Critical genetic program for Drosophila imaginal disc regeneration revealed by single-cell analysis

Whether regeneration is primarily accomplished by re-activating gene regulatory networks used previously during development or by activating novel regeneration-specific transcriptional programs remains a longstanding question. Currently, most genes implicated in regeneration also function during development. Using single-cell transcriptomics in regenerating Drosophila wing discs, we identified two regeneration-specific cell populations within the blastema. They are each composed of cells that upregulate multiple genes encoding secreted proteins that promote regeneration. In this regenerative secretory zone, the transcription factor Ets21C controls the expression of multiple regenerationpromoting genes. While eliminating Ets21C function has no discernible effect on development, it severely compromises regeneration. This Ets21C-dependent gene regulatory network is also activated in blastema-like cells in tumorous discs, suggesting that pro-regenerative mechanisms can be co-opted by tumors to promote aberrant growth.

A long-standing question in the field of regenerative biology is whether regeneration is mainly accomplished by reactivation of gene regulatory networks (GRNs) used during earlier stages of development or, alternatively, by GRNs that are specifically activated during regeneration. Studies of regenerating tissues have provided evidence for cellular states that are not observed during normal development and for patterns of gene expression that seem specific for regeneration (for example Gerber et al., 2018;Aztekin et al., 2019). However, so far, there is little evidence for genes that are needed for regeneration but not for normal development.
To identify transcriptional programs initiated during regeneration, we examined regeneration of Drosophila larval wing imaginal discs, the epithelial tissues that differentiate into the adult wings and thorax. Imaginal discs are capable of regenerating after damage through the formation of a blastema, defined by localized proliferation and increased cellular plasticity (reviewed by Worley et al., 2012). To search for regeneration-specific GRNs, we compared regenerating and developing wing discs using single-cell transcriptomics. Tissue damage was induced by temporarily expressing the pro-apoptotic TNF ortholog eiger within the wing pouch, the portion of the disc that generates the wing blade (Smith-Bolton et al., 2009) (supplemental fig. 1). Subsequent regeneration occurs by localized cell proliferation and cell-fate re-specification. We collected wing discs after 24 hours of regeneration, approximately one third of the way through the regenerative process, and sequenced a total of 14,320 cells from two biological replicates, with an average of >3,000 genes detected per cell. Three major cell types were identified: epithelial cells, myoblasts, and hemocytes (supplemental fig. 2). Since imaginal disc regeneration is driven by epithelial cell proliferation (Smith-Bolton et al., 2009;Worley et al., 2012), we focused further analysis on these cells.
To identify potential regeneration-specific GRNs, we harmonized data from epithelial cells from regenerating discs with our previously collected data from undamaged discs (Everetts et al., 2021) using scVI (Gayoso et al., 2021) (see . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi. org/10.1101org/10. /2021 Materials and Methods) (Figure 1A, B). We assigned cell clusters to specific subregions of the wing disc epithelium based on the expression of known marker genes (Bageritz et al., 2019;Deng et al., 2019;Zappia et al., 2020;Everetts et al., 2021) (Figure 1B, C; supplemental fig. 3). As expected, cell clusters with pouch identity were underrepresented in the regenerating sample, as this portion of the tissue was ablated (supplemental fig. 3). From our single-cell analysis, we observed two clusters, denoted Blastema1 and Blastema2, that were almost exclusively composed of cells from the regenerating sample (181/186 and 519/564 cells, respectively) ( Figure 1B; supplemental fig. 3). Within these two regeneration-specific clusters, we observed the upregulation of genes known to be induced around the site of damage, including the Wnt ligands wingless (wg) and Wnt6 (Harris et al., 2016), Matrix metalloproteinase 1 (Mmp1), and Insulin-like peptide 8 (Ilp8) (Figure 1C, D).
