%0 Generic
%A Kasem, Mhd Haidar
%C Heidelberg
%D 2025
%F heidok:34798
%R 10.11588/heidok.00034798
%T Intra-clonal gene expression heterogeneity in human   colorectal cancer
%U https://archiv.ub.uni-heidelberg.de/volltextserver/34798/
%X Colorectal cancer (CRC) harbors heterogeneous subclonal cell populations that contribute to   tumor progression, metastasis and therapy resistance. Subclonal heterogeneity arises not only   from genetic and epigenetic differences, but also from functional heterogeneity of subclones   sharing homogeneous genetic features. Genetic barcoding approaches have shown that   subclones with long-term tumor-initiating cell (LT-TIC) activity are capable of fueling CRC   expansion to metastasis and that CRC subclones can enter a drug tolerant persister state to   tolerate chemotherapy. Nevertheless, subclonal transcriptional features which characterize   functional states, such as treatment response and metastasis, are underexplored. Characterizing   how subclonal gene expression contributes to functional states of CRC enhances understanding  the dynamics of tumor progression. Here, I describe subclonal dynamics in addition to associated   single cell transcriptomics of treatment response and metastasis in barcoded CRC patient-derived   organoids (PDOs) and xenografts, respectively.   To study subclonal gene expression heterogeneity in metastatic and therapy resistant populations,   I transduced three CRC PDOs with a genetic RNA-expressed barcoding library for unique single   cell marking. Barcoded PDOs were either orthotopically serially transplanted into NOD.Cg-Prkdc   scid Il2rg tm1Wjl /SzJ (NSG) mice or treated with chemotherapy to study clonal and transcriptional   dynamics in tumor progression to metastasis and in treatment response. I performed single cell   RNA (scRNA) sequencing in addition to target barcode DNA sequencing on cells harvested from   xenografts or treated PDOs to analyze clonal dynamics and subclonal gene expression programs.   I found that combined scRNA data from more than 3.5x10E5 cells in xenografts demonstrated a   drop in the number of clones by which 0.05-1% of detected clones formed the majority of cancer   cells in serial transplantations of tumor, liver and lung metastasis, with an enrichment of few   individual clones that drive tumor progression. The frequency of these enriched clones varied   across serial transplantations of different tumor locations such as the tumor and metastases,   suggesting a dynamic subclonal expansion in different microenvironments. Subclonal cell   populations in both tumor and liver or lung metastases reflected distinct gene expression profiles   of differentiated and undifferentiated intestinal cells. Differentiated epithelial profiles were   reminiscent of epithelial intestinal cells such as absorptive enterocytes and secretory cells while   undifferentiated cells shared transcriptional features of invasive and fetal/embryonic cell states.   Individual metastatic subclones of liver metastasis showed higher expression of particular genes   such as SOX4, VEGFA and PLCG2 which were enriched in the invasive fetal cell phenotype. Lung   metastasis cells of the same subclones demonstrated higher expression of S100A11, ANXA1 and   KRT19, which were elevated in the differentiated epithelial cell phenotype. In addition, clones   which represent more than 25% of the tumor or metastasis cell population did not show   reproducible gene expression differences in comparison to smaller clones forming less than 5%   of the cell population. Upon chemotherapy treatment and recovery of PDOs, the frequency of   individual clones was marginally affected in comparison to control DMSO treated cells, without a   strong selection of particular clones. In response to treatment, cells reflected abundant pre-existing   gene expression profiles reminiscent to those of intestinal enterocytes, undifferentiated cycling   and progenitor cells and a low frequency of transcriptional profiles related to proliferative and   secretory cells. One week after drug washout, stem cell transcriptional clusters, in addition to   transcriptional clusters abundant in treatment response, were enriched. The functional relevance   of 181 marker genes of treatment response-associated transcriptional clusters was assessed   through a custom clustered regularly interspaced palindromic repeats interference (CRISPRi)  screen, which demonstrated the depletion of 43 out of 52 significantly affected genes after   chemotherapy treatment. Further, transposase-accessible chromatin single cell (scATAC)  sequencing data from two PDOs treated with 5-Fluorouracil (5-FU) showed an open chromatin   status in the region of genes detected to be differentially expressed in response-associated   transcriptional clusters identified in scRNA data, thus linking gene transcription to epigenetic   chromatin regulation.  Overall, this study links clonal dynamics to transcriptional states, thereby dissecting gene   expression of subclonal functional states of tumor progression. This analysis reveals that gene   expression states and marker genes associated with metastasis and treatment response could be   implemented in further development and validation of innovative treatment strategies in CRC.