 31.91,

29.69, 29.67, 29.65, 29.61, 29.50, 29.47, 29.43, 29.35, 29.28, 29.26, 29.10, 29.07, 24.89,

24.85, 22.67, 14.10 ( - ) (28 signals overlapping).




1-stearoyl-2-myristoyl-3-palmitoyl-sn-glycerol (19i)


triacylglycerols, stearic acid was used for ring opening of

glycidyl palmitate and myristic acid was used for the final

esterification. Reaction afforded 720 mg of desired product as white solid (79%)

NOTE: small amount of THF (ca 0.5 ml) is necessary for dissolving the starting

materials


85

4


J = 11.9, 6.0, 2H), 2.31 (td, J = 7.6, 2.4, 6H), 1.62 (m, 6H), 1.26 (m, 72H), 0.88 (t, J

= 6.8, 9H).

13C NMR (101 MHz, CDCl3, δ): 173.28, 173.28, 172.86, 68.85 ( - ), 62.08, 34.21,

34.04, 31.91, 29.69, 29.65, 29.61, 29.49, 29.47, 29.35, 29.28, 29.26, 29.11, 29.07, 24.89,

24.85, 22.68, 14.10 ( - ) (29 signals are overlapping).




1-oleoyl-2-oleoyl-3-palmitoyl-sn-glycerol (19j)



palmitate and myristic acid was used for the final esterification.


liquid (92%)


Hz, 2H), 4.13 (dd, J = 11.9, 4.3 Hz, 2H), 2.29 (td, J = 7.6, 2.2 Hz, 6H), 2.01 (m, 8H),

1.60 (m, 6H), 1.26 (m, 64H), 0.86 (t, J = 6.8 Hz, 9H).

13C NMR (101 MHz, CDCl3, δ): 173.13, 173.10, 172.70, 129.94 ( - ), 129.92 ( - ),

129.63 ( - ), 129.60 ( - ), 68.85 ( - ), 62.03, 29.74, 29.64, 29.51, 29.45, 29.32, 29.27,

29.17, 29.09, 29.05, 27.15, 25.59, 24.83, 22.65, 22.31, 14.06 ( - ), 14.02 ( - ) (31 signals

overlapping).

HRMS-ESI+ (m/z): [M + H]+ calculated for C55H103O6, 859.774; found: 859.776


1-butyryl-2-palmitoyl-3-stearoyl-sn-glycerol (19k)


triacylglycerols, butyric acid was used for ring opening of

Chapter 4

86

4

glycidyl stearate and palmitic acid was used for the final esterification. Reaction

afforded 546 mg of desired product as white solid (82%)


materials


J = 11.9, 6.0, 1.3, 2H), 2.34 (m, 6H), 1.64 (m, 6H), 1.25 (m, 54H), 0.95 (t, J = 6.1,

3H), 0.88 (t, J = 6.8, 6H).

13C NMR (101 MHz, CDCl3, δ): 173.28, 173.07, 172.86, 68.85 ( - ), 62.07, 36.04,

35.89, 34.20, 34.03, 31.91, 29.68, 29.66, 29.64, 29.62, 29.61, 29.59, 29.53, 29.52, 29.47,

29.45, 29.34, 29.27, 29.25, 29.24, 29.10, 29.09, 29.06, 24.88, 24.84, 22.67, 18.35, 18.32,

14.10 ( - ), 13.59 ( - ), 13.53 ( - ) (6 signals are overlapping).




1-myristoyl-2-palmitoyl-3-stearoyl-sn-glycerol (19l)


triacylglycerols, myristic acid was used for ring opening of

glycidyl stearate and palmitic acid was used for the final

esterification. Reaction afforded 546 mg of desired product as white solid (80%)


materials


J = 11.9, 6.0, 2H), 2.32 (td, J = 7.6, 2.4, 6H), 1.60 (m, 6H), 1.28 (m, 72H), 0.88 (t, J

= 6.8, 9H).

13C NMR (101 MHz, CDCl3, δ): 173.56, 173.30, 172.88, 68.86 ( - ), 62.10, 49.40,

44.67, 34.23, 34.09, 34.06, 31.93, 29.71, 29.67, 29.65, 29.63, 29.60, 29.51, 29.49, 29.45,

29.37, 29.30, 29.28, 29.25, 29.13, 29.09, 24.92, 24.87, 22.70, 14.12 ( - ) (22 signals

overlapping)


87

4

HRMS-ACPI (m/z): [M + Na]+ calculated for C51H98O6Na, 829.725; found, 829.725.



1-oleoyl-2-oleoyl-3-stearoyl-sn-glycerol (19m)



stearate and oleic acid was used for the final esterification.


liquid (80%)



1.26 (m, 68H), 0.85 (t, J = 6.8 Hz, 9H).

13C NMR (101 MHz, CDCl3, δ): 173.13, 173.10, 172.70, 129.94 ( - ), 129.92 ( - ),

129.63 ( - ), 129.60 ( - ), 68.85 ( - ), 62.03, 29.74, 29.64, 29.51, 29.48, 29.45, 29.34,

29.26, 29.15, 29.08, 29.05, 27.15, 25.67, 24.85, 22.71, 22.33, 14.08 ( - ), 14.04 ( - )

(31 signals overlapping).



1-palmitoyl-2-stearoyl-3-oleoyl-sn-glycerol (19n)


triacylglycerols, palmitic acid was used for ring opening of

glycidyl oleate and stearic acid was used for the final

esterification. Reaction afforded 681 mg of desired product as

white solid (80%)

1H NMR (400 MHz, CDCl3, δ): 5.34 (m, 2H), 5.25 (m, 1H), 4.29 (dd, J = 11.9, 4.3,

2H), 4.14 (dd, J = 11.9, 6.0, 2H), 2.31 (t, J = 7.6, , 6H), 2.00 (m, 4H), 1.61 (s, 6H),

1.25 (m, 72H), 0.88 (t, J = 6.8, 9H).

Chapter 4

88

4

13C NMR (101 MHz, CDCl3, δ): 173.24, 173.21, 172.83, 129.97 ( - ), 129.69 ( - ), 68.84

( - ), 62.08, 34.20, 34.03, 34.01, 31.91, 31.89, 29.75, 29.69, 29.65, 29.63, 29.61, 29.51,

29.48, 29.47, 29.35, 29.30, 29.28, 29.26, 29.16, 27.20, 27.15, 24.89, 24.85, 22.67, 22.66,

22.32, 14.09 ( - ), 14.03( - ) (21 signals are overlapping)

HRMS-ESI+ (m/z): [M + H]+ calculated for C55H105O6, 861.790; found: 861.790.



Synthesis of the modified Jacobsen ligand (Scheme 8-I)

2-(tert-butyl)-4-benzyloxyphenol (29)

A roundbottom flask was charged with tert-butyl hydroquinone (7.51 g,

45 mmol) and potassium iodide (375 mg, 5.0 mol%). Solids were dissolved

in CH3CN (150 ml), and the remaining compounds were added in the

following order: benzyl bromide (6.5 ml, 54 mmol, 1.2 equiv) and K2CO3

(7.5 g, 54 mmol, 1.2 equiv). The reaction mixture was immersed into a preheated oil

bath (80 °C) and the mixture was refluxed for 3 h. The oil bath was removed and

reaction mixture was allowed to cool to RT. Once cooled down, the acetonitrile was

evaporate in vacuo. The crude residue was suspended in CH2Cl2 and filtered. The

filtrate was concentrated in vacuo and the crude residue was further purified by flash

chromatography on silica gel using 10% Et2O in pentane. Fractions with Rf = 0.36

in 10% Et2O in pentanes were collected and evaporated to afford 5.7 g of the desired

product monobenzylated 29 (50%) as a brown solid.

5.48 g of the bisprotected derivative was also isolated (35%).

1H NMR (400 MHz, CDCl3, δ): 7.48 (d, J = 7.3 Hz, 2H), 7.42 (t, J = 7.3 Hz, 2H),

7.36 (t, J = 7.1 Hz, 1H), 7.00 (d, J = 2.9 Hz, 1H), 6.70 (dd, J = 8.5, 2.8 Hz, 1H), 6.59

(d, J = 8.5 Hz, 1H), 5.03 (s, 2H), 4.83 (bs, 1H), 1.44 (s, 9H).

13C NMR (101 MHz, CDCl3, δ): 152.61, 148.56, 137.64, 137.40, 128.58 ( - ), 127.94

( - ), 127.68 ( - ), 116.82 ( - ), 115.25 ( - ), 111.78 ( - ), 70.84, 34.75, 29.53 ( - ), ppm.


Melting point: 82 – 83 °C


89

4

Bisprotected derivative

1H NMR (400 MHz, CDCl3, δ): 7.48 (t, J = 5.3 Hz, 4H), 7.41 (t, J = 7.2 Hz, 4H),

7.35 (d, J = 6.9 Hz, 2H), 7.04 (d, J = 2.9 Hz, 1H), 6.87 (d, J = 8.8 Hz, 1H), 6.77 (dd,

J = 8.7, 2.9 Hz, 1H), 5.09 (s, 2H), 5.03 (s, 2H), 1.43 (s, 9H).

13C NMR (101 MHz, CDCl3, δ): 152.61, 152.02, 139.96, 137.65, 137.39, 129.03,

128.55, 128.50, 127.88, 127.61, 127.21, 115.37, 113.00, 110.89, 70.58, 70.55, 35.04,

29.76.


5-(benzyloxy)-3-(tert-butyl)-2-hydroxybenzaldehyde (31), conditions a.

A roundbottom flask was charged with the benzyl derivative 29 (256 mg,

1.0 mmol) and hexamine (280 mg, 2.0 mmol, 2.0 equiv). These compounds were

dissolved in glacial acetic acid (10 ml) and immersed into preheated oil bath (120 °C).

After heating for 4 h, TLC indicated full conversion of the starting material (29). The

flask was removed from the oil bath and allowed to cool to RT. The reaction mixture

was then poured into water and extracted with Et2O (3 x 30 ml). Combined organic

layers were washed with water and brine, dried over MgSO4 and evaporated to

dryness. The crude residue was purified by flash chromatography on silica gel using

10% Et2O in pentanes.

Reaction afforded 168 mg of the desired aldehyde 31 as a brown solid (53%)

1H NMR (400 MHz, CDCl3, δ): 11.44 (s, 1H), 9.73 (s, 1H), 7.36 (t, J = 5.9 Hz, 2H),

7.32 (d, J = 7.3 Hz, 1H), 7.27 (dd, J = 6.1 Hz, 2H), 7.17 (d, J = 3.7 Hz, 1H), 6.81 (d,

J = 3.0 Hz, 1H), 4.97 (s, 2H), 1.33 (s, 9H).

13C NMR (101 MHz, CDCl3, δ): 196.58 ( - ), 156.33, 151.14, 140.19, 136.67, 128.66

( - ), 128.55, 128.15 ( - ), 127.57 ( - ), 119.80 ( - ), 113.26 ( - ), 70.82, 35.02, 29.11( - )

ppm.

Chapter 4

90

4


Melting point: 55 – 56 °C

5-(benzyloxy)-3-(tert-butyl)-2-hydroxybenzaldehyde (31), conditions c.

A dried three-necked 250 ml flask equipped with a reflux condenser and connected

to a nitrogen line was charged with phenol 29 (5.12 g, 20 mmol), anhydrous MgCl2

(3.8 g, 40 mmol, 2.0 equiv) and paraformaldehyde (loading based on monomeric

formaldehyde, 1.8 g, 60 mmol, 3.0 equiv, dried overnight over P2O5 and high

vacuum) under a positive stream of nitrogen. Solids were dissolved in dry THF (150

ml) and Et3N (5.5 ml, 40 mmol, 2.0 equiv) was added. The resulting mixture was

immersed into a preheated oil bath (72 °C) and refluxed for 20 h. The oil bath was

removed and the reaction mixture was allowed to cool to RT. Aqueous HCl (ca 1 M,

150 ml) was carefully added. The aqueous layer was further extracted with Et2O

(3 x 100 ml), washed with brine, dried over Mg2SO4 and evaporated. The crude

residue was chromatographed using 5% Et2O in pentanes to afford 4.5 g of the

desired compound as yellow-brown solid (80%) containing 8% of methoxymetylated

derivative based on 1H NMR.

NOTE: 32 (methoxymethylated derivative) which occasionally forms has the same

Rf as the desired product 31, but it can be identified in 1H-NMR based on the

following set of signals: (400 MHz, CDCl3) 6.93 (d, J = 3.0 Hz, 1H), 6.53 (d, J = 3.0

Hz, 1H), 4.61 (s, 2H), 3.44 (s, 3H). Trituration with cold pentanes is sufficient to

remove contaminant 32 from the desired aldehyde 31.

1H NMR (400 MHz, CDCl3, δ): 11.44 (s, 1H), 9.73 (s, 1H), 7.36 (m, 5H), 7.17 (d, J =

3.0 Hz, 1H), 6.80 (d, J = 3.0 Hz, 1H), 4.96 (s, 2H), 1.33 (s, 9H).

13C NMR (101 MHz, CDCl3, δ): 196.58 ( - ), 156.33, 151.14, 140.19, 136.67, 128.66

( - ), 128.55, 128.15 ( - ), 127.57 ( - ), 119.80 ( - ), 113.26 ( - ), 70.82, 35.02, 29.11( - ).


3-(tert-butyl)-2,5-dihydroxybenzaldehyde (20)

An oven-dried roundbottom flask was charged with aldehyde 31 (2.5 g,

8.8 mmol). This was dissolved in EtOH (150 ml) and Pd/C (Degussa type

E101 NE/W, 10 mol%, 930 mg) and cyclohexa-1,4-diene (8.3 ml, 10

equiv) were added. The mixture was stirred for 24 h until full conversion


91

4

of 31 was achieved (TLC). Reaction mixture was filtered over a pad of celite. The

filtrate was evaporated to dryness and the crude residue was purified by flash

chromatograph using 30% AcOEt in pentanes to afford 1.62 g of desired aldehyde

20 as yellow solid (95%).

1H NMR (400 MHz, CDCl3, δ): 11.38 (s, 1H), 9.79 (s, 1H), 7.10 (d, J = 3.0 Hz, 1H),

6.83 (d, J = 3.1 Hz, 1H), 4.71 (bs, 1H), 1.41 (s, 9H).

13C NMR (101 MHz, CDCl3, δ): 196.54 ( - ), 155.90, 147.60, 140.09, 123.34 ( - ),

120.08, 115.39 ( - ), 34.94, 29.06 ( - ).



Melting range: 140 – 141 °C

Immobilization of ligand 24 on solid support (scheme 7)

Salen ligands 27, 37, 38

A roundbottom flask was charged with 3-tert-butyl-2,5-dihydroxybenzaldehyde 26

(620 mg, 3.2 mmol) and 3,5-di-tert-butylsalicylaldehyde 25 (2.25 g, 9.6 mmol,

3.0 equiv). These solids were dissolved in CH2Cl2 (30 ml), and (1R,2R)-cyclohexane-

1,2-diamine 24 (730 mg, 6.4 mmol, 2.0 equiv) was added. The reaction mixture was

stirred for 24 h at RT. The solvent was removed under reduced pressure to afford a

mixture of salen ligands 27, 37, and 38 in statistical ratio of 6:1:9 as a yellow solid.

This mixture was used without further purification in the next reaction.

Resin bounded [R,R]-Co(Salen) catalyst 41

Chapter 4

92

4

A roundbottom flask was charged with poly-4-

hydroxystyrene resin (Advanced ChemTech, 1% cross-

linked, 90 µm, 3.6 mmol/g, 36 mg what is considered as

0.13 μmol). Beads were further suspended in CH2Cl2

(2.5 ml) and 4-nitrophenyl chloroformate (400 mg, 2.0

mmol, 15 equiv) and DMAP (61 mg, 0.46 mmol, 3.5 equiv) were added in this order.

The mixture was stirred for 1 h at RT. The solids were filtered using a glass filter

(por 4) and washed with CH2Cl2. The resulting white beads were dried for 1 h under

high vacuum (oil pump).The IR spectrum showed a band at 1765 cm-1 corresponding

to the carbonate stretch vibration and a prominent band at 1528 cm-1 corresponding

to the asymmetric stretch vibration of the nitro group. The white beads were further

suspended in DMF (2.5 ml) and the mixture of salen ligands 27, 37 and 38 followed

by DIPEA (180 μl, 1.0 mmol, 8.0 equiv) and DMAP (61 mg, 0.46 mmol, 3.5 equiv)

were added. The mixture was stirred for 1.5 h. An immediate color change to yellow

attributed to p-nitro-phenolate was observed. After 1.5 h of stirring, the resin was

filtered using a glass filter (por 4) and washed with solvents in the following order:

DMF, CH2Cl2, MeOH and CH2Cl2. The resulting yellow beads were subjected to IR

analysis. In the IR spectrum the band corresponding to carbonate shifted towards a

lower wave number (1745 cm-1) and the band corresponding to asymmetric stretch

vibration of the nitro group disappeared. Furthermore, a new band corresponding

to the stretch vibration of an imine appeared in the IR spectrum (1630 cm-1). The

resulting yellow beads were dissolved in MeOH / toluene (1 / 1, 2.5 ml) and

Co(OAc)2.4H2O (249 mg, 1.0 mmol, 8.0 equiv). After a short period of time, the

beads turned dark red. After 1.5 h of stirring, the beads were filtered using a glass

filter (por 4) and washed with solvents in the following order: MeOH, CH2Cl2 /

toluene 9 / 1, AcOH, CH2Cl2, MeOH and CH2Cl2 to yield dark red/brown beads.

The nitrogen content N = 2.33% of the resin corresponds to 0.83 mmol of ligand

per 1 g of the resin. Assuming the quantitative incorporation of the Co2+ the final

loading of the Co3+[R,R-(SALEN)] is 0.83 mmol/g.

Anal. Calcd found: N, 2.33; C, 66.56; H, 6.37.

Ring opening using immobilized catalyst (scheme 10)

To a 10 ml round-bottom flask, resin bound Co[R,R-Salen]

catalyst 35 (144 mg of the resin corresponds to 10 mol% of

the Co catalyst was added. This was suspended in small

amount of CH2Cl2 (0.5 ml) and oleic acid (320 µl, 1.0 mmol,


93

4

1.0 equiv), followed by glycidyl butyrate, 1 (102 µl, 1.0 mmol) and Hünigs base (170

µl, 1.0 mmol, 1.0 equiv) were added in this order. The reaction was followed by 1H

NMR. After 6 days, full conversion of glycidyl butyrate was achieved. The resin

bound catalyst was removed by filtration and the filtrate was diluted with n-heptane

(10 ml). To this mixture, myristic acid (274 mg, 1.2 mmol, 1.2 equiv) and DMAP (6.0

mg, 50 μmol, 5 mol%) followed by DCC (227 mg, 1.1 mmol, 1.1 equiv) were added,

and the resulting reaction was stirred for 20 h. The crude residue was directly

transferred onto a silica column and chromatographed using 10% Et2O in pentane

to afford 411 mg of the desired product as a yellowish liquid (80% yield).

Spectral data corresponded to those reported in table 2, entry 4.


(1) Baer, E.; Fischer, H. O. L. J. Biol. Chem. 1939, 128, 475.

(2) The band corresponding to the symmetric stretching vibration of the nitro group

was overlapping with other bands.

(3) Craven, R. J.; Lencki, R. W. Cryst. Growth Des. 2011, 11, 1723.

(4) Schlenk, W., Jr. J. Am. Oil Chem. Soc. 1965, 42, 945.

(5) Patent, O. G. G. a. W. N. G., 1920.

(6) (a) Sreenivasan, B. J. Am. Oil Chem. Soc. 1978, 55, 796 (b) Konishi, H.; Neff, W.;

Mounts, T. J. Am. Oil Chem. Soc. 1993, 70, 411 (c) Eckey, E. W. Ind. Eng. Chem. 1948,

40, 1183.

(7) (a) Jandacek, R. J.; Webb, M. R. Chem. Phys. Lipids 1978, 22, 163 (b) Smith, R. E.;

Finley, J. W.; Leveille, G. A. J. Agric. Food Chem. 1994, 42, 432.

(8) (a) Xu, X. Eur. J. Lipid Sci. Technol. 2000, 102, 287 (b) Mu, H.; Porsgaard, T. Prog.

Lipid Res. 2005, 44, 430 (c) Lee, K. T.; Akoh, C. C. Food Rev. Inter. 1998, 14, 17.

(9) Chandler, I.; Quinlan, P.; McNeill, G. J. Am. Oil Chem. Soc. 1998, 75, 1513.

(10) Kristinsson, B.; Haraldsson, G. G. Synlett 2008, 2178.

(11) (a) Halldorsson, A.; Magnusson, C. D.; Haraldsson, G. G. Tetrahedron 2003, 59, 9101

(b) Halldorsson, A.; Magnusson, C. D.; Haraldsson, G. G. Tetrahedron Lett. 2001, 42,

7675 (c) Haraldsson, G. G. In Biocatalysis and Bioenergy; John Wiley & Sons, Inc.,

2008.

(12) (a) Stamatov, S. D.; Stawinski, J. Synlett 2006, 2251 (b) Stamatov, S. D.; Stawinski, J.

Bioorg. Med. Chem. Lett. 2006, 16, 3388 (c) Stamatov, S. D.; Stawinski, J. Tetrahedron

Lett. 2006, 47, 2543 (d) Stamatov, S. D.; Stawinski, J. Synlett 2007, 439 (e) Stamatov,

S. D.; Stawinski, J. Org. Biomol. Chem. 2007, 5, 3787 (f) Stamatov, S. D.; Stawinski, J.

Eur. J. Org. Chem. 2008, 2635 (g) Stamatov, S. D.; Stawinski, J. Org. Biomol. Chem.

2010, 8, 463.

(13) 9 common fatty acids in milk fat correspond to 729 different possible triacylglcerols.

(21/729) x 100 = 2.8%

Chapter 4

94

4

(14) (a) Annis, D. A.; Jacobsen, E. N. J. Am. Chem. Soc. 1999, 121, 4147 (b) Baleizão, C.;

Garcia, H. Chem. Rev. 2006, 106, 3987.

(15) Hofsløkken, N. U.; Skattebøl, L. Acta Chem. Scand. 1999, 53, 258.

(16) Hansen, T. V.; Skattebøl, L. Org. Synth. 2005, Vol. 82, p. 64-68; 2009, Coll. Vol. 11,

p. 267-271. Addendum: Org. Synth. 2012, 89, 220-229.

