RESUMEN
Pancreatic cancer is one of the deadliest cancers worldwide, mainly due to late diagnosis. Therefore, there is an urgent need for novel diagnostic approaches to identify the disease as early as possible. We have developed a diagnostic assay for pancreatic cancer based on the detection of naturally occurring tumor associated autoantibodies against Mucin-1 (MUC1) using engineered glycopeptides on nanoparticle probes. We used a structure-guided approach to develop unnatural glycopeptides as model antigens for tumor-associated MUC1. We designed a collection of 13 glycopeptides to bind either SM3 or 5E5, two monoclonal antibodies with distinct epitopes known to recognize tumor associated MUC1. Glycopeptide binding to SM3 or 5E5 was confirmed by surface plasmon resonance and rationalized by molecular dynamics simulations. These model antigens were conjugated to gold nanoparticles and used in a dot-blot assay to detect autoantibodies in serum samples from pancreatic cancer patients and healthy volunteers. Nanoparticle probes with glycopeptides displaying the SM3 epitope did not have diagnostic potential. Instead, nanoparticle probes displaying glycopeptides with high affinity for 5E5 could discriminate between cancer patients and healthy controls. Remarkably, the best-discriminating probes show significantly better true and false positive rates than the current clinical biomarkers CA19-9 and carcinoembryonic antigen (CEA).
RESUMEN
Mucin-1 (MUC1) glycopeptides are exceptional candidates for potential cancer vaccines. However, their autoantigenic nature often results in a weak immune response. To overcome this drawback, we carefully engineered synthetic antigens with precise chemical modifications. To be effective and stimulate an anti-MUC1 response, artificial antigens must mimic the conformational dynamics of natural antigens in solution and have an equivalent or higher binding affinity to anti-MUC1 antibodies than their natural counterparts. As a proof of concept, we have developed a glycopeptide that contains noncanonical amino acid (2S,3R)-3-hydroxynorvaline. The unnatural antigen fulfills these two properties and effectively mimics the threonine-derived antigen. On the one hand, conformational analysis in water shows that this surrogate explores a landscape similar to that of the natural variant. On the other hand, the presence of an additional methylene group in the side chain of this analog compared to the threonine residue enhances a CH/π interaction in the antigen/antibody complex. Despite an enthalpy-entropy balance, this synthetic glycopeptide has a binding affinity slightly higher than that of its natural counterpart. When conjugated with gold nanoparticles, the vaccine candidate stimulates the formation of specific anti-MUC1 IgG antibodies in mice and shows efficacy comparable to that of the natural derivative. The antibodies also exhibit cross-reactivity to selectively target, for example, human breast cancer cells. This investigation relied on numerous analytical (e.g., NMR spectroscopy and X-ray crystallography) and biophysical techniques and molecular dynamics simulations to characterize the antigen-antibody interactions. This workflow streamlines the synthetic process, saves time, and reduces the need for extensive, animal-intensive immunization procedures. These advances underscore the promise of structure-based rational design in the advance of cancer vaccine development.