RFC1 regulates the expansion of neural progenitors in the developing zebrafish cerebellum

Nobilleau, Fanny; Audet, Sébastien; da Silva Babinet, Alexandra; Tork, Sanaa; Zaouter, Charlotte; Liao, Meijiang; Pilon, Nicolas; Tétreault, Martine; Patten, Shunmoogum A.; Samarut, Éric

doi:10.1038/s41467-025-60775-5

Download PDF

Article
Open access
Published: 01 July 2025

RFC1 regulates the expansion of neural progenitors in the developing zebrafish cerebellum

Nature Communications volume 16, Article number: 6019 (2025) Cite this article

2925 Accesses
2 Altmetric
Metrics details

Subjects

Abstract

DNA replication and repair are basic yet essential molecular processes for all cells. RFC1 encodes the largest subunit of the Replication Factor C, an essential clamp-loader for DNA replication and repair. Intronic repeat expansion in RFC1 has recently been associated with so-called RFC1-related disorders, which mainly encompass late-onset cerebellar ataxias. However, the mechanisms making certain tissues more susceptible to defects in these universal pathways remain mysterious. Here, we provide the first investigation of RFC1 gene function in vivo using zebrafish. We showed that RFC1 is expressed in neural progenitor cells within the developing cerebellum, where it maintains their genomic integrity during neurogenic maturation. Accordingly, RFC1 loss-of-function leads to a severe cerebellar phenotype due to impaired neurogenesis of both Purkinje and granule cells. Our data point to a specific role of RFC1 in the developing cerebellum, paving the way for a better understanding of the pathogenic mechanisms underlying RFC1-related disorders.

Phenotypic, metabolic, and biogenesis properties of human stem cell-derived cerebellar spheroids

Article Open access 27 July 2022

Cellular development and evolution of the mammalian cerebellum

Article Open access 29 November 2023

Phenotypic analysis of catastrophic childhood epilepsy genes

Article Open access 03 June 2021

Introduction

The cerebellum is the main derivative of the embryonic hindbrain in all vertebrates. It receives and processes information from sensory systems, the spinal cord and other parts of the brain^1,2,3,4, using these inputs to ensure that motor movements are smooth, coordinated and efficient⁵. Best known for its motor coordination role, the cerebellum also influences cognitive processes such as attention, language, and modulation of emotional responses^6,7,8. Due to all these essential roles, dysfunctions in the development or activity of the cerebellum lead to pathological syndromes associated with movement disorders, problems of balance, posture, and even learning^8,9,10.

Ataxias are neurodegenerative diseases of the cerebellum and/or brainstem that are characterized by a lack of muscle control or coordination of voluntary movements. These conditions have been linked to hundreds of genes with high phenotypic overlap^11,12. Particularly, late-onset cerebellar ataxias (LOCA) are a heterogeneous group of cerebellar progressive neurodegenerative diseases that manifest in adulthood with unsteadiness. It is estimated that up to three in 100,000 people worldwide will develop a LOCA. This prevalence is likely underestimated because of the great heterogeneity of clinical presentations, especially in aging patients. The genetic and phenotypic heterogeneity of LOCA contributes to our inability to diagnose ataxia patients, and most LOCA patients remain without a genetic diagnosis¹³. This dampens our understanding of the underlying pathological mechanisms involved in these diseases. However, in 2019, a breakthrough in the genetics of ataxia identified biallelic repeat expansions in the second intron of the RFC1 gene, encoding Replication Factor C subunit 1. These pentanucleotide repeat expansions are primarily associated with the cerebellar ataxia-neuropathy-vestibular areflexia syndrome (CANVAS)^14,15. Although it is a rare syndrome, recent findings suggest that RFC1 expansions are a major cause of a broader phenotypic spectrum of late-onset progressive ataxias. Moreover, the pathogenic scope of these expansions has been expanded to other movement syndromes, now collectively referred to as RFC1-related disorders, which include specific forms of Parkinsonism and multiple systems atrophy (MSA)^16,17,18,19. Although the pathogenicity of RFC1 repeat expansions is now well established, the underlying mechanisms remain unknown. Recent studies propose that RFC1 loss-of-function could be at the basis of the repeat expansion pathogenicity^20,21,22,23, as the recessive inheritance pattern also suggests. However, other studies showed no specific neurodegenerative phenotype in RFC1-depleted neuronal cultures in vitro²⁴. This questions the direct involvement of the RFC1 gene in so-called RFC1-related pathologies.

RFC1 encodes the largest subunit of the replication factor C, a DNA-dependent ATPase which loads the DNA-clamp protein PCNA (Proliferating Cell Nuclear Antigen) and recruits DNA polymerases onto DNA undergoing replication or repair^25,26,27. The Human Protein Atlas database indicates a ubiquitous tissue expression in adults, which is consistent with its assumed housekeeping function as a DNA repair and replication regulator. Nevertheless, the accurate expression of RFC1 has not yet been assessed in mammals or vertebrates in vivo, particularly during embryogenesis. Since its cloning in humans in the 90s²⁸, the function of RFC1, particularly in the context of DNA replication and repair, has only been studied in vitro^25,29,30 or in plants^31,32,33. Detailed in vivo analysis of its function in animals is needed to better understand RFC1-associated diseases.

In this work, we studied the gene and protein expression patterns of RFC1 in vertebrates, in the developing cerebellum of both mice and zebrafish. This analysis revealed specific expression patterns in neuronal progenitor populations of both Purkinje and granule cell lineages. To circumvent the early embryonic lethality of Rfc1 knockout mouse models, we generated a zebrafish model of rfc1 loss-of-function using CRISPR/Cas9 targeted mutagenesis. Although rfc1^−/− zebrafish larvae die prematurely after 10 days of age, we found that this does not prevent the functional analysis of RFC1 during neurodevelopment. This model uncovered a key role for rfc1 in the proper development of both granule and Purkinje cell populations. Notably, single-cell transcriptomic approaches showed that rfc1 regulates the expansion and differentiation of the neuronal progenitor pools. In the absence of rfc1, these progenitors accumulate DNA damage and stop proliferating, leading to their death and ultimately severely affecting cerebellum neurogenesis. This specific role of RFC1 during neurodevelopment opens new doors regarding its potential involvement in cerebellar pathologies.

Results

RFC1 is expressed in the developing cerebellum

To evaluate the tissue distribution of RFC1 in vivo, we first evaluated its mRNA expression in zebrafish embryos via whole-mount in situ hybridization. We found an early broad expression in the developing brain, from 1 day post-fertilization (dpf), with a robust expression in the midbrain-hindbrain boundary (mhb) starting at 2 dpf (Fig. 1A). Using fluorescent in situ RNAscope assays, we showed that at 2 dpf, rfc1 is expressed on both sides of the midbrain-hindbrain boundary: anteriorly in the developing optic tecum (ot), particularly in the tectal periventricular layer (PVL); posteriorly in the upper rhombic lip (URL), in the forming valvula cerebelli (Va) and in the ventral and caudal LRL. We confirmed the identity of these structures using ptf1a and atoh1a probes as specific markers³⁴ (Fig. 1B–D and Fig. S1A, B). Rfc1 diffused expression is sustained in the tectal periventricular layer (PVL) and the developing cerebellar plate corpus cerebelli (CCe) from 3 dpf, and its expression overlaps with cerebellar neuronal markers neurod1 (granule cells) and pvalb7 (Purkinje cells) (Fig. 1E, F and Fig. S1C, D). Although specific, the expression of rfc1 in the developing cerebellum is not exclusive especially, at early developmental stages, as an expression signal is also observed in the eye, floor plate and surrounding pharyngeal tissues. To complement these observations at the cellular level and in another vertebrate model, we examined the RFC1 protein distribution in the developing cerebellum of wild-type mice at the cellular level (Fig. 1G–J). Using co-immunolabelling of RFC1 expression with specific markers of Purkinje cells (PC, Calbindin) or granule cells (GC, Pax6), we showed that RFC1 is expressed in early Purkinje cells progenitors (PCp) from P0. Its expression persists in the Purkinje Cell Layer (PCL) at P7, P11, and P60 (Fig. 1G–J, white arrows). Of note is that the nuclear expression of RFC1 in PCp gradually becomes cytoplasmic in more mature PC from P11 onwards. We also found that RFC1 is not expressed in early granule cell progenitors at P0, as shown by the absence of colocalization with the PAX6 marker. However, it is transiently expressed in some migratory granule cells (mGCs) at P11. This stage corresponds to the moment when GCs migrate from the external granule layer (EGL) towards the internal granular layer (IGL), passing through the molecular layer (ML) and the Purkinje cell layer (PCL) (Fig. 1G–J, white arrowheads). Once in the IGL, we could not detect colocalization between RFC1 and the GC marker PAX6. However, we did detect RFC1 expression in some PAX6-positive GCs scattered in the ML. Initially described as ectopic GCs (eGCs), these cells have been recently renamed molecular GCs (mGCs) based on their proven function within the cerebellum circuitry³⁵. Altogether, our results show that RFC1 is expressed in the developing cerebellum, both in PC and transiently in a subset of GCs in vertebrates.

