https://orcid.org/0009-0000-2774-071X
This is an uncorrected proof.
Figures
Abstract
Equine Juvenile Spinocerebellar Ataxia (EJSCA) is a novel autosomal recessive neurologic disease in Quarter Horses. Affected foals display a progressive proprioceptive ataxia by 1–5 weeks of age, leading to recumbency and necessitating euthanasia. Whole genome sequencing was performed on 7 EJSCA cases and unaffected horses that included 4 obligate carriers, 4 unaffected half or full-siblings, and 28 unrelated, unaffected control Quarter Horses. An 82 kb region of association was identified (EquCab3.0, chr11: 6963986–7045999), containing 9 candidate SNPs across four genes (FADS6, FDXR, GRIN2C and TMEM104). Decreased FDXR mRNA expression and a cryptic exon was identified in spinal cord tissue from EJSCA cases via RNA-sequencing. One of the 9 associated SNPs (FDXR-203 c.177 + 1778G > C) was the eighth base pair of this cryptic exon. Affected foals were all homozygous for the variant. Protein concentrations of FDXR were lower in EJSCA cases in spinal cord and liver compared to unaffected controls. The FDXR-203 c.177 + 1778G > C mutation represents the first non-coding neurological genetic variant in horses. Additionally, this is the first genetic cause of a degenerative axonopathy in the horse and a spontaneous disease model to study FDXR pathology in humans.
Author summary
The American Quarter Horse is the most popular horse breed in North America. A rapidly progressive spinocerebellar ataxia, termed equine juvenile spinocerebellar ataxia (EJSCA), was first identified in Quarter Horse foals in 2020. EJSCA was shown to be inherited as an autosomal recessive trait. Whole-genome sequencing on 7 ESJCA cases and RNA-sequencing of spinal cord identified an intronic variant in Ferredoxin reductase (FDXR) creating a cryptic third exon in affected foals. Four additional affected foals were subsequently genotyped and were all homozygous for the intronic variant. Decreased transcript and protein expression of FDXR was observed in spinal cord and liver of affected animals. The FDXR-203 c.177 + 1778G > C mutation represents the first disease-causing genetic variant for an inherited degenerative axonopathy in the horse and a spontaneous disease model to study FDXR pathology in humans.
Citation: Brown BN, Dahlgren AR, Ghosh S, Durbin-Johnson B, Willis A, Olivas C, et al. (2026) An intronic variant in Ferredoxin Reductase (FDXR) creates a cryptic exon in Quarter Horses with Equine Juvenile Spinocerebellar Ataxia. PLoS Genet 22(5): e1012158. https://doi.org/10.1371/journal.pgen.1012158
Editor: Giovanni Stevanin, INCIA: Institut de Neurosciences Cognitives et Integratives d’Aquitaine, FRANCE
Received: December 12, 2025; Accepted: May 7, 2026; Published: May 20, 2026
Copyright: © 2026 Brown et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The whole-genome sequence data generated in this study is available in the NCBI BioProject database (https://www.ncbi.nlm.nih.gov/bioproject/) under accession number PRJNA1395672. RNA-sequencing of spinal cord tissue from foals is available in the NCBI BioProject database (https://www.ncbi.nlm.nih.gov/bioproject/) under accession number PRJNA1395676.
Funding: Support for this study was provided by UC Davis Center for Equine Health (CEH) to C.J.F (CEH https://ceh.vetmed.ucdavis.edu/) with funds provided by the State of California pari-mutuel fund and contributions by private donors and funds from UC Davis Veterinary Genetics Laboratory. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: I have read the journal’s policy, and the authors of this manuscript have the following competing interests: A portion of the revenue generated by UC Davis Veterinary Genetics Lab (UCD-VGL) for tests discovered in the Finno Laboratory (MYHM and EJSCA) provides funding for additional research conducted in the Finno Laboratory. RRB and RG are affiliated with the UC Davis Veterinary Genetics Laboratory offering commercial diagnostics testing for horses and other species.
Introduction
In 2020, the first identified case of equine juvenile spinocerebellar ataxia (EJSCA) in the American Quarter Horse (QH) was recognized [1]. Over the next 4 years, an additional eleven cases were identified. All foals were presented between 1–5 weeks of age with an acute history of neurological deficits, with the pelvic limbs more severely affected. The spinal ataxia rapidly progressed to include paraparesis, resulting in recumbency within 0–18 days and necessitating euthanasia [1]. Postmortem lesions were confined to histologic evidence of dilated myelin sheaths and digestion chambers throughout the dorsal spinocerebellar tract of entire spinal cord, but most severe in the cervicothoracic region. Some foals had additional lesions within the fasciculus cuneatus and ventromedial tracts [1]. The pedigree analysis of cases supported an autosomal recessive mode of inheritance, with common ancestors tracing back within 4–6 generations on both sides of the pedigree [1].
Inherited defects that result in a spinocerebellar ataxia have been reported in dogs [2–9], cattle [10–15] and sheep [16–18]. In horses, a distinct spinocerebellar ataxia, termed equine neuroaxonal dystrophy/ degenerative myelopathy, has been described but horses develop less severe clinical signs later in life (typically 1–3 years of age) [19]. Despite several intensive genetic investigations, there is currently no genetic mutation identified for eNAD/EDM. Further, EJSCA is a clinically and histologically distinct disorder [1].
Based on the apparent familial basis for EJSCA, we aimed to identify putative genetic variants associated with the phenotype. We performed whole-genome sequencing in seven affected QH foals with EJSCA and explored differences in gene expression within our candidate region in spinal cord tissue. Segregating coding variants were excluded, and we identified an intronic variant, resulting in a cryptic exon in Ferredoxin reductase (FDXR c.177 + 1778G > C), that significantly associated with the EJSCA phenotype. PromethION Nanopore sequencing was used to clearly define the boundaries of the cryptic exon and to determine that inclusion of the in-frame cryptic exon led to a premature stop codon. Decreased FDXR transcript and protein expression supported loss-of-function in the spinal cord and liver of affected foals, which aligned with clinical findings.
Results
Animals
Clinicopathologic findings of foals affected with EJSCA were previously described [1]. Briefly, the study included n = 11 confirmed EJSCA foals (n = 9 fillies and n = 2 colts, which were included in the previous report [1]) that appeared healthy at birth but developed an acute onset of neurologic deficits within the first months of life (S1 Table). Deficits consisted of a spinocerebellar ataxia of all four limbs, with mild asymmetric paraparesis in some cases. Clinicopathologic findings included elevated gamma-glutamyl transferase (GGT) and hyperglycemia on biochemical profiles, with normal muscle enzymes. All foals rapidly progressed to recumbency and were euthanized. Postmortem evaluation identified dilated myelin sheaths with digestion chambers throughout the entire spinal cord but most severe in the dorsal spinocerebellar tracts of the cervicothoracic region [1]. Control horses were considered healthy if they had lived past six months of age with no evidence of ataxia.
