• No results found

Multicyclic peptides via CLIPS and oxime ligation: Synthesis of acid labile amino-oxy protection groups

N/A
N/A
Protected

Academic year: 2021

Share "Multicyclic peptides via CLIPS and oxime ligation: Synthesis of acid labile amino-oxy protection groups"

Copied!
40
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Bachelor Thesis Scheikunde

Multicyclic peptides via CLIPS and oxime ligation

Synthesis of acid labile amino-oxy protection groups

door

Filip Aleksic

14-07-2017

Studentnummer 10754830 Onderzoeksinstituut

Van ’t Hoff Institute for Molecular Sciences (HIMS)

Onderzoeksgroep

Synthetic Organic Chemistry (SOC)

Verantwoordelijk docent

Prof. Dr. J. H. van Maarseveen

Begeleider

(2)
(3)

3

Abstract

The synthesis of multi-cyclic peptide chains has proven to be important for possible applications in the antibody design field. CLIPS technology has shown to be an effective way to constrain peptide chains into cyclic structures. However, when tricyclic structures are attempted to be synthesized using CLIPS regioisomers are rapidly formed. To avoid the formation of these regioisomers, a second orthogonal coupling reaction was used: oxime ligation. For a sequential orthogonal coupling of CLIPS and oxime ligation the pH has to be mediated to fit the reaction conditions of both coupling methods. CLIPS is performed at pH = 8, after which the pH is lowered to <1 to remove the acid labile protection group (Boc) on the amino-oxy moiety. The pH is then increased to pH = 4 to meet the optimal oxime ligation conditions. The large volumes of acid and base needed are not practical to work with when applying this method to screening a large peptide library. It was envisioned that through the use of a highly acid labile protection group a pH buffer could simply be used to mediate the reactions and deprotect the acid labile protection group, thereby eliminating the need for large volumes of acid and base. The aim of this project was to find a highly acid labile protection group for an amino-oxy moiety. Two acid labile protection groups were tested; Mtt and t-Bumeoc. It was found that the free amino-oxy rapidly reacts with trace amounts of acetone, thereby nullifying the reactivity of the amino-oxy. The removal of the acetone adduct was successful by allowing the acetone adduct to react with methoxyamine. The Mtt protected amino-oxy was synthesized. However, the high acid lability of the compound resulted in rapid degradation on acidic silica when purification was attempted using column chromatography. Crystallisation was additionally attempted in Et2O and pentane with DCM as a co-solvent, however

neither showed any product. The tertiary alcohol of t-Bumeoc was synthesized. However, the instability of the chloroformate, and imidazole carbamate of t-Bumeoc lead to the inability to couple the protection group to an amino-oxy.

(4)

4

Samenvatting

Het nabootsen van de in het immuunsysteem aanwezige moleculen om deze als medicijn toe te dienen is een zeer effectieve manier bij bepaalde ziektes die het immuunsysteem aantasten. HIV en influenza zijn bijvoorbeeld twee virussen die op deze manier bestreden kunnen worden. Echter is het maken van dit soort mimicry moleculen geen simpele taak omdat processen in ons lichaam met een hoge specificiteit plaatsvinden. In het geval van het immuunsysteem reageert een antilichaam en een antigen met elkaar met behulp van het sleutel-slot-principe. Als men een antigen synthetisch wil namaken moet de ‘sleutel’ van dit molecuul (die selectief in het slot van een antilichaam past) zo goed mogelijk lijken op het door het lichaam gemaakte antigen. Antigenen hebben echter vaak een complexe structuur met veel verschillende gebogen bindingsplekken waar een keten aminozuren in de vorm van meerdere lussen op past.

Het maken van deze lussen kan bereikt worden door de keten aminozuren vast te maken aan de buitenkant van een rigide steiger molecuul. Bij het koppelen van de keten aan de steiger wordt gebruik gemaakt van twee verschillende koppelingsreacties die individueel plaatsvinden bij verschillende condities, gehete CLIPS (Chemical Linkage of Peptides onto Scafolds) en oxime ligatie. De CLIPS reactie vindt eerst plaats, waarna oxime ligatie wordt geïnitieerd (Figuur 1).

Figuur 1: Een schematische weergave van de twee koppelings reacties. Links is een T4 rigide steiger molecuul te zien met 2 rode CLIPS bindingsplekken en 2 groene oxime ligatie bindingsplekken. Op de

lineaire aminozuurketen zitten ook 2 CLIPS en oxime ligatie bindingsplekken in rood en blauw respectievelijk. Allereerst wordt een CLIPS reactie uitgevoerd gevolgd door oxime ligatie.

De koppelingsreacties vinden plaats bij verschillende condities. De CLIPS koppelings reactie wordt geïnitieerd bij een licht basische pH van rond de 8, terwijl oxime ligatie het snelst is bij licht zure condities plaatsvindt van een pH waarde rond de 4. Men wil echter uiteraard niet dat de koppelings reacties door elkaar heen kunnen plaatsvinden.

(5)

5 Daarom wordt gebruik gemaakt van een speciale groep die wordt vastgeplakt aan de reactieve plek van oxime ligatie aan het rigide steiger molecuul die de reactiviteit van deze plek compleet tenietdoet. Zo een speciale groep wordt ook wel een beschermgroep genoemd (immers beschermt men de oxime ligatie reactieve plek zodat deze niet meer kan reageren).

In dit geval kan deze beschermgroep weer worden verwijdert met behulp van geconcentreerd zuur. De ontscherming vindt plaats bij een pH waarde lager dan 1, dit is te laag voor oxime ligatie en de pH moet vervolgens weer omhoog gebracht worden met een base. Deze grote hoeveelheden zuur en base die moeten worden toegevoegd zijn praktisch niet handig wanneer op kleine schaal veel verschillende aminozuurketens worden getest. De buisjes waarin deze testjes worden gedaan zijn zo klein dat ze simpelweg overvloeien nadat al het zuur en base is toegevoegd.

Een oplossing voor dit probleem is het gebruik maken van een beschermgroep die onder minder zure condities er al af gaat. Dit zou het benodigde volume zuur verminderen en het gebruik van een base zou dan niet meer nodig zijn (Figuur 2).

