MSc in Chemistry
Molecular design, synthesis and catalysis
Master Thesis
TOWARDS THE SYNTHESIS OF NEW HIGHLY
FUNCTIONALIZED MOLECULAR SCAFFOLDS FOR
PEPTIDE CYCLIZATIONS
by
Maria Luisa Corrado
10524835
July 2015 36 + 6 EC
December 2014-July 2015
Examiners: Daily supervisor:
Prof. dr. J. H. van Maarseveen G.J.J.Richelle, MSc
Dr. E. Ruijter
Synthetic Organic Chemistry group
2
Index
List of abbreviations 3
Abstract 5
1. General Introduction: mimicry of protein binding surfaces 6
1.1 Linear versus cyclic peptides 7
1.2 CLIPS™ technology 9
1.3 Aim of the research project 12
2. Results and discussion 15
2.1 Negishi cross-coupling 15
2.2 C-C bond formation via organolithium reagent 19
2.3 C-C bond formation via Gilman reagent 24
2.4 Towards the synthesis of an alternative scaffold 25
3. Conclusions 31
4. Future prospects 33
5. Experimental section 34
6. References 41
3
List of abbreviations
AcOH acetic acid
Bu4NF Tetra-n-butylammonium fluoride
CCl4 carbon tetrachloride
CDCl3 deuterated chloroform
CLIPS™ Chemical Linkage of Peptides onto Scaffolds
13C-NMR Carbon-13 Nuclear Magnetic Resonance
CuAAC copper-catalyzed azide-alkyne cycloaddition CuI copper(I) iodide
DBE 1,2-dibromoethane DBP dibenzoyl peroxide
DCM dichloromethane DMA dimethylacetamide
EAS electrophilic aromatic substitution Et2O diethyl ether
HBr hydrobromic acid
HCl hydrochloric acid
1H-NMR proton nuclear magnetic resonance
H2SO4 sulfuric acid
I2 iodine
IC50 half-maximal inhibitory concentration
4
KI potassium iodide
KMnO4 potassium permanganate
LiBr lithium bromide
MeCN acetonitrile MeOH methanol
MgSO4 magnesium sulfate
Mp melting point
NaOAc sodium acetate
n-BuLi n-butyl lithium
OAc acetate
Pd(dppf)Cl2 [1,1′-Bis(diphenylphosphino)ferrocene]dichloropalladium(II)
PE petroleum ether
OPiv pivalate PivOH pivalic acid
PPIs protein-protein interactions Rf retention factor
TBMB 1,3,5-tris(bromomethyl)benzene TMED N,N,N’,N’-tetramethylethylene-diamine THF tetrahydrofuran
TLC thin layer chromatography
TMS trimethylsilyl
5
Abstract
The high potential of cyclic peptides as therapeutics has revolutionized the design of new drug molecules. They have already shown to be good candidates as receptor agonists and now scientists are also trying to develop receptor antagonists or complex structures that can interfere with damaging protein-protein interactions by mimicking protein binding surface with cyclic peptides1. Among several cyclization
strategies, scaffold-assisted peptide cyclization has shown to be a robust approach to facilitate the constraining of linear peptide precursors; furthermore, multicyclic peptides can also be achieved via this method. Studies demonstrate that higher binding interactions can derive from the presence of multiple sites where such interactions can take place 2. Just as in nature, where binding sites on protein surface
are usually present as multiple discontinuous regions resembling the shape of multicyclic peptides. The final scope of the project is the synthesis of a highly functionalized benzene-based scaffold bearing orthogonal reactive moieties for CLIPS™/CLICK reactions, bromomethyl-groups and alkyne functionalities respectively, to generate up to five peptide loops that can eventually lead to multicyclic peptides able to mimic protein binding sites. Different routes have been explored towards the synthesis of T6-1 and T6-2 scaffolds: mainly Negishi cross-coupling and SN2 substitution with both in situ formed
organolithium compound and propargylic alkoxide for the formation of C-C bonds and free radical bromination to install bromomethyl-groups. Unfortunately, the routes that were investigated did not lead to the desired scaffolds; however, several new approaches like Suzuki cross-coupling will be tested.
6
1. General introduction: mimicry of protein binding surfaces
Protein-protein interactions (PPIs) regulate most of the vital functions in the human body. A significant example is the immune system that protects from exogenous proteins, also called antigens, by producing complementary proteins, antibodies, which are able to recognize and capture antigens via epitope-paratope interactions 3. As malfunction in vital processes like the immune response almost always leads
to cell death, scientists all over the world are looking for suitable drug molecules which assure that the activity of foreign proteins is inhibited4. Indeed, the understanding of such interactions has become an
attractive and important research field for drug discovery. In general, PPIs are established via non-covalent bonds, such as hydrogen-bonds, electrostatic-and van der Waals interactions, on specific sites of the protein surface. These can be present as continuous or discontinuous regions; however, the latter are more commonly encountered in nature (Scheme 1)5. The mimicry of such binding surfaces and the
manipulation of such protein-protein interactions may lead to the design of new pharmaceutical active compounds1. A large amount of small molecules were developed for such purposes ,however, molecules
that possess more similar structures to those of proteins may result in more selective binding ability 6,7.
Scheme 1: topology of protein binding sites, continuous and discontinuous respectively
Peptides, as well as proteins, are composed of amino acids residues; peptides are indeed small proteins and both of them have specific functions in the human body. Nevertheless, peptides are much more easily accessible in synthesis compared to proteins due to their smaller and less complex structures. Indeed, peptide libraries can be generated and eventually screened for biological activity. Therefore, scientist worldwide are investigating the design of new drug molecules by taking inspiration from nature. Mimicking the binding sites of proteins with peptide-based ‘’small molecules” continues to be a key challenge in protein mimics5,8. One way to achieve such goal could be to develop complex structures that
7
1.1 Linear versus cyclic peptides
Linear peptides are barely considerate good drugs candidates due to the low metabolic stability and high flexibility. On the other hand, cyclic peptides display higher enzymatic stability and lower sensitivity towards proteolytic degradation because of their more rigid structures and lack of free C-and N-termini. The fixed conformation allows to the amino acids side chains to be settled in well-defined positions in space leading to improved pharmacodynamics and kinetic properties 9. Furthermore, cyclic peptides will
exhibit higher affinity for a specific molecular target because of the less entropy upon interaction compared to their linear conformation10,11. Many natural cyclic peptides showing biological activity were
identified, such as vancomycin12, a powerful glycopeptide antibiotic used for treatments of bacterial
infections and cyclosporin13, which is currently used as immune-suppressor in organ transplants to avoid
body rejection of the implanted organ (Figure 1). Most peptide-based drugs derive from natural peptides, however, analogues or surrogates have also shown to possess pharmaceutical properties14.
.
Figure 1: examples of cyclic peptides available as drugs: vancomycin and cyclosporin respectively.
Nature cyclizes peptides with the use of enzymes, which facilitate ring closure by recognizing the N-and C-termini of a linear peptide precursor and bringing the reactive moieties in close proximity. Nevertheless, cyclization of peptides can also be achieved by synthetic approaches 15. The most common way is via
macrolactamization, where the free C-terminal carboxylic acid and N-terminal amine are reacted to form a lactam thus resembling a natural peptide bond. However, this strategy presents some drawbacks not only because of the side reactions that can occur, such as dimerization and isomerization, but also because macrolactamization depends on the ring size that has to be formed 16. Moreover, epimerization
8 at C-terminus is a major disadvantage; this is due to the necessary activation of the carboxylic acid in order to undergo cyclization17, while enzymes facilitate the ring closure by allowing the reactive termini to
be close to each other and thus overcoming the entropically unfavoured interaction that is usually synthetically difficult to achieve. As depicted in Scheme 2, peptides can cyclize in four main ways according to the position of the reactive moieties on the linear sequence: head-to-tail, head-to-side chain, side chain-to-side chain, side chain-to-tail 15.