Whole-Genome Genome-Wide Association Study (WGS-GWAS)
A whole-genome association genome-wide association (WGS-GWAS) study was performed to investigate the genetic cause of EJSCA. Significantly associated variants were identified, and their effects predicted using the whole-genome sequence of n = 7 EJSCA foals, n = 3 carrier dams, n = 1 carrier sire, n = 3 unaffected half-siblings, n = 1 unaffected full-sibling and n = 28 unrelated control Quarter Horses (S1 Table). Seventy-seven variants were associated with EJSCA at p < 1 x 10-11, and all were homozygous in EJSCA foals and heterozygous in the carriers. The 77 associated variants were in an 82 kb region on chr11 from 6963986-7045999 bp (S2 Table). The variants within this region were further filtered by effect (‘MODERATE’ and ‘HIGH’) and screened in public databases (https://ncbi.nlm.nih.gov/subs/sra/ and [20]). No ‘HIGH’ effect nonsense variants were identified. All ‘MODERATE’ effect variants were found in other breeds with no reported cases of this neurological condition (S2 Table). Of the remaining single nucleotide variants (SNVs), only 9 were not identified in any other breed (Table 1). These variants spanned four genes: Fatty acid desaturase 6 (FADS6, 1 synonymous), Ferredoxin reductase (FDXR, 4 intronic SNVs), Glutamate ionotropic receptor NMDA type subunit 2C (GRIN2C, 1 intronic SNV) and Transmembrane protein 104 (TMEM104 (3 intronic SNVs) (Fig 1). All genes are expressed in human spinal cord RNA (https://www.proteinatlas.org/) [21]. FADS6, FDXR and GRIN2C are expressed at the RNA level in equine spinal cord and cerebellum, but TMEM104 is not [22]. Scores were determined for each orthologous human variant using the 100-vertebrate score by phastCons (https://genome.ucsc.edu/). All conservation scores were 0, except for chr11:6973334, which was 0.05.
Whole genome association testing identified an 82 kb candidate region of interest (p = 2.9 x 10-12) that was homozygous in all affected foals. After filtering by effect (‘MODERATE’ and ‘HIGH’) and screening in public databases (https://ncbi.nlm.nih.gov/subs/sra/ and [20]), nine homozygous variants across the genes Fatty acid desaturase 6 (FADS6, 1 synonymous SNV), Ferredoxin reductase (FDXR, 4 intronic SNVs), Glutamate ionotropic receptor NMDA type subunit 2C (GRIN2C, 1 intronic SNV) and Transmembrane protein 104 (TMEM104 (3 intronic SNVs) were only present in EJSCA foals. Top tracks represent equine FAANG data tracks.
mRNA-sequencing
To quantify gene expression across the four genes of interest in the candidate region (FADS6, FDXR, GRIN2C and TMEM104), mRNA-sequencing on spinal cord tissue was performed on n = 5 EJSCA cases and n = 6 unaffected age-matched controls (S1 Table). There were no genes that were significantly different at PFDR < 0.05. However, of the four positional candidate genes (FADS6, FDXR, GRIN2C and TMEM104), only FDXR was significantly decreased in EJSCA foals as compared to healthy foals at Punadjusted<0.01 (p = 0.0008) (Fig 2A). Mapped RNA-seq reads were then visualized for FDXR, and a cryptic exon was identified in intron 2, which contained the intronic variant identified at chr11:6973334 (G/C) (Fig 2B).
(A) Bulk mRNA sequencing of spinal cord tissue performed on n = 5 EJSCA cases and n = 6 unaffected age-matched controls identified significantly decreased FDXR expression in EJSCA foals at Punaadjusted = 0.0008). (B) Mapped mRNA reads over FDXR identified a cryptic exon (indicated in red box) in intron 2 in EJSCA foals, corresponding to one of the FDXR deep intronic variants identified in the WGS-WGAS (chr11:6973334, G/C). (C) Representative chromatograms from Sanger sequencing of n = 11 EJSCA foals, n = 12 dams, n = 4 sires, n = 11 siblings, and n = 8 unrelated healthy control horses confirmed perfect segregation of the c.177 + 1778G > C variant. (D) RT-PCR of FDXR exons 2-5 identified three faint amplicons in EJSCA foals (n = 3) and two amplicons (one bold, one faint) in unrelated control horses (n = 3). GAPDH is the positive control. The ladder is on the left of the gel, indicating band sizes from 100 bp to 1500 bp.
qRT-PCR Validation of FDXR expression
To validate the decreased expression of FDXR observed in spinal cord from EJSCA foals, qRT-PCR was performed in n = 3 EJSCA foals and n = 3 age-matched unrelated healthy control foals. FDXR expression was decreased in EJSCA affected foals (fold-change = 0.44, p = 0.02).
Variant Validation via Sanger Sequencing
DNA from n = 11 EJSCA foals, n = 12 dams, n = 4 sires, n = 11 siblings, and n = 8 unrelated healthy control horses underwent Sanger sequencing to confirm the putative FDXR variant detected from whole genome sequencing data (c.177 + 1778G > C). All EJSCA foals were homozygous for the variant while all dams and sires were heterozygous. Unrelated control samples were homozygous wildtype (Fig 2C). An additional 1,060 available presumed unaffected Quarter Horses were randomly selected from the biorepository at the UC Davis Veterinary Genetics Laboratory and genotyped for this variant. From the 1,060 horses, 25 were heterozygous and the remainder of the horses were homozygous wildtype. The allele frequency for the FDXR variant was calculated from this sample set for the Quarter Horse population (q = 0.012). An additional 181 horses across 26 breeds from the University of California Veterinary Genetics Laboratory were screened and no other carriers were identified in other breeds (S3 Table).
FDXR c.177 + 1778G > C functional validation
Since the putative variant for EJSCA was deep within an intron, 1778 bp away from exon 2, and the RNA-sequencing data did not clearly define the boundaries of the cryptic exon, we pursued reverse-transcriptase PCR (RT-PCR) in spinal cord, liver and gluteal muscle from EJSCA affected foals and age-matched controls. Primers flanking exons 2 and 3 of FDXR-201, as annotated by Ensembl EquCab3.0, (https://useast.ensembl.org/Equus_caballus/Info/Index), amplified two PCR products, one prominent band and one band with less intensity in control horses (~350 bp and ~480 bp, respectively, Fig 2D). Three products were observed in EJSCA foals (~350, 480 and 650 bp, Fig 2D).