Both Blastema1 and Blastema2 clusters express Ilp8, which is strongly upregulated around the site of damage in the regenerating disc ( Figure 1F, G). However, Blastema2 showed a higher expression of hinge-identity markers, such as transcription factor Zn finger homeodomain 2 (zfh2), than Blastema1 ( Figure 1C), suggesting that these cells might occupy a more proximal (outer) position. Indeed, in regenerating tissue, we observed higher Zfh2 expression in the outer ring of Ilp8expressing cells ( Figure 1G). In contrast, Blastema1 cells expressed higher levels of the unpaired (upd1, upd2, upd3) ligands, asperous (aspr), and PDGF-and VEGFrelated factor 1 (Pvf1) (Figure 1C, E). The Upd ligands activate the JAK/STAT pathway, which is important for cellular plasticity and regeneration (Katsuyama et al., 2015;Santabarbara-Ruiz et al., 2015;La Fortezza et al., 2016;Worley et al., surrounded by cells that express Ilp8 but not Pvf1 or upd3. Thus, the Blastema1 cells are located at the center of the blastema and are surrounded by Blastema2 cells; cells in both regions secrete ligands, some of which are known to promote regeneration, and are likely acting on the surrounding tissue. We refer to these regions together as the regenerative secretory zone. From our single-cell analysis, we used gene signatures to determine that the cells within the regenerative secretory zone were in an intermediate state between hinge and pouch identities ( Figure 1J; supplemental fig. 5). This finding suggested that these cells were derived from the surrounding inner-hinge region and were in the process of acquiring more distal pouch fates. To investigate this process, we examined the location of proliferating cells and found high levels of EdU incorporation surrounding the regenerative secretory zone ( Figure 1K) (Cosolo et al., 2019). As regeneration proceeded, the EdU incorporation extended more centrally to occur within the regenerative secretory zone ( Figure 1L). To determine if these proliferating cells are reprogrammed during regeneration to replace the ablated pouch, we performed a lineage-tracing experiment with an enhancer that is normally only expressed in a ring of cells of the inner-hinge. In the absence of pouch ablation, these cells and their progeny remain confined to the hinge ( Figure 1M, N). However, after regeneration following pouch ablation, most of the regenerated pouch was derived from cells that once expressed this enhancer ( Figure 1O). Thus, the ablated pouch is regenerated by the proliferation and reprogramming of more proximally fated inner hinge cells, likely driven by the ligands secreted by the regenerative secretory zone ( Figure 1P).
To search for a regulator of these regeneration-specific transcriptional changes, we analyzed our single-cell data for a transcription factor that was specifically expressed within the blastema cells. We found that Ets at 21C (Ets21C) was specifically expressed during regeneration, primarily within the cells of the regenerative secretory zone, and not in cells from developing wing discs which were undamaged (Figure 2A-C). Ets21C was also upregulated after physically . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi. org/10.1101org/10. /2021 wounding of the wing disc ( Figure 2D), implying that Ets21C is involved in a general regeneration response. Ets21C had previously been shown to be upregulated during disc regeneration by bulk sequencing of blastema-enriched cells (Khan et al., 2017). Our single-cell data indicates that Ets21C expression is highly correlated with Ilp8 and Mmp1 expression during regeneration (supplemental fig. 6). Ets21C expression was induced during the genetic ablation period and was maintained throughout regeneration (supplemental fig. 7), suggesting that Ets21C could function at multiple stages of regeneration.
