Unnatural glycopeptides and proteins and their use in antigen detection and immunization

Carbohydrates are essential for the survival and proliferation of many microorganisms, for cancer progression, and they play a fundamental role in virtually all human physiological processes, most importantly in immunity. Consequently, these macromolecules are at the centre of many vaccines already commercially available, and are crucial for many clinical diagnostics tests. MUC1 is a glycoprotein overexpressed in about 80% of human cancers. However, while in healthy cells the MUC1 backbone displays complex oligosaccharides, in tumours it is decorated with truncated carbohydrates. Consequently, different tumour-associated antigens become exposed, and immunogenic. Over the years, several studies have demonstrated that circulating anti-MUC1 antibodies in serum may be a favourable prognosis for patients with different types of cancer, because these antibodies can limit tumour outgrowth and dissemination. Therefore, efforts have been devoted towards the rational design of MUC1-based antigens to be used as diagnostic tools for detection of anti-MUC1 antibodies in human serum. In this context, our collaborators from the University of La Rioja designed and synthesized various MUC1 antigens that featured a specific hydrogen-by-fluorine substitution, which displayed enhanced affinity towards two commercially available anti-MUC1 antibodies. Naturally, we wanted to assess whether this improved binding would allow for these MUC1 variants to be used to detect circulating anti-MUC1 in the sera of cancer patients. We demonstrated that these novel derivatives are more efficient than natural antigens in detecting circulating anti-MUC1 antibodies in human serum of patients with prostate cancer. However, this was not transposed to the case of breast cancer patients. We hypothesize that this can be due to differences in glycosylation patterns of MUC1 in different cancers, and we are exploring this possibility. It is clear from our results that there are advantages from using unnatural glycopeptides in the search for improved diagnostics with higher affinity, and potentially higher sensitivity. However, it is highly unlikely that there is a “one size fits all” test for all cancers, or even for all cancers that are characterized by aberrant MUC1 glycosylation and expression. Therefore, this search must be an orchestrated multidisciplinary effort, one that combines the areas of fundamental cancer biology, structural biology, and biochemistry, among others not covered in this work. Improving antibody-antigen binding affinity may also be useful for the generation of more immunogenic agents. We are exploring new synthetic glycoproteins as anti-cancer immunizing agents and developing methods of conjugating known protein carriers to unnatural MUC1 variants or other relevant TACAs. In this work, we focused on one of the most used protein carriers for vaccine production (CRM197), and developed a cheap, reproducible, and simple method for the expression and purification of this model protein that requires very rudimentary equipment and common BSL-2 facilities. This protein (iCRM197) has been shown to be identical to one of the commercially available variants (that requires BSL-3 facilities), as shown by biophysical characterization methods. We have also shown that iCRM197 can be used for site-selective modification, using a proof-of-concept reaction with an irreversible Cys-selective carbonylacrylic reagent previously described by our group. This cheaply produced CRM197 variant is currently being conjugated, at the Dr Bernardes group, with high-affinity MUC1 variants (such as the ones studied in Chapter 1) to produce glycoprotein-based cancer vaccine candidates, and might also be useful as model proteins to assess Cys-directed chemical modifications, or even to study carbohydrate-associated immunogenicity. We have also generated and purified a mutant CRM197 bearing an extra Cys residue (i5CysCRM197), making it amenable to Cys-selective chemical conjugation of immunizing antigens without disruption of the native disulfide bonds. This mutant protein has also been characterized and been shown to maintain a similar conformation to CRM197, a requirement for efficient antibody production, as well as its immunogenic character. We have also shown that both iCRM197 and i5CysCRM197 are capable of eliciting specific antibody production in BalB/C mice in a manner similar to CRM197, and that this response can be boosted by subsequent immunizations, showing that these could be efficient as protein carriers in glycovaccine candidates. Therefore, this work sets the foundation for the establishment of a novel alternative protein carrier for chemically-defined glycovaccines