Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Mar 28;57(6):799-819.e6.
doi: 10.1016/j.devcel.2022.03.001. Epub 2022 Mar 21.

Single-cell transcriptome analysis reveals three sequential phases of gene expression during zebrafish sensory hair cell regeneration

Affiliations

Single-cell transcriptome analysis reveals three sequential phases of gene expression during zebrafish sensory hair cell regeneration

Sungmin Baek et al. Dev Cell. .

Abstract

Loss of sensory hair cells (HCs) in the mammalian inner ear leads to permanent hearing and vestibular defects, whereas loss of HCs in zebrafish results in their regeneration. We used single-cell RNA sequencing (scRNA-seq) to characterize the transcriptional dynamics of HC regeneration in zebrafish at unprecedented spatiotemporal resolution. We uncovered three sequentially activated modules: first, an injury/inflammatory response and downregulation of progenitor cell maintenance genes within minutes after HC loss; second, the transient activation of regeneration-specific genes; and third, a robust re-activation of developmental gene programs, including HC specification, cell-cycle activation, ribosome biogenesis, and a metabolic switch to oxidative phosphorylation. The results are relevant not only for our understanding of HC regeneration and how we might be able to trigger it in mammals but also for regenerative processes in general. The data are searchable and publicly accessible via a web-based interface.

Keywords: hair cell death; hair cell regeneration; lateral line system; mechanosensory organ; neomycin; pseudotime analysis; scRNA-seq atlas; stem cells; support cells; zebrafish.

PubMed Disclaimer

Conflict of interest statement

Declaration of interests The authors declare no competing interests.