To determine if Ets21C was important either for normal development or for regeneration, we turned to mutant analysis. First, we observed that homozygous Ets21C -/null mutants generate viable and fertile adults, as previously noted (Mundorf et al., 2019), whose wings were of normal size and shape ( Figure 2E-G).
By generating mosaic eyes composed of marked wild-type cells and Ets21C -/mutant cells, we found that mutant cells did not display defects in cell proliferation even in a competition scenario with wild-type cells (Figure 2H, I). Thus, Ets21C is dispensable for normal development and its absence does not impair cell proliferation.
Next, we tested if Ets21C is essential for imaginal disc regeneration. Following our genetic ablation assay, homozygous null Ets21C -/mutants showed a dramatic defect in the extent of wing regeneration when compared to either Ets21C +/heterozygotes or wild-type controls ( Figure 2J). This effect was observed with the null mutation in trans to a chromosomal deletion ( Figure 2J, K), indicating that the effect was indeed due to the loss of Ets21C function. Thus, Ets21C is required for effective regeneration.
Ets21C is part of the Ets-family of DNA binding transcription factors that are broadly conserved in animals. The Ets21C mammalian orthologs are Ets-related gene (ERG) and Friend Leukemia Integration 1 Transcription Factor (FLI1), both of which can act as proto-oncogenes (Kar and Gutierrez-Hartmann, 2013). In . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10. 1101 Drosophila, although Ets21C is not expressed in undamaged third instar wing discs, its expression is upregulated in tumorous imaginal discs (Kulshammer et al., 2015;Toggweiler et al., 2016) and it has been shown to be involved in adult midgut homeostasis (Jin et al., 2015;Mundorf et al., 2019). Ets21C is a downstream target of JNK/AP1 signaling in these contexts (Kulshammer et al., 2015;Toggweiler et al., 2016;Mundorf et al., 2019). Similarly, we found that even in undamaged discs, activation of the JNK pathway induces Ets21C expression (supplemental fig. 8).
Thus, we hypothesized that Ets21C could be functioning downstream of JNK/AP1 signaling to activate a regeneration-specific GRN in the blastema.
We investigated if Ets21C mutant tissues would fail to upregulate secreted molecules during regeneration. Pvf1 and upd3 are expressed in the inner region of the regenerative secretory zone. In wild-type regenerating discs, Pvf1 and upd3-lacZ (Bunker et al., 2015) expression was detected within the center of the blastema and also within cellular debris (Figure 3A, C). In contrast, regenerating Ets21C -/mutant discs showed a substantial decrease both in Pvf1 and upd3-lacZ expression ( Figure 3B, D), indicating that Ets21C is required to initiate expression of both ligands in response to damage. In addition, we found that this reduced expression of upd3 and possibly its paralogs impacted downstream JAK/STAT signaling. In control tissues, the STAT activity reporter was expressed in the center of the blastema ( Figure 3E) as well as in the surrounding hinge regions where the JAK/STAT pathway is active during development (Ayala-Camargo et al., 2013;La Fortezza et al., 2016). In contrast, Ets21C -/mutant tissues failed to activate the JAK/STAT reporter within the cells at the center of the blastema (Figure 3F), while JAK/STAT signaling in the hinge remained unaffected. Thus, Ets21C is required for localized expression of Pvf1 and Upd ligands within the inner regenerative secretory zone. The disruption of the inner regenerative secretory zone is also observed when we assess the pattern of cell proliferation in regenerating Ets21C mutant discs, which showed a reduced central non-proliferating zone ( Figure 3G, . