The glycosylation of the proteins is a common post-translational modification found in eukaryotic cells. Glycosylated proteins play a crucial role in cell adhesion, cell-matrix interaction, and immune response. Several studies suggest that the overexpression of these glycoconjugates is linked to many diseases, inflammation, viral infections, cancer metastasis, and cellular apoptosis among other pathologies. Over the years, the scientific community has tried to answer questions about the influence of the oligosaccharides on the protein structure or the relationship between the oligosaccharide structure and glycoproteins function. The main limitation to study these important aspects is the difficulty to obtain homogeneous isoforms of glycoproteins due to the lack of genetic control in the biosynthesis. The crescent interest in the understanding of structure-activity relationship of glycoproteins has led to the development of new technologies to obtain complex glycans and proteins. In this work, different methodologies were investigated to synthesize defined isoforms of the ubiquitous glycoprotein called Thy-1 or CD90. This heavily glycosylated protein was first discovered in 1964 in mouse T lymphocytes and later detected in human fibroblast neurons, blood stem cells, and endothelial cells. The 25kDa glycoprotein is translated as a 161 and 160 amino acids length protein in humans and mouse, respectively. The protein is transferred from ribosomes into the ER and receives several post-translational modifications including N-glycosylation at three different asparagine residues and a glypiation at the C-terminal cysteine residue. Several studies showed the association between the expression of Thy-1 and T cell activation, neurite outgrowth, apoptosis, leukocyte and melanoma cell adhesion and migration, tumor suppression, and fibroblast proliferation. The exploration of the structure-activity relationship to determine the role of the glycosylation on this protein required defined glycoprotein and motivated this work to develop a de novo synthesis of the Thy-1 pure glycoforms. The retrosynthesis of the mature glycoprotein was designed considering three key steps: 1) the assembly of the primary sequence of the protein; 2) the introduction of glycans into the protein; 3) the differentiation of the oligosaccharides at the glycosylation sites. The primary sequence of the glycoprotein was obtained using a sequential native chemical ligation (NCL) of three segments: the peptide fragment (1-18), the glycopeptide (19-84) having two N glycosylations, and the glycopeptide (85-110) containing one N glycosylation. The three peptide fragments were synthesized by solid-phase peptide synthesis (SPPS) and were obtained having a hydrazide moiety on the C-terminus that was converted into a thioester prior to the NCL. For this purpose, a Wang and a trityl resin were functionalized with hydrazine and modified manually with the amino acid on the C-termini. The protecting acetamidomethyl group was used to mask the thiol of the N-terminal cysteine residues in the fragment (19-84) and (85-110) to allow the stepwise ligation of the fragments. The elongation of the polypeptide chains was carried out in the microwave-assisted synthesizer and careful optimization of deprotection and coupling cycles was executed for the three fragments. The incorporation of amino acids was performed considering the common problems that affect the peptide synthesis including the cyclization reaction on asparagine and aspartic acid, and the racemization of cysteine and histidine at temperatures higher than 50°C. These precautions were not enough to obtain the Thy-1 glycopeptide fragments in high yield and with desired quality. The synthesis of the designed peptides required multiple optimizations to avoid side reactions. Some general requirements were established. Hindered amino acids (glutamic acid, isoleucine, and phenylalanine) required double coupling to assure complete attachment to the growing chain. Vicinal identical amino acids (e.g. serine 25-serine 26) were introduced using single coupling for the first residue in the sequence and a double coupling for the following residue. One of the most crucial point for the optimization of the synthesis was the minimization of the aspartimide formation involving the cyclization of the aspartic acid. Among different cocktails for the removal of the Fmoc group promising the suppression of this side reaction, the best results were obtained by using 20% piperidine and 0.7% of formic acid in dimethylformamide, which reduced the aspartimide formation up to 70%. The glycosylations were introduced following two approaches to allow differentiation between the glycan at each glycosylation site. In the cassette-method, a glycosylated asparagine was synthesized and incorporated in the corresponding glycosylation site during the assembly of the peptides. Different procedures were evaluated and optimized for the synthesis of the N-acetyl-O-per-acetylated glucosaminyl asparagine 2 and its introduction into SPPS to get the fragments 6a-c (19-84) and 37a-b (85-110) with a peracetylated glucosamine on each glycosylation site. The convergent approach was used to synthesize fragment 6b (19-84) having the peracetylated