Research On Dielectrophoresis Trapping-microwell Transfer Based Single Cell Sequencing Technology And Its Application In CAR-T Immunotherapy

Single cell m RNA sequencing(sc RNA-seq)has become a powerful tool to elucidate cell-to-cell heterogeneity,identify new cell types,reconstruct organ developmental landscape and understand tumor microenvironment.The established high-throughput sc RNA-seq platforms isolate single cells in a passive and stochastic way,thereby preventing further increasing their cell capture efficiency and limiting their application in studying precious clinical samples.Dielectrophoresis(DEP)has been broadly used in single cell trapping,sorting and pairing attributed to its ability for precise cell manipulation.However,the size of DEP microwells(10～20μm)dedicated for single-cell capture is not sufficient for accommodation of barcoded sequencing beads(40～50μm),hence the extension of the success in DEP single-cell trapping to sc RNA-seq is still challenging.By utilizing a “DEP trapping-microwell transfer” approach,this dissertation develops an active microfluidic sc RNA-seq platform,DTMT-seq,and applies it to characterize clinical Chimeric antigen receptor-modified T(CAR-T)cells derived from patients treated with this immunotherapy.The core innovative work includes:The microfluidic array chip was designed.Through a pre-aligned double-layer construct,cell trapping and beads loading were separately implemented to enable a successful “Sub-Poisson” proactive single cell isolation for sc RNA-seq.A DTMT-seq chip composed of 3600 electroactive microwell units was fabricated,and an optical surface profiler was used to qualify the fabrication.Additionally,the impact of microwell depth on single cell trapping performance was analyzed.Under a doublet rate of less than 2%,DTMT-seq demonstrated a single-cell trapping rate of 91.84%,a transfer efficiency of82%,and a routine bead loading rate of 99%.The stringent human-mouse species mixture experiment was conducted to validate the sequencing performance of the device,and a comparable gene and transcript detection capability with current mainstream sc RNA-seq platforms was achieved.GSEA enrichment analysis verified that DTMT-seq can identify subtle differences in gene expression patterns underlying genotypically homogeneous samples.Through comparison with non-DEP microwell-based method in terms of the major biological activities of single cells,it is certified that the use of DEP electric field has negligible alternation on cell functional states at the transcriptional level.The overall performance of DTMT-seq in analyzing CAR-T cells generated from relapsed/refractory acute lymphoblastic leukemia patients was systematically evaluated.We experimentally verified the consistency of different chips in analyzing a group of identical CAR-T cells,confirmed its ability in multi-omics co-detection of m RNA and cellular protein expression,and demonstrated its capability of precisely dissecting the CAR-T intrinsic heterogeneities in a multi-condition/multi-sample setting.The genetic information of over 100,000 single cells from 48 samples was obtained employing DTMT-seq,which demonstrated its capacity in profiling large-scale clinical specimens.The in-depth analyses of sequencing data verified that DTMT-seq was capable of delineating the intrinsic mechanisms of CAR-T cells and correlating them with clinical responses.Through unbiased cytokine co-expressed module identification,unsupervised clustering and canonical pathway analyses,we analyzed the molecular characteristics,phenotypic composition and functional states of CAR-T pr-infusion products upon CD19 antigen-specific stimulation in comparison with non-target stimulation,T cell receptor-mediated activation and unstimulated controls.The cross comparison with published literatures and fundamental immunology confirmed the accuracy of DTMTseq in dissecting gene profiles and its potential in identifying new gene targets or crucial signaling pathways.