Figuur 2: Schema van toevoegingen zuur en base die nodig zijn. Bij een beschermgroep die er onder minder zure condities al af gaat is geen base nodig om oxime ligatie te initiëren.

In dit project zijn twee beschermgroepen gebruikt: Mtt en t-Bumeoc. De Mtt beschermde oxime ligatie bindingsplek was gesynthetiseerd, echter was het zuiveren van dit molecuul tot dusver niet gelukt. Het zuiveringsproces dat gebruikt werd vindt plaats onder licht zure condities en dit was al genoeg om de beschermgroep van het molecuul af te splitsen. De tweede beschermgroep t-Bumeoc was ook gesynthetiseerd. De volgende stap zou zijn om de beschermgroep te koppelen aan de oxime ligatie bindingsplek, echter is dit niet gelukt. De beschermgroep bleek na meerdere verschillende manieren geprobeerd te hebben om het te koppelen simpelweg niet geschikt voor dit project.

(6)

6

List of abbreviations

Adpoc 1-(1-Adamantyl)-1-methylethoxycarbonyl carbamate

Boc Tert-butoxycarbonyl

Cbz Carboxybenzyl

CLIPS Chemical Linkage of Peptides onto Scaffolds

CDI 1,1-Carbonyldiimidazole

d Doublet (by NMR)

DCM Dichloromethane

DIAD Diisopropyl azordicarboxylate

DIPEA N,N-Diisopropylethylamine

DMSO Dimethylsulfoxide

Equiv Equivalents

EtOAc Ethyl acetate

EtOH Ethanol

IR Infrared spectroscopy

m Multiplet (by NMR)

Mtt 4-Methyltrityl

NBS N-Bromosuccinimide

NMR Nuclear Magnetic Resonance

OSu N-Hydroxysuccinimide

pAcF p-Acetylphenylalanine

Pd/C Palladium on activated carbon

PE Petroleum ether (40-60) PG Protection Group phth Phthalimide py Pyridine q Quartet (by NMR) s Singlet (by NMR)

t-Bumeoc 1-(3,5-Di-t-butylphenyl)-1-methylethyl carbamate

TFA Trifluoroacetic acid

THF Tetrahydrofuran

TLC Thin layer chromatography

(7)

7

Table of contents

Abstract. . . 3

Samenvatting . . . 4

List of abbreviations . . . 6

Table of contents . . . 7

1. Introduction . . . 8

1.1 Chemical Linkage of Peptides onto Scaffolds (CLIPS) . . . 9

1.2 Orthogonal coupling reaction: Oxime ligation . . . 10

1.3 Results in the group . . . 12

1.4 Aim of the project. . . 13

2 Experimental. . . 14

2.1 Synthesis of scaffold: CLIPS half. . . 14

2.2 Synthesis of scaffold: Protected amino-oxy. . . 15

2.2.1 Acetone adduct formation and prevention . . . 16

2.2.2 Purification of the Mtt protected amino-oxy. . . 18

2.3 Alternative protection group: t-Bumeoc . . . 19

2.3.1 Mechanistic discussion of the Friedel-Crafts alkylation . . . 19

2.3.2 Synthesis towards the chloroformate. . . 21

2.3.3 Coupling t-Bumeoc to an amine: Literature approach . . . 22

2.3.4 Coupling t-Bumeoc to an amine: The use of CDI. . . 23

3 Conclusion . . . 23

4 Future prospects . . . 24

5 Acknowledgements . . . 25

6 References . . . 26

7 Experimental. . . 27

7.1 General remarks. . . 27

8 Appendix . . . 33

(8)

8

1. Introduction

Immunotherapy is widely regarded as one of the most promising ways to cure patients from chronic diseases which target the immune system.1 Therefore, understanding the complex peptide interactions of the immune system can be tremendously beneficial to the design of new immunotherapy drugs. Typically, immunotherapy drugs consist of artificial antibodies which replace the antibodies that have been destroyed by a debilitating disease. Antigens contain binding sites (epitopes) to which an antibody can bind with high affinity and selectively with regions on the antibody, called paratropes (Figure 1).

Figure 1: Schematic view of an antigen and antibody. The paratrope and epitope bind to each other with remarkable specificity due to their pairing 3D-structure.

There are two different types of epitopes: continuous and discontinuous. Continuous epitopes are linear peptide chains which interact with the paratrope on an antibody, whereas discontinuous epitopes consist of multiple peptide chains (Figure 2).

Figure 2: Left: Continuous epitope. A single loop on a linear peptide forms the binding site for the paratrope. Right: Discontinuous epitope. Multiple loops on different parts of the epitope form the

binding site.2

In vitro experiments have been performed to synthesize peptide loops which can effectively mimic

continuous epitopes with the use of a peptide which was constrainted into 1 loop.3 However, discontinuous epitopes have proven to be particularly difficult to emulate. It has also been reported that the vast majority, if not all, epitopes are discontinuous to some extent.4

(9)

9 Therefore, an extensive amount of research has been conducted to find ways to efficiently synthesize peptides into a multi-loop structure. Various methods to constrain linear peptides into the correct conformation have since been developed, such as CLIPS (Chemical Linkage of Peptides onto Scaffolds) technology, oxime ligation, Click-chemistry, and disulfide formation.5, 6, 7, 8

1.1 Chemical Linkage of Peptides onto Scaffolds (CLIPS)

CLIPS was found to be particularly promising due to the fact that it is a remarkably clean and fast reaction under mild reaction conditions (Figure 3).9, 10, 11 As the name suggests the main synthetic strategy is the constraining of the linear peptide chain onto a scaffold by substituting the bromine on the scaffold with the thiol of a cysteine residu of the peptide chain.

Figure 3: Schematic view of CLIPS technology.5

The potential of CLIPS was explored by Heinis et al. who showed that the binding affinity of a bicyclic system to an enzyme dramatically increased relative to the mono-cyclic and linear systems.12

(10)

10 This raises the question about whether or not a tricyclic, tetracyclic, or even a pentacyclic peptide system would show an even better activity. However, this is where the limitations of CLIPS technology come into view. When a tricyclic (or more) system is synthesized using CLIPS regioisomers are rapidly formed (Figure 4).

