Friendly Strategy to Prepare Encoded One bead-One Compound Cyclic Peptide Library

GIUDICESSI, S. L.; GUREVICH-MESSINA, J. M.; MARTINEZ CERON, M. C.; ERRA-BALSELLS, R.; ALBERICIO, F.; CASCONE, O.; CAMPERI, S. A.

ACS Combinatorial Science

ACS

Año: 2013 vol. 15 p. 525 - 529

2156-8952

One-bead-one-peptide libraries allow the screening of suitable ligands for any target protein. Short cyclic peptides are ideal ligands for affinity chromatography due to their high affinity and selectivity for the target protein and stability against proteases. We designed a library synthesis strategy to facilitate the identification of cyclic peptides by MS consisting in: a) sequential incorporation of a mixture of Fmoc-Ala-OH and Fmoc-Asp[2-phenylisopropyl (OPp)]- OH (15:85) to Gly-hydroxymethylbenzoyl-ChemMatrix resin, b) synthesis of the combinatorial library on the resin by the divide-couple-recombine method, c) removal of OPp with 4% TFA d) peptide cyclization on solid phase through side-chain Asp and amino terminus, e) removal of side chain protecting groups with a 95% TFA cocktail. Peptides were cleaved from the beads with ammonia and the linear code was sequenced by MALDI-TOF MS/MS. The high capacity of ChemMatrix resin together with the sensitivity of MS allows code sequencing from a single bead. Small peptides consisting of a few amino acids represent promising affinity ligands candidates for industrial separations by affinity chromatography. Peptide ligands are much more physically and chemically stable and resistant to proteolytic cleavage than antibodies. They can be readily synthesized by standard chemistry in bulk amounts at a lower cost under good manufacturing practices. Furthermore, peptides allow site-directed immobilization and high ligand density and the matrices are more robust during elution and regeneration than protein-based affinity matrices such as monoclonal antibodies. Particularly, cyclic peptides have higher selectivity and resistance to enzymatic degradation than linear peptides due to their structural constraint which reduces the conformational flexibility, thus potentiating their use as affinity ligands for protein purification. The application of combinatorial peptide synthesis strategies greatly facilitates the discovery of suitable ligands for any given protein of interest. Among the solid-phase strategies available for peptide library preparation, the divide-couple-recombine (DCR), also known as the split-and-mix method, is the most advantageous. This approach entails the following: (i) dividing the solid support (resin beads) into equal portions; (ii) coupling each portion individually with a different amino acid (building block); and (iii), mixing the portions. This procedure assures a theoretically even representation of the library members and a ?one-bead-one-peptide? distribution. To screen these combinatorial libraries, tens of thousands to millions of compound beads are first mixed with the probe molecule. The beads that interact with it will be identified and isolated for peptide structure determination. The high sensitivity and mass accuracy of mass spectrometry (MS), together with the speed of the analyses and the large amount of information generated in each experiment made it a technique of choice for peptide sequencing. Although linear peptides may be readily sequenced by MS, cyclic peptides typically yield complex fragment ion mass spectra, arising from ring opening at multiple positions to afford mass degenerate ions, which undergo subsequent loss of residues. One-bead-two-peptide libraries in which each bead contains both the cyclic peptide and its linear counterpart as a coding tag facilitate MS analysis. For peptide identification by MS a cleavage site to release the peptides from the beads is required. The cyclic peptide as well as its linear counterpart anchored on each bead are released and analyzed together by MS. In current approaches, methionine is commonly used as linker.8-11 The main drawback is that the selective cleavage of the peptides is performed with CNBr which is a very hazardous reactive, especially when working with hundreds of individual beads selected from the screening process.12 We have previously proposed the use of 4-hydroxymethylbenzoic acid (HMBA) linker13 coupled to ChemMatrix (CM) resin for one-bead-one-peptide library synthesis14 and we have developed a suitable cleavage method to release the peptide from the linker HMBA by vapor phase ammonolysis using NH4OH 30%, which is much more economic and safer than CNBr. The method allows MS analysis from each single bead isolated after one-bead-one-peptide library screening. The aim of this work was to design a new strategy to prepare beads with homodetic cyclic peptides together with a minimum proportion of its linear peptide code suitable for one-bead-one-peptide libraries. The