Summary
The underlying molecular foundation of neural progenitor diversity and plasticity is critical for understanding repair processes. The neonatal cerebellum with multiple progenitor populations has high regenerative potential. Upon ablation of cerebellar granule cell progenitors at birth, a normally gliogenic Nestin-expressing progenitor (NEP) in the Bergmann glia (Bg) layer (BgL-NEPs) undergoes adaptive reprograming to restore granule neuron production while a white matter NEP (WM-NEPs) reduces interneuron production. However, the cellular states and genes regulating the NEP fate switch are not known. Here using scRNA-seq and fate-mapping, we defined the molecular subtypes of NEPs and their lineages under homeostasis and repair. Five NEP populations comprising two molecular subtypes, Hopx-expressing gliogenic- and Ascl1-expressing neurogenic-NEPs were identified in both states. Furthermore, in the WM, distinct NEP populations generate interneurons or astrocytes, and amongst gliogenic-NEPs, astrocyte and Bg lineages are molecularly separable. Importantly, we uncovered that after injury a new transitory cellular state arises from Hopx-NEPs in the BgL that is defined by initiation of expression of the neurogenic gene Ascl1. Moreover, Ascl1 is required for adaptive reprogramming and the full regenerative capacity of the cerebellum. We thus define new populations of NEPs and identifed the transcription factor responsible for inducing a transitory cell critical for a glial to neural switch in vivo following injury.
Introduction
Progenitor cells must have a multiplicity of subtypes and flexibility in fate choices for complex tissues to be generated and repair of injuries to be effective. Development of single cell approaches has been instrumental in dissecting progenitor diversity and identifying transitory cell states critical to fate decisions (Llorens-Bobadilla et al., 2015; Wang and Navin, 2015). The neonatal cerebellum (CB) represents a valuable system to identify molecular mechanisms that drive plasticity after injury due to its high regenerative potential and diversity of progenitor populations. The CB, a folded hindbrain structure that houses the majority of the neurons in the brain (Azevedo et al., 2009; Herculano-Houzel et al., 2006), is important for motor and higher order cognitive functions (Caligiore et al., 2017; Fatemi et al., 2012; Koziol et al., 2014; Lackey et al., 2018). The CB has prolonged development compared to the rest of the brain as production of a majority of the cells occurs after birth in mammals. The postnatal progenitors of the CB continue proliferating up to 6 months after birth in humans (Rakic and Sidman, 1970) and 2 weeks in mice (Altman and Bayer, 1997). The late development of the CB leads to increased susceptibility to injury around birth and cerebellar hypoplasia is the second leading risk factor of autism spectrum disorders (Wang et al., 2014). Interestingly, the newborn rodent CB can efficiently replenish at least two of its main cell types when they are ablated (Altman and Anderson, 1971; Altman et al., 1969; Bayin et al., 2018; Wojcinski et al., 2017; Wojcinski et al., 2018), and one repair process involves unexpected progenitor plasticity and a glial to neural fate switch after injury. The molecular mechanisms that drive progenitor plasticity in vivo however are unknown in the neonatal CB.
After birth, several distinct cerebellar progenitor populations derived from the rhombic lip (RL) or the ventricular zone (VZ) continue to proliferate and generate late-born cells. The rhombic lip-derived Atoh1-expressing granule cell progenitors (GCPs) proliferate in the external granular layer (EGL) on the surface of the CB and upon their last cell division, produce excitatory neurons that migrate inwards to form the inner granular layer (IGL) (Machold and Fishell, 2005; Wang et al., 2005; Wingate and Hatten, 1999). A poorly defined group of cerebellar ventricular zone-derived nestin-expressing progenitors (NEPs) give rise to astrocytes, inhibitory interneurons (INs) and Bergmann glia (Bg), a specialized polarized glial cell with fibers extending to the cerebellar surface (Cerrato et al., 2018; Fleming et al., 2013; Lee et al., 2005; Milosevic and Goldman, 2004; Parmigiani et al., 2015; Wojcinski et al., 2017). Fate mapping studies have indicated that the lineage propensity of NEPs depends on their position, such that NEPs in the layer that houses the Bg (BgL) or in the prospective white matter (WM) produce Bg or IN, respectively, and that both produce astrocytes (Cerrato et al., 2018; Fleming et al., 2013; Parmigiani et al., 2015; Wojcinski et al., 2017). Furthermore, there are NEPs in the deep WM of the lateral CB that contains the cerebellar nuclei (Cerrato et al., 2018). However, whether deep and lobule WM-NEPs are molecularly and functionally distinct is unknown. It is assumed, however that only NEPs in the WM of the lobules give rise to the INs that migrate to the molecular layer (ML) and populate it in an inside to outside manner (Figure 1A, B) (Brown et al., 2019; Sudarov et al., 2011). The full extent of the diversity of postnatal NEPs, and the molecular signature, lineage propensity and location of each NEP subtype are unknown.
When the proliferating GCPs are depleted in the newborn CB either via genetic approaches or by irradiation, the EGL recovers within a week and the CB grows to almost a normal size (Altman and Anderson, 1971; Altman et al., 1969; Wojcinski et al., 2017; Wojcinski et al., 2018). A key aspect of the regenerative process involves BgL-NEPs, thought to be gliogenic, undergoing adaptive reprograming to replenish the lost GCPs. The first response of the BgL-NEPs is an increase in proliferation, followed by migration to the injured EGL where they initiate expression of Atoh1 and become GCPs (Andreotti et al., 2018; Jaeger and Jessberger, 2017; Wojcinski et al., 2017; Wojcinski et al., 2018). Simultaneously, the WM-NEPs in the lobules reduce their proliferation and differentiation until the EGL is restored, leading to an overall delay in CB development, likely to ensure proper scaling of the different cell types of the cerebellar cortex (Wojcinski et al., 2017). Mosaic mutant analysis revealed that Sonic hedgehog signaling (SHH) is important for the adaptive reprograming of NEPs (Wojcinski et al., 2017). However, it is not known what molecular event induces the glial to neural fate change of BgL-NEPs upon injury and whether a new transitory cellular state facilitates the fate switch.
The commitment of neural stem cells to a neural fate is dependent upon a group of proneural basic helix-loop-helix (bHLH) transcription factors (Dennis et al., 2019; Guillemot and Hassan, 2017). In the CB, RL-derived excitatory neurons and VZ-derived inhibitory neurons are specified by distinct bHLH proteins typified by ATOH1 and ASCL1, respectively (RL-bHLH and VZ-bHLH). Fate mapping in the CB has shown that Ascl1 is transiently expressed sequentially in a subset of VZ- and then NEP-derived cells that generate inhibitory neurons and is never involved in GCP production (Brown et al., 2019; Sudarov et al., 2011). Moreover, in cell culture ASCL1 is one of the primary proteins found to transform non-neural cells into neurons in vitro. For example, mis-expression of Ascl1 along with Brn2 and Myt1l converts fibroblasts into functional neurons, and ASCL1 acts as a pioneer factor that initiates the reprograming of a fate change via inducing a neurogenic transcriptional program (Vierbuchen et al., 2010; Wapinski et al., 2017). In vivo, ectopic expression of ASCL1 or other VZ-bHLH proteins is able to reprogram astrocytes to form neural cells upon injury (Grande et al., 2013; Heinrich et al., 2014; Liu et al., 2015; Zamboni et al., 2020). In contrast to the required mis-expression of VZ-bHLH proteins in all these studies, during GCP regeneration BgL-NEPs do not require ectopic expression of any genes to undergo a glial-to-neural fate switch upon injury, highlighting their high plasticity compared to other injury responsive glial cells outside the CB. However, whether and which proneural bHLH genes play a role in the acquisition of a neural fate by BgL-NEPs during GCP regeneration is unknown.
