Expanding the genetic spectrum of hereditary motor sensory neuropathies in Pakistan

Abstract

Background

Hereditary motor and sensory neuropathy (HMSN) refers to a group of inherited progressive peripheral neuropathies characterized by reduced nerve conduction velocity with chronic segmental demyelination and/or axonal degeneration. HMSN is highly clinically and genetically heterogeneous with multiple inheritance patterns and phenotypic overlap with other inherited neuropathies and neurodegenerative diseases. Due to this high complexity and genetic heterogeneity, this study aimed to elucidate the genetic causes of HMSN in Pakistani families using Whole Exome Sequencing (WES) for variant identification and Sanger sequencing for validation and segregation analysis, facilitating accurate clinical diagnosis.

Methods

Families from Khyber Pakhtunkhwa with at least two members showing HMSN symptoms, who had not previously undergone genetic analysis, were included. Referrals for genetic investigations were based on clinical features suggestive of HMSN by local neurologists. WES was performed on affected individuals from each family, with Sanger sequencing used to validate and analyze the segregation of identified variants among family members. Clinical data including age of onset were assessed for variability among affected individuals, and the success rate of genetic diagnosis was compared with existing literature using proportional differences and Cohen’s h.

Results

WES identified homozygous pathogenic variants in GDAP1 (c.310 + 4 A > G, p.?), SETX (c.5948_5949del, p.(Asn1984Profs*30), IGHMBP2 (c.1591 C > A, p.(Pro531Thr) and NARS1 (c.1633 C > T, p.(Arg545Cys) as causative for HMSN in five out of nine families, consistent with an autosomal recessive inheritance pattern. Additionally, in families with HMSN, a SETX variant was found to cause cerebellar ataxia, while a NARS1 variant was linked to intellectual disability. Based on American College of Medical Genetics and Genomics criteria, the GDAP1 variant is classified as a variant of uncertain significance, while variants in SETX and IGHMBP2 are classified as pathogenic, and the NARS1 variant is classified as likely pathogenic. The age of onset ranged from 1 to 15 years (Mean = 5.13, SD = 3.61), and a genetic diagnosis was achieved in 55.56% of families with HMSN, with small effect sizes compared to previous studies.

Conclusions

This study expands the molecular genetic spectrum of HMSN and HMSN plus type neuropathies in Pakistan and facilitates accurate diagnosis, genetic counseling, and clinical management for affected families.

Background

Hereditary motor and sensory neuropathy (HMSN), commonly referred to as Charcot-Marie-Tooth disease (CMT), entails progressive nerve diseases marked by demyelination and/or axonal degeneration [1,2,3]. This often results in peripheral nerve dysfunction, occasionally accompanied by a noticeable slowing in nerve conduction velocity (NCV) [4]. Symptoms typically manifest in the first two decades, leading to significant disability and healthcare burden [5, 6]. Initial presentations involve distal-to-proximal strength and sensation loss, predominantly in the legs, and are often accompanied by complications such as foot deformities and fine motor skill impairments in the hands [7]. The risk of falls escalates with fatigue and proprioceptive loss, with symptoms overlapping with hereditary motor neuropathy (HMN), motor neuron diseases and other neurodegenerative diseases [1, 2, 8,9,10], thereby accelerating disease progression. The hereditary pattern of HMSN significantly influences its progression, with autosomal recessive manifestations generally exhibiting earlier onset and more severe symptoms compared to autosomal dominant forms [11]. It is noteworthy that HMSN is observed globally, with an estimated prevalence of 1 in 3,300 individuals [12, 13]. However, in regions where consanguineous unions are prevalent, such as Pakistan with a 65% incidence rate, autosomal recessive inheritance could notably contribute to HMSN cases [14]. Given this cultural practice, it is plausible that the prevalence of familial autosomal recessive forms of HMSN may be notably elevated in such populations, presenting an opportunity for genetic analysis and further research.