Long-read PromethION Nanopore sequencing of the resulting RT-PCR products from spinal cord tissue was performed in n = 2 EJSCA foals and n = 4 controls to define these fragments. A consensus sequence was identified, and two clusters (Cluster 24 and 26, both 306 bp on average) demonstrated allele sharing across all six samples (EJSCA and controls). The two other clusters (Cluster 0, 398 bp and Cluster 144, 541 bp) were quantified at 30-2880x only in EJSCA foals (Fig 3A and S4 Table). Clusters 0 and 144 each contained sequences that aligned over the cryptic exon that had been identified with the RNA-sequencing data. Cluster 144 also contained a longer 5’ end of canonical exon 3. Within both clusters 0 and 144, the G > C substitution was evident at 8 bp (Cluster 0) or 29 bp (Cluster 144) from the start of the reads in EJSCA foals only. Based on the PromethION Nanopore sequencing, the boundaries of the cryptic exon were clearly defined (Fig 3B) and the resulting amino acid sequence determined (Fig 3C). Inclusion of the in-frame cryptic exon led to a premature stop codon after 24 amino acids (Fig 3C).
(A) Intronic FDXR reads were identified in intron 2 only in EJSCA samples, consistent with cryptic exon localization. (B) The PromethION Nanopore sequencing allowed for accurate detection of the cryptic exon boundaries. (C) When the cryptic exon is transcribed, the additional amino acids are in-frame but result in a premature stop codon (“TGA”) after 24 amino acids.
Western blotting of spinal cord demonstrated decreased FDXR protein expression in EJSCA foals compared to unaffected control horses (p = 0.04, Fig 4A and 4B). Additionally, decreased expression was observed in myocardium and gluteal muscle from EJSCA foals; however, the overall expression in these tissues across all horses was lower (Fig 4C). Biological replicates of western blotting from spinal cord and original gel images are included in S1A, S1B, S1C, S1D, and S1E Fig.
(A) FDXR protein expression was decreased in spinal cord of FDXR foals (n = 3 biologic replicates) as compared to age-matched unrelated control foals. (B) Quantification of decreased spinal cord expression of FDXR. (C) FDXR protein expression was decreased in all EJSCA foal tissues but is particularly evident in spinal cord and liver. Overall expression was decreased in myocardium and gluteal muscle. Beta-actin was the loading control. Biological replicates and original gel images are included in S1 Fig. SC = spinal cord, MyoC = myocardium, GM = Gluteal muscle.
Discussion
There are currently six well-characterized disease variants in the American Quarter horse that are recommended for testing as part of the American Quarter Horse Association Health Panel. These include glycogen branching enzyme deficiency (GBED), hereditary equine regional dermal asthenia (HERDA), hyperkalemic periodic paralysis (HYPP), malignant hyperthermia (MH), polysaccharide storage myopathy Type 1 (PSSM1) and myosin-heavy chain myopathy (MYHM). Testing for this panel is required by the breed association for stallions before resulting foals are registered. The variants for these six diseases are all coding in nature, including missense (HERDA, HYPP, MH, PSSM1, MYHM) and nonsense (GBED) mutations. The FDXR c.177 + 1778G > C mutation identified in this study represents the first intronic variant that results in a severe fatal neurologic disease in Quarter Horses. Additionally, this FDXR variant represents the first identified cause for an inherited degenerative axonopathy in the horse and a spontaneous disease model to study FDXR pathology in humans.
While horses are not a common model for studying human diseases, the discovery of a missense variant in Myosin heavy chain 1 (MYH1) [24] in horses led to the prioritization of MYH1 as a candidate gene for recurrent rhabdomyolysis in humans [25]. Additionally, horses suffering from malignant hyperthermia have a mutation in the Ryanodine receptor 1 (RYR1) [26], like humans [27]. A pathogenic variant in the ER-associated degradation pathway protein SEL1L was identified in a rare platelet disease in horses [28], which spurred a cross-species approach that has established SEL1L variants as candidates for human bleeding disorders [29]. Thus, leveraging genetic findings in the horse can lead to new discoveries with strong translational relevance.
In humans, FDXR mutations are associated with mitochondriopathies defined by decreased electron transport chain function, increased reactive oxygen species and iron overload [30]. The first link between FDXR and mitochondrial disease in humans was published in 2017, where recessive mutations in FDXR resulted in optic atrophy, ataxia and hypotonia in a cohort of individuals [31]. Functional studies in cells from these patients showed reduced FDXR activity and mitochondrial overload. That same year, eight children with auditory neuropathy and bilateral defects of the optic nerves were identified with recessive FDXR mutations [32]. Since then, the disease has been renamed FDXR-related mitochondriopathy (FRM) and symptoms were expanded to include movement disorders and global developmental delay [33]. All the pathogenic variants described to date in humans are missense, in frame deletions, or nonsense variants (https://www.ncbi.nlm.nih.gov/clinvar), with no splice site variants identified to date. Foals with EJSCA did not appear to have visual or auditory deficits, but electroretinography and brainstem auditory-evoked potential testing was not performed. Optic nerves are not routinely collected during veterinary necropsies, but should be evaluated in any future cases of this disorder
Deep intronic mutations can disrupt normal splicing patterns by creating new splice sites, altering splicing enhancers or silencers or interfering with canonical splice sites. Deleterious DNA variants that are more than 100 bp away from exon-intron junctions commonly lead to pseudo-exon inclusion [34]. With EJSCA, the deep intronic FDXR c.177 + 1778G > C mutation creates a cryptic splice site, which likely activates nonsense-mediated decay, as evidenced by reduced mRNA expression and the fainter RT-PCR bands, as well as reduced protein observed. The advantage of performing whole-genome sequencing, versus whole exome sequencing, was clearly highlighted in this study, as an exome approach would have failed to detect this variant. Pairing our whole-genome results with reads obtained from RNA-sequencing was critical in identifying a putative functional intronic variant in FDXR.
While EquCab3.0 has undoubtedly improved the reference assembly for the horse, annotation of specific isoforms still relies primarily on in silico predictions through Ensembl. There are six transcripts of FDXR annotated in Ensembl (FDXR-201,202,203,204,205 and 206), with FDXR-201 labeled as the “canonical” isoform. Using the NCBI annotation, which relies more on RNA-seq data to identify coding regions, there are five transcripts (XM_023652066.1 or variant X4, XM_023652067.1 or variant X5, XM_023652064.1 or variant X1, XM_023652065.1 or variant X2 and XM_001492615.6 or variant X2). Using the Ensembl annotation, the main variation between isoforms occurs in the 5’ UTR region and 3’ to exon 3, with alternative splicing annotated between exons 3 and 4, 4 and 5 and after exon 10. In NCBI, the variation is 3’ to exon 10, with alternative splicing denoted between exons 11, 12 and 13. Notably, none of these isoforms have coding sequence predicted between canonical exons 2 and 3, and none of these isoforms support a longer 5’ end to exon 3 that was noted in our EJSCA foals (as shown by PromethION long-read sequencing results). Within the human genome, there are eight isoforms and none include an additional exon between canonical exons 2 and 3. Thus, this disease in the horse is associated with a novel splice variant of FDXR.
Within NCBI, two proteins are predicted for FDXR in the horse: XP_070082099.1 (isoform 1, 1918 nucleotides) and XP_001492665.3 (isoform 2, 1912 nucleotides), which only differ by 6 nucleotides.