(17) Hansen, T. V.; Skattebøl, L. Tetrahedron Lett. 2005, 46, 3829.

(18) Bagno, A.; Kantlehner, W.; Kress, R.; Saielli, G. Z. Naturforsch. B 2004, 59, 386.

(19) The other explored hydrogen sources were ammonium formate and H2 (balloon).

(20) the ratio of 25 : 26 is 1 : 3, translated to % this means the mixture consists of 25%

of 25 and 75% of 26. Therefore ligand 37 will correspond to (0.25 x 0.25) x 100 =

6.25%, ligand 38 (0.75 x 0.75) x 100 = 56.3% and the desired C2 symmetrical ligand

27 will be present in 2 x (0.25 x 0.75) x 100 = 37.5%

(21) The band corresponding to the symmetric vibration of the nitro group was

overlapping with other bands.

(22) Hirschmann, H. J. Biol. Chem. 1960, 235, 2762.

95

Pag

e95

Chapter 5 A Methyl Matters

Abstract: Methyl branched fatty acids and their corresponding phospholipids are

found as components of bacterial membranes. Despite their occurrence, it is

challenging to isolate a reasonable amount of these lipids and use them for studies.

Therefore, these amounts have been obtained by chemical synthesis.

This work has been carried out in collaboration with Mac Donald José, and Dr. A.

Koçer (Department of Medical Physiology, UMCG), and Dr. H. I. Ingólfsson, Dr.

A. de Vries and Prof. Dr. S.-J. Marrink (Department of Molecular Dynamics, GBB,

University of Groningen).

Chapter 5

96

5

Introduction

Methyl-branched fatty acids, introduced in Chapter 2, are components of

phospholipids that occur in a number of specific bacterial species. The influence of

the methyl branch on the physical properties of membranes composed of these lipids

is a topic of discussion. Currently, it is accepted that the main function of the methyl

group is to increase the fluidity of the membranes, acting as a chemically stable

analogue of a double bond. However, methyl branching can have more roles, for

example, the corresponding membranes might be more stable and less permeable.

In this context, the study of Elferink et al.1 is relevant. The authors compared the

stability and the proton permeability of liposomes from linear, branched and archaeal

(that is; multimethyl-branched) lipids, which were obtained as lipid extracts from

Escherichia coli, Bacillus stearothermophilus (Figure 1) and Sulfolobus acidocaldarius (Figure

2).

Figure 1. Typical lipid and fatty acid composition of E. coli grown at 25 °C2; ( II ) typical

lipid and fatty acid composition of B. Stearothermophilus grown at 55 °C.3

The composition of the lipids differs in the headgroups, chain lengths, unsaturation

and branching. The study of Elferink et al.1 showed that the liposomes of B.

A Methyl Matters

97

Pag

e97

5

stearothermophilus lipids were more stable and less proton permeable than the

liposomes of E. coli lipids.

Figure 2. Typical lipid composition of S. acidocaldarius.4

The liposomes composed of archaeal lipids showed even greater stability. From the

compared lipid mixtures it is difficult to conclude, if methyl branching has any effect

on the stability of the liposomes.

Due to their difficult accessibility, branched-chain lipids are studied often in

silico.5 Recently, Lim and Klauda6 published a computational study of the influence

of branched lipids (Figure 3) from Chlamidia trachomatis on the properties of the

corresponding bilayer. One of their conclusion was that the branched lipids form

stiffer membranes. However, from their publication it is not clear whether the

authors considered the stereochemistry of the lipids.

Figure 3. Membrane-forming lipids studied by Lim and Klauda by molecular dynamics

simulation.

Stereochemistry of the branched lipids might have an impact not only on the

properties of the bilayers (an ensemble - or macroscopic effect), but also on lipid-

protein interactions in a more molecular effect. Substitution of the glycerol with a

chiral, branched fatty acid introduces an additional stereogenic center to the

Chapter 5

98

5

molecule. The resulting lipid then comes in 4 stereoisomers. One might expect, that

the protein (which is chiral and enantiopure) will interact with each stereoisomer in

a different way.

In this context it is interesting to study the interaction of the

mechanosensitive protein channels with lipids. Mechanosensitive channels regulate

the pressure inside the cell by opening when this gets to life-threating levels. High

turgor pressure stretches the lipid membrane, what triggers a conformational change

in the protein that opens the channel’s pore. Through the opened pore an unselective

efflux takes place. When the pressure is below the life-threating value, the channel

closes. One of the best understood mechanosensitive channels is the

mechanosensitive channel of large conductance7 (MscL) from Mycobacterium

tuberculosis.

The crystal structure of M. tuberculosis MscL revealed that it is a

homopentamer (Figure 4-I, 4-II) with a simple topology.

Figure 4. Schematic representations of the MscL channel from the M. tuberculosis crystal

structure. ( I ) side view of the MscL pentamer; ( II ) top view from the periplasmic side, (

III ) individual MscL subunit with the positions of the two transmembrane helices, TM1 and

TM2.

The cytoplasmic N-terminus is followed by a transmembrane helix (TM1), which is

connected to the second transmembrane helix (TM2), followed by the cytoplasmic

C-terminus (Figure 4-III). The MscL is embedded in the inner membrane of the M.

tuberculosis cell wall, which is largely composed by lipids bearing the methyl branched

fatty acid tuberculostearic acid.

A Methyl Matters

99

Pag

e99

5

There are several techniques to study the MscLs. Probably, the most widely

used technique is patch-clamp.8 As this is experimentally difficult, other, more

convenient techniques have been developed. One of the most versatile is a calcein

efflux essay.9 The idea of the calcein efflux essay is the preparation of liposomes filled

with calcein (Figure 5-I), which is a self-quenching fluorescent dye.

Figure 5 ( I ) Principle of the calcein efflux essay; ( II ) release triggered by MTSET; ( III )

release triggered by LPC; ( IV ) chemical structures of the compounds used in the calcein

efflux essay.

Subsequently, the MscL is reconstituted into these dye-filled liposomes. The

efflux of the dye is triggered either by introduction of a charge into the pore of the

MscL or by deformation of the bilayer, which triggers a change in the MscL

conformation or leads to the opening of the pore. To trigger the MscL by charge

(Figure 5-II), the proteoliposomes with a reconstituted mutated MscL are treated

Chapter 5

100

5

with 2-(trimethylammonium)ethyl methanethiosulfonate (MTSET). MTSET labels

the cysteine residues of TM1, which are in the middle of the pore. In the same

manner, the charge is introduced onto the remaining 4 MscL subunits. The

introduced charge leads to mutual repulsion of the subunits, thus forcing the pore to

open. Through the opened pore, the calcein can efflux, which is recorded as an

increase in the fluorescence. Triggering of the MscL through membrane induced

deformation is achieved by addition of lysophosphatidylcholine (LPC). LPC (Figure

5-III) has a conical shape and can insert into the outer leaflet of the bilayer. The

insertion leads to the deformation of the membrane. This deformation leads to a

change in the conformation of the MscL, which again leads to the opening of the

pore. The effluxed calcein again increases the fluorescence. At the final stage of the

calcein efflux assay, the amount of the effluxed dye is determined after bursting of

the liposomes induced with a detergent.

Until now, there is no study that directly compares branched and non-

branched lipids in their influence on an MscL, although the best studied MscL

originates from M. tuberculosis. Recently, Ter Horst et al. showed that 1 (Figure 6) is

the most abundant lipid in M. tuberculosis by comparison of the MS/MS spectra of

synthetic and natural 1. The authors needed 1 g of lipid extract (corresponding to 20

g of bacteria) for the isolation of 50 μg of natural 1.

Figure 6. The most abundant membrane lipid in M. tuberculosis.

Extrapolating, for 30 mg of 1, what is an acceptable amount to study MscL

incorporation and function in liposomes, one would need 600 g of the extract,

corresponding to 12 kg of M. tuberculosis.

The previous chapters (chapter 2 and 3) showed the efficient synthesis of

branched fatty acids and their phospholipids. These methods can be applied to the

synthesis of 1 providing sufficient amounts of the lipid for further studies.

A Methyl Matters

101

Pag

e10

1

5


Chemical synthesis of the phospholipids

The phospholipids used in this study were prepared according to the

synthetic routes developed in the previous chapters (Chapter 2, Chapter 3).

Figure 7. ( I ) (R)-tuberculostearic acid and its enantiomer; ( II ) branched lipids used in this

study; ( III ) linear chain lipids used in this study.

Chapter 5

102

5

Commercially available nonanal was converted into (R)-tuberculostearic acid in 5

steps. Application of the opposite enantiomer of the ligand, used for the introduction

of the methyl branch, led to the non-natural (S)-tuberculostearic acid (Figure 7-I).

Both fatty acids were obtained in comparable yields and in the same optical purity.

The fatty acids could be further converted into the corresponding

phospholipids (Figure 7-II). As Okuyuma10 and Ter Horst11 independently reported,

the phospholipids of M. tuberculosis have the sn-1 position mainly substituted by (R)-

tuberculostearic acid and the sn-2 position by (saturated) palmitic acid. Therefore,

lipid 1 (Figure 1-II) was synthetized and used for the study. To facilitate the

formation of the liposomes, a PC analogue 2 was also prepared. The non-natural (S)-

tuberculostearic acid was converted into the lipid 3. 3 is a chain epimer of 1. Lipids

1 and 3 allow a controlled study of the influence of the configuration of the methyl

branch on the properties of the liposomes. The chemical synthesis afforded 32.6 mg

of lipid 1, 73.5 mg of lipid 2 and 50.0 mg of lipid 3.

The methyl branched lipids were compared to their desmethyl analogues 4

and 5 (Figure 1-III). Lipid 4 is commercially available, and the synthesis of lipid 5 is

described in the previous chapter (Chapter 3).

Formation of the liposomes and proteoliposomes

The initial experiments with lipids 1 and 3 showed that these do not form

liposomes. Therefore, the lipids were studied in two-component liposomal

formulations which were prepared as a 1/1 (w/w) ratio of the PE and PC

phospholipids. This resulted in the following mixtures:

1:1 mixture of compounds 1 and 2, designated as RR

1:1 mixture of compounds 3 and 2, designated as SR

1:1 mixture of compounds 1 and 4, designated as R0

1:1 mixture of compounds 3 and 4, designated as S0

1:1 mixture of compounds 5 and 4, designated as 00

The lipid mixtures are designated based on the presence and configuration of the

methyl branch in the lipid. The first letter stands for the configuration of the PE

component and the second letter for the configuration of the PC component.

A Methyl Matters

103

Pag

e10

3

5

Formation of the liposomes already revealed different behavior of the

branched and non-branched lipids. The mixtures consisting only of branched lipids

(RR, SR) smoothly afforded liposomes at room temperature. From the “half-

branched” mixtures R0 and S0, only S0 afforded liposomes, after heating above the

transition temperature of the PC component. At the same conditions, R0 did not

yield any liposomes. The lipids 4 and 5 displayed low solubility in organic solvents,

what together with their high transition temperatures resulted in failure to form

liposomes from the 00 formulation.

The SR and S0 liposomes were subsequently studied by cryo-electron

microscopy (Figure 8). This revealed, that while the SR mixture yields round, well-

shaped liposomes, the S0 mixture yields edgy, uneven liposomes, probably due to

phase separation in the membrane.

Figure 8. ( I ) uneven liposomes from the SR mixture, ( II ) well defined liposomes from

the S0 mixture (courtesy of Dr. M. Stuart).

The MscL proteins from E. coli and M. tuberculosis were reconstituted into liposomes

based on the branched (RR, SR) and half-branches (S0) mixtures. The efficiency of

the reconstitution was estimated using SDS-PAGE (Figure 9). This showed, that

roughly the same amount of E. coli and M. tuberculosis protein was reconstituted in

the SR and RR liposomes (lanes 2-5). However, when the proteins were

reconstituted into the S0 mixture, the MscL from M. tuberculosis reconstituted with

greater efficiency (lane 13 and 14), than the MscL from E. coli. The proteins were

reconstituted also in liposomes based on soy-extract (20%) (lanes 6-9 and 11-12),

which was used as a control experiments in the calcein efflux assays.

Chapter 5

104

5

Figure 9. SDS-PAGE of the reconstituted proteoliposomes; ( I ) reconstitution of the E.

coli and M. tuberculosis proteins in the RR and SR -based liposomes; ( II ) reconstitution of

the E. coli and M. tuberculosis proteins in S0-based liposomes.

Molecular dynamics study of the bilayers

This part of the chapter briefly summarizes results obtained by Dr. Helgi I. Ingólfsson and Dr. Alex de Vries, department of Molecular Dynamics.

Simulation of a phospholipid bilayer using molecular dynamics (MD) is a

powerful technique. One of the big advantages is its predictive power. Out of the 5

studied liposomal formulations, only 3 afforded liposomes. However, thanks to the

MD simulations all 5 formulations could be compared. In the case of the fully

branched formulations SR and RR, the bilayer resides in a liquid phase at 25 °C

(Table 1).12 In the case of S0 and R0, where only 50% of the lipids are branched, the

bilayer resides in a gel phase.13 The lipid mixture 00, which is composed only from

linear lipids is also in the gel phase. The phase in which lipids reside is closely related

to the area per lipid. Table 1 further shows, that the removal of the methyl branches

leads to tighter packing of the membrane. The mixtures SR and RR (0.634 ± 0.012

and 0.640 ± 0.012 nm2 area per lipid) show very similar packing. The removal of a

portion of methyl branched lipids as in the mixtures S0 and R0 leads to tighter

packing compared to the SR and RR bilayers.

Table 1. Summary of the molecular dynamics study.

Lipid mixture Phase Area per lipid / nm2 Phosphate distance / Å

SR Liquid 0.634 ± 0.012 39.2 ± 0.7

RR Liquid 0.640 ± 0.012 38.8 ± 0.6

S0 Gel 0.544 ± 0.021 44.0 ± 1.3

R0 Gel 0.496 ± 0.009 47.1 ± 0.7

00 Gel 0.475 ± 0.003 47.0 ± 0.3

A Methyl Matters

105

Pag

e10

5

5

Interesting is the difference between the S0 and R0 mixtures. Despite being

diastereomeric, the stereogenic centers are distant. However, mixture S0 shows a

higher area per lipid (0.544 ± 0.021 nm2) compared to R0 (0.496 ± 0.009 nm2). The

subsequent removal of a methyl-branch as in mixture 00 leads to an even tighter

packing (0.475 ± 0.003 nm2). The MD simulations also provided information about

the thickness of the bilayers (expressed as the transbilayer distance of the phosphate

head-groups). The thickness of the membrane in the SR and RR bilayers is

comparable (39.2 ± 0.7 Å and 38.8 ± 0.6 Å). The bilayers composed of the

diastereomers S0 and R0 again were notably different. The bilayer corresponding to

R0 was thicker than the one of S0 (47.1 ± 0.7 Å and 44.0 ± 1.3 Å). In comparison

to the previous mixtures, the 00 mixture formed the thickest bilayer (47.0 ± 0.3 Å).

The simulations of the studied lipid bilayers have been summarized in an illustrative

fashion. The residues in pink correspond to the PE components and the blue

residues correspond to PC. In the case of the branched SR and RR mixtures (Figure

10), the fatty acid residues are unorganized and randomly ordered.

Figure 10. Simulated bilayers of SR and RR lipid mixtures.

However, the mixtures S0 and R0 (Figure 11) are considerably more ordered. The

aggregation of the PC (blue) and PE (pink) residues in the mixture S0 might explain

the formation of the observed edgy uneven liposomes, as observed by cryo-electron

microscopy (Figure 8-II). Comparison of the diastereomers S0 a R0 suggests that

despite the fact that the stereogenic centers (of the methyl-branch and the head

group) are distant, they have an impact on the macroscopic properties of the bilayer.

Chapter 5

106

5

Figure 11. Simulated bilayers of S0 and R0 mixtures.

Simulation of the bilayer from the linear lipid mixture 00 (Figure 12) showed that

this is highly organized.

Figure 12. Simulated bilayer of the 00 mixture.

Calcein efflux assay

This part of the research has been conducted by Mac Donald José and Dr. Armagan Koçer, department of medical physiology, UMCG.

In the calcein efflux assay, MscLs from 2 different species were studied. The

MscL from M. tuberculosis and the one from from E. coli. In Nature, the membranes

in which these MscLs are embedded differ. The M. tuberculosis channel is embedded

in a membrane consisting of methyl-branched lipids and the E. coli channel is

embedded in lipids from linear and unsaturated fatty acids. The first studied trigger

in the calcein efflux essay was MTSET. The increase of fluorescence was plotted

A Methyl Matters

107

Pag

e10

7

5

against time (Figure 13). After the fluorescence did not increase anymore, the total

amount of released calcein was determined after bursting of the liposomes. The

measured data points were fitted by an exponential function. The initial rate and the

total amount of effluxed calcein are the studied parameters. These are summarized

in Table 2.

Figure 13. Graphs of the calcein efflux essay triggered by MTSET.

Chapter 5

108

5

Table 2. Summary of the calcein efflux essay triggered by MTSET.

M. tuberculosis MscL E. coli MscL

Lipid mixture Release Initial rate (s-1) Release Initial rate (s-1)

Soy extract 94.0 ± 1.0% 1.13 ± 0.5 98 ± 1.0% 1.15 ± 0.03

SR 94 ± 1.0% 1.13 ± 0.04 79 ± 3.0% 0.92 ± 0.3

RR 95 ± 1.0% 0.78 ± 0.1 85 ± 7.0% 0.53 ± 0.06

S0 90 ± 1.0% 1.2 ± 0.6 22 ± 3.0% 0.06 ± 0.01

As the graph (figure 13-I) and Table 2 show, the M. tuberculosis MscL is not much

affected by the different lipid mixtures. The release of the calcein is the same in all 4

cases and the rate is only slightly slower in the case of the RR lipid mixture. However,

the difference in the rate is not significant. The E. coli MscL shows comparable

release in soy extract, SR and RR mixture. The release in S0 mixture is dramatically

lower (22 ± 3.0%). Furthermore, E. coli MscL shows only negligible difference in the

rates in soy extract and SR mixture. However, the rate of the RR mixture is 2 times

slower compared to the soy extract. The rate in the S0 mixture is again significantly

lower (0.06 ± 0.01 s-1). Both the lower release and rate of E. Coli MscL in the S0

mixture are probably caused by much lower reconstitution efficiency as shown on

SDS-PAGE above (Figure 7, lane 14).

A Methyl Matters

109

Pag

e10

9

5

Release triggered by LPC

The second studied trigger was LPC. The experimental set up and the output

are very similar to the previous MTSET-experiment. Following figure (Figure 14)

summarizes the results in graphs.

Figure 14. Graphs of the calcein efflux essay triggered by LPC.

Chapter 5

110

5

The releases and rates are summarized in Table 3.

Table 3. Summary of the calcein efflux assay triggered by LPC.

M. tuberculosis MscL E. coli MscL

Lipid mixture Release Initial rate (s-1) Release Initial rate (s-1)

Soy extract 82 ± 1.0% 1.6 ± 0.1 59 ± 4.0% 1.10 ± 0.01

SR 81 ± 1.0% 1.8 ± 0.1 34 ± 1.0% 1.14 ± 0.01

RR 90 ± 1.0% 1.0 ±0.2 19 ± 1.0% 1.09 ± 0.01

S0 70 ± 1.0% 0.4 ± 0.1 29 ± 6.0% 1.09 ± 0.01

Using LPC as the trigger, the M. tuberculosis MscL shows similar

characteristics as in the MTSET experiments. The release is comparable (though not

identical, see the RR mixture) in all studied liposomes. The lower release and rate of

the M. tuberculosis MscL in the S0 mixture can be attributed to a hydrophobic

mismatch; the hydrophobic region of the M. tuberculosis MscL spans 35 Å. The

phosphate distances calculated by MD simulations of the mixture SR and RR is

comparable to this; 39.2 ± 0.7 Å and 38.8 ± 0.6 Å, respectively. The S0 mixture,

however, forms bilayers of 44.0 ± 1.3 Å.

Another aspect which has to be considered in the case of liposomes based

on the S0 mixture is the concentration of liposomes. The concentration of added

LPC, which was kept constant in all the experiments, leads in the case of the S0

mixture to a greater ratio of the trigger to liposomes. The E. coli MscL shows

different, deviating behavior in the calcein efflux assay when LPC is used as the

trigger (Figure 11-II). Only the control experiment with soy extract shows the same

kinetics. The apparent linear dependence of the release against time in the SR, RR

and S0 mixtures suggest, that in these mixtures the release follows a different kinetics.

The total release is in all cases lower than the corresponding M. tuberculosis MscL

experiments.

Conclusions and outlook

The methyl-branched phospholipids were synthetized according to procedures

described in the previous chapters. Both natural (R)-tuberculostearic and non-natural

(S)-tuberculostearic acid were prepared in the same optical purity and comparable

yields. Both fatty acids were further converted to a set of epimeric PE lipids. The

natural (R)-tuberculostearic acid was also converted to a PC lipid. All of the lipids

were obtained in reproducible yields and in sufficient amounts (up to 100 mg) to

conduct further studies, which showed that these lipids do not form single

A Methyl Matters

111

Pag

e11

1

5

component liposomes. Therefore, the liposomes were formed as 2 component

PE/PC 1/1 (w/w) mixtures. The mixtures consisting of lipids with branched fatty

acids formed regular round liposomes. From the mixtures of linear and branched

lipids, only 1 mixture afforded liposomes. These, however had uneven edgy shapes.

The mixtures were studied by molecular dynamics simulations, which provided

insight into the organization and thickness of the bilayers. MscLs from 2 different

species, M. tuberculosis and E. coli were reconstituted in these liposomes. The function

of the MscLs was further studied in a calcein efflux assay. While the M. tuberculosis

MscL showed similar behavior with both studied triggers, the E. coli MscL did not

perform well in the lipid mixture consisting of methyl branched lipids.

As such, the use of natural (synthetic) lipids in this study more closely mimics

the natural environment of in particular the M. tuberculosis MscL. Despite the studied

lipids in this chapter are natural, they were not the best choice for a comparison of

the properties of branched and linear lipids. The low solubility and high transition

temperature of the linear lipids are the probable causes for their reluctance to form

liposomes. In future research, this can be addressed by studies of shorter analogues.

For linear lipids this would lower the transition temperature and increase the

solubility. Furthermore, in future research, the MscLs should be studied by patch-

clamp as this gives more insight into the properties of the bilayer.