Previous studies investigating clinical and molecular aspects within HMSN families with autosomal recessive inheritance have revealed numerous associated genes with the condition [15, 16]. Despite these analyses, a significant number of patients remain genetically undiagnosed, underlining the critical need for ongoing exploration into HMSN families [10, 15] to establish molecular diagnoses, expand the spectrum of genetic variants, and provide comprehensive carrier counseling. For demyelinating forms of HMSN, genes such as GDAP1, MTMR2, SBF2, NDRG1, EGR2, SH3TC2, PRX, FGD4, PMP22, GJB1, MPZ, MFN2, MED25 and FIG4 have been identified [17, 18]. Similarly, for axonal forms of HMSN, genes including LMNA, MED25, HINT1, GDAP1, LRSAM1, NEFL, HSPB1, MFN2, PLA2G6, PNKP, AIFM1, COA7, and KIF1A have been implicated [19, 20]. Additionally, the detection of causal gene variants exhibits a notable discrepancy between demyelinating and axonal forms of HMSN, with rates reaching up to 87% in the former and 36% in the latter [21, 22]. Research findings suggest that a significant majority, approximately two-thirds, of HMSN cases can be attributed to sequence variations in key genes such as PMP22, GJB1, MPZ, and MFN2 [10, 23]. In recessive variants of the condition, albeit less common, variants in genes like SH3TC2 or the recently identified SORD gene emerge as principal genetic determinants [10, 24, 25].

Genomic technologies, especially Whole Exome Sequencing (WES), have transformed the diagnosis of HMSN by enabling thorough analysis of genetic variants throughout the exome [26, 27]. These methods are essential for detecting causative variants in complex cases where conventional diagnostic approaches may be insufficient or unavailable. In Pakistan, several studies have leveraged WES to explore the genetic landscape of HMSN, revealing diverse genetic variants within the local population. Notably, genes such as IGHMBP2, HSPB1, MTMR2, AMACR, GDAP1, MFN2, SH3TC2, HK1, REEP1, MPV17, SACS, PRX, RTN2, and GJB1 have been exclusively linked to HMSN, as documented by various researchers [28,29,30,31,32,33,34,35]. Rare cases of HMSN in Pakistani families reveal genetic associations with additional phenotypes: a FIG4 variant with both HMSN and Yunis-Varón syndrome [29], MPV17 and SACS variants with both HMSN and ARSACS [35], and NARS1 variants with HMSN and other neurodevelopmental disorders [36]. Notably, pathogenic variants in families with autosomal recessive forms of HMSN across different regions of Pakistan have been reported. However, the genetic spectrum of HMSN in families from the Khyber Pakhtunkhwa region of Pakistan, known for its consanguineous and tribal endogamy practices, has rarely been investigated. In the current study, nine Pakistani families affected by HMSN from the Khyber Pakhtunkhwa region were analyzed using WES to identify disease-associated causative variants. Subsequently, causative biallelic variants were identified in GDAP1, SETX, IGHMBP2, and NARS1 in five out of nine families, contributing to autosomal recessive forms of HMSN or HMSN plus type of neuropathies within Pakistani families.

Methods

Study Design

This study was specifically designed to find the genetic causes of HMSN in families from Khyber Pakhtunkhwa, Pakistan. Ethics was approved by the Kohat University of Science and Technology (KUST) ethics committee [KUST/Ethical Committee/2241]. The research adhered to the ethical guidelines established by the Declaration of Helsinki. Written informed consent was obtained from adult participants and the parents of affected children in the participating families. The data were obtained through clinical evaluations and genetic investigations performed from March 2022 to December 2023.

Inclusion and exclusion criteria

Families were eligible for inclusion if they met the following criteria:

a. Inclusion criteria

• Resided in Khyber Pakhtunkhwa, Pakistan, with at least two members exhibiting symptoms characteristic of HMSN.

• Had not previously participated in genetic testing or provided blood/DNA samples for genetic research.

b. Exclusion criteria

• Exhibited symptoms of HMSN but had previously undergone genetic testing or provided blood/DNA samples for genetic research.