Scheme 2: four main ways for peptide cyclization
Macrolactamization is not the only way to achieve peptide cyclization. Indeed, the N-and C-termini of a linear peptide can be capped with protective groups and other reactive moieties can be installed on the amino acid side chains. Many ligation strategies have been developed in the last decades for facilitating peptide cyclization such as oxime ligation, CLICK chemistry, native chemical ligation 15. These reactions
have first been reported as strategies for intermolecular peptide ligations. Nevertheless, if the reactive functionalities are present on the same linear peptide precursors then the ligations can also be performed in an intramolecular fashion. The main advantages of most of these ligation strategies is that they can usually be conducted under mild conditions and their chemo selectivity allows to combine such reactions without the need of any protective groups on the amino acid side chains.
Of high relevance for scientists worldwide is that protein-binding regions display loop-like structures resembling the 3D conformations of peptides. These similarities and the advantages of peptide synthesis could allow the discovery of new drug molecules not only by synthesizing cyclic peptides for the direct use as receptor agonists but also by constructing multiple peptide domains in order to mimic protein binding sites and eventually act as therapeutics (Scheme 3)18,19. This would potentially lead to higher
9 molecules already available as drugs 20. Indeed, the surface involved in protein-protein interactions is
quite large, usually in the range of 750-1500 Å221.Scaffold-assisted peptide cyclization has revealed to
be an important tool to achieve multi-cyclic peptides showing high potential for the mimicry of proteins binding surface.
Scheme 3: schematic protein binding surface and PPIs
1.2 CLIPS™ technology
An alternative way to synthesize cyclic peptides is via scaffold-assisted peptide cyclization. Recently, CLIPS™ technology, which was developed by Timmerman and co-workers in 2005, has shown to be a robust and straightforward approach for the cyclization of peptides22. The technology is based on the
chemical linkage of a linear peptide precursor onto a molecular scaffold via nucleophilic attack of thiol groups present on the linear peptide precursor, usually as cysteine residues, to activated bromides present on the molecular scaffold. CLIPS™ technology provides not only a way to fix the conformation of a linear unprotected peptide but it also contributes to the preparation of polycyclic peptide structures with applications in peptidomimetics. One of the first reported benzene-based scaffold is depicted in Scheme 4, which bears two bromomethyl-groups leading to the formation of bicyclic peptides.
10
Scheme 4: CLIPS™ technology
CLIPS™ technology has shown to be a reliable and fast approach for rigidifying the conformation of amino acids sequences. In general, the linear peptide precursor does not require any protection scheme allowing all the possible unprotected side-chain groups to be present on the backbone, except for free cysteine residues that can interfere with the cyclization process. CLIPS™ is a fast process that can be conducted in aqueous solution, at pH around 8; however, an organic solvent, such as MeCN, can also be present up to 80%.The achieved results demonstrate that in this method the size of the linear peptide precursor is not as important as the nature of the synthetic scaffold. Most of the scaffolds employed for CLIPS™ reactions are based on the dibromoxylene-type structure reported by Timmerman and coworkers.
Protein-binding surfaces have usually specific cavities where interactions with complementary proteins occur. With linear peptides is difficult to achieve protein surface mimics because of their high flexibility and lack of tertiary structures. Furthermore, monocyclic peptides can cover only a small portion of a protein-binding surface. On the other hand, the presence of a molecular scaffold that can lead to multiple, rigid peptide loops may allow the mimic of protein-binding sites. Therefore, in order to achieve maximal binding, different cyclic peptide domains need to be properly organized around a core scaffold (Scheme 5) 19,22.
11
Scheme 5: schematic scaffolded-bicyclic peptide for mimicry of protein binding sites
In 2012, Heinis and coworkers reported the synthesis of a potent bicyclic peptide showing high inhibition activity towards human urokinase-type plasminogen activator (uPA)23. The two-looped peptide was
achieved via CLIPS™ technology and 1,3,5-tris(bromomethyl)benzene (TBMB) has been used as template to assist the constrain of the linear peptide sequence (Scheme 6). The reported half maximal inhibitory concentration, IC50, of the bicyclic peptide towards human uPA is in the nanomolar range (IC50=
102 nM); on the other hand, the linear sequence did not show any activity towards uPA indicating that the
fixed conformation is fundamental for successful mimicry of the protein binding sites. Furthermore, the interactions cover a quite large surface of about 701Å2 and consist of more than eight hydrogen bonds
12
Scheme 6: scaffolded-bicyclic peptide showing human uPA inhibition activity
1.3 Aim of the research project
This research project is an extension of the work of Timmerman and the final scope would be to increase the number of chemically linked peptides onto a molecular scaffold up to five loops. The state-of-art of multicyclic peptides via CLIPS™ technology includes the formation of up to bi-cyclic peptideswithout any detection of possible isomers24,25. Indeed, tricyclic peptides can also be achieve via CLIPS™ technology
however, formation of regioisomers is the major drawback. Furthermore, in order to be able to mimic protein-binding sites, which in nature are more often present as multiple discontinuous regions, the presence of multiple loops could lead to stronger interactions towards specific protein targets 5. Therefore,
in order to increase the number of possible loops around a molecular scaffold, avoiding the formation of regioiosomers, CLIPS™ technology needs to be combined with a bio-orthogonal ligation strategy. CLICK chemistry, and in this particular case copper-catalyzed alkyne-azide cycloaddition (CuAAC), shows high potentiality to be combined with CLIPS™. The hypothesis is that these reactions could be carried in sequence in a regioselective fashion without any interference with each other and the need of any protective groups on the amino acid side chains.
13 The idea behind the final scope of this project is depicted in Scheme 7. In order to accomplish the formation of five loops, CLIPS™ should be run first, leading to bicyclic peptides (middle structure), followed by CuAAC which would then lead to the formation of up to pentacyclic peptides.
Scheme 7: design for the formation of pentacyclic peptides via CLIPS™/CLICK reactions
The focus of this project has been first on the synthesis of a suitable highly functionalized molecular scaffold, which would then lead to multicyclic peptide structures. The goal is to synthesize a symmetrical benzene-based scaffold bearing the reactive functionalities needed for CLIPS™ and CuAAC reactions, respectively bromomethyl-groups and alkyne moieties. A series of T6 scaffolds have been already synthesized in our laboratory and they will be tested for CLIPS™/CLICK cyclization to form multicyclic peptides. One of these scaffolds is depicted in Figure 2.
14 Even though T6 scaffold already possesses the functionalities needed for CLIPS™ and CuAAC reactions, the hypothesis is that the formation of pentacyclic peptides may be limited because the side chains bearing the alkyne functionalities are quite rigid. Furthermore, the alkyne moieties are directly conjugated with the benzene ring and this may lower their reactivity. Therefore, the aim of the research project is to synthesize a highly functionalized benzene-based scaffold bearing longer alkynyl chains, two carbon atoms more, in order to have a more flexible scaffold for preventing steric hindrance upon cyclization and the loss of conjugation between the alkyne moieties and the benzene ring.
Figure 3: design of the desired highly functionalized benzene-based scaffold
As shown in Figure 3, the desired scaffold needs to contain three activated bromides for CLIPS™ and three alkyne moieties for CuAAC. The mutual functionalities for closing the rings will be present on the amino acid sequence of the peptide precursor: azides will be present as homo-azidoalanine residues, a modified amino acid, while nucleophilic thiols will derive from cysteine residues. The complexity in such synthesis may be encountered in the installation of bromomethyl-groups which are very reactive and at the same time maintaining the benzylic positions of the other side chains on the benzene ring.
After this general introduction about the topic and the final goal of the research project, the investigated routes towards the synthesis of the highly functionalized benzene-based scaffold will be described in details in chapter two while chapter three will deal with the conclusions of the research project. Some other routes that could be investigated for the synthesis of the desired scaffold are described in chapter four. Finally, the experimental section is reported in chapter five.
15
2. Results and discussion
2.1 Negishi cross-coupling
The key step in the first designed approach towards the synthesis of T6-1 scaffold was palladium-catalyzed Negishi cross-coupling. This is the reaction of an organohalide with an organozinc compound catalyzed by palladium to give the corresponding coupled product 26.