Experimental part

Note: The synthesis of (R)-tuberculostearic acid is described in previous chapters (Chapter 2)

(S)-S-ethyl 3-methylundecanethioate

A dried Schlenk flask was charged with CuBr.SMe2 (Sigma-Aldrich, 24.7 mg, 2

mol%) and (R,S)-Josiphos.EtOH adduct (Sigma-Aldrich, 92.2 mg, 2.4 mol%). The

flask was evacuated in three cycles and MTBE (54 ml) was added. The resulting

solution was stirred for 30 min at rt (21 °C) before cooling to −78 °C. After stirring

for 10 min at −78 °C a solution of MeMgBr (Sigma-Aldrich, 3 ml, 1.5 equiv) was

added dropwise during 10 min accompanied by formation of a voluminous yellow

precipitate. To this suspension, (E)-S-ethyl undec-2-enethioate (1.37 g, 6 mmol) was

added as a solution in MTBE (6 ml) during 3 h. The reaction was stirred for an

additional 16 h and then poured into an NH4Cl/ice mixture. The organic layer was

separated, and the aqueous layer was extracted with Et2O (3 x 20 ml). The combined

organic layers were dried over MgSO4 and evaporated. The residual orange liquid

was chromatographed on silica gel using 3% of Et2O in pentane.

Chapter 5

112

5

The reaction afforded the desired (S)-S-ethyl 3-methylundecanethioate as colorless

liquid (1.38 g, 95%)

1H NMR (400 MHz, CDCl3) δ 2.87 (q, J = 7.4 Hz, 3H), 2.52 (dd, J = 14.4, 6.0 Hz,

1H), 2.33 (dd, J = 14.4, 8.1 Hz, 1H), 2.00 (broad s, J = 5.9 Hz, 1H), 1.33 – 1.17 (m,

17H), 0.92 (d, J = 6.7 Hz, 3H), 0.87 (t, J = 6.8 Hz, 3H).

13C NMR: (101 MHz, CDCl3) δ 199.32, 51.40, 36.61, 31.86, 31.06, 29.69 ( - ), 29.55,

29.27, 26.82, 23.23 , 22.65, 19.51 ( -), 14.78 ( - ), 14.08 ( - ).

HRMS-ESI+ (m/z): [M + H]+ calculated for C14H29OS: 245.193 found 245.193.

[α]D +3.32° (c = 1.00, CHCl3).

The enantiomeric excess was determined on the corresponding carbamate, obtained

by LiAlH4 reduction, treatment with phosgene and subsequent treatment with (S)-1-

(1-naphthyl)ethylamine.

Chiracel OD-H, flow=1 ml/min, tminor = 11.4 min, tmajor = 12.4 min, e.r. = 95 : 5.

(S)-3-methylundecanal

A solution of (S)-S-ethyl 3-methylundecanethioate (1.2 g, 4.9 mmol) in CH2Cl2 (30

ml) in a 100 ml roundbottom flask was immersed into a – 40 °C bath. After 5 min

of stirring, a solution of DIBAL-H (Sigma-Aldrich, 1 M in CH2Cl2, 1.2 equiv) was

introduced slowly via the wall of the flask to cool down the reagent. The reaction

was stirred for 1 h (full conversion, TLC) before Rochelle salt (saturated aqueous

solution, 10 ml) was added to stop the reaction. The flask was removed from the

cooling bath and stirred at rt (21 °C) until the layers fully separated. The organic layer

was diluted with additional CH2Cl2 (50 ml), and filtered over a Whatman® 1PS phase

separator filter paper. The filtrate was carefully evaporated (maximum vacuum 100

mbar).

The reaction afforded 915 mg (100%) of the desired aldehyde as a pleasantly smelling

colorless liquid. This was used without further purification in subsequent step.

(S)-10-methyloctadec-7-enoic acid (E/Z mixture)

A dry Schlenk flask was charged with 7-(bromotriphenylphosphoranyl)-heptanoic

acid (Wittig reagent) under a stream of nitrogen. After degassing in three cycles,

LiHMDS (Sigma-Aldrich, 1 M solution in THF, 2.2 equiv) was added dropwise

(addition time ca 3 min, the Wittig reagent is treated without previous suspending or

dissolving). The obtained suspension was stirred for 15 min at rt (21 °C) until it

A Methyl Matters

113

Pag

e11

3

5

turned into a cloudy (unreacted phosphonium salt) deep red solution. This was

further cooled to 0 °C by immersion into an ice/water bath. After 5 min of stirring,

(S)-3-methylundecanal (915 mg, 4.9 mmol) in THF (1 ml added to rinse the flask)

was added dropwise over 15 min. After addition of ca 30% of the aldehyde,

triphenylphosphine oxide started to precipitate. The reaction was stirred for 2 h at 0

°C, and then for 30 min at rt (21 °C). Subsequently it was cooled again by immersing

into the ice/water bath and water was added (5 ml). The pH of the solution was

adjusted to 1 by careful addition of concentrated HCl. The reaction mixture was

transferred into a separatory funnel and extracted with CH2Cl2 (3 x 20 ml), the

combined organic layers were dried over MgSO4 and evaporated with a small amount

of silica gel (10 g, column chromatography using solid loading). The silica gel with

the adsorbed compound was transferred onto a silica gel column and was further

chromatographed with 5% of Et2O in pentane (to elute unreacted aldehyde). After

complete elution of the aldehyde the polarity of the eluent was increased to 60%

Et2O in pentane.

The reaction afforded a mixture of E and Z (S)-10-methyloctadec-7-enoic acid (1.04

g, 71% over 2 steps) and 150 mg of unreacted aldehyde (yield based on recovered

starting material 83% over two steps).

1H NMR (400 MHz, CDCl3) δ 5.53 – 5.23 (m, 2H), 2.35 (t, J = 7.5 Hz, 2H), 2.12 –

1.92 (m, 2H), 1.65 (dd, J = 9.9, 5.0 Hz, 1H), 1.53 – 1.01 (m, 23H), 0.86 (dt, J = 14.7,

7.3 Hz, 6H).

13C NMR (101 MHz, CDCl3) δ 173.49, 130.11, 128.81, 36.70, 34.52, 33.40, 31.91,

29.96 ( - ), 29.66, 29.34, 29.30, 28.72, 27.17, 27.07, 24.58, 22.67, 19.58 ( - ), 14.10 ( - ).

(S)-Tuberculostearic acid

In a 500 ml roundbottom flask, (S)-10-methyloctadec-7-enoic acid (940 mg, 3 mmol)

was dissolved in MeOH (20 ml). To this solution, Pt/C (Sigma-Aldrich, 10% loading,

589 mg, 5 mol%) was added with caution (PYROPHORIC). The flask was capped

with a H2 balloon and vigorously stirred for 20 h at rt (21 °C). The MeOH was

evaporated and the black oily residue was transferred directly onto a silica column.

The desired compound was eluted with 60% Et2O in pentane.

1H NMR (400 MHz, CDCl3) δ 2.35 (t, J = 7.5 Hz, 2H), 1.63 (dt, J = 15.0, 7.5 Hz,

2H), 1.42 – 1.00 (m, 27H), 0.88 (t, J = 6.9 Hz, 3H), 0.83 (d, J = 6.5 Hz, 3H).

Chapter 5

114

5

13C NMR (101 MHz, CDCl3) δ 180.82, 37.08, 37.05, 34.06, 32.73 ( - ), 31.91, 30.02,

29.92, 29.68, 29.45, 29.35, 29.23, 29.05, 27.07, 27.02, 24.65, 22.67, 19.69 ( - ), 14.09

( - ).


[α]D +0.2 (c = 3.0, CHCl3).

Synthesis of the 3-protected-1,2-diacylglycerols

(S)-(R)-3-((tert-butyldimethylsilyl)oxy)-2-(stearoyloxy)propyl tuberculostearate

A dry flask was charged with (R,R)-(−)-N,N′-Bis(3,5-di-tert-butylsalicylidene)-1,2-

cyclohexanediaminocobalt(II) complex (12 mg, 2 mol%) and (R)-TBSA (298 mg, 1

mmol) and guarded with a CaCl2 tube. The mixture was stirred at rt (21 °C) for 30

min (change of color from brick red to brown). Hünigs base (175 µl, 1 equiv) was

added, followed by tert-butyldimethylsilyl (R)-glycidyl ether (Sigma-Aldrich, 210 µl,

1 equiv). The mixture was stirred under a nitrogen atmosphere until full conversion

of the epoxide (followed by 1H NMR). All volatiles were removed under vacuum.

The brownish liquid residue was dissolved in heptane (5 ml), and palmitic acid (332

mg, 1.3 equiv) and DMAP (12 mg, 0.1 equiv) were added sequentially. To the

brownish solution, dicyclohexylcarbodiimide (247 mg, 1.2 equiv) was added in one

portion. The solution turned into a suspension in less than 1 min. The mixture was

further stirred at rt (21 °C) for 16 h (the reaction is complete after ca 4 h). After full

conversion of the alcohol (TLC, NMR), the reaction mixture was directly transferred

onto a silicagel column and chromatographed using 5% Et2O in pentane.

The reaction afforded 658.8 mg of desired product as colorless liquid (89% yield)

1H NMR (400 MHz, CDCl3) δ 5.15 – 4.97 (m, 1H), 4.34 (dd, J = 11.8, 3.7 Hz, 1H),

4.16 (dd, J = 11.8, 6.3 Hz, 1H), 3.71 (m, 2H), 2.33 – 2.27 (m, 4H), 1.60 (d, J = 6.8

Hz, 4H), 1.27 (d, J = 11.1 Hz, 48H), 0.88 (s, 16H), 0.83 (d, J = 6.5 Hz, 4H), 0.05 (s,

6H).

13C NMR (101 MHz, CDCl3) δ 173.57, 173.22, 71.82 ( - ), 62.59, 61.61, 37.25, 34.50,

34.31, 32.91 ( - ), 32.08, 30.20, 30.14, 29.86, 29.82, 29.79, 29.68, 29.64, 29.52, 29.47,

29.45, 29.30, 29.26, 27.25, 27.23, 25.90 ( - ), 25.10, 25.07, 22.84, 19.85( - ), 18.35,

14.26 ( - ), -5.33 ( - ), -5.37 ( - ).

A Methyl Matters

115

Pag

e11

5

5

HRMS-ESI+ (m/z): [M + H]+ calculated for C44H89O5Si: 725.647, found: 725.649

[α]D = 5.1 (c = 2.0, CHCl3)

(R)-(R)-3-((tert-butyldimethylsilyl)oxy)-2-(palmitoyloxy)propyl tuberculostearate

The reaction afforded 632 mg of the desired product (86%) as a colorless liquid.

1H NMR (400 MHz, CDCl3) δ 5.15 – 4.97 (m, 1H), 4.34 (dd, J = 11.8, 3.7 Hz, 1H),

4.16 (dd, J = 11.8, 6.3 Hz, 1H), 3.71 (m, 2H), 2.33 – 2.27 (m, 4H), 1.60 (d, J = 6.8

Hz, 4H), 1.27 (d, J = 11.1 Hz, 48H), 0.88 (s, 16H), 0.83 (d, J = 6.5 Hz, 4H), 0.05

(s, 6H).

13C NMR (101 MHz, CDCl3) δ 173.57, 173.22, 71.82 ( - ), 62.59, 61.61, 37.25, 34.50,

34.31, 32.91 ( - ), 32.08, 30.20, 30.14, 29.86, 29.82, 29.79, 29.68, 29.64, 29.52, 29.47,

29.45, 29.30, 29.26, 27.25, 27.23, 25.90 ( - ), 25.10, 25.07, 22.84, 19.85 ( - ), 18.35,

14.26 ( - ), -5.33 ( - ), -5.37 ( - ).

HRMS-ESI+ (m/z): [M + H]+ calculated for C44H89O5Si: 725.647, found: 725.649.

[α]D = +4.1 (c = 2.3, CHCl3)

Major lipid precursor

To a solution of (R,R)-3-((tert-butyldimethylsilyl)oxy)-2-(palmitoyloxy)propyl 10-

methyloctadecanoate (81.3 mg, 112 µmol) in CH2Cl2 (1 ml), CH3CN●BF3 (100 µl,

123 µmol, 1.1 eq) was added and the resulting light yellow reaction mixture was

stirred for 5 min, carefully monitored by TLC. Upon full conversion, the reaction

mixture was diluted with Et2O (15 ml) and poured onto chilled sodium phosphate

buffer (1 M, 5 ml). The organic layer was separated and washed with saturated brine

(5 ml), dried and evaporated to dryness to afford (R, S)-3-hydroxy-2-

(palmitoyloxy)propyl 10-methyloctadecanoate (65.6 mg, 96%) as a colorless oil. The

compound was used directly without delay and further purification.

Chapter 5

116

5

1H NMR (400 MHz, CDCl3) δ 5.09 (m, 1H), 4.32 (m, 1H), 4.26 (s, 1H), 3.75 (s, 1H),

2.34 (dd, J = 15.8, 8.1 Hz, 4H), 2.01 (s, 1H), 1.63 (s, 4H), 1.54 (dd, J = 8.2, 4.9 Hz,

2H), 1.27 (d, J = 8.1 Hz, 47H), 0.94 – 0.79 (m, 9H).

To a stirred solution of 1-(R)-TBSA-2-palmitoyl glycerol (63.6 mg, 104 µmol) in

CH2Cl2 (0.5 ml), phosphoramidite 15 (54.0 mg, 125 µmol, 1.2 eq) was added. The

mixture was cooled to 0 °C and 1H-imidazole-4,5-dicarbonitrile (15.4 mg, 0.13 µmol,

1.3 eq) was added in one portion. The reaction was stirred until complete conversion

of the starting diacylglycerol (monitored by TLC - typically 30 min). Subsequently,

the mixture was cooled to –20 °C and tBuOOH (ca 5 M in decane, 38 µl, 0.208

mmol) was added, followed by stirring for 30 min. Then the reaction was diluted

with 10 ml of CH2Cl2 and poured into aqueous NaHCO3 (1 M, 10 ml). The organic

layer was washed with aqueous HCl (1 M, 10 ml), brine, dried and evaporated.

The resulting crude yellow oil was purified by column chromatography on SiO2

(CHCl3 : pentane 9 : 1) to afford the desired product (85 mg, 85%) as a colorless thick

liquid, together with a co-eluting impurity.

1H NMR (400 MHz, CDCl3) δ 7.47 – 7.19 (m, 1H), 5.34 (broad s, 1H), 5.18 (dd, J =

9.6, 5.3 Hz, 1H), 5.11 – 5.02 (m, 4H), 4.31 – 4.22 (m, 1H), 4.17 – 4.01 (m, 4H), 3.42

(m, 2H), 2.32 – 2.24 (m, 4H), 1.57 (d, J = 7.0 Hz, 4H), 1.46 – 0.99 (m, 52H), 0.88 (t,

J = 6.8 Hz, 6H), 0.83 (d, J = 6.5 Hz, 9H).

13C NMR (101 MHz, CDCl3) δ 173.19, 172.81, 128.80( - ), 128.67( - ), 128.47 ( - ),

128.09 ( - ), 128.03 ( - ), 69.80, 66.78, 65.49, 61.56, 41.33, 37.09, 34.10, 33.97, 32.75 (

- ), 31.91, 30.02, 29.97, 29.68, 29.64, 29.52, 29.47, 29.35, 29.28, 29.11, 29.06, 27.08,

24.81, 22.67, 19.69 ( - ), 14.10 ( - ).

31P NMR (162 MHz, CDCl3) δ 0.81, 0.83.

HRMS-ESI+ (m/z): [M + H]+ calculated for C55H93O10NP: 958.653, found 958.653.

[α]D = +7.1 (c = 2.3, CHCl3)

Epimer of the major TBSA lipid

Starting from 103 mg of protected diacylglycerol, the reaction afforded 120 mg (93%)

of the desired product.

A Methyl Matters

117

Pag

e11

7

5

1H NMR (400 MHz, CDCl3) δ 7.47 – 7.19 (m, 10H), 5.34 (broad s, 1H), 5.18 (dd,

J = 9.6, 5.3 Hz, 2H), 5.11 – 5.02 (m, 4H), 4.31 – 4.22 (m, 1H), 4.17 – 4.01 (m, 4H),

3.42 (m, 2H), 2.32 – 2.24 (m, 4H), 1.57 (d, J = 7.0 Hz, 4H), 1.46 – 0.99 (m, 52H),

0.88 (t, J = 6.8 Hz, 6H), 0.83 (d, J = 6.5 Hz, 9H).

13C NMR (101 MHz, CDCl3) δ 173.19, 172.81, 128.80 ( - ), 128.67 ( - ), 128.47 ( - ),

128.09 ( - ), 128.03 ( - ), 69.80, 66.78, 65.49, 61.56, 41.33, 37.09, 34.10, 33.97,

32.75( - ), 31.91, 30.02, 29.97, 29.68, 29.64, 29.52, 29.47, 29.35, 29.28, 29.11, 29.06,

27.08, 24.81, 22.67, 19.69 ( - ), 14.10 ( - ).

31P NMR (162 MHz, CDCl3) δ 0.81, 0.83.

HRMS-APCI+ (m/z): [M + Na]+ calculated for C55H92NO10PNa: 980.635 found

980.635.

Synthesis of the phospholipids

Phospholipid 1

To a stirred solution of the TBSA lipid precursor (50 mg, 52 µmol) in MeOH/formic

acid (2 ml, 96/4), Pd/C (Sigma-Aldrich, Degussa Type E101 NE/W, 3 mg, 2.6 µmol,

5 mol%) was added. The suspension was stirred under a hydrogen atmosphere

(balloon) until complete conversion of the starting material (typically 2 h, according

to TLC). Subsequently, the solution was diluted with CH2Cl2 (10 ml) and SiO2 (2 g)

was added followed by evaporation of the volatiles. The SiO2 with adsorbed

phospholipid was transferred onto a short (5 g) SiO2 column (Sigma-Aldrich, Silica

gel, high-purity grade (Davisil Grade 633), pore size 60 Å, 200-425 mesh particle

size), impurities were eluted with Et2O (100 ml) followed by elution of the

phospholipid with CHCl3/MeOH/H2O (65 /35/7) to afford the M. tuberculosis

phospholipid 1 (32.6 mg, 0.044 mmol, 87%) as a white sticky solid.

1H NMR (400 MHz, CDCl3/CD3OD/D2O, v/v, 95/35/2) δ 5.19 (s, 1H), 5.07 (s,

1H), 4.36 (d, J = 11.3 Hz, 1H), 4.21 – 3.80 (m, 4H), 3.64 (s, 4H), 3.37 (s, 1H), 3.09

(s, 1H), 2.27 (dd, J = 15.5, 8.1 Hz, 4H), 1.56 (s, 4H), 1.48 – 0.98 (m, 44H), 0.98 –

0.77 (m, 9H).

Chapter 5

118

5

13C NMR (101 MHz, CDCl3/CD3OD/D2O v/v 95/35/2 ) δ 173.42, 173.14, 128.41

( - ), 127.97 ( - ), 37.14, 34.24, 34.06, 32.79 ( - ), 31.92, 30.05, 29.75, 29.70, 29.42,

29.37, 29.25, 27.15, 27.11, 24.94, 24.87, 22.68, 19.65 ( - ), 14.10 ( - ). 31P NMR (162 MHz, CDCl3) δ +0.29

HRMS-ESI+ (m/z): [M + H]+ calculated for C40H81NO8P: 734.570, found 730.569.

[α]D = +7.0 (c = 0.3, CHCl3).

Spectral data correspond to those previously published17

Phospholipid 2

Starting from 72 mg of the precursor reaction afforded 50 mg (91%) of the desired

compound.

1H NMR (400 MHz, CDCl3/CD3OD/D2O, v/v, 95/35/2) δ 5.24 (d, J = 4.2 Hz,

1H), 4.40 (dd, J = 12.1, 2.9 Hz, 2H), 4.17 (dd, J = 12.1, 7.1 Hz, 1H), 4.08 (t, J = 8.4

Hz, 1H), 3.99 (t, J = 6.0 Hz, 2H), 3.24 – 3.13 (m, 2H), 2.33 (dd, J = 15.7, 8.2 Hz,

4H), 1.66 – 1.56 (m, J = 5.7 Hz, 4H), 1.27 (bs, 50H), 0.89 (t, J = 6.8 Hz, 6H), 0.84

(d, J = 6.5 Hz, 3H)

13C NMR (101 MHz, CDCl3/CD3OD/D2O, v/v, 95/35/2) δ 173.90, 173.55, 77.48,

77.16, 76.84, 62.50, 61.59, 40.17, 36.92, 34.04, 33.90, 32.57 ( - ), 31.73, 29.83, 29.51,

29.49, 29.46, 29.39, 29.37, 29.16, 28.98, 28.94, 26.89, 24.73, 24.67, 22.47, 19.41 ( - ),

13.78 ( - ).

31P NMR (162 MHz, CDCl3/CD3OD/D2O, v/v, 95/35/2) δ 0.65.

HRMS-ESI+ (m/z): [M + H]+ calculated for C40H81NO8P: 734.569, found 734.569.

Phospholipid 3

To a solution of (R)-(R)-3-((tert-butyldimethylsilyl)oxy)-2-(palmitoyloxy)propyl 10-

methyloctadecanoate (200 mg, 280 µmol) in CH2Cl2 (2.8 ml), CH3CN●BF3 (220 µl,

1.1 eq) was added and the resulting light yellow reaction mixture was stirred for 5

A Methyl Matters

119

Pag

e11

9

5

min carefully monitored by TLC. Upon full conversion, the reaction mixture was

diluted with Et2O (30 ml) and poured onto chilled phosphate buffer (1 M, 10 ml).

The organic layer was separated and washed with saturated brine (10 ml), dried and

evaporated to dryness to afford (R)-(S)-3-hydroxy-2-(palmitoyloxy)propyl 10-

methyloctadecanoate (151 mg, 90%) as a colorless oil. The compound was used

directly without delay and further purification.

1H NMR (400 MHz, CDCl3) δ 5.09 (m, 1H), 4.32 (m, 1H), 4.26 (s, 1H), 3.75 (s, 1H),

2.34 (dd, J = 15.8, 8.1 Hz, 4H), 2.01 (s, 1H), 1.63 (s, 4H), 1.54 (dd, J = 8.2, 4.9 Hz,

2H), 1.27 (d, J = 8.1 Hz, 47H), 0.94 – 0.79 (m, 9H)

2-Chloro-2-oxo-1.3.2-dioxaphospholane (distilled under reduced pressure at 150 °C,

and 3.6 mbar, before use, 80 μl, 3 equiv) was added to a chilled solution of

diacylglycerol. To this solution, DMAP (103 mg, 3 equiv) was added. The reaction

was allowed to reach rt and stirred for 24 h at rt (21 °C). The reaction was diluted

with Et2O (20 ml), and the organic layer was washed with water and brine, dried and

evaporated. The crude residue (ca 170 mg) was dissolved in CH2Cl2 (2 ml) and cooled

to 0 °C with an ice bath. To this solution, NMe3 (125 μl, 4 equiv) was added using a

plastic syringe wrapped in cotton previously dipped in acetone/liquid N2.