Outcome Measures

a. Clinical assessment and genetic referral

The probands’ detailed medical history, including age of onset, symptomatology, and family history of neurological and genetic disorders, was carefully documented from medical records. Local neurologists at district hospitals in Khyber Pakhtunkhwa region of Pakistan then conducted comprehensive physical and neurological examinations, assessing muscle strength, sensation, deep tendon reflexes, and coordination. These evaluations were based on clinical criteria for diagnosing peripheral neuropathy, specifically targeting symptoms such as muscle weakness, sensory loss, diminished or absent reflexes, and motor coordination difficulties, characteristic of HMSN. When available, nerve conduction studies and imaging tests (such as magnetic resonance imaging) were included to further support the diagnosis. The detailed documentation of the probands’ medical and family history, alongside the clinical identification of HMSN symptoms, facilitated the referral of the probands and their family members for genetic investigations. Blood samples from probands and their accompanying family members were collected by trained phlebotomists immediately following clinical evaluations and referrals in the hospital. Samples from remaining family members were collected at their residences.

b. Genetic investigations

Genomic DNA was extracted from blood samples via the Mag-Bind^® (Blood & Tissue DNA Miniprep System) according to the manufacturer’s instructions. To identify the causative gene, WES was performed on a single affected individual in each family (Fig. 1) to develop a profile of variants not present in publicly available databases and rare sequence variants. Coding regions were captured by HiSeq2000 using paired-end (2 × 100) protocol at a mean coverage depth of 30X at the Otogenetics Corporation (Norcross, GA, USA). The Agilent SureSelect Human All ExonV4 (51 Mb) enrichment kit was used for exome enrichment. The sequence reads were aligned to the human genome reference sequence [hg38] and read alignment, variant calling, and annotation were performed by DNAnexus (DNAnexus Inc., Mountain View, CA; https://dnanexus.com).

On the WES data of probands, various filters were applied for identification of potentially causative variants. Homozygous and compound heterozygous variants were prioritized due to the observed autosomal recessive inheritance pattern in the families. Additionally, non-synonymous variants in exonic or splice site regions with an allele dosage of 50:50 were considered. The variants were also filtered according to minor allele frequency (MAF) ≥ 0.01% in the 1000 Genomes Project, Genome Aggregation Database (gnomAD) v 4.0, and Exome Aggregation Consortium (ExAC) databases to identify potentially deleterious changes. Furthermore, variants showing associations with disease phenotypes in the HGMD (http://www.hgmd.cf.ac.uk/ac/search.php), Clinvar (https://www.ncbi.nlm.nih.gov/clinvar/), PubMed (https://www.ncbi.nlm.nih.gov/pubmed/), and OMIM (https://www.ncbi.nlm.nih.gov/omim/) databases were compiled to create a list of genetic variants as potential causative factors for further analysis.

Validation segregation studies involved polymerase chain reaction (PCR) and Sanger (dideoxy) sequencing using allele-specific primers designed using primer3 software (https://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) using standard protocols. The amplified PCR products were sequenced by Source BioScience LifeSciences (https://www.sourcebioscience.com/) and Beijing Tsingke Biotech Co., Ltd. (https://tsingke.com/pages/about-us-1). The Sanger sequencing analysis was conducted using Chromas Lite version 2.6.6, a chromatogram viewer software. In silico variant pathogenicity prediction tools included PredictSnp (https://loschmidt.chemi.muni.cz/predictsnp/), Sift (https://sift.bii.a-star.edu.sg/), SpliceAI (https://spliceailookup.broadinstitute.org/), MaxEntScan (http://hollywood.mit.edu/burgelab/maxent/Xmaxentscan_scoreseq.html), SPiP (https://sourceforge.net/projects/splicing-prediction-pipeline/) and Franklin (https://franklin.genoox.com/). REVEL score with a threshold of ≥ 0.7 was used, where approximately 95% of benign variants are excluded at this threshold level [37] to assess pathogenic potential of segregating variants. Evolutionary conservation was evaluated using PhyloP scores obtained from the UCSC Genome Browser. Genetic variants were categorized and interpreted according to the standards and guidelines set by the American College of Medical Genetics and Genomics (ACMG) [38].

c. Data analysis and comparative evaluation

The average age of onset and its standard deviation were calculated to evaluate the variability among the individuals affected. The efficiency of the genetic diagnostic method was evaluated by comparing the success rate of this study with those published in previous research, notably the 45.95% rate reported by Gonzaga et al. (2015) and the 58% rate reported by Candayan et al. (2021). Proportional differences and Cohen’s h were calculated to quantify the effect size and assess the relative effectiveness [39, 40].

