c-Src Increases the Sensitivity to TKIs in the EGFR-Mutant Lung Adenocarcinoma

Exosomal lncCRLA is predictive for the evolvement and development of lung adenocarcinoma

Lung adenocarcinoma, the most common histological subtype of non-small cell lung cancer, exhibits heterogeneity that enables adaptability, limits therapeutic success, and remains incompletely understood. Our team uncovers that lncRNA related to chemotherapy resistance in lung adenocarcinoma (lncCRLA) is preferentially expressed in lung adenocarcinoma cells with the mesenchymal phenotype. lncCRLA can not enhance chemotherapy resistance in lung adenocarcinoma due to its binding to RIPK1 in exosomes, which is released into intercellular media and transferred by exosomes from mesenchymal-like to epithelial-like cells. However, plasmatic lncCRLA corresponding to tissue lncCRLA functions as a preferred biomarker to reflect the response to chemotherapy and disease progression of lung adenocarcinoma. Through single-cell sequencing, RNA-Mutect technique and spatial transcriptomics, a handful of hybrid EMT cells with elevated lncCRLA are characterized as the origin of lung adenocarcinoma, which are indiscriminated from hybrid EMT cells by the in-depth sequencing. Plasmatic lncCRLA is properly predictive for the preinvasive lesion of lung adenocarcinoma that would evolve to invasive lesion. That notion is confirmed by a brand-new transgenic mouse model in which EMT is tracked by Cre and Dre system. Dasatinib is potential to hinder the spontaneous progression from preinvasive to invasive lesion of lung adenocarcinoma. Together, plasmatic lncCRLA is defined as a brand-new circulating biomarker to predict the occurrence and evolvement of lung adenocarcinoma, a light for early detection of lung adenocarcinoma.

Exosomes secreted by metastatic cancer cells promotes epithelial mesenchymal transition in small cell lung carcinoma: The key role of Src/TGF-β1 axis

Exosome-mediated epithelial mesenchymal transition (EMT) is key to cancer metastasis. c-Src is involved in the secretion of exosomes and initiation of EMT. Effects of exosomes from metastatic non-small cell lung carcinoma (NSCLC) cells on the EMT process in primary NSCLC cells were assessed. Levels of c-Src in NSCLC tissues were detected and the influence of exosomes from metastatic NSCLC cells on the exosome secretion and EMT process in primary NSCLC cells was assessed. The expression of c-Src was modulated, and the influence on the secretion of exosomes and EMT initiation was evaluated. The level of c-Src was higher in NSCLC specimen and NSCLC cells with promoted EMT process. The suppression of c-Src inhibited secretion of exosomes. Exosomes from metastatic NSCLC cells enhanced migration and invasion abilities of primary NSCLC cells, which had identical effects to c-Src overexpression. The suppression of c-Src inhibited growth and metastasis of solid tumors as well as secretion of exosomes, while the injection of exosomes with c-Src overexpression promoted lung metastasis. TGF-β1 restored the invasion and migration abilities even with c-Src knockdown. The exosomes from metastatic NSCLC cells with high c-Src expression of can increase c-Src level in primary NSCLC cells, contributing to the promoted EMT process through TGF-β1 pathway.

DEML: Drug Synergy and Interaction Prediction Using Ensemble-Based Multi-Task Learning

Synergistic drug combinations have demonstrated effective therapeutic effects in cancer treatment. Deep learning methods accelerate identification of novel drug combinations by reducing the search space. However, potential adverse drug-drug interactions (DDIs), which may increase the risks for combination therapy, cannot be detected by existing computational synergy prediction methods. We propose DEML, an ensemble-based multi-task neural network, for the simultaneous optimization of five synergy regression prediction tasks, synergy classification, and DDI classification tasks. DEML uses chemical and transcriptomics information as inputs. DEML adapts the novel hybrid ensemble layer structure to construct higher order representation using different perspectives. The task-specific fusion layer of DEML joins representations for each task using a gating mechanism. For the Loewe synergy prediction task, DEML overperforms the state-of-the-art synergy prediction method with an improvement of 7.8% and 13.2% for the root mean squared error and the R2 correlation coefficient. Owing to soft parameter sharing and ensemble learning, DEML alleviates the multi-task learning 'seesaw effect' problem and shows no performance loss on other tasks. DEML has a superior ability to predict drug pairs with high confidence and less adverse DDIs. DEML provides a promising way to guideline novel combination therapy strategies for cancer treatment.