Loss of rfc1 causes survival and morphological anomalies in zebrafish

The Jackson Laboratory Phenotyping Center generated a knock-out (KO) Rfc1 allele in mice (Rfc1^{tm1b(KOMP)Mbp}), for which homozygosity was reported to lead to embryonic and/or pre-weaning lethality (no homozygotes over 66 pups at weaning; MGI ID:5495286³⁶). In an effort to circumvent this roadblock and with the goal to study the global impact of RFC1 loss-of-function in vivo in vertebrates, we generated a zebrafish rfc1 KO model using the CRISPR/Cas9 technology. We targeted the 5^th exon and selected a founder line with a 20-bp deletion mutation, which leads to a premature stop codon at position 182 in the zebrafish RFC1 protein (Fig. 2A, B). Of note is that zebrafish have a single rfc1 ortholog, and it encodes a protein sharing 57% of sequence identity (69% of sequence similarity) with the human RFC1, encompassing the same functional domains (Fig. 2c and Fig. S2). We confirmed rfc1 loss-of-function at the transcript level both in heterozygous (+/−) and homozygous (−/−) zebrafish larvae (Fig. 2D). Since heterozygous (rfc1^+/–) zebrafish survive to adulthood without overt phenotypes, and retain ~75% of rfc1 transcript levels (Fig. 2C) compared to wild-type, this suggests that rfc1 is haplosufficient. In other words, a single functional allele of rfc1 provides sufficient gene dosage to maintain normal development and survival. However, we found that rfc1^−/− larvae die prematurely from 8 days of age with no survivors past day 10 (n = 168; Fig. 2E). Moreover, rfc1^−/− larvae depict gross morphological abnormalities from 2 days post-fertilization (dpf). In particular, they display a smaller head, reduced eye diameter and a shorter otic vesicle than their siblings (Figs. S3A, 2B and Fig. 2F–H). These morphological defects become more obvious from 3 dpf and worsen until death. Noteworthy, we did not observe differences in the average body size (Fig. S3B), suggesting that the head morphological phenotype observed in rfc1^−/− larvae is very specific and not due to developmental delay. Finally, we monitored the swimming behavior of 4 dpf larvae and found that rfc1^−/− larvae are significantly hypoactive compared to their siblings as assessed by the total distance swum during 60 min (Fig. 2I, J). Altogether, these data show that RFC1 is necessary for normal development, motor behavior and survival in zebrafish. Moreover, its loss-of-function leads to head morphological deformities that we aim to characterize further.

**Fig. 2: *Rfc1* loss-of-function leads to premature death, head morphological anomalies and motility defects in zebrafish.**

Rfc1 ^−/− zebrafish larvae depict severe cerebellar defects

As we showed that RFC1 is expressed in the developing brain, particularly at the level of the hindbrain/midbrain boundary, we monitored brain development, particularly the cerebellum, in our rfc1^−/− zebrafish model. We confirmed that the reduced head size of rfc1^−/− larvae is accompanied by a general reduction in brain size, particularly at the level of the midbrain as revealed by hematoxylin-eosin staining of brain cross-sections (Fig. 3A). More importantly, unlike in control rfc1^+/− and wild-type larvae, we were not able to identify any distinct cerebellar structures in 4 dpf rfc1^−/− larvae cross-sections at the level of the hindbrain (asterisks in Fig. 3A and Fig. S4A). To confirm this observation, we followed by in toto immunolabelling the development of the two main cerebellar neuronal cell types: Parvalbumin7⁺ PCs and neurod1⁺/vglut1⁺ GCs (Fig. 3B–F and Figs. S4B, S3C). Consistent with our histological observations, we showed that GCs fail to develop in rfc1^−/− larvae compared to control siblings (Fig. 3B, C). Indeed, the number of neurod1⁺ GCs is drastically reduced at 3 dpf in rfc1^−/− larvae (Fig. 3B), also remaining low afterwards (Fig. 3B, E and Figs. S4B, S3C). This is consistent with the absence of vglut1⁺ posterior axonal projections in rfc1^−/− larvae, which are normally observed from 3 dpf and thicken with age in siblings (Fig. 3C and Fig. S4C). Similarly, albeit less severe, the number of developing parvalbumin7⁺ PCs is significantly reduced in rfc1^−/− from 3 dpf onwards (Fig. 3D, F and Fig. S4D). Moreover, the general organization of the PC layer is affected in 5 dpf larvae with the absence of the Valvula cerebelli structure, a cerebellar structure unique to ray-finned fishes that controls non-locomotor-related functions³⁷(Fig. 3D and Fig. S3D). To confirm the specificity of these observations, we generated F0-injected rfc1-CRISPant larvae targeting each RFC1 functional domain individually (i.e., BRCT, ATPase or DNA replication domain) using specific guide RNAs (Fig. S5A). We showed that all CRISPants phenocopy the small-head and small-eye phenotypes of rfc1^−/− mutants (Fig. S5B–D), as well as the defects in cerebellar integrity (Fig. S5E, F). Using these F0-CRISPant larvae, we also showed that the integrity of olig2⁺ eurydendroid cells (i.e., the equivalents of the mammalian deep cerebellar nuclei in teleosts³⁸) is affected upon rfc1 loss-of-function (Fig. 3G, H). Overall, these data confirm that rfc1 is involved in cerebellum development, particularly in populating its neuronal subtypes in vertebrates.

**Fig. 3: *Rfc1−/−* zebrafish larvae exhibit severe cerebellar defects.**

Rfc1 is necessary for early neural progenitor proliferation

To complement our previous observations and in order to dig further into rfc1-regulated brain cell populations, we performed deep RNA sequencing of single cells (scRNA-seq) from larval brains rfc1^−/− and wild-type animals (Fig. 4A). We obtained the single-cell transcriptome of 16140 cells from wildtype and 6583 cells from rfc1^−/− 4 dpf brains. We processed these data according to the Seurat standard procedure of filtering and clustering to determine cell type identity based on genetic markers of zebrafish brain cell populations at relevant developmental timepoints^39,40.

**Fig. 4: *RFC1* is expressed in early neural progenitors and is required for their proliferation.**

Consistent with our previous observations, we confirmed the general reduction in the number of cerebellar neuronal clusters defined as PCs (pvalb7⁺, ca8⁺, aldoca⁺, and ITPR1⁺) and GCs (neurod2⁺, neurod1⁺, zic5⁺, FAT2⁺, nebl⁺, olfm2b⁺, cbln12⁺, and fabp4b⁺) in rfc1^−/− brains (Fig. S6A, 4B). Moreover, we found a significant loss of cell populations constituting clusters identified as neural progenitors (her4.2⁺, her2⁺, and sox2⁺), cycling neural progenitors (pcna⁺, mki67⁺, and npm1a⁺), neuronal progenitors (dla⁺, dlb⁺, elavl3⁺, and tubb5⁺) and early neurons (lhx9⁺, elavl3⁺, and tubb5⁺) (Fig. 4B, C and Fig. S6B). Our findings suggest that early processes of neurodevelopment are perturbed in rfc1^−/− fish.

Given that cerebellar neurogenesis occurs as early as 2 dpf^38,41, we next sought to perform scRNA-seq at 2 dpf. We dissociated the heads of 2-dpf rfc1^−/− and wild-type larvae and sequenced their individual transcriptomes (18,632 cells from wildtype and 9805 cells from rfc1^−/−, Fig. 4A). Although less severe than at 4 dpf, we noticed a significant reduction in the cellular density of the same clusters defined as neural progenitors, cycling neural progenitors, neuronal progenitors and early neurons (Fig. 4D, E and Fig. S6C).

Consistent with our prior expression analyses during neurodevelopment, we found that rfc1 is mainly expressed in the neural progenitor and cycling neural progenitor clusters at 4 dpf and 2 dpf, respectively (Fig. S6B, C). Altogether, these observations suggest that RFC1 regulates the development of early neural progenitors, particularly in the developing cerebellum. To confirm this hypothesis, we compared the expression level of specific cerebellar neurogenesis markers³⁸ (neurog1, ascl1a; atoh1a, atoh1b, atoh1c, zic1, zic4, pax6a for GC neurogenesis^42,43; ptf1a, rorb for PC neurogenesis^44,45 and olig2 for eurydendroid cell neurogenesis). We found that the expression of these markers is decreased at 3 dpf in rfc1^−/− larvae compared to their control siblings by RT-qPCR (although not significant for zic1 and zic4) (Fig. 4F and Fig. S7B).

We confirmed this decrease in rfc1^−/− larval brain by whole-mount in situhybridization and fluorescent RNAscope at 3 dpf (Fig. 4G, H). We also checked the expression of these markers at 2 dpf and 4 dpf and found that while only atoh1a, ptf1a and olig2 were significantly reduced in mutants compared to wild-type at 2 dpf, they were all significantly decreased at 4 dpf (Fig. S7A–C). This suggests that these changes caused by the loss of RFC1 arise from 2 dpf with sustained consequences on cerebellar neurogenesis at 3 dpf. Altogether, these results indicate that RFC1 is expressed in early neural progenitors and is necessary for their neurogenesis, particularly in cerebellar neural progenitors.

