MOLECULAR SCAFFOLDS FOR PROTEIN
SURFACE MIMICRY. NEW ERA IN
PHARMACEUTICALS.
Abstract
Proteins are important for preserving life. Their highly complex surface imparts functionalities that are only found in proteins. Through specific sites on their surface they can interact with one another to carry out several biological functions, such as DNA replication, signal transduction, immune response and apoptosis. However, infectious proteins can also interact with physiological proteins inside a living organism. This results in harming the organism which can even lead to death. Nevertheless, by administrating a molecule that generates therapeutic antibodies can save the day. The problem is the easiest way that can be done is by using a natural protein, however, proteins are known as bad drugs candidates, basically due to degradation that exhibit inside the body. A novel solution would be to re-create the “hot spots” of a protein in a considerable smaller molecule, i.e. to mimic the protein.
A good way to do that is to connect linear peptides, which bear the antigenic sites of a protein, on an organic molecule (scaffold). This molecular scaffold has to induce certain conformation on the peptides and stability. Moreover, reactions that are applicable to unprotected peptides (“Click Chemistry”) are needed so that protective groups are absent in the synthesis.
Contents
cover page
Abstract
Abbreviations
1.Introduction
Protein – Protein Interactions (PPIs) and their Importance.
Artificial Proteins and the expertise of mimicking.
2.Discrimination: The ability to distinguish minor Differences – Selectivity in Reactions
3.Re-creating Nature in a Flask
Divergent Approach for mimicking protein binding sites.
3.1Scaffold-induced conformation to protein mimics.
3.1.2Mimicking α-Helix secondary structure.
3.1.3Mimicking β-Turn and β-Sheet Secondary Structures.
3.1.4Mimicking Looped conformations.
3.1.5Mimicking Discontinuous Protein Binding Sites.
4.Starting from Simple Organic Molecules towards Drug Candidates
Discontinuous Protein Binding Site Mimics.
4.1The Triazacyclophane (TAC) – Scaffold.
4.2Functionalized cyclic peptides as scaffolds.
4.3CLIPSTM - Technology.
4.3.1Peptide ligation via Copper(II)-catalyzed Azide-alkyne Cycloaddition.
4.3.2Peptide ligation via Strain-promoted Alkyne-Azide Cycloaddition.
4.3.3Peptide ligation via Thiol-ene Reaction.
4.3.4Peptide Ligation via Oxime Formation.
4.3.5Constrain-induced via Single and Double Disulfide Bond.
4.3.6Double-CLIPS Technology in the Mimicry of Discontinuous Binding Sites.
4.4Other Multifunctional Scaffolds.
References
Abbreviations
1 |Molecular scaffolds for Protein Surface Mimicry. New Era in Pharmaceuticals.
AAPH / 2,2'-Azobisisobutyramidinium chlorideAcm / Acetamidomethyl
AFB2 / Aflatoxin B2
AIDS / Acquired immunodeficiency syndrome
Ala / Alanine
Alloc / Allyloxycarbonyl
Arg / Arginine
Asp / Aspartic acid
Bn / Benzyl
Boc / Di-t-butyl dicarbonate
BSA / Bovine Serum
albumin
Cbz / Benzyl chloroformate
CD3 / Cluster of differentiation 3
CH3CN / Acetonitrile
Cu / Copper
CuAAC / Copper(I)-catalyzed azide-alkyne cycloaddition
CuSO4 / Copper(II) sulfate
Cys / Cysteine
DAVE / Asp-Ala-Val-Glu
DNA / Deoxyribonucleic acid
DPAP / 2,2-Dimethoxy-2-Phenyl Acetophenone
DTT / Dithiothreitol
ELISA / Enzyme-linked immunosorbent assay
EVH / Ena/Vasp homology proteins
FAKL / Phenylalanine-Alanine-Lysine-Leucine
FRET / Förster resonance energy transfer
Fmoc / Chloroformic acid 9H-fluoren-9-ylmethyl ester
Glu / Glutamic acid
Gly / Glycine
gp120 / Envelope glycoprotein GP120
HIV / Human immunodeficiency virus
H2O / Water
HPLC / High-performance liquid chromatography
I2 / Iodine
IgG / Immunoglobulin G
KLA / Lys-Leu-Ala
LC-MS / Liquid chromatography–mass spectrometry
Leu / Leucine
Lys / Lysine
MBP / Maltose-Binding Protein
MeOH / Methanol
NaHCO3 / Sodium hydrogen carbonate
Na2S4O6 / Sodium tetrathionate
OKT 3 / Muromonab-CD3/ Orthoclone OKT3
PBS / Phosphate-buffered saline
PPIs / Protein-protein interactions
RAFT / Regioselectively addressable functionalized templates
RGD / Arg-Gly-Asp
r.t. / Room temperature
SN2 / Bi-molecular nucleophilic substitution
SPAAC / Strain-promoted azide-alkyne cycloaddition
tBuOH / 2-methyl-2-propanol
TCEP / Tris(2-carboxyethyl)phosphine
TFA / Trifluoroacetic acid
Tris / tris-(Hydroxymethyl)-aminomethane
Val / Valine
VEGF-A / Vascular endothelial growth factor A
VP1 / Polyomavirus Capsid Protein
1 |Molecular scaffolds for Protein Surface Mimicry. New Era in Pharmaceuticals.