Figures

Figure 1.
Figure 1.. Spatio-temporal scRNA-Seq analysis during lateral line HC regeneration.
(A) 5dpf Tg(she:H2A-mCherry) larvae were used for FACS of lateral line cells. (B - D) Illustration of homeostatic lateral line cell types. (B and D) dorsal view and (C) lateral view. (E) Experimental timeline of scRNA-Seq sample collection. Maximum intensity projection of confocal images of Tg(myo6b:H2B-mScarlet-I/Et20:GFP) neuromasts during HC regeneration. This transgenic line shows the HCs more clearly than the Tg(she:H2A-mCherry) line that was used for sorting. Scale bar: 20μm, arrow indicates two new HCs. (F) Integration of scRNA-Seq samples of seven different time points in UMAP latent space. (G) UMAP of a cluster analysis identifies eight different neuromast cell types. (H) Heatmap visualizing the top 25 genes in each timepoint during HC regeneration. (Differential gene expression analysis of each time point vs. all other time points) (I - M) UMAPs of subsetted (I) central SCs, (J) A/P cells, (K) D/V cells, (L) HC progenitors and (M) young HCs with bar graphs showing the number of up and down-regulated genes during regeneration (differential gene expression analysis of each time point vs. homeostasis in individual cell type).
Figure 2.
Figure 2.. Downregulation of progenitor maintenance/placodal genes, Notch and Fgf- signaling 0min after HC death.
(A) Heatmap of selected cell type-specific downregulated genes 0min after neomycin treatment. (B) Dot plot of the average expression of progenitor/stem cell maintenance/placodal genes during HC regeneration. The size of the dots represents the proportion of cells expressing the gene. (C) ISH images of gata2b, gpx1a, and sox2 expression during HC regeneration. Bottom two rows: ISH of sox2 in siblings and atoh1a CRISPR mutants during homeostasis and 0min after neomycin treatment. (D and E) Dot plots of the average expression of Notch (D) and Fgf (E) pathway members during HC regeneration. (F) Enrichment term analysis of downregulated genes 0min after HC death compared to all other time points.
Figure 3.
Figure 3.. Injury/inflammatory response genes are rapidly upregulated 0min after HC death.
(A - B) Heatmap of the expression of selected genes upregulated in all SCs at (A) 0min and (B) 0–30min. (C - F) Heatmaps of genes upregulated in different cell types at 0min and 30min after HC death. (G and H) ISH images of (G) fosab and (H) nr4a1 in neuromasts at homeostasis, 0min, 30min, and 1h after HC death. (I and J) ISH images of (I) fosab and (J) nr4a1 expression in neuromasts of sibling and atoh1a CRISPR mutants at homeostasis and 0min after HC death. (K) Confocal images of Tg(she:H2A-mCherry) neuromast cells and transmitted light during laser ablation in 5dpf sibling and atoh1a CRISPR mutants. (L) ISH images of nr4a1 expression in laser-ablated and non-ablated neuromasts of sibling and atoh1a CRISPR mutants. (M - O) Dot plots of the relative expression of (M) AP-1 TF network members, (N) Glucocorticoid receptor and (O) IFN signaling in neuromasts during HC regeneration. (P) Gene enrichment term analysis for genes upregulated 0min after HC death compared to all other time points.
Figure 4.
Figure 4.. Upregulated genes 30min–1h after HC death.
(A) Heatmap of the average expression of selected upregulated genes in several SC types and HC progenitors at 30min-1h. (B - E) Heatmaps of the average expression of genes enriched in different cell types at 30min-1h after HC death. (F and G) ISH images of genes upregulated at 30min-1hr in all SCs (F) and central SCs (G). (H and I) Dot plots visualizing the average expression of (H) ‘regeneration enriched’ genes across regeneration time points and (I) neuromast cell types. (J and K) Dot plots of the average expression of RA pathway genes across (J) regeneration time points and (K) cell types. (L) Gene enrichment term analysis of up- and downregulated genes at 30min and 1h after HC death compared to all other time points.
Figure 5.
Figure 5.. Ribosome biogenesis, spliceosome, HC genes, Wnt, Notch and cell cycle genes are enriched 3–10h after HC death.
(A) Heatmap of the average expression of selected genes upregulated in several cell types at 3h-10h after HC death. (B - E) Heatmaps of the average expression of selected cell type-specific genes upregulated 3h-10h after HC death. (F) ISH images of nop58, dla, and mcm3 in neuromasts during HC regeneration. (G - K) Dot plots displaying the average expression per time point of (G) ribosome biogenesis, (H) spliceosome, (I) HC lineage, (J) Wnt signaling and (K) cell cycle genes. The dot size shows the percentage of cells that express the gene. (L) Cell cycle genes by cell type enrichment. Amplifying SCs omitted. (M - O) Gene enrichment term analysis of up- and downregulated genes at (M) 3h, (N) 5h, and (O) 10h after HC death compared to all other time points.
Figure 6.
Figure 6.. Pseudotime analysis (injury response + progenitor activation).
(A - C) UMAPs of pseudotime trajectories of subsetted central SC- and HC lineage cells. (A) Dark purple shows cells at the beginning of pseudotime and yellow cells at the end. (B) UMAP with the location of cells from different time points highlighted in different colors. (C) Cluster analysis groups cells with similar transcriptomes into ten different clusters. After the branch point, cells either follow the central SC lineage (blue arrow), or they follow the HC lineage (green arrow). (D) Feature plot of cd44a that is enriched in cells before the branch point. (E) gata2b is enriched in the central SC lineage. (F) rbm24a labels the HC lineage. (G) Heatmap of markers enriched in the injury response + progenitor activation population, partitioned into clusters. (H - J) Feature plots of genes expressed in the injury response + progenitor activation central SC population. (K and L) Line plots showing the scaled expression of fosab, irg1l, and lima1a along the pseudotime trajectory in the (K) central SC lineage and (L) HC lineage. Individual cells are represented by color-coded tics on the x-axis.
Figure 7.
Figure 7.. Pseudotime analysis (central SC lineage after the branch point).
(A) Heatmap of the average expression of genes selectively enriched in the central SC lineage, partitioned by Louvain clusters. (B) Feature plot of sox3 and list of other expressed genes. (C) Feature plot of hey1 and list of Notch pathway members that are also expressed. (D) Line plots of the scaled expression of sox3, prox1a, and hey1 along the pseudotime trajectory in cells of the central SC lineage and HC lineage cells. Individual cells are represented by color-coded tics on the x-axis.
Figure 8.
Figure 8.. Pseudotime analysis (HC lineage after the branch point).
(A) Heatmap of the average expression of genes selectively enriched in HC lineage cells, partitioned by Louvain clusters. (B-D) Feature plots of sox4a (B), atoh1a (C), dld (D). (E and F) Fluorescent ISH (HCR) images of sox4a (E) and atoh1a (F) with DAPI in homeostasis, 1h and 3h after HC death. Scale bar: 20μm. (G and H) Line plots of sox4a, atoh1a, and dld scaled expression along the pseudotime trajectory in the (G) central SC lineage and (H) HC lineage. Individual cells are represented by color-coded tics on the x-axis. (I-L) ISH images of her4.1, cdkn1bb, sox4a, and atoh1a in 5dpf DMSO- or LY411575-treated (to inhibit Notch signaling) neuromasts. Larvae were treated for 6h before fixation. (M and N) Line plots of the scaled expression of atoh1a, insm1b, myo6b, gadd45gb.1, rbm24a, and gfi1aa along the pseudotime trajectory in the (M) central SC lineage and (N) HC lineage. Individual cells are represented by color-coded tics on the x-axis. (O and P) Line plots of dlb, dld, dlc, hes2.2, pcna, and stmn1a scaled expression along the pseudotime trajectory in the (O) central SC lineage and (P) HC lineage. (Q) UMAP of the distribution of HC populations comprised of HC prog., young HCs and mature HCs. (R - T) UMAP of the distribution of 3h, 5h, and 10h HC progenitor populations, respectively.
Figure 9.
Figure 9.
Overview of the molecular processes occurring during HC regeneration.

Similar articles

Cited by

References

    1. Appel B, and Eisen JS (1998). Regulation of neuronal specification in the zebrafish spinal cord by Delta function. Development 125, 371–380. - PubMed
    1. Bahrami S, and Drablos F (2016). Gene regulation in the immediate-early response process. Adv Biol Regul 62, 37–49. - PubMed
    1. Basch ML, Brown RM 2nd, Jen HI, and Groves AK (2016). Where hearing starts: the development of the mammalian cochlea. J Anat 228, 233–254. - PMC - PubMed
    1. Beauchemin M, Smith A, and Yin VP (2015). Dynamic microRNA-101a and Fosab expression controls zebrafish heart regeneration. Development 142, 4026–4037. - PMC - PubMed
    1. Behra M, Bradsher J, Sougrat R, Gallardo V, Allende ML, and Burgess SM (2009). Phoenix is required for mechanosensory hair cell regeneration in the zebrafish lateral line. PLoS Genet 5, e1000455. - PMC - PubMed

Publication types