Figure 4: Regioisomer formation for CLIPS reaction on a scaffold with 4 binding sites. After the first CLIPS couples, the next thiol on the peptide can react with any of the 3 remaining bromines on the

scaffold, thereby forming regioisomers.

1.2 Orthogonal coupling reaction: Oxime ligation

For potential applications of CLIPS in the pharmaceutical industry it is crucial to have a process which is able to produce a product without regioisomers to avoid costly large scale separation procedures (if the separation is even possible at all). Combining CLIPS with another orthogonal coupling reaction with each only having 2 reactive sites on the scaffold could prevent the formation of regioisomers. In our group prevention of regioisomer formation is explored by combining CLIPS technology with an orthogonal coupling reaction; oxime ligation (Figure 5).

Figure 5: Schematic view of the process of an orthogonal CLIPS and oxime ligation reaction with a 4 functional group scaffold forming a peptide chain suspended into 3 loops.

Oxime ligation is the reaction of an amino-oxy with a ketone resulting in the formation of an oxime with the removal of H2O (Figure 6). The reaction is typically executed in acidic conditions of a pH

value at around 4. Performing the reaction at an even more acidic pH would lower the reaction rate significantly, as the free amino-oxy would become protonated which would reduce its nucleophilic strength.

(11)

11

Figure 6: Oxime ligation mechanism in an acidic environment. An amino-oxy group performs a nucleophilic attack on a ketone in an acidic environment. After elimination of water the oxime is

formed.

When a primary amine reacts with a ketone forming an imine, the equilibrium heavily favours the ketone. However α-effect nitrogens, such as amino-oxy groups, favour the formation of the oxime (Figure 7).13 The high reactivity and selectivity, coupled with the mild reaction conditions makes oxime ligation a highly effective orthogonal coupling reaction.

Figure 7: Above: Formation of an imine. The equilibrium is favoured for the ketone. Below: Oxime ligation. The equilibrium is favoured for the oxime. Both E- and Z-isomers are formed.

By adding an amino-oxy functional group to the scaffold, which can react with a ketone on the side chain of a peptide, oxime ligation can be used as an orthogonal coupling reaction. However, no natural amino acids contain a ketone group. Therefore the synthetic amino acid p-acetylphenylalanine (pAcF) was used in the synthesis of the linear peptide chain to allow for the coupling reaction (Figure 8).

(12)

12

1.3 Results in the group

The 4 functional group scaffold T4-N3 was tested in the group showing promising results (Figure 9). The linear peptide chain with the amino acid sequence AcC-E-pAcF-A-pAcF-A-K-CNH2 (C=

Cysteine, E = Glutamate, A = Alanine, K = Lysine) coupled with high efficiency to the scaffold for the CLIPS reaction. However, following the oxime ligation it was found that both E- and Z-isomers are formed of the oxime bond. The problem of E- and Z- isomers forming was not in the scope of this project and will therefore not be discussed in more detail.

Figure 9: The orthogonal coupling of the linear peptide chain AcC-E-pAcF-A-pAcF-A-K-CNH2 onto the T4-N3 scaffold. The CLIPS reaction was performed under basic conditions, after which the pH

was reduced to remove the Boc group. The pH was then increased slightly for an efficient oxime ligation reaction.

A problem arose with the Boc PG used to protect the amino-oxy, as the conditions to remove this PG are relatively harsh. Typically concentrated HCl or pure trifluoroacetic acid (TFA) is used. These conditions are not practical to work with when screening a large number of different linear peptide chains in peptide library format. These screening reactions are done in microwell plates which fill up rapidly and can even overflow due to the large volumes of acid and base needed to mediate the conditions for the different coupling, and deprotection reactions.

Therefore a new, more acid labile amino-oxy PG is desired. The research question of this project is: can a scaffold be synthesized with a highly acid labile PG on the amino-oxy substituent for use in an orthogonal CLIPS and oxime ligation?

(13)

13

1.4 Aim of the project

For this project the exceptionally acid labile 4-methyltrityl (Mtt) will be used to protect the amino-oxy group (Figure 10). The Mtt group was chosen not only for its acid lability, but also for the fact that it is well described in literature and is readily commercially available as MttCl.14

Figure 10: The T4-N2 scaffold with the Mtt protected amino-oxy group.

To synthesize the scaffold the synthetic approach showed in Figure 11 and 12 will be followed. Diethanolamine was protected with carboxybenzyl (Cbz) followed by a Mitsunobu reaction with n-hydroxyphthalimide to yield product 3. After treatment with hydrazine the free amino-oxy is subsequently protected with MttCl. After hydrogenation over Pd/C to remove the Cbz PG the Mtt protected amino-oxy 6 is obtained. The other part of the scaffold was synthesized from 3,5-dimethylbenzoic acid which was converted to the t-Bu ester 7. After bromination the carboxylic acid is formed which can then be coupled to product 6 yielding the scaffold.

(14)

14

Figure 12: Synthetic route of the CLIPS reactive group part of the scaffold, followed by the coupling of both scaffold parts to each other.

2 Experimental

2.1 Synthesis of scaffold: CLIPS half

The top half of the scaffold was synthesized from 3,5-dimethylbenzoic acid which converted to the t-Bu ester 8 followed by the bromination of the product to the di-brominated product 9 (Scheme 1).

Scheme 1: t-Bu ester synthesis followed by the bromination towards product 9.

This reaction had to be followed using NMR as the bromination as the mono-brominated compound is formed before the desired product 9. Additionally, bromination does not stop after compound 9 is formed, as 9 is also converted by NBS into the tri-brominated compound 9-1 (Scheme 2).

(15)

15

Scheme 2: Bromination of 9 towards the tri-, and tetra-brominated products 9-1 and 9-2.

An equilibrium is desired where most of the starting material 8 is converted to the product 9 whereby most of 9-1 is converted and the concentration of 9-2 is kept at a minimum. Typically crystallization in n-hexane is carried out to purify the product 9. However, after 3 attempts no crystallization was observed. Co-crystallization with left over product 9 from previous experiments was also unsuccessful. Therefore, it was decided to continue the synthesis of the scaffold with the crude product. Following the bromination the t-Bu ester was converted to the carboxylic acid 10 using formic acid in dichloromethane (DCM) (Scheme 3).