- Introduction
Life on Earth emerged when highly complex molecules were formed;these were given the namemacromolecules(from the Greek macro which means long).These make upthe backbone of life; the cells (1).In living cells, macromolecules are found in four different types: proteins, nucleic acids, carbohydrates and lipids. These are essential for all living organisms, as they control all life processes. However, proteins by themselveshold a key role in life’s cycle, as they are the most abundant macromolecules found in living cells; the diversity of functional proteins in cells and tissues is incomparably larger than of any other class of macromolecules and they are involved in practically all life processes (1).
However, the question is“why a synthetic chemist should trouble his/her mind with developingartificial proteins”. Some scientists find it challenging to devise and synthesize new molecules, but, the most valid answer could be, the need of constructing molecules (usually smaller than those found in nature) that bear comparable activity with nature’s macromolecules such as proteins.
Protein – Protein Interactions (PPIs) and their Importance.
Protein – protein interactions (PPIs)govern most of biological processes in a living organisms (2). Such interactions take part in the signal transduction, immune response, DNA replication and apoptosis(3; 4; 5). Hence, even the slightest alternation of these interactions can be fatal to an organism (source of HIV, cancer, leukemia, malaria and other auto-immune or infectious diseases) (6; 7; 8; 9). This fact unveils the significance of vaccines and drugs.
Although, the idea of using proteins as drugs is tempting, proteins in general are poor drug candidates due to bioavailability problems that arise from conformational instability (proteins readily lose their three-dimensional structure under a variety of conditions), susceptibility to proteolytic degradation, poor membrane penetration and low cellular uptake, and unfavorable pharmacokinetics due to rapid clearance or instability of peptide bonds to degradation by peptidases inside the human body (10).
The old fashioned view that a protein must keep its native morphology (primary, secondary, tertiary and quaternary structure) to be functional, is wrong; recent studies have shown that about 30% of eukaryotic proteins are composed of proteins that do not have a well-defined structure prior to interaction with their binding partner (11; 12; 13).Moreover, their biological activity is not globally extended on all protein surface and only small regions of their folded surfaces exhibit any activity (14), hence, their functionality can be reproduced in smaller molecules (protein mimics) that retain these bioactive surfaces(10).
Artificial Proteins and the expertise of mimicking.
Protein mimics have been used in the past and have proved to be highly selective and, therefore, successful in the treatment of several life-threatening diseases(15). In the literature, several cases of linear peptides being used as drugs have been documented(16). However, further research on the subject indicates that certain structural fixation is required for biological activity (17). It is well known nowadays among scientists in the peptide field that proteins recognize (interact) one another via defined parts on their surfaces, these constitute the antigenic sites of a protein and aredefined asepitopes(18). A protein mimic is usually a smaller compound than a protein, which bears enough information to reproduce the functionalities of a protein or peptide (19).
Epitopes are distinguished in two types; the “continuous epitopes”, which are seldom in proteins and the “discontinuous epitopes”, which are most abundant antigens in natural proteins. A contiguous epitope is defined by consecutive residues directly linked in the sequence (primary structure) by peptide bonds, whereas, a discontinuous epitope is defined as spatially adjacent group of surface amino acid residues that are partially or totally without any peptide linkage directly connected (20).
Since, epitopes cover a small part of a protein’s surface, an artificial and, at the same time, active drug requires no more than the antigen sites of the molecule. Chemists try to achieve that goal, the past few decades, by binding only the active parts of a protein on a small organic molecule, i.e. “scaffold” (or “template”). Scaffolds are useful especially for mimicking discontinuous epitopes. The latter can only bereproduced effectively when their three-dimensional topology is retained in the artificial protein (21).
The origins of immunotherapy are along with the discovery of the structure of antibodies and the development of hybridoma technology, with which the first monoclonal antibody became reality (22). The latter, was approved by the FDA (Food and Drug Administration) in 1986 and was a murine IgG2a CD3 specific transplant rejection drug, OKT3 (brand name muromonab). This drug found use in solid organ transplant recipients who became steroid resistant (23).
