Novel approaches to study immune responses to post translational modifications for cancer detection

Despite advances in screening and treatment, mortalities from breast and lung cancers have remained high in the US over the last 20 years. It is widely accepted that early detection is critical to improving outcomes in both diseases. Both also rely on imaging for screening, but false positive and false negative detection are associated with unnecessary biopsies, missed diagnoses, and costs. There is an urgent need for biochemical markers that improve the performance of imaging technologies. Our laboratories have been successful at identifying useful cancer biomarkers by exploiting patients’ own ability to produce antibodies against tumorassociated antigens (TAA), referred to as tumor-associated autoantibodies (TAAb). With prior EDRN support, we developed high-throughput programmable protein display methods for the rapid detection and validation of autoantibody biomarker signatures in breast and lung cancers. Our breast cancer TAAb biomarkers have been licensed and integrated into Videssa™ Breast that is now available as CLIA-certified test. Our triple negative breast cancer markers have been validated in blinded phase 2 multicenter validation studies. These demonstrate the great utility of TAAb in cancer early detection. However, the sensitivities of most TAAbs is moderate and there is a suggestion that greater sensitivity and specificity could be obtained by examining TAAb directed at aberrantly modified proteins in cancers. Our central hypothesis is that aberrant protein glycosylation, a hallmark of breast and lung cancers, induces glycoprotein-specific TAAb that can be measured as specific serum biomarkers of these cancers. Alterations in glycosylation are highly immunogenic, and there is strong historical evidence for significant antibody responses to cancer-altered glycoproteins. However, all current protein (or polypeptide) display tools allow limited or no post-translational modification. This historical roadblock has prevented the identification of these biomarkers because of the lack of screening methods that test immunogenic structural glycoproteins. We introduce a tool for the high-throughput display of full-length proteins decorated with cancer-specific O-glycan structures. This will revolutionize the opportunity to screen glycan-protein epitopes in their natural context. Our team comprises strong expertise in functional proteomics, biomarker development, glycoproteomics, medical oncology and biostatistics. Targeted proteins will include: the extra-cellular domains of relevant single pass membrane proteins, proteins known to be O-glycosylated and overexpressed in the two cancers, and all known mucins. They will be translated in situ using human ribosomes and chaperone proteins and then systematically decorated with Tn and STn O-GalNAc-type glycans by consecutive addition of recombinant glycosyltransferases and sugar nucleotides to mirror what occurs in the two cancers. Adhering to the principles of PRoBE design, we will screen these arrays with cancer patient and control sera. Our study will focus on cancer patients and non-cancer subjects with positive imaging findings. Study design will include Phase I discovery (arrays/ELISA) and Phase II validation using ELISA.