Establishment Of A Stromal Cell Infiltration Intensity Score Model,and Identification Of Chemoresistance Genes And Chemosensitization Strategies Based On CRISPR/Cas9 Library Screening

Tumor microenvironment(TME)cells constitute a vital element of tumor tissue.Increasing evidence has indicated their clinicopathological significance in predicting outcomes and therapeutic efficacy.In colon cancer,the immunoscore model based on intensity IHC staining of CD3 and CD8 is considered an important way for estimating TME cell infiltration patterns.However,neglecting the assessment of stromal cell infiltration may limit immunoscore prediction accuracy,since our previous studies have found that the infiltration level of stromal cells is closely related to the prognosis and chemotherapeutic benefit of colon cancer patients.The objective of the current study was:i)the development of a novel model for assessing the stromal infiltration intensity;ii)evaluating the clinical significance of the novel model;iii)exploring the molecular mechanism related to the chemoresistance caused by stromal infiltration based on CRISPR/Cas9 library screening.The results are summarized as followed:1)We used the unsupervised hierarchical clustering method to classify tumour samples from GEO colon cancer databases into three different TME subtypes,based on absolute abundance of TME cell infiltration quantified from transcriptomic data.The following analysis revealed significant survival and chemotherapeutic response differences between the three subtypes.2)By consistently using the bootstrap and LASSO regression methods,we selected 30 genes to build a novel model,named as "stromal cell infiltration intensity score,SIIS",to identify patients with higher degree of stromal cell infiltration.The SIIS model could not only distinguish patients with different recurrence and mortality risks,but also effectively screened chemotherapy-responsive patients in both GEO and TCGA datasets.3)By applying the SIIS signature to the TCGA pan-cancer datasets,we could still observe that the SIIS value was significant positively correlated with stromal cell infiltration abundance and IC50 value of chemotherapeutic drugs,indicating a broad applicability of this signature in multiple types of solid tumors.4)We also retrieved two public immunotherapy datasets,and found that SIIS value decreased with the activation of inflammation.Moreover,the SIIS model enabled prediction of the outcomes immune checkpoint blockage therapy in urothelial carcinoma patients and melanoma patients5)We retrospectively analyzed the association between the treatment outcome data and transcriptomic profile of 11 metastatic colorectal cancer patients,who received treatment of mFOLFOX6 regimen with or without Bevacizumab.we also observed that higher SIIS value was correlated with worse clinical response status of Bevacizumab treatment.6)By using CRISPR-Cas9 activation library screening and multiple liner regression analysis,we identify GPX3 and PRICKLE2,as critical drivers for mediating chemoresistance in colon cancer patients with higher SIIS value.PRICKLE2 is a core molecules of WNT/PCP signaling pathway.We uncovered a significantly positive correlation between SIIS value and activation level of WNT/PCP pathway,and found that targeting WNT/PCP pathway using the Rho kinase inhibitor is an effective appoach to overcome chemoresistance in vivo.