Gene-expression profiling of single cells from archival tissue

RNA sequencing (RNA-seq) is a sensitive and accurate method for quantifying gene expression. Small samples or those whose RNA is degraded, such as formalin-fixed, paraffin-embedded (FFPE) tissue, remain challenging to study with nonspecialized RNA-seq protocols. Here we present a new method, Smart-3SEQ, that accurately quantifies transcript abundance even with small amounts of total RNA and effectively characterizes small samples extracted by laser-capture microdissection (LCM) from FFPE tissue. Here we obtain distinct biological profiles from FFPE single cells, which have been impossible to study with previous RNA-seq protocols, and detect both gene-expression differences and copy-number alterations in single macrophages vs. ductal carcinoma in situ from a breast-cancer case. We propose Smart-3SEQ as a highly scalable method to enable large gene-expression profiling experiments unconstrained by sample size and tissue availability. In particular, Smart-3SEQ''s compatibility with FFPE tissue unlocks an enormous number of archived clinical samples, and combined with LCM it allows unprecedented studies of small cell populations and single cells isolated by their in situ context.