The protein sequence for FDXR is well-conserved between horses and humans (89% identity, 100% query cover, E value = 0.0 via BLAST search (https://blast.ncbi.nlm.nih.gov). Loss-of-function variants in FDXR are associated with mitochondrial iron overload, and this accumulation results in reactive oxygen species accrual and ferroptosis [30]. Recent evidence suggests that Nuclear factor erythroid 2-related factor 2 (NRF2), a transcription factor that is activated during oxidative stress, is downregulated in FDXR-related mitochondrial disease in humans [35].
Since we were unable to amplify cDNA from the region between canonical exons 2 and 3 of FDXR using standard nested PCR and Sanger sequencing techniques, we elected to leverage cDNA Nanopore long-read sequencing technology to sequence each fragment. Major isoforms of genes have been readily identified in human cell lines using Nanopore long-read sequencing protocols [36]. Gene expression estimates from Nanopore long-read RNA-seq data showed the lowest estimation error overall when compared to various RNA-seq protocols [36]. In our study, this long-read sequencing technology provided detailed insight into the alternative splicing that resulted from the FDXR-203 c.177 + 1778G > C variant. This study represents the first use of Nanopore long-read sequencing technologies in the horse to identify novel splice events resulting from intronic variants.
Clinically, foals with EJSCA are presented with a spinocerebellar ataxia and clinicopathologic evidence of liver injury, as evidenced by high GGT concentrations. Muscle enzymes (creatine kinase [CK] and aspartate aminotransferase [AST]) are within normal limits on these foals. This clinical presentation is reflected in the tissue-specific expression that we identified for FDXR. Expression was highest, at both the RNA and protein level, in spinal cord and liver and lower in gluteal muscle and myocardium. In the human protein atlas (https://www.proteinatlas.org), FDXR expression is noted to be highest in adrenal gland (RNA and protein levels), testis and epididymis (protein only), and fallopian tube and ovary (protein only). Within the nervous system, expression is quantifiable, but much lower, in human cerebellum and spinal cord at the RNA level only. We were able to leverage the functional atlas of animal genomes (FAANG) dataset that was created in the horse to evaluate equine-specific tissue patterns of gene expression and regulation (https://www.ebi.ac.uk/ena/data/view/ERA1487553) [22]. FDXR was expressed at the RNA level in equine spinal cord and cerebellum, with no functional elements (promoters, enhancers, silencers) predicted at the site of the EJSCA variant (FDXR-203 c.177 + 1778G > C). Since FDXR expression was identified in equine spinal cord and cerebellum, the FAANG dataset provided species-specific gene expression information that was advantageous to this investigation.
It is worthwhile noting that, of the 9 putative variants (Table 1), the FDXR-203 c.177 + 1778G > C variant was the only one that did not have a conservation score of 0. While variants that disrupt canonical splice site elements often adhere to a highly conserved consensus sequence, de novo mutations that create cryptic exons do not [37]. Thus, while using conservation scores to prioritize putative variants remains an important tool, variants should not be excluded as candidates simply due to low conservation scores.
Limitations of this study include the small sample sizing of EJSCA affected foals, limited frozen spinal cord tissue for extensive qRT-PCR validation of all cases, inclusion of only female cases in RNA-sequencing and qualitative analysis of FDXR protein in the liver. With the family-based cohort, a WGS-GWAS was not the ideal model for performing association testing and either family-based association testing or homozygosity mapping could have been pursued, with the likely autosomal recessive mode of inheritance. We did not have a large multi-generational family to track disease segregation and, with the high background inbreeding in this family of horses, homozygosity mapping could have resulted in spurious associations.
In summary, this study identified a putative variant associated with EJSCA in the American Quarter Horse (FDXR-203 c.177 + 1778G > C). Supporting evidence for causality include: (1) this variant is the 8th base pair of a cryptic exon that occurs only in EJSCA affected foals, (2) there is an overall decrease in FDXR expression in spinal cord of EJSCA foals, (3) perfect genotyping concordance was observed between the FDXR-203 c.177 + 1778G > C variant and the EJSCA phenotype and (4) comparative knowledge of a similar phenotype in humans with genetic mutations in FDXR. As such, to avoid producing affected offspring, utilization of genetic testing results for the FDXR-203 c.177 + 1778G > C variant is recommended for any Quarter Horse breeding program. The overall low allele frequency across the QH breed and evidence of all affected foals tracing back to one prominent sire within 6 generations [1] suggests that this is a relatively new genetic variant. Notably, identification of this deep intronic variant was only possible with whole-genome sequencing. This FDXR variant represents the first to explain an inherited degenerative axonopathy in the horse and a spontaneous disease model to study FDXR pathology in humans.
Materials and methods
Ethics statement
All procedures were approved by the University of California-Davis Institutional Animal Care and Use Committee (#23015) and carried out in accordance with guidelines and regulations. Written consent from the owners was obtained for all sample collections.
Animals
Clinicopathologic findings of affected foals were previously described [1]. The study included n = 11 confirmed EJSCA foals (n = 9 fillies and n = 2 colts) that appeared healthy at birth but developed severe neurologic deficits within the first months of life (S1 Table). A total of n = 7 EJSCA affected foals underwent whole-genome sequencing (S1 Table). For RNA-sequencing, spinal cord tissue from n = 5 affected EJSCA foals was profiled (all fillies, median age 30 days, range 14–30 days, S1 Table). For RT-PCR of spinal cord, liver and gluteal muscle, a subset of n = 3 affected EJSCA foals were used (S1 Table). For Western blot analysis, spinal cord, liver, and gluteal muscle from n = 3 affected foals, and cardiac muscle from n = 1 affected foal were used (S1 Table).
For the genetic investigation, blood samples were provided on all the affected foals’ dams, which were neurologically normal, of which n = 3 were used for whole-genome sequencing, along with n = 1 sire, n = 3 unaffected half-siblings and n = 1 unaffected full-sibling (S1 Table). Siblings had lived past six months of age with no evidence of ataxia. Publicly available whole-genome sequences from unaffected unrelated Quarter Horses (https://www.ebi.ac.uk/ena/data/view, SUB11328561, SUB13501748 and SUB15353402) was leveraged as an additional set of control samples. For RNA-sequencing, spinal cord tissue from n = 5 unaffected age-matched control foals was profiled (2 fillies, 3 colts, median age 180 days, range 21–2190 days, S1 Table). For RT-PCR of spinal cord and gluteal muscle, a subset of n = 3 control foals were used. For RT-PCR of liver, a subset of n = 2 control foals were used. For Western blot analysis, spinal cord and gluteal muscle from n = 3 control foals, liver from n = 2 control foals and myocardium from n = 1 control foal were used (S1 Table).
Blood DNA extraction
Genomic DNA was isolated from whole blood samples according to the WIZARD Blood DNA Extraction Kit protocol (Promega, Madison, WI). Sample DNA concentrations were measured using QIAxpert (QIAGEN, Hilden, Germany).
