RNase H-dependent PCR-enabled T-cell receptor sequencing for highly specific and efficient targeted sequencing of T-cell receptor mRNA for single-cell and repertoire analysis

Shuqiang Li, Jing Sun, Rosa Lundbye Allesøe, Krishnalekha Datta, Yun Bao, Giacomo Oliveira, Juliet Forman, Roger Jin, Lars Rønn Olsen, Derin B. Keskin, Sachet A. Shukla, Catherine J. Wu, Kenneth J. Livak

Research output: Contribution to journalJournal articleResearchpeer-review


RNase H-dependent PCR-enabled T-cell receptor sequencing (rhTCRseq) can be used to determine paired alpha/beta T-cell receptor (TCR) clonotypes in single cells or perform alpha and beta TCR repertoire analysis in bulk RNA samples. With the enhanced specificity of RNase H-dependent PCR (rhPCR), it achieves TCR-specific amplification and addition of dual-index barcodes in a single PCR step. For single cells, the protocol includes sorting of single cells into plates, generation of cDNA libraries, a TCR-specific amplification step, a second PCR on pooled sample to generate a sequencing library, and sequencing. In the bulk method, sorting and cDNA library steps are replaced with a reverse-transcriptase (RT) reaction that adds a unique molecular identifier (UMI) to each cDNA molecule to improve the accuracy of repertoire-frequency measurements. Compared to other methods for TCR sequencing, rhTCRseq has a streamlined workflow and the ability to analyze single cells in 384-well plates. Compared to TCR reconstruction from single-cell transcriptome sequencing data, it improves the success rate for obtaining paired alpha/beta information and ensures recovery of complete complementarity-determining region 3 (CDR3) sequences, a prerequisite for cloning/expression of discovered TCRs. Although it has lower throughput than droplet-based methods, rhTCRseq is well-suited to analysis of small sorted populations, especially when analysis of 96 or 384 single cells is sufficient to identify predominant T-cell clones. For single cells, sorting typically requires 2-4 h and can be performed days, or even months, before library construction and data processing, which takes ~4 d; the bulk RNA protocol takes ~3 d.
Original languageEnglish
JournalNature protocols
Pages (from-to)2571-2594
Publication statusPublished - 2019

Cite this