Scheme 3: Carboxylic acid synthesis from the t-Bu ester.

The next step of the synthesis is the coupling of the carboxylic acid to the free secondary amine of the Mtt protected amino-oxy. The synthesis of this compound will be discussed next.

2.2 Synthesis of scaffold: Protected amino-oxy

The sysnthesis was started with a straightforward protection of the secondary amine of diethanolamine to yield product 2 followed by a Mitsunobu reaction with n-hydroxyphthalimide to introduce the amino-oxy moiety, which yielded the diphthalimide product 3 (Scheme 4).

(16)

16

Scheme 4: Cbz protection of diethanolamine.

Afterwards, the phth groups were removed with hydrazine to yield the free amino-oxy 4 which allows us to add the desired acid labile PG. However, after analysis of the product it became evident that no free amino-oxy was present, instead the acetone adduct 4-1 was formed (Scheme 5).

Scheme 5: Removal of phth groups with hydrazine followed by the unwanted side reaction with trace amounts of acetone towards the oxime 4-1.

2.2.1 Acetone adduct formation and prevention

Although no acetone was used in any steps of the synthesis trace amounts appeared to be present in the reaction mixture, either from cleaning glassware, vapours which entered the flask during the use of the rotary evaporator, or simply trace amounts present in the solvents used. Ethanol specifically commonly has ketone and aldehyde impurities which react with the free amino-oxy. The rapid formation of the acetone adduct showed the strength of the oxime ligation reaction, however the formed oxime nullifies the reactivity of the scaffold towards future oxime ligation, therefore it is a highly undesired product.

(17)

17 In an attempt to recover the free amino-oxy from the acetone adduct methoxyamine was added in a large excess and after 2 days at room temperature the free amino-oxy 4 was obtained in poor yield of 9% (Scheme 6).

Scheme 6: Recovering the free amino-oxy from the acetone adduct with the use of methoxyamine

The low yield could be a result of low solubility of the protonated free amino-oxy into the water phase during the acid-base extraction which was performed. The large, apolar Cbz was likely too difficult to solute in water despite the highly polar protonated amino-oxy groups. However, the aim of this reaction was to recover some of the lost product, therefore a low yield was acceptable.

Avoiding the formation of the acetone adduct was a pressing issue and several special measures were taken to make sure no acetone was present in reactions involving free amino-oxy groups. All glassware which was used was either flame dried or kept in an oven at 150°C before use. Additionally, distilled tetrahydrofuran (THF) was chosen as a solvent for future reactions, as it has been shown in our group to be an effective solvent when working with free amino-oxy containing compounds. Further prevention of the acetone adduct formation was successfully attempted by performing a one-pot deprotection - protection sequence from the di-phth compound 3 to the Mtt protected amino-oxy 5. Previously hydrazine soluted in water was used for the removal of the phth groups. However, this resulted in a 2 phase system with THF, which was not ideal. Methylhydrazine was instead used to avoid this problem. Additionally, the use of methylhydrazine resulted in reaching full conversion towards the free amino-oxy after only 5 to 10 min. Afterwards, MttCl was added alongside N,N-diisopropylethylamine (DIPEA) in 10 and 5 equiv respectively and left to stir overnight to yield the Mtt protected amino-oxy 5 (Scheme 7).

(18)

18 Analysis of the product was performed with NMR whereby three signals were used to determine conversion of the product (Figure 13).

Figure 13: 1H-NRM spectrum of the crude product from Scheme 10.

The 1H-NMR signals of the hydrogens marked with 1 and 2 did not only shift to a lower ppm value compared to the phth protected compound from 4.44 ppm to 3.63 ppm and from 3.91 ppm to 3.18 ppm respectively, but the splitting pattern also changed. Due to hindered rotation around carbamate Cbz bond the phth protected compound showed doublet of triplet-like structures for signals 1 and 2. After the Mtt protection signal 1 changed into a broad singlet, and 2 into two separate broad signals. This change indicated a more rotationally hindered compound, which is what is expected given the more bulky Mtt PG.

2.2.2 Purification of the Mtt protected amino-oxy

To purify the Mtt protected compound column chromatography was not the preferred method given that the acid lability of the product could result in rapid degradation on the acidic silica. It was argued that due to the large aromatic structure of the Mtt PG the isolated product could be crystalline. Therefore, crystallization was firstly attempted in Et2O, and pentane with DCM as a co-solvent.

(19)

19 NEt3 treated silica was therefore used despite the possible degradation problems. After two sequential

attempts to isolate the product using column chromatography no product could be isolated due to the similarity in polarity of the by-products and main product. The structure of the compounds which are formed as by-products has not been determined. However, due to the large signals in the aromatic region of the 1H-NMR spectrum in Figure 13 it was assumed that the by-products are likely Mtt-containing compounds. Following the second column it also became evident that the mixture became more complicated than before, as more spots appeared on TLC even closer together than previously observed after the first column. This change likely occurred due to degradation of the Mtt protected compound.

2.3 Alternative protection group: t-Bumeoc

Following the difficulties to isolate the Mtt protected compound alternative acid labile PGs were investigated. Different carbamates were explored as alternatives, as they would likely be more robust towards acidic degradation on silica, which would allow for the purification of the protected product. 1-(3,5-Di-t-butylphenyl)-1-methylethyl carbamate (t-Bumeoc) (Figure 14) was chosen as an alternative, as it has been shown to be removed from an amine in an 80% AcOH/H2O solution 4000

times faster than Boc.15, 20 The synthesis of t-Bumeoc will now be discussed in more detail.

Figure 14: t-Bumeoc protection group coupled to a generic amino-oxy RONH2.

2.3.1 Mechanistic discussion of the Friedel-Crafts alkylation

The first step of the t-Bumeoc synthesis is a Friedel-Crafts alkylation of toluene with tert-butyl chloride (t-BuCl) (Scheme 8).

