Current technologies for high-throughput single-cell RNA sequencing (scRNA-seq) are based upon stochastic pairing of cells and barcoded beads in nanoliter droplets or wells. They are limited by the mathematical principle of the Poisson statistics such that the utilization of either cells or beads or both is no more than ∼33%. Despite the versatile design of microfluidics or microwells for high-yield loading of beads that beats the Poisson limit, subsequent encapsulation of single cells is still determined by stochastic pairing, representing a fundamental limitation in the field of single-cell sequencing. Here, we present dTNT-seq, an integrated dielectrophoresis (DEP)-trapping-nanowell-transfer (dTNT) approach to perform cell trapping and bead loading both in a sub-Poisson manner to facilitate scRNA-seq. A larger-sized 50 μm microwell array was prealigned precisely on top of the 20 μm DEP nanowell array such that single cells trapped by DEP can be readily transferred into the underneath larger wells by flipping the device, followed by subsequent hydrodynamic bead loading and coisolation with transferred single cells. Using a dTNT device composed of 3600 electroactive DEP-nanowell units, we demonstrated a single-cell trapping rate of 91.84%, a transfer efficiency of 82%, and a routine bead loading rate of >99%, which breaks the Poisson limit for the capture of both cells and beads, thus called double-sub-Poisson distribution, prior to encapsulating them in nanoliter wells for cellular mRNA barcoding. This approach was applied to human (HEK) and mouse (3T3) cells. Comparison with a non-DEP-based method through gene expression clustering and regulatory pathway analysis demonstrates consistent patterns and negligible alternation of cellular transcriptional states by DEP. We envision the dTNT-seq device can be modified for studying cell-cell interactions and enable other applications requiring active manipulation of single cells prior to transcriptome sequencing.
Keywords: dielectrophoresis; microfluidic; nanowell; single-cell RNA sequencing; sub-Poisson distribution.