Scheme 8: retrosynthetic strategy for the synthesis of T6-1 scaffold via Negishi cross coupling
As shown in Scheme 8, T6-1 scaffold should be obtained from Negishi product 5 by transesterification of acetates using K2CO3 in MeOH to yield the primary alcohols. Subsequent reaction with PBr3 would
eventually afford compound T6-1. The building blocks needed for the cross coupling to obtain 5 have been synthetized according to literature procedures.
16
Scheme 9: synthesis of Negishi coupling partner 3
Compound 3 was prepared in two steps starting from commercially available mesitylene 1 which undergoes electrophilic aromatic substitution (EAS) upon treatment with iodine and bis-(trifluoroacetoxy)iodobenzene in CCl4 at room temperature yielding compound 7 in 86% (Scheme 9).
The product was easily collected as white solid by simply filtration and washing with cold CCl4. Compound
7 was then treated with potassium permanganate and acetic anhydride in acetic acid and concentrated sulfuric acid at room temperature in order to convert the methyl groups into benzylic acetates. During reaction with potassium permanganate, the three-methyl groups present on compound 7 are first oxidized to primary alcohols and then converted into acetates because of the presence of acetic anhydride. The product was recrystallized from acetone obtaining compound 3 as white powder in 30% yield which was in agreement with the literature 27.
Once the first Negishi partner has been prepared in acceptable yields and high purity, the synthesis of organozinc 4 was carried out. First the precursor (4-bromobut-1-yn-1-yl) trimethylsilane9 was prepared in two steps starting from 4-(trimethylsilyl)but-3-yn-1-ol 6 (Scheme 10).
Scheme 10: synthesis of (4-bromobut-1-yn-1-yl)trimethylsilane 9
First, the hydroxyl group was transformed into a better leaving group by treatment with methanesulfonyl chloride. The final compound 9 was obtained in good yield by reacting the crude mesylated compound 8 in an SN2 reaction with LiBr in acetone at room temperature 28. Nevertheless, other approaches to get to
compound 9 has been investigated, such as direct bromination of alcohol 6 with CBr4 and PPh3 in DCM29.
17 After the synthesis of 9, the Negishi cross-coupling could be performed starting with the synthesis of the organozinc reagent 4 according to a literature procedure 30. This involves first the activation of zinc dust
with catalytic iodine and then dropwise addition of compound 9 to the activated zinc (Scheme 11). Because organozinc reagents are highly water-sensitive compounds, the formation of 4 was always performed under inert atmosphere. However, fully formation of the desired reagent 4 was not observed. Another route has been explored for the formation of organozinc 4 via zinc activation with droplets of TMS-Cl and DBE followed by drying on vacuum and dropwise addition of 9. However, the one depicted in Scheme 11 seemed to be the most promising approach.
Scheme 11: in situ preparation of organozinc reagent 4
The presence of the organozinc was verified by iodine titration according to a literature procedure reported by Knochel and coworkers31. Upon addition of the organometallic compound, the iodine solution
in THF has to gradually change colour from dark red to orange and eventually to colourless when titration is completed. However, a colourless solution of iodine in THF upon addition of the in situ formed organozinc 4 has never been observed in our case. To make sure that the titration was done properly, a commercially available 1 M solution of diethylzinc in hexane was titrated and indeed a colourless solution was observed. Therefore, it could be concluded that organozinc reagent 4 was not formed.
Therefore, since compound 4 was actually not formed and because zinc can insert easier into a C-I bond (weaker bond than C-Br), it has been decided to synthesis compound 10 for the in situ generation of the organozinc reagent. In this case, the synthesis started from but-3-yn-1-ol and the alkyne functionality was first protected with trimethylsilyl group 32. The mesylate 8 was achieved under the same conditions as for
the synthesis of compound 9 (vide supra). Potassium iodide in MeCN was used in the final step obtaining the desired product in 65% yield. For the formation of organozinc 11 the same literature procedure has been applied as for the generation of 4 (Scheme 12).
18
Scheme 12: synthesis of (4-iodobut-1-yn-1-yl)trimethylsilane 10 followed by in situ organozinc formation
In this case, the presence of organozinc 11 was detected at first attempt by iodine titration obtaining a colourless solution of iodide in THF upon addition of 11. The organozinc reagent was formed as solutions 0.1-0.5 M in THF.
With a procedure in hand for the in situ generation of organozinc 11, the synthesis of compound 5 via Pd-catalyzed Negishi cross-coupling can now be accomplished. However, before testing the cross-coupling on the desired substrate 3 which means three times Negishi coupling on the same molecule, the reaction has first been tested on a model substrate following the same literature procedure where the generation of the organozinc reagent is described30. This procedure involves the coupling of phenyl iodide with
organozinc 11 formed in situ in the presence of Pd(dppf)Cl2 catalyst and copper(I) iodide as co-catalyst.
Therefore, first the organozinc has to be formed in situ (vide supra) and then the coupling reagents are added to the same pot. Because of disambiguates in the literature procedure, both THF and DMA were tested as solvent for the cross-coupling process (Table 1).
19 Entry # Substrate Pd cat. CuI Solvent T °C
1 PhI 5% mol 5 % mol DMA 80-90
3 PhI 5% mol 5% mol THF Reflux
2 PhI 7% mol 7% mol THF 80-90
4 3 2% mol 2% mol THF Reflux
Table 1: screened conditions for Negishi cross-coupling
The authors reported that the reaction has been conducted at 80-90 °C; for this reason, when THF has been tested as solvent the reaction was conducted in sealed tubes in order to be able to reach higher T than THF boiling point. All the coupling reactions have always been performed under nitrogen or argon atmosphere. Furthermore, all the reagents involved in the coupling were pre-treated: THF and DMA were freshly distilled and stored under nitrogen atmosphere; the substrate phenyl iodide was also distilled prior to use; a new amount of palladium catalyst was purchased and the CuI co-catalyst was vacuum-dried. However, even if the coupling reaction was always performed under inert atmosphere and run under optimal conditions, compound 14 has never been observed and the literature procedure was unreproducible.
Therefore, other routes towards the synthesis of scaffold T6-1 were considered as reaction via Negishi cross coupling did not yield the desired product.
2.2 C-C bond formation via organolithium reagent
An alternative approach for introducing the alkyne moiety is by generating the anion of trimethyl(prop-1-yn-1-yl)silane followed by nucleophilic attack on benzylic bromide 33.
Model compound 14 was synthesized via reaction of benzyl bromide 15 with the organolithium reagent formed in situ from trimethyl(prop-1-yn-1-yl)silane and n-BuLi (1.6 M in hexane) which led to the formation of the desired product in quite good yield, 80% (Scheme 13).
20
Scheme 13: model compound for C-C bond formation
Therefore, a new approach towards the desired scaffold has been investigated. It involved, first, SN2
substitution with (3-(trimethylsilyl)prop-2-yn-1-yl)lithium on compound 16 followed by installation of bromo methyl-groups via radical bromination (Scheme 14). It should be stressed that the benzene side chains are methyl groups and terminal protected alkynyl moieties, therefore it might be that radical bromination may lead to a mixture of products.
Scheme 14: alternative route towards T6-1 scaffold
Compound 16 was synthesized starting from mesitylene 1 by treatment with p-formaldehyde, hydrobromic (acid 33% wt. in acetic acid) and acetic acid as solvent heated to 100 °C34. Subsequent
reaction of 16 with(3-(trimethylsilyl)prop-2-yn-1-yl)lithium led to the formation of the desired product 17 in high yield, 87%. The last step of the route consists in the free-radical bromination of compound 17
21 however, the formation of the desired product was not observed. The screened conditions are summarized in Table 2.
