Introduction

The term “neuromuscular disorder” covers a broad category describing diseases that affect the peripheral nervous system, neuromuscular junction, or skeletal muscle. While a small percentage of neuromuscular disorders result from environmental factors such as exposure to environmental toxins [1], a greater percentage of neuromuscular disorders arise due to genetic variants [2]. It was estimated that 1 out of 3500 individuals worldwide may have a genetic neuromuscular disorder manifested during childhood or at later stages of life [3]. However, additional research has suggested that the actual number of individuals suffering from genetic neuromuscular disorders is much higher [4]. A study estimated world-wide prevalence rates to vary between 1 to 10 per 100,000 individuals for common neuromuscular disorders such as Charcot-Marie-Tooth disease, Myotonic dystrophy, Becker muscular dystrophy and facioscapulohumeral dystrophy [4]. The diverse symptoms manifested by patients with neuromuscular disorders and the absence of exact estimates about the incidence rate of rare neuromuscular disorders pose a major challenge in predicting the worldwide prevalence rate for the NMDs.

The heterogeneous nature of neuromuscular disorders also poses a problem in the correct diagnosis and possible treatment of the disorder. Moreover, as neuromuscular disorders involve progressive degeneration of muscles, leading to paralysis in extreme cases, therefore these are often difficult to differentiate from central nervous system disorders, hence posing another barrier in making a correct diagnosis [5, 6]. With the advent of massively parallel sequencing technologies, the diagnosis of neuromuscular disorders has become relatively easier. Neuro-Muscle DB (http://yu-mbl-muscledb.com/NeuroMuscleDB, accessed May 2022), Gene Table of Neuromuscular disorders (http://www.musclegenetable.fr /index.html, accessed May 2022) and DisGeNet (https://www.disgenet.org/home/ accessed May 2022) list a number of genetic variants in roughly around 170 genes associated with different types of neuromuscular disorders ranging from simple gait/muscular disorders to more life-threatening disorders such as amyotrophic lateral sclerosis. Genetic therapies based on compounds inducing exon skipping, or nonsense read-through and gene replacement are now being used to provide genetic treatment for some neuromuscular disorders [7, 8].

We report a family in which multiple individuals suffered from a neuromuscular disorder characterized by a waddling gait, limb deformities, muscular weakness and facial palsy. We identified a novel biallelic missense variant in DCAF13 segregating with this phenotype.

Materials and methods

Family ascertainment and exome sequencing analyses

This study was conducted after approval from the Institutional Review Board (IRB# 00005281, FWA 00010252) of School of Biological Sciences, University of the Punjab, Lahore, Pakistan. Written informed consents for participation as well as publication of images were obtained from all participants or the parents for their minor children. Family RDHR-03 had four affected individuals born to parents who were first cousins (Fig. 1A). Parents and their offspring participated in the study. Movements of all affected individuals were recorded according to a standardized video protocol. These videos were evaluated by a neurologist (KRK). MRI of brain and spinal cord was performed for one patient while nerve conduction studies, needle electromyography, and muscle biopsy could not be performed for any participant. Blood samples were collected from the available family members and genomic DNA was extracted by a standard protocol.

Fig. 1: Pedigree RDHR-03, phenotype, sequence traces, and depiction of DCAF13 with the encoded protein.
figure 1