Subsequently, TMSOTf (100 μl, 2 equiv) was added. The mixture was stirred until

full conversion of the starting material. All volatiles were evaporated and the crude

residue was purified by column chromatography on silica gel (Davisil high purity

silica gel Grade 633, pore size 150 Å, 200-425 mesh particle size) carefully using a

gradient from 1% MeOH to 30% MeOH in CHCl3. The reaction afforded 73.5 mg

of desired product (34% over three steps).

1H-NMR was inconclusive because of strong significant signal-broadening.

13C NMR (101 MHz, CDCl3/CD3OD/D2O, v/v, 95/35/2) δ 173.54, 173.18, 70.14,

70.04, 69.95, 63.12, 62.37, 58.87, 58.82, 55.07, 55.03, 53.72 ( - ), 51.15 ( - ), 36.68,

33.85, 33.70, 32.33 ( - ), 31.50, 29.60, 29.28, 29.26, 29.23, 29.16, 29.13, 28.93, 28.75,

28.70, 26.66, 24.54, 24.47, 22.25, 19.22 ( - ), 13.60 ( - ).

31P NMR (162 MHz, CDCl3/CD3OD/D2O, v/v, 95/35/2) δ 2.94.

HRMS-ESI+ (m/z): [M + H]+ calculated for

This part of the research was conducted by Mac Donald José

Preparation of large unilamellar vesicles (LUV) or liposomes

Chapter 5

120

5

20% total soy extract (Avanti polar lipids) was dissolved in lipid buffer (10 mM

sodium phosphate and 150 mM NaCl, pH 8) to 20 mg/ml. The dissolved lipid was

then passed through an alternate five cycles of freezing in liquid nitrogen and thawing

in a water bath set at 50 oC. The sample was stored at -20 oC until use.

For the synthetic lipids, both pure lipids and lipid mixtures were dissolved first in

chloroform and dried using a rotatory evaporator under vacuum for 1.5 h. After

drying, the lipids were dissolved (10 mg/ml) in lipid buffer (10 mM sodium

phosphate and 150 mM NaCl, pH 8) and passed through five cycles of freeze and

thaw just like soy lipid extract. The liposomes were composed as follows:

Lipid mixture SR: 1:1 mixture of compounds 2 and 3 (18:(10-(S)Me)16:0 PE 18:(10-

(R)Me)16:0 PC)

Lipid mixture RR: 1:1 mixture of compounds 1 and 2 (18:(10-(R)Me)16:0 PE and

18:(10-(R)Me)16:0 PC)

Lipid mixture S0: 1:1 mixture of compounds 3 and 4 (18:(10-(S)Me)16:0 PE and

18:0 16:0 PC)

Soy extract (control-100%)

Proteoliposomes reconstitution

The lipids were then extruded through a polycarbonate filter of 400 nm by passing

to and fro eleven times. Two aliquots of the extruded lipids (300 l) were destabilized

by addition of Triton X-100 for five min in a waterbath set to 50 oC (the amount of

detergent for destabilization was different for each lipid and was found by prior

titration). To one portion of lipid sample MscL cysteine mutant protein from M.

tuberculosis (A20C) was added and to another portion of the same lipid MscL cysteine

mutant protein from E. coli (G22C)was added to a protein to lipid ratio of 1 to 50

(wt/wt). A control sample was prepared which contained no protein for monitoring

unspecific release from the liposomes. This was followed by incubation at 50 oC for

30 min. After incubation, one volume of calcein (200 mM calcein, 10 mM sodium

phosphate pH 8) was added to each portion. About 200 mg (wet weight) of biobeads

(SM-2-Absorbents) were added to each tube to remove the detergent during

overnight incubation at 4 oC. The excess calcein was removed from the

proteoliposomes by passing the samples through a Sephadex G50 size exclusion

column using efflux buffer (150 mM NaCl, 10 mM sodium phosphate, 1 mM EDTA

pH 8).

Calcein efflux assay

A Methyl Matters

121

Pag

e12

1

5

The activity of the protein in the calcein filled proteoliposomes was determined by

following the increase in fluorescence at an excitation wavelength of 495 nm and an

emission wavelength of 515 nm. Into 2.2 ml of efflux buffer (150 mM NaCl, 10 mM

sodium phosphate, 1 mM EDTA pH 8) 2-5 l proteoliposomes were added and

allowed to equilibrate for about 50 s before adding the trigger. Two triggers were

used, MTSET to a final concentration of 1 mM and LPC (lysophosphatidyl choline)

to a final concentration of 4.5 M. The release was followed for about 12 min at

which point complete release was observed. Another portion of detergent (Triton)

to a final concentration of about 0.5% was added to destroy all the proteoliposome

to obtain maximum calcein release.


(1) Elferink, M. G. L.; de Wit, J. G.; Driessen, A. J. M.; Konings, W. N. Biochim. Biophys.

Acta - Biomembranes 1994, 1193, 247.

(2) Marr, A. G.; Ingraham, J. L. J. Bacteriol. 1962, 84, 1260.

(3) (a) Bezbaruah, R. L.; Pillai, K. R.; Gogoi, B. K.; Baruah, J. N. Antonie van Leeuwenhoek

1988, 54, 37(b) Jurado, A. S.; Pinheiro, T. J. T.; Madeira, V. M. C. Arch. Biochem.

Biophys. 1991, 289, 167.

(4) Langworthy, T. A. J. Bacteriol. 1977, 130, 1326.

(5) Gabriel, J. L.; Chong, P. L. G. Chem. Phys. Lipids 2000, 105, 193.

(6) Lim, J. B.; Klauda, J. B. Biochim. Biophys. Acta - Biomembranes 2011, 1808, 323.

(7) Chang, G.; Spencer, R. H.; Lee, A. T.; Barclay, M. T.; Rees, D. C. Science 1998, 282,

2220.

(8) Hamill, O. P.; Marty, A.; Neher, E.; Sakmann, B.; Sigworth, F. J. Pflugers Archiv - EJP

1981, 391, 85.

(9) Kocer, A.; Walko, M.; Feringa, B. L. Nat. Protoc. 2007, 2, 1426.

(10) Okuyama, H.; Kankura, T.; Nojima, S. J. Biochem. 1967, 61, 732.

(11) ter Horst, B.; Seshadri, C.; Sweet, L.; Young, D. C.; Feringa, B. L.; Moody, D. B.;

Minnaard, A. J. J. Lipid Res. 2010, 51, 1017.

(12) This is partially supported by DSC measurment where no transition was observed

between -10 and 80 °C.

(13) transition temperature at 43 °C

Chapter 5

122

5

123

Chapter 6 Synthesis of a Cyclooctyne–based Lipidation Probe

Abstract: Protein lipidation is an important, relatively unexplored posttranslational

modification. Usually, the protein lipidation is studied with radioactive or fluorescent

radioactive probes. With the development of bioorthogonal chemical reporters,

azides and alkynes fatty acid analogues became important tools for the lipidation

studies. However, a probe which could be used in living cells is still lacking. This

chapter describes the design and synthesis of a cyclooctyne–based fatty acid that can

be potentially applied in the study of the protein lipidations.

Chapter 6

124

6

Introduction

Previous chapter (Chapter 5) discussed the influence of lipids on the function

of the membrane-embedded proteins via lipid-protein interactions. Lipids can

control the function also in a different way. Protein lipidation is a posttranslational

modification in which hydrophobic groups are attached to the protein, thus altering

its activity and subcellular location. Depending on the lipidation pattern, modified

proteins can be associated with the inner or the outer membrane leaflet.

Proteins modified with a glycosylphosphatidylinositol-anchor (GPI-anchor,

Figure 1) are localized at the outer leaflet of the membrane towards the extracellular

space.

Figure 1. GPI-anchor modification.

This very complex anchor consists of a lipophilic portion, a glycan core and

a phosphoethanolamine linkage. The phosphoethanolamine linkage attaches to the

C-terminus of the protein. GPI-anchor modified proteins fulfil diverse biological

roles. Usually, they are signal transduction enzymes, antigens or receptors.1

Synthesis of a Cyclooctyne-based Lipidation Probe

125

Pag

e12

5

6

The lipidation patterns for anchoring to the inner leaflet of the membrane

are less complex (Figure 2). Typically, these are acyl residues bound to the N-

terminus of the protein (N-myristoylation or N-palmitoylation), prenyl residues

bound to the sulfur of a terminal cysteine (S-farnesylation and S-geranylgeranylation)

or a phosphatidylethanolamine residue bound to the C-terminus of the protein (C-

phosphatidylethanolaminylation).

Figure 2. Lipidation patterns for anchoring to the intracellular leaflet of the membrane.

Another common lipidation is acylation of the sulphur of cysteine (S-palmitoylation).

In comparison to the other lipidation patterns, the S-palmitoylation is specific,

because it is reversible, thus more challenging to study.

Chapter 6

126

6

Lipidations play a central role in many physiological processes and diseases.2

For example, Fearson et al.3 reported that inhibitors of N-myristoyltransferase of

Trypanosoma brucei are promising pharmaceutical leads for treatment of the African

sleeping sickness. S-farnesylation of lamin A protein plays an important role in the

Hutchinson-Gilford progeria syndrome.4 Inhibitors of the corresponding farnesyl

transferase5 improved the disease symptoms and survival in mice. Proteins modified

by C-phosphatidylethanolamination are important in autophagy, cellular homeostasis

and infection resistance.6 The cycle of S-palmitoylation, depalmitoylation and

repalmitoylation is vital for the localization and signalling activity of Ras proteins.7

Inhibition of the corresponding palmitoylase reduced the growth of human tumour

cells.8

The importance of lipidated proteins motivated the development of many

tools and procedures for their study.9 Traditionally, palmitoylation has been studied

by radioactively-labelled palmitic acid analogues 1 and 2 (Figure 3).10

Figure 3. Radioactively-labelled palmitic acid analogues.

Although this method is very sensitive and the change in the molecular

properties are minimal, 1 and 2 are radioactive, which can be seen as a disadvantage.

To address the radioactivity issue, Dursina et al. 11 developed the fluorescently–

labelled prenyl analogue 3.11 The authors also applied 3 for the development of

farnesyl and geranylgeranyl transferase inhibitors.

Figure 4 . Fluorescently–labelled prenyl analogue.

The development of bio-orthogonal chemical reporters brought new

opportunities for the lipidation studies. Hang et al. 12 successfully applied 4 and 5

(Figure 5) in the studies of the dynamic S-palmitoylation and protein turnover in a

cell lysate.


127

Pag

e12

7

6

Figure 5. Bio-orthogonal probes used to study protein lipidation.

Logical progress in the field would be the utilization of a copper-free

cycloaddition to monitor the lipidation in the living cells. However, this tool is still

lacking. An unique example of application of the copper-free cycloaddition in a lipid

related study was reported by Schultz and Neef.13 The authors developed and applied

6 (Figure 6) to visualize membrane lipids in living cells.

Figure 6. An analogue of a diacylphosphoglycerol to visualize membrane lipids in living cells.

One of the reasons for the slow development of strain-promoted probes in

the lipid research is the general trend towards reactive and hydrophilic probes14 for

labelling of proteins, whereas for lipid studies hydrophobic probes are required.

Design of a lipophilic lipidation probe

The design of a suitable lipophilic lipidation probe can be based on the

recently reported15 bicyclononanes 7 or 8 (Figure 7).

Figure 7. Strained cyclooctynes.

Chapter 6

128

6

These are readily available (4 steps) and undergo a fast, strain promoted dipolar

cycloaddition. Even though cyclooctynes 9, 10 and 11 display higher rate of dipolar

cycloaddition, their synthesis is longer and their structure significantly deviates from

the hydrocarbon nature of fatty acids. The characteristics cyclooctynes 7-11 are

summarized in the following table (Table 1).

Table 1. Comparison of the rate and synthesis of cyclooctynes 7-11.

Entry Cyclooctyne number of steps Yield Ratea

(10-3 M-1s-1)

115 7 4 15% 110

215 8 4 35% 140

316 9 6 11% 240

417 10 9 41% 410

518 11 6 18% 960 arate constant determined in reaction with benzyl azide as a model compound.


12 (Figure 8) can be a suitable lipophilic lipidation probe. The bicyclononyne

can be attached to a fatty acid by a Wittig reaction, thus giving 12 (Figure 8) as the

desired compound. Given that the strained triple bond is the most sensitive

functionality in 12 (Figure 8), this can be introduced in the final step using a double

Figure 8. Retrosynthesis of 12.


129

Pag

e12

9

6

HBr elimination. Dibrominated 13 can be prepared by a Wittig reaction and double

bond reduction sequence. Synthesis of functionalized ylide 14 from 15 was described

in a previous chapter (Chapter 2). The aldehyde 16 can be easily obtained from

bicyclononenol 17 via bromination and oxidation. And finally, 17 is available via

cyclopropanation of 18 and subsequent reduction.15

The synthesis towards 12 (Scheme 1) started with the cyclopropanation of 18

according to previously described conditions.15 Purification by chromatography

afforded 20 and 21 in 30% and 51% yield, respectively.

Reagents and conditions: a) ethyl diazoacetate (0.13 equiv) added over 12 h, Rh2(OAc)4 (2.0 mol%), CH2Cl2, 24 h, 21 °C; b) LiAlH4 (1.0 equiv), 0 °C, then 10 min at 21 °C, Et2O; c) pyridinium bromide perbromide (1.2 equiv), CH2Cl2, 1 h, 21 °C; d) TPAP (5.0 mol%), NMO (1.5 equiv), 3 Å molecular sieves, CH2Cl2, 1 h.

Scheme 1. Attempted synthesis of the lipidation probe.

The synthesis continued with reduction of the exo ester 21 with LiAlH4, yielding

alcohol 17 in quantitative yield. The subsequent bromination with Br2 according to

the described conditions15 resulted in decomposition of 21. A slight excess of

pyridinium bromide perbromide was an excellent alternative for Br2 affording the

desired 22 in 95% yield. Oxidation of 22 with TPAP/NMO afforded the aldehyde

16 in 65% yield. Despite a substantial effort, the planned Wittig reaction did not

afford the desired 23, but resulted in the decomposition of the starting material.

The unsuccessful Wittig reaction forced a new design of the lipidation probe,

in which the bicyclononyl moiety is bound to the linker via an ether bond (24 in

Figure 9). The strained triple bond is available (again) via a debromination of 25.

Chapter 6

130

6

Figure 9. Retrosynthetic analysis of a fatty acid probe with an ether linkage.

Acid 25 can be synthetized from alcohol 26 by oxidation, which in turn, can be

obtained by bromination and deprotection of 27. Finally, 27 is available from 17 and

28 via a Williamson ether synthesis.

The synthesis of the ether-linked clickable fatty acids was straightforward

(Scheme 2). The steps of the synthesis were carried out with both the endo and the

exo stereo-isomer (separately), but for clarity only the exo-isomer is discussed. The

yields for the endo stereoisomer are showed in brackets. Alkylation of 17 in basic

conditions afforded 30 in 65% yield (endo-30 in 48%) based on recovered 17. Despite

a substantial number of experiments, the competing elimination of 29 could not be

fully suppressed. Ether 30 was further converted into dibromide 26 in 2 steps. First,

the tetrahydropyranyl (THP) group was removed by refluxing in methanol in

presence of Amberlite (acidic form). Amberlite was removed by filtration, and the

intermediate alcohol was brominated using pyridinium bromide perbromide.

Dibromide 26 was isolated in 86% yield over 2 steps after flash chromatography

(endo-26 in 95% yield). Subsequent oxidation of 26 to acid 25 was carried out with

TPAP/NMO to the corresponding aldehyde followed by treatment with oxone in

DMF, affording 25 in 80% yield (endo-25 in 71%). Finally, 25 was debrominated by

refluxing in THF in the presence of excess KOtBu. The lipidation probe exo-24 was

obtained in 99% yield (endo-24 71%).


131

Pag

e13

1

6

Reagents and conditions: a) NaH (1.2 equiv) DMF 2 h, then 29 (1.2 equiv), 16 h, 0- 21 °C, b) Amberlite (100 mg/1 mmol), MeOH, 4 h, reflux c ) TPAP (5 mol%), NMO (2.0 equiv), CH2Cl2, 4 h, 0 °C then Oxone (3.0 equiv), DMF, 16 h, 21 °C; d) KOtBu (3.3 equiv), THF, 2 h, reflux.

Scheme 2. Synthesis of clickable fatty acid 24.

The reactivity of the clickable fatty acid 24 was explored in a dipolar

cycloaddition in an aqueous solution (Scheme 3). 24 underwent full conversion with

a stoichiometric amount of 31 in phosphate buffer (100 mM NaHPO4, pH 7.4) at 37

°C, in 20 min. Triazole 32 was isolated in 99% yield.

Reagents and conditions: a) 31 (1.0 equiv), sodium phosphate buffer (100 mM, pH = 7.4), 20 min, 37 °C.

Scheme 3. Strain promoted dipolar cycloaddition of 24.

Conclusions and outlook

A cyclooctyne based fatty acid is accessible in 6 steps from 1,5-cycloocta-

diene. The lipidation probe exo-24 undergoes smooth and fast dipolar cycloaddition

with an azide, resulting in a fluorescent adduct.

Chapter 6

132

6

Preliminary study showed, that 24 is metabolised by eukaryotic cells.

However, a further study is required in order to determine if 24 is incorporated to

the lipids or proteins.

Experimental part

Synthesis of bicyclononanols (scheme 1).

Exo- and endo- ethyl bicyclo[6.1.0]non-4-ene-9-carboxylate (20 and 21)

Prepared as a mixture according to Van Delft et al. with a modification in

the reagent addition time and the eluent for flash chromatography.

To the stirred mixture of freshly distilled cycloocta-1,5-diene (120 ml, 0.96 mol.

8.0 equiv) and Rh2(OAc)4 (106 mg, 0.2 mmol, 0.2 mol%) in CH2Cl2 (120 ml) ethyl

diazoacetate (15 ml, ca 0.12 mmol) was added dropwise over 12 h. After the addition

was complete, the mixture was stirred for 24 h at RT (21 °C). The crude reaction

mixture was filtered over a short silica pad. The resulting filtrate was fractionally

distilled to remove CH2Cl2 (800 mbar, 40 °C), and the unreacted 1,5-cyclooctadiene

(20 mbar, 70 °C) on a rotatory evaporator. The distillation residue was purified by

flash chromatography over silica gel using 2% diisopropyl ether in pentanes.

The reaction afforded 7.2 g (30%) of exo-isomer 11, 10.0 g (43%) of endo-isomer 12

and 2.2 g (10%) of a mixed fraction that was further separated by a second column

chromatography using the same eluent to afford 2 g of 12 (8%)

exo-ethyl bicyclo[6.1.0]non-4-ene-9-carboxylate (20)

1H NMR (400 MHz, CDCl3, δ): 5.62 (m, 2H), 4.08 (q, J = 7.1, 2H), 2.29 (m, 2H),

2.18 (m, 2H), 2.16 (m, 2H), 1.49 (m, 4H), 1.23 (t, J = 7.1, 3H), 1.17 (t, J = 4.5, 1H).

13C NMR (100 MHz, CDCl3, δ): 174.6, 130.1( - ), 60.4, 28.4, 28.11, 27.95, 26.87 ( - ),

14.52 ( - ).

endo-Ethyl bicycle[6.1.0]non-4-ene-9-carboxylate (21)


133

Pag

e13

3

6

1H NMR (400 MHz, CDCl3, δ): 5.60 (m, 2H), 4.10 (q, J = 7.1, 2H), 2.49 (m, 2H), 2.19 (m, 2H), 2.04 (m, 2H), 1.81 (m, 2H), 1.68 (m, 1H), 1.39 (m, 2H), 1.25 (t, J = 7.1, 3H).

13C NMR (100 MHz, CDCl3, δ): 172.47, 129.63 ( - ), 59.90, 27.27, 24.37,

22.85, 22.13 ( - ), 14.60 ( - ).

The NMR data are in agreement with previously reported.15

exo-bicyclo[6.1.0]non-4-en-9-ylmethanol (17)

20 (9.65 g, 50 mmol) was dissolved in Et2O (300 ml). The solution was

cooled on an ice bath, and LiAlH4 (920 mg, 50 mmol, 1.0 equiv) was added

in small portions (GAS EVOLUTION) over 10 min. After complete

addition, the cooling bath was removed, and the resulting grey suspension

was stirred for 15 min at RT (21 °C). The reaction was monitored by TLC. As soon

as the starting material was consumed, the reaction mixture was immersed again into

an ice bath. Water (1 ml) was added with great caution (GAS EVOLUTION). After

stirring for 10 min, aqueous NaOH (15%, 3 ml) was added. The grey suspension

turned into a white suspension to which water (1 ml) was added. The mixture was

stirred for 15 min and then filtered over a Celite pad.

After evaporation of the volatiles, 7.35 g (99 %) of alcohol 13 was obtained.

1H NMR (400 MHz, CDCl3, δ): 5.63 (m, 2H), 3.71 (d, J = 7.6, 2H), 2.36 (m, 2H),

2.10 (m, 2H), 1.98 (m, 2H), 1.58 (m, 2H), 1.31 (s, 1H), 1.13 (m, 1H), 1.00 (m, 2H).

13C NMR (100 MHz, CDCl3, δ): 130.22 ( - ), 60.73, 28.17, 24.38, 21.25 ( - ), 19.49

( - ).

Anal. Calcd for C10H16O: C, 78.90; H, 10.59. Found C, 78.70; H, 10.77.

endo-bicyclo[6.1.0]non-4-en-9-ylmethanol.

The same procedure as for the exo isomer was employed. Endo-12 (6.84 g,

35 mmol) was dissolved in Et2O (200 ml) and treated with LiAlH4 (650 mg, 35 mmol,

1.0 equiv).

Chapter 6

134

6

The reaction afforded 5.31 g (99%) of alcohol 14.

1H NMR (400 MHz, CDCl3, δ): 5.62 (m, 2H), 3.45 (d, J = 6.9, 2H), 2.26 (m, 2H),

2.15 (m, 2H), 2.5 (m, 2H), 1.40 (m, 3H), 0.76 (m, 2H), 0.64 (m, 1H).