(20)

20 Interestingly the t-Bu groups are added at the meta positions relative to the methyl of toluene, which is odd because typically methyl substituents are ortho-para direction groups. The exact mechanism of this reaction was not found in literature. However, two different mechanisms will be proposed. The first mechanism relies on the addition and removal of t-Bu groups being in equilibrium with each other. Due to steric hindrance between the t-Bu and methyl group the ortho addition is rare. On the contrary, the meta addition of a t-Bu group results in no steric interaction with the methyl group. This results in the reaction eventually moving towards the 3,5-di-tert-butylated compound which can be seen as a thermodynamic product (Figure 15).

Figure 15: First proposal of the mechanism of the Friedel-Crafts alkylation of toluene with t-BuCl.

A second mechanistic explanation can be a methyl shift from the 2,4-di-tert-butylated product towards the 3,5- substituted product.

(21)

21 To determine which of these mechanistic explanations is correct two experiments could be performed. Firstly, the Friedel-Crafts alkylation of 4-tert-Butyltoluene can be performed to see if any 3,5- product is formed, which would indicate a methyl shift. Additionally, the alkylation of toluene can be performed using deuterated toluene to see if any deuterium scrambling takes place. Deuterium scrambling would occur when the t-Bu groups detach and re-attach from the toluene repeatedly and would therefore indicate the presence of the thermodynamic product 3,5-di-tert-Butyltoluene.

2.3.2 Synthesis towards the chloroformate

3,5-Di-tert-butyltoluene was oxidized to the carboxylic acid 15 using potassium permanganate. The low yield for this reaction could be contributed to incomplete oxidation of 14. However, a large portion of the product was likely also lost during the difficult workup due to the formed MnO2, which

stained the glassware that was used, and made it difficult to see a phase separation during extractions. Afterwards, the methyl ester 16 was formed which was reacted with the Grignard reagent MeMgBr to yield the tertiary alcohol 17 (Scheme 9).

Scheme 9: Synthesis of the Me ester, followed by a Grignard reaction to yield the tertiary alcohol 17.

The next step of the synthesis was to convert the tertiary alcohol into an activated compound which could then be coupled to the amino-oxy 4. Firstly, the chloroformate was attempted to be synthesized using triphosgene (Scheme 10).

(22)

22 However, no conversion towards the chloroformate was observed. It was not clear what had happened, as the only difference in the 1H-NMR spectrum was a shift in ppm of all the signals. IR and 13C-NMR were also not conclusive about what had happened.

2.3.3 Coupling t-Bumeoc to an amine: Literature approach

In literature the tertiary alcohol of t-Bumeoc is coupled to an amine using the fluoroformate (Figure 17).

Figure 17: Fluoroformate of t-Bumeoc.

Theoretically it is expected that the fluoroformate is more stable than the chloroformate due to electronic effects. In the article where t-Bumeoc was synthesized and used it was reported that the fluoroformate is only stable at -18°C, which would explain why no conversion could be seen when attempting to synthesize the chloroformate, as it likely decomposes instantly upon formation at room temperature. The fluoroformate is synthesized using carbonic chloride fluoride which was formed from reacting anhydrous sulphuric acid with trichlorofluoromethane (Figure 18).16

Figure 18: Synthesis of the fluoroformate from the tertiary alcohol using carbonic chloride fluoride.

Due to the highly hazardous reagents and the practically difficult procedure it was chosen not to follow this synthetic route.

(23)

23

2.3.4 Coupling t-Bumeoc to an amine: The use of CDI

Another method which was attempted to form an imidazole carbamate with the use of 1,1-carbonyldiimadazole (CDI), as it has been shown in literature that CDI can be used to couple tertiary alcohols to amines.17 The reaction was followed with TLC which indicated full conversion to the elimination product 19-1 (Scheme 11).

Scheme 11: Synthesis of the CDI activated ester and the following elimination. After elimination of CO2 the highly stabilised carbocation of t-Bumeoc is formed, after which a hydrogen is abstracted by

the imidazole anion to yield the alkene elimination product.

3 Conclusion

The goal of this project was to find highly acid labile amino-oxy PG for the use in orthogonal CLIPS and oxime ligation. Two amino-oxy protection groups were attempted to be synthesized with these applications in mind: Mtt and t-Bumeoc. The Mtt protected amino-oxy was synthesized, however purification of the product was so far unsuccessful due to similar polarities of the by-products and product, alongside difficulties with product degradation on the acidic silica.

The tertiary alcohol of t-Bumeoc was synthesized. However, coupling of the PG to an amino-oxy using a chloroformate, or by using CDI was unsuccessful. The fluoroformate was used in literature to couple the tertiary alcohol to an amine, though this synthesis deals with highly hazardous reagents and was therefore not attempted.

(24)

24

4 Future prospects

Using an Mtt PG for orthogonal CLIPS and oxime ligation reactions might still be an attractive option despite difficulties with isolation of the product. By limiting the amount of by-products formed by optimizing the coupling reaction separation might become more straightforward. It is envisioned that the main reaction condition which should be changed is the amount of MttCl added. The products which were attempted to isolate were all synthesized with 10 equiv MttCl. A coupling reaction with 2.1 equiv of MttCl can be attempted to analyse the conversion to the desired protected amino-oxy. Although t-Bumeoc turned out not to be a usable protection group other carbamates could potentially still be used due to them being more robust towards acidic degradation compared to the Mtt PG. For example 1-(1-Adamantyl)-1-methylethoxycarbonyl carbamate (Adpoc) is another carbamate PG which is highly acid labile and could serve at a potential alternative to Mtt (Figure 19).15

Figure 19: Adpoc PG coupled to a generic amino-oxy R-ONH2.

The adamantane carboxylic acid can be purchased commercially after which it is chlorinated and reacted with MeMgI to obtain the tertiary alcohol.18 The coupling of the tertiary alcohol to an amine is once again performed in literature using the carbonic chloride fluoride reagent.19 If it is decided that Adpoc is to be researched an alternative to the fluoroformate should first be looked for.