A Family RDHR-03. The genotypes for DCAF13 ENST00000616836.4, c.1363 G > A:p.(Asp455Asn) or ENST00000612750.5, NM_015420.7, c.907 G > A:p.(Asp303Asn) variant are indicated below the individual symbols of all participants. *Individuals who participated in this study. Exome sequencing was performed for individual IV:1, IV:2 and IV:3. B Photographs of a patient from family RDHR-03 showing lower limb atrophy, tip-toe walking and a waddling gait. C Partial electropherograms of DCAF13 exon 8 sequence analyses. The site of the variant is indicated by an arrow. D–I Schematic representation of DCAF13 isoforms. Filled boxes represent translated exons and unfilled boxes denote the 5ʹ and 3ʹ untranslated regions. Introns are depicted by horizontal lines (not drawn to scale). Three isoforms, GENCODE transcript ID, ENST00000616836.4 ENST00000297579.9 and ENST00000612750.5 (NM_015420.7) consist of 11 exons with the latter encoding a 445 amino acids protein while the other two encode 597 amino acids proteins. The 445 amino acids protein lacks some amino acids from the N-terminal domain. The other shorter three isoforms contain four to five exons and will ultimately give rise to smaller proteins, lacking all but one WD40 domain. (J) Diagrammatic representation of domain structure of DCAF13 Q9NV06 isoform (encoded by ENST00000612750.5, NM_015420.7), consisting of 445 amino acid residues. The site of the variant in the protein p.(Asp303Asn) is represented by an asterisk. (K) Diagrammatic representation of domain structure of DCAF13 A0A087WT20 isoform (encoded by ENST00000297579.9 or ENST00000616836.4), consisting of 597 amino acids. The site of the variant in the protein p.(Asp445Asn) is represented by an asterisk.

Full size image

Exome sequencing was performed for affected individuals IV:1, IV:2 and IV:3 by using Agilent SureSelect Human All Exon V5 capture library and Illumina Hiseq 2000 sequencer with a coverage of 100X (Macrogen Inc, Seoul, South Korea and 3billion inc. Seoul, South Korea). We analyzed the variant call file data using wANNOVAR (https://wannovar.wglab.org, accessed April 2019) and Franklin (https://franklin.genoox.com/clinical-db/home, accessed January 2021) as well as by 3billion automated prioritization system. The analyzed data contained allele frequencies of the detected variants in the public databases as well as predictions of the pathogenicity from several online software. We filtered the exonic and intronic variant data against all known variants in public databases and those with minor allele frequencies equal to or greater than 0.01 were removed. Homozygous variants present in the data of all three individuals were prioritized and we also examined compound heterozygous variants.

Primers were designed using primer3 (https://bioinfo.ut.ee/primer3-0.4.0/, accessed April 2019) for amplification and subsequent Sanger sequencing on samples from available family members to confirm the segregation of the candidate variants (Table S1). DNA from 150 unrelated ethnically matched controls was genotyped by Tetra-ARMS PCR [9] to check the frequency of the variant in the Pakistani population using a primer set for a constant product and two primers for allele specific products amplification (Table S1). These primers were designed using Primer1 (http://primer1.soton.ac.uk/primer1.html, accessed April 2019). Frequency of the variants was also assessed by observing their presence in the data of 150 unrelated ethnically matched in-house exome data. Thus a total of 300 controls were assessed (150 + 150). Variant data has been deposited in LOVD (https://www.lovd.nl/, in August 2022), ID: 0000877191.

The exome vcf data was analyzed using AgileVCFMapper [10] to identify regions of homozygosity from the exome sequences in order to verify that the disease variant was present in a chromosomal region identical by descent.

Expression analysis of Dcaf13

Tissues from wild type mice were obtained after ethical review from the animal use committee at the School of Biological Sciences, University of the Punjab. P7 mice brains were dissected. Tissue was disrupted by repeated aspiration with a sterile medical syringe and RNA was extracted by TriZol (Invitrogen). Random primed cDNA libraries were synthesized using RevertAid First Strand cDNA Synthesis Kit #K1622 (Thermo Fisher Scientific). The 1338 bp open reading frame of Dcaf13 was amplified from the prepared cDNA library by PCR with Phusion DNA polymerase using specific primers (Table S1). A separate nested PCR was also performed to amplify part of Dcaf13 to further confirm findings of expression (Table S1).

Evolutionary conservation analyses

The gene models from the various eukaryote species all encode the shorter isoform of 445 residues ENST00000612750.5 (NM_015420.7); therefore it was used for conservation analysis. We used a phylomedicine approach [11] to determine whether Asp303 in DCAF13 is evolutionarily conserved across other organisms. We BLASTed (discontiguous megablast) the human DCAF13 sequence in GenBank (NM_015420.7) across a targeted set of species in NCBI’s nucleotide collection. The general strategy for taxonomic sampling was to have a high representation of mammals, but then expand species diversity such that we had a mix of species more closely related to humans (e.g., Primates and other mammals) as well as highly divergent taxa (e.g., insects, cnidarians, plants, algae). In instances where there were negative BLAST results (e.g., due to sequence divergence or potentially gene absence), we manually searched NCBI’s nucleotide collection using taxonomic terms + DCAF13 or “DDB1 and CUL4 associated factor 13”. Note that while some DCAF13 orthologues in NCBI are the products of mRNA (cDNA) sequencing, the vast majority are derived from NCBI’s eukaryotic genome annotation (EGA) pipeline, and therefore represent predicted gene sequences that need experimental validation.