Whole-genome sequencing
Whole-genome sequencing was performed on 15 horses on the Illumina HiSeq4500 platform using paired-end, 2 x 150 bp chemistry. Average coverage depth was 30x across the Equus Caballus genome. Quality control (FastQC) indicated that % of bases possessed a Phred score > 30. Reads were mapped to the EquCab3.0 equine reference sequence [38] using BWA for Illumina mapping program [39]. Mapping quality was assessed using Samtools Flagstat [40]. SNP, INDEL discovery, and genotyping across all samples was performed for the discovery of a variant using Genome Analysis Toolkit (GATK, v4.2) [41] Variant calling was performed using GATK’s HaplotypeCaller to generate genomic variant call format (gVCF) files for each sample. Joint genotyping was conducted with GenotypeGVCFs to create a multi-sample VCF. The resulting VCF was filtered using SNPSift [23] to filter the resulting.vcf file by quality, using a variant Phred threshold of 30 (Q ≥ 30). Whole genome sequences were deposited in the NCBI Sequence Read Archive (https://ncbi.nlm.nih.gov/subs/sra/) (PRJNA1395672). For publicly available files, fastq files were obtained from the SRA archive (https://www.ebi.ac.uk/ena/data/view, SRR19364590, SRR28735411 and SRR34301010) and trimmed for quality.
A whole-genome genome-wide association study (WGS-GWAS) was performed using n = 7 EJSCA foals, n = 4 parents (coded as unknown) and n = 32 control Quarter Horses. Case/control status was assigned using SNPSift caseControl [42], followed by filtering based on an allelic P value of 1 x 10-11. In addition to assessing significantly associated variants called with FreeBayes, raw bam files were visually inspected using Integrative Genome Viewer [43] in the candidate regions for any structural variants, including duplications, inversions, and large deletions or insertions. SNPEff was then used to predict the effect on protein function of the alternate variants, classifying variants as having ‘HIGH’, ‘MODERATE’, ‘LOW’, or ‘MODIFIER’ putative effects [42]. Chromosomal distribution of the resulting significantly associated variants (P < 1 x 10-11) was assessed. Variants with predicted ‘HIGH’ or ‘MODERATE’ effects were then screened in the publicly available database of equine whole-genome sequences using NCBI’s Sequence Read Archive (SRA) database (https://www.ncbi.nlm.nih.gov/sra) and excluded if found in breeds other than the Quarter Horse. Functional effects were predicted using the Ensembl Variant Effect Predictor (https://www.ensembl.org).
mRNA-sequencing
To quantify gene expression across the four genes of interest in the candidate region (FADS6, FDXR, GRIN2C and TMEM104), mRNA-sequencing on spinal cord tissue was performed on n = 5 EJSCA cases and n = 6 unaffected age-matched controls (S1 Table). RNA was isolated from spinal cord tissue as previously described [22] and RNA integrity scores (RIN) assessed using the BioAnalyzer High Sensitivity RNA Assay. RIN scores below 7.0 were considered to have failed quality control. Libraries were prepared using PolyA selection with the NEBNext Ultra II Directional Prep Kit. The resulting libraries were sequenced on Illumina, with 150 base pair paired end reads at a depth of 40 million reads per sample. Adapter trimming, poly A trimming, quality trimming, length filtering, and removal of PCR duplicates were conducted using HTStream, version 1.3.3 [44]. The resulting reads were aligned to EquCab.3.0 using STAR, version 2.7.10b [45]. Differential expression analyses were conducted using the limma-voom Bioconductor pipeline (limma version, edgeR version, in R version 4.2.2 (2022-10-31) [46]), using a model with effects for disease status, sex, and age. Low expressed genes were filtered prior to analysis using the function filterByExpr in edgeR, leaving 16,449 genes. Raw bam files of FDXR were visually inspected using Integrative Genome Viewer [43]. RNA-sequences were deposited in the NCBI Sequence Read Archive (https://ncbi.nlm.nih.gov/subs/sra/) (PRJNA1395676).
qRT-PCR Validation of FDXR expression
To validate decreased FDXR expression in spinal cord from EJSCA affected foals, qRT-PCR was performed using a subset of animals from the mRNA-seq study (n = 3 EJSCA affected foals and n = 3 age-matched unrelated unaffected control foals, S1 Table). Primer3Plus software [47] and DNA oligonucleotides were synthesized by Integrated DNA Technologies, with primers flanking exons 3 and 4 of FDXR (F 5’CCCATGTGGACATCTTCGAGAA3’, R 5’ GAAGGCACAGCGGTCAGAG3’) and Beta-actin (ACTB) was used as a reference gene (F 5’AAGGAGAAGCTCTGCTATGTCG3’, R 5’GGGCAGCTCGTAGCTCTTC3’). The reverse transcription of the two-step RT-qPCR was performed using Superscript III (Invitrogen) and random hexamers with 1.5 µg total RNA. Quantitative PCR Reactions were performed in a 10 µL reaction volume using the QIAGEN Rotor-Gene SYBRGreen PCR Kit (QIAGEN, Valencia, CA, USA. Each tube contained 1:10 dilution of total converted cDNA, and 1 µm final primer concentration for each forward and reverse primer. PCR was performed on a Rotor-Gene Q 72-well thermocycler (QIAGEN, Valencia, CA, USA) as follows: 5 min at 95 °C; 35 cycles of 5 s at 95 °C and 10 s at 58 °C; melt curve ramping from 60 °C to 95 °C, rising by 1 °C at each step. Each reaction was run in triplicate and each run included a no-template control. Efficiencies for each gene were calculated at 90–110%. Relative quantitation of gene expression was calculated by comparative threshold cycle method (2 − ΔΔCt) using the Ct of the housekeeping gene with the lowest variance. Data were analyzed using a student’s T test, with significance set at p < 0.05.
Sanger sequencing validation
To confirm the putative variants identified on whole-genome sequencing, primers were designed using Primer3Plus software [47] and DNA oligonucleotides synthesized by Integrated DNA Technologies (F 5’TGCATCCTCAAGACGACCTG3’, R 5’TTCCACCCTCATGCATTCCC3’). Amplification of products was performed using end-point PCR. Products were loaded with Invitrogen TrackIt Cyan/Orange Loading Buffer and electrophoresis was performed using a 1.5% agarose gel stained with SYBR Safe DNA Gel Stain (Invitrogen; s33102). Fragments were visualized using the Azur c150 Gel Imaging Workstation. PCR was performed using the Invitrogen Platinum Taq DNA Polymerase kit (10966). Each 25-µl PCR reaction comprised 0.1 ul of Platinum Taq DNA Polymerase, 2.5 µl of 10x PCR Buffer -Mg, 0.75 ul of 5 mM MgCl2, 0.5 ul of 10 mM dNTPs, 0.25 ul of 20 uM forward and reverse primers (Integrated DNA Technologies), 1 ul KB Extender, and 1 µl of 50 ng/ul genomic DNA. Standard PCR conditions were performed as follows: 94 °C for 2 min and 35 cycles of 94 °C denaturation for 30 s, 57 °C annealing for 30 s, and 72 °C extension for 1 min. Sanger sequencing was performed using the commercially available services provided by the UC Berkeley DNA Sequencing Facility (https://ucberkeleydnasequencing.com/home). Resulting sequences were aligned to EquCab3.0 (http://www.ncbi.nlm.nih.gov/genome/145) and analyzed with SEQUENCER software (Gene Codes Corp.). The n = 11 affected foals, n = 12 dams, n = 4 sires, n = 11 siblings, and n = 8 unrelated control horses were genotyped (S1 Table).