We used scRNA-seq in combination with genetic inducible fate mapping (GIFM) and loss of function studies to define molecularly distinct NEP populations in the postnatal developing CB and uncover a transcription factor involved in the fate switch of a VZ-derived gliogenic BgL-NEP to a RL-derived GCP. Five molecularly distinct NEP populations were identified at postnatal day (P) 5 and based on their lineage propensity they comprise 2 NEP subtypes. GIFM revealed that during homeostasis a Hopx-expressing NEP subtype is primarily gliogenic and resides in both the BgL and WM, whereas an Ascl1-expressing NEP subtype resides in the lobule WM and is neurogenic (produces INs). Furthermore, amongst the NEP subtypes seen after injury a new transitory state was identified in cells derived from Hopx-expressing BgL-NEPs and defined by Ascl1 expression. Conditional mutagenesis showed that Ascl1 is not only a marker of the transitory cell state but also plays a crucial role in GCP regeneration. Collectively, our results reveal the molecular diversity and the cellular plasticity of cerebellar NEPs and identify a context- (injury-) dependent transitory cellular state reponsible for a glial to neural fate switch that enables neonatal cerebellar regeneration.
Results
Identification of NEPs using scRNA-seq of whole P5 cerebellum during homeostasis and repair
In order to identify molecularly distinct NEP subtypes that are present in the postnatal CB and distinct cellular states that are acquired during adaptive reprogramming to become GCPs, we performed scRNA-seq on dissociated cells from 4 male P5 Nes-Cfp/+ pups that were irradiated (IR) at P1 or control littermates that were not irradiated (nonIR) (Figure S1A). An automated microwell-array-based platform was used to simultaneously profile 13,000 single cells from nonIR and IR animals (Yuan and Sims, 2016), followed by subclustering to identify pervasive and specific markers for molecularly distinct NEP subtypes (Figure S1A-C). Clustering of cells (6,000 nonIR, 7,000 IR, Figure S1D) using Phenograph (Levine et al., 2015) revealed the expected variety of cell types, with the greatest number of cells being neurons and the majority being GCPs and GCs based on the expression of Atoh1 (GCP) and Barhl1 (GCP/GC) (Figure S1D-G, Table S1). The other major cell types that were identified include astroglia (Bergmann glia, astrocytes, C7, Figure S1F), oligodendrocyte progenitors/oligodendrocytes, microglia and other parenchymal cells (Figure S1G). We used Cfp transcripts as a surrogate to identify cluster(s) containing NEPs and found Cfp expression was enriched only in the astroglial cluster C7 (Figure S1F). Interestingly, the Cfp+ cells in the astroglial cluster only represented 32% of all the Cfp+ cells, as scattered Cfp-expressing cells were observed in the neural clusters and the oligodendrocyte cluster (Figure S1G), suggesting that the transcript is retained in some of the immediate differentiated progeny of NEPs. Henceforth, we define NEPs as the cells in cluster 7 and all other Cfp+ cells present in the neural clusters (Figure S1A). Finally, our analysis of sections of P5 Nes-Cfp/+ cerebella showed that in addition to the BgL and the lobule WM, there are also CFP+ cells in the deep WM that houses the cerebellar nuclei (Figure S1H-J). These cells represent <10% of all the CFP+ cells in the CB, and therefore are likely poorly represented in our scRNA-seq data.
scRNA-seq identifies two major subtypes and 5 populations of NEPs in the P5 CB at steady state
In order to determine whether distinct NEP subpopulations exist at steady state, we clustered the 1,067 NEPs from nonIR P5 mice (Figure 1D-F, Table S1). Phenograph analysis revealed five clusters (nIR1-5 for nonIR NEPs) and a small cluster of 13 ependymal cells (Foxj1+, nIR6) that cover the wall of the 3rd ventricle (Figure 1G-H). Analysis of the Allen Brain Atlas RNA in situ mouse P4 data set for the significantly enriched genes in each cluster and lineage markers allowed further validation of cluster identities. nIR5 is enriched for Bg lineage genes (Hopx, Gdf10) (Carter et al., 2018; Heng et al., 2017; Mecklenburg et al., 2014), nIR2 also is enriched for these genes but has lower expression of them and both nIR2 and nIR5 express pan-astrocyte genes (Slc1a3, Aqp4, Fabp7) (Figure 1I). One gene enriched specifically in nIR2 is Slc6a11, an astrocyte specific GABA transporter (Boddum et al., 2016). Interestingly, Slc6a11 was expressed in the subset of nIR2 cells that do not express Cfp, indicating that they are differentiated astrocytes and the rest of the cluster might be astrocyte progenitors (Figure 1G,H). On the other hand, the high expression of Bg genes in nIR5 could indicate the cells are Bg progenitors. Three clusters (nIR1, 3, 4) showed similar patterns of expression of neural genes (e.g. Rbfox3, Tubb3) and a gene enriched in the WM (Igbfbp5) based on RNA in situ analysis (Allen Brain Atlas, Figure 1G-I). Interestingly, the proneural VZ-bHLH gene, Ascl1, was enriched in nI4, although it is present in all three clusters at low levels (Figure 1G-I). Other immature IN marker genes such as Pax2 were not detected at high levels in our dataset. Furthermore, differential expression analysis of the genes expressed in the gliogenic clusters (nIR2, 5) compared to neurogenic clusters (nIR1, 3, 4) highlight further the gliogenic and neurogenic gene signature of each NEP subtype (Figure 1J). In summary, scRNA-seq analysis identified two major NEP subtypes, gliogenic (nIR2, 5) and neurogenic (nIR1, 3, 4), which appear to be further subdivided but into less distinct populations.
To identify differences between the neurogenic clusters (nIR1, 3 and 4) we performed differential expression analysis for pairwise comparisons of the three clusters. The results further highlighted the similarities between nIR1, 3 and 4, as there were few significantly differentially expressed genes (FDR<0.1, Log2FC>1, n=39 genes nIR1 vs. nIR3, n=16 genes nIR1 vs. nIR4 and n=28 genes nIR3 vs. nIR4, Figure 1K) and the genes were related to cell cycle (Table S2, Figure 1K). The results indicate that nIR1 represents a more proliferative population, consistent with its slightly higher Cfp expression, whereas the nIR4 and nIR3 could represent NEP populations transitioning into a more differentiated state.
