Hereditary spastic paraplegia (HSP) is a group of clinically and genetically heterogeneous neurodegenerative diseases with a global incidence of 4.3-9.8 per 100,000 individuals (1). Four inheritance patterns have been identified: i) Autosomal-dominant (AD), ii) autosomal-recessive, iii) X-linked recessive and iv) mitochondrial; additionally, de novo mutations have been found in a number of patients with HSP (1). Thus far, >70 related pathogenic genes have been identified in patients with HSP (1). According to the clinical phenotype, HSP can be divided into pure and complex forms. The pure form is characterized by progressive spasticity and weakness of the lower limbs, with occasional sensory impairment or bladder dysfunction; the complex form includes the pure phenotype with additional abnormalities, including mental and cognitive changes, optic atrophy, muscular atrophy, ataxia, deafness, ichthyosis and/or peripheral neuropathy (2).

Spastic paraplegia type 4 (SPG4) is the most common subtype of AD-SPG; it is present in ~40% of patients with AD-SPG. SPG4 is caused by mutations in the spastin (SPAST) gene and usually manifests as pure HSP (3). Mutations in SPAST are the most common causes of familial and sporadic HSP. In China, the frequency of SPAST mutations is 40% in patients with familial HSP and 33.33% in patients with sporadic HSP (2). The SPAST gene is located on chromosome 2p22.3 and contains 17 exons; it encodes the Spastin protein, a member of the ATPase associated with diverse cellular activities (AAA) protein family. Spastin comprises four functional domains: i) Hydrophobic domain (HD; amino acid residues 1–86); ii) microtubule interacting and trafficking domain (MIT; amino acid residues 116–194), which is involved in cytokinesis and endosomal-tubule recycling (4); iii) microtubule-binding domain (MTBD; amino acid residues 270–328), which enables Spastin to bind to microtubules; and iv) AAA domain (amino acid residues 342–599), which is required for Spastin hexamerization and microtubule-severing activity (5). There are two translation initiation codons in SPAST, which encode the full-length isoform M1 (expressed in the spinal cord) and the slightly shorter isoform M87 (expressed in the spinal cord and cerebral cortex) (5). HD is present only in the M1 Spastin isoform; MIT, MTBD and AAA are present in both Spastin isoforms (4).

Spastin is an ATPase that severs microtubules. During the severing process, the six Spastin subunits assemble into a ring-shaped hexamer that is attached to the microtubule and energy from adenosine triphosphate (ATP) hydrolysis is used to sever the microtubule by pulling the negatively charged C-terminus of tubulin through the central pore of the hexamer (4). In neurons, this process can convert long microtubules into short microtubules that move rapidly and harmoniously within axons, thereby enabling efficient microtubule transport that is crucial to axonal growth and axon branch formation; microtubule transport also maintains neurite complexity (5). Mutations in SPAST are presumed to cause partial loss of Spastin microtubule-severing activity or the production of a neurotoxic mutant protein; these changes contribute to the onset of SPG4 (5). Additionally, SPG4 penetrance may be influenced by biological sex and age (6).

In the present study, the clinical characteristics of affected individuals and sequencing analysis of a mutation that caused SPG4 only among male members of a family are reported.

The proband was a 30-year-old male who was admitted in November 2021 to The Affiliated Hospital of Jining Medical University with a >10-year history of unstable gait; the patient first began to experience difficulty walking because of tremors in the lower limbs. Subsequently, weakness and tremors in the lower limbs during walking gradually worsened, and the patient could not bend his knees; however, he could stand and walk slowly unaided at the time of admission. His urinary and bowel functions were normal. Neurological examination revealed increased muscle tone in both lower limbs, muscle strength grade 3–4 (normal muscle strength is grade 5), tendon hyperreflexia and bilateral Babinski sign positive; lower limb sensation was normal. Examination of cranial nerves and upper limbs revealed no abnormalities. General physical examination, routine laboratory tests and magnetic resonance imaging of the brain, cervical spine and thoracic spine did not reveal any obvious pathognomonic alterations (Fig. 1). The patient had been born through full-term vaginal delivery, and he had exhibited good health from birth until the onset of the present symptoms.