RFC1 loss halts cerebellar progenitor proliferation, leading to their death

We next asked what molecular functions are regulated by RFC1 in neural progenitor cells in which it is expressed. To do so, we performed a differential expression analysis on our scRNA-seq datasets at 2 and 4 dpf comparing rfc1^−/− to wild-type samples (Fig. 5A, B and Supplementary Data 1, 2). Within early progenitor clusters, we found multiple genes involved in early DNA damage response, cell cycle checkpoints, cell proliferation homeostasis, DNA repair and apoptosis to be differentially expressed between rfc1^−/− and wild-type samples. In particular, we noted the overexpression of early DNA-damage sensor genes in neural progenitor clusters in rfc1^−/− brains at 2 and 4 dpf (mdc1, atm, atr, mre11, and Fanconi anemia complementation group genes (fancd2 and fancl)). These genes are activated in response to DNA damage or replication blockage, and they halt cell cycle progression by controlling critical cell cycle regulators. Particularly, ATM and ATR are thought to be master controllers of cell cycle checkpoint signaling pathways that are required for cell response to DNA damage and genome stability.

**Fig. 5: RFC1 loss-of-function impairs neural progenitor proliferation and induces cell death.**

Consistently, we noticed an increased expression of many cell cycle checkpoints in the same mutant clusters such as brca2, chek1/2, cdk6, cdk2, or cdkn1a. Notably, these genes are downstream DNA-damage triggers which stall cell cycle progression. Moreover, several genes involved in the negative regulation of cellular proliferation and cell cycle stalling, such as nupr1a, yap1, and btg2^46,47,48, were found to be increased in mutant progenitor clusters.

This is consistent with the decrease of proliferation markers such as cenpf or cdca3 in these cell populations. We also found increased expression of multiple genes involved in DNA repair processes, such as cenpx, xpc, ercc5/6 and rpa1/2 (Fig. 5A, B). Those DNA repair and tumor-suppressor genes are activated by cell cycle checkpoint controllers upon alteration in DNA integrity^{49,50,51,52,53}. Finally, we showed that the expression of multiple pro-apoptotic genes (tp53, baxa, casp8, mdm2, and mdm4)^54,55,56,57 and p53 target genes (rrm2, ccng1, rps27l, sesn1, and sesn3) is increased in these same clusters (Fig. 5A, B). Notably, these gene expression changes are predominant in neural progenitor clusters but also in retinal progenitors at 2 dpf and glia progenitors at 4 dpf.

We wanted to confirm these transcriptomic changes with a series of immunofluorescence labelings in toto. To do so, we first immunolabelled embryos using an antibody anti-phosphorylated histone (gH2AX), a marker of DNA damage at the level of the chromatin⁵⁸. We found a significant increase in the number of gH2AX-positive cells in the developing brain of rfc1^−/− embryos compared to control siblings (Fig. 5C, F, I and Fig. S8A).

Moreover, we assessed cell proliferation using an anti-phospho-histone 3 (PH3) antibody and showed that from 2 dpf, the number of proliferative cells in the developing hindbrain is significantly reduced in rfc1^−/− embryos (Fig. 5D, G, J and Fig. S8B). Notably, these changes are exacerbated and sustained at 3 dpf in rfc1^−/− larvae compared to their siblings (Fig. 5E, H, K and Fig. S8C). Finally, when quantifying the number of apoptotic cells at the same stages using an anti-activated caspase-3 antibody, we found a significant increase in the number of dying cells both at 2 and 3 dpf in rfc1^−/− specifically (Fig. 5E, H).

Interestingly, we do not detect this increase in DNA damage and apoptosis at later stages (4, 5 dpf), reinforcing that RFC1 plays a role during early cerebellar neurogenesis (Fig. S8A–C). We confirmed that these changes occur broadly in the developing brain of rfc1^−/− embryos, including in the developing cerebellum (Fig. 5F–K). Particularly, by co-labeling cerebellar neural progenitors (atoh1a and ptf1a) with γH2AX foci, we confirmed that DNA damages accumulate in cerebellar neural progenitors at 2 dpf (Fig. 6A, B and S8D, E).

**Fig. 6: Cerebellar neural progenitors accumulate DNA damage and undergo apoptosis in the absence of RFC1.**

Finally, we correlated these molecular events (DNA damages and apoptosis) at the level of individual cells within the developing cerebellum, by co-labeling apoptotic DNA fragmentation using TUNEL and gH2AX-positive DNA damage foci in rfc1^−/− at 2 dpf (Fig. 6C, D, white arrows). Although we could not detect any colocalization of these sparse signals in wild-type brains, we observed overlapping γH2AX and TUNEL signals throughout the midbrain and hindbrain in rfc1^−/− embryos (Fig. 6C, D).

Additionally, we identified the characteristic γH2AX apoptotic ring pattern, a hallmark of late-stage apoptosis associated with global chromatin fragmentation^59,60; the latter being the main effector for the apoptotic γH2AX phosphorylation. The presence of this ring, together with apoptotic marker colocalization, and the concomitant reduction in cell proliferation, reinforces that rfc1^−/− neural progenitor cells accumulate DNA damages, which trigger the stalling of cell cycle progression and, eventually, programmed cell death.

Discussion

The RFC1 gene has recently caught attention due to the recent identification of biallelic repeat expansions associated with a wide range of movement disorders, including late-onset cerebellar ataxias such as CANVAS^14,16,17,19. However, the underlying pathogenic mechanisms were unknown, and this lack of knowledge is largely explained by the fact that the RFC1 gene function has never been investigated in vivo. Our study is the first to investigate the function of RFC1 during neurodevelopment, particularly in the cerebellum, the region primarily affected by ataxic cerebellar disorders. We showed that RFC1 is expressed during cerebellar embryonic development, particularly in expanding neural progenitors in the early zebrafish embryo. Importantly, we showed that the loss of RFC1 led to a severe cerebellar phenotype characterized by a significant reduction in PCs and GCs and a consistent motor dysfunction.

We unveiled that the loss of RFC1 jeopardizes the genomic integrity of these progenitor cells, which triggers cell cycle arrest and ultimately leads to their programmed death. Indeed, the spatial and temporal correlation between γH2AX-positive DNA damage accumulation and apoptosis underscores the functional link between DNA damage and cell death fate in the absence of RFC1. Thus, our work is the first to show that RFC1 plays a crucial role in cerebellar neurogenesis in vivo.

There are multiple DNA repair genes which have been associated with hereditary ataxias: TDP1 is associated with spinocerebellar ataxia with axonal neuropathy-1 (SCAN1)⁶¹; XRCC1 is associated with autosomal recessive cerebellar ataxia^62,63; PNKP is associated with ataxia-oculomotor apraxia-4 (AOA4)^64,65; APTX is associated with ataxia-oculomotor apraxia-1 (AOA1); PCNA is associated with ataxia-telangiectasia–like disorder-2 (ATLD2)⁶⁶; MRE11 is associated with ataxia-telangiectasia–like disorder-1 (ATLD1)^67,68 and ATM is associated with ataxia-telangiectasia⁶⁹. The mechanism by which ataxia-associated DNA repair genes lead to preferential cerebellar dysfunction is unknown. It has been suggested that the cerebellum and its main functional components, Purkinje cells, are particularly vulnerable to DNA damage and are therefore more sensitive to changes in these proteins’ functions. In this work, we showed that the expression pattern of RFC1 is not ubiquitous during embryo development, as it could have been expected. Moreover, in the postnatal cerebellum, RFC1 expression remains restricted to Purkinje Cells in the cerebellar cortex in mice (Fig. 1G–J).

Thus, this restricted and specific expression pattern could explain how deleterious genetic variations in such universal genes can lead to tissue-specific defects. As a result, instead of explaining the cerebellar-specific effects of these variants by a greater sensitivity of this tissue/cells to DNA damage, it is possible that these genes are expressed in a very specific way at the tissue and cellular level, which explains the specificity of the defects when altered. Because they are involved in late-onset disorders, the mechanisms by which genetic variations in these DNA repair genes, such as RFC1, are mainly speculated in the context of post-mitotic neurons. However, our work suggests that these genes can be specifically expressed in the developing cerebellum and are essential for normal cerebellar neurogenesis. Notably, the investigation of these causative DNA repair genes’ function in vivo is hampered by the embryonic lethality of mouse KO models⁷⁰. Our study reveals that it could be insightful to investigate their function in vivo in other vertebrate species, such as zebrafish.

Our data showed that the developing cerebellum is primarily affected by RFC1 loss-of-function, which is consistent with RFC1’s expression pattern in the developing hindbrain. However, it is worth noting that RFC1 is also expressed in the periventricular layer of the optic tectum as well as in the developing eye, particularly at the level of the ciliary marginal zone. This suggests that it plays a broader role in neurodevelopment, particularly in retinal neurogenesis and explains why the main transcriptomic changes we describe within brain progenitor populations are also found in retinal progenitor cells (Fig. 5A, B). These results are also consistent with the strikingly reduced eye size phenotype observed in rfc1^−/− larvae compared to their siblings. Although no retinal syndrome has previously been associated with RFC1 hypomorphism, our data suggest that genetic variants in RFC1 could be associated with it. However, understanding the mechanisms underlying this phenotype in the eye warrants future studies.

