Patterns of IgM Binding to Tumor Associated Antigen Peptides Correlate with the Type of Brain Tumors

Autor: Velizar Shivarov, M. Y. Hadzhieva, D. V. Ferdinandov, Anastas Pashov, A. V. Bussarsky, Victor Kostov
Rok vydání: 2020
Předmět:
Popis: The immune system can be used as a biosensor of the internal environment. Changes in the reactivities of the antibody repertoire can be used as a readout for a wide range of disturbances including various inflammatory conditions and malignant tumors. Extending our previous work based on IgM mimotope libraries, here we report our studies on the interpretability of profiles of IgM reactivities to a library of natural 15-mer peptides derived from 20 tumor associated antigens and 193 linear B cell epitopes involved in tumor pathogenesis. Sera from 21 patients with glioblastoma multiforme (GBM, n=10), brain metastases of other tumors (n=5) and non-tumor bearing neurosurgery patients (n=6) were used to probe their IgM reactivity with an array of 4526 peptide sequences. Using feature selection algorithms, we were able to extract profiles that separated well the three diagnostic groups with accuracy of up to 0.9. A key feature of the profiles extracted was their size (138 peptides for differentiating GBM and 340 – for tumor bearing patients) and origin from practically all tested antigens. Comparable numbers of reactivities were gained or lost in tumor bearing patients. A minimal set of the most significant 41 reactivities from 16 antigens contained disproportionately large number of epitopes from stromelysine-3 and erbB2 receptor with some of the reactivities gained and other lost in cancer patients. Epitopes from human papilloma virus 16 and HTLV-1 were included too. Some of the reactivities were readily interpretable both as antigen source and structural context (signal peptides). The interpretation of the rest requires further confirmatory studies. Thus, a set of natural peptides from tumor antigens readily provides profile of interpretable IgM reactivities which can serve as classifiers for clinically relevant patient stratification.
Databáze: OpenAIRE