We examined DCAF13 homologs in 97 species (spp.) of eukaryotes, consisting predominantly of vertebrates (55 spp.), as well as invertebrates (24 spp.), plants (seven spp.) and unicellular to colonial eukaryotes, such as molds, amoebae and algae (11 spp.) (Table S2). We extracted the coding region for all obtained DCAF13 sequences (Table S2) and imported them into Geneious Prime ver. 2019.2.3 [12]. Sequences were successively aligned to human DCAF13 using Muscle ver 3.8.425 [13] in Geneious, beginning with more closely-related species and working towards distantly-related taxa. We typically began by using the nucleotide alignment function due to higher sequence similarity for closely-related species, but for more divergent sequences we primarily used translation alignment to align based on amino acid similarity.

After completing the alignment (Dataset S1), we performed a RAxML ver. 8.2.11 [14] gene-tree analysis (GTR GAMMA, rapid hill-climbing) to validate that the obtained gene sequences are most likely orthologues of human DCAF13. We then compared the human Asp303 residue and flanking amino acids to all other species in the alignment to determine whether this position is evolutionarily conserved.

Predicted function, domains, and structure of DCAF13

Human Protein Atlas (https://www.proteinatlas.org/, accessed January 2022) and Expression Atlas (https://www.ebi.ac.uk/gxa/home, accessed January 2022) were queried to identify the cellular and subcellular expression and localization of DCAF13. Human Brain Transcriptome database (https://hbatlas.org/pages/hbtd, accessed January 2022) was used to identify the brain regions with high DCAF13 gene expression in adult and developing human brain. STRING analysis (https://string-db.org/, accessed January 2022) of human DCAF13 (Q9NV06) was performed to pin down the functional association network and co-expression profile for DCAF13.

A combination of three programs; InterPro (https://www.ebi.ac.uk/interpro/, accessed January 2022), Pfam (http://pfam.xfam.org/, accessed January 2022) and ScanProsite (https://prosite.expasy.org/, accessed January 2022) were used for performing conserved domain analysis of the protein. Using these data, we mapped the identified variant position to the predicted domain structure of DCAF13. To find out the secondary structure element of DCAF13 affected by the amino acid change, we utilized protein secondary structure prediction tools i.e., PDBsum (http://www.ebi.ac.uk/thornton-srv/databases/cgi-bin/pdbsum/GetPage.pl?pdbcode =index.html, accessed January 2022) and PSIPRED (http://bioinf.cs.ucl.ac.uk/psipred/, accessed January 2022).

We also performed in-silico structural analysis of the mutant DCAF13 protein (UniProt IDs: Q9NV06 and A0A087WT20 corresponding to 445 and 597 amino acids proteins respectively). No structures were available for DCAF13 in Protein Data Bank (https://www.rcsb.org/, accessed January 2022) for any of the protein isoforms. Since both isoforms Q9NV06 and A0A087WT20 contained similar domains, we performed all subsequent analysis using the A0A087WT20 corresponding to the longest 597 amino acids isoform. We carried out homology modeling for A0A087WT20 using Swiss Model (https://swissmodel.expasy.org/, accessed January 2022) and Phyre2 (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index, accessed January 2022). Since the built structures had low percentage identity and low coverage, therefore we utilized i-TASSER (https://zhanglab.dcmb.med.umich.edu/I-TASSER/, accessed January 2022), online tool for ab-initio modeling of unknown protein structures [15] for more accurate structure prediction of DCAF13 (A0A087WT20). After modeling a native protein structure of DCAF13 isoform as described above, the p.(Asp455Asn) amino acid change was introduced using the Swiss-PDB viewer (https://spdbv.vital-it.ch/, accessed January 2022). Energy minimization for the mutant DCAF13 structure p.(Asp455Asn) was performed using GROMOS96 force field [16, 17]. 3D structure visualization of DCAF13 wild type and mutant structure was performed using iCn3D (https://www.ncbi.nlm.nih.gov/Structure/icn3d/full.html, accessed January 2022).