13C NMR (100 MHz, CDCl3, δ): 130.65 ( - ), 67.69, 29.51, 29.34, 27.57 ( - ), 22.60 ( - ).

Anal. Calcd for C10H16O: C, 78.90; H, 10.59. Found: C, 78.60; H, 10.78.

Synthesis of lipid probe 24 (scheme 2)

2-((9-bromononyl)oxy)tetrahydro-2H-pyran (29)

In a 250 ml flask, 9-bromo nonanol (10.4 g, 46 mmol) was

dissolved in CH2Cl2 (46 ml). To this solution, 3.4-dihydro-2H-pyran (13 ml, 0.14 mol,

3 equiv) and pyridinium p-toluene sulphonate (1.17 g, 4.6 mmol, 10 mol%) were

added. The resulting suspension was immersed into a pre-heated oil bath (65 °C) and

refluxed until full conversion of the starting alcohol as monitored by TLC

(typically 3 h). The flask was removed from the bath and allowed to cool down. All

volatiles were evaporated using a rotatory evaporator. The resulting slurry was

dissolved in Et2O (150 ml), washed with water, brine, dried and evaporated. The

crude liquid residue was purified by column chromatography using 5% Et2O in

pentane.

The reaction afforded 14.0 g (>99%) of the desired product as a colorless liquid.

1H NMR (400 MHz, CDCl3, δ): 4.57 (m, 1H), 3.87 (m, 1H), 3.73 (m, 1H), 3.50

(m, 1H), 3.38 (m, 3H), 1.84 (m, 3H), 1.71 (m, 1H), 1.56 (m, 6H), 1.35 (m, 10H).

13C NMR (100 MHz, CDCl3, δ): 99.02 ( - ), 77.48, 77.16, 76.84, 67.80, 62.53, 34.20,

32.97, 30.94, 29.88, 29.51, 28.85, 28.30, 26.34, 25.66, 19.87.

NMR data correspond to those previously published.19

Endo-(Z)-2-((9-(bicyclo[6.1.0]non-4-en-9-ylmethoxy)nonyl)oxy)tetrahydro-2H-

pyran (30)

A dry Schlenk flask was charged with NaH (60% in oil,

672 mg, 17 mmol, 1.2 equiv). The mineral oil was

removed by washing with pentane (4 x 15 ml), and the resulting white powder was


135

Pag

e13

5

6

suspended in DMF (dry, 70 ml). This suspension was immersed to an ice bath

(ice/brine) and 17 (2.12 g, 14 mmol) was added dropwise (GAS EVOLUTION).

The mixture was removed from the ice bath and stirred at RT (21 °C) until the gas

evolution ceased (2 h). The mixture was again cooled in an ice bath, and 29 (5.14 g,

17 mmol, 1.2 equiv) was added. The mixture was allowed to slowly reach RT and

stirred overnight (16 h). The obtained solution was diluted with Et2O (75 ml),

transferred into a separatory funnel and washed with water, brine, dried and

evaporated. The resulting liquid was purified using the Reveleris® X2 Flash

Chromatography System with the following gradient: 10 column volumes 5% Et2O

in pentane followed by 3 column volumes of 60% Et2O in pentane.

The reaction afforded 2.38 g (yield based on 65% of recovered starting material) of

desired 30 as colourless liquid and 406 mg of the starting alcohol.


1H), 3.40 (m, 6H), 2.32 (m, 2H), 2.07 (m, 2H), 1.94 (m, 2H), 1.76 (m, 1H), 1.68 (m,

1H), 1.54 (m, 10H), 1.30 (m, 10H), 1.09 (m, 1H), 0,96 (m, 2H).

13C NMR (400 MHz, CDCl3, δ): 129.90 ( - ), 99.00 ( - ), 71.12, 68.04, 67.84, 62.51,

30.94, 29.96, 29.91, 29.71, 29.59, 27.90, 26.38, 26.35, 25.66, 24.06, 19.87, 18.87 ( - ),

18.12 ( - ).

IR (cm-1): 2927, 2853, 1462, 1352, 1106, 1078, 1032.


HRMS-APCI (m/z): [M + Na]+ calculated for C24H43NaO3, 401.303; found, 401.301.

Exo-(Z)-2-((9-(bicyclo[6.1.0]non-4-en-9-ylmethoxy)nonyl)oxy)tetrahydro-2H-pyran

The same procedure as for the synthesis of 30 was

employed, starting from endo-17 (1.23 g, 7.9 mmol).

The reaction afforded 963 mg (48% yield based on the recovered starting material)

of desired endo-17 as colorless liquid and 406 mg of the starting alcohol.


1H), 3.48 (m, 1H), 3.36 (m, 3H), 3.27 (d, J = 6.8, 2H), 2.27 (m, 2H), 2.15 (m, 2H),

Chapter 6

136

6

2.02 (m, 2H), 1.81 (m, 1H), 1.70 (m, 1H), 1.53 (m, 8H), 1.32 (m, 12H), 0.73 (m, 2H),

0.55 (m, 1H).

13C NMR (100 MHz, CDCl3, δ): 130.18 ( - ), 98.83 ( - ), 74.88, 70.41, 67.68, 62.34,

30.78, 29.75, 29.74, 29.53, 29.44, 29.42, 28.96, 27.12, 26.22, 26.16, 25.89, 25.49, 22.31

( - ), 19.70 ( - ).

IR (cm-1): 3008, 2927, 2854, 1454, 1102, 1032.

HRMS-APCI (m/z): [M + Na]+ calculated for C24H43NaO3, 401.303; found, 401.301.


Endo- 9-((4,5-dibromobicyclo[6.1.0]nonan-9-yl)methoxy)nonan-1-ol.

30 (2.38 g, 6.3 mmol) was dissolved in MeOH (63

ml), and Amberlite IR 120 in its H+ form (630 mg,

100 mg/1 mmol of the starting material) was

added. The mixture was immersed into a pre-heated oil bath and refluxed for 4 h.

After full conversion (TLC), the reaction mixture was allowed to cool to RT. All

volatiles were evaporated, and the crude residue was suspended in Et2O (100 ml).

The precipitate was removed by filtration, and the filtrate was evaporated to dryness.

The crude residue was used without further purification in the subsequent reaction.

The crude product from the previous step was dissolved in CH2Cl2 (63 ml) and

pyridinium tribromide (3.02 g, 9.5 mmol, 1.5 equiv) was added in one portion. The

mixture was stirred until full conversion of the starting material (4 h). The conversion

was determined by 1H NMR of reaction mixture aliquots. After full conversion, the

reaction mixture was transferred into a separatory funnel, washed with water, brine,

dried and evaporated to dryness. The crude residue was purified by flash

chromatography over silica using 50% Et2O in pentanes.

The reaction afforded 2.47 g (86%) of the desired product as a colourless liquid.

1H NMR (400 MHz, CDCl3, δ): 4.81 (m, 2H), 3.41 (t, J = 6.7, 2H), 3.33 (d, J = 6.9,

2H), 2.63 (m, 2H), 2.34 (t, J = 7.5, 2H), 2.25 (m, 1H), 2.08 (m, 3H), 1.60 (m, 4H),

1.44 (m, 1H), 1.31 (m, 9H), 0.86 (m, 2H), 0.61 (m, 1H).


137

Pag

e13

7

6

13C NMR (100 MHz, CDCl3, δ): 179.31, 74.55, 70.70, 56.54 ( - ), 53.61 ( - ), 35.18,

35.08, 34.10, 29.87, 29.46, 29.39, 29.20, 26.31, 25.45( - ), 24.86, 24.56, 23.80, 22.93

( - ), 20.23 ( - ).

Anal. Calcd for C19H34Br2O2: C, 50.23; H, 7.53. Found: C, 50.49; H, 7.56.

endo-9-((4,5-dibromobicyclo[6.1.0]nonan-9-yl)methoxy)nonan-1-ol.


employed. Starting from endo-26 (963 mg, 2.6 mmol).

The reaction afforded 1.10 g (95%) of endo-26 as colourless liquid.

1H NMR (400 MHz, CDCl3, δ): 4.82 (m, 2H), 3.64 (t, J = 6.6, 2H), 3.50 (d, J = 7.2,

2H), 3.42 (t, J = 6.7, 2H), 2.67 (m, 2H), 2.21 (m, 2H), 1.91 (m, 2H), 1.57 (m, 5H),

1.45 (m, 2H), 1.32 (m, 10H), 1.19 (m, 2H), 1.07 (m, 1H).

13C NMR (100 MHz, CDCl3, δ): 71.10, 67.37, 63.32, 56.70 ( - ), 53.67( - ), 35.23,

33.02, 30.03, 29.78, 29.63, 29.60, 26.42, 25.95, 20.21, 20.07( - ), 19.41( - ), 19.28,

17.11( - ).


Exo-9-((4,5-dibromobicyclo[6.1.0]nonan-9-yl)methoxy)nonanoic acid (25)

Dibromide 26 (2.03 g, 4.4 mmol) was dissolved in

CH2Cl2 (44 ml). This solution was cooled to 0 °C (brine/ice) and stirred for 5 min.

Then TPAP (77.3 mg, 0.2 mmol, 5.0 mol%) followed by NMO (1.03 g, 8.8 mmol,

2 equiv) were added. The mixture was stirred for 4 h. After full conversion of the

starting material, all volatiles were evaporated, and the crude black residue was used

in the next step.

The crude product from the previous step was dissolved in DMF (44 ml). Oxone (4

g, 13.2 mmol, 3.0 equiv) was added, and the mixture was stirred overnight (16 h) at

RT (21 °C). The reaction mixture was diluted with Et2O (200 ml), transferred into a

separatory funnel and washed with water and brine. The organic layer was dried and

Chapter 6

138

6

evaporated. The crude residue was purified by a Reveleris® X2 Flash

Chromatography System with the following gradient: from 20% Et2O to 60% Et2O.

The reaction afforded 1.64 g of the desired product (80% over 2 steps) as colourless

thick liquid.

1H NMR (400 MHz, CDCl3, δ): 4.82 (dd, J = 12.5, 7.0 Hz, 2H), 3.41 (t, J = 6.8 Hz,

2H), 3.32 (d, J = 6.9 Hz, 2H), 2.64 (m, 2H), 2.34 (t, J = 7.5 Hz, 1H), 2.23 (m, 1H),

2.09 (dd, J = 15.8, 9.9 Hz, 3H), 1.59 (m, 4H), 1.37 (m, 10H), 0.85 (ddd, J = 20.4,

16.1, 7.8 Hz, 2 H), 0.61 (m, 1H).

13C NMR (100 MHz, CDCl3, δ): 179.69, 74.45, 70.61, 56.47 ( - ), 53.53 ( - ), 35.10,

35.00, 34.13, 29.81, 29.40, 29.33, 29.13, 26.24, 25.38 ( - ), 24.80, 24.47, 23.72, 22.83

( - ), 20.12 ( - ).


Endo-9-((4,5-dibromobicyclo[6.1.0]nonan-9-yl)methoxy)nonanoic acid


employed, starting from endo-26 (1.32 g, 2.9 mmol).

The reaction afforded 970 mg (71%) of endo-25 as a colourless thick liquid.

1H NMR (400 MHz, CDCl3, δ): δ 4.88 – 4.77 (m, 2H), 3.50 (d, J = 7.1 Hz, 2H), 3.41

(t, J = 6.7 Hz, 2H), 2.77 – 2.57 (m, 2H), 2.34 (t, J = 7.5 Hz, 2H), 2.31 – 2.20 (m, 1H),

2.14 (dt, J = 15.3, 5.1 Hz, 1H), 1.98 – 1.83 (m, 2H), 1.69 – 1.45 (m, 6H), 1.31 (s, 8H),

1.23 – 1.01 (m, 3H).

13C NMR (100 MHz, CDCl3, δ): 179.63, 70.82, 67.15, 56.44, 53.41, 35.01 ( - ), 34.97

( - ), 33.98, 29.72, 29.24, 29.19, 28.99, 26.11, 24.65, 19.95, 19.83, 19.13, 19.02 ( - ),

16.87 ( - ).


Exo-9-(bicyclo[6.1.0]non-4-yn-9-ylmethoxy)nonanoic acid (24)


139

Pag

e13

9

6

25 (960 mg, 2.0 mmol) was dissolved in THF. The

solution was cooled in an ice bath, and KOtBu (Sigma-Aldrich 1 M solution, 0.75 ml,

6.6 mmol, 3.3 equiv) was added. After complete addition, the reaction mixture was

immersed in a pre-heated oil bath, and the reaction was refluxed for 2 h. After

cooling to RT, the reaction mixture was quenched with saturated aqueous NH4Cl

solution. The mixture was transferred to a separatory funnel and extracted with

CH2Cl2. The combined organic layers were washed with brine, dried and evaporated.

The reaction afforded 630 mg of 24 (99%).

24 could be further chromatographed on silica using a Reveleris® X2 Flash

Chromatography System with following gradient: from 20% Et2O to 60% Et2O.

The chromatography usually resulted in a decreased yield (ca 50%).

1H NMR (400 MHz, CDCl3, δ): 3.47 (d, J = 7.7, 2H), 3.40 (t, J = 6.7, 2H), 2.26 (m,

6H), 1.59 (m, 6H), 1.31 (m, 10H), 0.88 (m, 3H).

13C NMR (100 MHz, CDCl3, δ): 179,44, 99.18, 71.10, 67.76, 34.13, 29.93, 29.45,

29.40, 29.37, 29.21, 26.35, 24.87, 21.72, 20.04 ( - ), 18.89( - ).

HRMS-ESI (m/z): [M + H]+ calculated for C19H31O3, 307.226; found, 307.225.

Endo-9-(bicyclo[6.1.0]non-4-yn-9-ylmethoxy)nonanoic acid


employed, starting from endo-25 (870 mg, 2.6 mmol).

The reaction afforded 570 mg (99 %) of desired endo-22 as a colourless liquid.

1H NMR (400 MHz, CDCl3, δ): 3.42 (t, J = 6.7, 2H), 3.36 (d, J = 6.4, 2H), 2.28 (m,

6H), 1.60 (m, 4H), 1.33 (m, 12H), 0.63 (m, 3H).

13C NMR (100 MHz, CDCl3, δ): 179.44, 99.12, 75.01, 70.71, 34.08, 33.61, 29.94,

29.48, 29.40, 29.21, 26.34, 24.89, 24.67( - ), 22.98( - ), 21.74.

Chapter 6

140

6

HRMS-ESI (m/z): [M + H]+ calculated for C19H31O3, 307.226; found, 307.225.

Strain promoted dipolar cycloadition (scheme 3)

Adduct 32

Round-bottom flask containing sodium

phosphate buffer of pH = 7.4 (100 mM, 5.0 ml) was warmed to 37 °C. In this

solution, 24 (16 mg, 50 μmol) was dissolved. To this solution, a solution of 3-azido-

7-(diethylamino)-2H-chromen-2-one (12.9 mg, 1 equiv) in DMSO (100 μl) was

added. The progress of the reaction was monitored by TLC. After 20 min, full

conversion was reached. Water was evaporated and the crude residue was purified

by flash chromatography using 1% MeOH in CH2Cl2.

The reaction afforded 28 mg of 32 (99%) as orange viscous liquid.

1H NMR (400 MHz, CDCl3, δ): δ 7.75 (s, 1H), 7.28 (s, 1H), 6.45 (d, J = 2.3 Hz, 1H),

6.41 (d, J = 2.5 Hz, 1H), 3.30 (m, 8H), 2.74 (m, 4H), 2.18 (t, J = 7.6 Hz, 2H), 1.20

(m, 22H), 0.66 (m, 3H).

13C NMR (100 MHz, CDCl3, δ): 176.01, 164.13, 158.10, 156.78, 141.89, 128.93,

128.13, 125.20, 116.05, 109.76, 97.17, 79.03, 74.02, 70.39, 44.94, 44.70, 37.72, 34.09,

32.37, 29.60, 29.19, 29.18, 29.16, 29.02, 24.84, 12.35, 12.33.

HRMS-ESI (m/z): [M + H]+ calculated for C32H44N4O5, 556.338; found, 556.335.

References

(1) (a) Ferguson, M. A. J. J. Cell Sci. 1999, 112, 2799(b) Paulick, M. G.; Bertozzi, C. R.

Biochemistry 2008, 47, 6991(c) Low, M. G. Biochim. Biophys. Acta 1989, 988, 427(d)

Ikezawa, H. Biol. and Pharm. Bull. 2002, 25, 409.

(2) Resh, M. D. Trends in Molecular Medicine 2012, 18, 206.

(3) Frearson, J. A.; Brand, S.; McElroy, S. P.; Cleghorn, L. A. T.; Smid, O.; Stojanovski,

L.; Price, H. P.; Guther, M. L. S.; Torrie, L. S.; Robinson, D. A.; Hallyburton, I.;

Mpamhanga, C. P.; Brannigan, J. A.; Wilkinson, A. J.; Hodgkinson, M.; Hui, R.; Qiu,

W.; Raimi, O. G.; van Aalten, D. M. F.; Brenk, R.; Gilbert, I. H.; Read, K. D.;


141

Pag

e14

1

6

Fairlamb, A. H.; Ferguson, M. A. J.; Smith, D. F.; Wyatt, P. G. Nature 2010, 464,

728.

(4) Eriksson, M.; Brown, W. T.; Gordon, L. B.; Glynn, M. W.; Singer, J.; Scott, L.;

Erdos, M. R.; Robbins, C. M.; Moses, T. Y.; Berglund, P.; Dutra, A.; Pak, E.; Durkin,

S.; Csoka, A. B.; Boehnke, M.; Glover, T. W.; Collins, F. S. Nature 2003, 423, 293.

(5) Yang, S. H.; Meta, M.; Qiao, X.; Frost, D.; Bauch, J.; Coffinier, C.; Majumdar, S.;

Bergo, M. O.; Young, S. G.; Fong, L. G. J. Clin. Invest. 2006, 116, 2115.

(6) Levine, B.; Mizushima, N.; Virgin, H. W. Nature 2011, 469, 323.

(7) Rocks, O.; Gerauer, M.; Vartak, N.; Koch, S.; Huang, Z.-P.; Pechlivanis, M.;

Kuhlmann, J.; Brunsveld, L.; Chandra, A.; Ellinger, B.; Waldmann, H.; Bastiaens, P.

I. H. Cell 2010, 141, 458.

(8) Ducker, C. E.; Griffel, L. K.; Smith, R. A.; Keller, S. N.; Zhuang, Y.; Xia, Z.; Diller,

J. D.; Smith, C. D. Mol. Cancer Ther. 2006, 5, 1647.

(9) Hang, H. C.; Linder, M. E. Chem. Rev. 2011, 111, 6341.

(10) Resh, M. D. Methods 2006, 40, 191.

(11) Dursina, B.; Reents, R.; Delon, C.; Wu, Y.; Kulharia, M.; Thutewohl, M.; Veligodsky,

A.; Kalinin, A.; Evstifeev, V.; Ciobanu, D.; Szedlacsek, S. E.; Waldmann, H.; Goody,

R. S.; Alexandrov, K. J. Am. Chem. Soc. 2006, 128, 2822.

(12) Zhang, M. M.; Tsou, L. K.; Charron, G.; Raghavan, A. S.; Hang, H. C. Proc. Nat.

Acad. Sci 2010, 107, 8627.

(13) Neef, A. B.; Schultz, C. Angew. Chemi. Int. Ed. 2009, 48, 1498.

(14) Debets, M. F.; Prins, J. S.; Merkx, D.; van Berkel, S. S.; van Delft, F. L.; van Hest, J.

C. M.; Rutjes, F. P. J. T. Org. Biomol. Chem. 2014, 12, 5031.

(15) Dommerholt, J.; Temming, R.; Hendriks, L. J. A.; Rutjes, F. P. J. T.; Van Hest, J. C.

M.; Van Delft, F. L.; Schmidt, S.; Friedl, P.; Lefeber, D. J. Angew. Chem., Int. Ed.

2010, 49, 9422

(16) Varga, B. R.; Kállay, M.; Hegyi, K.; Béni, S.; Kele, P. Chem. Eur. J. 2012, 18, 822.

(17) Debets, M. F.; van Berkel, S. S.; Schoffelen, S.; Rutjes, F. P. J. T.; van Hest, J. C. M.;

van Delft, F. L. Chem. Commun. 2010, 46, 97.

(18) Jewett, J. C.; Sletten, E. M.; Bertozzi, C. R. J. Am. Chem. Soc. 2010, 132, 3688.

(19) Sharma, A.; Chattopadhyay, S. J. Org. Chem. 1998, 63, 6128.

Chapter 6

142

6

143

Chapter 7 A Missing Link in Archaeal Lipid Biosynthesis; a Contribution from Organic Synthesis

Abstract: The Archaea form a separate domain of Life known to occupy extreme

ecological niches. Their extremophilicity is often associated with their characteristic

membrane lipids, which require different biosynthetic pathways than eukaryotes and

bacteria. Given that the in vivo identification of the involved enzymes is ambiguous,

their function needs to be confirmed by in vitro experiments. However, development

of the detection methods, assay conditions and isolation of the substrates are

challenges on their own. At least with the substrates, organic synthesis can help. This

chapter describes a chemical synthesis of an intermediate in the biosynthesis of

archaeal lipids. This intermediate was essential for an in vitro assay, which revealed

one of the missing links in biosynthesis of archaeal lipids. A second contribution of

this chapter is a total synthesis of cycloarcheol and its ß-glucosyl analogue, which are

important taxonomic tools in Archaea.

Parts of this chapter have been published.

Jain, S.; Caforio, A.; Fodran, P.; Lolkema, J. S.; Minnaard, A. J.; Driessen, A. J. M.

Chemistry & Biology, 2014, 21, 1392.

Ferrer, C.; Fodran, P.; Barroso, S.; Gibson, R.; Hopmans, E. C.; Sinninghe Damsté,

J.; Schouten, S.; Minnaard, A. J. Org. Biomol. Chem. 2013, 11, 2482.