(25)

25

5 Acknowledgements

Firstly I would like to thank Prof. Dr. H. Hiemstra, Prof. Dr. J. H. van Maarseveen and Dr. S. Ingemann Jorgensen for allowing me to do this project in the Synthetic Organic Chemistry group. Additionally I would like to thank Prof. Dr. A. M. Brouwer for being the second reviewer of this project. I would also like to express my deepest gratitude and appreciation to my daily supervisor Dieuwertje Streefkerk. She has taken a vast amount of time out of her busy schedule to guide me through any difficulties I may have had. Not only has she helped me tremendously in the lab by teaching me how to perform experimental work both safely and efficiently, but she was also more than willing to guide me with the work outside the lab. I have learned a great deal during this project because of her willingness to actively help me with anything I may have needed help with.

Lastly I would like to thank everyone in the Synthetic Organic Chemistry group for creating a wonderful atmosphere which made the long days in the lab more enjoyable.

(26)

26

6 References

1: Waldmann, T. A. Nat. Med. 2003, 9 (3), 269–277.

2: Werkhoven, P. R.; Elwakiel, M.; Meuleman, T. J.; Quarles, H. C.; Quarles van Ufford, H. C.; Kruijtzer, J. A. W.; Liskamp, R. M. J. Org. Biomol. Chem., 2016, 14, 701–710.

3: Scott, J. K.; Smith, G. P. Science 1990, 249, 386–390.

4: Barlow, D. J.; Edwards, M. S.; Thornton, J. M. Nature 1986, 322, 747–748.

5: Timmerman, P.; Beld, J.; Puijk, W. C.; Meloen, R. H. ChemBioChem 2005, 6, 821–824. 6: Tang, L.; Yin, Q.; Xu, Y.; Zhou, Q.; Cai, K.; Yen, J.; Dobrucki, L. W.; Cheng, J. Chem. Sci.

2015, 6, 2182–2186.

7: Angell, Y. L.; Burgess, K. Chem. Soc. Rev. 2007, 36, 1674–1689.

8: Zeng, W.; Ghosh, S.; Macris, M.; Pagnon, J.; Jackson, D. C. Vaccine 2001, 19, 3843–3852. 9: Timmerman, P.; Puijk, W. C.; Meloen, R. H. J. Mol. Recognit. 2007, 20, 283–299.

10: Smeenk, L. E. J.; Dailly, N.; Hiemstra, H.; van Maarseveen, J. H.; Timmerman, P. Org. Lett.

2012, 14 (5), 1194–1197.

11: Timmerman, P.; Puijk, W. C.; Boshuizen, R. S.; van Dijken, P.; Slootstra, J. W.; Beurskens, F. J.; Parren, P. W. H. I.; Huber, A.; Bachmann, M. F.; Meloen, R. H. Open Vaccine J. 2009, 2, 56–67. 12: Heinis, C.; Rutherford, T.; Freund, S.; Winter, G. Nat. Chem. Biol. 2009, 5 (7), 502–507. 13: Dirksen, A.; Hackeng, T. M.; Dawson, P. E. Angew. Chem. Int. Ed. 2006, 45, 7581–7584. 14: Liu, F.; Thomas, J.; Burke Jr., T. R. Synthesis 2008, 15, 2432–2438.

15: Wuts, P. G. M. Greene’s Protective Groups In Organic Synthesis, 5th ed.; Wiley: Michigan, USA. 16: Voelter, W.; Müller, J. Liebigs Ann. Chem., 1983, 248–260.

17: Stoddart, A.; Feast, W. J.; Rannard, S. P. Soft Matter 2012, 8, 1096–1108.

18: Kolocouris, A.; Busath, D. D.; Johnson, B. Antiviral compounds for treating amantadine resistance influenza A. PCT Int. Appl. 2014121170, 2014.

19: Kalbacher, H.; Voelter, W. J. Chem. Soc., Chem. Commun. 1980, 72, 1265–1266. 20: Voelter, W.; Müller, J. Liebigs Ann. Chem. 1983, 248–260.

(27)

27

7 Experimental

7.1 General remarks

Reagents which were purchased were used as supplied. Unless stated otherwise, reactions were performed at room temperature without any special precautions such as drying or N2 atmosphere.

Dried solvents were obtained by distillation with sodium. Reactions were followed with thin layer chromatography (TLC) carried out on 0.25 mm E. Merck silica gel plates (60F-254). TLC analysis of acid labile compounds such as the Mtt protected amino-oxy were performed by first hanging the TLC plate in a NEt3 vapour rich atmosphere for several seconds and adding 1-10% of NEt3 to the liquid

phase. SilaFlash® P60 (particle size 40-63μm) was used for column chromatography. NMR spectra were recorded on Bruker DRX-400, and 300 MHz instruments and calibrated on residual undeuterated solvent signals as internal standard. IR spectra were recorded on a Bruker Alpha FTIR.

benzyl bis(2-hydroxyethyl)carbamate

Diethanolamine (4.8 mL, 50 mmol, 1 equiv), 1,4-dioxane (85 mL) and sat. NaHCO3 (aq., 100 mL)

were mixed and cooled with an ice bath. To the white suspension a solution of Cbz-OSu (13.08 g, 52.5 mmol, 1.05 equiv) dissolved in acetone (60 mL) was added and the flask was stirred at rt overnight. The volatiles were removed until ca. 100 mL solvent remained. The solution was extracted three times with ethyl acetate, which was subsequently washed six times with KHSO4 (aq., 1M), and brine. After

drying over MgSO4, the volatiles were removed under reduced pressure yielding the colourless oil

(9.11 g, 38.1 mmol, 76.2%). 1H NMR (400 MHz, Chloroform-d) δ 7.39 – 7.31 (m, 5H), 5.15 (s, 2H), 3.96 (s, 2H), 3.84 (s, 2H), 3.77 (s, 2H), 3.57 – 3.42 (m, 4H)

benzyl bis(2-((1,3-dioxoisoindolin-2-yl)oxy)ethyl)carbamate

Cbz-diethanolamine (5.73 g, 23.97 mmol, 1.0 equiv), PPh3 (13.19 g, 50.34 mmol, 2.1 equiv) and

N-hydroxyphthalimide (8.19 g, 50.34 mmol, 2.1 equiv) were dissolved in distilled THF (370 mL) in a flame dried flask.