To further understand the possible implications of the novel DCAF13 variant on the structural integrity of the encoded protein, we utilized online protein stability prediction server tools I-Mutant (https://folding.biofold.org/i-mutant/i-mutant2.0. html, accessed January 2022), MUpro (http://mupro.proteomics.ics.uci.edu/, accessed January 2022) and CUPSAT (http://biosig.unimelb.edu.au/mcsm/, accessed January 2022). We determined the effect of the identified amino acid change on protein stability using Q9NV06 (445 amino acids) and A0A087WT20 (597 amino acids) protein isoforms of DCAF13 (NP_056253).

Results

Clinical data

Family RDHR-03 comprises of four affected individuals which includes three males and one female (Fig. 1A, Table 1). The ages of the participants ranged between 7 to 20 years at the time of sampling in 2017. The onset of the disease was reported to be at the age of 1 year. Individual IV:1 had right lower (complete) facial nerve palsy, although we were unable to exclude an alternative explanation such as Bell’s palsy. Individual IV:2 and IV:3 had a severe waddling gait (Fig. 1B) which is probably due to proximal muscle weakness (favoring a myopathic process rather than a neuropathic disorder). There was no definite foot drop to suggest distal weakness. Individual IV:2’s brain MRI findings were suggestive of demyelination process in the both frontal lobes. Cervicodorsal and lumbosacral spine MRI features of individual IV:2 were suggestive of cystic dilated CSF cavity with no myelomalacia changes (data not shown). Individual IV:4 had a similar waddling type gait with lower limb wasting, equinus deformity/toe walking/achilles tendon contracture, which could be a myopathic or neuropathic disorder. The phenotype was most consistent with a myopathic disorder given the severe waddling gait in multiple affected family members. However, the creatine kinase was normal and we were not able to proceed with nerve conduction studies, needle electromyography or muscle biopsy due to unwillingness of the patients to undergo further testing.

Table 1 Summary of clinical features in family RDHR-03.
Full size table

Exome analyses implicate a DCAF13 variant in pathogenesis of the disease

Only three different homozygous exonic variants were shared by the three affected individuals as identified after the filtering criteria were applied to the analyzed exome data (Table 2). Two of these (in CC2D2A & LRP12) were not considered further due to non-segregation of the former with the phenotype in the family or prediction to be benign of the latter by multiple software including low REVEL and CADD scores (Table 2). Though predicted to be benign by eight software, a modifying effect of LRP12 variant cannot be excluded. However, only the missense variant in DCAF13 (ENST00000616836.4, NM_015420.6, c.1363 G > A:p.(Asp455Asn), ENST00000612750.5, NM_015420.7, c.907 G > A:p.(Asp303Asn), rs1209794872), located in a region of homozygosity at Chr 8: 74 Mb -108Mb (Fig. S1), was predicted pathogenic and segregated with the phenotype. The region of homozygosity on Chr 4: 7.9 Mb - 37.63 Mb (Fig. S1) was not supported by the data of the fourth affected individual who was heterozygous for a variant in this region (Table 2).

Table 2 Filtered homozygous variants after exome sequencing.
Full size table

The variant in DCAF13 was homozygous in the four affected individuals, heterozygous in the parents and heterozygous in the unaffected sibling (Fig. 1A, Fig. 1C). The variant was absent from 600 ethnically-matched chromosomes and was extremely rare in all public databases (gnomAD, 0.000007081, none homozygous). It was predicted to be damaging to protein by multiple software and had a high pathogenicity score (CADD = 34, ClinPred = 0.9973, REVEL = 0.889). None of the three pairs of heterozygous variants (Table S3); likely compound heterozygous in all three affected individuals, were predicted to be pathogenic in causing a recessive disorder (Table S4). No phenotypic match from additional patients has been received to date after we added the name of DCAF13 to GeneMatcher (https://genematcher.org/, added in April 2019).