Differential expression analysis was also performed on the two gliogenic clusters and confirmed that nIR5 has significantly higher expression of Bg associated genes (Hopx and Gdf10), supporting that nIR5 represents Bg progenitors and their immediate Bg progeny, whereas nIR2 expresses astrocyte genes at a higher level and thus could represent astrocyte progenitors and mature astrocytes (Figure 1L). nIR2 also shows expression of Rbfox3 (and Tubb3) in some cells and the WM enriched Igfbp5 gene (Figure 1G,H). Therefore nIR2 could include a more plastic group of NEPs including ones in the WM that can form astrocytes and possibly INs (Figure 1G,H). Collectively, our data show that the primary molecular signatures that distinguish NEP subpopulations is based on their lineage propensities, gliogenic vs neurogenic subtypes. In addition, neurogenic NEPs exhibit additional molecular diversity seemingly associated with their differentiation state, whereas the two gliogenic NEPs populations could be dedicated to producing Bg (nIR5) or astrocytes (nIR2).
Ascl1- and Hopx-expressing NEPs have distinct cell lineages
In order to confirm the identities of the NEP subtypes identified via scRNA-seq and study their lineages, we performed GIFM for the neurogenic-NEPs using an Ascl1CreERT2 allele and for the gliogenic-NEPs using a HopxCreERT2 allele in combination with a R26lox-STOP-loxTdTomato reporter (Ascl1-TdT and Hopx-TdT, mice respectively). When Tamoxifen (Tm) was administered at P5 to label cells at a similar time point as our scRNA-seq data (Figure 2A), short-term GIFM (analysis 2 days after Tm injection) in Ascl1-TdT mice showed that the majority of labeled cells were restricted to the lobule WM at P7 with scattered cells in the IGL that had the appearance of migrating immature ML INs and interestingly no cells in the deep WM (Figure 2B-E, Figure 2SA-D). Double labeling with the immature IN marker PAX2 confirmed that most of the TdT+ cells in the WM are INs (Figure S2I). Analysis of long-term GIFM (P30) showed that almost all of the progeny of the Ascl1+ WM-NEPs are ML INs (97.9 ± 2.3 %, n=5), specifically the stellate cells which are the latest born INs located in the outer ML (Figure 2F-I). Only rare astrocytes in the lobule WM or IGL and no GCs were labeled in the adult CB. These results not only confirm there is a neurogenic NEP subtype in the lobule WM, but that it has a restricted lineage and at P5 gives rise almost exclusively to Stellate INs.
Short-term labeling in P7 Hopx-TdT animals showed that all of the TdT+ cells were in the lobules and the majority were in the BgL (72.4 ± 6.2%, n=3, Figure 2J-Q and S2E-H). Marker analysis showed that both NEPs (SOX2+/S100β−) and Bg (SOX2+/S100β+) in the BgL were TdT+ (Figure S2J). Curiously, a subpopulation of lobule WM-NEPs was also TdT+ (26.6 ± 6.1%, n=3) along with very rare SOX2+ cells in the inner EGL (1.1 ± 0.3%, n=3, Figure 2J-M). Analysis of the Hopx-TdT mice at P30 showed that 98.2 ± 0.6% (n=3) of the TdT+ cells were astroglia (71.1 ± 2.7% Bg, 25.8 ± 2.1% astrocytes in the IGL and WM and 1.3 ± 1.0 % oligodendrocytes in the WM) and only ~1.8% were INs (1.10 ± 0.3 % in the ML) (Figure 2N-Q) and no labelled cells were seen in the deep WM (Figure SE-H).
RNA in situ analysis of Igfbp5 showed expression in the deep and lobule WM, whereas Hopx and Ascl1 expression were restricted to the lobules at P5 (Figure S2M-O). As there was no labeling in the deep WM cells at both P7 and P30 both in the Ascl1-TdT and the Hopx-TdT cerebella, this reveals that the lineages of NEPs expressing these genes are restricted to the lobules (Figure 2SA-H). Given that the INs of the cerebellar nuclei are generated during embryonic development directly from the VZ, the deep WM-NEPs should only be gliogenic but a specific marker gene is yet to be determined. Overall the GIFM studies validate the identities of the gliogenic- vs. neurogenic-NEP subtypes obtained via scRNA-seq (Figure 1E), uncover a WM-NEP population in the lobules that expresses Hopx and demonstrate that Ascl1-expressing NEPs in the neurogenic clusters have an ML IN-restricted lineage at P5 (Figure 2P).
Irradiation induces changes in the NEP populations and the emergence of an Ascl1+ transitory NEP
The P5 time point was selected for scRNA-seq because it should allow a variety of cellular states during regeneration to be identified since in IR mice the NEPs and their immediate progeny that express CFP are undergoing all stages of adaptive reprogramming, including proliferation, a fate change, and migration to the EGL (Figure S3A, Figure 3A-D). When the 1,472 NEPs from the P5 IR cerebella were analyzed using Phenograph, 6 clusters were obtained (IR1-6, for IR NEP clusters, Figure 3E, Figure S1A,C, Figure S3B-G, Table S1). Gene set enrichment analysis (GSEA) of the differentially expressed genes between nonIR and IR NEPs, ascertained by single cell differential expression analysis (SCDE) (Kharchenko et al., 2014), identified many cellular processes altered 4 days after injury, including upregulation of neural fate commitment and glial differentiation in the IR NEPs compared to increased synapse maturation in the nonIR NEPs (NES>1.5, p<0.05, Figure S3H, Table S3).
The clustering of the IR NEPs, as with the nonIR NEPs, was dictated by a gliogenic or neurogenic transcriptional signature. In addition, 2 gliogenic clusters (IR2 and 6, Figure 3E-G) and 3 neurogenic clusters (IR1, 3 and 4, Figure 3E-G) were again identified. However, importantly an additional distinct cluster was identified in IR NEPs (IR5). Most interestingly, the new cluster in the IR NEPs (IR5) showed enrichment for the neurogenic VZ-bHLH factor Ascl1 as well as glial genes such as Slc1a3 and Fabp7 (Figure 3F-G). In terms of the gliogenic clusters, IR2 showed enrichment for Bg genes (Hopx and Gdf10) similar to nlR5, whereas IR6 cells were enriched for astrocyte genes (Slc1a3, Aqp4 and Fabp7) but not Bg genes and were mainly Cfp- (Figure 3E-G). Although such a distinct cluster was not detected in nonIR NEPs, a group of similar mature astrocytes was identified as a subgroup within nIR2 (Figure 1E,H). RNA in situ analysis of the BgL-NEP markers Gdf10 and Hopx on sections from nonIR and IR P5 cerebella showed that their expression pattern was not changed by IR, as expression remained in the BgL, although as expected the BgL was thicker reflecting the increased proliferation of BgL-NEPs upon irradiation (Figure 3J-K) (Wojcinski et al., 2017). In terms of the neurogenic clusters (IR1,3 and 4), the WM-NEP marker Igfbp5 was expressed in a smaller proportion of the IR NEPs (270/1067 nonIR NEPs, Figure 1G vs. 137/1472 IR NEPs, Figure 3F). RNA in situ analysis of sections from IR P5 cerebella validated the reduction in Igfbp5 expression in both the lobule and deep WM cells after IR compared to nonIR controls (Figure S3I-M).