Results

A total of ten families with symptoms of HMSN were identified during the study period, but nine were recruited for the study based on the inclusion criteria, as one family had a known disease-causing variant from prior genetic testing. Ultimately, the study included 28 affected individuals (18 males and 10 females) from the nine recruited families. The age of onset for these individuals ranged from 1 to 15 years, with a mean age of onset of 5.13 years and a standard deviation of 3.61 years. A genetic diagnosis was achieved in 55.56% of the families presenting with HMSN phenotypes. In comparison, Gonzaga et al. (2015) reported a genetic diagnosis rate of 45.95%, while Candayan et al. (2021) reported a rate of 58%. The proportion difference between this study and Gonzaga et al. (2015) was 9.61%, and the difference compared to Candayan et al. (2021) was − 2.44%. Cohen’s h, calculated to assess the effect size, was approximately h ≈ 0.193 for the comparison with Gonzaga et al. (2015) and h ≈ -0.049 for the comparison with Candayan et al. (2021). This suggests that the difference in the genetic diagnosis rate between the current study and Gonzaga et al. (2015) is more significant, while the difference with Candayan et al. (2021) is negligible.

Table 1 lists the pathogenic variants identified in five families with features of HMSN, along with their ACMG classification, genomic database allele frequencies, Revel and PhyloP scores. Figure 1 presents the pedigrees of these families and demonstrates the segregation of pathogenic variants where causal genetic variants for HMSN were identified. Notably, the parents of affected individuals in Families 1, 3, 4, and 5 engaged in consanguineous marriages, while the parents in Family 2 practiced tribal endogamy, both contributing to autosomal recessive inheritance patterns. The supplementary data (Figure S1) include pedigrees of four families and a list of variants analyzed for segregation, though a genetic diagnosis could not be confirmed.

Table 1 Genetic findings in investigated families with HMSN

Clinical findings

In Family 1, the male proband (V:3) shown in Fig. 1, currently aged 5year, demonstrated normal developmental milestones initially, including walking at 16 months, without any additional features noted. However, clinical examination at the age of 2.5 years revealed frequent falls, slowly progressive weakness in his lower limbs, a high-steppage gait with foot drop, absent ankle reflexes, but normal knee jerk. By the age of 3 years and 2 months, he had progressed to an inability to walk. Nerve conduction studies (NCS) confirmed severe sensory, and motor axonal peripheral neuropathy and he was diagnosed with HMSN (Table S2). Affected male sibling (V:1) had slowly progressive distal weakness, muscle atrophy, and sensory loss and only features of hand tremor was found in older family members.

Fig. 1

Pedigrees of the Pakistani families investigated, and genetic findings. A. Pedigrees showing segregation of the variants identified in each case. B. Sequence chromatograms of the identified variants showing both wild-type and mutant homozygous alleles as well as heterozygous alleles. C. Multiple sequence alignment of different species showing evolutionary conservation of mutated amino acid residues

In Family 2, five affected individuals were identified (Fig. 1), a male proband (II:5) presented at the age of 15 years with progressive difficulties in walking with frequent falls since early childhood and normal cognitive function. Examination showed a broad-based gait, generalized reduced power (grade 4/5), and absent/reduced reflexes with down going plantar responses and normal tone and sensation. Magnetic resonance imaging (MRI) revealed mild cerebellar atrophy, characterized by a mild prominence of the cerebellar folia, with no other structural abnormalities detected at the age of 15 years. This finding aligns with the clinical presentation of a fine tremor of the upper limbs and ataxia, supporting the diagnosis of cerebellar ataxia. Other affected family members (II:3, II:4, and III:2) presented with comparable symptomatology, with individual III:3 also having a fine tremor of the upper limbs.