Entry # Radical initiator NBS equiv. Solvent T °C Time (h) 1 DBP (0.2 equiv.) 5 CCl4 (0.1 M) reflux 22
2 hν 5 DCM (0.2 M) - 3-4
3 DBP (0.2 equiv.) 3.1 CCl4 (0.1 M) reflux 22
4 hν 3 CCl4 (0.1 M) - 2-3
Table 2: screened conditions for the free-radical bromination step
The radical bromination on compound 17 was first performed with dibenzoylperoxide as radical initiator and NBS in CCl4 (Entry 1, table 2). The reaction led to complete conversion of NBS to succinimide,
disappearance of the two triplet signals coming from the methylene groups on the alkynyl side chains while the TMS and methyl signals were still visible on 1H-NMR spectrum. These results were taken into
account before performing the bromination because of the presence of methylene groups on the side chains bearing the alkyne moieties. Indeed, the presence of methyl signal and the disappearance of signals from methylene groups in the 1H-NMR may indicate the possible bromination of such groups
rather than methyl side chains. It was not possible to investigate further the products because a complex mixture formed with very similar retention factor. The radical bromination was also tested with light instead of DBP as radical initiator with the same amount of NBS but the results did not improve (Entry 2, Table 2). Later on, it has been noted that the free-radical bromination have always been performed with large excess of NBS, therefore, it was then tested with less amount of NBS and dibenzoyl peroxide as radical initiator (Entry 3, Table 2). Under such conditions from 1H-NMR of the crude mixture it was possible to
22 observe the presence of the starting material together with other signals in the expected regions of bromomethyl side chains. It was attempted to separate the mixture by column chromatography in order to be able to confirm at least a small conversion into the desired product. However, it was not possible to obtain a clear 1H-NMR spectrum because the product eluted with similar retention time as the starting
material, so they could not be separated. As last attempt, the radical bromination on 17 was performed with three equivalents of NBS but now light was the radical initiator (Entry 4, Table 2). However, neither these conditions led to the desired product T6-1.
Another route was then investigated. This approach involves first the conversion of bromomethyl-groups into acetates followed by radical bromination on compound 19. Then, substitution of bromides with (3-(trimethylsilyl)prop-2-yn-1-yl)lithium on compound 20 would yield 5 (Scheme 15).The final product could then be achieved by conversion of the acetates into alcohols and eventually into bromides with PBr3 as
described for the first route via Negishi cross coupling.
Scheme 15: alternative approach towards T6-1 scaffold
Conversion of compound 16 into compound 19 was straightforward and the product was obtained in 90% yield. Compound 19 was then subjected to radical bromination and the results are reported in Table 3.
23 Entry # Radical initiator NBS equiv. Solvent Time (h) Yield
1 DBP (0.2 equiv.) 5 CCl4 (0.1 M) 24 -
2 DPB (0.2 equiv.) 5 CCl4 (0.01 M) 24 -
3 hν 5 CCl4 (0.1 M) 1.5 -
4 hν 3.9 DCM (0.1 M) 2 4%
5 hν 5 DCM (0.2 M) 3-4 14-25%
Table 3: results of free-radical bromination on 19
As first attempt, the radical bromination on compound 19 was performed with NBS and dibenzoylperoxide as radical initiator in CCl4, however even after overnight stirring the only formation of mono-and
di-brominated compound was observed (Entry 1, Table 3). The results were comparable when the reaction was performed under higher dilutions (Entry 2, Table 3). In addition, the radical bromination has been performed with light as radical initiator in CCl4; however, the desired product did form neither in this case
(Entry 3, Table 3). It might be that the radical bromination did not work under such conditions because of poor solubility of the starting material in CCl4. For this reason, the reaction was tested in another
solvent. Eventually, the desired product 20 was obtained when radical bromination was performed with light and NBS in DCM even though the yield was low due to the formation of unwanted side products (Entry 4-5, Table 3).
Now, the SN2 substitution with (3-(trimethylsilyl)prop-2-yn-1-yl)lithium on compound 20 is the crucial step
in this approach (scheme 16). When the substitution was carried out, it was observed from in time TLC monitoring the disappearance of the starting material after few minutes and from 1H-NMR it was possible
24
Scheme 16: SN2 reaction with TMS-protected propyne anion
Eventually the desired T6-1 scaffold was not obtained neither via this route and an alternative approach has been investigated.
2.3 C-C bond formation via Gilman reagent
To circumvent the issue with the presence of the strong organolithium base, which can interfere with the acetate groups present on the molecule, an alternative approach has been explored. As transmetallation of the organolithium compound with copper(I) iodide would yield a less basic reagent, it has been decided to proceed via Gilman reagent by in situ formation of lithium dialkyl cuprate 22 following a literature procedure (Scheme 17) 35.
Scheme 17: in situ generation of Gilman reagent
The first step involved the formation of the organolithium intermidiate with n-BuLi in dry diethyl ether leading to a bright yellow solution. After that, the organolithium reagent was transferred via cannula to a suspension of CuI in diethyl ether obtaining an intense orange mixture. The coupling of an organohalide with Gilman reagents is known as Corey–Posner, Whitesides–House reaction in which a lithium dialkyl cuprate reacts with an organo halide, in this case benzyl bromides, to form a new C-C bond, an
25 organocopper compound and a lithium halide 36. Compound 22 was then coupled directly with 20: the
lithium dialkyl cuprate was transferred via cannula under nitrogen atmosphere to a solution of the substrate in dry diethyl ether (Scheme 18).
Scheme 18: Corey–Posner, Whitesides–House reaction
However, this approach was unsuccessful. Compound 20 was decomposed after few minutes of reaction with the Gilman reagent 22 as in the case with organolithium reagent in scheme 17 (vide supra). The transmetallation with copper(I) iodide did not prevent the side reaction with acetate groups present on the molecule. Therefore, it has been decided to move on torwards an alternative scaffold.
2.4 Towards the synthesis of an alternative scaffold
Since the formation of new C-C bonds via the routes that were taken into consideration did not lead to the desired product, the design of T6-1 scaffold has been slightly altered. This includes the introduction of an ether bridge on the alkynyl side chains on the benzene ring. The final product would eventually be obtained in two steps under the same conditions as the approach via Negishi cross-coupling (vide supra). The first investigated route is depicted in Scheme 19.
26 With compound 20 in hand, SN2 substitution was performed with propargylic alkoxide formed in situ. The
formation of the desired compound 23 was unsuccessful even though it should be mentioned that the reaction was run overnight and maybe already after few hours the reaction could have been stopped. An alternative way to synthesize scaffold 24 could be by performing first SN2 with propargylic alkoxide
formed in situ on compound 16 followed by radical bromination on the resulting product (Scheme 20). This approach is similar to the one described in Scheme 14 (vide supra) but now propargylic alcohol is used as nucleophile instead of trimethyl(prop-1-yn-1-yl)silane.
Scheme 20: alternative approach towards the synthesis of T6-2 scaffold
The first step worked very well obtaining compound 25 in 92% yield. As first attempt the radical bromination was performed under the same conditions as for compound 20 (NBS 5 equiv. and light in DCM 0.2 M). After few hour, a peak in the region of interest was observed by in time 1H-NMR and the
reaction was worked up to confirm at least the mono-bromination of the starting material. The crude mixture was then washed with H2O to remove succinimide, however, at this point the 1H-NMR became
very complex. The first hypothesis was that the species shown in Figure 4 may have formed.
Figure 4: possible cationic species formed after H2O washing on brominated compound 25
However, when the radical bromination was performed again under the same conditions the desired product was not observed while the starting material could be recovered in high purity. Therefore, it looked like the radicals formed did not affect the substrate at all. The reaction was monitored by in time 1H-NMR:
27 after six hours, the starting material was always present in the mixture with only a few conversion of NBS into succinimide while after overnight stirring, either with the lamp or without, the crude 1H-NMR became
very complex. The free radical bromination was investigated further trying to find the optimal conditions for such process. For this purpose, a model compound bearing only one propargylic ether bridge and one methyl group on the benzene ring has been synthesized for better following and understanding the bromination reaction. As depicted in Scheme 21, 1-(bromomethyl)-2-methylbenzene 26 was reacted with propargylic alkoxide formed in situ leading to compound 27 in 85% yield.