To further investigate how the IR NEP clusters are related to the nonIR NEP subtypes, we projected the IR cells (density plots) onto the nonIR (dot plots) UMAP embedding (Figure 3H). This analysis led to multiple interesting observations. First, the 3 IR neurogenic-NEP clusters (IR1, 3 and 4) overlapped specifically with nIR1, which appears to represent the most proliferative and least differentiated (Cfp high) steady-state neurogenic-NEP subpopulation (Figure 3H). Secondly, IR2 overlapped with nIR5, the gliogenic NEP cluster with Bg characteristics, suggesting that these NEPs primarily generate Bg. Consistent with the astrocyte expression signature of IR6, it overlapped specifically with the subregion of nIR2 that has Cfp- astrocytes (Figure 3H). It is curious that very few IR cells in IR2 or IR6 projected to the subregion of nIR2 that does not contain astrocytes and contains progenitors. Moreover, thirdly, there was no clear nonIR counterpart to lR5, which projected between nIR2 and the three neurogenic-NEPs clusters (Figure 3H). Thus, IR5 could represent a transitory cell population induced upon injury with a new gene expression signature that shares some similarities with the nonIR Bg-like gliogenic-NEPs (nIR2) and the neurogenic-NEP subtype. Differential expression analysis of IR5 compared to gliogenic- (IR2) or neurogenic- (IR1, 3 and 4) NEPs further revealed IR5 has aspects of both the VZ-bHLH neurogenic and glial signatures (Figure S3N-Q, Table S4). Finally, as a negative control, no ependymal cells were detected in the IR NEP dataset (based on Foxj1 expression), and none of the IR NEPs projected onto the small cluster of ependymal cells observed in the nonIR data (nIR6).
Transcriptional programs of Ascl1+ NEPs diverge between nonIR and IR conditions
We next investigated the transcriptional differences between all the Ascl1+ cells in nonIR compared to IR NEPs. We assessed co-expression and mutual exclusivity for all gene pairs and calculated probability ratios of detecting two genes in the same cell (Mizrak et al., 2019) for the top 50 genes with high Ascl1 co-expression probability ratios (15/100 showed high co-expression probability ratios with Ascl1 in both IR and nonIR cells, Table S5). Among the genes co-expressed with Ascl1 in both conditions were three other proneural VZ-bHLH genes, Ptf1a and Neurog1/2, (Figure 3I). Interestingly, we also detected sets of genes that were co-expressed with Ascl1 only in nonIR or in IR NEPs (Figure 3I). Analysis of P4 RNA in situ data (Allen Brain Atlas) for Sall1 that is co-expressed with Ascl1 in nonIR-NEPs revealed its expression is restricted to the WM where Ascl1+ NEPs normally reside (Figure 3I). Interestingly, Grin3a that is co-expressed with Ascl1 in IR NEPs is expressed in the BgL, consistent with the idea that after irradiation Ascl1 is induced in some BgL-NEPs (Figure 3I). Furthermore, some of the other genes co-expressed with Ascl1 in the IR NEPs, Grin3a, Arntl and S1pr3 have been implicated in astrocyte injury-related functions (Lananna et al., 2018; Rothhammer et al., 2017; Rusnakova et al., 2013), indicating the new cluster derives from a gliogenic NEP subtypes. These results further highlight that the Ascl1+ NEPs in the IR cerebella have a glial gene signature, leading us to suggest that the injury responsive gliogenic BgL-NEPs turn on Ascl1 upon injury.
To test whether injury does indeed induce VZ-bHLH gene expression in BgL-NEPs, we performed RNA in situ hybridization for Ascl1 on P5 nonIR and IR cerebellar sections and observed a clear increase in expression of Ascl1 in the BgL (Figure 3L). IF analysis of P5 Nes-Cfp/+ cerebella also confirmed an increase in the number of ASCL1+ CFP+ cells in the BgL (3.1 ± 0.4 –fold, n=3) of IR cerebella compared to nonIR P5 littermates (Figure 3M). Interestingly, ASCL1 expression was not observed in most of the CFP+ cells in the EGL indicating that Ascl1 is transiently induced upon injury in BgL NEPs prior to migration to the EGL (Figure 3M). Finally, quantification of the number of ASCL1+ cells every 24 hours between P2 and P6 and at P8 showed that expression in BgL-NEPs (CFP+ cells) peaks 3-4 days after IR (Figure 3N). Based on our scRNA-seq comparison of P5 IR and nonIR NEPs and in vivo identification of the location of the NEP populations, we propose that that upon injury, normally gliogenic BgL-NEPs turn on a new transcriptional program involving proneural VZ-bHLH transcription factors that allows a switch to a neural fate to occur prior to their mobilization to the EGL.
ASCL1+ transitory cells and granule cells are derived from Hopx-NEPs after injury
In order to demonstrate that Hopx-derived BgL-NEPs give rise to ASCL1+ cells that produce GCPs and GCs after irradiation, we performed GIFM with Hopx-TdT animals. When Tm was administered at P0 (Figure 4A), analysis of P7 cerebella revealed a significant increase in the percentage of ASCL1+ TdT+ cells of all TdT+ cells in the BgL in IR pups (12.7 ± 3.1%, n=3) compared to nonIR littermates (1.4 ± 1.4%, n=3, Figure 4B-H). In the nonIR brains, as expected ASCL1 expression was restricted to the WM and IGL and showed almost no overlap with Hopx-derived TdT+ cells in the WM (asterisks, Figure 4B-G). Interestingly, in IR pups the percentage of the ASCL1+ TdT+ double positive cells amongst all the ASCL1+ cells in the BgL was higher than the percentage of SOX2+ TdT+ double positive cells amongst all the SOX2+ cells (a pan-NEP/Bg/astrocyte marker) in the BgL (~39.2±7.7% in IR and 6.5 ± 1.9% in nonIR vs 15.5 ± 6.7% IR and 26.1 ± 8.8% in nonIR, respectively, n=3, Figure 4I). This result suggests that HopxCreERT2 preferentially marks the BgL-NEPs at ~P1 with the ability to undergo adaptive reprograming in response to loss of GCPs. Importantly, TdT+ cells were present in the EGL of the IR P7 cerebella, and most expressed SOX2, indicating that progeny of SOX2+ Hopx-TdT+ labeled BgL-NEPs migrated to the site of injury (Figure 4J-P). At P30, we observed a significant increase in the density of TdT+ cells in IR Hopx-TdT brains compared to nonIR (Figure 4Q-S), which was primarily driven by a significant increase in the density of TdT+ GCs in the IGL (Figure 4S-T). Although when Tm was given at P0 compared to P5 in Hopx-TdT mice we found a higher proportion of the TdT+ cells to be WM-NEPs at P7 (49.1% vs. 26.6%, Tm at P0 vs. P5, n=3) and ML INs at P30 (31.4% vs. 1.1%, Tm at P0 vs. P5, n=3), there was no significant difference in the density of TdT+ ML INs between IR and nonIR mice (Figure 4S-T). Finally, no TdT+ cells were detected in the deep WM of IR or nonIR Hopx-TdT animals that were given Tm at P0 (Figure S4A-B). Overall, these results provide experimental evidence that Hopx+ BgL-NEPs give rise to ASCL1-expressing transitory cells that change their fate and become GCPs and then GCs.