(28)

28 At 0°C DIAD (9.86 mL, 50.34 mmol, 2.1 equiv) was added dropwise and left to stir for 45 minutes after which the reaction mixture was allowed to warm to rt. The reaction mixture was left to stir overnight after which the volatiles were removed under reduced atmosphere and the yellow oil crude product was purified by column chromatography (silica powder, 2:1 PE: EtOAc) to yield an off white solid (6.62 g, 12.50 mmol, 53%). 1H NMR (400 MHz, Chloroform-d) δ 7.77 (m, J = 19.3, 5.5, 2.9 Hz, 8H), 7.42 – 7.23 (m, 5H), 5.18 (s, 2H), 4.45 (dt, J = 11.9, 5.4 Hz, 4H), 3.92 (dt, J = 16.3, 5.5 Hz, 4H);

13

C NMR (101 MHz, Chloroform-d) δ 163.22, 155.89, 136.28, 134.34, 128.75, 128.36, 127.85, 127.69, 123.38, 77.28, 76.96, 76.64, 67.35, 47.73, 47.14; IR (cm-1): 1733, 903, 878, 723, 700, 648.

benzyl bis(2-(aminooxy)ethyl)carbamate

Acetone derivate oxime 4-1 (3.74 g, 13.9 mmol, 1.0 equiv) was suspended in distilled THF (80 mL), after which methoxyamine (25-35 wt% HCl salt) (21 mL, 69.12 mmol, 4.97 equiv) was added. The solution was left to stir at r.t. for 48h. The organic layer was extracted with strongly acidic water. The water layer was subsequently made strongly basic with NaHCO3. The water layer was then extracted

with EtOAc. The organic layer was dried with MgSO4 and the volatiles were removed to yield the

product as a colourless oil (310 mg, 1.15 mmol, 8.29%). 1H NMR (400 MHz, Chloroform-d) δ 7.38 – 7.27 (m, 6H), 5.14 (s, 2H), 3.76 (dt, J = 20.0, 5.2 Hz, 4H), 3.54 (dt, J = 25.8, 5.4 Hz, 4H).

benzyl bis(2-(((diphenyl(p-tolyl)methyl)amino)oxy)ethyl)carbamate

In a flame dried flask phth protected compound 3 (1.50 g, 2.84 mmol, 1 equiv) was dissolved in distilled THF (45 mL) after which methylhydrazine (0.18 mL, 3.41 mmol, 1.2 equiv) was added. After 2 hours of stirring a white solid had formed and DIPEA (4.9 mL, 14.2 mmol, 5.0 equiv) and MttCl (4.15 g, 28.4 mmol, 10 equiv) were added and the reaction mixture was left to stir overnight.

The reaction mixture was washed twice with H2O and brine. After drying with K2SO4 the volatiles

were removed under reduced atmosphere to yield a yellow oil crude product which was attempted to purify by column chromatography (silica powder pre-treated with 10% NEt3, 6:1 PE:EtOAc).

(29)

29 No pure product was obtained, however NMR indicated product formation. 1H NMR (300 MHz, Chloroform-d) δ 7.47 – 6.98 (m, 116H), 5.06 (d, J = 5.8 Hz, 2H), 3.69 – 3.58 (m, 4H), 3.30 – 3.19 (m, 2H), 3.19 – 3.08 (m, 2H), 2.45 – 2.29 (m, 23H).

tert-butyl 3,5-dimethylbenzoate

3,5-dimethyl benzoic acid (7.25g, 48.25 mmol, 1.00 equiv) was suspended in toluene (6 mL) and thionyl chloride (7.23 mL, 77.20 mmol, 1.60 equiv) was added and the solution was stirred at reflux for 4 hours. The volatiles were removed and t-BuOH (9.3 mL, 96.5 mmol, 2.00 equiv) and pyridine (4.12 mL, 51.1 mmol, 1.1 equiv) were added. After 100 hours the solution was washed with HCl (aq., 4M), water, NaOH (aq., 2M) and water. After drying with K2CO3 the volitiles were removed to

yield a yellow oil (9.06 g, 43.92 mmol, 91%). 1H NMR (400 MHz, Chloroform-d) δ 7.65 (s, 2H), 7.16 (s, 1H), 2.38 (s, 6H), 1.63 (s, 9H).

tert-butyl 3,5-bis(bromomethyl)benzoate

To a flame dried flask under N2 atmosphere tBu ester 8 (8.90 g, 43.20 mmol, 1.00 equiv) was

dissolved in DCM (200 mL) and NBS (16.50 g, 90.76 mmol, 2.1 equiv) was added. The yellow suspension was irradiated with a 500W lamp to reflux and the reaction was monitored using 1H-NMR. After 1.5h it was determined the reaction was complete and the mixture was diluted with DCM and washed with water. After drying with Na2SO4 the volatiles were removed to yield a yellow oil (5.82 g,

15.98 mmol, 31%). 1H NMR (400 MHz, Chloroform-d) δ 7.93 (s, 2H), 7.60 (s, 1H), 4.50 (s, 4H), 1.61 (s, 9H).

(30)

30

3,5-bis(bromomethyl)benzoic acid

t-Bu ester 9 (610 mg,1.68 mol, 1.00 equiv) was dissolved in DCM (6 mL) and formic acid (6 mL) was added. The solution was stirred overnight after which the volatiles were removed to yield the product as an off white solid (516 mg, 1.67 mmol, 99%). 1H NMR (400 MHz, Chloroform-d) δ 8.14 – 8.07 (s, 2H), 7.71 (s, 1H), 4.54 (s, 4H).

1,3-di-tert-butyl-5-methylbenzene

t-BuCl (143 mL, 1.3 mol, 2.6 equiv) and toluene (53 mL, 0.5 mol, 1.0 equiv) were added to a flame dried 3-neck flask fitted with an oil bubbler. AlCl3 (6.07 g, 0.046 mol, 0.09 equiv) was added in small

batches and the reaction was left to stir overnight. Water was added to quench the reaction and the product was extracted in Et2O. The crude mixture was distilled at reduced pressure to yield the product

as a colourless liquid (37.3 g, 0.18 mol, 37%). 1H NMR (300 MHz, Chloroform-d) δ 7.30 (s, 1H), 7.10 (s, 2H), 2.42 (s, 3H), 1.38 (s, 18H).