Scheme 21: model compound to monitor radical bromination
However, when radical bromination was performed under the same conditions as for compound 25 the same results were obtained without any detection of the desired product. Different conditions were screened and these are reported in Table 4.
Entry # Bromine source Solvent T °C Time (h) 1 NBS (1.5 equiv.) DCM (0.2 M) - 2 2 NBS (2.5 equiv.) DCM (0.2 M) - 16 3 NBS (1.6 equiv.) CCl4 (0.1 M) - 4.5 4 NBS (2.0 equiv.) Benzene - 6 5 Br2 (1.5 equiv.) DBE 100 4 6 Br2 (1.5 equiv.) DCM (0.2 M) - 4 7 NBS (2.5 equiv.) DBE (0.2 M) 100 5 8 NBS (2 equiv.) CHCl3 70 4
28 As shown in Table 4, various solvents have been tested because the failure of the radical bromination could derive from low solubility of the substrate in the solvent. However, none of them led to the desired product. Furthermore, bromine source has been varied: mainly NBS has been tested varying the amount of equivalents, but changing these conditions did not lead to the wanted product. In all screened conditions, the starting material could be recovered and only a small portion of NBS was found to be converted into succinimide if the reaction time did not exceed 6 hours. When the reaction was left overnight, either with the lamp still on or only with stirring, from 1H-NMR complex mixture formed and the
starting material was not visible anymore. Moreover, to replace NBS, bromine was instead used (Entry 5,6; Table 4). With high boiling solvents, the reaction mixture was also heated up. However, a complex mixture formed and the formation of the desired product was not observed.
In the meanwhile, it has been investigated the protection of bromomethyl groups present on compound 16 with a more hindered moieties for the synthesis of T6-2 scaffold. A bigger group such as pivalate moieties could maybe prevent possible transesterification during the most crucial step which is the SN2
reaction. The idea would be to proceed via a route similar to the one depicted in Scheme 15 (vide supra) but in this case the SN2 is performed with in situ propargylic alkoxide after free-radical bromination on 29.
The final T6-2 scaffold could then be achieved by converting the pivalate side chains into bromomethyl moieties (Scheme 22).
29 Conversion of compound 16 to 29 by treatment with pivalic acid and n-tetrabutylammonium fluoride eventually led to the desired product in high yield (94%) after overnight stirring at 60-65 °C 37. For the
free-radical bromination on compound 29 various conditions have been tested and are reported in Table 5.
Entry # Radical initiator Bromine source Solvent T °C Time
1 hν NBS (5 equiv.) DCM (0.1 M) - 3h
2 hν NBS (3.2 equiv.) DCM (0.15 M) - 1.5h
3 hν NBS (3 equiv.) DCM (0.1 M) - 2h
4 hν Br2 (3.1 equiv.) DBE (0.1 M) - 1.5h
5 DBP (0.2 equiv.) NBS (3.1 equiv.) CCl4 (0.1 M) Reflux Overnight
Table 5: screened conditions for the free.radical bromination on compound 29
When radical bromination was performed on compound 29 with NBS and light in DCM the results were not very clear (Entry 1, Table 5). From 1H-NMR of the crude mixture it was observed the disappearance
of methyl groups’ signal of the starting material and the appearance of weak peaks in the regionwhere the signal from protons next to bromide would be expected. Furthermore, NBS was completely converted into succinimide and multiple singlets arised next to the methyls signal from pivalate groups indicating possible bromination of such moieties. However, from 1H-NMR it was not possible to identify compound
30. The radical bromination was investigated further by lowering NBS equivalents and varying the radical initiator and the solvent. With 3-3.2 equivalents of NBS and light in DCM a peak integrating four protons arised in the region of interest after half an hour (Entry 2-3, Table 5). However, NBS was completely consumed and also in this case multiple singlets appeared next to t-butyl signal from pivalate groups. Therefore, a milder radical initiator (DBP) was used instead of light with NBS in CCl4 (Entry 5, Table 5).
30 region of interest (4-5 ppm). Furthermore, the signal from methyl groups on the benzene side chains splitted and the same happened for the t-butyl signal from pivalate. This may indicate the possible bromination not only of some methyl groups on the benzene ring but also bromination of pivalate groups. At this point NBS and the starting material were not fully converted and the reaction was stirred overnight. However, the 1H-NMR became complex and the desired product was not observed.
The free-radical bromination step could be investigate further with DBP instead of light as radical initiator in order to clarify if this can lead to the desired compound 30. Furthermore, even if the t-butyl group on pivalate moieties were affected by the radical bromination this would not be a big issue because pivalates will be converted again into bromomethyl-groups in later stages of the synthesis.
31
3. Conclusions
To conclude, the goal of the research project has been the synthesis of highly functionalized benzene-based scaffolds bearing two orthogonal functional groups, bromomethyl-groups and alkynyl moieties, for CLIPS™ and CLICK reactions respectively. The final scope would be to form up to five peptide loops by chemical linkage of a linear peptide precursor onto the synthesized scaffold.
Different routes have been tested towards the synthesis of T6-1 scaffold and they are summarized in Scheme 23. The milder approach via Negishi cross coupling was unsuccessful; however, neither the other investigated routes led to the desired product mainly because of the presence of sensitive side chains on the benzene ring, such as acetate groups in the synthetic approaches via organolithium or lithium dialkyl cuprate reagents and methylene groups for the free-radical bromination.
32 Since the approaches for the synthesis of T6-1 did not lead to the desired product, a slight modification was made to the design of the scaffold by introducing an ether bridge on the alkynyl side chains on the benzene ring. However, the approaches towards the synthesis of the alternative T6-2 scaffold were also unsuccessful as depicted in Scheme 24. As for the synthesis of T6-1, the presence of acetate groups disfavoured the SN2 reaction while the presence of ether bridges did not allow the radical bromination to
occur at all.
Scheme 24: key steps towards the synthesis of T6-2 scaffold
Therefore, none of the two designed scaffolds were successfully synthesized however the last attempt towards the synthesis of T6-2 could be further investigated and in particular the radical bromination step to yield compound 30 (Scheme 25).
Scheme 25: radical bromination step towards the synthesis of T6-2 scaffold
The synthesis of the desired scaffolds revealed to be challenging, however other design could be investigated to lead to a flexible, non-conjugated benzene-based scaffold for CLIPS™/CLICK peptide cyclizations.
33
4. Future prospects
Besides the possible investigation of free-radical bromination on compound 29 (vide supra) another design that could be investigated to get to a more flexible, non-conjugated scaffold than T6 is via Suzuki cross-coupling. This reaction consists in the coupling of a boron compound with organohalides, in this case compound 3, in the presence of a palladium catalyst The final T6-3 scaffold may be obtained by converting acetates into bromomethyl moieites (Scheme 26).
Scheme 26: Suzuki cross-coupling as key step for the synthesis of T6-3 scaffold
The synthesis of the boron compound 32 is reported in literature38 and the main difference with the other
coupling reactions explored so far for the synthesis of T6-1 scaffold is that this boron compound can be easily isolated and further purified. The only difference between T6-1 and T6-3 scaffold is that the alkynyl side chains would be one carbon atom less therefore only one carbon more than T6 (vide supra). However, this could already provide the desired flexibility and the loss of conjugation between alkyne moieties and the benzene ring in order to facilitate the cyclization of linear peptides onto the synthetic scaffold leading, eventually, to pentacyclic peptides.
34
5. Experimental section
General methods
All reactions were carried out in vacuum-dried glassware, unless otherwise stated. Reagents and solvents were purchased from Sigma Aldrich and FluoroChem and used as received. Tetrahydrofuran, dichloromethane, diethyl ether, acetonitrile, dimethylacetamide were freshly distilled prior to use. All reactions involving oxygen- or moisture-sensitive compounds were carried out under nitrogen or argon atmosphere, unless otherwise specified. Flash chromatography purification was performed with SiliaFlash® P60, 40-63µm (230-400mesh). Reactions were followed by in time 1H-NMR or by TLC on
Silica Gel F254. 1H-and 13C-NMR data were recorded at 25-30 °C on Bruker Avance 400 instruments
operating at 400 and 300 MHz in CDCl3 solutions. Chemical shifts are reported in ppm (δ) with respect
to chloroform (CDCl3) as internal standard. Coupling constants are reported as J-values in Hz. IR data
were acquired using a Bruker Alpha-P spectrometer and wavenumbers (ṽ) are reported in cm-1. Melting
points were determine with Polytherm A microscope.