ASCL1 marks a neurogenic BgL-NEP population during adaptive reprogramming that gives rise to GCs
We next tested whether Ascl1 marks a new BgL-NEP transitory cellular state after irradiation that generates GCs by analyzing nonIR and IR Ascl1-TdT animals given Tm at P5 (Figure 5A), the stage with the largest number of injury-induced ASCL1+ BgL-NEPs (Figure 3N). Analysis of P6 and P7 cerebella revealed a population of TdT+ BgL-NEPs that are SOX2+/S100β− and have glial fibers projecting to the pial surface only in pups irradiated at P1 (Figure 5C,E,G,I). In contrast, the few TdT+ cells in the BgL of nonIR cerebella had a morphology of migrating progenitors (Figure 5B,D,F,H). Quantification of the density of TdT+ cells at P7 revealed a significant increase in the density of TdT+ cells in the EGL and BgL (Figure 5J). Furthermore, the number of labeled cells in the EGL increased significantly between 24 and 48 after Tm in IR cerebella (Figure 5C, J). Analysis of the proliferation of the TdT+ cells showed that in the IR pups, Ascl1-TdT+ cells in the BgL (both BgL-like and IN progenitor-like) and EGL incorporated EdU, as well as the Ascl1-TdT+ cells in the IGL+WM (Figure S4C-D). On the other hand in the nonIR brains, Ascl1-TdT+ IN progenitors in the IGL+WM and some that started migrating outwards towards the ML incorporated EdU, suggesting that all the Ascl1-expressing nonIR and IR NEPs are proliferative (Figure S4C-D).
Analysis of P30 IR Ascl1-TdT cerebella revealed several key findings. Firstly, the P30 IR brains had a significant and large increase in the density and the proportion of the TdT+ cells that were GCs (13.8 ± 2.0 % of all TdT+ cells, n=4 IR compared to 1.0 ± 1.4% of TdT+ cells, n=5 nonIR, Figure 5R,T,U). Secondly, almost no TdT+ Bg were detected in the P30 IR Ascl1-TdT cerebella, suggesting that Ascl1 induces the BgL-NEPs to switch from making Bg to GCPs and that the switch is irreversible (Figure 5M-N, U). Thirdly, there was a delay in production of ML INs after irradiation, as demonstrated by an increase in the density of TdT+ ML INs per lobule that included both earlier born Basket INs and later born Stellate INs (Figure 5K-L, U, Figure S4E-G). Finally, we observed an increase in astrocyte production in the IR Ascl1-TdT brains, primarily in the WM (Figure 5M-P, U). Surprisingly, after irradiation we detected an emergence of TdT+ astrocytes in the deep WM in Ascl1-TdT cerebella at P30, unlike in nonIR Ascl1-TdT cerebella or IR and nonIR Hopx-TdT brains where no TdT+ deep WM astrocytes were detected (Figure S4A-B, H). Since no TdT+ cells were detected in the deep WM of IR Ascl1-TdT cerebella at P6 and P7, this indicates that the TdT+ cells must originate from the lobules (Figure S3H). This result shows that two different NEP subtypes (Ascl1+ vs Hopx+) have different lineage responses to injury.
Finally, live imaging of lobule 3 in thick cerebellar sections from P7 cerebella showed that the Ascl1-TdT+ cells do indeed migrate from the BgL to the EGL in the IR cerebella, whereas in nonIR P7 slices no cells migrate from BgL to EGL (Figure S4A-C, and movie 1-2). Collectively, these results demonstrate that GCPs are generated from an Ascl1-expressing Bgl-NEP upon injury. Furthermore, irradiation leads to a delay in production of INs and ectopic astrocyte production, likely from normally neurogenic WM-NEPs.
Ascl1 is required in Hopx+ NEPs for adaptive reprogramming and production of GCs
Given the surprising upregulation of Ascl1 in the normally gliogenic BgL-NEPs upon death of GCPs, we tested whether Ascl1 is required for repair of the GC lineage after cerebellar irradiation using a conditional knockout (CKO) approach. HopxCreERT2/+; Ascl1fl/fl animals injected with Tm at P0 (Hopx-Ascl1 CKOs) were used to delete Ascl1 primarily from BgL-NEPs and their regenerative capacity was compared to Ascl1fl/fl littermate controls after irradiation at P1 (Figure 6A). Analysis of P30 cerebella showed a significantly greater reduction in cerebellar area at the midline in IR Hopx-Ascl1 CKOs compared IR controls (Figure 6B-E, J). Furthermore, the IGL was less organized in some lobules of the IR mutants than in controls, further demonstrating impaired regeneration (Figure 6F-I). In order to rule out that the impairment of the regeneration is due to a loss of one copy of the Hopx gene in the Hopx-Ascl1 CKO animals, we compared cerebellar areas of IR and nonIR Hopx-TdT animals and their R26TdT/+ littermates and found no significant changes between the genotypes, showing that loss of one copy of Hopx does not impair regeneration (Figure S6A-D, I). Interestingly, the loss of one copy of Ascl1 resulted in a mild but significant reduction in the area of the CB of Ascl1-TdT animals compared to R26TdT/+ littermate controls, further confirming that ASCL1+ is required for regeneration after irradiation (Figure S6E-H, J).
We next asked whether the number of Bg or astrocytes generated was altered in P30 Hopx-Ascl1 CKOs. Interestingly, there was a mild but significant increase in the density of Bg in IR Hopx-Ascl1 CKOs compared to IR controls, but no differences between the other three groups. One possibility is that after irradiation the mutant BgL-NEPs generate Bg rather than GCPs (Figure S6L-P). In addition, the density of the astrocytes in the lobule WM in the IR Hopx-Ascl1 CKOs was increased compared to nonIR mutants and IR controls (Figure S6Q-R). Finally, we quantified the density of ML INs, because we observed a population labeled with HopxCreERT2 upon TM administration at P0. No significant changes in the densities of ML INs were observed after irradiation in the Hopx-Ascl1 CKOs compared to their control littermates and their respective nonIR controls (Figure S6S-T). However, since the size of the mutant CB is significantly reduced, by extrapolation the total number of ML INs is reduced in IR mutants.