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint H; supplemental fig. 9). This observation suggests that Ets21C is required for proper function of the inner regenerative secretory zone and potentially for its initial establishment.
We next examined if Ets21C regulated the expression of genes expressed in the outer regenerative secretory zone. Wg expression during regeneration was unaffected in Ets21C mutants (supplemental fig. 10). While Mmp1 expression was observed at early stages of regeneration, it was prematurely absent by 24 hours in  . 11). Ets21C -/mutants, however, showed premature expression of Broad during regeneration (supplemental fig. 11). In addition, a regeneration-induced developmental delay that occurs in more distant tissues, specifically the eye disc, was also reduced in Ets21C -/larvae (supplemental fig.   11). Ilp8 is crucial for delaying pupariation (Colombani et al., 2012;Garelli et al., 2012), and this delay is correlated with regeneration outcomes (Smith-Bolton et al., 2009;Halme et al., 2010;Katsuyama et al., 2015;Harris et al., 2016). Indeed, Ets21C animals ended the larval phase of development approximately 30h before regenerating controls (supplemental fig. 12), which is likely the result of a decrease in Ilp8 levels and other signaling molecules. Thus, Ets21C mutants have both local and systemic defects in their regenerative response that collectively contribute to the reduced regeneration.
To test whether Ets21C regulates Ilp8 cell-autonomously, we generated mosaic discs containing patches of both Ets21C -/mutant and wild-type cells prior to tissue . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint damage (see Materials and Methods). In regenerating discs, clones of Ets21C cells (marked by the absence of RFP) are comparable in size to wild-type clones, indicating that Ets21C does not have a cell-autonomous function in regulating cell proliferation during regeneration ( Figure 3O). However, the Ets21C -/mutant cells showed a cell-autonomous decrease in Ilp8-GFP expression after 48h of regeneration compared to wild-type cells ( Figure 3O''), indicating that as with Mmp1, Ets21C is required to sustain Ilp8 expression. We propose that Ets21C works downstream of AP-1 in a type-1 coherent feed-forward loop (Alon, 2007), where the target genes that require Ets21C for initiating expression have enhancers with AND logic gates and target genes that require Ets21C for sustained expression during regeneration have enhancers with OR logic gates ( Figure 3P). This GRN model suggests that Ets21C is critical for cells to interpret sustained JNK/AP-1 signaling and activate pro-regenerative pathways.
While Ets21C function is required for regeneration and not necessary for normal development, it is known to be expressed in tumorous imaginal discs that have mutations that disrupt apicobasal polarity (Kulshammer et al., 2015;Toggweiler et al., 2016). Moreover, in one study, reducing Ets21C function was shown to reduce was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint models (Toggweiler et al., 2016), the presence of blastema-like cells could be critical for promoting the overgrowth of tumorous discs.
In conclusion, we have discovered a GRN that is dispensable for normal development yet essential for regeneration. Regeneration-specific GRNs may also exist in vertebrates and their reactivation could be valuable for regenerative medicine. Finally, the role of pro-regenerative GRNs in oncogenesis merits further exploration.