In Family 3, an affected individual (V:5) displayed overlapping features of HMSN with the 20-year-old male proband (V:4) in Fig. 1. These features included childhood onset of leg weakness and gait abnormality, later progressing to severe muscle wasting and atrophy in both upper and lower limbs, generalized areflexia, bilateral foot and hand drops resulting in profound mobility impairment and eventual wheelchair dependence, along with sensory loss primarily affecting the lower limbs.

In Family 4, there are 2 affected individuals (Fig. 1) exhibiting features of muscle wasting, atrophy, and bilateral foot drop. Lower limb power is reduced with distal areflexia and hypotonia, alongside down-going plantar responses, resulting in profound mobility impairment and eventual wheelchair dependence. Sensory loss primarily affects the lower limbs in affected individuals within this family. The male proband (IV:2) in Fig. 1 presented at 7 years with gait abnormality, lower limb weakness, sensation loss and was clinically diagnosed with HMSN.

Two affected brothers from Family 5 (Fig. 1) presented with lower limb weakness and walking difficulties onset at the age of 3–4 years, both accompanied by mild to moderate intellectual disability. The male proband (IV:1), aged 12, exhibits severe lower limb weakness with muscle atrophy, hypotonia, and depressed reflexes, leading to progressive loss of independent ambulation. Additionally, he had a healed neuropathic ulcer over the right shin and requires a wheelchair for mobilization and has been clinically diagnosed with peripheral neuropathy. The other affected individual (IV:4), aged 18, remains mobile independently, although he experiences lower limb weakness with muscle atrophy. His gait is characterized by a high step page, broad-based gait with foot drop, and he also shows milder upper limb weakness and wasting with areflexia.

Genetic findings

WES data analysis identified homozygous variants in 4 different HMSN-associated genes (Table 1). All variants segregated with complete penetrance in these families (Fig. 1). In Family 1, a novel homozygous splice site variant [Chr8(GRCh38): g.75,263,705 A > G NM _018972.4: c.310 + 4 A > G, p.?] in GDAP1 was identified in a site with a moderate level of evolutionary conservation (PhyloP score = 3.808), with a very low allelic frequency (MAF = 0.00000006238) allelic frequency in gnomAD (v.4.1). In silico analysis utilizing MaxEntScan, SpliceAI, and SPiP predicted this variant to be damaging or disease-causing, with scores of 0.33, 0.601, and 1(98%), respectively, indicating a high likelihood of affecting splicing with a risk of altering the splice site. The GDAP1 c.310 + 4 A > G variant is classified as a Variant of Uncertain Significance (VUS) following ACMG guidelines (Table 1).

In Family 2, a novel homozygous frameshift variant [Chr9 (GRCh38): g. 132296887-132296888del NM_015046.7: c.5948_5949del, p.(Asn1984Profs*30)] in SETX was detected was detected in a site with a moderate level of evolutionary conservation (PhyloP score = 5.828). This variant was absent in both the gnomAD and 1000 Genomes Project databases. SIFT prediction classified it as damaging with a score of 0.529, while the Franklin analysis suggested a likely pathogenic nature of the variant. Based on ACMG criteria, the variant is predicted to be “pathogenic” (Table 1).

Families 3 and 4 shared a previously reported [41] homozygous IGHMBP2 missense variant [Chr11(GRCh38): g. 68934517 C > A NM_002180.3: c.1591 C > A, p.(Pro531Thr)] within a site with a low level (PhyloP score = 2.743) of evolutionary conservation and with a low allele frequency (MAF = 0.00001736) in gnomAD database. However, the REVEL score prediction is close to 0.7 (0.678), indicating a likelihood of being causal. It’s noteworthy that according to ACMG criteria, the variant is predicted to be “pathogenic” (Table 1).