Dcaf13 is expressed in the mouse brain and muscles

We detected Dcaf13 specific products from cDNA library constructed from mouse brain extracted RNA (data not shown). We did not have human cDNA libraries available for expression analysis. However, Human Protein Atlas and Expression Atlas demonstrated high expression of DCAF13 in the brain, especially in the regions of cerebellum, cerebral cortex, amygdala, and caudate nucleus (Fig. S2). Moreover, DCAF13 is transcribed in both the developing and the adult brain (Fig. S3). Furthermore, muscle cells; smooth, and skeletal muscles, showed medium levels of DCAF13 (data not shown). Expression Atlas search revealed differential RNA expression levels of DCAF13 across different brain regions in unaffected individuals versus patients of Alzheimer’s disease (https://www.ebi.ac.uk/gxa/home, accessed January 2022), (data not shown).

DCAF13 has multiple isoforms

DCAF13 (DDB1 and CUL4 associated factor 13) resides on chromosome 8q22.3 and is transcribed into multiple isoforms (Fig. 1D-I). Three of these isoforms are comprised of eleven exons each (Fig. 1D–F). Two isoforms, ENST00000616836.4 (NM_015420.6), and ENST00000297579.9, encode a 597 amino acids protein, while one ENST00000612750.5 (NM_015420.7) encodes a 445 amino acids protein. There are other isoforms as well; all encoding shorter proteins of 96, 185 or 221, amino acids respectively (Fig. 1G–I), missing two of the three WD40 and multiple other domains of the longer isoforms (Fig. 1J, K). In addition, there are three processed pseudogenes located on chr 7, chr 5, chr 15, DCAF13P1, DCAF13P2 and DCAF13P3, respectively (data not shown).

DCAF13 is predicted to interact with multiple proteins

DCAF13 is a member of protein orthologous group KOG0268, which includes Sof1-like rRNA proteins. STRING analysis suggested that DCAF13 [Q9NV06 (NP_056235)] may interact with UTP18, WDR46, and RPR9 (Fig. S4). These proteins have a role in the processing of rRNA.

Asp303has been conserved during the evolution of eukaryotes

The amino acid Asp303 affected by the variant has been absolutely conserved during the evolution of many eukaryotes, including invertebrates, plants, molds and algae (Fig. 2). We did not find any DCAF13-like sequences in Bacteria or Archaea, suggesting that this gene is unique to eukaryotes. The annotations for nearly all of the animals and plants support the conclusion that the genes we obtained are orthologues of human DCAF13 (Table S2). The remaining species have more ambiguous annotations (e.g., “hypothetical protein”), but our alignment (Fig. 2, Dataset S1) and gene tree analysis (data not shown) support the interpretation that these homologous genes are also DCAF13. Furthermore, recent robust phylogenetic analyses of eukaryotes suggest that the most recent common ancestor of animals and plants is the ancestor for all eukaryotes [18, 19], suggesting that DCAF13 was present in this ancestor, further increasing the likelihood that these ambiguously annotated genes indeed represent DCAF13. Importantly, for all species in which we examined these DCAF13-like sequences, Asp303 was universally conserved (Fig. 2, Dataset S1). This suggests that during the last 1.21–2.38 billion years since the origin of eukaryotes [18, 19], natural selection has maintained Asp303 in DCAF13, pointing to the functional significance of aspartic acid at this position.

Fig. 2: Conservation of DCAF13 Asp303 in a sampling of eukaryote species.
figure 2

Hypothesized evolutionary relationships and divergence times of focal species are shown on the left, with a partial alignment of DCAF13 (positions 294–312) shown on the right. Asp303 (indicated by an arrow) is conserved in all species examined, including unicellular eukaryotes (e.g., alveolate, red alga, euglenozoan). For full details, see Table S2 and Dataset S1.