Our results can also be discussed regarding the uncertain pathogenic mechanisms underlying CANVAS. To date, no consensus exists on whether repeated expansions in RFC1 can cause a gene loss-of-function or a toxic gain of function. Indeed, although truncating variants in RFC1 have been found in CANVAS patients^20,21 and that in silico analyses suggest repeat-formed secondary DNA structures which could prevent normal transcription⁷¹, no irrefutable evidence of a loss of expression in RFC1 has yet been provided in post-mortem tissues or in vitro cultures of patient cells. Nevertheless, our data provide evidence that RFC1 is essential for cerebellar development by ensuring genomic integrity during the expansion of neural progenitor populations. Thus, one can speculate that early deleterious changes in RFC1’s function or expression during development could affect the proper development of the cerebellum and that these developmental defects could also contribute, at least in part, to the pathogenicity of these repeat expansions. Moreover, it would be interesting to know if structural defects in cerebellar cortical organization could be present before the appearance of symptoms and contribute to the increased sensitivity of the system to DNA damage, for example. This would require generating relevant biological models of the disease, which are complex to mimic, particularly for large repeat expansions.

Finally, our expression data in mice show that RFC1 is expressed in post-mitotic calbindin-positive Purkinje cells at P60 in a mature cerebellum (Fig. 1J). However, we noticed that RFC1 expression in mature PCs is mainly cytoplasmic. It is, therefore, possible to speculate that RFC1 might play a distinct role beyond preserving genomic integrity in these post-mitotic neurons. It is worth noting that other DNA replication and repair, protein such as PCNA have been found cytoplasmic with a role in several cellular processes in differentiated cells^72,73.

In sum, our data improve our general knowledge of the in vivo function of the RFC1 gene and open new avenues of study that may help elucidate the pathogenic mechanisms underlying so-called RFC1-related disorders.

Limitations of the study

Late-onset ataxia is, by definition, a disorder occurring at an adult stage. Therefore, our results show that the role of RFC1 during cerebellum neurogenesis cannot be directly transposed to explain the pathogenic mechanisms underlying CANVAS and other late-onset RFC1-related disorders. More work is needed to determine the function of RFC1 in post-mitotic cerebellar neurons, particularly for regulating the maintenance of genomic integrity against DNA damage, against which the cerebellum seems particularly sensitive. Moreover, our functional investigation was carried out on zebrafish, which is a vertebrate. While the cellular organization of the cerebellum has been well conserved from teleosts to mammals, there are some obvious anatomical differences. For instance, the molecular pathways equivalents of the mammalian deep cerebellar nuclei reside in isolated eurydendroid cells (which we show to be affected in rfc1^−/− larvae, Fig. 3G, H) located within the cerebellar cortex in teleosts³⁸. Nevertheless, the molecular pathway regulating its neurogenesis and maturation (including GC internal migration) are well conserved. Moreover, the cerebellum’s neuronal cell types, connectivity, and function are highly conserved from teleosts to humans, making the cerebellum one of the highest evolutionary conserved tissues throughout jawed vertebrate brains.

Methods

Experimental models

All experimental procedures involving mice were approved by the institutional ethics committee of the Université du Québec à Montréal (CIPA # 650) in accordance with the biomedical research guidelines of the Canadian Council of Animal Care (CCAC). FVB/N mice initially purchased from Charles River (strain code 207) were maintained and bred in individually ventilated cages within the conventional animal facility of the Université du Québec à Montréal, under 12 h light–12 h dark cycles (7AM to 7PM) at 21.5 °C and 50% humidity and with ad libitum access to regular chow diet (Charles River Rodent Diet #5075, Cargill Animal Nutrition). The brain was collected from newborns, juveniles (postnatal day P7 and P11) and adults (P60) animals of both sexes. Newborns were sacrificed by decapitation, while all P7-P11-P60 mice were sacrificed by transcardiac perfusion with 4% PFA after isoflurane-mediated anesthesia.

Routine zebrafish (Danio rerio) maintenance was performed following standard procedures (Westerfield) at 28.5 °C under 12/12 h light/dark cycles at the animal facility of the University of Montreal Hospital Research Center (CRCHUM), Montreal, QC, Canada. All experiments complied with the guidelines of the Canadian Council for Animal Care. The primary model used for this study was rfc1 loss-of-function mutants. For generation of the rfc1 loss-of-function model, zebrafish optimized Cas9 mRNA was synthesized using the mMESSAGE mMACHINE SP6 kit (Ambion) and a pCS2-nCas9n (Addgene #47929) template linearized with NotI. A single-guideRNA sequence (5’-TGGGGTTGGTGCCACCTTAG-3’) targeting the 5th exon of rfc1 was designed using the online tool CRISPRscan and ordered from Synthego, CA, USA. (Tubingen long fin (TL) wild-type embryos were collected for microinjection. A 1nL drop of a mix of 100 ng/µL of Cas9 mRNA and 30 ng/µL of gRNA was injected into one-cell stage embryos using a Picospritzer III pressure ejector. Genotyping of rfc1+/+ (wild-type), +/− (heterozygous) and −/− (homozygous) animals was performed by high-resolution melting analysis (HRM) using genomic DNA extracted by boiling the embryo/larva/clipped caudal fin in 50 mM NaOH for 10 min and then neutralizing it with 100 mM Tris HCl (pH 8). HRM primers were designed using the Universal Probe Library Assay Design Center (Fwd: 5’ GCCATTACTGATGTGGGCATCTG 3’ and Rev: 5’ TGAAAGTCTCCTCCAGCAAATCC 3’). All primer sets are available upon request. The PCR reactions were made with 5 µL of the Precision Melt Supermix for HRM analysis (Bio-Rad #172-5112), 0.5 µL of each primer (10 µM) and 2 µL of genomic DNA and water up to 10 µL. The PCR was performed in a LightCycler 96 Instrument (Roche) using white 96-well plates. The two-step Evagreen PCR reaction protocol was 95 °C for 2 min, then 45 cycles of 95 °C for 10 s and 60 °C for 30 s, followed by 95 °C for 30 s, 60 °C for 60 s, the temperature was increased by 0.02 °C/s until 95 °C for 10 s, then cooling at 40 °C. Curves were analyzed using the Roche LightCycler 96 software. Moreover, polymerase chain reaction (PCR) and sequencing were used to verify the mutation’s exact sequence. The following primers were used for genomic PCR: Fwd 5′ TCACGCCTGTAATCCCAGCATTG 3′ and Rev 5′ TCTTGAAGAATAGCTGTGTTGTCCTGTCAC 3′. PCR amplicons were loaded on an agarose gel 1.5% (A87-500G, FroggaBio), and single-band samples were sent to CES Génome Québec (Montréal, Canada) for Sanger sequencing.

To eliminate any putative off-target mutations, the F0 founder was outcrossed with Tubingen long fin wild-type fish for at least three generations before phenotyping the embryos. All experiments were performed on larvae, and the zebrafish sex is not yet determined at this stage. rfc1^+/− fish were incrossed to obtain rfc1+/+, +/−, and −/− from the same cross for experiments. Rfc1+/− fish of different generations have been used to verify the consistency of the phenotype.

Zebrafish morphological analysis

Rfc1^+/+, ^+/−, and ^−/− larvae were anesthetized with buffered tricaine methanesulfonate (MS222) and dorsal and lateral images of the whole body were acquired using a 4x magnification under a stereomicroscope (Leica). Body length, eye size, head size and surface area were measured manually using ImageJ software (version 2.3.0). Body length was measured from the anterior tip of the head and the posterior top of the tail. Eye size and otic vesicle diameter or surface area were measured by specifying the eye and vesicle boundaries.

Survival monitoring

About 50 embryos were raised in a dedicated glass beaker, and death was monitored daily. Dead larvae were collected and processed for genomic DNA extraction and high-resolution melting genotyping on the same day. At least two independent batches of larvae were monitored, each containing at least 50 larvae. Data were gathered in GraphPad Prism (version 10) for Kaplan–Meier survival analysis.

Locomotion behavior

At 4 dpf, larvae were separated into single wells of a 96-well plate containing ~200 µL of E3 medium. The plate was placed inside a Daniovision® recording chamber (Noldus). Before the start of the experiment, larvae were habituated for 30 min in the dark. Larval locomotor activity was monitored over 1 h light, 1 h dark cycles over 4 h. The Ethovision XT13 software (Noldus) was used to quantify the total swimming distance per period. Moreover, the swimming tracks were extracted from the integrated view panel of the Ethovision software.

Mouse perfusion and brain cryosection

Mice brains were removed after either head decapitation (P0) or after isoflurane anesthesia and transcardiac perfusion (P7, P11, and P60). Transcardiac perfusion was performed with 1X phosphate-buffered saline (PBS), followed by 4% (w/v) paraformaldehyde solution (PFA). PFA is commonly used to remove blood and preserve the brain for immunostaining. Removed mouse brains were postfixed in 4% PFA solution and left for 24 h at 4 °C. The PFA solution was replaced the next day with a 30% (w/v) sucrose solution (cryoprotection), and the brains were stored at 4 °C until they sank to the bottom of the tube. A cryostat (LEICA CM 1950) was used to slice brain tissues into 10 µm sagittal brain sections (on glass slides) that were kept at −80 °C.