In Family 5, a previously reported [42] homozygous missense variant [Chr18(GRCh38): g. 57601666 G > A NM_004539.4:c.1633 C > T, p.(Arg545Cys)] in NARS1 was identified within a highly evolutionarily conserved site (PhyloP score = 9.491), with a low allele frequency (MAF = 0.00001363) in the gnomAD database. The REVEL score (0.814) predicted it to be a disease-causing variant, and according to ACMG criteria, the variant is predicted to be “likely pathogenic” (Table 1).

Discussion

A current challenge in interpreting the clinical relevance of genetic variants present in individuals from Pakistani populations stems from the poor representation of Pakistani individuals in population genomic databases, and despite efforts to improve global diversity in gnomAD v4, only 0.327% (2637 out of 807162) of exomes/genomes are from individuals of Pakistani origin. As a result, the absence of a variant in publicly available genomic databases is currently of limited value for indicating pathogenicity, underscoring the importance of continued efforts to diversify the genomic evidence-base [43]. Pakistan still lacks a national genomic sequencing service, but research studies involving Pakistani populations, mostly in collaboration with US, UK, and European institutions, aim to increase knowledge of genomic variation in Pakistan [44]. In this study, as part of an international collaboration, nine consanguineous/endogamous Pakistani families from Khyber Pakhtunkhwa with HMSN and related neuropathic symptoms in some cases were analyzed to determine the genetic basis of the disease. The relatively small sample size of nine families may limit the generalizability of the findings to the broader Pakistani population. Despite this limitation, the study identified novel and rare causative biallelic variants in the GDAP1, SETX, IGHMBP2, and NARS1 genes in five families, achieving a genetic diagnosis in 55.56% of the cases. This study’s genetic diagnosis rate (55.56%) is higher than the 45.95% reported by Gonzaga et al. (2015) but slightly lower than the 58% reported by Candayan et al. (2021). The effect size (Cohen’s h ≈ 0.193) suggests a significant difference in diagnosis rates between this study and Gonzaga et al. (2015), while the difference with Candayan et al. (2021) (h ≈ -0.049) is negligible. These findings emphasize the value of genetic screening in consanguineous populations and highlight the importance of further expanding genetic databases to include underrepresented populations, like those from Pakistan, for more accurate variant interpretation and diagnosis.

A novel GDAP1 variant (c.310 + 4 A > G, p.?) was found to be associated with autosomal recessive HMSN in Family 1, with the proband suspected to have CMT following NCS showing severe sensory and motor axonal neuropathy (Table S 2). Biallelic GDAP1 variants, associated with autosomal recessive HMSN, have been previously reported in families from various ethnic backgrounds, including Tunisian, Turkish, Moroccan, Amish, Spanish, Polish, Brazilian, Iranian, Bangladeshi, and Pakistani populations [34, 45,46,47,48,49,50,51]. In addition, GDAP1 variants have also been reported in families of European origin to cause autosomal dominant HMSN [45, 46]. Autosomal recessive GDAP1-HMSN is generally more severe than dominant forms, with biallelic truncating variants causing an early-onset severe phenotype, with rapid disease progression leading to the inability to walk in early childhood and wheelchair-dependency before the end of the third decade [52] Although, clinical heterogeneity is common in patients with GDAP1 variants, with significant variability in the age of onset and functional disability between affected individuals within the same Family [53, 54].

GDAP1 is a 358-amino-acid outer mitochondrial membrane protein that has been shown to be largely expressed in neurons [55, 56] and is crucial for several aspects of mitochondrial morphology and functioning in neurons [57,58,59]. Loss of function variants in GDAP1 result in decreased mitochondrial membrane potential and ATP production that may lead to mitochondrial dysfunction as well as disturbed mitochondrial dynamics. In addition, interaction of GDAP1 with transport proteins induces mitochondrial transport within the cell [60], and GDAP1 variants may disrupt mitochondrial transport leading to axonal loss in patients with HMSN [61]. GDAP1 belongs to a subfamily of glutathione-S-transferases (GSTs), containing two typical GST domains, domain I (GST-N) at the N-terminal region and domain II (GST-C) at the C-terminal region, a hydrophobic domain (HD) and a single transmembrane domain (TMD) [62]. The GDAP1 c.310 + 4 A > G splice variant is predicted to be VUS by multiple splicing prediction tools and ACMG criteria. However, the study did not include functional validation of the identified novel GDAP1 variant, another limitation of the study, which could impact the robustness of conclusions regarding its pathogenicity. This is particularly important given that seven splice site variants are listed as pathogenic in HGMD Professional, leading to exon skipping, intron retention, or activation of cryptic sites predicted to affect pre-mRNA splicing [60,61,62]. Functional studies are essential to confirm the impact of these variants on GDAP1 protein function and to establish their role in disease pathology.

