Single-cell RNA-seq analysis of cutaneous wound healing in mouse

Studying the transcriptomic response of different epidermal stem cell populations to wounding has been difficult due to intermixing of wound healing and homeostastic cells from different stem cell pools in bulk-cell sequencing setups. Here, we circumvent those problems by using a single-cell sequencing approach. We randomly sequenced the traced progeny of either Lgr5 or Lgr6 stem cells isolated from wounded or unwounded skin 0 day (control), 1 d, 4 d, 7 d, 10 d or more than 1 month after wounding. We then identified Lgr5 or Lgr6 wound cells using a computational approach. Wound cells were defined by using a negative binominal Naïve Bayes classifier with 0-day control cells (unwounded mice) as reference. Wound cell populations were defined by clustering wound cells in t-SNE space using k-means clustering.