Bioinformática aplicada ao desenvolvimento de drogas contra o câncer gástrico

Abstract

Gastric cancer (GC) represents a significant public health challenge, being ranked as the fifth most common malignancy and the fourth leading cause of cancer-related mortality worldwide. In 2020, the year in which the latest estimates were released by Globocan, there were an estimated 1,089,103 new cases and 768,793 deaths from gastric cancer. These data highlight the urgent need for advances in treatment to combat GC. Although treatment has gradually improved in recent decades, the median survival of patients with advanced GC is still very low, less than 12 months. The heterogeneity of tumors, resistance to chemotherapy agents, lack of selectivity and, consequently, toxicity to patients' healthy cells limit the treatment of this neoplasm and contribute to the prognosis of gastric carcinogenesis being quite challenging. In this context, it is urgent to search for molecules with low levels of toxicity that explore new therapeutic targets in cancer. To this end, microarray transcriptome studies combined with the SBVS technique (Structure-Based Virtual Screening) show promise. Additionally, another important strategy is the repositioning of synthetic antimicrobial peptides (SAMPs) for the treatment of gastric cancer. These represent an alternative therapeutic intervention for GC as they are strategically designed, based on protein sequences, to have low levels of toxicity, multifaceted performance and high selectivity. Therefore, the present work had two main objectives (1) to develop, through bioinformatics, an efficient pipeline to identify new therapeutic targets in gastric cancer, in addition to searching for inhibitory compounds for the identified targets; (2) carry out in silico, in vitro and in vivo studies to investigate the repositioning of SAMPs as a potential therapeutic approach in GC. In the first part of this work, data were collected in the Gene Expression Omnibus (GEO) database from three arrays of data sets that analyzed gastric tumor tissue versus normal gastric tissue, performed by microarrays using the GPL570 platform. Data were analyzed using the GEPIA tool for differential expression and KMPlot for patient survival analysis. For greater robustness, GC data from the TCGA database (The Cancer Genome Atlas Program) were used to corroborate the GEO analysis. The genes found in the in silico analysis, through GEO and TCGA, were confirmed in different GC cell lines by RT-qPCR. The AlphaFold database was used to find the 3D structure of the proteins encoded by the overexpressed genes. Then, the SBVS technique was used to find molecules with high binding affinity to the selected proteins, and, therefore, molecular docking analysis was performed using the DockThor server. In the second part of this work, the knowledge acquired through bioinformatics studies was used to evaluate the repositioning of synthetic antimicrobial peptides against gastric cancer. Using the sequences of synthetic peptides, it was possible to conduct a predictive analysis on their anticancer potential. Furthermore, using bioinformatics, a toxicity prediction study was carried out with different parameters, including LD50. The peptides selected by in silico tests were tested through in vitro experiments against several GC strains. The peptides that demonstrated an IC50 against gastric tumor cells were then used in studies of mechanisms of action, which included assessments of increased permeability and formation of pores in the membrane, morphological analyzes by atomic force microscopy (AFM) and inhibition of migration cell phone. Still in vitro, toxicity tests were carried out against healthy human cells to evaluate the safety of the peptide(s) with anticancer action. In vivo toxicity tests against zebrafish were performed to evaluate the potential application of this peptide(s) in GC therapy. Finally, the selected peptide(s) were subjected to molecular docking assays to evaluate their interaction with the new targets identified in part I of this work and overexpressed in the AGP-01 strain. Regarding the results, in the first part of the work, in silico and RT-qPCR (in vitro) analyzes confirmed the high expression of the AJUBA, CD80 and NOLC1 genes in gastric tumor lines. Subsequently, the 3D structures of the proteins corresponding to these genes were used in the SBVS analysis,

resulting in the selection of three molecules, one for each target: MCULE-2386589557-0-6, MCULE-9178344200-0-1 and MCULE-5881513100- 0-29. All molecules demonstrated favorable pharmacokinetic, pharmacodynamic and toxicological properties. Molecular docking analysis revealed that the molecules interacted with target proteins at sites critical for their activity, altering their three-dimensional structures and, possibly, their function in gastric tumor cells. In the second part of the study, six synthetic peptides were evaluated by bioinformatics in relation to their anticancer potential and obtained positive prediction results. Thus, all of them were subjected to in vitro assays against different gastric cancer cell lines. Four of these peptides showed IC50 values against GC cells. Among these four, the RcAlb-PepI peptide demonstrated a low IC50 value (53.87 μg/mL) against the metastatic GC line AGP-01. Furthermore, RcAlb-PepI exhibited low toxicity in healthy gastric cells (MNP-01) compared to the other peptides. This peptide induced an increase in membrane permeability, followed by the

formation of a 6 kDa pore in the AGP-01 cell membrane. AFM analyzes revealed that RcAlb- PepI was able to modify the ultrastructural and nanomechanical properties of AGP-01 cells,

resulting in increased roughness, adhesion, cell area and volume, as well as reduced cellular plasticity and cytoskeletal organization. These data corroborated the inhibitory effect, observed in the cell migration assay, of RcAlb-PepI on the AGP-01 cell. In vitro and in vivo toxicity results showed that RcAlb-PepI does not exhibit toxicity against healthy zebrafish cells, embryos or larvae, confirming its safety. The molecular docking assay demonstrated that RcAlb-PepI showed a strong interaction with the AJUBA and NOLC1 proteins, identified in part I of this work and overexpressed in the AGP-01 strain. In conclusion, this work presents a pipeline built by bioinformatics that proved to be efficient in identifying new therapeutic targets in GC and inhibitory compounds for the identified targets, with the possibility of extending this

pipeline to other types of cancer. Furthermore, it reveals the potential application of RcAlb- PepI against the metastatic gastric lineage AGP-01 and provides insights into the mechanism

of action of this synthetic peptide behind its anticancer activity.