Full size image

Other amino acids surrounding Asp303 are not similarly conserved (for example Phe302, Lys304) which further points to importance of Asp303 at this position. In addition, comparing all probable homologous sites of DCAF13 across multiple subgroups of eukaryotic species, we found that amino acid conservation becomes increasingly infrequent for sites other than Asp303: for the Primates examined, 389 of 445 sites (87.4%) are identical; for mammals, 303 of 445 sites are identical (68.1%); for vertebrates, 190 of 445 sites (42.7%) are identical; for animals, 79 of 445 (17.8%) are identical; and among all eukaryotes examined, only 68 of the 445 (15.3%) total residues are identical.

The DCAF13 p.(Asp455Asn, Asp303Asn) variant affects a WD40 domain

InterPro, Pfam and ScanProsite revealed presence of three WD40 domains, one eukaryotic initiation factor 2 A like domain and one Sof1 like domain in DCAF13 (Fig. 1J, K). The amino acid change p.(Asp455Asn) or (Asp303Asn) affected one of the three WD40 domains of DCAF13 (Fig. 1J, K). According to PDBsum and PSIPRED analysis of DCAF13 secondary structure, this Asp residue resides in a β hairpin loop structure of WD40 domain (Fig. S5, Fig. S6).

Variant DCAF13 p.(Asp455Asn, Asp303Asn) may not cause structural changes in the protein

A comparison of Swiss Model, Phyre2 and i-TASSER built structures for DCAF13 (A0A087WT20) showed that i-TASSER predicted structure has a greater accuracy as compared to the others (Fig. S7). A comparison of the native with the mutant DCAF13 proteins revealed no visible structural changes due to the amino acid change p.(Asp455Asn) or p.(Asp303Asn) (Fig. S8).

The DCAF13 p.(Asp455Asn, Asp303Asn) variant may decrease protein stability

I-Mutant and MUpro predicted a decrease in protein stability of DCAF13 (A0A087WT20) due to the p.Asp455Asn change. CUPSAT predicted the change of Asp455 to Asn455 as destabilizing to the protein with an unfavorable ΔΔG value of -2.54 kcal/mol. Similarly, mCSM predicted a ΔΔG value of -0.649 kcal/mol indicating high instability of the protein.

Discussion

DCAF13 functions in ubiquitination of proteins and rRNA processing, as it is part of the DCAF subfamily proteins which act as substrate receptors for Cullin-RING E3 ubiquitin ligase complexes such as CUL4, assisting in the ubiquitination of proteins [20, 21]. Recent studies suggest that DCAF proteins not only serve the role of receptor adaptors imparting substrate specificity to ubiquitin machinery but also participate in the degradation of the proteins involved in cell proliferation and growth at different time points in the cell cycle [22]. DCAF13 localizes to numerous cellular compartments such as the nucleolus, the centrosome, the nuclear lumen, and the CUL4-RING E3 ubiquitin ligase complex [23].

DCAF13 has three WD40 repeats. These repeats are a feature common to all DCAF subfamily proteins. The WD40 domain is a doughnut-shaped beta-propeller domain having seven blades, involved in protein-protein interactions [24]. The presence of WD40 repeats in DCAF13 has been shown to enable it to assist protein-protein interactions in the ubiquitin proteasome pathway (UPP) [25]. The identified variant c.1363 G > A:p.(Asp455Asn) or c.907 G > A:p.(Asp303Asn) affected the third WD40 repeat of DCAF13 and therefore the variant may disrupt protein-protein interactions. Moreover, the presence of p.Asp455 or p.Asp303 residue within the β-hairpin loop, a motif known to play an important role in protein folding [26, 27], suggests that the variant could also affect this process. Many previous studies show a link between misfolded proteins and the onset of neuromuscular and neurodegenerative disorders [28, 29]. In addition, the p.Asp455Asn or p.Asp303Asn involves a change from a polar negatively charged (acidic) amino acid, Asp to polar uncharged (amide containing) amino acid, Asn. This change in the polarity and the charge of the concerned amino acid residue, combined with the fact that the variant is present in the β-hairpin loop, strongly supports the in-silico predictions that the variant disrupts the stability of DCAF13.