Mouse brain Immunostaining

Brain slices on slides were permeabilized for 2 h in a blocking solution (10 % fetal bovine serum and 1% Triton X-100 in 1X PBS). Sections were then incubated overnight at 4 °C with specific primary antibodies: anti-Calbindin D-28K (cb300, Swant, Switzerland; 1:200) or anti-PAX6 (Biolegend, UK; 1: 200). Subsequently, brain sections were incubated with secondary antibodies, diluted in the same blocking buffer, at room temperature for 2 h. All sections were mounted and counterstained with DAPI (1:1000). A Nikon A1 confocal microscope equipped with NIS-Elements C software (Nikon) was used to acquire images, which were then analyzed using ImageJ.

Zebrafish in toto immunofluorescent staining

Larvae were fixed in 4% paraformaldehyde (28908, Thermo Scientific) overnight at 4 °C. The day after, PBS1X-Triton 1% (X-100-500ML, Sigma-Aldrich) was used for several washes over 1 h. Larvae were then incubated in a blocking solution (2% NGS, 1% BSA, 1% DMSO, and PBS1X-Triton 1%) (005-000-121, Jackson Laboratory) for 30 min, and then rinsed several times with PBS1X-Triton 1 %. Larvae were incubated in primary antibodies: anti-pvalb7 (gift from M. Hibi; 1:1000), anti-neurod1 (gift from M. Hibi; 1:500), anti-vglut1 (gift from M. Hibi; 1:1000), anti-gH2AX (GTX127342, GenTex; 1:1000), anti-PH3 (Millipore 09838; 1:500), or anti-cleaved caspase-3 (9661, Cell Signaling; 1:200). The following day, after several washes in PBS1X-Tween 0.1% (P5927, Sigma), larvae were incubated with a secondary antibody diluted in blocking solution: goat anti-mouse Alexa Fluor 488 (A11029, Millipore Sigma; 1:500), goat anti-rabbit Alexa Fluor 488 (A11034, Millipore Sigma; 1:500). For imaging, larvae were positioned in 0.5% low-melting point agarose (16520-050, Thermo Scientific). and acquired using a Zeiss or an Olympus confocal microscope at a 20X or 40X magnification. Images were processed using ZEN (Zeiss) or Volocity (Quorum Technologies) software. Cell counting was performed manually using Image J.

Hematoxylin & Eosin coloration

Briefly, whole 4 dpf larvae or microdissected adult brains were embedded in paraffin and sectioned transversally at 5 µm using Leica microtome (Jung Biocut). The hematoxylin staining was incubated for 4 min, washed with alcohol-acid and rinsed with tap water. Finally, slides were incubated for 10 s in lithium carbonate solution and for 2 min in Eosin staining. Stained samples were mounted under a coverslip using Permount mounting media (3989, Sigma-Aldrich).

Probe cloning and in situ hybridization (RNAscope)

A Specific 504 bp rfc1 probe was cloned from 2 dpf zebrafish cDNAs, using the following primers: Fwd- 5’ TCACGCCTGTAATCCCAGCATTG 3’; Rev- 5’ TCTTGAAGAATAGCTGTGTTGTCCTGTCAC 3’) within the pCS2+ vector using TOPO TA cloning kit (Invitrogen). Digoxygenin(DIG)-labeled sense and anti-sense RNA probes were in vitro synthesized using the DIG RNA Labeling Kit (Millipore Sigma, Roche-11175033910).

Whole-mount in situ hybridization was performed as described previously⁷⁴. Briefly, larvae were previously fixed in PFA 4% overnight at 4 °C. Embryos were digested with proteinase K (AM2546, Invitrogen): 1 and 2 dpf embryos were incubated at 10 µg/mL for 20 and 30 min, respectively; 3, 4, and 5 dpf larvae were incubated at 20 µg/mL for 30, 35, and 40 min, respectively. After several washes of PBS1X-Tween 0.1%, embryos were fixed with PFA 4% for 20 min. A prehybridizing step was performed at 69 °C for 4 h in MH Buffer (50% formaldehyde (FOR001.1, Bioshop), 0.1% Tween, 5X SSC (880-040-LL, Multicell), 11.5 mM citric acid pH6, 50 µg/mL heparin and 500 µg/mL tRNA). DIG probes (11093274910, Sigma-Aldrich) were incubated in MH Buffer at 69 °C overnight. The day after, several washes with a prewarm medium were performed. Then, embryos were blocked for 1–4 h at room temperature in Blocking Buffer (2% sheep serum, 2 mg/ml BSA, PBT). Preabsorbed anti-DIG antibody was added overnight at 4 °C incubation under gentle agitation. On the third day, samples were washed with PBS1X-Tween 0.1% and colored using BCIP/NBT (11681451001, Sigma-Aldrich). Larvae were imaged in 80% glycerol (G5516-1L, Sigma).

RNA extraction and qPCR

RNA was isolated from pools of five larvae using the RNeasy Plus Mini Kit (74134, Qiagen), according to the manufacturer’s protocol. Briefly, 500 ng of total RNA was used for cDNA synthesis using the SuperScript®Vilo™ Kit (11754050, Invitrogen).

The cDNA was diluted 1:10 and used for qPCR with SYBR Green I Master (Roche) on a LightCycler 96 instrument. Polr2d and actin B were used as reference genes for normalization. Primers were designed using A plasmid editor (ApE) and span different exons: rfc1 (Forward) 5′-CCACAGAGAAGAAACACCCAGTAAC-3′ and rfc1 (Reverse) 5′-GCCGTTCCATTCTCTTTAGGTTTG-3′; atoh1a (Forward) 5′-GAGGCGAAGAATGCACGGATTG-3′ and atoh1a (Reverse) 5′-TAGTAAGTCGGACAGGGCGTTG-3′; atoh1b (Forward) 5′-TGGGAGTAAAGCATCTGTCAGTGG-3′ and atoh1b (Reverse) 5′-TCTCCAATGAAGGAATGACGCTTCTC-3′; atoh1c (Forward) 5′-GCTAACGCCCGAGAGAGGAG-3′ and atoh1c (Reverse) 5′- TCTGTGCCATCTGAAGCGTCTC-3′; ptf1a (Forward) 5′-CCACACAGTGACGCCTTAAACC-3′ and ptf1a (Reverse) 5′-GAGAGTGTCCTGCGAGAGGAG-3′; polr2d (Forward) 5′-AACGCAAAGTGGGAGATGTG-3′ and polr2d (Reverse) 5′-AGCGTCTCTGCGTTCTCAA-3′); actinb (Forward) 5′-CCAGCTTTTCAGCCTCACTT-3′ and actinb (Reverse) 5′-CGGCAATTTCATCATCCAT-3′. The 2^−∆∆CT algorithm was used to analyze the relative changes in gene expression between samples.

Single-cell dissociation

Larval brains were microdissected from 4 dpf pre-genotyped larvae using fine forceps in calcium-free Ringers’ solution. For 2 dpf embryos, heads were cut off using sharp microblades. A total of four brains or three heads were pooled per sample and transferred into a 1.5 ml tube containing 250 µl of dissociation solution containing 0.02% papain (LS003124, Worthington), 0.02% DNase I (LK003170, Worthington), and 12 mg/ml L-cysteine into 5 mL of DMEM/F12 (11320-033, Gibco). Tissues were digested at 37 °C for 10 min and slowly pipetted up and down ten times, then incubated for an additional 10 min followed by pipetting up and down 10–13 times. To stop the enzymatic digestion, 1 ml of washing solution (30% glucose, 0.5% HEPES (15630-080, Gibco), and 5% FBS in DPBS 1X (14040-141, Gibco)) was added. Samples were then centrifuged at 4 °C at 800×g for 5 min. The supernatant was discarded and 50 µl of washing solution was added to resuspend the pellet. The cell suspension is filtered out with a cell strainer (136800040, SP Bel-Art, Wayne, USA) to remove aggregates or duplets. Single-cell suspension and cell number was assessed on a Kova slide, and cell mortality was assessed using Trypan blue. Fresh single-cell samples were sent to the Montreal Clinical Research Institute (IRCM) genomic platform. Libraries were prepared using the Chromium Single Cell 3’ reagent kits. Reads obtained from the 10X Genomics platform were processed using the 10X Genomics Cell Ranger v7.1.0 pipeline and Cloud Analysis⁷⁵, which includes STAR v2.7.2a alignment⁷⁶ against the Danio Rerio GRCz11 genome with the corresponding Ensembl annotations (release 110)⁷⁷.

Single-cell data analysis

The bioinformatics analysis was performed in R (v4.3.1) using packages centered around Seurat v5.0.0⁷⁸. Quality control steps included cell scoring with PercentageFeatureSet for mitochondrial genes (^mt-), ribosomal genes (^rp[sl]), and hemoglobin genes (^hba.*$|^hbb.*$). Cells were removed when they had poor sequencing depth (features ≤300; counts ≤500), abnormally high mitochondrial content (≥20%)⁷⁹, or significant hemoglobin content (≥1%). Potential multiplets were removed by filtering out cells with outlier sequencing depth (features ≥6000; counts ≥50,000). Features not present in at least three cells were also excluded from the analysis.

The Seurat RNA assay data were then log-normalized with NormalizeData default parameters, and scaled across all remaining genes. Additionally, an SCT assay was generated using SCTransform (vst.flavor = v2), and mitochondrial content was regressed during the gene expression transformation (vars.to.regress)⁸⁰.

