Figure 7.
Coding gene repression stalls or aborts differentiation, while lncRNA gene repression permits a greater diversity of cell states
(A) Cell states associated with target perturbations, on UMAP.
(B) Normalized density heatmaps of positive and negative coding and lncRNA hits with similar magnitudes of effect from the genome-wide screens.
(C) Heatmap of gene expression signatures of cell states observed in (B), with examples of enriched ontology terms (FDR < 0.01).
(D) Pie charts showing the proportion of negative hits classified as non-productive or multiple trajectories. Coding and lncRNA hits were significantly different in the proportion of these phenotypes. ∗p < 1 × 10−5 by Fisher’s exact test.
(E) Diagram of internally controlled differentiation assay, with sgRNA-containing cells labeled by GFP or RFP.
(F) Overall percentage of GFP and RFP cells that express TAGLN protein (n = 3 replicates per condition). GFP/RFP double-positive cells (expressing both sgRNAs) were excluded. Error bars denote 1 SD. ∗p < 0.01 by t test.
(G) Relative expression of SERTAD4-AS1 and SERTAD4 coding gene in Perturb-seq experiment in cells with sgSERTAD4-AS1 compared with sgControl. Compared with control cells, complete knockdown was achieved for SERTAD4-AS1, with no significant change in SERTAD4 coding gene expression. Error bars show mean ± SEM (n = 6 pseudobulk samples). ∗p = 0.004 by t test.