Chapter 7

144

7

Introduction

Archaea form the third domain of life, comprising up to 20% of the biomass

on Earth. Since 1977, when the domain Archaea was first described,1 their

evolutionary origin has been a topic of intense debate. Recently, Forterre2 presented

a hypothesis in which the Archaea and the Eukarya evolved from a common

ancestor. He further hypothesized, that the ancestors of the Archaea escaped from

their proto-eukaryotic predators by invading ecological niches with harsh

environmental conditions. Surroundings like hydrothermal vents,3 geysers3-4 with

temperatures over 120 °C, highly acidic (pH = 0)5 or alkaline (pH >10)6 springs, or

lakes with salt7 concentrations 10 times higher than sea water required a lot of

adaptation. One of the features that distinguishes Archaea from Bacteria and

Eukaryota is their cell envelope. This lacks a general cell wall polymer, and contains

membrane phospholipids, which differ from bacterial and eukaryotic lipids in three

aspects (Figure 1-I). First, in Archaea the hydrophobic part comprises of two

terpenoid (mostly phytanyl) chains, while this part is composed of two fatty acid

residues in bacterial and eukaryotic lipids.

Figure 1. ( I ) Comparison of bacterial and eukaryotic lipids with archaeal lipids;

( II ) common archaeal lipid backbones.

A Missing Link in Archaeal Lipid Biosynthesis; a Contribution from Organic Synthesis

145

Pag

e14

5

7

Second, in Archaea the phytanyl chains are linked to the glycerol via an ether bond,

while in Bacteria and Eukarya the fatty acids bind via an ester bond. Third, the

stereogenic centre in the glycerol moiety of archaeal phospholipids has the opposite

configuration compared to that in bacterial and eukaryotic lipids. Furthermore,

Archaea display a greater variability of the lipid backbones compared to Bacteria and

Eukarya. Structures like cycloarcheol (1) (Figure 1-II) and caldarcheol (2) are

common. Despite substantial investigations on archaeal membrane lipid

biosynthesis, several steps and their corresponding enzymes remain unknown.

The extremophilic nature of Archaea is often associated with their unique

lipids. Driessen et al.8 studied the stability of liposomes from lipid extracts of

Escherichia coli (mesophilic Bacteria), Bacillus stearothermophilus (thermophilic Bacteria)

and Sulfolobus acidocaldarius (thermophilic Archaea). The Archaea-derived liposomes

(archeosomes) showed significantly higher stability at all studied temperatures. In the

same study, the authors also reported that while the bacterial liposomes gradually

released about 50% of their content (fluorescent dye) over 62 days, the archeosomes

showed only 8-10% release over the same period of time. The higher stability of

archeosomes already found application in bioelectronics9, gene delivery10 and

vaccination.11 A bottleneck for their wider application is their limited availability.

Growing Archaea is technically more difficult than growing Bacteria or eukaryotic

cells. Furthermore, the yields of lipids are low. Typically, 1 g of lyophilized archaeal

cells affords only 0.11-54 mg of crude lipid extracts.12

Biosynthesis of archaeal membrane lipids

Complete genome sequencing13 of Archaeoglobus fulgidus allowed a better

understanding of the lipid metabolism in Archaea. Archaea species have a complete

set of genes encoding fatty acid metabolism,14 similar to the bacterial and eukaryotic

metabolism. Nevertheless, the fatty acids in Archaea are not used for the synthesis

of membrane lipids, but for posttranslational modification of proteins. As mentioned

above, the lipophilic portion of archaeal membrane lipids is exclusively terpenoid-

based. These terpenes are biosynthesized15 in a pathway that is similar to the bacterial

and eukaryotic pathways (Figure 2). Diphosphate 3 (coming from the mevalonic acid

pathway) is isomerised to diphosphate 4. A three- or fourfold extension of 4 affords

geranylgeranyl diphosphate (5) or farnesylgeranyl diphosphate (6), which

subsequently enters the lipid biosynthetic pathway.

Chapter 7

146

7

Figure 2. Biosynthesis of terpenes in Archaea.

The biosynthesis of the phospholipids starts with glycerol monophosphate 8

(Figure 3), which is obtained by the reduction of 7 (a product of glycolysis). Given

that the absolute configuration of glycerol phosphate (8) is opposite in Bacteria and

Eukarya, the corresponding reductase was considered unique for Archaea. Babinger

et al.16 recently showed that Bacillus subtilis produces a homologous enzyme. The

biosynthesis of the phospholipids continues with the attachment of the terpenoid

(here geranylgeranyl) chains to 8 (Figure 3). First, the primary hydroxyl group is

prenylated in the cytosol. The significantly more lipophilic 9 is transferred to the cell

membrane where the second prenylation takes place. Both prenylations proceed via

a same SN1 type mechanism. In the active site of the enzyme, the diphosphate of 5

is cleaved with the assistance of Mg2+. Subsequently, the resulting allylic carbocation

reacts with a nucleophilic hydroxyl group of the glycerol. With the bis-prenylated

glycerol 10, the biosynthesis of the archaeal phospholipids continues by attachment

of the polar head group, which is achieved in two steps (Figure 3). The first step is

an activation of 10 with cytidine triphosphate (CTP). Although an analogous step

takes place in all three domains of life, the corresponding archaeal enzyme has been

elusive until now. The identification of this enzyme is the topic of the first part of

this chapter. The second step of the attachment of the phosphorous headgroup is

the conversion of 11 to the final phospholipid 12, 13 or 14. With the headgroup

attached, 12, 13 or 14 need to undergo 1 or 2 more transformations, depending on

the species. First, the reduction of the double bonds – a transformation that is

common in all Archaea species. The reduction was studied by Nishimira and Eguchi,17

who purified and characterized the corresponding enzyme in Thermoplasma

acidophilum. This enzyme did not show any preference for a headgroup –

phospholipids 12, 13 and 14 were reduced in a similar rate.


147

Pag

e14

7

7

Figure 3. Biosynthesis of phospholipids in Archaea.

While for some of the Archaea, the biosynthesis ends with the reduction of the

prenyl chains (from 15 to 16 in Figure 4-I), some other members further modify the

chain by dimerization (17 in Figure 4-I), cyclization (18 in Figure 4-I), or dimerization

followed by cyclization (19 in Figure 4-I). The exact mechanism of these

transformations remains unknown. A study by Eguchi et al.18 suggests that the

terminal double bonds are crucial for the dimerization. These findings are however

Chapter 7

148

7

in contrast to the findings of Nemoto et al.,19 who reported that the dimerization

takes place at fully saturated precursors. Fitz and Arigoni20 studied an analogous

dimerization in Butyrvibrio fibrisolvens (a genus of Bacteria), which produces a

membrane spanning diabolic acid (21) (Figure 4-II).

Figure 4. ( I ) Final steps in the biosynthetic pathway of membrane-spanning Archaea-

lipids; ( II ) dimerization of palmitic acid in Butyrivibrio fibrisolvens.


149

Pag

e14

9

7

The authors showed that diabolic acid is synthetized by dimerization of a fully

saturated 20.

Archaeal lipids as taxanomic markes

The second part of this chapter is dedicated to the total synthesis of the

cycloarcheol lipid core 1 and its glycolipid analogue 22. The lipids in Archaea fulfil

also an important taxonomic function which highlights the necessity of their

unambiguous structural determination. A total synthesis of the lipid is one of the

options. Further comparison of the HPLC chromatograms and mass spectra of the

synthetic and natural samples gives a high confidence in determining its presence. A

total synthesis of cycloarcheol 1 (Figure 1) and its ß-glucosyl analogue 22 (Figure 5)

is interesting in this context. Cycloarcheol was detected for the first time in 198321 in

a deep sea hydrothermal vent.

Figure 5. ß-glucosyl analogue of 1.


Synthesis of 2,3-bis-O-(geranylgeranyl)-sn-glycero-1-phosphate

Retrosynthetic analysis 10 (Figure 6) suggests its preparation by

phosphorylation of 2,3-bis-O-geranylgeranyl-sn-glycerol (23). 23 can be synthetized

from a suitably protected glycerol derivative 24 and geranylgeranyl halide 25 via a

Williamson ether synthesis. Another alternative would be ring-opening of protected

enantiopure glycidol 26 (more details in chapter 3) with geranylgeraniol (27),

followed by etherification of the formed secondary alcohol. Given that allyl halides

are excellent partners in Williamson’s reaction, this is the method of choice for the

construction of the unsaturated derivatives. This strategy was already recognized by

Morii, Nishihara and Koga.22In their synthesis, the authors used geranylgeranyl

bromide and enantiopure benzylglycerol. The benzyl group was subsequently

removed by Na/NH3 (liq). Phosphorylation with dimethyl chlorophosphate in basic

conditions and subsequent demethylation with TMSBr afforded 10 in <6% overall

yield. The authors explained their low overall yield by instability of the intermediate

Chapter 7

150

7

compounds. In an unrelated publication, Dannenmuller et al.23 reported the synthesis

and properties of archaeal membrane phospholipids analogues of 10.


The authors prepared bisgeranylgeranyl glycerol 23 via Wiliamson reaction of

geranylgeranyl chloride and dimethoxybenzyl protected glycerol. Application of this

protecting group is advantageous compared to the benzyl group because it can be

removed using mild oxidative conditions. The reported conditions were applied to

the synthesis with some minor modifications.

The glycerol derivative 29 was prepared in 2 steps from commercially

available (R)-solketal.24 The geranylgeranyl chloride as etherification partner was

prepared by treatment of geranylgeraniol with N-chlorosuccinimide and Me2S.25

Reaction of both 29 and geranylgeranyl chloride (Scheme 1) in the presence of dimsyl

sodium (sodium methylsulfinylmethylide) in DMSO afforded the desired diallylated

30 in 61% yield, together with monoallylated 31 in 16% yield. The yield of 30 is in

perfect agreement to that reported by Dannenmuller et al.23 (60%). Deprotection of

30 (Scheme 1) with DDQ in CH2Cl2/H2O (40/1) afforded 23 in 60% isolated yield,

again in a very good agreement with the literature (60%).23


151

Pag

e15

1

7

Reagents and conditions: a) NaDMSO (2.1 equiv), DMSO, then geranylgeranyl chloride (2.0 equiv.), RT, 16 h; b) DDQ (2.0 equiv), CH2Cl2, H2O, 0 °C

Scheme 1. Synthesis of bisgeranylgeranyl glycerol 13.

The phosphorylation (Scheme 2) of 23 turned out to be a challenging step.

Phosphorylation of 23 with POCl3 (Scheme 2) and subsequent hydrolysis in the

presence of AgNO326 resulted only in decomposition of the starting material. The

procedure reported by Morii, Nishihara and Koga22 (Scheme 2) afforded

dimethylphosphate 32, but all attempted demethylations resulted in its

decomposition. Phosphoramidites could be another viable option. First explored

reagent 33 (scheme 2) underwent phosphoramidite coupling with 23 in the presence

of tetrazole, and subsequent in situ oxidation with a solution of tBuOOH in decane

afforded bisprotected 34 in 54% yield. However, all the explored deprotection

methods of 34 resulted only in monodeprotected 35. Next, phosphoramidite 36 was

explored. The pKa value of the corresponding phosphate suggests a greater base-

lability.27 Synthesis of 36 was straightforward, but all attempts to purify the reagent

resulted in its decomposition. Finally, a reaction of 23 with an excess of crude 36 in

the presence of tetrazole and subsequent oxidation with tBuOOH, afforded

bisprotected 37 in 94% yield. Reaction of 37 with excess Et3N resulted again in

monodeprotected 38. Treatment of 38 with aqueous NaOH (1 M) resulted in the

removal of the second fluorenylmethyl group, affording 10 in 48% yield. After

further optimization, both protecting groups could be cleaved in 1 reaction. Stirring

37 in a 1 M aqueous NaOH in dioxane mixture followed by acidification and column

chromatography on 130 Å Davisil silica gel afforded 10 in 68% yield. Overall, 10 was

prepared in 4 steps and 23% overall yield, starting from 29.

Chapter 7

152

7

Reagents and conditions: a) OP(OMe)2Cl, Et3N CH2Cl2, 21 °C, 2h; b) 33, tetrazole 4h, 21 °C then tBuOOH (2.0 equiv), -10 °C, 15 min; c) 36 (3.0 equiv), tetrazole (3.0 equiv), 24 h, 21 °C, then tBuOOH (4.0 equiv) -10 °C, 1 h; d) Et3N (20 equiv), 21 °C, 18 h; e) 1 M aqueous NaOH; f) dioxane/1 M aqueous NaOH.

Scheme 2. Explored phosphorylations methods of 23.


153

Pag

e15

3

7

Identification of CDP-archaeol synthase

This paragraph summarizes the experiments of Dr. Samta Jain, and Dr.

Antonella Caforio, from the department of Molecular Microbiology of the

Rijksuniversiteit Groningen.

Based on the analogy between the biosynthesis of CDP-activated precursor

11 (Figure 3) in Archaea and Bacteria, bioinformatic analysis could identify a putative

CDP-archaeol synthase in Archaea. The sequence of bacterial phosphatidate

cytidylyltransferase28 (CDP- diacylglycerol synthase) served as input for an NCBI-

BLAST analysis. This resulted in a list of hypothetical proteins. Their sequences were

aligned to an averaged hydropathy (hydrophobicity) profile. The alignment revealed

common structural features of the hypothetical proteins – an extracellular N-

terminus and 5 transmembrane helices. Although the bacterial enzymes are longer

than the archaeal ones, the alignment of the family averaged hydropathy profile of

the two showed a common pattern at the C-terminal region. Furthermore, analysis

of the sequence of one of the protein loops revealed a consensus sequence between

the archaeal and bacterial enzyme. This putative enzyme could be the CDP-archaeol

synthase. The corresponding amino acid sequence was codon optimized for

expression in E. coli and the C-terminus of the protein was equipped with an

octahistidin tag. The enzyme was isolated after affinity chromatography. The

predicted function of the enzyme was confirmed in two assays. Synthetic 10, the

natural substrate for the enzyme, was incubated in the presence of Mg2+ salts and

cytidine triphosphate. LC/MS analysis of the reaction mixture confirmed the

presence of CDP-archaeol 11. When 10 was incubated with 2′-deoxycytidine 5′-

triphosphate under identical conditions, LC/MS analysis confirmed the presence of

deoxy-CDP-archaeol. Application of the other nucleosides did not lead to the

corresponding products. In the second assay, 10 was incubated with a radiolabelled

cytidine triphosphate ([5-T]CTP) under the same conditions. TLC analysis of the

reaction mixture showed only a single radioactive spot.

With the identified enzyme CDP-archaeol synthase, the archaeal lipid

biosynthesis could be reconstituted in vitro (Figure 7). After combining isopentenyl

diphosphate (3), dihydroxyacetone phosphate (7), farnesyl diphosphate 38, five

enzymes catalyzing the steps of the biosynthesis and NADH, LC/MS analysis

confirmed the presence of 11.

Chapter 7

154

7

Figure 7. In vitro reconstitution of the biosynthesis of 11.

Catalytic alcoholysis of benzylglycidol as a key step in the synthesis of cyclo-

archaeol and -glucosyl-cyclo-archaeol

The following part of the chapter summarizes research performed together

with Dr. Catalina Ferrer and Dr. Santiago Barroso.

Although bisgeranylgeranyl glycerol 10 and cycloarchaeol 1 (as their

corresponding phosphates) are part of the same biosynthetic route, the synthetic

challenges in 1 are considerably larger. Enzymatic reduction of the double bonds

introduces 8 new stereogenic centers, making the synthesis of the hydrocarbon chain

a challenge. A second, frequently underestimated hurdle is the construction of the

ether bonds. While in the case of reactive, unsaturated allylic derivatives (as in the

case of 10) the Wiliamson synthesis is straightforward, in the case of the saturated

alkylsulphonates or alkyl iodides, the competing elimination is a problem frequently

resulting in low yields of the etherification. At least a partial solution can be

alcoholysis of an enantiopure glycidyl ether catalysed by Jacobsen’s catalyst.

A two-fold conjugate addition (Scheme 3) on cyclo-octadienone (39),

followed by ozonolysis and esterification, afforded hydroxyl ester 40 with two

methyl-branched stereogenic centres. One portion of 40 was converted in 3 steps to

protected tetrazole 41, the second portion of ester 40 was oxidized to aldehyde 42.

41 and 42 were coupled in a Julia-Kocienski reaction. Hydrogenation using in situ

generated diimide using the aforementioned flavine catalyst, afforded alcohol 43. A

part of 43 was converted to iodide 44. The first ether bond was constructed (Scheme

4) by alcoholysis of (R)-benzylglycidol 45 with alcohol 43 using 8.0 mol% of the


155

Pag

e15

5

7

Co[R,R-(salen)]OTs, affording the desired ring-opened product 46 in 87% yield. The

second ether bond was constructed via Williamson reaction. After testing a series of

reaction conditions, best results were obtained when a mixture of 46 and the iodide

44 were treated with freshly ground KOH and catalytic nBu4NBr under solvent-free

conditions. In a slow reaction, this provided the desired product 47 with yields

varying from 35% to 55%. These values are in good agreement to the literature.29.

The contrast between the two applied etherification methods is noteworthy. While

the epoxide alcoholysis is a clean reaction with 87% yield, Wiliamson reaction affords

the ether in significantly lower 35 to 55% yield, together with side products comming

from the elimination.

Reagents and conditions: a) Me2Zn (3.0 equiv), Cu(OTf)2 (5.0 mol%), L-Phos (10 mol%), 39 added over 6 h, toluene, -25 °C, overnight; b) Me2Zn (1.5 equiv), Cu(OTf)2 (2.5 mol%), L-Phos (5.0 mol%), substrate added over 6 h, toluene, -25 °C, overnight then Et3N (3.5 equiv), TMSCl (5.0 equiv), 2 h c) crude TMS enol ether dissolved in MeOH, CH2Cl2, O3, -78 °C, then NaBH4; d) p-toluenesulfonic acid (5.0 mol%), MeOH, reflux, 24 h; e; TBDPSCl (1.6 equiv), 1H-imidazole (2.0 equiv), DMF, rt, 16 h; f) DIBAL (5.0 equiv), THF, -78 °C, 2 h; g) 1-phenyl-1H-tetrazole-5-thiol (2.0 equiv), PPh3 (1.5 equiv), DIAD (1.8 equiv), rt, overnight, then mCPBA (5.0 equiv), rt, overnight; h) TPAP (5.0 mol%), NMO (1.5 equiv), CH2Cl2, rt, overnight; i) LiHMDS (1.0 equiv), 41 (1.0 equiv), then 42 added, THF, -78 °C to rt, overnight; j) DIBAL (5.0 equiv), THF, -78 °C, 2 h; k) NH2NH2.H2O (20 equiv) added over 10 h, L-flav (2.0 equiv), EtOH, rt, 2 h; l) N,N,-dimethyl-N-(methansulfanylmethylene)ammonium iodide (1.5 equiv), 1H-imidazole (0.5 equiv), toluene, 85 °C, 16 h.

Scheme 3. Synthesis of methyl-branched precursors 43 and 44.

Chapter 7

156

7

However, the amounts of the building blocks were sufficient to complete the

synthesis of 1 (Scheme 4). The deprotection, oxidation and a Wittig reaction

sequence afforded bis-alkene 48, which was cyclized by ring closing metathesis. The

resulting double bond was reduced by hydrogenation over Pt/C catalyst because the

flavin generated diimide did not result in a full conversion.

Reagents and conditions: a) 43 (0.55 equiv), Co[R,R-(salen)]OTs (4.5 mol%), O2 (ballon), rt, 16 h, b) 44 (1.1 equiv), nBu4Br (0.5 equiv), KOH (2.7 equiv), 42 °C, 48 h; c) TBAF (4.0 equiv), THF, rt, overnight, d) Dess-Martin periodinane (2.5 equiv), CH2Cl2, rt, 1 h; e) Me3PPh3Br (4.5 equiv), KHMDS (4.2 equiv), THF, rt, 1 h; f) 2nd Grubbs catalyst (15 mol%), CH2Cl2 (0.002 M), reflux, 48 h; g) Pt/C (20 mol%), MeOH/CH2Cl2 (2/1), H2 (ballon), rt, 16 h; Pd/C (Degussa type E101 NE/W, 25 mol%), H2 (ballon), EtOAc, rt, 16h; i) 49 (3.5 equiv), AgOTf (3.5 equiv), tetramethylurea (4.5 equiv), toluene/CH2Cl2 (1/1), 0 °C; j NaOMe (30 equiv), MeOH, rt.

Scheme 4. Final steps of synthesis of 1 and 22.


157

Pag

e15

7

7

Final debenzylation afforded cycloarcheol 1. The Koenigs–Knorr glycosylation

followed by deprotection of the hydroxyl groups afforded the desired ß-glucosyl

derivative 22.

Detection of 1 and 22 in the deep sea samples

Both compounds were used to confirm their presence in hydrothermal vents.

The analyzed sample was collected from the Rainbow hydrothermal vent (36°14′N)

field located on the Mid-Atlantic Ridge. Samples were collected during a sampling

campaign in 2008 using the remotely operated vehicle Jason. The sample is

composed of material from the interior of a vent chimney collected at a depth of

2293 meters below the sea level. Analysis by GC-MS (in the case of 1) and

HPLC/ESI/MS (in the case of 22) showed that synthetic and natural compounds

co-eluted and that their mass spectra were identical. This suggest the presence of

methanogenic Archaea in the Rainbow.

Conclusion

The chemical synthesis of unsaturated archaeatidic acid has been important

in the identification of CDP-archaeol synthase, one of the missing links in the

biosynthesis of archaeal membrane lipids. The synthetically challenging step was the

phosphorylation of bisgeranylgeranyl-glycerol. This was achieved by the application

of bisfluorenylmethyl substituted phosphoramidite, in situ oxidation, and subsequent

deprotection under basic conditions.

In the second part of this chapter, a key step in the synthesis of cyclo-archaeol

is described. As in chapter 3, the catalytic regioselective ring opening of a protected

glycidol is successfully applied as an alternative for a Williamson ether synthesis with

a glycerol derivative. A versatile method for the subsequent alkylation of the

secondary hydroxyl group is still lacking, but the currently applied procedure is

acceptable. The synthesis of cycloarchaeol and ß-glucosyl cycloarchaeol allowed to

unambiguously establish their presence in a sample taken from a hydrothermal vent

field.

Experimental part

(S)-4-(((3,4-dimethoxybenzyl)oxy)methyl)-2,2-dimethyl-1,3-dioxolane

Chapter 7

158

7

A dry three-necked round-bottom flask equipped with a reflux condenser

was charged with (R)-1,2-isopropylidene glycerol (744 mg, 4.0 mmol) and nBu4Itetra-n-butylammonium iodide (147 mg, 0.4 mmol, 10 mol%,).

Solids were degassed in three cycles before dry THF (12 ml) was added.