Single-cell data collection
For each sample, approximately 300 regenerating wing-imaginal discs were collected after 24 hours of regeneration. Regenerating discs were dissected within 1 hour in Supplemented Schneider's Medium. The samples were then processed according to the protocol outlined in (Everetts et al., 2021). Briefly, we used a mixture of trypsin and collagenase to enzymatically dissociate the tissues. Then we used FACS to eliminate both apoptotic cells and cellular debris. Because our tissue dissociation protocol enriched for myoblasts, we decided to specifically sort out myoblasts during the collection of our second regeneration sample. This was done with a Holes in muscle (Him)-GFP construct that specifically labeled the myoblasts (Rebeiz et al., 2002). The myoblasts represented 75.67 % of the cells from the sample without Him-GFP and 12.71 % cells of the sample with Him-GFP (as determined by single-cell analysis). Single-cell suspensions were barcoded for single-cell RNA sequencing with the 10X Chromium Single Cell platform (v2 chemistry). Barcoded samples were sequenced on an Illumina NovaSeq (S2 flow cell) to over 60% saturation.

Single-cell data analysis
Single-cell sequencing reads were aligned with the 10x Genomics CellRanger pipeline (v.2.2.0) to the Drosophila melanogaster transcriptome (version 6.24, . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint FlyBase). Analysis of the single-cell data was conducted in the R and Python programming languages, primarily using the packages scvi-tools v0.9.1 (Gayoso et al., 2021) and Seurat v3 (Stuart et al., 2019).
Before cell filtering, we used scvi-tools to harmonize our single-cell data from regenerating wing discs with the single-cell data from developing wild-type wing discs presented in our previous study (accession number GSE155543) (Everetts et al., 2021). We used Seurat's variance-stabilizing transformation method to select 1000 variable genes for each batch, and the scVI VAE model was trained on the union of these genes with the following parameters: n_latent = 15, n_layers = 2, gene_likelihood = "nb", max_epochs = 400, and train_size = 0.8. The scVI latent space was used as the input for Seurat's clustering algorithm (parameters: clustering k.param = 35, clustering resolution = 2.0; default parameters otherwise).
Known transcriptional markers were used to classify cell clusters: SPARC and twist for AMPs, Fasciclin 3(Fas3) and narrow for the disc epithelium, and regucalcin and Hemese (He) for hemocytes (supplemental fig. 2). We also identified an unusual cell cluster that expressed both AMP and epithelium markers, with slightly elevated average nGene and nUMI counts, which suggested that these cells were actually doublets. When we applied the tool DoubletFinder (McGinnis et al., 2019) (parameters: 30 PCs, pN = 0.25, expected doublets (nExp) = 7.5% of total cells in each batch, and pK determined by the recommended BCmvn method) to each individual batch, the majority of cells within this cluster had been classified as potential doublets within their respective batches. We determined this cluster to represent AMP-epithelium doublets and removed it from subsequent analysis. We isolated the remaining disc epithelium and AMP clusters and filtered these cell types separately.
When filtering the disc epithelium cells, we first processed each batch using the standard Seurat pipeline (parameters: nfeatures = 1000, npcs = 30, k.param = 20, clustering resolution = 2.0; default parameters otherwise) and removed low-quality clusters. We classified low-quality clusters as having: 1) an average nGene less than . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. 1 standard deviation below the average nGene of all cells, 2) an average percent.mito greater than 1 standard deviation above the average percent.mito of all cells, and 3) an abundance of negative marker genes compared to positive markers genes (as calculated by a Wilcoxon test). After removing low-quality clusters, we marked potential doublets within the epithelium cells by applying DoubletFinder (same parameters as described initially) to the epithelium cells in each batch. Batches were harmonized with scVI (same parameters as described initially), trained on the union of the top 1000 variable genes within the epithelium cells for each batch as determined by Seurat. After harmonization, we used the scVI latent space as a basis for Seurat clustering (parameters: k.param = 35, resolution = 2.0; default parameters otherwise). We removed a cluster that we determined to be epithelium-epithelium doublets, based on the following characteristics: (1) a noticeably higher average nGene compared to all other clusters (the only cluster with an average nGene > 1 standard deviation above the average nGene of all cells), (2) an extreme abundance of potential doublets as classified by DoubletFinder from each batch (~70% of all potential doublets classified were contained within this cluster), and (3) a lack of marker genes (both positive and negative) when compared to other clusters. We also removed a cluster that we determined to represent a small number of trachea cells, based on the unique expression of marker genes tracheal-prostasin and waterproof. We re-ran our variable gene selection, scVI harmonization, and Seurat clustering. Data was visualized in 2 dimensions with UMAP (parameters: min.dist = 0.1; default parameters otherwise).

