Understanding scRNA-seq data from Abbasi et al. 2020 Cell Stem Cell
A new and interesting study was recently published by Abbasi et al. in Cell Stem Cell 2020 . This study presents data suggesting that large wound environments, which can regenerate hair follicles (WIHN), induces fibroblasts to differentiate into a specialized cell called dermal papilla (DP). Dermal papilla cells are unique and are thought to be a major cell type for inducing hair growth, cycling, and regeneration.
The authors utilized a novel mouse model ‘Hic1CreERt2’ that can genetically label what appears to be the reticular dermis with tdTomato labeling during neonatal development (P3/4). When the authors injured the Hic1CreERt2/tdTomato labeled adults with large wounds, the regenerated hair follicles were thought to contain tdTomato cells (Figure 3 above). The authors then performed an analysis of large wounds by scRNA-seq and indicated that the regenerating dermal papilla were also tdTomato positive. These results suggested that Hic1 cells acted as a progenitor for most fibroblasts subtypes in a wound including dermal papilla. Our re-analysis of the Abbasi et al scRNA-seq data provides a new and exciting look into their dataset.
Here, we show evidence by our re-analysis of the published scRNA-seq data suggesting that Hic1 lineages may not contribute to regenerating dermal papilla because they are negative for tdTomato. Our code to perform this analysis can be found on our Github page.
We have been collaborating with the Biernaskie lab who has graciously taken the time and effort to discuss possible explanations of why Hic1CreERt2-tdTomato lineage cells are sparsely detected in neogenic dermal papilla (DP) following large wounds. After rigorous discussion, two prevailing hypotheses of equal value have been agreed upon, as below:
Hypothesis 1: ROSA-CAG-driven reporter transgenes (i.e. tdTomato, YFP) may be transcriptionally downregulated in dermal papilla cells (please see our analysis page to compare UMAPS of sorted tdTomato separation to actual tdTomato expression).
Hypothesis 2: The Hic1CreERt2 mouse model may not specifically label distinct fibroblast lineages at P3/4 (Figure 2Q below from Abbasi et al.2020). To understand the significance of this revelation please read (Driskell et al. 2013).
We are happy to announce that the Driskell lab and the Biernaskie lab will be co-publishing our findings with an in-depth analysis and discussion, in an upcoming special issue of Experimental Dermatology focusing on fibroblast biology. Furthermore, we have reposted our analysis on tintyn.com. Please contact tintyn if you would like your data to be hosted there.
Stay tuned for the announcement of our manuscript.