Clustering, annotations, and population composition analysis

Dimensionality reduction calculations, such as the principal component analysis (PCA) and subsequent uniform manifold approximation and projection (UMAP), were utilized in the SCT assay. The appropriate number of dimensions for the initial clustering was determined using a combination of ElbowPlot and Jackstraw functions to assess the significance of the top 50 principal components⁸¹.

The FindMarkers function identified differentially expressed genes within each resulting cluster, which served to assess cell identities. Cell type annotations were curated manually by comparing top markers to annotations databases such as Daniocell and Zebrahub^40,82. Subclustering was conducted to ensure cluster identity was not driven by small populations of specialized cells. Cell types were also validated by projecting online annotations onto the cells to observe correlating results³⁸. While top markers of each presented cluster can be visualized through dot plots (Fig. S4B, C), more extensive marker lists are available (Supplementary Data 1, 2). Markers that define cell identity are provided for 2 dpf and 4 dpf with a cutoff of log fold change >0.5, and a minimum percentage of expression across cells of 0.2.

Determination of differential cell population abundance between conditions was computed using the DCATS v0.99.6 beta-binomial generalized linear statistical model to compare annotated clusters⁸³. Cell ratio bar plots utilized the DittoSeq v1.12.2 package⁸⁴. Subsequently, the FindMarkers function (logfc.threshold = 0.5; min.pct = 0.2) was employed to identify differentially expressed genes associated with significantly altered cell types between wild type and knock-out samples.

Imaging

Imaging was performed using either a Zeiss Axio Observer.Z1 equipped with a Yokogawa CSU-X1 spinning disk confocal module (Zeiss, Germany) and a motorized stage (ASI system with an MCL nano-drive piezo for precise focus control) or an Olympus confocal microscope. For spinning disk confocal imaging, images were acquired using ZEN 2.6 (Blue Edition) software (version 2.6.76.00000). Excitation was provided by a diode laser at 488 nm, controlled by acousto-optic tunable filters (AOTF), and emission was optimized for Alexa Fluor 488 using filters designed for a peak emission around 517 nm (DBP 527/54 + 645/60 from Chroma). Detection was performed using an EMCCD Evolve 512 monochrome camera (Photometrics) (Photometrics, PvCamEvolve_512) with a native pixel size of 16 µm, a 512 × 512 resolution, 16-bit depth, and a 1.2x EMCCD camera adapter. A Plan-Apochromat 20×/0.8 M27 objective (Zeiss, Germany) was used, with a physical pixel size of 0.67 × 0.67 µm in the XY plane. Z-stacks were acquired with an interval of 0.54 µm. Tiling was used to acquire larger fields of view, and the images were stitched using the ZEN 2.6 (blue edition) Tiling module, ensuring seamless reconstruction of the larger areas with 10% overlap. Post-acquisition image processing was carried out using ImageJ or Volocity (Quorum Technologies). Manual cell counting was performed with ImageJ/FIJI. Brightness and contrast adjustments were applied uniformly across datasets to enhance image clarity while preserving data integrity.

Quantification and statistical analysis

Morphological and locomotion parameters were quantified using DanioScope and EthoVision software (Noldus) as described above. All data acquisition and analyses were performed by experimenters blinded to the genotype of the larvae. For all experiments, “N” refers to the number of independent biological replicates (i.e., distinct batches of embryos), and “n” refers to the number of individual larvae analyzed per condition. Each experiment was independently repeated at least twice. Before conducting parametric tests, data were tested for normality using the Shapiro–Wilk test. When data met the assumptions of normality, an unpaired Student’s t-test (for two-group comparisons) or a one-way analysis of variance (ANOVA) (for comparisons among more than two groups) was performed. When ANOVA revealed a significant effect, Tukey’s post hoc test was used for multiple comparisons. The exact statistical test used for each analysis is now indicated in the corresponding figure legends. All statistical analyses and graphs were generated using GraphPad Prism software. Data were expressed as mean ± standard error of the mean (SEM). Significance thresholds were set as follows: ns (p > 0.05); * (p ≤ 0.05); ** (p ≤ 0.01); *** (p ≤ 0.001); **** (p ≤ 0.0001).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Single-cell RNA-seq data have been deposited in the NCBI BioProject PRJNA1126282 and are publicly available as of the date of publication. Original in situ hybridization images and microscopy data reported in this paper can be shared by the lead contact upon request. The rfc1-KO zebrafish line generated in this study is cryopreserved at the University of Montreal Hospital Research Center (CRCHUM) and is available upon request and signature of a Material Transfer Agreement between the CRCHUM and the requestor. Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Dr. Éric Samarut (eric.samarut@umontreal.ca). Source data are provided with this paper.

Code availability

All original code have been deposited and made publicly available on Zenodo as of the date of publication. (https://doi.org/10.5281/zenodo.15499729). Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