To the obtained solution, KH (50% in paraffin, 370 mg, 4.6 mmol, 1.15 equiv) was

added in small portions. The mixture was stirred for 10 min before 4-(chloromethyl)-

1,2-dimethoxybenzene30 (860 mg, 4.6 mmol, 1.15 equiv) was added in one portion.

The so-obtained reaction mixture was immersed into a preheated oil bath (87 °C)

and refluxed for 16 h. After removal from the oil bath and cooling down to rt, solid

NH4Cl (1 g) was added. The mixture was stirred for 15 min, filtered and the collected

filtrate was evaporated to dryness. The yellow liquid residue was further purified by

flash chromatography using 50% Et2O in pentane. Fractions with an Rf = 0.37 (50%

Et2O in pentane) were collected and concentrated to afford 1.04 g of the desired

compound as colourless thick liquid (92%).

1H NMR (400 MHz, CDCl3, δ): 6.82 (m, 3H), 4.49 (m, 2H), 4.27 (m, 1H), 4.03 (dd,

J = 8.2, 6.5 Hz, 1H), 3.71 (dd, J = 8.2, 6.4 Hz, 1H), 3.51 (dd, J = 9.8, 5.8 Hz, 1H),

3.44 (dt, J = 12.3, 4.7 Hz, 1H), 1.40 (s, 3H), 1.34 (s, 3H).

13C NMR (100 MHz, CDCl3, δ): 149.20, 148.83, 130.66 ( - ), 120.51, 111.21, 111.02

( - ), 109.54 ( - ), 74.93 ( - ), 73.56, 70.96, 67.00, 56.06 ( - ), 55.99 ( - ), 26.95 ( - ), 25.54

( - ).

αD =+15.9 (c = 0.067, CHCl3).

Anal. Calcd for C15H22O5: C, 63.81; H, 7.85. Found: C, 63.51; H, 7.88%.

The spectroscopic data correspond to previously published24

(R)-3-((3,4-dimethoxybenzyl)oxy)propane-1,2-diol

The corresponding acetonide (950 mg, 3.4 mmol) was dissolved in

CH2Cl2/MeOH (10 ml/10 ml). Amberlite IR120 (acid form, 100 mg)

was added, and the mixture was stirred at RT until full conversion (36

h). The catalyst was filtered off and the filtrate was concentrated in vacuo.

The reaction afforded 815 mg of desired product (>99%) as colourless very thick

liquid.


159

Pag

e15

9

7

1H NMR (400 MHz, CDCl3, δ): 6.85 (m, 3H), 4.47 (s, 2H), 3.87 (m, 7H), 3.69 (dd, J

= 11.4, 3.9 Hz, 1H), 3.61 (dd, J = 11.4, 5.5 Hz, 1H), 3.57 – 3.48 (m, 2H), 2.31 (s,

2H).

13C NMR (100 MHz, CDCl3, δ): 149.02, 148.76, 130.18, 120.47, 111.16 ( - ), 110.94

( − ), 73.45, 71.44, 70.71 ( - ), 64.02, 55.89 ( - ), 55.86 ( - ).

The spectral data corresponds to previously reported.24

Geranylgeraniol prepared following the literature procedure

Farnesyl bromide

A dry flask was charged with farnesol (2.0 g, 8.9 mmol). Dry THF (30 ml)

was added, and resulting solution was immersed into a -47 °C bath (ethanol,

cryostat). After stirring for 10 min, freshly distilled MsCl (900 µl, 12 mmol, 1.3 equiv)

was added via syringe over 5 min. Subsequently, Et3N (2.5 ml, 18 mmol, 2.0 equiv)

was added over another 5 min. After complete addition, the mixture was stirred for

45 min at –47 °C. To the resulting suspension, a solution of LiBr (3.0 g, 36 mmol,

4.0 equiv) in dry THF (10 ml) was added dropwise over 5 min. After complete

addition, the reaction vessel was transferred to a 0 °C bath (ice/water) and stirred

for 1 h. The reaction mixture was poured into chilled saturated NaHCO3 solution.

The organic layer was separated, the aqueous layer was extracted with cold Et2O (a

mixture of Et2O with pieces of ice, 3 x 25 ml), the combined organic layers were

washed with cold water, brine, dried over MgSO4 and evaporated. The crude farnesyl

bromide was obtained as a yellow liquid and used without further manipulation.

The reaction afforded 1.91 g of the desired product as a light yellow oil (75% yield)

which was stored at -80 °C, and used within 1 day.

(6E,10E)-ethyl 7,11,15-trimethyl-3-oxohexadeca-6,10,14-trienoate

A dry Schlenk flask was charged with NaH (60%

dispersion, 880 mg, 22 mmol, 3.3 equiv). The

mineral oil was removed by three washings with pentane (3 x 10 ml). The resulting

white solid was dried in vacuum. Dry THF (16 ml) was added and the resulting white

suspension was cooled to 0 °C (ice/water bath). To this suspension, freshly distilled

ethyl acetoacetate (2.6 ml, 3.0 equiv) was added dropwise over 5 min. After complete

addition the suspension turned into a light yellow solution. To this solution was

added a solution of nBuLi in hexanes (2.5 M, 8.3 ml, 3.1 equiv) over 15 min. The

resulting orange solution was stirred for additional 15 min at 0 °C before a solution

Chapter 7

160

7

of farnesyl bromide (from the previous experiment 1.91 g, 6.7 mmol) in dry THF

(3.5 ml) was added dropwise over 5 min. The resulting solution was stirred at 0 °C

for 15 min, during which formation of a precipitate was observed. The reaction was

quenched by careful addition of chilled aqueous HCl (1 M, 10 ml,

EXOTHERMIC). The mixture was transferred into a separatory funnel, where the

organic layer was separated, the aqueous layer was extracted with Et2O (3 x 10 ml),

the combined organic layers were washed with brine, dried and evaporated.

The title compound was obtained after column chromatography using 10% Et2O in

pentane as 1.67 g of a light yellow liquid (55% starting from farnesol, lit 70%).

1H NMR (400 MHz, CDCl3) δ 12.09 (s, 0.2H), 5.08 (d, J = 6.5 Hz, 3H), 4.19 (dt, J

= 7.2, 5.3 Hz, 2H), 3.42 (s, 2H), 2.56 (t, J = 7.4 Hz, 2H), 2.38 – 2.17 (m, 3H), 2.05 -

1.98 (m, 8H), 1.68 (s, 3H), 1.64 – 1.58 (m, 9H), 1.28 (t, J = 7.1 Hz, 3H).

The spectral data corresponds to previously reported.31

Ethyl (2Z,6E,10E)-3-((diethoxyphosphoryl)oxy)-7,11,15-trimethylhexadeca-

2,6,10,14-tetraenoate

A dry Schlenk flask was charged with NaH (60%

dispersion, 228 mg, 1.15 equiv). The mineral oil was

removed by three successive washings with hexane (3 x 5 ml) and the resulting white

solid was suspended in dry Et2O (21 ml). The suspension was immersed in an

ice/water bath of 0 °C and a solution of (6E,10E)-ethyl 7,11,15-trimethyl-3-

oxohexadeca-6,10,14-trienoate (1.65 g, 5.0 mmol) was added as a solution in dry

Et2O (7 ml) over 15 min. After the addition was complete, the resulting light yellow

solution was stirred for 15 min at 0 °C and for 15 min at 21 RT. Then the solution

was again cooled in the ice bath and neat (EtO)2P(O)Cl (1.1 ml, 7.5 mmol, 1.5 equiv)

was added dropwise. The resulting reaction mixture was stirred for 15 min at 0 °C

and subsequently quenched by addition of saturated aqueous NH4Cl solution (15

ml). The organic layer was separated, and the aqueous layer was extracted with Et2O

(3 x 15 ml). The combined organic layers were washed with saturated aqueous

NaHCO3 (3 x 15 ml), brine (2 x 15 ml), dried over MgSO4 and the solvent was

removed in vacuo.

The resulting crude (2.08 g) was obtained as a yellow liquid and used without further

purification in the following step.

ethyl (2E,6E,10E)-3,7,11,15-tetramethylhexadeca-2,6,10,14-tetraenoate


161

Pag

e16

1

7

A dry Schlenk flask was charged with CuI (1.7 g,

8.9 mmol, 1.8 equiv) which was suspended in dry Et2O

(5.5 ml) and cooled to 0 °C. The resulting suspension was treated with MeLi (1.6 M

in Et2O, 11.0 ml, 18 mmol, 3.6 equiv). The suspension turned yellow after initial

addition of MeLi, after complete addition the CuI fully dissolved affording a nearly

colourless solution.

The reaction vessel with Me2CuLi was immersed in a cryostat at –78°C32 and

a solution of the phosphate (2.08 g, ca 5.0 mmol) in Et2O (dry, 7 ml) was added

dropwise via the cold wall of the Schlenk flask (in order to cool the solution of the

phosphate). After complete addition, the colour changed to orange/red and the

resulting solution was stirred at –78 °C. After 1 h the bath was allowed to warm to –

47 °C and the reaction mixture was stirred at –47 °C for 2 h. After this time TLC

showed full conversion of the phosphate and a new spot had appeared on TLC. MeI

(630 µl) was added to quench the unreacted cuprate. After stirring for 10 min, the

reaction mixture was carefully poured into a solution of NH4Cl (24 ml) and NH4OH

(6 ml) (can be exothermic with gas evolution). The mixture was stirred until all

solids dissolved. Layers were separated, the aqueous layer was extracted with Et2O

(3 x 20 ml) and the combined organic layers were washed with NH4OH (10%, 2 x

40 ml), brine (2 x 40 ml), dried and evaporated.

The reaction afforded 1.15 g of a yellow liquid which was used without further

purification.

Geranylgeraniol

The ethyl ester from the previous step (1.15 g, 4.5

mmol) was dissolved in toluene (p.a. grade, 17 ml).

This solution was cooled to –78 °C (N2/acetone bath) and a solution of DIBAL (1M

in hexane, 14.0 ml, 14 mmol, 3.0 equiv) in hexane (15 ml) was added dropwise. The

mixture was stirred until complete consumption of the starting material (TLC). The

reaction was quenched by careful addition of MeOH (3.0 ml, added over 10 min,

EXOTHERMIC, gas evolution). When gas evolution ceased, the mixture was

removed from the bath and stirred for 10 min at rt. The reaction mixture was poured

into saturated NH4Cl (50 ml)/HCl (50 ml) solution and stirred until clear separation

of the layers took place (ca 30 min). The aqueous layer was extracted with Et2O (3 x

50 ml). The combined organic layers were washed with water (2 x 50 ml) and brine

(2 x 50 ml), dried over MgSO4 and evaporated. The residual thick liquid was further

purified by flash chromatography on silica using 30% Et2O in pentane as the eluent.

The reaction afforded 857 mg of geranylgeraniol (60% over three steps) with >99%

double bond isomer purity according to GC analysis.

Chapter 7

162

7

The spectral data corresponds to previously reported31

Geranylgeranyl chloride

N-chlorosuccinimide (1.2 g, 9.0 mmol, 1.3 equiv) was

suspended in dry CH2Cl2 (15 ml). The suspension was

cooled to –30 °C (acetone/liquid N2 bath) and after stirring for 5 min, dimethyl

sulfide (750 µl, 10 mmol, 1.5 equiv) was added. The reaction was stirred for 10 min

at –30 °C and 10 min at 0 °C. Then the solution was cooled to –40 °C. A solution

of geranylgeraniol (2.0 g, 6.9 mmol) in CH2Cl2 (dry, 5 ml) was added dropwise. The

resulting suspension was allowed to warm during 150 min to 0 °C, turning into a

cloudy solution at –15 °C. The reaction mixture was poured into pentane (150 ml),

The organic layer was washed with water (2 x 50 ml), brine (50 ml), dried over MgSO4

and evaporated.

The reaction afforded 1.94 g of geranylgeranyl chloride as a colourless liquid which

was used without further purification.

Synthesis of 2,3-bisgeranylgeranyl-sn-glycerol (23), (scheme 1)

1-((3,4-dimethoxybenzyl)oxy)- 2,3-bisgeranylgeranyl-sn-glycerol (30)

A dry Schlenk flask was charged

with NaH (60% dispersion in mineral oil, 265 mg, 6.6 mmol, 2.1 equiv). The mineral

oil was removed by washing with pentane (3 x 5 ml) and the white solid was dried in

high vacuum before suspending in DMSO (5 ml). The obtained suspension was

immersed in a preheated oil bath (70 °C) and stirred for 40 min during which the

suspension turned into a pale yellow solution. The flask was removed from the bath

and allowed to cool to rt (21 °C). To this solution, a solution of (R)-3-((3,4-

dimethoxybenzyl)oxy)propane-1,2-diol (595 mg, 3.2 mmol) in dry DMSO (5 ml) was

added carefully. After the complete addition, the reaction mixture was stirred for 1

h at rt (21 °C). Then a mixture of geranylgeranyl chloride (1.94 g from the previous

experiment, ca 6.3 mmol) in a small amount of DMSO (2 ml) was added. The

resulting solution was stirred for 16 h before pouring into saturated aqueous NH4Cl

solution (20 ml). The aqueous layer was extracted with Et2O (3 x 50 ml). The

combined organic layers were washed with brine, dried and evaporated. The crude

residue was further chromatographed using 30% Et2O in pentane to afford the

desired product and the product of the mono-alkylation.


163

Pag

e16

3

7

Dialkylated product 1.52 g (61%) as colourless liquid

1H NMR (400 MHz, CDCl3, δ): 6.97 – 6.75 (m, 3H), 5.35 (dt, J = 13.5, 6.6 Hz, 2H),

5.10 (d, J = 5.8 Hz, 6H), 4.49 (s, 2H), 4.16 (d, J = 6.7 Hz, 2H), 4.01 (d, J = 6.7 Hz,

2H), 3.85 (s, 3H), 3.89 (s, 3H), 3.68 (dt, J = 10.0, 5.1 Hz, 1H), 3.62 – 3.46 (m, 5H),

2.16 – 1.90 (m, 23H), 1.68 (s, 6H), 1.65 (s, 26), 1.59 (s, 16H)

13C NMR (100 MHz, CDCl3, δ): 149.06, 148.60, 140.20, 139.95, 135.43, 135.40,

135.06, 131.39, 131.10, 124.51 ( - ), 124.32 ( - ), 124.31 ( - ), 124.03 ( - ), 123.99 ( - ),

121.33 ( - ), 120.99 ( - ), 120.31 ( - ), 111.07 ( - ), 110.91 ( - ), 77.03( - ), 73.40, 70.35,

70.19, 68.02, 66.95, 56.03 ( - ), 55.92 ( - ), 39.86, 39.84, 39.78, 26.89, 26.78, 26.53,

26.50, 25.85( - ), 17.83 ( - ), 16.69( - ) , 16.66 ( - ), 16.15 ( - ).

αD =+5.2 (c = 1.0, CHCl3)

NMR data correspond to those previously published23.

Monoalkylated product 256 mg

(16%) as colourless liquid.

1H NMR (400 MHz, CDCl3, δ): 6.84 (m, 3H), 5.34 (d, J = 7.1 Hz, 2H), 5.10 (s, 3H),

4.48 (d, J = 5.0 Hz, 2H), 4.01 (m, 3H), 3.85 (s, 3H), 3.89 (s, 3H), 3.76 – 3.40 (m, 5H),

2.05 (m, 12H), 1.66 (d, J = 8.3 Hz, 6H), 1.59 (s, 9H).

13C NMR (100 MHz, CDCl3, δ): 149.11, 148.76, 148.73, 140.72, 140.67, 135.49,

135.45, 135.05, 131.36, 130.62, 124.47 ( - ), 124.26 ( - ), 123.88 ( - ), 123.85 ( - ),

120.79 ( - ), 120.59 ( - ), 120.48 ( - ), 120.38 ( - ), 111.16 ( - ), 111.02 ( - ), 110.95 ( - ),

110.93 ( - ), 77.61 ( - ), 73.53, 73.48, 71.34, 71.22, 69.93, 69.70 ( - ), 67.94, 66.67,

63.01, 56.01 ( - ), 55.93 ( - ), 55.92 ( - ), 39.83, 39.80, 39.72, 26.86, 26.72, 26.43, 25.82

( - ), 17.80 ( - ), 16.66 ( - ), 16.63 ( - ), 16.13 ( - ) 16.12 ( - ).

αD =+9.2 (c = 1.0, CHCl3).

HRMS-APCI (m/z): [M + Na]+ calculated for C32H50O5Na, 538.354; found, 538.355.

Chapter 7

164

7

2,3-bisgeranylgeranyl-sn-glycerol (23)

30 (1.44 g, 1.8 mmol) was dissolved in

CH2Cl2 (30 ml). To this solution, H2O (0.75 ml) was added. The obtained biphasic

solution was cooled in an ice bath (0 °C) and DDQ (830 mg, 3.6 mmol, 2.0 equiv)

was added. The reaction was stirred for 4 h at 0 °C until TLC showed full conversion

of the starting material. The crude reaction mixture was filtered over a small silica

pad and washed with CH2Cl2 (200 ml). The washings were combined and evaporated.

The obtained yellow liquid was further purified by column chromatography (30%

Et2O in pentane).

The reaction afforded 705 mg of the desired product (60%) as yellow thick liquid

containing traces of unidentified co-eluting impurities.

1H NMR (400 MHz, CDCl3, δ): 5.35 (dt, J = 13.5, 6.7 Hz, 2H), 5.11 (t, J = 6.5 Hz,

6H), 4.26 – 4.06 (m, 2H), 4.02 (d, J = 6.7 Hz, 2H), 3.77 – 3.41 (m, 5H), 2.16 – 1.93

(m, 24H), 1.68 (s, 12H), 1.60 (s, 21H)

13C NMR (100 MHz, CDCl3, δ): 140.68, 135.49, 135.47, 135.07, 131.37, 124.52 ( - ),

124.32 ( - ), 123.9 ( - )3, 120.91 ( - ), 120.69 ( - ), 77.61 ( - ), 70.16, 68.08, 66.66, 63.21,

39.88, 39.86, 39.83, 39.75, 26.90, 26.77, 26.50, 26.46, 25.83 ( - ), 17.82 ( - ), 16.68 ( - ),

16.65 ( - ), 16.15 ( - ), 16.14 ( - ).

HRMS-APCI+ (m/z): [M + Na]+ calculated for C43H72O3Na, 659.536; found,

659.537.

Synthesis of 2,3-bis-O-(geranylgeranyl)-sn-glycero-phosphate (10) (Scheme 2)

bis((9H-fluoren-9-yl)methyl) diisopropylphosphoramidite

PCl3 (7 ml, 80 mmol) was dissolved in pentane (700 ml).

Via an addition funnel, a solution of diisopropylamine

(distilled from CaH2, 23 ml, 0.16 mol, 2.0 equiv) in

pentane (100 ml) was added dropwise over 15 min.

After this time, a significant amount of white precipitate had formed. The suspension

was stirred for 2 h. The mixture was transferred to a separatory funnel. The pentane

layer was washed with acetonitrile (the pentane layer stays on top, 5 x 100 ml of

acetonitrile, after the washing the pentane layer was fully transparent). Pentane was

subsequently evaporated. The reaction afforded 1,1-dichloro-N,N-

diisopropylphosphinamine (6 g, 38%) as colourless liquid


165

Pag

e16

5

7

1H NMR (400 MHz, CDCl3, δ): 4.02 – 3.83 (m, 1H), 1.28 (d, J = 6.8 Hz, 6H),

31P NMR (400 MHz, CDCl3, δ): 169.59.

From the obtained liquid, an aliquot was taken (2.0 g, 10 mmol). This was dissolved

in dry THF (40 ml) and DIPEA (3.5 ml, 20 mmol, 2.0 equiv) was added. The solution

was cooled in an ice bath and (9H-fluoren-9-yl)methanol (3.92 g, 20 mmol, 2.0 equiv)

in dry THF (10 ml) was added. The solution was stirred for 10 h during which a

white precipitate formed. The reaction was poured into aqueous phosphate buffer

(1 M, pH = 7) and extracted with ethyl acetate (4 x 50 ml). The combined organic

extracts were washed with the same phosphate buffer, brine, dried and evaporated.

The crude product was used without further purification due to its sensitivity.

31P NMR (162 MHz, CDCl3) δ 148.01.

bis((9H-fluoren-9-yl)methyl) ((S)-2,3-bisgeranylgeranyl)oxy)propyl) phosphate 37

(R)-2,3-bisgeranylgeranyl glycerol

(66.0 mg, 0.1 mmol)was dissolved

in CH2Cl2/CH3CN (0.5 ml/0.5

ml) and 36 (154 mg, 0.3 mmol,

3.0 equiv) was added. The mixture was cooled to 0 °C and tetrazole (21 mg, 0.3

mmol, 3.0 equiv) was added. The mixture was allowed to gradually warm to rt (21

°C) and stirred overnight. When full conversion of starting material was observed

(TLC), the mixture was cooled and a solution of tBuOOH (5 M in decane, 64μl, 0.3

mmol, 3.2 equiv) was added in one portion followed by stirring for 45 min. The

mixture was subsequently poured into aqueous phosphate buffer (1 M, pH = 7) and

extracted with Et2O (4 x 20 ml). The combined extracts were washed with brine,

dried and concentrated. The crude residue was purified by flash chromatography

using (50% Et2O in pentane). Fractions with an Rf = 0.4 (50% Et2O in pentane)

were collected to afford 102 mg (94%) of the desired compound.

1H NMR (400 MHz, CDCl3, δ): 7.71 (dd, J = 12.2, 4.3 Hz, 4H), 7.56 (m, 4H), 7.38

(m, 4H), 7.22 (m, 4H), 5.27 (q, J = 6.8 Hz, 2H), 5.09 (m, 6H), 4.28 (m, 4H), 4.14 (m,

3H), 3.97 (m, 5H), 3.59 (dd, J = 9.7, 4.9 Hz, 1H), 3.54 (m, 2H), 2.03 (m, 24H), 1.68

(s, 6H), 1.59 (dd, J = 11.2, 6.0 Hz, 24H)

13C NMR (100 MHz, CDCl3, δ): 147.46, 147.40, 145.60, 144.69, 144.56, 139.59,

139.57, 139.21, 135.52 ( - ), 132.10 ( - ), 131.36 ( - ), 129.48 ( - ), 129.43 ( - ), 128.66

Chapter 7

166

7

( - ), 128.47 ( - ), 128.09 ( - ), 124.89 ( - ), 124.86 ( - ), 124.25 ( - ), 124.21 ( - ), 80.44

( - ), 80.36 ( - ), 73.59, 73.53, 73.07, 72.22, 71.45, 71.40, 71.07, 52.24 ( - ), 52.16 ( - ),

44.00, 43.98, 43.87, 31.04, 30.93, 30.70, 30.65, 29.98 ( - ), 21.97 ( - ), 20.78 ( - ), 20.29

( - ).