Population allele frequency
To estimate the Quarter Horse population allele frequency and test additional animals for the presence of this variant, 1,060 presumed unaffected Quarter Horses with samples banked at the UC Davis Veterinary Genetics Laboratory (VGL) were genotyped for the FDXR intronic variant. Samples were randomly selected from those with consent from owners or breed organizations to use for research purposes and individuals with known first degree relationships were removed. Genotyping was performed using the assay developed and now commercially available at the VGL (https://vgl.ucdavis.edu/test/equine-juvenile-spinocerebellar-ataxia-ejsca). The allele frequency for the FDXR variant was calculated as:
Where:
fD represents the frequency of the disease allele
SEfD represents the standard error of fD
HD/N represents the number of horses with one disease allele (heterozygous)
HD/D represents the number of horses with two disease alleles (homozygous)
n represents tested population number
An additional 181 horses across 26 breeds were screened and no other carriers were identified in other breeds (S3 Table).
FDXR conservation and expression
Scores were determined for each orthologous human variant using the 100-vertebrate score by phastCons (https://genome.ucsc.edu/) since conservation scores are not available in the EquCab3.0 genome browser within UCSC (https://genome.ucsc.edu/). FDXR expression was evaluated in tissues available from the equine FAANG initiative (https://www.ebi.ac.uk/ena/data/view/ERA1487553) [22].
FDXR c.177 + 1778G > C functional validation
RT-PCR.
To define the boundaries of the cryptic exon identified in FDXR, RT-PCR was performed on spinal cord, liver and gluteal muscle of n = 3 EJSCA foals and n = 3 controls (S1 Table). Primers were designed to anneal to exon 2 and exon 5 of FDXR-203, as annotated in the Ensembl database for EquCab3.0 (https://useast.ensembl.org/Equus_caballus/Info/Index). Primer sequences were F 5’CTTCTACACGGCCCAACACC 3’, R 5’ GAGCCCATTGTACCAGCCC 3’. Gel extraction and Sanger sequencing of the resulting bands was unsuccessful. Therefore, long-read sequencing was performed on the resulting bands from EJSCA and control horses.
PromethION nanopore long-read sequencing.
Long-read PromethION Nanopore sequencing of the resulting RT-PCR products from spinal cord tissue was performed in n = 2 EJSCA foals and n = 4 controls (S1 Table). Barcoding PCR was performed using the PBC001 kit, barcoded samples were pooled and the sequencing library was prepared using LSK114 kit. The libraries were sequenced on R10.4.1 PromethION flow cell. Data was base called in the super accurate mode using dorado v4.3.0, 400 bps. The yield was 12M reads with more than 1M reads per barcode. Only data >Q10 was used for downstream analysis. For analysis, reads were randomly subsampled at 100,000 reads per sample using seqtk [48] and primer sequences trimmed using a custom Python script. Reads within an inexact primer match were omitted from the analysis (S5 Table). Samples were then merged into a single file and the clustering algorithm CD-HIT [49] was used to cluster reads that were 98% similar to each other based on global sequence identify. Significant clusters were defined as those that had > 10% of at least one sample’s total reads in them. With the resulting four clusters, representing 82.5% of all analyzed reads, 25 sequences from each cluster were again randomly subsampled and ClustalW [50] was used to identify a consensus sequence using default parameters. Sequences were aligned to EquCab3.0 using BLAST.
Western blotting.
For total protein extraction, 10–20 mg of gluteal muscle, liver, and heart and 30–40 mg of spinal cord was used. Tissues were transferred to 1.5 mL microcentrifuge tubes on ice, 600 uL RIPA lysis buffer was added, and tissues were homogenized with plastic pestles. Contents were agitated on a plate shaker for 2 h at 4°C, centrifuged at 13,000 x g for 20 minutes at 4°C, and supernatant was collected. Isolated protein sample concentrations were quantified using the Pierce BCA Protein Assay Kit (Thermo Scientific; Cat no. 23225) and samples were stored at -80°C. All protein concentrations were diluted to 1 ug/1 ul with 1x Tris Glycine SDS and 15 ul lysate samples mixed with 7.13 ul 4x Laemmli sample buffer (Bio-Rad; Cat no. 1610747) and 0.75 ul 2-Mercaptoethanol (Bio-Rad; Cat no. 1610710). Samples were heated for 5 minutes at 90°C. SDS-PAGE and blotting was performed using the Bio-Rad Mini-PROTEAN Tetra cell, 10% Mini-PROTEAN TGX Precast Protein Gels, 10-well, 30 µl (Cat no. 4561033), and Bio-Rad Nitrocellulose Membranes, Precut, 0.2 µm, 7 x 8.4 cm (Cat no.16201460). 15 ug total protein was loaded for each sample. Biological replicates included n = 3 controls and n = 3 cases (spinal cord) and n = 2 controls and n = 3 cases (liver) (S1A, S1B, S1C, and S1D Fig).
To detect FDXR protein, a polyclonal rabbit IgG primary antibody (Novus Biologicals NBP2–38706) was used at a dilution ratio of 1/10000, and a goat anti-rabbit HRP-linked secondary antibody (Cell Signaling Technology 7074P2) was used at a dilution ratio of 1/10000. To detect actin, a mouse monoclonal IgG1 κ beta Actin (C4) primary antibody (Santa Cruz Biotechnology sc-47778) was used at a dilution ratio of 1/10000 and a goat anti-mouse IgG (H + L) linked cross-adsorbed HRP-linked secondary antibody (Invitrogen A16072) was used at a dilution ratio of 1/5000. Imaging was performed using the Bio-Techne ProteinSimple FluorChem E System according to the manufacturer’s instructions.
In the initial set of Western Blots (S1A, S1B, S1C, and S1E Fig), both human colon cell lysate and mouse lung lyase were used as positive controls, since human is the target species of the rabbit polyclonal anti-FDXR antibody (Novus NBP2‑38706) with confirmed reactivity at ~54 kDa. Mouse lung lyase produced a band at~54 kDa and therefore this antibody recognized both human and murine FDXR under our Western blot conditions and was subsequently included in all experiments as a positive control. Alignment of the FDXR immunogen peptide region to the predicted equine FDXR sequence shows 94.6% amino acid identity with the antibody epitope and alignment was 89.2% for the predicted murine FDXR protein sequence [51]. This indicates that mouse and horse FDXR are sufficiently conserved at the epitope level for reliable cross-species detection.
