Keywords: spinal cord / regeneration / Foxm1 / progenitor / Xenopus / differentiation Pelzer et al. Role of Foxm1 during spinal cord regeneration 2 Summary Mammals have limited tissue regeneration capabilities, particularly in the case of the central nervous system. Spinal cord injuries are often irreversible and lead to the loss of motor and sensory function below the site of the damage [1]. In contrast, amphibians suchas Xenopus tadpoles can regenerate a fully functional tail, including their spinal cord, following amputation [2,3]. A hallmark of spinal cord regeneration is the re-activation of Sox2/3+ progenitor cells to promote regrowth of the spinal cord and the generation of new neurons [4,5]. In axolotls, this increase in proliferation is tightly regulated as progenitors switch from a neurogenic to a proliferative division via the planar polarity pathway (PCP) [6][7][8]. How the balance between self-renewal and differentiation is controlled during regeneration is not well understood. Here, we took an unbiased approach to identify regulators of the cell cycle expressed specifically in X.tropicalis spinal cord after tail amputation by RNAseq. This led to the identification of Foxm1 as a potential key transcription factor for spinal cord regeneration. Foxm1-/-X.tropicalis tadpoles develop normally but cannot regenerate their spinal cords. Using single cell RNAseq and immunolabelling, we show that foxm1+ cells in the regenerating spinal cord undergo a transient but dramatic change in the relative length of the different phases of the cell cycle, suggesting a change in their ability to differentiate. Indeed, we show that Foxm1 does not regulate the rate of progenitor proliferation but is required for neuronal differentiation leading to successful spinal cord regeneration. Pelzer et al. Role of Foxm1 during spinal cord regeneration 3
Results
Foxm1 is specifically expressed in the regenerating spinal cordWe compared the transcriptome of isolated spinal cords at 1day post amputation (1dpa) and 3dpa to spinal cord from intact tails (0dpa, Figure 1A). Principle component plot, dendogram of sample-to-sample distances and MA-plot of the log fold change (FC) of expression in relation to the average count confirmed the quality of the data (Figures S1A-D). Between 0dpa and 1dpa, 5129 differentially expressed (DE) transcripts (FC> 2 and FDR<0.01) were identified (2074 down-, 3055 up-regulated). Between 0dpa and 3dpa, 9787 genes are differentially expressed (4609 down and 5178 up-regulated, Figure S1E).To identify the most enriched biological processes by gene ontology (GO), a nonbiased hierarchical cluster for all DE genes was performed ( Figure 1B). We observed three phases: first an increase in expression of genes involved in metabolic processes (cluster I), then a strong upregulation of genes associated with cell cycle regulation (cluster II and III) and finally, a downregulation of expression of genes involved in nervous system development (Cluster IV and V, Figure 1B).Using Ingenuity Pathway Analysis (IPA), we identi...