The preclinical study of the mechanism of action of anticancer small molecules is challenging due to the complexity of cancer biology and the fragmentary nature of available data. With the aim of identifying a protein subset characterizing the cellular activity of anticancer peptides, we used differential mass spectrometry to identify proteomic changes induced by two peptides, LR and [d-Gln(4)]LR, that inhibit cell growth and compared them with the changes induced by a known drug, pemetrexed, targeting the same enzyme, thymidylate synthase. The quantification of the proteome of an ovarian cancer cell model treated with LR yielded a differentially expressed protein data set with respect to untreated cells. This core set was expanded by bioinformatic data interpretation, the biologically relevant proteins were selected, and their differential expression was validated on three cis-platinum sensitive and resistant ovarian cancer cell lines. Via clustering of the protein network features, a broader view of the peptides' cellular activity was obtained. Differences from the mechanism of action of pemetrexed were inferred from different modulation of the selected proteins. The protein subset identification represents a method of general applicability to characterize the cellular activity of preclinical compounds and a tool for monitoring the cellular activity of novel drug candidates.

Hepatocellular carcinoma (HCC) ranks fifth in frequency worldwide amongst all human cancers causing one million deaths annually. Despite many promising treatment options, long-term prognosis remains dismal for the majority of patients who develop recurrence or present with advanced disease. Notch signaling is an evolutionarily conserved pathway crucial for the development and homeostasis of many organs including liver. Herein we showed that aberrant Notch1 is linked to HCC development, tumor recurrence and invasion, which might be mediated, at least in part, through the Notch1-E-Cadherin pathway. Collectively, these findings suggest that targeting Notch1 has important therapeutic value in hepatocellular carcinoma. In this regard, comparative analysis of the secretome of HepG2 and HepG2 Notch1 depleted cells identified novel secreted proteins related to Notch1 expression. Soluble E-Cadherin (sE-Cad) and Thrombospondin-1 (Thbs1) were further validated in human serum as potential biomarkers to predict response to Notch1 inhibitors for a tailored individualized therapy.

Proteomics and bioinformatics are a useful combined technology for the characterization of protein expression level and modulation associated with the response to a drug and with its mechanism of action. The folate pathway represents an important target in the anticancer drugs therapy. In the present study, a discovery proteomics approach was applied to tissue samples collected from ovarian cancer patients who relapsed after the first-line carboplatin-based chemotherapy and were treated with pemetrexed (PMX), a known folate pathway targeting drug. The aim of the work is to identify the proteomic profile that can be associated to the response to the PMX treatment in pre-treatement tissue. Statistical metrics of the experimental Mass Spectrometry (MS) data were combined with a knowledge-based approach that included bioinformatics and a literature review through ProteinQuest™ tool, to design a protein set of reference (PSR). The PSR provides feedback for the consistency of MS proteomic data because it includes known validated proteins. A panel of 24 proteins with levels that were significantly different in pre-treatment samples of patients who responded to the therapy vs. the non-responder ones, was identified. The differences of the identified proteins were explained for the patients with different outcomes and the known PMX targets were further validated. The protein panel herein identified is ready for further validation in retrospective clinical trials using a targeted proteomic approach. This study may have a general relevant impact on biomarker application for cancer patients therapy selection.

No abstract