In order to understand why the regeneration is impaired in the Hopx-Ascl1 CKOs, we analyzed nonIR and IR cerebella at P7, and determined whether proliferation of NEPs and/or their migration to the EGL were affected. Quantification of the density of ASCL1+ cells in the BgL in P7 IR and nonIR cerebella confirmed loss of ASCL1-expressing transitory cells after IR and thus deletion of Ascl1 in mutants (Figure 6K-N, S). Significantly, Hopx-Ascl1 CKOs had a lower density of SOX2+ cells in the EGL (BgL-derived cells that had migrated to the EGL) after injury compared to their control IR littermates (Figure 6O-R, T). A previous study suggested that proliferation genes are direct targets of ASCL1 in embryonic stem cell-derived neural stem cell cultures or the embryonic ventral brain (Castro et al., 2011). To test if loss of Ascl1 in the Hopx-expressing NEPs impairs their proliferation, we injected EdU 1h prior to sacrificing P7 pups (Figure 6A) and assed the percentage of EdU+ SOX2+ cells in the BgL. A significant decrease in the percentage of EdU+ SOX2+ cells was detected in the BgL of IR Hopx-Ascl1 CKOs compared to controls (Figure 6O-R, U). In summary, these results show that ASCL1 is required for the full reprograming of Hopx-expressing BgL-NEPs after depletion of GCPs, including efficient proliferation and migration to the EGL, and as a consequence of impaired replenishment of GCPs the cerebella of P30 IR Ascl1 CKOs are reduced in size compared to control IR mice.
Discussion
The neonatal mouse CB has remarkable regenerative potential and cellular plasticity upon injury. Using scRNA-seq and GIFM, we defined two NEP subtypes and the transcriptional signatures of their 5 subpopulations during homeostasis, and also identified a key new transitory cellular state that is necessary for adaptive reprogramming of BgL-NEPs into GCPs following EGL injury. Our results reveal that lineage propensity of NEPs (astroglial vs neural) is the primary molecular factor that differentiates NEP subtypes. At steady state Hopx-expressing NEPs are proliferative, primarily gliogenic and found in both the BgL and WM, whereas the Ascl1-expressing NEPs are proliferative but restricted to the lobule WM and dedicated to making ML INs at P5. Importantly, we discovered that upon depletion of the GCPs Hopx-expressing gliogenic BgL-NEPs transition to a new state where proneural VZ-bHLH genes are activated prior to their migration to the EGL. Furthermore, upon injury, we not only confirmed a delay in production of INs by neurogenic NEPs using GIFM but uncovered that a subset of neurogenic-NEPs switch to producing astrocytes, including some that become ectopically located in the deep WM. Finally, given that ASCL1 is a pioneer transcription factor (Wapinski et al., 2017) and we discovered it is required for full regeneration of the GCPs, we propose that ASCL1 is involved in erasing the gliogenic differentiation program of the BgL-NEPs, thus allowing them to acquire a GCP identity (Figure 7).
An important question is whether within the gliogenic-NEP subtype there are NEP subpopulations that are dedicated to making only Bg or astrocytes, and if so whether the latter are located in both the WM and BgL. Based on our scRNA-seq, there are two gliogenic-NEP subpopulations (nIR2 and nIR5) present at steady state and one is more related to astrocytes and the other to Bg. Whereas Bg only are located in the BgL, Hopx-derived astrocytes are located in the IGL and lobule WM, which raises the question of whether both Hopx+ BgL-NEPs and/or WM-NEPs generate astrocytes. For example, Hopx-BgL NEPs might generate IGL astrocytes and WM Hopx-NEPs generate the WM astrocytes. Elegant clonal labeling of apparently only BgL-NEPs using GlastCreER/+; R26Confetti/+ mice and application of Tm to the surface of the P6 CB in vivo found that the majority of clones had both Bg and astrocytes located in the IGL (~55%), although a large proportion were dedicated to making only Bg (~39%) and the rest to making only IGL astrocytes (Cerrato et al., 2018). This result indicates that WM astrocytes must be generated from WM-NEPs, likely the Hopx-expressing WM-NEPs we identified. Indeed, when WM progenitors were fluorescently labelled directly by injection of virus, some P3-P5 WM-NEPs gave rise to WM astrocytes (Parmigiani et al., 2015). Furthermore, our comparison of labeling before irradiation in IR and nonIR Hopx-TdT animals revealed preferential labeling of the BgL-NEPs that become Ascl1+ after injury, providing evidence that that not all BgL-NEPs are eqivalent in their extent of plasticity. Thus, molecular and lineage analysis shows that WM astrocytes and Bg have distinct progenitors, whereas IGL astrocytes might come from both a bipotent progenitor and lineage restricted progenitor in the BgL.
At steady state, Ascl1 is expressed in all three neurogenic-NEP clusters, and our GIFM showed that Ascl1+ cells are exclusively in the lobule WM and give rise to ML INs at P5, in agreement with previous findings at other postnatal ages (Brown et al., 2019; Sudarov et al., 2011). Our differential expression analysis of the neurogenic clusters (nIR1, 3 and 4) showed they are very similar and suggests the main differentially expressed genes are related to cell cycle or Cfp level. One interpretation is that the three neurogenic WM-NEP subpopulations represent progressive differentiation states. Interestingly, we found that Hopx-expressing NEPs give rise to a small number of INs at P5 and the proportion and the number is higher when Tm is injected at P0, indicating there could be a bipotent Hopx+ WM-NEP that produces INs. Further clonal analysis is required to resolve whether gliogenic WM-NEPs can give rise to both INs and WM astrocytes.
We showed that induction of the proneural VZ-bHLH transcription factor ASCL1 in BgL-NEPs is a key step in their adaptive reprograming upon GCP death. Of importance, this Ascl1-driven gliogenic to neurogenic switch occurs without outside intervention such as the forced ectopic expression of proneural transcription factors required in other CNS regions (Grande et al., 2013; Heinrich et al., 2014; Zamboni et al., 2020). Further investigation is needed to identify the injury induced signal(s) that activate Ascl1 expression. Since upregulation of Ascl1 peaks 3-4 days after injury, the critical events required to produce the Ascl1+ transitory cell state must occur befor P5. Interestingly, in another study we showed that upon depletion of Purkinje cells in the newborn CB, although new Purkinje cells are generated, BgL-NEPs do not undergo adaptive reprogramming to produce them (Bayin et al., 2018). Thus, the type of injury/cell type killed is instrumental in determining the downstream regenerative response of NEPs.
We previously showed that at the same time that the BgL-NEPs are generating new GCPs after irradiation, the lobule WM-NEPs reduce their proliferation and production of INs and to a lesser extent astrocytes (Wojcinski et al., 2017). Our finding that expression of the WM-NEP marker Igfbp5 is reduced in WM-NEPs provides molecular evidence for such a response to EGL injury. Importantly, a surprising observation was that after injury Ascl1-expressing NEPs produce a significant number of astrocytes located not only in the IGL and lobule WM, but some that migrate ectopically to the deep WM. Although the transitory Ascl1+ BgL-NEPs might be the source of the IGL astrocytes, it is likely that Ascl1+ WM-NEPs in the lobules have sufficient plasticity to switch to producing astrocytes in the lobules and deep WM upon injury. The reason for this neural to glial fate switch after injury and whether the same NEPs retain their neurogenic potential (i.e. are bipotent) or whether a distinct subpopulation of the Ascl1+ WM-NEPs become gliogenic remains to be explored.