A novel homozygous SETX variant p.(Asn1984Profs*30) was identified in 5 affected individuals from Family 2 diagnosed with HMSN and cerebellar ataxia, indicating cerebellar ataxia as an additional phenotype associated with HMSN in these cases. The observed phenotypes of fine tremor and ataxia in these Family members align with the mild CA detected on neuroimaging, supporting the diagnosis of cerebellar ataxia. Contrastingly, in a previous study involving siblings of a Pakistani consanguineous Family, phenotypes of progressive CA, such as nystagmus, elevated α-fetoprotein, and limb weakness in patients with peripheral neuropathy and cerebellar ataxia, were noted [63]. Additionally, another study identified CA phenotypes of mild truncal ataxia, bilateral sensorineural deafness, and learning difficulties in a Spanish patient with axonal neuropathy and cerebellar ataxia [64]. These findings collectively emphasize the importance of recognizing cerebellar atrophy phenotypes in patients with SETX-associated recessive HMSN and cerebellar ataxia, reinforcing the association of these conditions with SETX variants observed in previous studies [65, 66]. Additionally, biallelic SETX variants are associated with autosomal recessive spinocerebellar ataxia with axonal neuropathy-2 (SCAN2) also known as recessive ataxia with oculomotor apraxia type 2 (AOA2) [66,67,68], with an additional autosomal dominant phenotype of Juvenile amyotrophic lateral sclerosis- 4 (ALS4) associated with heterozygous SETX variants [67, 69]. SETX is expressed throughout the nervous system particularly within the cerebellum, spinal cord, and peripheral nerves conferring a broader spectrum of clinical features [70], although the exact mechanisms that lead to cerebellar damage and neurodegeneration are yet to be established. The SETX gene has 33 exons, encoding a large 2,677 amino acid Senataxin, a putative RNA/DNA helicase involved in both DNA and RNA processing in nerve cells [69]. Senataxin contains an N-terminal protein interaction domain, a helicase domain and a C-terminal Nuclear Localization Signal (NLS) domain [71]. The novel pathogenic SETX variant p.(Asn1984Profs*30) described here is located in the helicase domain and is predicted to undergo non-sense mediated decay (NMD) resulting in loss of senataxin function. Anheim et al. (2009) reported HMSN as the most predominant feature in 97.5% of individuals with SCAN2 with cerebellar atrophy identified on MRI neuroimaging in 96%, although cerebellar features on examination were more variable, consistent with our findings in Family 2 that have a predominant phenotype of HMSN with mild cerebellar features [65]. Interestingly, Senataxin along with another RNA/DNA helicase IGHMBP2 play a critical role in DNA damage repair pathways in nerve cells, and pathogenic variants in IGHMBP2 helicase are associated with CMT type 2 S (CMT2S) [72], suggesting similar underlying mechanisms of pathogenesis.