Supporting information
S1 Fig. Western blotting of FDXR in additional cases and controls.
https://doi.org/10.1371/journal.pgen.1012158.s001
(DOCX)
S1 Table. Overall demographics of horses used for the whole-genome sequencing and subsequent functional validation.
MC = male castrated, FS = female spayed, SC = spinal cord, GM = gluteal muscle, LIV = liver, HRT = myocardium.
https://doi.org/10.1371/journal.pgen.1012158.s002
(XLSX)
S2 Table. Partial vcf file and public dataset searches of 77 variants associated with EJSCA at at p < 1x10-11.
https://doi.org/10.1371/journal.pgen.1012158.s003
(XLSX)
S3 Table. Summary of breeds screened for the EJSCA mutation and corresponding genotype and estimated allele frequencies.
https://doi.org/10.1371/journal.pgen.1012158.s004
(XLSX)
S4 Table. Clusters identified by cDNA Nanopore long-read sequencing in control and EJSCA horses.
https://doi.org/10.1371/journal.pgen.1012158.s005
(XLSX)
S5 Table. Number of reads used for analysis after randomly subsampling 100,000 reads and filtering out inexact primer regions.
https://doi.org/10.1371/journal.pgen.1012158.s006
(XLSX)
Acknowledgments
We would like to acknowledge the owners and veterinarians of affected foals and healthy controls for allowing sampling of the foals, their dams and sires, and siblings.
References
- 1. Willis AT, Dahlgren AR, Woolard KD, Ghosh S, Donnelly CG, de la Concha-Bermejillo A, et al. Clinicopathological and pedigree investigation of a novel spinocerebellar neurological disease in juvenile Quarter Horses in North America. J Vet Intern Med. 2024;38(3):1808–14. pmid:38669583
- 2. Forman OP, De Risio L, Matiasek K, Platt S, Mellersh C. Spinocerebellar ataxia in the Italian Spinone dog is associated with an intronic GAA repeat expansion in ITPR1. Mamm Genome. 2015;26(1–2):108–17. pmid:25354648
- 3. Forman OP, De Risio L, Mellersh CS. Missense mutation in CAPN1 is associated with spinocerebellar ataxia in the Parson Russell Terrier dog breed. PLoS One. 2013;8(5):e64627. pmid:23741357
- 4. Forman OP, De Risio L, Stewart J, Mellersh CS, Beltran E. Genome-wide mRNA sequencing of a single canine cerebellar cortical degeneration case leads to the identification of a disease associated SPTBN2 mutation. BMC Genet. 2012;13:55. pmid:22781464
- 5. Letko A, Dietschi E, Nieburg M, Jagannathan V, Gurtner C, Oevermann A, et al. A missense variant in SCN8A in alpine dachsbracke dogs affected by spinocerebellar ataxia. Genes (Basel). 2019;10(5):362. pmid:31083464
- 6. Van Poucke M, Stee K, Sonck L, Stock E, Bosseler L, Van Dorpe J, et al. Truncating SLC12A6 variants cause different clinical phenotypes in humans and dogs. Eur J Hum Genet. 2019;27(10):1561–8. pmid:31160700
- 7. Agler C, Nielsen DM, Urkasemsin G, Singleton A, Tonomura N, Sigurdsson S, et al. Canine hereditary ataxia in old english sheepdogs and gordon setters is associated with a defect in the autophagy gene encoding RAB24. PLoS Genet. 2014;10(2):e1003991. pmid:24516392
- 8. Steinberg HS, Van Winkle T, Bell JS, de Lahunta A. Cerebellar degeneration in old english sheepdogs. J Am Vet Med Assoc. 2000;217(8):1162–5. pmid:11043686
- 9. Steinberg HS, Troncoso JC, Cork LC, Price DL. Clinical features of inherited cerebellar degeneration in Gordon setters. J Am Vet Med Assoc. 1981;179(9):886–90. pmid:7341602
- 10. Duchesne A, Vaiman A, Frah M, Floriot S, Legoueix-Rodriguez S, Desmazières A, et al. Progressive ataxia of Charolais cattle highlights a role of KIF1C in sustainable myelination. PLoS Genet. 2018;14(8):e1007550. pmid:30067756
- 11. Drögemüller C, Reichart U, Seuberlich T, Oevermann A, Baumgartner M, Kühni Boghenbor K, et al. An unusual splice defect in the mitofusin 2 gene (MFN2) is associated with degenerative axonopathy in Tyrolean Grey cattle. PLoS One. 2011;6(4):e18931. pmid:21526202
- 12. Thomsen B, Nissen PH, Agerholm JS, Bendixen C. Congenital bovine spinal dysmyelination is caused by a missense mutation in the SPAST gene. Neurogenetics. 2010;11(2):175–83. pmid:19714378
- 13. Timsit E, Albaric O, Colle MA, Costiou P, Cesbron N, Bareille N, et al. Clinical and histopathologic characterization of a central and peripheral axonopathy in Rouge-des-prés (Maine Anjou) calves. J Vet Intern Med. 2011;25(2):386–92. pmid:21281347
- 14. Duchesne A, Vaiman A, Castille J, Beauvallet C, Gaignard P, Floriot S, et al. Bovine and murine models highlight novel roles for SLC25A46 in mitochondrial dynamics and metabolism, with implications for human and animal health. PLoS Genet. 2017;13(4):e1006597. pmid:28376083
- 15. Harper PA, Healy PJ. Neurological disease associated with degenerative axonopathy of neonatal Holstein-Friesian calves. Aust Vet J. 1989;66(5):143–4, 145–6. pmid:2735893
- 16. Letko A, Strugnell B, Häfliger IM, Paris JM, Waine K, Drögemüller C, et al. Compound heterozygous PLA2G6 loss-of-function variants in Swaledale sheep with neuroaxonal dystrophy. Mol Genet Genomics. 2021;296(1):235–42. pmid:33159255
- 17. Harper PA, Plant JW, Walker KH, Timmins KG. Progressive ataxia associated with degenerative thoracic myelopathy in Merino sheep. Aust Vet J. 1991;68(11):357–8. pmid:1776933
- 18. Harper PA, Duncan DW, Plant JW, Smeal MG. Cerebellar abiotrophy and segmental axonopathy: Two syndromes of progressive ataxia of Merino sheep. Aust Vet J. 1986;63(1):18–21. pmid:3954688
- 19. Young AE, Finno CJ. Current insights into equine neuroaxonal dystrophy/equine degenerative myeloencephalopathy. Vet J. 2024;305:106129. pmid:38734404
- 20. Durward-Akhurst SA, Schaefer RJ, Grantham B, Carey WK, Mickelson JR, McCue ME. Genetic variation and the distribution of variant types in the horse. Front Genet. 2021;12:758366. pmid:34925451