Gene signature analysis of the blastema
For each identity combination (hinge-pouch, pouch-notum, and notum-hinge), gene signatures were constructed as follows: First, differential expression was performed between wild-type (non-regenerating) cells of each identity pair (e.g., for the hinge-pouch signature, differential expression was performed between cells from (Everetts et al., 2021) classified as hinge vs. cells classified as pouch). This was conducted using a Wilcoxon test via Seurat's FindMarkers function, selecting genes with a natural-log fold-change of greater than 0.25 (logfc.threshold = 0.25) and a . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint Bonferroni-corrected p-value of < 0.05. This provided three gene sets that differentiated hinge-pouch, pouch-notum, and notum-hinge identities. Second, principal component analysis was performed on all cells using each gene set. The first principal components from each analysis were defined as the gene signatures, as they best separated cells of the different identities. The signature scores of cell clusters were visualized in 2-dimensions using Seurat's VlnPlot function (Figure 1J;   supplemental fig. 5A, B) and in 3-dimensions using the R package Plotly (supplemental fig. 5C).

Gene signature of cellular maturity
To determine the relative cellular maturity (or developmental progression) of individual cells within the regenerating tissue we generate a gene signature score based on genes with differential expression during normal development. First, we selected genes with consistent differential expression between epithelial cells from mid (younger) and late (older) 3rd instar imaginal discs (with a threshold of greater than 0.25 natural-log fold-change). Second, this gene set was then used to perform principal component analysis to derive a cellular maturity score. The relative cellular maturity score of cells from normal developing and regenerating discs were visualized on the UMAP (supplemental fig 11B).

Single-cell comparison of regenerating and scrib tissues
The expression matrices for the scrib single-cell data were downloaded from GEO, accession number GSE130566 . Gene names were updated to match those within our regeneration and wild-type datasets. All scrib datasets (4d, 5d, 8d, and 14d) were harmonized with scVI (n_latent = 15, n_layers = 2, gene_likelihood = "nb", max_epochs = 400, and train_size = 0.8), trained on the union of the top 1000 variable genes for each batch as determined by Seurat.
Clustering was performed using Seurat, and we isolated the scrib epithelium clusters (identifiable by high expression of Fasciclin 3 and narrow) for subsequent comparison with the regeneration and wild-type epithelium data. No scrib epithelium cells were filtered during this comparative analysis.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint The epithelium data from regeneration, wild-type, and scrib samples was initially harmonized with scVI (n_latent = 15, n_layers = 2, gene_likelihood = "nb", max_epochs = 400, and train_size = 0.8), trained on the union of the top 1000 variable genes for each batch as determined by Seurat. The weights from this scVI model were used to initialize a scANVI model (using the from_scvi_model function) for semi-supervised training and label transfer. The cluster identities from our regeneration analysis ( Figure 1B) were supplied as input labels (via setup_anndata), with all scrib cells marked as "Unknown". The scANVI model was trained for 50 epochs (max_epochs = 50) to predict the probability of "Unknown" cells belonging to each of the input labels. Input labels were subsampled during training (n_samples_per_label = 150) to prevent the loss of labels with relatively few cells (e.g., Blastema1) during label transfer. After training, the scANVI latent space was used as a basis for UMAP (supplemental fig. 13B), and the transferred labels corresponded to the highest predicted identity for each cell (supplemental fig. 13C).
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint

EdU assay and quantification
For EdU staining, live discs were incubated in Schneider's medium (ThermoFisher 21720024) with EdU for 30 minutes, following the protocol for the Click-iT EdU Cell Proliferation Kit, Alexa Fluor 555 (ThermoFisher C10338) and Alexa Fluor 647 (ThermoFisher C10340). After the incubation, discs were fixed in 4% PFA for 15 min, before proceeding with standard antibody staining, as detailed above.
EdU intensity was quantified using the ImageJ software. For each regenerating disc, a square box was drawn, centered around the blastema. The length of the box was 140 microns for the 0h R discs, 160 microns for the 24h and 36h R discs, and 200 microns for the 48h R discs. The EdU intensity was measured at every pixel along the two diagonals of each box using ImageJ's "Plot Profile" function. Subsequent analysis was done using R software. The measured EdU intensities were first znormalized (i.e., for all values in a measured profile, subtract the mean and divide by the standard deviation) and then averaged across all diagonals from all processed discs at each regenerating time point. The average normalized (scaled) EdU intensity was plotted with the package ggplot2, and smoothed curves were added using the stat_smooth function with method = "gam".