It has not being a whilesince, artificial proteins were synthesized via long and repetitive routes, utilizing the properties of orthogonal protective groups (:groups that are selectively removed). Those routes were characterized by several limitationsspotted in a divergent approach; according to a divergent method, a central molecule (or scaffold) bearing orthogonally protected functional groups is used, to which multiple different peptides are being attached. A divergent approach suffers from limitations regarding deprotection reactions and purification steps. Moreover, organic molecules such as those exhibit strongly hydrophobic properties (21). To overcome theaforementioned limitations a convergent approach must be used instead.
Aconvergent approach, unlike to any divergent, relies on orthogonal ligation of pre-functionalized peptide fragments that simultaneously constitute the central scaffold. Moreover, protection or pre-activation of the functionalities is not required. Usually fewer steps are involved in convergent synthesis. (21).The introduction of the concept was made by Velluz et alin 1967 (24). According to him, in a convergent procedure various parts of the target molecule are assembled separately and independently and at the final steps are linked together. A convergent procedure has the effect of lowering the path length of the main line, which in turn makes the procedure more economical than a simple linear procedure (25).
Scheme 1.3: Convergent synthesis.
For the successful mimicry of all epitopes, three categories (or levels) must be ascribed (21): 1) “Continuous (Linear) Epitope”; it can be reproduced usually by alinear peptide, 2) “Continuous (Conformational) Epitope”; is necessary to have a spectacular structure to bind on the corresponding protein and, 3) “Discontinuous (Conformational) Epitope”; successful mimic of the epitope requires more than one separate peptide fragment on the scaffold and by forming the right structure (Scheme 1.3).
Scheme 1.4: Three levels of protein mimicry.
The scope of this project is to present the latest fashion in protein mimics. The focus will be upon orthogonal ligation utilizing scaffolds and also different approaches will be mentioned.Several examples from the recent literature will be demonstrated with the advantages and the disadvantages of eachmethodology.
In the final chapter itwill be attempted to showthebest possible way, if it’s feasible, for the synthesis of complex proteins and their mimics.
- Discrimination: The ability to distinguish minor Differences – Selectivity in Reactions
The problem with the old fashion syntheses towards protein mimics, was the large number of reactive centers present in a single reaction. To bypass this obstacles, chemists thought of deactivating certain groups selectively. This was made possible by the introduction of protective groups and especially orthogonal protective groups(26). However, it has been proved in practice that using protective groups creates some additional problems; i) incomplete coupling and deprotection reactions lead to truncated and deletion sequences, ii) accumulation of side products from incomplete reactions, impurities from reagents, solvents and protected amino acids, and iii) aggregation of growing peptides (27). Hence, strategies employing protective groups are limited in the number of reactive groups present in one reaction, i.e. small peptides had to be used with no exceptions.
The only solution is to rely on convergent approaches that are fully chemoselective and render protective groups obsolete. Those kind of strategies use reagents with unique functionalities and usually are facile and robust. But what do chemist imply when they refer to a reaction as chemoselective?
In 1983 Trost set it as follows, “The ability to discriminate among the reactive sites is referred to as chemoselectivity” (28). In simple words, single reaction of one functional group in the presence of others renders the reaction chemoselective. For example, the preferential reduction of reduction of a ketone in the presence of an ester and a carbon-carbon double bond, as stated in a recent publication by the group of Procter, is highly chemoselective (29).
Chemists were able to use convergent strategies for mimicking nature after the introduction of several relatively new chemical reactions, these belong to the more general group of reactions called Chemical Ligation. With chemical ligation approach large peptide fragments are joined together chemoselectively through the formation of an amide or non-amide linkage (26). Moreover, a chemoselective approach is usually based on thiol and imine functionalities for forming an amidic bond or by employing hydrazine, oxime, thioester and thioether functionalities, which results in a non-amidic bond formation (30; 31; 32; 33). However, a more recent approach, coined by the Nobel laureate in Chemistry in 2001 (34)Sharpless and co-workers, promises fast, facile and robust reactions towards the desire products (35). The new approach is termed Click Chemistry and is defined by a set of stringent criteria that a process must meet to belong in this context. A reaction must be i) modular, ii) wide in scope, iii) high in yields, iv) no by-products or produce harmless by-products that can be removed by nonchromatographic methods, v) stereospecific but not necessarily enantioselective, vi) simple reaction conditions, vii) readily available starting materials and reagents, viii) no solvent or a solvent that is benign or easily removed, and ix) simple product isolation. A reaction that fits perfect in this category is the azide-alkyne Huisgen cycloaddition or Click reaction. It is a 1,3-dipolar cycloaddition between an azide and an alkyne to give a 1,2,3-triazole (36). However, a variation to the Huisgen cycloaddition has shown better results; though, Huisgen reaction is devoid of side products, usually the outcome is a mixture of 1,4 and 1,5 regioisomers (Figure 2.1) and usually proceeds in elevated temperatures (37). This variant, on the other hand, is copper(I) catalyzed and proceeds in room temperature in a variety of solvents (including water with no organic co-solvent). It is better termed as Copper(I)-catalyzed Azide-Alkyne Cycloaddition or CuAAC. However, click chemistry includes several other chemical reactions such as the famous Diels-Alder reaction, nucleophilic substitution and particulary ring-opening of strained heterocyclic electrophiles, carbonyl chemistry of the non-aldol type (oxime ethers, urea formation, thiourea etc.) and additions to carbon-carbon multiple bonds (37; 38).