1,3,5-triiodomesitylene (7)
In a 250 ml two-neck round-bottom flask, iodine (21.6 g, 85.3 mmol) was dissolved in carbon tetrachloride (86 mL). Mesitylene (6.8 g, 57 mmol) and [Bis(trifluoroacetoxy)iodo]benzene (40 g, 93 mmol) were then added and the reaction mixture was stirred at room temperature for 4 hours. The precipitate was collected by filtration and washed with cold carbon tetrachloride. The recovered product was dried under high vacuum obtaining a white solid. Yield 24.3 g (86%). Rf 0.5 (DCM); IR (ṽ, cm-1): 1514, 1444, 1374, 1329, 936, 605; 1H-NMR
35 (1,3,5-Triiodo)-2,4,6-triacetoxymethylbenzene (3)
Compound 7 (5.1 g, 10 mmol) was added to a 250 mL round-bottom flask containing glacial acetic acid (51 mL), acetic anhydride (102 mL) and concentrated sulfuric acid (10 mL). Solid potassium permanganate (6.5 g, 40.8 mmol) was added in small portions over a period of 3 hours. After overnight stirring, the solvent was evaporated and water (1 ᵡ 50 mL) was added. The organic layer was extracted with dichloromethane (3 ᵡ 50 mL), washed with brine, dried over MgSO4 and the solvent was removed
under reduced pressure. The solid residue was suspended in acetone and the white powder was collected by filtration. Yield: 2.1 g (30%). Rf 0.3 (PE/EA 9:2); Mp=207-210°C; IR (ṽ, cm-1): 1727, 1515,
1461, 1433, 1347, 1232, 1032, 967, 944, 911, 732, 604; 1H-NMR (CDCl
3, 400 MHz): δ 5.72 (s, 6H), 2.14
(s, 9H); 13C-NMR (CDCl
3, 300 MHz): δ 170.43 (C), 142.44 (C), 108.01 (C), 78.89 (CH2), 20.68 (CH3)
(4-bromobut-1-yn-1-yl)trimethylsilane (9)
To a 50 mL two-neck round-bottom flask, a solution of mesyl chloride (6.2 g, 54 mmol) in dichloromethane (16 mL) was charged. A mixture of 4-trimethylsilyl-3-butyn-1ol (5.1 g, 36 mmol) and trimethylamine (5.5 g, 54 mmol) was added at 0°C. After overnight stirring, the organic layer was extracted with DCM (4 ᵡ 20 mL), washed with brine and dried over MgSO4. The solvent was removed under reduced pressure and a red liquid was obtained. The crude
mesylate (9.04 g, 41.1 mmol) was dissolved in acetone (69 mL) and added to a mixture of LiBr (13.5 g, 156 mmol) in acetone (42 mL). After overnight stirring at room temperature, the organic phase was extracted with DCM (4 ᵡ 50 mL), washed with brine, dried over MgSO4. The solvent was removed under
reduced pressure. The pure product was obtained by distillation under high vacuum (83 °C, 29 mbar) as a colorless liquid. Yield: 5.5 g (74%). Rf 0.3 ( PE/EA 9:1); IR (ṽ, cm-1): 2960, 2898, 2177, 1327, 1249,
1211, 998, 838, 759, 679, 636; 1H-NMR (CDCl
3, 400 MHz): δ 3.43-3.40 (t, J= 7.5 Hz, 2H), 2.78-2.74 (t, J= 7.5Hz, 2H), 0.15 (s, 9H); 13C-NMR (CDCl
3, 300 MHz): δ103.17 (C), 86.93 (C), 29.12 (CH2), 24.29
36 (4-iodobut-1-yn-1-yl)trimethylsilane (10)
To a 250 mL round-bottom flask containing 3-butyn-1-ol (4.04 g, 57.6 mmol) in THF (29 mL), n-BuLi in hexane (1.6 M, 115.3 mmol) was added at ‒78 °C. After 1 hour stirring at this temperature, the reaction mixture was treated with Me3SiCl (13.15 g,
121.0 mmol) and warmed to 25 °C over 1-2 hours. The reaction was quenched with H2O (1 ᵡ 60 mL),
extracted with diethyl ether (3 ᵡ 60 mL) and concentrated under vacuo. The concentrate was treated with HCl 3 N, extracted with diethyl ether (3 ᵡ 50 mL), washed with a saturated solution of NaHCO3 and brine.
The organic phase was dried over MgSO4 and the solvent removed under reduced pressure. Yield: 6.8
g (86%).
To a 50 ml round-bottom flask, a solution of mesylchloride (8.2 g, 72 mmol) in dichloromethane (21 mL) was charged and a mixture of 4-trimethylsilyl-3-butyn-1ol (6.8 g, 48 mmol) and trimethylamine (7.3 g, 72 mmol) was added at 0°C. After overnight stirring, the organic layer was extracted with DCM (4 ᵡ 20 mL), washed with brine and dried over MgSO4. The solvent was removed under reduced pressure and a red
liquid was obtained. The crude mesylate (10.63 g, 48.26 mmol) was treated with KI (22.43 g, 135.13 mmol) in MeCN (60 mL) and stirred overnight under refluxing conditions. The mixture was quenched with H2O (50 mL) and the organic layer was extracted with DCM (3 ᵡ 50 mL), washed with brine and dried
over MgSO4. The solvent was removed under reduced pressure and the crude mixture was subjected to
column chromatography (PE) obtaining a clear yellow liquid. Yield 7.9 g (65%). Rf 0.5 (PE); IR (ṽ, cm-1):
2957, 2896, 2349, 2173, 1248, 1215, 1171, 842, 760, 653; 1H-NMR (CDCl
3, 400 MHz): δ 3.22-3.18 (t, J=
7.5 Hz, 2H), 2.79-2.76 (t, J= 7.5 Hz, 2H), 0.15 (s, 9H); 13C-NMR (CDCl
3, 300 MHz): δ 105.07 (C), 86.79
(C), 25.09 (CH2), 1.06 (CH2), -0.02 (CH3)
General procedure to generate (4-(trimethylsilyl)but-3-yn-1-yl)zinc(II)iodide (11) in situ
To a suspension of zinc dust (2-10 equiv.) in THF (2 M), activated with iodide (2% mol), (4-iodobut-1-yn-1-yl)trimethylsilane 10 (1-2.5 equiv.) was added dropwise and the mixture was stirred overnight under refluxing conditions. The presence of the organozinc was verified by iodine titration. The organozinc reagent was formed as solutions 0.1-0.5 M in THF.