In previous studies, DCAF13 variants have been described to cause various human disorders, though a definitive link has not been established (Table S5) and none of them affect evolutionary conserved amino acids. A study employing exome sequencing to identify disease-causing variants in a Chinese family suffering from cortical myoclonic tremor with epilepsy suggested that a heterozygous variant in DCAF13 (ENST00000616836.4, NM_015420.6, c.20 G > C:p.(Trp7Ser), ENST00000612750.5, NM_015420.7, c.- 437 G > C) alone, or together with another heterozygous variant in CCN3, NM_002514.4, also called NOV, c.983 T > C:p.(Ile328Thr) causes the phenotype in five affected individuals [30]. Single nucleotide rare or common variants of DCAF13 have been associated with low bone mineral density, long QT interval, autism spectrum disorder and idiopathic scoliosis [31,32,33,34]. Scoliosis or autism was not observed in the participants of our family while other conditions could not be assessed. In addition, comparative in silico analyses of gene expression profiles of four common neurodegenerative disorders; Alzheimer’s disease, Parkinson’s disease, Frontotemporal Dementia and Amyotrophic lateral sclerosis, indicated DCAF13 as one of the differentially expressed gene common among all these disorders [35].

Dcaf13 knockout mice die at preimplantation stage [36]. A conditional oocyte-specific Dcaf13 deletion mouse model determined that DCAF13 is crucial for tight packaging of chromatin to avoid the loss or damage of chromatin material during the maternal-zygotic transition [37]. Conditional oocyte-specific Dcaf13 knockout female mice are infertile [38]. It is not known if a biallelic deletion of DCAF13 will cause early lethality in humans as well, and if not, will the loss cause female infertility. The single female patient in our family with the DCAF13 missense variant was too young to evaluate for reproductive issues.

A recent study has implicated RNAi knockdown of dcaf13 in C. elegans to be important for worm development and determining lengths of the larvae [39]. Fertility of the worms was also affected. This latter effect was probably not due to embryonic lethality but because of a specific role of DCAF13 in reproduction, which remains to be determined.

DCAF13 overexpression has been correlated with poor prognosis in various cancers [40, 41]. On the converse, as stated above, its deficiency impacts chromatin compaction and female fertility in conditional knockout mice [38]. Other genes involved in neurodevelopment also have this effect in which both increased and decreased dosage of the protein is harmful. An example is UBE3A which causes both Angelman syndrome or autism spectrum disorder depending on either loss of functional protein or its overexpression [42, 43].

Variants of other genes encoding DCAF subfamily proteins have been previously linked to a wide range of neurological disorders in humans. Woodhouse Sakati Syndrome; an autosomal recessive neuroendocrine disorder, is caused by biallelic variants of DCAF17 [44, 45]. Exome sequencing identified the cause of axonal hereditary sensory and motor neuropathy as a homozygous variant in DCAF8 p.(Arg317Cys) segregating with the disease [46]. A study on neuronal development in Drosophila melanogaster showed that DCAF12 is essential for the maintenance of optimal neuronal function and plasticity at neuromuscular junctions [47]. Exome sequencing of unrelated patients with common symptoms of Chung-Jansen syndrome (developmental delay, intellectual and cognitive decline, hypotonia, and dysmorphic facial features) pinpointed novel de novo heterozygous variants in DCAF14 (HGNC approved symbol PHIP) as the reason for the observed phenotype [48].

The combined information of the DCAF13 c.1363 G > A:p.(Asp455Asn) variant’s perfect segregation in multiple affected individuals of a family; the predicted effect on protein stability observed by in silico analysis, together with the previous evidence of an association between the gene’s heterozygous variants and neurological disorders, support the observation that biallelic DCAF13 defects in humans may lead to neuromuscular or developmental abnormalities. However, the functional bases of these disorders are yet to be deciphered fully and till then association of DCAF13 variants with disease remains tentative. In vitro studies with native and mutant DCAF13 protein are warranted. Though knockout of Dcaf13 causes lethality in mice, it is possible that a knock-in mouse model may be viable. Such a comparison would answer in detail the questions regarding DCAF13 functions and how the changes in protein can manifest themselves in the form of disorders.