The proband's deceased paternal grandfather had exhibited difficulty walking (Fig. 2); his age at death was not provided by the family. The proband's 66-year-old uncle (II-3) had begun to experience difficulty walking at the age of 36 and he began walking with crutches at the age of 56 (Video SI). The proband's 61-year-old father (II-4) had begun to experience difficulty walking at the age of 41 and is currently confined to a wheelchair. The proband's youngest uncle (II-5) died of diabetes at the age of 54; he could not walk before death, although his age at the onset of walking difficulty was unknown. The proband's 33-year-old brother (III-5) exhibited muscle weakness at the age of 15 and has been confined to a wheelchair since the age of 29. The urinary and bowel functions of these living relatives were normal. Moreover, neurological examinations of these relatives revealed increased muscle tone, reduced muscle strength, tendon hyperreflexia, positive bilateral Babinski sign and sensory deficits in both lower limbs. Muscle strength was normal in the upper limbs; examinations of the cranial nerves and upper limbs revealed no abnormalities. General physical examinations of these relatives also did not reveal any abnormalities. The proband's 72-year-old aunt (II-1), 70-year-old aunt (II-2), 51-year-old aunt (II-6) and 27-year-old sister (III-7) did not exhibit signs of difficulty walking or muscle weakness; general physical and neurological examinations revealed no abnormalities in these relatives. Because no equipment for MRI was available at the local hospital, magnetic resonance imaging scans of family members other than the proband could not be obtained. Patient information and clinical features of affected individuals in this family are summarized in Table I.

WES identified the missense mutation in the SPAST gene: c.1785C>A as a likely pathogenic variant according to ACMG guidelines (11). Sanger sequencing analysis of the proband, his father (II-4), uncle (II-3), third aunt (II-6), brother (III-5) and sister (III-7) revealed that all tested male family members carried the same heterozygous missense mutation in the SPAST gene: c.1785C>A in exon 17; genomic coordinates, Chr2:32379499 (Fig. 3). This mutation resulted in a serine to arginine substitution at position 595 (Ser595Arg) in the AAA ATPase cassette. However, the missense mutation was not present in the proband's aunt and sister.

According to American College of Medical Genetics and Genomics criteria, this variant was considered likely pathogenic: PM1 (located in mutation hotspots and/or in critical functional areas where no benign variation is known), PM2 (no variation or very low frequency in recessive mode was found in normal control populations), PP3 (variants that would have a deleterious effect on the function of the gene or gene product) and PP4 (the clinical phenotype or family history of the mutation carrier is highly consistent with the characteristics of a certain monogenic genetic disease). SIFT and PolyPhen2 analyses indicated that the Ser595Arg mutation is likely to be ‘deleterious’ and ‘probably damaging,’ respectively. MutationTaster software identified the Ser595Arg mutation as a ‘disease-causing’ mutation.

The RNAfold web server was used to predict the effects of the SPAST mutation on mRNA structure; the results revealed that it could lead to changes in mRNA structure, as shown by the increase of hairpin structure at 1564–1904 ribonucleic acid site (Fig. 4A and B). PSIPRED (V4.0) was used to predict and compare the mutant and wild-type Spastin secondary protein structure; the c.1785C>A missense mutation was predicted to cause changes in Spastin amino acid polarity, from neutral to positively charged (Fig. 4C and D), and in the secondary structure, which was manifested as the strand at amino acids 252–253 changed to coil, the helix shortened at amino acids 331–337, the helix shortened at amino acids 436–441, and the helix lengthened at amino acids 443–447 (Fig. 4E and F). RaptorX was used to predict the mutated Spastin tertiary structure, which revealed the spatial structure of the HD, MIT and MTBD domains was reversed compared with wild-type Spastin (Fig. 4G and H). Protein sequence alignment revealed that the Spastin amino acid sequence is highly conserved across species at amino acid 595 (red frame in Fig. 4I). Therefore, we hypothesized that the mutation causes a change in tertiary structure that affects protein function.

A summary of all mutations occurring in the four domains of spastin and the various pathogenic mutations (according to the clinical significance of mutations shown in the ClinVar and gnomAD databases) affecting the four domains of Spastin is provided in Table II. Among these pathogenic mutations, frameshift mutations were most common, followed by missense mutations, nonsense mutations were least common. Approximately 84% (78/93) of the frameshift mutations were pathogenic; most occurred in the AAA domain (47/78), followed by the MIT domain (13/78) and MTBD (13/78), and then the HD (5/78). Among all missense mutations, ~30% (65/208) were considered pathogenic; most missense mutations occurred in the AAA domain (63/65), and two occurred in the MTBD, whereas no pathogenic missense mutations occurred in the MIT domain or the HD. Similar to frameshift mutations, ~65% (32/49) nonsense mutations were pathogenic, and most occurred in the AAA domain.