References

Voogd, J. The human cerebellum. J. Chem. Neuroanat. 26, 243–252 (2003).
Article PubMed Google Scholar
Purves, D. et al. Circuits within the Cerebellum (Sinauer Associates, 2001).
Fine, E. J., Ionita, C. C. & Lohr, L. The history of the development of the cerebellar examination. Semin. Neurol. 22, 375–384 (2002).
Article PubMed Google Scholar
Roostaei, T., Nazeri, A., Sahraian, M. A. & Minagar, A. The human cerebellum: a review of physiologic neuroanatomy. Neurol. Clin. 32, 859–869 (2014).
Article PubMed Google Scholar
Manto, M. et al. Consensus paper: roles of the cerebellum in motor control-the diversity of ideas on cerebellar involvement in movement. Cerebellum 11, 457–487 (2012).
Article PubMed PubMed Central Google Scholar
Kim, L. H., Heck, D. H. & Sillitoe, R. V. Cerebellar functions beyond movement and learning. Ann. Rev. Neurosci. 47, 145–166 (2024).
Koziol, L. F. et al. Consensus paper: the cerebellum’s role in movement and cognition. Cerebellum 13, 151–177 (2014).
Article PubMed PubMed Central Google Scholar
Strick, P. L., Dum, R. P. & Fiez, J. A. Cerebellum and nonmotor function. Annu. Rev. Neurosci. 32, 413–434 (2009).
Article CAS PubMed Google Scholar
Schmahmann, J. D. Disorders of the cerebellum: ataxia, dysmetria of thought, and the cerebellar cognitive affective syndrome. J. Neuropsychiatry Clin. Neurosci. 16, 367–378 (2004).
Article PubMed Google Scholar
Bodranghien, F. et al. Consensus paper: revisiting the symptoms and signs of cerebellar syndrome. Cerebellum 15, 369–391 (2016).
Article PubMed PubMed Central Google Scholar
Tsang, H., Addepalli, K. & Davis, S. R. Resources for interpreting variants in precision genomic oncology applications. Front. Oncol. 7, 214 (2017).
Article PubMed PubMed Central Google Scholar
Ngo, K. J. et al. A diagnostic ceiling for exome sequencing in cerebellar ataxia and related neurological disorders. Hum. Mutat. 41, 487–501 (2020).
Article CAS PubMed Google Scholar
van Gaalen, J. & van de Warrenburg, B. P. A practical approach to late-onset cerebellar ataxia: putting the disorder with lack of order into order. Pr. Neurol. 12, 14–24 (2012).
Article Google Scholar
Cortese, A. et al. Biallelic expansion of an intronic repeat in RFC1 is a common cause of late-onset ataxia. Nat. Genet. 51, 649–658 (2019).
Article CAS PubMed PubMed Central Google Scholar
Rafehi, H. et al. Bioinformatics-based identification of expanded repeats: a non-reference intronic pentamer expansion in RFC1 causes CANVAS. Am. J. Hum. Genet. 105, 151–165 (2019).
Article CAS PubMed PubMed Central Google Scholar
Sullivan, R. et al. RFC1-related ataxia is a mimic of early multiple system atrophy. J. Neurol. Neurosurg. Psychiatry 92, 444–446 (2021).
Article PubMed Google Scholar
Ylikotila, P. et al. Association of biallelic RFC1 expansion with early-onset Parkinson’s disease. Eur. J. Neurol. 30, 1256–1261 (2023).
Article PubMed Google Scholar
Malaquias, M. J. et al. Multisystemic RFC1-related disorder: expanding the phenotype beyond cerebellar ataxia, neuropathy, and vestibular areflexia syndrome. Neurol. Clin. Pr. 13, e200190 (2023).
Article Google Scholar
Davies, K., Szmulewicz, D. J., Corben, L. A., Delatycki, M. & Lockhart, P. J. RFC1-related disease: molecular and clinical insights. Neurol. Genet. 8, e200016 (2022).
Article PubMed PubMed Central Google Scholar
Benkirane, M. et al. RFC1 nonsense and frameshift variants cause CANVAS: clues for an unsolved pathophysiology. Brain 145, 3770–3775 (2022).
Article PubMed Google Scholar
Ronco, R. et al. Truncating variants in RFC1 in cerebellar ataxia, neuropathy, and vestibular areflexia syndrome. Neurology 100, e543–e554 (2023).
Article CAS PubMed PubMed Central Google Scholar
Arteche-Lopez, A. et al. New cerebellar ataxia, neuropathy, vestibular areflexia syndrome cases are caused by the presence of a nonsense variant in compound heterozygosity with the pathogenic repeat expansion in the RFC1 gene. Clin. Genet. 103, 236–241 (2023).
Article CAS PubMed Google Scholar
Abdi, M. H., Zamiri, B., Pazuki, G., Sardari, S. & Pearson, C. E. Pathogenic CANVAS-causing but not non-pathogenic RFC1 DNA/RNA repeat motifs form quadruplex or triplex structures. J. Biol. Chem. https://doi.org/10.1016/j.jbc.2023.105202 (2023).
Maltby, C. J. et al. AAGGG repeat expansions trigger RFC1-independent synaptic dysregulation in human CANVAS neurons. Sci. Adv. 10, eadn2321 (2024).
Zhang, G., Gibbs, E., Kelman, Z., O’Donnell, M. & Hurwitz, J. Studies on the interactions between human replication factor C and human proliferating cell nuclear antigen. Proc. Natl Acad. Sci. USA 96, 1869–1874 (1999).
Article CAS PubMed PubMed Central Google Scholar
Bowman, G. D., O’Donnell, M. & Kuriyan, J. Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex. Nature 429, 724–730 (2004).
Article CAS PubMed Google Scholar
Zhuang, Z., Yoder, B. L., Burgers, P. M. & Benkovic, S. J. The structure of a ring-opened proliferating cell nuclear antigen-replication factor C complex revealed by fluorescence energy transfer. Proc. Natl Acad. Sci. USA 103, 2546–2551 (2006).
Article CAS PubMed PubMed Central Google Scholar
Bunz, F., Kobayashi, R. & Stillman, B. cDNAs encoding the large subunit of human replication factor C. Proc. Natl Acad. Sci. USA 90, 11014–11018 (1993).
Article CAS PubMed PubMed Central Google Scholar
Overmeer, R. M. et al. Replication factor C recruits DNA polymerase delta to sites of nucleotide excision repair but is not required for PCNA recruitment. Mol. Cell Biol. 30, 4828–4839 (2010).
Article CAS PubMed PubMed Central Google Scholar
Hashiguchi, K., Matsumoto, Y. & Yasui, A. Recruitment of DNA repair synthesis machinery to sites of DNA damage/repair in living human cells. Nucleic Acids Res. 35, 2913–2923 (2007).
Article CAS PubMed PubMed Central Google Scholar
Liu, Y., Deng, Y., Li, G. & Zhao, J. Replication factor C1 (RFC1) is required for double-strand break repair during meiotic homologous recombination in Arabidopsis. Plant J. 73, 154–165 (2013).
Article CAS PubMed Google Scholar
Yao, T., Jin, D., Liu, Q. & Gong, Z. Abscisic acid suppresses the highly occurred somatic homologous recombination in Arabidopsis rfc1 mutant. J. Genet. Genomics 40, 465–471 (2013).
Article CAS PubMed Google Scholar
Qian, J. et al. Arabidopsis replication factor C4 is critical for DNA replication during the mitotic cell cycle. Plant J. 94, 288–303 (2018).
Article CAS PubMed Google Scholar
Kani, S. et al. Proneural gene-linked neurogenesis in zebrafish cerebellum. Dev. Biol. 343, 1–17 (2010).
Article CAS PubMed Google Scholar
Dey, M. R. et al. Granule cells constitute one of the major neuronal subtypes in the molecular layer of the posterior cerebellum. eNeuro https://doi.org/10.1523/ENEURO.0289-21.2022 (2022).
Laboratory, T. J. Alleles produced for the KOMP project by The Jackson Laboratory. MGI Direct Data Submission (2012).
Chang, W., Pedroni, A., Koster, R. W., Giacomello, S. & Ampatzis, K. Purkinje cells located in the adult zebrafish valvula cerebelli exhibit variable functional responses. Sci. Rep. 11, 18408 (2021).
Article CAS PubMed PubMed Central Google Scholar
Pose-Mendez, S., Schramm, P., Valishetti, K. & Koster, R. W. Development, circuitry, and function of the zebrafish cerebellum. Cell Mol. Life Sci. 80, 227 (2023).
Article CAS PubMed PubMed Central Google Scholar
Raj, B. et al. Emergence of neuronal diversity during vertebrate brain development. Neuron 108, 1058–1074 e1056 (2020).
Article CAS PubMed PubMed Central Google Scholar
Sur, A. et al. Single-cell analysis of shared signatures and transcriptional diversity during zebrafish development. Dev. Cell 58, 3028–3047 e3012 (2023).
Article CAS PubMed PubMed Central Google Scholar
Hibi, M. & Shimizu, T. Development of the cerebellum and cerebellar neural circuits. Dev. Neurobiol. 72, 282–301 (2012).
Article PubMed Google Scholar
Machold, R., Klein, C. & Fishell, G. Genes expressed in Atoh1 neuronal lineages arising from the r1/isthmus rhombic lip. Gene Expr. Patterns11, 349–359 (2011).
Article CAS PubMed PubMed Central Google Scholar
Kidwell, C. U., Su, C. Y., Hibi, M. & Moens, C. B. Multiple zebrafish atoh1 genes specify a diversity of neuronal types in the zebrafish cerebellum. Dev. Biol. 438, 44–56 (2018).
Article CAS PubMed PubMed Central Google Scholar
Hoshino, M. et al. Ptf1a, a bHLH transcriptional gene, defines GABAergic neuronal fates in cerebellum. Neuron 47, 201–213 (2005).
Article CAS PubMed Google Scholar
Seto, Y. et al. Temporal identity transition from Purkinje cell progenitors to GABAergic interneuron progenitors in the cerebellum. Nat. Commun. 5, 3337 (2014).
Article PubMed Google Scholar
Clark, D. W. et al. NUPR1 interacts with p53, transcriptionally regulates p21 and rescues breast epithelial cells from doxorubicin-induced genotoxic stress. Curr. Cancer Drug Targets 8, 421–430 (2008).
Article CAS PubMed Google Scholar
Zhao, B. et al. TEAD mediates YAP-dependent gene induction and growth control. Genes Dev. 22, 1962–1971 (2008).
Article CAS PubMed PubMed Central Google Scholar
Rouault, J. P. et al. Identification of BTG2, an antiproliferative p53-dependent component of the DNA damage cellular response pathway. Nat. Genet. 14, 482–486 (1996).
Article MathSciNet CAS PubMed Google Scholar
Osman, F. & Whitby, M. C. Emerging roles for centromere-associated proteins in DNA repair and genetic recombination. Biochem. Soc. Trans. 41, 1726–1730 (2013).
Article CAS PubMed Google Scholar
Sugasawa, K. et al. Xeroderma pigmentosum group C protein complex is the initiator of global genome nucleotide excision repair. Mol. Cell 2, 223–232 (1998).
Article CAS PubMed Google Scholar
Joo, J. et al. Nucleotide excision repair gene ERCC2 and ERCC5 variants increase risk of uterine cervical cancer. Cancer Res. Treat. 48, 708–714 (2016).
Article CAS PubMed Google Scholar
Fan, T. et al. The chromatin remodeler ERCC6 and the histone chaperone NAP1 are involved in apurinic/apyrimidinic endonuclease-mediated DNA repair. Plant Cell 36, 2238–2252 (2024).
Article PubMed PubMed Central Google Scholar
Dueva, R. & Iliakis, G. Replication protein A: a multifunctional protein with roles in DNA replication, repair and beyond. NAR Cancer 2, zcaa022 (2020).
Article PubMed PubMed Central Google Scholar
Fritsch, M. et al. Caspase-8 is the molecular switch for apoptosis, necroptosis and pyroptosis. Nature 575, 683–687 (2019).
Article CAS PubMed Google Scholar
Levine, A. J. & Oren, M. The first 30 years of p53: growing ever more complex. Nat. Rev. Cancer 9, 749–758 (2009).
Article CAS PubMed PubMed Central Google Scholar
Pawlowski, J. & Kraft, A. S. Bax-induced apoptotic cell death. Proc. Natl Acad. Sci. USA 97, 529–531 (2000).
Article CAS PubMed PubMed Central Google Scholar
Dilla, T., Velasco, J. A., Medina, D. L., Gonzalez-Palacios, J. F. & Santisteban, P. The MDM2 oncoprotein promotes apoptosis in p53-deficient human medullary thyroid carcinoma cells. Endocrinology 141, 420–429 (2000).
Article CAS PubMed Google Scholar
Sharma, A., Singh, K. & Almasan, A. Histone H2AX phosphorylation: a marker for DNA damage. Methods Mol. Biol. 920, 613–626 (2012).
Article CAS PubMed Google Scholar
Prabhu, K. S. et al. H2AX: A key player in DNA damage response and a promising target for cancer therapy. Biomed. Pharmacother. 175, 116663 (2024).
Article CAS PubMed Google Scholar
Solier, S. et al. Heat shock protein 90alpha (HSP90alpha), a substrate and chaperone of DNA-PK necessary for the apoptotic response. Proc. Natl Acad. Sci. USA 109, 12866–12872 (2012).
Article CAS PubMed PubMed Central Google Scholar
Takashima, H. et al. Mutation of TDP1, encoding a topoisomerase I-dependent DNA damage repair enzyme, in spinocerebellar ataxia with axonal neuropathy. Nat. Genet. 32, 267–272 (2002).
Article CAS PubMed Google Scholar
O’Connor, E. et al. Mutations in XRCC1 cause cerebellar ataxia and peripheral neuropathy. J. Neurol. Neurosurg. Psychiatry 89, 1230–1232 (2018).
Article PubMed Google Scholar
Hoch, N. C. et al. XRCC1 mutation is associated with PARP1 hyperactivation and cerebellar ataxia. Nature 541, 87–91 (2017).
Article CAS PubMed Google Scholar
Rudenskaya, G. E. et al. Ataxia with oculomotor apraxia type 4 with PNKP common “Portuguese” and novel mutations in two Belarusian families. J. Pediatr. Genet 8, 58–62 (2019).
Article CAS PubMed PubMed Central Google Scholar
Paucar, M. et al. Expanding the ataxia with oculomotor apraxia type 4 phenotype. Neurol. Genet 2, e49 (2016).
Article PubMed PubMed Central Google Scholar
Magrino, J. et al. A thermosensitive PCNA allele underlies an ataxia-telangiectasia-like disorder. J. Biol. Chem. 299, 104656 (2023).
Article CAS PubMed PubMed Central Google Scholar
Mahale, R. R. et al. A rare case of ataxia-telangiectasia-like disorder with MRE11 mutation. J. Pediatr. Neurosci. 15, 283–285 (2020).
Article PubMed PubMed Central Google Scholar
Sedghi, M. et al. Ataxia-telangiectasia-like disorder in a family deficient for MRE11A, caused by a MRE11 variant. Neurol. Genet. 4, e295 (2018).
Article CAS PubMed PubMed Central Google Scholar
Sandoval, N. et al. Characterization of ATM gene mutations in 66 ataxia telangiectasia families. Hum. Mol. Genet. 8, 69–79 (1999).
Article CAS PubMed Google Scholar
Shin, U. & Lee, Y. Unraveling DNA repair processes in vivo: insights from zebrafish studies. Int. J. Mol. Sci. https://doi.org/10.3390/ijms241713120 (2023).
Wang, Y. et al. Structural investigation of pathogenic RFC1 AAGGG pentanucleotide repeats reveals a role of G-quadruplex in dysregulated gene expression in CANVAS. Nucleic Acids Res. 52, 2698–2710 (2024).
Ohayon, D. et al. Cytoplasmic proliferating cell nuclear antigen connects glycolysis and cell survival in acute myeloid leukemia. Sci. Rep. 6, 35561 (2016).
Article CAS PubMed PubMed Central Google Scholar
Tang, D. et al. Cytoplasmic PCNA is located in the actin belt and involved in osteoclast differentiation. Aging 12, 13297–13317 (2020).
Article CAS PubMed PubMed Central Google Scholar
Thisse, C. & Thisse, B. High-resolution in situ hybridization to whole-mount zebrafish embryos. Nat. Protoc. 3, 59–69 (2008).
Article CAS PubMed Google Scholar
Zheng, G. X. et al. Massively parallel digital transcriptional profiling of single cells. Nat. Commun. 8, 14049 (2017).
Article CAS PubMed PubMed Central Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Martin, F. J. et al. Ensembl 2023. Nucleic Acids Res. 51, D933–D941 (2023).
Article CAS PubMed Google Scholar
Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat. Biotechnol. 42, 293–304 (2024).
Article CAS PubMed Google Scholar
Jiang, M. et al. Characterization of the zebrafish cell landscape at single-cell resolution. Front. Cell Dev. Biol. 9, 743421 (2021).
Article PubMed PubMed Central Google Scholar
Hafemeister, C. & Satija, R. Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression. Genome Biol. 20, 296 (2019).
Article CAS PubMed PubMed Central Google Scholar
Chung, N. C. & Storey, J. D. Statistical significance of variables driving systematic variation in high-dimensional data. Bioinformatics 31, 545–554 (2015).
Article CAS PubMed Google Scholar
Lange, M. et al. A multimodal zebrafish developmental atlas reveals the state-transition dynamics of late-vertebrate pluripotent axial progenitors. Cell 187, 6742–6759.e17 (2024).
Lin, X., Chau, C., Ma, K., Huang, Y. & Ho, J. W. K. DCATS: differential composition analysis for flexible single-cell experimental designs. Genome Biol. 24, 151 (2023).
Article CAS PubMed PubMed Central Google Scholar
Bunis, D. G., Andrews, J., Fragiadakis, G. K., Burt, T. D. & Sirota, M. dittoSeq: universal user-friendly single-cell and bulk RNA sequencing visualization toolkit. Bioinformatics 36, 5535–5536 (2021).
Article PubMed Google Scholar