CuAAC reactions is perfect in every aspect except of one, the mandatory use of copper as catalyst renders the reactionuseless, due to the toxicity of copper, in both bacterial and mammalian cells(39). Bertozzi et al. thought of a different route to promote [3 +2] cycloadditions (40). By using strain were able to promote the reaction; the best candidate was already known since 1961, when Wittig and Krebs publish their findings, of a simple and facile reaction between cyclooctyne (the smallest stable cycloalkyne)and an azide, which proceed fast under physiological conditions in the absence of catalysts or other reagents to give one single product, the triazole. The reaction was studied further, from the Bertozzi group, in living cells with positive results and no harmful effects on the cells were observed.
A different and chemoselective route of connecting peptides on scaffolds is Native Chemical Ligation (NCL). Following Kemp’s “priop thiol capture”(41)approach and Kent’s chemical ligation method (33), Dawson et al. developed a procedure that is completely entropically driven, overcoming the problems associated with enthalpy driven approaches, which are due to low reactant concentrations and entropy barrier imposed by high molecular masses, especially when side-chain protection is extensively used (42).. NCL relies on the unique properties of sulfur chemistry; in a certain environment thiol groups can act as extreme nucleophiles and react chemoselectively. NCL employs two fully unprotected peptide fragments, on the C-terminal there is a thioester group and on the N-terminal is always a Cys moiety. Firstly, the sulfhydryl residue on the Cys attacks the thioester group on the other peptide fragment to give the chemoselective reaction, then an intramolecular acyl transfer results in the final product bearing a native amide bond in every case of NCL.
However, NCL has a Cys requirement at the N-terminal, which is quite a problem due to scarcity of Cys amino acids in proteins. A solution to this problem is by utilizing serine or threonine ligation methods along with a (removable) salicylaldehyde ester as scaffold(43).This procedure relies on the properties of salicylaldehyde ester, which renders the ester completely inert to amine nucleophiles. The first step involves imine ligation; nucleophilic attack of the amine moiety of the N-terminal fragment, followed by a reversible oxazolidine formation. Then a non-reversible 1,5 O/S → N acyl transfer forms an amide bond, then if it’s desirable the scaffold can be removed under acidic conditions. The reaction proceeds smoothly in pyridine/acetic acid (1:1) and usually takes 4 – 24 hours to completion in room temperature.
Though, is a purely chemoselective ligation proceduredue requirement of mainly organic solvents (pyridine – acetic acid buffer) the method is limited until it is applicable to physiological conditions.
All these approach have being applied in the synthesis of protein mimics.
Figure 2.2: 1) CLIPS scaffolds developed by Timmerman et al. 2) 2nd generation CLIPS scaffold. 2) Copper(I)-Catalyzed Azide-Alkyne cycloaddition. 3) Strain-Promoted Azide-Alkyne cycloaddition.
A novel method towards protein mimics was developed by Timmerman et al. of Pepscan Therapeutics in 2005, this is called CLIPS-technology (Chemical Linkage of Peptides onto Scaffolds)(44). The reaction has been found to be chemoselective to side chain unprotected dithiol-containing peptides. It was developed mainly for mimicking looped protein functionalitiesCLIPS reactions employ readily available materials such as 1,3-dibromoxylene or m-T2 (m stand for meta bromide positions, T for Template and 2 for the number of bromides), 1,3,5-tribromomesitylene or T3 and 1,2,4,5-tetrabromodurene or T4 (3Figure 2.2). However, further development, under the supervision of Timmerman, in the subject resulted in a second generation of scaffolds (4 Figure 2.2), where a second exquisite functionality was added to the scaffold. By differentiating the central molecules of a protein mimic they were able to add more peptide fragments on the central molecule and imitate protein functions more effectively (45).