In conclusion, our studies provide molecular insights into the make up of NEP subtypes and the endogenous permissive mechanisms that support a glial to neural switch crucial for enhancing repair after brain injury. A new transitory BgL-NEP population was identidied and we demonstrate that ASCL1, which normally drives cerebellar IN production (Sudarov et al., 2011), is required for adaptive reprograming of BgL-NEPs to produce excitatory granule cells upon EGL injury. The mechanisms and cellular states we identified may well have broader implications for gliogenic progenitors in other regions of the brain, and their potential responses to injury.
Methods
Animals
All the mouse experiments were performed according to protocols approved by the Memorial Sloan Kettering Cancer Center’s Institutional Animal Care and Use Committee (IACUC, protocol no 07-01-001). Animals were housed on a 12-hour light/dark cycle and given access to food and water ad libitum.
The following mouse lines were used: Nes-Cfp (Mignone et al., 2004; Wojcinski et al., 2017), HopxCreERT2 (Takeda et al., 2011), Ascl1CreERT2 (Kim et al., 2011; Sudarov et al., 2011), Ascl1fl/fl (Pacary et al., 2011; Sudarov et al., 2011) Rosa26lox-STOP-loxTdTomato (ai14, Stock no: 007909, The Jackson Laboratories)(Madisen et al., 2015). Animals were maintained on an outbred Swiss Webster background. Both sexes were used for analyses and experimenters were blinded for genotypes whenever possible.
Tamoxifen (200ug/g, Sigma) was injected subcutaneously at P0 or at P5. EdU (50ug/g) was injected intraperitoneally at P7.
Analysis of Ascl1-TdT and Hopx-TdT P30 brains that were not injected with Tm showed no ectopic TdT+ cells (Figure S2K-L).
Irradiation
P1 pups were anesthetized by hypothermia and a single dose of 4 Gy gamma-irradiation was provided using an X-RAD 225Cx (Precision X-ray) microirradiator in the Small Animal Imaging Core Facility at Memorial Sloan Kettering Cancer Center. The CB region was targeted using a collimator with 5 mm diameter.
Tissue Preparation and Histological analysis
For P5 and younger animals, brains were dissected and drop fixed in 4% PFA for 24-48 hours at 4°C. Animals older than P5 were systemically perfused with ice-cold PBS followed by 4% PFA, following anesthesia with a Ketamine (100mg/kg) and Xylaxine (10mg/kg) cocktail. After dissection brains were fixed for an additional 24-48 hours in 4% PFA. Fixed brains were switched to 30% Sucrose in PBS. Once they sunk, brains were embedded in OCT (Tissue-Tek) for cryosectioning. 14um-thick sections were obtained using a cryostat (Leica, CM3050S) and stored at −20°C.
Haematoxylene and Eosin (H&E) staining was performed for cerebellar area (size) measurements and assessment of the cerebellar cytoarchitecture.
Immunofluorescent (IF) analysis was performed on cyrosections. Slides were allowed to warm to room temperature (RT) and washed once with PBS. 1 hour blocking was performed at RT using 5% Bovine Serum Albumin (BSA, Sigma) in PBS-T (PBS with 0.1% Triton-X). For ASCL1 IF analysis, Mouse on Mouse Blocking reagent (Vector Labs) was applied for 2 hours at RT to reduce the background. Slides were incubated with primary antibodies diluted in blocking solution at 4°C overnight (Table S6). Slides were then washed with PBS-T (3 × 5 min) and incubated with fluorophore-conjugated secondary antibodies (1:500 in blocking buffer, Invitrogen). Hoechst 33258 (Invitrogen) was used to label the nuclei and the slides were mounted with Fluoro-Gel mounting media (Electron Microscopy Sciences). To detect EdU, a Click-it EdU assay with Sulfo-Cyanine5 azide (Lumiprobe corporation, A3330) was used.
RNA in situ hybridization
Specimen treatment and hybridization were performed as previously described (Blaess et al., 2011). Templates for the probes were in vitro transcribed either from PCR-amplified cDNAs obtained from neonatal cerebella extracts or synthesized DNA (GeneScript). An Ascl1 probe was generated from a plasmid as previously described (Guillemot and Joyner, 1993). Sequences of the probes are shown in Table S7.
Slice cultures and live imaging
P7 cerebella were dissected in ice-cold 1x HBSS (Gibco) and embedded in 4% low melting point agarose. 250 μm thick sagittal slices were obtained using a vibratome (Leica) and immediately placed on Millicell (Millipore) tissue culture inserts over Neurobasal media supplemented with 1x B27 and 1x N2 supplements (Gibco), 2mM L-glutamine and allowed to equilibrate for 30 minutes at 37°C in 5% CO2 prior to imaging. Image stacks were obtained every 3-4 minutes for 6-8 hours using a LSM880 (Leica) with an environmental chamber (37°C with 5% CO2). Movies were processed using ZEN (Leica) and ImageJ (NIH) software.
Image acquisition and analysis
Images were collected with a DM6000 Leica microscope or Zeiss LSM 880 confocal microscope. Images were processed using ImageJ Software (NIH).
Three near midline sections for each animal were quantified and averaged for all the analyses shown. Quantification of the number of cells was performed on lobules 3 and 4/5 at P7 and on lobule 3 at P30. Boundaries of the lobules to be quantified were decided based on straight lines drawn from the bases of the fissures. Cell densities were calculated by dividing the number of cells by the area of the lobule(s) quantified. CB area was measured on H&E stained slides from 3 near midline sections and the values were averaged. The number of animals that is used in each quantification is denoted in the figure legends and wherever summary statistics are presented.
Single cell sequencing and data analysis
Sample preparation
Four male P5 nonIR and IR cerebella were dissected into ice-cold 1x HBSS (Gibco) and were pooled for downstream analysis. All the steps were performed on ice when possible. Cerebella were minced with a clean blade and then dissociated in Accutase (Innovative Cell Tech.) at 37°C for 10-15 minutes. Following dissociation, accutase was washed out using Neural stem cell media (Neurobasal, supplemented with N2, B27 and non essentital amino acids, Gibco). Following filtering through a cell strainer and trituration in media to single cells, cells were layered over a 5mL density gradient (albumin-ovomucoid inhibitor solution, Worthington) and centrifuged at 70g for 6 minutes to remove debris. The cells were briefly treated with 1:2 1X red blood cell lysis buffer (Sigma) for 2 minutes at room temperature. Cells were washed twice (500g, 5 minutes at 4°C) and resuspended in 1X Tris-buffered Saline. Cells were stained with Calcein AM live stain dye (Fisher, 1:500) on ice for 30 minutes, and were passed through a 40 μm cell strainer to remove any cell clumps before loading onto a microwell device.