A homozygous missense IGHMBP2 variant, p.(Pro531Thr), was identified in two unrelated Pakistani families (Families 3 and 4) with HMSN. Both families exhibit common features typical of HMSN. However, the pronounced presence of severe muscle wasting and atrophy, affecting both upper and lower limbs in Family 3, led to the development of hand and foot drops, distinguishing it from Family 4 within the context of HMSN. Comparing the phenotypes of two Pakistani families with homozygous IGHMBP2 variant (p. Pro531Thr) to a previously reported Pakistani case of CMT2 plus Down Syndrome with compound heterozygosity of this variant reveals shared features like muscle weakness and neuropathic findings [41]. The occurrence of the IGHMBP2 variant, in both homozygous and compound heterozygous states, alongside the complexity of observed phenotypes, underlines the intricate relationship between genetic variations and phenotypic outcomes in HMSN among Pakistani individuals. Additionally, identification of the IGHMBP2 variant, p.(Pro531Thr), in both previous and present Pakistani families also suggests a likely founder effect within this population. Biallelic IGHMBP2 gene variants are associated with the neuromuscular disease spinal muscular atrophy with respiratory distress 1 (SMARD1) [73], usually caused by loss of function variants in IGHMBP2, as well as CMT2S, more commonly caused by missense or protein altering variants, as reported in Families 3 and 4. Although a specific diagnosis of CMT2S could not be confirmed with NCS as this was unavailable in local clinical testing facilities, clinical features in these patients were consistent with CMT2S [41, 74, 75] in both families confirming this diagnosis alongside our genetic findings.

The IGHMBP2 gene contains 15 exons, encoding a 993 amino acids ATP-dependent 5′→3′ helicase that facilitates effective transcription, translation as well as RNA metabolism [76]. IGHMBP2 shows high levels of expression in the cerebellar cortex and peripheral nerves [41]. The IGHMBP2 protein consists of a helicase domain, a single-stranded RNA and DNA binding R3H domain and an AN1-type zinc finger motif [77, 78]. The R3H domain is critical for function of protein as it specifically recognizes the phosphorylated 5′-ends of both single-stranded RNA and DNA [79], thus promotes RNA binding and the ATPase activity of this domain [77]. The IGHMBP2 variant, p.(Pro531Thr) is located within the helicase domain, alongside a number of other pathogenic biallelic missense IGHMBP2 variants in patients with confirmed CMT2S, suggesting disruption of helicase activity as the underlying pathomechanism of disease [80,81,82].

Affected individuals from Family 5 were identified to have a homozygous missense NARS1 p.(Arg545Cys) variant, as the cause of HMSN and intellectual disability. The NARS1 variant has previously been reported in a homozygous state in a cohort of 16 afflicted individuals originating from eight families of Pakistani and North Indian descent. These individuals manifested a spectrum of clinical characteristics encompassing microcephaly, neurodevelopmental delay, seizures, and ataxia [42, 83]. The findings of present and previous studies emphasize the crucial role of genetic variations in contributing to the diverse phenotypic spectrum observed in HMSN and related conditions within Pakistani and North Indian populations. Additionally, the discovery of the NARS1 p. (Arg545Cys) variant in another Pakistani family further confirms its potential as a likely founder variant within this population. Interestingly both de novo heterozygous and biallelic NARS1 variants have been associated with a similar clinical phenotype. NARS1 comprises 14 exons and encodes a 58 amino acid asparaginyl-tRNA synthetase (AsnRS) protein. AsnRS is a vital component of the aminoacyl-tRNA synthetase family, tasked with the essential role of catalyzing the precise attachment of asparagine to its corresponding transfer RNA (tRNA) during protein translation. This meticulous process ensures the accurate incorporation of asparagine into growing polypeptide chains, underpinning the fidelity of protein synthesis. Pathogenic variants within other cytoplasmic ARS-encoding genes have also been associated with HMSN disease, including variants in AARS1, GARS1, HARS, MARS1, SARS1, WARS and YARS1, frequently with additional phenotypic features [84].

In conclusion, this study emphasizes the need to address genomic healthcare disparities and challenges related to interpreting rare genetic variants in underrepresented populations. It highlights the utility of WES in identifying novel and rare variants associated with HMSN, including potential founder effects for certain gene variants within the Pakistani population. Our study also highlights the substantial impact of genetic variations on the varied phenotypic expressions observed in HMSN and HMSN plus type neuropathies within the Pakistani population. However, the study reveals limitations such as small sample sizes and the absence of functional validation of novel variants. Future research should focus on increasing the representation of underrepresented populations in genomic databases, exploring the functional implications of identified variants, and conducting studies with larger family cohorts. This approach will enhance the robustness and generalizability of genetic findings, ultimately improving the understanding and management of HMSN and related disorders.