method was based on the proper combination of the solid support, the linker for resin-peptide anchorage and protecting groups for N-alfa and side-chain protection. CM was used as solid support and HMBA as linker. CM is a highly crosslinked, amphiphilic resin composed entirely of polyethyleneglycol monomers that contain exclusively primary ether bonds. These bonds facilitate high resin loading, as compared to polystyrene resins, and high chemical and mechanical stability. CM is compatible with both organic and aqueous solvents, which are used for the peptide synthesis and for the screening step respectively, and therefore it is highly suitable for the whole process. The benzyl ester, which forms the HMBA linker with the first amino acid, is stable to piperidine used for Fmoc group removal making the linkage applicable for chemical elongation of a peptide sequence with Fmoc strategies. The HMBA inertness to trifluoroacetic acid (TFA) allows side-chain deprotection without releasing the peptide from the resin. The removal of protecting groups is required since screening is normally carried out on fully unprotected peptides. At the C-terminal, Gly was incorporated with the aim of increasing the molecular weight of the final peptide to facilitate its identification by MS and overcoming the poor cleavage efficiency of esters of Ile and Val. For solid phase ring formation Fmoc-Asp[2-phenylisopropyl (OPp)]-OH was added to the peptide sequence. The semi-permanent group OPp was cleaved with 4% TFA and an amide linkage was formed between the amino terminus and carboxylic acid function to give homodetic cyclic peptides. Most permanent protecting groups commonly used in solid phase peptide synthesis are stable to 4% TFA. However, these conditions are sufficient to remove the side-chain trityl (Trt) protection of His residue. Thereafter, in the case of using His in the peptide library, Fmoc-His(Boc)-OH must be used instead of Fmoc-His(Trt)-OH. 4-[N-(l-(4,4-dimethyl-2,6-dioxocyclohexylidene)-3-methylbutyl)amino]-benzyl (ODmab), commonly used as a semi-permanent protecting group for Asp in solid phase cyclization, cannot be used in the present approach because its cleavage is achieved with hydrazine, reactive which also removes the peptides from the HMBA-CM resin. In order to prepare a peptide code together with the cyclic peptide, a minor proportion of Fmoc-Ala-OH was incorporated instead of Fmoc-Asp(OPp)-OH. For the purpose of reducing interference during the screening step due to the interaction of the linear peptides with the target, the minimum proportion of the linear code necessary for its MS analysis was synthesized. When adding 0.2 eq of Fmoc-Ala-OH to the resin previous to Fmoc-Asp(OPp)-OH incorporation, only 0.15 eq of Fmoc-Ala-OH was incorporated as was measured by Fmoc analysis. Then, Fmoc-Asp(OPp)-OH was coupled to the remaining resin equivalents. Peptides bounded through HMBA to the resin beads can be released with nucleophiles. CM contains exclusively primary ether bonds and so is stable to nucleophilic conditions. The ammonia vapor has the advantage over other nucleophiles such as NaOH that it is easily removed by evaporation, thus avoiding the presence of contaminants that could interfere with the MS peptide analysis. The small proportion of the code (non-cycled Ala containing peptide) was enough for MS analysis due to the high sensitivity of the MS technique together with the high loadings of HMBA-CM (0.67 mmol/g substitution). High quality mass spectra were obtained. Peptide signals were high and the cyclic peptide sequence could be deduced from the tandem mass spectra of the linear peptide. Figure 1 shows the resulting MALDI-mass spectrum and MALDI-tandem mass spectrum performed, obtained by adding 1 μL of the eluted peptides ARYKSRVAG-NH2 and ciclocyclo-(1,8)-ARYKSRVDG-NH2 from a single bead. To prepare the linear code Fmoc-Asp(tBu)-OH together with Fmoc-Asp(OPp)-OH could have been used instead of Fmoc-Ala-OH but the second amino acid was cheaper. Moreover, when preparing a cyclic peptide library of tens of thousands to millions of compounds, the yield of cyclization of some minor sequences may be incomplete but undetected by the Kaiser test. If Fmoc-Asp(tBu)-OH together with Asp(OPp)-OH is used, it is not possible to be aware of incomplete cyclization of the peptide-bead analyzed specially because MS is not quantitative. Besides, when using Fmoc-Ala-OH together with Fmoc-Asp(OPp)-OH, incomplete cyclization can be clearly visualized with MS because there will be three different signals corresponding to the cyclic peptide, its linear counterpart and its linear code clearly identified by MS. The strategy developed is suitable for the synthesis of one-bead-one-peptide libraries in which each bead has a cyclic peptide together with a minimum proportion of its linear code for MS identification