Download references

Acknowledgements

We thank Pr Masahiko Hibi (Tokyo, Japan) for providing antibodies against pvalb7, neurod1, and vglut1 and associated protocols. We thank Dr Gary Armstrong (McGill University) for the use of the olig2:GFP transgenic line. We thank the CRCHUM animal facility, particularly the zebrafish platform, for their help and support in maintaining our fish colonies. We also thank the CRCHUM Cell Imaging Core Facility for access to confocal microscopy (Canadian Innovation Foundation) and particularly Aurélie Cleret-Buhot for her help in image acquisition and analysis. Finally, we thank the IRCM genomic platform for its rigorous work preparing single-cell RNA libraries. This work was funded by the Centre d’excellence en recherche sur les maladies orphelines—Fondation Courtois (CERMO-FC), by Génome Quebec and Ataxie Canada. E.S. and S.A.P. are supported by the Natural Science and Engineering Research Council (NSERC), Canadian Institutes of Health Research (CIHR), and are FRQS research scholars. M.T. is a Junior 2 FRQS research scholar. F.N. received M.Sc scholarships from CRCHUM and the Faculty of Medicine/Université de Montreal. S.A. received doctoral scholarships from CRCHUM, Faculty of Medicine/Université de Montreal, CERMO-FC, FRQS, and CIHR.

Author information

Authors and Affiliations

Department of Neuroscience, Faculty of Medicine, Université de Montréal, Montreal, QC, Canada
Fanny Nobilleau, Sébastien Audet, Alexandra da Silva Babinet, Meijiang Liao, Martine Tétreault, Shunmoogum A. Patten & Éric Samarut
University of Montréal Hospital Research Center (CRCHUM), Montreal, QC, Canada
Fanny Nobilleau, Sébastien Audet, Alexandra da Silva Babinet, Meijiang Liao, Martine Tétreault & Éric Samarut
Molecular Genetics of Development Laboratory, Département des Sciences Biologiques, Université du Québec à Montréal (UQAM), Montréal, QC, Canada
Sanaa Tork & Nicolas Pilon
Centre d’excellence en recherche sur les maladies orphelines – Fondation Courtois (CERMO-FC), Université du Québec à Montréal, Montreal, QC, Canada
Sanaa Tork, Nicolas Pilon, Martine Tétreault, Shunmoogum A. Patten & Éric Samarut
INRS- Centre Armand-Frappier Santé Biotechnologie, Laval, QC, Canada
Charlotte Zaouter & Shunmoogum A. Patten
Department of Pediatrics, Faculty of Medicine, Université de Montréal, Montreal, QC, Canada
Nicolas Pilon

Authors

Fanny Nobilleau
View author publications
Search author on:PubMed Google Scholar
Sébastien Audet
View author publications
Search author on:PubMed Google Scholar
Alexandra da Silva Babinet
View author publications
Search author on:PubMed Google Scholar
Sanaa Tork
View author publications
Search author on:PubMed Google Scholar
Charlotte Zaouter
View author publications
Search author on:PubMed Google Scholar
Meijiang Liao
View author publications
Search author on:PubMed Google Scholar
Nicolas Pilon
View author publications
Search author on:PubMed Google Scholar
Martine Tétreault
View author publications
Search author on:PubMed Google Scholar
Shunmoogum A. Patten
View author publications
Search author on:PubMed Google Scholar
Éric Samarut
View author publications
Search author on:PubMed Google Scholar

Contributions

Conceptualization, E.S., S.A.P., and F.N.; Methodology: E.S., S.A.P., F.N., A.DSB and M.L; Investigation: F.N., S.A., A.d.S.B., S.T., and C.Z.; Formal analysis and data curation: S.A.; Resources: N.P., M.T., and E.S.; Writing original draft: E.S. and F.N.; Writing—review & editing: S.A.P., N.P., and M.T.; Visualization: N.P., E.S., and S.A.; Supervision: E.S., S.A.P., N.P., and M.T.; Project administration and funding acquisition: E.S.

Corresponding author

Correspondence to Éric Samarut.

Ethics declarations

Competing interests

E.S. is a co-founder of DanioDesign Inc. (QC, Canada) and of Osta Therapeutics (France); N.P. is a co-founder of Neurenati Therapeutics (QC, Canada). These commercial affiliations did not play any role in investigational design, data collection and analysis, the decision to publish or the preparation of the manuscript. The remaining authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review file

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Nobilleau, F., Audet, S., da Silva Babinet, A. et al. RFC1 regulates the expansion of neural progenitors in the developing zebrafish cerebellum. Nat Commun 16, 6019 (2025). https://doi.org/10.1038/s41467-025-60775-5

Download citation

Received: 30 July 2024
Accepted: 04 June 2025
Published: 01 July 2025
DOI: https://doi.org/10.1038/s41467-025-60775-5