3,5-di-tert-butylbenzoic acid

In a three neck flask fitted with a reflux condenser 14 (35.99 g, 0.176 mol, 1.0 equiv) was added to a solution of KOH (15.8 g, 0.282 mol, 1.6 equiv), pyridine (130 mL) and H2O (35 mL) and the solution

was heated to 95°C under vigorous mechanical stirring. KMnO4 (70.0 g, 0.443 mol, 2.52 equiv) was

added in portions and the suspension was left to stir overnight. The reaction mixture was then filtrated and washed with NaOH (aq., 1M). After removal of ca. 200 mL of volatiles the mixture was made strongly acidic with HCl (aq., 6M) and EtOAc (100 mL) was added.

(31)

31 The organic phase was separated and the water phase was washed with EtOAc. The organic phases were combined and washed with water until neutral. After drying over Na2SO4 the volatiles were

removed to yield a white solid which was washed with EtOH to yield the product 15 as a white crystal (8.36 g, 0.00357 mol, 20%). 1H NMR (300 MHz, Chloroform-d) δ 8.02 (m, 2H), 7.72 (m, 1H), 1.41 (s, 18H).

methyl 3,5-di-tert-butylbenzoate

Benzoic acid 15 (8.36 g, 35.67 mmol, 1.0 equiv) was suspended in MeOH (17 mL) and SOCl2 (1.5

mL, 20.56 mmol, 0.58 equiv) was added and the solution was heated to reflux overnight. The reaction mixture was diluted with Et2O and washed with H2O, Na2CO3 and H2O. After drying with MgSO4 the

volatiles were removed to yield 16 as white crystals (8.16 g, 32.86 mmol, 92%). 1H NMR (300 MHz, Chloroform-d) δ 7.89 (s, 2H), 7.62 (s, 1H), 3.91 (s, 3H), 1.35 (s, 18H).

2-(3,5-di-tert-butylphenyl)propan-2-ol

In a flame dried flask methyl-ester 16 (7.19 g, 28.95 mmol, 1.0 equiv) in distilled Et2O (25 mL) was

slowly added to MeMgBr (Et2O, 3M) (29.0 mL, 86.85 mmol, 3.0 equiv). The solution was then heated

to 35°C overnight, after which the solution was poured over ice and made strongly acidic with KHSO4

1M. The organic layer was separated and the water layer was washed with Et2O. After combining the

organic layers they were washed with Na2CO3 and H2O and dried with MgSO4. The volatiles were

removed to yield the tertiary alcohol 17 as a yellow oil (6.83 g, 27.49, 95%). 1H NMR (400 MHz, Chloroform-d) δ 7.37 (m, 2H), 7.35 (m, 1H), 1.62 (s, 6H), 1.36 (s, 18H).

(32)

32

2-(3,5-di-tert-butylphenyl)propan-2-yl carbonochloridate

At 0°C pyridine (0.78 mL, 9.66 mmol, 1.2 equiv) and triphosgene (1.00 g, 3.36 mmol, 0.4 equiv) were added to a solution of 17 (2.09 g, 8.40 mmol, 1.0 equiv) in DCM (50 mL). The solution was allowed to come up to r.t. and was left overnight. IR showed no product was formed. 1H NMR (300 MHz, Chloroform-d) δ 7.54 – 7.30 (m, 3H), 2.04 (s, 6H), 1.37 (s, 18H).

2-(3,5-di-tert-butylphenyl)propan-2-yl 1H-imidazole-1-carboxylate

To a solution of 17 (200 mg, 0.80 mmol, 1.0 equiv) in toluene (2 mL) was added CDI (182 mg, 1.14 mmol, 1.43 equiv) and the suspension was heated to 60°C overnight. Full conversion to the elimination product 19-1 was observed. 1H NMR (400 MHz, Chloroform-d) δ 7.48 – 7.17 (m, 62H), 5.37 (s, 1H), 5.11 (s, 1H), 2.22 (s, 3H), 1.39 (s, 18H).

(33)

33

8 Appendix

Figure 20: 1H-NMR spectrum of compound 2.

(34)

34 Figure 22: 13C-NMR spectrum of compound 3.

(35)

35 Figure 24: 1H-NMR spectrum of compound 4.

(36)

36 Figure 26: 1H-NMR spectrum of compound 5.

(37)

37 Figure 28: 1H-NMR spectrum of compound 9.

(38)

38 Figure 30: 1H-NMR spectrum of compound 14.

(39)

39 Figure 32: 1H-NMR spectrum of compound 16.

(40)

40 Figure 34: 1H-NMR spectrum of attempted synthesis of compound 18.

Referenties

GERELATEERDE DOCUMENTEN

5/20/2015 Welcome

Ten eerste zou het kunnen zijn dat er wel degelijk sprake is van een significant effect van politieke onvrede op zowel de links- als de rechts- populistische partij, maar dat de

Daarna is nog een worstcase scenario berekend, waarbij een keer alle dijkvakken falen langs het Volkerak-Zoommeer (maximale waterstand NAP + 2,75 m) en een keer dat alle

Naar aanleiding van de verschillen die er bestaan in emotieregulatie in de klinische en niet- klinische groepen van sociale angst en autisme, wordt verwacht dat sociaal

La levée a été édifiée au moyen d'une terre argileuse rouge, rapportée, contenant quelques petits cailloux de schiste rouge et parfois mêlée à un peu de terre noiratre et de

De Eerste Wereldoorlog zorgde echter voor een malaise, onder meer door een gebrek aan grondstoff en, maar na de oorlog kende het bedrijf dankzij investeringen vanuit de

While we can think of a parametric vocoder as a form of vector quantisation since the vector of possible input segments is assigned to a single integer index - if we regard e

wijkgebondenheid groot. Dit zijn dan ook de groepen die weinig buiten de stad komen voor hun openluchtrecreatie.. Onder de bevolkingscategorieën, die op basis van de enquête zijn