HRMS-APCI (m/z): [M + Na]+ calculated for C71H93O6PNa, 1095.660; found

1095.660.

2,3-bis-O-(geranylgeranyl)-sn-glycero-phosphate (10)

Bisprotected phosphoric ester 37 (102

mg, 0.10 mmol) was dissolved in acetonitrile (5.0 ml). To this solution, Et3N (280

μmol, 2.0 mmol, 20 equiv) was added and the resulting mixture was stirred overnight.

All volatiles were evaporated and the crude mono deprotected phosphoric ester (as

assumed from the TLC) was suspended in aqueous NaOH (1 M, 5.0 ml) until full

conversion of the monoprotected ester was observed (TLC, 3 h). The mixture was

acidified with HCl (1 M) to pH = 1 and extracted with Et2O (3 x 20 ml). The

combined extracts were dried over MgSO4 and concentrated. The crude residue was

purified on a silica column using a carefully established gradient of 2%

MeOH/CHCl3 33% MeOH/CHCl3.

Reaction afforded 34.3 mg (48%) of the desired compound as a colourless liquid.

One pot procedure for the deprotection:

Bisprotected phosphoric ester (102 mg, 0.10 mmol) was dissolved in dioxane (1 ml).

To the stirred solution aqueous NaOH soulution (1 M, 1 ml) was added. The mixture

was stirred until full conversion of the starting material (3 h). The mixture was

acidified with HCl (1 M) to pH = 1 and extracted with Et2O (3 x 20 ml). The

combined extracts were dried over MgSO4 and concentrated. The crude residue was

purified on a 120 Å Davisil silica column using a carefully established gradient of 2%

MeOH/CHCl3 33% MeOH/CHCl3.

The reaction afforded 48.6 mg (68%) of the desired compound as a colourless oil

NOTE: The final product was stored in the freezer (–20 °C).

1H NMR (400 MHz, CDCl3, δ): 5.32 (m 2H), 5.09 (m, 6H), 4.15 (m, 2H), 4.00 (m,

4H), 3.57 (m, 3H), 2.01 (m, 24H), 1.77 – 1.49 (m, 30H).


167

Pag

e16

7

7

13C NMR (100 MHz, CDCl3, δ): (101 MHz, CDCl3, the spectrum shows

a considerable number of overlapping signals) δ 135.23, 134.85, 131.16, 124.38,

124.22, 123.88, 120.71, 39.75, 39.72, 26.76, 26.61, 25.67, 17.66, 16.53, 16.00, 15.97.

31P NMR (162 MHz, CDCl3) δ 1.37.

HRMS-ESI+ (m/z): [M + H]+ calculated for C43H72O6P, 715.507; found, 715.506.

Total synthesis of cycloarcheol 1 (Scheme 4)

(4S,9R,13R,17S,21S)-9,13,17,21,25,25-hexamethyl-1,24,24-triphenyl-2,6,23-trioxa-

24-silahexacosan-4-ol (46)

To a roundbottom flask containing

neat 13 (488 mg, 0.9 mmol) was

added neat benzyl-(S)-glycidol (250 μl, 1.6 mmol, 1.65 equiv) and Co[R,R-

(salen)]OTs (35 mg, 8.0 mol%). An atmosphere of dry oxygen was applied (balloon,

1 bar). The mixture was stirred for 16 h and then purified by silica gel

chromatography using 20% Et2O in hexane. The reaction afforded 533 mg of the

desired product as colourless liquid (85%)

1H NMR (400 MHz, CDCl3, δ): 7.69 – 7.65 (m, 4H), 7.45 – 7.27 (m, 11H), 4.57 (s,

2H), 3.98 (br s, 1H), 3.59 – 3.40 (m, 8H), 2.47 (br d, J = 3.1 Hz, 1H), 1.70 – 1.47 (m,

4H), 1.46 – 1.14 (m, 20H), 1.06 (s, 9H), 0.92 (d, J = 6.7 Hz, 3H), 0.88 (d, J = 6.6 Hz,

3H), 0.85 (d, J = 5.9 Hz, 3H), 0.83 (d, J = 6.0 Hz, 3H).

13C NMR (100 MHz, CDCl3, δ): 138.0, 135.6, 134.1, 129.4, 128.4, 127.7, 127.7, 127.5,

73.4, 71.8, 71.4, 70.0, 69.5, 68.9, 37.5, 37.4, 37.4, 36.6, 35.7, 33.5, 32.8, 32.8, 29.9,

26.9, 24.5, 24.4, 24.4, 19.8, 19.7, 19.3, 17.0.

HRMS-APCI (m/z): [M + Na]+ calculated for C46H72O4SiNa, 739.509; found:

739.509.

[]D -0.3 (c = 1.2, CHCl3).


(1) Woese, C. R.; Fox, G. E. Proc. Nat. Acad. Sci. 1977, 74, 5088.

(2) Forterre, P. Archaea 2013, 2013, 18.

(3) Blöchl, E.; Rachel, R.; Burggraf, S.; Hafenbradl, D.; Jannasch, H. W.; Stetter, K. O.

Extremophiles 1997, 1, 14.

Chapter 7

168

7

(4) Barns, S. M.; Fundyga, R. E.; Jeffries, M. W.; Pace, N. R. Proc. Nat. Acad. Sci. 1994,

91, 1609.

(5) Schleper, C.; Puehler, G.; Holz, I.; Gambacorta, A.; Janekovic, D.; Santarius, U.;

Klenk, H. P.; Zillig, W. J. Bacteriol. 1995, 177, 7050.

(6) Duckworth, A. W.; Grant, W. D.; Jones, B. E.; van Steenbergen, R. FEMS Microbiol.

Ecol. 1996, 19, 181.

(7) Brisou, J.; Courtois, D.; Denis, F. App. Microbiol. 1974, 27, 819.

(8) Elferink, M. G. L.; de Wit, J. G.; Driessen, A. J. M.; Konings, W. N. Biochim. Biophys.

Acta - Biomembranes 1994, 1193, 247.

(9) De Rosa, M.; Morana, A.; Riccio, A.; Gambacorta, A.; Trincone, A.; Incani, O.

Biosens. Bioelectron. 1994, 9, 669.

(10) Benvegnu, T.; Lemiegre, L.; Cammas-Marion, S. Recent Pat. Drug Deliv. Formul. 2009,

3, 206.

(11) Sprott, G. D.; Yeung, A.; Dicaire, C. J.; Yu, S. H.; Whitfield, D. M. Archaea 2012,

513231.

(12) (a) Nishihara, M.; Koga, Y. J. Biochem. 1987, 101, 997(b) Hedrick, D. B.; Guckert, J.

B.; White, D. C. J. Lipid Res. 1991, 32, 659.

(13) Klenk, H.-P.; Clayton, R. A.; Tomb, J.-F.; White, O.; Nelson, K. E.; Ketchum, K.

A.; Dodson, R. J.; Gwinn, M.; Hickey, E. K.; Peterson, J. D.; Richardson, D. L.;

Kerlavage, A. R.; Graham, D. E.; Kyrpides, N. C.; Fleischmann, R. D.;

Quackenbush, J.; Lee, N. H.; Sutton, G. G.; Gill, S.; Kirkness, E. F.; Dougherty, B.

A.; McKenney, K.; Adams, M. D.; Loftus, B.; Peterson, S.; Reich, C. I.; McNeil, L.

K.; Badger, J. H.; Glodek, A.; Zhou, L.; Overbeek, R.; Gocayne, J. D.; Weidman, J.

F.; McDonald, L.; Utterback, T.; Cotton, M. D.; Spriggs, T.; Artiach, P.; Kaine, B.

P.; Sykes, S. M.; Sadow, P. W.; D'Andrea, K. P.; Bowman, C.; Fujii, C.; Garland, S.

A.; Mason, T. M.; Olsen, G. J.; Fraser, C. M.; Smith, H. O.; Woese, C. R.; Venter, J.

C. Nature 1997, 390, 364.

(14) Peretó, J.; López-García, P.; Moreira, D. Trends Biochem. Sci. 2004, 29, 469.

(15) Boucher, Y.; Kamekura, M.; Doolittle, W. F. Mol. Microbiol. 2004, 52, 515.

(16) Guldan, H.; Matysik, F.-M.; Bocola, M.; Sterner, R.; Babinger, P. Angew. Chem.

Internat. Ed. 2011, 50, 8188.

(17) Nishimura, Y.; Eguchi, T. J. Biochem. 2006, 139, 1073.

(18) Eguchi, T.; Nishimura, Y.; Kakinuma, K. Tetrahedron Lett. 2003, 44, 3275.

(19) Nemoto, N.; Shida, Y.; Shimada, H.; Oshima, T.; Yamagishi, A. Extremophiles 2003,

7, 235.

(20) Fitz, W.; Arigoni, D. J. Chem. Soc., Chem. Commun. 1992, 1533.

(21) Comita, P. B.; Gagosian, R. B. Science 1983, 222, 1329.

(22) Morii, H.; Nishihara, M.; Koga, Y. J. Biol. Chem. 2000, 275, 36568.

(23) Dannenmuller, O.; Arakawa, K.; Eguchi, T.; Kakinuma, K.; Blanc, S.; Albrecht, A.-

M.; Schmutz, M.; Nakatani, Y.; Ourisson, G. Chem. Eur. J. 2000, 6, 645.

(24) Alcaraz, M.-L.; Peng, L.; Klotz, P.; Goeldner, M. J. Org. Chem. 1996, 61, 192.


169

Pag

e16

9

7

(25) Davisson, V. J.; Woodside, A. B.; Neal, T. R.; Stremler, K. E.; Muehlbacher, M.;

Poulter, C. D. J. Org. Chem. 1986, 51, 4768.

(26) Modro, A. M.; Modro, T. A. Org. Prep. Proced. Int. 1992, 24, 57.

(27) Watanabe, Y.; Nakamura, T.; Mitsumoto, H. Tetrahedron Lett. 1997, 38, 7407.

(28) Sparrow, C. P.; Raetz, C. R. J. Biol. Chem. 1985, 260, 12084.

(29) Eguchi, T.; Terachi, T.; Kakinuma, K. J. Chem. Soc., Chem. Commun. 1994, 137.

(30) Howell, S. J.; Spencer, N.; Philp, D. Tetrahedron 2001, 57, 4945.

(31) Jin, Y.; Roberts, F. G.; Coates, R. M. Org. Synth. 2007, 84, 43.

(32) Keeping the solution of the cuprate too long at -78 °C leads to the formation of a

heavy precipitate.

Chapter 7

170

7

Summary

171

Summary

The idea that saponifiable phospholipids are only building blocks and energy

source of a cell has been refuted by the discovery of phospholipid signalling.

Currently, these lipids are widely studied, mainly because of their potential

undiscovered functions. One aspect that makes the study of phospholipids difficult,

is their limited availability in pure form. Biological membranes, which are the main

source of phospholipids, are typically composed of several tens to hundreds very

similar species. In this mixture, only a minute amount of a single species might have

a specific physiological property. To find, and to isolate this lipid reminds of the

phrase “search the needle in a haystack”. Despite organic chemistry cannot help

finding the needle directly, it offers a different solution; to prepare a new needle.

However, this requires efficient strategies that allow the synthesis of defined,

chemically pure lipids and their derivatives. This is what this thesis presents.

Chapter 2 describes a modular synthesis of branched fatty acids, which are

mainly lipid components of the membranes of various bacterial pathogens. The

described synthesis allows preparation of a fatty acid bearing the methyl-branch at

any position of the linear chain together with the desired absolute configuration.

Figure 1. Modular synthesis of branched fatty acids.

This modular approach was successfully applied in the synthesis of 3 different fatty

acids in the amounts relevant for further studies.

Despite the fact that glycerophospholipids seem relatively simple

compounds, their synthesis might be surprisingly complex. Chapter 3 is devoted to

the synthesis of these glycerophospholipids. A cobalt catalyst allowed the

regioselective ring opening of a silyl protected glycidol. The obtained protected

Summary

172

monoacylglycerol was further esterified in the same pot. The resulting silyl protected

diacylglycerol was deprotected without any notable migration, and subsequently

converted to a phospholipid.

Figure 2. One pot synthesis of protected diacylglycerols and the subsequent conversion to

a phospholipid.

Chapter 4 is an extension of this methodology to the synthesis of

triacylglycerols. When glycidyl esters were used as the starting material, the same

conditions afforded enantiopure triacylglycerols. This methodology was

demonstrated in the synthesis of 18 different triacylglycerols.

Figure 3. An efficient, 3 step synthesis of triacylglycerols.

Furthermore, this chapter described the first steps towards the automated synthesis

of triacylglycerols by a liquid-handling platform. The obtained triacylglycerols can be

applied as analytical standard in the analysis of triacylglycerols mixtures such as milk

fat.

The synthetic solutions described in the chapters 2 and 3 allowed the

synthesis of phospholipids bearing methyl branched fatty acids. These lipids were

Summary

173

prepared in amounts, which allowed initial studies of their properties as membrane

components.

Figure 4. Study of the influence of a methyl-branch on the organization of a bilayer.

The phospholipids were converted into 2-component liposomes. The bilayers of

these liposomes were studied by molecular dynamics simulations. Another studied

aspect of these lipids was their interaction with membrane proteins.

Chapter 6 is also dedicated to the influence of lipids on protein function, but from a

different perspective. Protein lipidation is a posttranslational modification, by which

lipids control the activity of proteins. The chapter describes the synthesis of a lipid

probe, which might be useful in protein lipidation studies in living cells.

Figure 5. A probe for the study of protein lipidation displays a rapid and clean dipolar

cycloaddition.

Summary

174

Chapter 7 is dedicated to (phospho)lipids found in Archaea. Archaea form a

domain of Life, which early in evolution strayed from the remaining 2 domains.

Figure 6. ( I ) Synthesis of a biosynthetic intermediate in archaeal lipid biosynthesis for the

identification of an involved enzyme; ( II ) Epoxide ring opening as a key step in the total

synthesis of cycloarchaeol.

The separated evolution granted, that the species from this domain

developed their own lipid biosynthetic pathways. Study of these pathways is

challenging because there are few assays and the corresponding biosynthetic

intermediates are not available. Chapter 7 presents a synthesis of one of these

intermediates, which was necessary for the identification of one of the missing

enzymes in archaeal lipid biosynthesis. Furthermore, chapter 7 describes a key step

in the total synthesis of cycloarchaeol – an archaeal lipid backbone.

Samenvatting

175

Samenvatting

Het idee dat verzeepbare fosfolipiden alleen functioneren als bouwstenen

en energiebron voor de cel werd weerlegd door de ontdekking van signaalfuncties

van fosfolipiden. Tegenwoordig worden deze lipiden uitgebreid bestudeerd, en

steeds worden nieuwe functies ontdekt. Een aspect wat de studie van fosfolipiden

bemoeilijkt, is hun gelimiteerde beschikbaarheid in zuivere vorm. Biologische

membranen, de hoofdbron van fosfolipiden, bestaan vaak uit honderden zeer

vergelijkbare lipiden. Daarin is soms maar een zeer geringe hoeveelheid van een

bepaald lipide met een specifieke fysiologische eigenschap aanwezig. Het vinden en

isoleren van dit lipide doet denken aan het spreekwoord “een speld in een hooiberg

zoeken”. Hoewel de organische chemie niet direct kan helpen naar de zoektocht

naar de speld, geeft het wel een andere oplossing; het maken van een zo’n speld.

Dit vereist wel een efficiënte strategie voor de synthese van gedefinieerde,

chemisch zuivere lipiden en hun derivaten. Dat is wat dit proefschrift beschrijft.

Hoofdstuk 2 beschrijft de modulaire synthese van vertakte vetzuren, vaak

hoofdbestanddelen van membranen van bacteriële pathogenen. De beschreven route

maakt het mogelijk om vetzuren te synthetiseren die methylsubstituenten hebben op

elke positie van de lineaire keten, in combinatie met de gewenste absolute

configuratie.

Figure 1. De modulaire synthese van vertakte vetzuren

Deze modulaire benadering is succesvol toegepast in de synthese van 3 verschillende

vetzuren in hoeveelheden die toereikend zijn voor verdere studies.

Samenvatting

176

Ondanks het feit dat glycerofosfolipiden een vrij eenvoudige structuur lijken

te hebben, is hun synthese verrassend complex. Hoofdstuk 3 is dan ook gewijd aan

de synthese van deze glycerofosfolipiden. Met behulp van een kobalt katalysator kan

een beschermd glycidol regioselectief worden geopend. Het verkregen

monoacylglycerol wordt nogmaals veresterd in dezelfde reactiekolf. Het resulterende

beschermde diacylglycerol wordt ontschermd zonder acylmigratie en vervolgens

omgezet naar het fosfolipide.

Figure 2. Eénpotssynthese van beschermde diacylglycerolen en de omzetting naar een

fosfolipide

Hoofdstuk 4 is een uitbreiding van deze methodologie tot de synthese van

triacylglycerolen. Door glycidyl-esters te gebruiken als startmateriaal, kunnen onder

dezelfde condities enantiozuivere triacylglycerolen worden geïsoleerd. Deze

methodologie is gedemonstreerd in de synthese van 18 verschillende

triacylglycerolen.

Figure 3. Een efficiënte, 3-stapssynthese van triacylglycerolen.

Samenvatting

177

Bovendien beschrijft dit hoofdstuk ook de eerste stap in de richting van de

geautomatiseerde synthese van triacylglycerolen met behulp van een pipetteer robot.

De verkregen triacylglycerolen kunnen gebruikt worden als standaarden in de analyse

van triacylglycerolmengsels zoals in melkvet.

Hoofdstuk 2 en 3 beschrijven een strategie voor de synthese van fosfolipiden

met methylvertakte vetzuren. Deze lipiden zijn gesynthetiseerd in hoeveelheden die

initiële studies toelieten naar hun eigenschappen als membraanlipiden.

Figure 4. De invloed van methylvertakkingen op de organisatie van de lipide bilaag, links

de moleculaire dynamicasimulaties.

De fosfolipiden werden gebruikt in liposomen. De bilaag van deze liposomen is

gesimuleerd met moleculaire dynamica. Het blijkt dat kanaalvormende

membraaneiwitten succesvol in deze liposomen kunnen worden gereconstitueerd.

Hoofdstuk 6 is ook gewijd aan de invloed van lipiden op de functie van eiwitten,

maar dan vanuit een ander perspectief. Eiwitlipidering is een posttranslationele

modificatie waarbij lipiden de activiteit van eiwitten sturen. Dit hoofdstuk beschrijft

Samenvatting

178

de synthese van een lipide-sonde, die gebruikt zou kunnen worden in levende cellen

voor de modificatie van eiwitten met lipiden om ze dusdanig te kunnen bestuderen.

Figure 5. Een sonde voor de studie van eiwitlipidering laat een schone en snelle

dipolaire cycloadditie zien.

Hoofstuk 7 is gewijd aan (fosfo)lipiden die gevonden worden in Archaea.

Archaea vormen één van de domeinen van het leven, die in een beginstadium van de

evolutie is afgesplitst van de andere twee domeinen.

Figure 6. ( I ) Synthese van een intermediair uit de biosynthese van archaiële lipiden in

verband met de identificatie van het betrokken enzym. ( II ) Epoxide ringopening als de

belangrijkste stap in de totaalsynthese van cycloarchaeol.

Samenvatting

179

De gescheiden evolutie van deze domeinen liet toe dat Archaea hun eigen

route hebben ontwikkeld voor lipidensynthese. Het bestuderen van deze

syntheseroute is uitdagend, omdat er maar enkele analyses bestaan en de bijhorende

biosynthese-intermediairen niet beschikbaar zijn. Hoofdstuk 7 beschrijft de synthese

van één van deze intermediairen, die leidde tot de identificatie van een enzym in de

lipide-biosynthese. Hoofdstuk 7 beschrijft ook een belangrijke stap in de

totaalsynthese van cycloachaeol, een membraanlipide uit Archaea.

Acknowledgements

180

Acknowledgements

Možno netypicky, ale v prvom rade sa chcem podakovať mojím rodičom. Tí

si so mnou užili viac než dosť počas môjho predchádzajúceho štúdia. Mama, ty si

ma naučila, že ak sa chcem niekam dostať, tak nemôžem byť čajová nula, ktorá sa

vzdá pri prvej, druhej, alebo dvadsiatej prekážke. Oco, nakoniec som sa predsa len

“potatil”. Zrejme ta to môžem poďakovať genetickej výbave, ktorú som po tebe

zdedil. Keď teba môže baviť labák a pokusy po viac ako 50 rokoch, prečo by

nemohol aj mňa...? Kuku (Tomáš)! Bez teba by som bol úplne iný človek! Dúfam, že

úspešne doštuduješ, potom úspešne skončíš PhD a potom ovládneme svet...muhaha!

Leti, existen un millón de razones por las que te estoy inmensamente

agradecido, dos de las cuales son: primero, te agradezco todo tu esfuerzo que has

puesto en la corrección de tanto mi tesis como de mis artículos. Segundo y

muchísimo más importante, te quiero agradecer la oportunidad que me has regalado

al elegir compartir tu vida conmigo. Estoy ansioso por conocer que nos va a regalar

el futuro juntos.

Adri, I would like to thank you for at least dozen of different things. First of

all, thank you for the opportunity to join your group and do PhD. under your

supervision. Then, thank you for your supervision, care, trust when you offered me

to work on Mycolic acid on everything you did for me. Dank je wel!

Niek and Manuel, where to start. Well, first of all, thank you for being my

paranymphs, friends and gym sparring partners. You deserve all my respect for all

the “gins at the bar”, “just 2 more repetitions”, “I know an awesome

exercise”...gentlemen, you are the real champions!

Selma, Nick, Jelle and Steven. With students like you I kind off won a lottery.

It was a pleasure to supervise a bunch of motivated and enthusiastic students. Thank

you!

I would also like to thank to the group of people joining for the Friday

dinners at Papa Joe’s, Klein Moghul, Tel Aviv and other places we visited. Paul, Ilse,

Mattia, Mannathan, Francesco, Guilloume, Melanie, Hylke and all those, which

I forgot (sorry). Thank you for your company.

Thanks also goes to all the members of Stratingh institute, which I had the

pleasure to meet.

university of groningen stereoselective synthesis of

Documents