37 Trimethyl(4-phenylbut-1-yn-1-yl)silane (14)
To a 25 mL round-bottom flask, n-BuLi in hexane (1.6 M, 0.66 mmol) was added dropwise at ‒78°C to a solution of trimethyl(prop-1-yn-1-yl)silane (80.8 mg, 0.72 mmol) in THF (6 mL) leading to the formation of the corresponding organolithium compound. The temperature was maintained at ‒78 °C. After 2 hours stirring, phenyl iodide (103 mg, 0.61 mmol) in THF (2 mL) was added and the mixture was stirred for 1 hour at ‒78°C. The reaction mixture was warmed to room temperature and quenched with brine (1 ᵡ 10 mL). The organic layer was extracted with diethyl ether (3 ᵡ 10 mL), washed with brine, dried over MgSO4 and concentrated
under reduced pressure obtaining a colorless oil. Yield: 97 mg (80%). Rf 0.3 (PE/EA 9:1); 1H-NMR (CDCl3,
400 MHz): δ 7.30-7.28 (d, J= 6.6Hz, 2H), 7.24-7.22 (d, J= 6.9Hz, 3H), 2.86-2.83 (t, J= 7.6 Hz, 2H), 2.53-2.49 (t, J=7.6 Hz, 2H), 0.16 (s, 9H)
(1,3,5-tris(bromomethyl))-2,4,6-trimethylbenzene (16)
To a 250 mL round-bottom flask equipped with a cooler, mesytilene (7.8 g, 65 mmol), AcOH (34 mL), HBr 33 wt. % in AcOH (46 mL) and p-phormaldehyde (7.2 g, 239 mmol) were added. The mixture was heated at 100 °C and after overnight stirring was crashed into ice. The desired product was recrystallized from DCM/PE. White needles were recovered by filtration and dried under high vacuum. Yield: 23.5 g (91%). Rf 0.6 (
PE/EA 2:1); IR (ṽ, cm-1): 3002, 2915, 1567, 1446, 1410, 1378, 1205, 1010, 787, 572, 537, 469; 1H-NMR
(CDCl3, 400 MHz): δ 4.59 (s, 6H), 2.46 (s, 9H); 13C-NMR (CDCl3, 300 MHz): δ 137.91 (C), 133.27 (C),
29.92 (CH2), 15.42 (CH3)
(1,3,5-tris(but-1-yne)(trimethylsilane))-2,4,6-trimethylbenzene (17)
To a solution of trimethyl(prop-1-yn-1-yl)silane (101 mg, 0.15 mL) in dry THF (7.5 mL), n-BuLi (1.6 M in hexane, 0.83 mmol) was added dropwise at ‒78 °C. After 2 hours stirring a bright yellow solution was obtained. The temperature was maintained at ‒78 °C and compound 16 (100 mg, 0.25 mmol) in dry THF (0.85 mL) was added and the reaction mixture was stirred at that temperature for 1 hour. The reaction mixture was quenched with brine (4 mL) after overnight stirring. The organic layer was extracted
38 with diethylether (3 ᵡ 20 mL), dried over MgSO4 and the solvent removed under reduced pressure. No
further purification was needed and a white solid was obtained. Yield: 107 mg (87%). Rf 0.4(PE/EA 9:2);
Mp= 85-88 °C; IR (ṽ, cm-1): 2958, 2900, 2173, 1428, 1248, 1036, 883, 758, 661, 635; 1H-NMR (CDCl 3, 400 MHz): δ 2.95-2.91 (t, J= 7.6 Hz, 6H), 2.36-2.30 (t, J= 7.6 Hz, 6H), 2.32 (s, 9H), 0.17 (s, 27H); 13 C-NMR (CDCl3, 300 MHz): δ 135.66 (C), 132.91 (C), 106.75 (C), 84.69 (C), 30.15 (CH2), 19.87 (CH2), 16.07 (CH3), 0.17 (CH3) (1,3,5-tris(methylene) triacetate)-2,4,6-trimethylbenzene (19)
To a 250 mL dried round-bottom flask, equipped with a cooler, compound 16 (5.0 g, 13 mmol) was added together with NaOAc (5.8 g, 71 mmol) and acetic acid (127 mL) as solvent. After overnight stirring under refluxing conditions, the reaction mixture was suspended in H2O (150 mL) and the organic layer was
extracted with DCM (3 ᵡ 80 mL), washed with a saturated solution of NaHCO3, dried over MgSO4 and the
solvent removed under reduced pressure. The recovered solid was dried under high vacuum obtaining a white powder. Yield: 3.83 g (90%). Rf 0.6 (PE/EA 1:1); IR (ṽ, cm-1): 2992, 2936, 1729, 1373, 1227, 1024,
980, 949, 917, 812.17; 1H-NMR (CDCl
3, 400 MHz): δ 5.24 (s, 6H), 2.40 (s, 9H), 2.07 (s, 9H); 13C-NMR
(CDCl3, 300 MHz): δ 170.82 (C), 139.49 (C), 131.06 (C), 61.34 (CH2), 20.64 (CH3), 15.65 (CH3)
(1,3,5-tris(methylene) triacetate)-2,4,6-tris(bromomethyl)benzene (20)
To a 5 mL round-bottom flask equipped with a cooler, N-bromosuccinimide (269 mg, 1.50 mmol) and compound 19 (100 mg, 0.30 mmol) in DCM (1.5 mL) were added. The mixture was irradiated with an halogen lamp (HL 400, 220-240 V, 50 Hz) for 3-4 hours. When the reaction was completed, the mixture was cooled down to room temperature and the solvent removed under reduced pressure. The crude was washed with H2O to remove succinimide and the product was obtained as a white solid after column
chromatography (PE/EA 9:2). Yield: 40 mg (23%). Rf 0.36 (PE/EA 9:2); Mp=106-110 °C; IR (ṽ, cm-1):
2924, 2853, 1723, 1500, 1447, 1374, 1207, 1022, 963, 912, 732; 1H-NMR (CDCl
3, 400 MHz): δ 5.42 (s,
6H), 4.80 (s, 6H), 2.11 (s, 9H); 13C-NMR (CDCl
3, 300 MHz): δ 170.61 (C), 140.59 (C), 135.44 (C), 59.08
39 Lithium bis(3-(trimethylsilyl)prop-2-yn-1-yl)cuprate(I) intermediate (22)
To trimethyl(prop-1-yn-1-yl)silane (20 mg, 0.2 mmol) and TMED (90 µM, 1.9 µmol) stirred in dry Et2O (0.8 mL), in a dry ice/acetone bath, n-BuLi was added (1.6 M in hexane, 0.2 mmol). The reaction was
allowed to warm to ‒15 °C for 45 minutes to generate the organolithium. This was then transferred via cannula to a stirred suspension of CuI (15 mg, 76 µmol,) in Et2O (0.8 mL) kept in a dry ice/acetone bath. The reaction was then warmed to ‒15 °C and stirred for 30
minutes to generate the lithium dialkyl cuprate 22 . 1-methyl-2-((prop-2-yn-1-yloxy)methyl)benzene (27)
To a 25 mL round-bottom flask, propargylic alcohol (454 mg, 8.10 mmol) was added dropwise to a suspension of NaH (324.2 mg, 8.1 mmol) in THF (11 mL) at 0 °C. After 30 minutes 1-(bromomethyl)-2-methylbenzene (1.0 g, 5.4 mmol) was added to the reaction mixture. The reaction was allowed to stir overnight at room temperature and then diluted with ethyl acetate (1 ᵡ 10 mL) and washed with HCl 1 M (2 ᵡ 10 mL). The collected organic layer was dried over MgSO4 and
the solvent was removed under reduced pressure. The crude product was subjected to column chromatography (PE/EA 8:1) obtaining a clear yellow liquid. Yield: 740 mg (85%). Rf 0.3 (PE/EA 8:1); 1
H-NMR (CDCl3, 400 MHz): δ 7.34 – 7.32 (d, J=6.8Hz, 1H), 7.23-7.18 (m, 3H), 4.62 (s, 2H), 4.20-4.19 (d, J=
2.3 Hz, 2H), 2.49-2.47 (t, J= 2.3 Hz, 1H), 2.38 (s, 3H)
(1,3,5-trimethyl)-2,4,6-tris((prop-2-yn-1-yloxy)methyl)benzene (25)
To a 25 mL round-bottom flask, propargylic alcohol (465 mg, 8.30 mmol) was added dropwise to a suspension of NaH (332 mg, 8.30 mmol) in THF (5 mL) at 0 °C. After 30 minutes compound 16 (1.0 g, 2.5 mmol) was added to the reaction mixture. The mixture was stirred overnight at room temperature and then diluted with ethyl acetate (1 ᵡ10 mL) and washed with HCl 1 M (2 ᵡ 10 mL). The collected organic layer was dried over MgSO4 and
the solvent was removed under reduced pressure obtaining a clear brown solid. Yield: 750 mg (92%). Mp= 92-97 °C; Rf 0.33 (PE/EA 9:2); IR (ṽ, cm-1): 3295, 3241, 2925, 2852, 1346, 1092, 1059, 928, 903; 1H-NMR (CDCl
40 (s, 9H); 13C-NMR (CDCl
3, 300 MHz): δ 138.99 (C), 131.99 (C), 80.05 (C), 74.49 (CH), 66.38 (CH2), 57.34
(CH2), 15.70 (CH3)
(1,3,5-tris(methylene)tris(2,2-dimethylpropanoate))-2,4,6-trimethylbenzene (29)
To a 25 mL round-bottom flask, compound 16 (300 mg, 0.76 mmol) and PivOH (697 mg, 6.80 mmol) were added followed by Bu4NF (1M in THF, 4.56 mmol).