N acetylglucosamine 2 in one glycosylation site and an unprotected N-acetylglucosamine, installed via Lansbury aspartylation on the second site. The Lansbury strategy required the formation of the amide bond between the free carboxylate at an aspartic acid and the N-acetylglucosamine amine. The reaction was performed on the fully protected peptide fragment and required the synthesis of two building blocks, an orthogonal protected aspartic acid 3, and a pseudoproline dipeptide 4. The aspartic acid was synthesized with a photolabile protecting group on the side chain, which was selectively removed on resin to give a free carboxylic acid. To avoid the undesired rearrangement of the amino acid, the protected pseudoproline dipeptide Thr-Ser(Me,MePro) was incorporated before the aspartic acid. The three (glyco)peptides were efficiently synthesized by a combination of manual and automated processes and were characterized by their challenging purifications. The difficulties in obtaining the isoforms from the Thy-1 fragment (19-84) glycopeptides 6a-b, were related to the low solubility of the generated glycopeptides and the loss of acetyl groups on the peracetylated glucosamine. The analysis of these problems led to the design of the optimal strategy to synthesize in high yield the fragment 6c having a protected and non-protected N‑acetylglucosamine. The synthesis of glycoform 6c involved the coupling of a per-acetylated N-acetylglucosaminyl asparagine building block in the peptide sequence that was de-acetylated before the introduction of the second glycosylated asparagine residue. The assembly of the Thy-1 glycoprotein involved the ligation of the synthesized fragments by the chemoselective reaction between the C-terminal peptide thioester in one peptide fragment and the free N-terminal cysteine residue of another fragment. The process required the conversion of the synthesized peptide hydrazides into the corresponding thioesters prior to each ligation and the removal of the Acm-group of the fragments having the N-terminal cysteine. Two combinations were studied to ligate the three fragments: from C- to N-terminus and from N- to the C-terminus. The first approach consisted of the initial ligation of the glycopeptides 6a-c (19-84) and 37a-b (85-110) and the subsequent ligation with the peptide 1 (1-18). The second strategy involved the reaction of peptide 1 (1-18) with the glycopeptide 6a-c (19-84) and the following ligation with the fragment 37a-b (85-110). The ligation conditions were investigated, and the strategy was selected considering factors such as the availability and solubility of the fragments, number of steps and the efficiency of the processes. The N-to C-strategy was the first approach adopted. The hydrazide precursor 6a was converted into the thioester fragment 38 by formation of the peptide azide at low pH and thiolysis with the corresponding thiol. The Acm group from the cysteine in 37b was removed with mercury (II) acetate and reduction of the sulfur-mercury complex with mercaptoethanol. The ligation of the fragments delivered the glycoprotein 40 and the hydrolysed fragment 38 as by-product. The low solubility of the glycoprotein 40 hampered the isolation of the product and suggested the need of more polar fragment. For this purpose, glycoform 6b was converted into the thioester 41 and reacted with fragment 37a having a free N-terminal cysteine. Surprisingly, this ligation did not proceed suggesting that the treatment with mercury (II) acetate could be beneficial for the ligation. A N- to C- ligation of fragment 6c treated with the mercury (II) acetate with the thioester of fragment 1 proceeded successfully and the isolation of the ligation product yielded the glycoprotein 45, confirming the better behaviour of polar fragments and reactivity of the cysteine-containing fragments treated with mercury. A ligation of the MPAA thioester of Thy-1 fragment (1-84) with 37b to obtain the full glycoprotein Thy-1 (1-120) was hindered by the poor solubility of thioester 46 in the ligation media. A new C- to N- ligation involving the polar glycoform 6c with the fragment 37b delivered the glycoprotein 49 that was treated in one-pot with PdCl2 to release the N-terminal cysteine and get 50. Finally, a methyl 3-mercaptopropionate MMP thioester of fragment 1-18 (51) was ligated with glycoprotein fragment 50 to obtain the desired Thy-1 glycoprotein (1-120) having three glycosylations. In this work was presented the design and evaluation of different strategies for the synthesis of the glycoprotein Thy-1. The assembly of the 13kDa glycoprotein required different steps and the optimized synthesis of amino acid building blocks for the solid-phase assembly of glycopeptides. Various methods were applied for the generation of glycopeptides, the exploration of the chemical properties of the obtained fragments, and the modulation of the chemical conditions for the ligation of the peptide fragments to get the target glycoprotein. This work focused on the production of homogeneous glycoforms of Thy-1. However, the synthetic methods and protocols established in this work are applicable for the synthesis of any peptides and glycopeptides and contribute to the chemical synthesis of other important glycoproteins.
