Single cell transcriptomics reveal trans-differentiation of pancreatic beta cells following inactivation of the TFIID subunit Taf4

Abstract

Regulation of gene expression involves a complex and dynamic dialogue between transcription factors, chromatin remodelling and modification complexes and the basal transcription machinery. To address the function of the Taf4 subunit of general transcription factor TFIID in the regulation of insulin signalling, it was inactivated in adult murine pancreatic beta cells. Taf4 inactivation impacted the expression of critical genes involved in beta-cell function leading to increased glycaemia, lowered plasma insulin levels and defective glucose-stimulated insulin secretion. One week after Taf4-loss, single-cell RNA-seq revealed cells with mixed beta cell, alpha and/or delta cell identities as well as a beta cell population trans-differentiating into alpha-like cells. Computational analysis of single-cell RNA-seq defines how known critical beta cell and alpha cell determinants may act in combination with additional transcription factors and the NuRF chromatin remodelling complex to promote beta cell trans-differentiation.

Fig. 1. Inactivation of Taf4 in beta cells.

A Immunostaining of Langerhans islets from mice with the indicated genotypes for Taf4 or insulin as indicated. The number of weeks after Tam injection are indicated. B Immunostaining of Langerhans islets as above but with the addition of the Taf4-DAPI merge. Representative Taf4-positive nuclei at 34 and 55 weeks are indicated by arrows, not to be confounded with the stronger non-specific staining indicated by *. Scale bar = 100 μM.

Fig. 2. Physiological parameters of Taf4^b−/− animals.

A Weight of Taf4^b+/+ and Taf4^b−/− animals at the indicated number of weeks following Tam injection. B, C Blood glucose and insulin levels of Taf4^b+/+ and Taf4^b−/− animals at the indicated number of weeks following Tam injection. The * above the bars indicate the significant difference between Taf4^b+/+ and Taf4^b−/− animals at each time point, whereas * and NS above/below the lines indicate significant or non-significant differences between the Taf4^b−/− values at the indicated times. D ATP content of isolated islets from control animals or from Taf4^b−/− animals at the indicated number of weeks following Tam injection. Values are expressed as a % of the Taf4^b+/+ values at each time point taken as 100%. E Blood glucose levels at the indicated times following injection of glucose in control and Taf4^b−/− animals. Mice at 14 weeks were injected with Tam and glucose administered 3 weeks following Tam treatment. F Plasma insulin levels from the same protocol. T-test with two-tailed P-value analyses and confidence interval 95% were performed by Prism 5. P-values: *p < 0.05; **p < 0.01; ***p < 0.001. Data are mean ± SD from N = 8 or N = 3 as indicated.

Fig. 3. Effects of Taf4 inactivation on islet gene expression.

A Heatmap showing the expression of the 100 genes most up and downregulated 1 week after Taf4 inactivation. Right hand panels show the GSEA and ontology analyses of the de-regulated genes. B Heatmap showing the expression of the 100 genes most up and downregulated 5 weeks after Taf4 inactivation. Right hand panels show the GSEA and ontology analyses of the de-regulated genes.

Fig. 4. Sc-RNA-seq of WT and week 5 mutant islets.

A–C Violin plots of expression of the indicated genes in each cluster from WT islets (A), 5-week mutant islets (B) and the aggregate (C). D UMAP representations of the aggregate data, illustrating origin of cells from WT and mutant animals and the expression of the genes indicated in each panel.

Fig. 5. Sc-RNA-seq of week 1 mutant islets.

A Violin plots of expression of the indicated genes in each cluster. B UMAP (left) and tSNE (right) representations of the cell populations. C UMAPs illustrating the expression of the genes indicated in each panel.

Fig. 6. Transcription factors regulating cell state in week 1 mutant islets.

A SCENIC-based tSNE representation colouring cells based on the binary activities of the transcription factor regulons. B, D–E Binary activities of the transcription factor regulons in the different cell populations. C Heatmap of regulon activities in the different cell populations were quantified using AUCell. G A flow-chart of the progression of cell state from differentiated beta cells to trans-differentiated alpha cells. The SCENIC regulons marking the transitions of each state are indicated.