- 21. Uhlén M, Fagerberg L, Hallström BM, Lindskog C, Oksvold P, Mardinoglu A, et al. Proteomics. Tissue-based map of the human proteome. Science. 2015;347(6220):1260419. pmid:25613900
- 22. Peng S, Dahlgren AR, Donnelly CG, Hales EN, Petersen JL, Bellone RR, et al. Functional annotation of the animal genomes: An integrated annotation resource for the horse. PLoS Genet. 2023;19(3):e1010468. pmid:36862752
- 23. Cingolani P, Patel VM, Coon M, Nguyen T, Land SJ, Ruden DM, et al. Using drosophila melanogaster as a model for genotoxic chemical mutational studies with a new program, SnpSift. Front Genet. 2012;3:35. pmid:22435069
- 24. Finno CJ, Gianino G, Perumbakkam S, Williams ZJ, Bordbari MH, Gardner KL, et al. A missense mutation in MYH1 is associated with susceptibility to immune-mediated myositis in Quarter Horses. Skelet Muscle. 2018;8(1):7. pmid:29510741
- 25. Alsaif HS, Alshehri A, Sulaiman RA, Al-Hindi H, Guzmán-Vega FJ, Arold ST, et al. MYH1 is a candidate gene for recurrent rhabdomyolysis in humans. Am J Med Genet A. 2021;185(7):2131–5. pmid:33755318
- 26. Aleman M, Riehl J, Aldridge BM, Lecouteur RA, Stott JL, Pessah IN. Association of a mutation in the ryanodine receptor 1 gene with equine malignant hyperthermia. Muscle Nerve. 2004;30(3):356–65. pmid:15318347
- 27. Kruijt N, den Bersselaar LV, Snoeck M, Kramers K, Riazi S, Bongers C. RYR1-related rhabdomyolysis: A spectrum of hypermetabolic states due to ryanodine receptor dysfunction. Curr Pharm Des. 2022;28(1):2–14.
- 28. Dahlgren AR, Careddu F, Norris JW, Di Buduo CA, Stanger L, Adili R, et al. ER-associated degradation pathway protein SEL1L plays an evolutionarily conserved role in platelet adhesion. J Clin Invest. 2026;136(4):e191433. pmid:41697739
- 29. Schneider CD, Luyendyk JP. One Health horsepower in hemostasis: SEL1Lecting stable platelet adhesion. J Clin Invest. 2026;136(4):e202267. pmid:41697747
- 30. Slone JD, Yang L, Peng Y, Queme LF, Harris B, Rizzo SJS, et al. Integrated analysis of the molecular pathogenesis of FDXR-associated disease. Cell Death Dis. 2020;11(6):423. pmid:32499495
- 31. Peng Y, Shinde DN, Valencia CA, Mo J-S, Rosenfeld J, Truitt Cho M, et al. Biallelic mutations in the ferredoxin reductase gene cause novel mitochondriopathy with optic atrophy. Hum Mol Genet. 2017;26(24):4937–50. pmid:29040572
- 32. Paul A, Drecourt A, Petit F, Deguine DD, Vasnier C, Oufadem M, et al. FDXR mutations cause sensorial neuropathies and expand the spectrum of mitochondrial fe-S-synthesis diseases. Am J Hum Genet. 2017;101(4):630–7. pmid:28965846
- 33. Stenton SL, Piekutowska-Abramczuk D, Kulterer L, Kopajtich R, Claeys KG, Ciara E, et al. Expanding the clinical and genetic spectrum of FDXR deficiency by functional validation of variants of uncertain significance. Hum Mutat. 2021;42(3):310–9. pmid:33348459
- 34. Vaz-Drago R, Custódio N, Carmo-Fonseca M. Deep intronic mutations and human disease. Hum Genet. 2017;136(9):1093–111. pmid:28497172
- 35. Campbell T, Slone J, Vu J, Liu W, Yang L, Dourson A, et al. Ferroptosis is a novel pathogenic mechanism of FDXR-related disease via disruption of the NRF2 pathway. Cell Death Discov. 2025;11(1):563. pmid:41436440
- 36. Chen Y, Davidson NM, Wan YK, Yao F, Su Y, Gamaarachchi H, et al. A systematic benchmark of Nanopore long-read RNA sequencing for transcript-level analysis in human cell lines. Nat Methods. 2025;22(4):801–12. pmid:40082608
- 37. Anna A, Monika G. Splicing mutations in human genetic disorders: Examples, detection, and confirmation. J Appl Genet. 2018;59(3):253–68. pmid:29680930
- 38. Kalbfleisch TS, Rice ES, DePriest MS, Walenz BP, Hestand MS, Vermeesch JR, et al. Improved reference genome for the domestic horse increases assembly contiguity and composition. Commun Biol. 2018;1:197. pmid:30456315
- 39. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25(14):1754–60. pmid:19451168
- 40. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009;25(16):2078–9. pmid:19505943
- 41. Van der Auwera GA, Carneiro MO, Hartl C, Poplin R, Del Angel G, Levy-Moonshine A, et al. From FastQ data to high confidence variant calls: The Genome Analysis Toolkit best practices pipeline. Curr Protoc Bioinformatics. 2013;43(1110):11.10.1-11.10.33. pmid:25431634
- 42. Cingolani P, Platts A, Wang LL, Coon M, Nguyen T, Wang L, et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly (Austin). 2012;6(2):80–92. pmid:22728672
- 43. Robinson JT, Thorvaldsdóttir H, Winckler W, Guttman M, Lander ES, Getz G, et al. Integrative genomics viewer. Nat Biotechnol. 2011;29(1):24–6. pmid:21221095
- 44. HTStream. https://github.com/s4hts/HTStream. 2020.
- 45. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21. pmid:23104886
- 46. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47. pmid:25605792
- 47. Untergasser A, Nijveen H, Rao X, Bisseling T, Geurts R, Leunissen JAM. Primer3Plus, an enhanced web interface to Primer3. Nucleic Acids Res. 2007;35(Web Server issue):W71-4. pmid:17485472
- 48. Li H. Seqtk: A fast and lightweight tool for processing FASTA or FASTQ sequences. 2013. https://github.com/lh3/seqtk
- 49. Fu L, Niu B, Zhu Z, Wu S, Li W. CD-HIT: Accelerated for clustering the next-generation sequencing data. Bioinformatics. 2012;28(23):3150–2. pmid:23060610
- 50. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994;22(22):4673–80. pmid:7984417
- 51. Ahmad S, Jose da Costa Gonzales L, Bowler-Barnett EH, Rice DL, Kim M, Wijerathne S, et al. The UniProt website API: Facilitating programmatic access to protein knowledge. Nucleic Acids Res. 2025;53(W1):W547–53. pmid:40331428