Regeneration experiments
Unless otherwise noted, the genetic ablation system used to study regeneration was rn-GAL4, tub-GAL80 ts , UAS-eiger (Smith-Bolton et al., 2009). Genetic ablation experiments were conducted by synchronizing development by collecting eggs on grape plates and picking 55 L1 larvae into vials with yeast paste. Temperature shifts to induce ablation (from 18 C to 30 C) were conducted on day 7 after egg lay (AEL) for 40 hours. The extent of adult wing regeneration was scored by binning the resulting wings into 5 categories (0%, 25%, 50%, 75%, and 100%) (4). The resulting regeneration scores were calculated per population. Experimental replicates were done on separate days with a minimum of 2 vials per genotype and three replicates per genotype. Statistical comparison performed on regeneration scores using ANOVA followed by Tukey test for significance.

Mitotic clones during regeneration
Mosaic tissues were generated by recombinase-driven (FLP/FRT) mitotic recombination within the genetic background of the ablation system. The expression of hsFLP was induced by an 1h heat-shock at 37 C on day 3 AEL, which generated clones throughout the imaginal discs prior to genetic ablation and regeneration. Mutant cells were labeled by the absence of RFP and wild-type cells were marked by 2X RFP. The genotype of the experimental larvae used to generate Ets21C mutant clones during regeneration: hsFLP; Ets21C Δ10 , FRT40A / ubi-RFPnls, FRT40A; rn-GAL4, tub-GAL80 ts , UAS-eiger / Ilp8-GFP ( Figure 3O).

Lineage-tracing experiments
We identify an enhancer for the gene grain (grn) that was primarily expressed in the inner-hinge, GMR26E03-lexA (Pfeiffer et al., 2010) during normal development ( Figure 1N). Lineage-tracing was performed by permanently labelling the cells that . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint

Pupariation timing experiments
Images were taken every 20 minutes of vials that contained animals as they transitioned between larva to pupa. This was performed at 18 C with a wide-angle camera (Arducam). Pupariation was scored by observing when the animals stopped moving and darken in color.

Physical wounding assay
Wing discs were physically wounded in situ as described in (Yoo et al., 2016).
Briefly, L3 larvae with the wing pouch fluorescently labeled (rn-GAL4, UAS-his2A::RFP) were visualized using a fluorescence microscope. The right wing pouch was wounded by carefully applying pressure on the larval cuticle using a thin gauge insulin needle without penetrating the larval cuticle. Larvae were then returned to vials containing Bloomington food and dissected 6 hours or 24 hours later. Competing interests: Authors declare that they have no competing interests.