After 24 hours stirring at 65 °C, the reaction mixture was diluted with ethyl acetate and washed sequentially with H2O (3 ᵡ 10 mL), saturated solution of NaHCO3 (3 ᵡ 10mL) and
brine (3 ᵡ 10 mL) to get rid of pivalic acid. The solvent was then removed under reduced pressure obtaining a white powder. Yield: 0.33 g (94.5%). Mp= 121-124 °C; IR (ṽ, cm-1): 2971, 2933, 2872, 1722,
1576, 1479, 1460, 1277, 1140, 1031, 770; 1H-NMR (CDCl
3, 400 MHz): δ 5.20 (s, 6H), 2.38 (s, 9H), 1.19
(s, 27H); 13C-NMR (CDCl
3, 300 MHz): δ 178.86 (C), 139.44 (C), 131.11 (C), 38.97 (C), 61.73 (CH2),
41
6. References
1. S. Fletcher , AD Hamilton, Journal of the Royal Society Interface, 2006, 3, 215-233. 2. E. S. Day, S. M. Cote and A. Whitty, Biochemistry, 2012, 51, 9124–9136
3. Jr CA Janeway, P. Travers , M. Walport et al., Immunobiology: The Immune System in Health and Disease. 5th edition. New York: Garland Science 2001.
4. M. Pellecchia, J.Med. Chem., 2013, 56, 13-14
5. C. Chamorro, J. A. W. Kruijtzer, M. Farsaraki, J. Balzarinib and R. M. J. Liskamp, Chem. Commun., 2009, 821-823
6. Michelle R. Arkin & James A. Wells, Nature Reviews Drug Discovery, 2014, 3, 301-317 7. U. Reineke, R. Sabat, R. Misselwitz, H. Welfle, H.D. Volk,and J. S. Mergener, Nature
biotechnology, 1999, vol. 17, 1-95
8. M. C. Smith and J. E. Gestwicki, Expert Rev.Mol.Med., 2012,14, e6, 1-17 9. A. Roxin, G. Zheng, Future Med Chem., 2012, 4,1601-18
10. Sang Hoon Joo, Biomol Ther., 2012, 20, 19-26.
11. M. Katsara, T.Tselios, S. Deraos, G.Deraos, M.T. Matsoukas, E. Lazoura, J. M. Matsoukas, E. Lazoura, J. Matsoukas, V. Apostolopoulos, current medical chemistry, 2006, 13, 2221-2232 12. K. Hiramatsu, Drug Resistance Updates, 1998, 1, 135-150
13. K.Heusler, A.Pletscher, SWISS MED WKLY, 2001, 131, 299-302
14. A. S. Gokhale and S. Satyanarayanajois, Immunotherapy, 2014, 6, 755–774. 15. C. J. White and A. K. Yudin, Nature Chemistry, 2011, 3, 509-524
16. F. Diederich, P. J. Stang and R. R. Tykwinski , Modern Supramolecular Chemistry: Strategies for Macrocycle Synthesis, 2008 WILEY-VCH
17. A.Thakkar, T.B. Trinh, D. Pei, ACS Comb. Sci., 2013, 15, 120–129.
18. B. Alberts.et al., Molecular Biology of the Cell (5th ed), Garland Science: New York, 2008. 19. L.E.J.Smeek, D.Timmers-Parohi, J.J.Benshop, W.C.Puijk, H.Hiemstra, J.H.van Maarseveen,
P.Timmerman, ChemBioChem, 2015, 16, 91-99
20. A. G Cochran, Current Opinion in Chemical Biology, 2001, 5, 654–659. 21. J.Janin, C.Chothia, J.Biol.Chem, 1990, 265, 16027-16030.
42 23. A. Angelini, L. Cendron, S. Chen, J. Touati, G. Winter, G. Zanotti, C. Heinis, ACS Chem. Biol.,
2012, 7, 817−821
24. P. Timmerman, W. C. Puijk1 and R. H. Meloen, J. Mol. Recognit., 2007, 20, 283–299. 25. P.Timmerman, Innovations in Pharmaceutical Technology, issue 43, 48-52.
26. A. O. King, N. Okukado and Ei-ichi Negishi , J. Chem. Soc., Chem. Commun., 1977, 683-684 27. T. Almen, S. Andersson, L.G. Wistrand, K. Golman, O. Antonsen, R. Fossheim, U. N. Wiggen,
H. Wikstrom, T. Klingstedt, I. Luenbach, A. Berg, H. Dugstad, US Patent 6,310,243 B1, 2001 28. G. Lutteke, R. AlHussainy, P. J. Wrigstedt, B. T. Buu Hue, R. de Gelder, J. H. van Maarseveen
and H. Hiemstra, Eur. J. Org. Chem., 2008, 925–933
29. T.W.Baughman, J.C.Sworen and K.B.Wagener, Tetrahedron, 2004, 60, 10943-10948 30. J.Storch, M.Bernard, J.Sykora, J.Karban and J.Cermàk, Eur.J.Org.Chem., 2013, 260-263 31. A.Krasovskiy and P.Knochel, Synthesis, 2006, 5, 890-891
32. E. Negishi, L.D.Boardman, H. Sawada, V. Bagheri, A.T. Stoll, J.M.Tour and C.L.Rand, J.Am.Chem.Soc., 1998, 110, 5383-5396
33. Da-Peng Li, X.Q.Pan, Li-Tao An, J.P.Zou and W.Zhang, J. Org. Chem., 2014, 79, 1850−1855 34. J. L. Nallasivam and R. A. Fernandes, Eur. J. Org. Chem. , 2015, 9, 2012–2022
35. P.J.Bednarski and S.D.Nelson, J.Med.Chem, 1989, 32, 203-213
36. G.M.Whitesides, W.F.Fischer, Jr., J.San Filippo, Jr., R.W.Bashe and H.O.House, J.Am.Chem.Soc.,1969,91,4871-4882
37. C-Y.Wu, A.Brik, S-K.Wang, Y-H.Chen and C-H. Wong, ChemBioChem, 2005, 6, 2176-2180 38. N. W. Mszar, F.Haeffner and A. H. Hoveyda, J. Am. Chem. Soc., 2014, 136 (9), 3362–3365
43
Acknowledgments
I wish to thank Prof.dr.H.Hiemstra who allowed me to join the synthetic organic chemistry group and Prof.dr.J.H.van Maarseveen for being the supervisor of this master project and for his support and encouragements throughout the internship. In addition, I would like to thank dr. E. Ruijter for his kindness to be the second reviewer of this master thesis. A special thanks to Gaston Richelle, MSc for his daily supervision and for his support during the practical work. Furthermore, I would like to thank dr. S. Ingemann Jorgensen and dr. M. A Fernandez-Ibanez for their suggestions during the periodical research meeting together with the entire SOC group (Dieuwertje, Roel, Luuk, Linda, Martin, Hans). I had really a great time in the group with all of you.
Furthermore, I wish to thank Marco, Roel and Gaston with whom I enjoyed most of my last year in this amazing city. We had a lot of fun together and you definitely made the difference during my stay here in Amsterdam.
Last but not least, a big thank goes to my family and to my beloved parents for their encouragement and support to undertake such amazing experience abroad and for always believing in me. Their support has been fundamental to follow my dreams and improve my career prospects.