Single Cell library preparation and sequencing
On-chip reverse transcription, cDNA amplification and sequencing library preparation were performed as described previously (Mizrak et al., 2020; Mizrak et al., 2019; Yuan and Sims, 2016). Briefly, reverse transcription reactions were performed on DropSeq beads (Chemgenes, MACOSKO-2011-10 (V+)). The microwell devices were scanned during the RNA capture to check for lysis efficiency. Following PCR amplification, the libraries were prepared using Nextera XT kit. DNA purifications were performed using Ampure XP beads. cDNA and library amount and quality were assessed using Qubit and a Bioanalyzer (Agilent). High quality samples were sequenced on a NextSeq 500 using a High Output 75 cycle kit (Read 1: 21 Cycles, Index: 8 Cycles, Read 2: 63 cycles).
Single Cell RNA-sequencing Data Analysis
The sequencing reads were demultiplexed, aligned, and quantified as described previously (Mizrak et al., 2020). 12 nucleotide cell barcodes and 8 nucleotide unique molecular identifiers were extracted from Read 1, and trimmed Read 2 reads were aligned to a mouse genome (GRCm38, Gencode annotation vM10) using STAR v.2.5 (Dobin et al., 2013). Cfp sequence was added in the annotation and denoted as NestinCFP in the expression matrices.
Unsupervised clustering was carried out as described previously (Levitin et al., 2018; Mizrak et al., 2019). The clustering was performed on protein-coding genes only, including NESTINCFP. After computing cell-by-cell Spearman correlation matrix, k-nearest neighbors graph was constructed with k set to 30. The resulting graph was used as input for Louvain clustering with Phenograph (Levine et al., 2015). Cluster-specific genes were identified using a binomial-specificity test (Shekhar et al., 2016). Uniform Manifold Approximation and Projection (UMAP) was used for all data visualization (Becht et al., 2018).
Differentially expressed genes in nonIR compared to IR conditions were identified using Single-cell differential expression (SCDE) analysis (Kharchenko et al., 2014). The resulting gene lists were pre-ranked based on the effect size, and were inputted to GSEA (Subramanian et al., 2005) using the following parameters: xtools.gsea.GseaPreranked -scoring_scheme weighted –setmin 1 –setmax 1000 –nperm 1000.
To address co-expression and mutual exclusivity of genes detected in more than 25 cells, we calculated the marginal detection probabilities for each gene pair as described previously (Mizrak et al., 2020). The top 50 genes that were co-expressed with Ascl1 in nonIR and IR cerebella were identified for downstream analysis. The co-expression values were plotted as a scatter plot to identify condition specific co-expressed genes. We projected the scRNA-seq profiles from each IR cluster into the UMAP embedding of the nonIR profiles as described previously (Szabo et al., 2019) using the transform function in Python implementation of UMAP (Becht et al., 2018). The code for this analysis is available at www.github.com/simslab/umap_projection.
We performed differential expression analysis between clusters within the IR or nonIR data as follows: First, we sub-sampled each pair of clusters to the same number of cells. Next, we sub-sampled the counts for each pair of clusters to the same average number of counts per cell, keeping only transcripts annotated as protein-coding. We normalized the resulting sub-sampled count matrices using the computeSumFactors function (Lun et al., 2016) in scran and then conducted differential expression analysis with the Mann-Whitney U-test as implemented in the mannwhitneyu function in SciPy and corrected the resulting p-values using the Benjamini-Hochberg method as implemented in the multipletests function in the Python module StatsModels.
Statistical analysis
Prism (GraphPad) was used for all statistical analysis. Statistical tests performed in this study were Student’s two-tailed t-test, Two-way analysis of variance (ANOVA), followed by post hoc analysis with Tukey’s multiple comparison tests. p-values of the relevant post hoc analyses are shown in the figures and in the Table S8. The statistical significance cutoff was set at p<0.05 and the data is presented as mean ± standard deviation (SD) of the mean. F-statistics and p-values are stated in the figure legends and relevant post hoc comparisons are shown in the figures. n≥3 mice were used for each experiment and the sample size for each experiment is stated in the figure legends and the text.
List of additional supplementary materials
Movie 1. Live imaging of the lobule 3 of a P7 nonIR Ascl1-TdT CB that was given Tm at P5. Outline shows EGL. Time lapse images are shown in Figure S5.
Movie 2. Live imaging of the lobule 3 of a P7 IR Ascl1-TdT CB that was IR at P1 and given Tm at P5. Outline shows EGL. Time lapse images are shown in Figure S5.
Table S1. Binomial test results showing the cell-type specific genes in the P5 nonIR and IR whole cerebella and in nonIR and IR NEPs (related to Figures 1, 3, S1, S3)
Table S2. Differential expression analysis amongst nonIR clusters (related to Figure 1)
Table S3. GSEA results of the differentially expressed genes between the nonIR and IR NEPs (related to Figure S3)
Table S4. Differential expression analysis amongst IR clusters (related to Figure 3)
Table S5. Top genes that are co-expressed with Ascl1 in nonIR and IR cells (related to Figure 3)
Table S6. List of antibodies
Table S7. Sequences complementary to the probes used for RNA in situ hybridization.
Table S8. Summary of the statistics
Code Availability
Data analysis code used for scRNA-seq is available at https://github.com/simslab.
Movie 1. Live imaging of the lobule 3 of a P7 nonIR Ascl1-TdT CB that was given Tm at P5. Outline shows EGL. Time lapse images are shown in Figure S5.
Movie 2. Live imaging of the lobule 3 of a P7 IR Ascl1-TdT CB that was IR at P1 and given Tm at P5. Outline shows EGL. Time lapse images are shown in Figure S5.
Table S1. Binomial test results showing the cell-type specific genes in the P5 nonIR and IR whole cerebella and in nonIR and IR NEPs (related to Figures 1, 3, S1, S3)
Table S2. Differential expression analysis amongst nonIR clusters (related to Figure 1)
Table S3. GSEA results of the differentially expressed genes between the nonIR and IR NEPs (related to Figure S3)
Table S4. Differential expression analysis amongst IR clusters (related to Figure 3)
Table S5. Top genes that are co-expressed with Ascl1 in nonIR and IR cells (related to Figure 3)
Acknowledgements
We thank past and present members of the Joyner laboratory for discussions and technical help. We are grateful to the MSKCC Small Animal Imaging and Mouse Genetics Core Facilities for technical services and support. A XRad 225Cx Microirradiator was purchased by support from a Shared Resources Grant from the MSKCC GeoffreyBeene Cancer Research Center. Finally, we thank the Sulzberger Columbia Genome Center, supported by a National Cancer Institute Cancer Center Support Grant (P30CA013696), for technical services for the scRNA-seq experiments. This work was supported by grants from the NIH to ALJ (NINDS R01NS092096, NIMH R37MH085726) and a National Cancer Institute Cancer Center Support Grant (P30 CA008748-48). NSB was supported by postdoctoral fellowships from NYSTEM (C32599GG) and NIH/NINDS (K99/R00 NS112605-01). PAS was supported by R01NS103473 from NIH/NINDS.