Data and materials availability:
The single-cell RNA sequencing data presented in this study and analysis code will be made available upon publication.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Supplementary Figures: Supplementary figure 1. Schematic of genetic ablation system to study imaginal disc regeneration. (A) Schematic of genetic ablation system. The canonical domains of the wing disc and the adult structures to which they give rise are colored in green (notum), white (hinge), and purple (pouch). Expression of the pro-apoptotic gene eiger is targeted to the wing pouch using rn-GAL4 and UAS-eiger. Gal4 function is inhibited at 18 o C and permitted at 30 o C by a ubiquitouslyexpressed temperature-sensitive Gal80 (tub-Gal80 ts ). (B) Imaginal discs can be dissected and analyzed during and after the ablation period. The half-way point through the 40h ablation is indicated by "½". Other times refer to the time after the downshift to 18 o C -the phase when regeneration occurs.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint Supplementary figure 2. Three major cell types were identified from scRNAseq of wing imaginal discs. (A) Harmonized UMAP of scRNA data from wing imaginal discs. Data colored by sample of origin and labeled. Samples were derived from developing discs at the middle at late stages of the third larval instar (L3), as described previously (Everetts et al., 2021), and from regenerating discs 24h after the downshift to 18 o C. Two biological replicates were obtained for each sample (see Materials and Methods). The three major cell types identified were epithelial cells, myoblasts and hemocytes. In addition, a few trachea cells were also identified. The cell counts from the regenerating discs were: 6,613 epithelial cells, 7,466 myoblasts, 224 hemocytes and 17 trachea cells. (B, C) Wing-imaginal discs stained with anti-H2 to label the hemocytes and anti-Twist to label the myoblasts in developing (B) and regenerating (C) wing discs. (D) UMAP colored by major cell types: myoblasts, epithelial cells, and hemocytes are shown in different colors (E-G) Expression of marker genes for the three major cell types: (E) Fasciclin 3 (Fas3) expression marks the epithelium; (F) Secreted protein, acidic, cysteine-rich (SPARC) expression marks the myoblasts; and (G) Hemese (He) marks the hemocytes.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint Supplementary figure 3. Composition of cell clusters in developing and regenerating discs. (A) UMAP of harmonized data from epithelial cells from regenerating and developing (from two time points) samples. Each dataset, including replicates, is represented in a distinct color. (B) Composition of cell clusters, as shown in Figure 1B. Note the underrepresentation of cells assigned to pouch clusters in regenerating discs and the near absence of cells assigned to the Blastema1 and Blastema2 clusters in developing discs. (C) Heatmap showing differential expression of marker genes through the different cell clusters of the harmonized epithelial cell object. Gene expression with individual cells from the single-cell data for the hinge makers Zn finger homeodomain 2 (zfh2) (D) and Sox box protein 15 (Sox15) (E) and for the pouch markers defective proventriculus (dve) (F) and rotund (rn) (G).
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 9, 2021. ; https://doi.org/10.1101/2021.07.08.451678 doi: bioRxiv preprint Supplementary figure 4. Co-expression of Blastema-1 and Blastema-2 marker genes within the regenerating epithelium. (A, B) Regenerating wing-imaginal disc, after 24h of regeneration, with transcriptional reporters and antibody staining to highlight the nested position of Blastema-1 and Blastema-2 cells within the regenerating epithelium. Ilp8-GFP expression is shown in yellow, anti-Pvf1 stain shown in magenta, and upd3-lacZ (Bunker et al., 2015) shown in cyan. (B) Magnification of blastema. Orange arrow highlights the region of Ilp8 expression and the red dotted line highlights the region of higher Pvf1 and upd3-lacZ expression that is in the surviving epithelial cells. Note that the cellular debris shows evidence of expression of all three of these marker genes (white arrowhead). Microscopy scale bars = 100 μm.

Pvf1 upd3
Blastema-1 Blastema-2 Supplementary figure 9. Cell proliferation during regeneration for wild type and Ets21C mutant wing discs. (A-D, I-K) Imaginal discs at different time points during regeneration (0R, 24R, 36R, 48R) for wild-type (A-D) or Ets21C mutant (I-K) tissues. Cell proliferation is marked by the incorporation of the thymidine analog EdU. Note that for the Ets21C mutants, the regeneration period is ended prematurely by pupariation (Supplementary fig. 12). (E-H, L-N) Profiles of average EdU intensity within the blastema at each regeneration time point for wildtype (E-H) or Ets21C mutant (L-N) discs. The y-axis corresponds to the average of z-normalized EdU intensity (see Materials and Methods). The x-axis corresponds to measured pixels centered around the blastema (i.e., the center of the distance axis is the center of the blastema). Note that in the wild-type tissues there is reduced EdU intensity within the center of the blastema, as evidenced by the bimodality of the EdU intensity profiles (most notable at 0R, 24R, and 36R). This pattern of proliferation is less robust within Ets21C mutant tissue, as evidenced by the unimodality of the EdU intensity profiles at 24R and 36R.

Supplementary figure 12. Relative pupariation timing for wild type (Wt) and
Ets21C mutant animals following damage and regeneration. Images of vials were taken at 10-minute intervals and then scored based on coloration changes during pupariation. Replicates were biological replicates conducted on separate days. The relative difference in pupariation timing was calculated based on the point when one-half of the animals scored had pupariated. Following ablation and regeneration, Ets21C mutant larvae formed pupa 31.5 hours and 29 hours earlier than the wildtype controls.