Pathogenic mutations in exons that encode AAA functional structure and splicing mutations in introns that may affect AAA were identified as shown in Table III. Further analysis revealed that point mutations (specifically missense mutations) were most common overall (Table III). Base deletions were also common. Analysis of exons 7–17 encoding the AAA domain did not reveal obvious differences in the distribution of pathogenic mutation types among exons or differences in the locations of splice site mutations.

All affected family members in this report were <40 years old at the time of onset and presented with clinical manifestations of spastic paraplegia, including lower limb spasticity and weakness, increased muscle tone in both lower limbs, varying degrees of reduction in muscle strength, tendon hyperreflexia and bilateral Babinski sign positive and had a family history suggesting autosomal dominant inheritance; furthermore, all affected family members carried the c.1785C>A missense mutation in the SPAST gene, met the diagnostic criteria of HSP-SPG4 (13). Although all affected family members showed progressive spasticity and weakness of the lower limbs, they did not exhibit mental, cognitive or other abnormalities that have been associated with complex HSP (2). The clinical manifestations of the patients in the present belong to simple HSP. However, this diagnosis is limited by the lack of MRI results for other family members. There is no cure for the disease, and only symptomatic treatment is available, such as intrathecal baclofen administration, botulinum toxin injections and functional electrical stimulation (13). Medication and surgical treatment were not provided to the proband or his affected family members.

The c.1785C>A missense mutation occurs in the AAA domain, which is involved in Spastin hexamerization, microtubule-severing activity and ATP hydrolysis (4). This mutation leads to changes in mRNA structure, causes a p.Ser595Arg amino acid change in the AAA ATPase cassette that may alter amino acid polarity and modify both the secondary and tertiary structures of the Spastin protein. The mutation is suspected to influence normal protein function, resulting in reduced microtubule-severing activity, which affects axon growth and development, ultimately leading to disease. Additionally, some missense mutations in the M1 isoform can reduce the activity of wild-type Spastin in a dominant-negative manner, thereby exacerbating the disease (4). The results of recent studies have suggested that the neurotoxic effects of mutant Spastin (mainly the M1 isoform) contribute to disease onset (14–16). Therefore, the neurotoxicity of this mutant protein is presumed to serve a role in the pathogenesis experienced by the proband's affected family members.

SPG4 is an autosomal dominant disease (1); in the present study, it was noted that all male members of this family carried the pathogenic gene and exhibited varying degrees of disease presentation. Among the female family members who provided samples for sequencing, none carried the SPAST mutation. There is a need to investigate whether the high heritability of the pathogenic gene among men in this family is a chance event. Moreover, the family members differed in terms of age at onset, disease progression and symptom severity, indicating substantial intrafamilial variability.

The present study determined that the proportion of pathogenic mutations was greatest in the AAA domain, probably because the long sequence of nucleotides encoding the AAA domain carries a greater risk of mutation. The high rates of pathogenicity for frameshift and nonsense mutations may be related to the early termination of translation caused by frameshift and nonsense mutations, which seriously affect protein structure and function, thus leading to the onset of disease. Although missense mutations occur most frequently, pathogenic missense mutations comprise ~30% of all missense mutations. These findings suggested that the reduced microtubule-severing activity or loss of AAA domain function is a critical component of SPG4 pathogenesis. No significant difference was identified in the distribution of pathogenic mutation types among exons or differences in the locations of splice site mutations. However, few pathogenic mutations occurred in exon 16, although this finding requires confirmation in future studies.

Approximately 75% of HSP-SPAST cases are inherited, and the remaining 25% of cases involve de novo mutations; patients with SPG4 mainly receive symptomatic treatment because no cure is available (17). Therefore, genetic counseling is essential for affected families. When progressive walking difficulties, lower limb spasticity and other symptoms, such as tendon hyperreflexia, occur in patients without a family history, HSP should also be suspected and genetic screening should be conducted. In the present report, the clinical characteristics and sequencing analysis results of a family with SPG4 were reported. The identification of a novel SPAST mutation expands the spectrum of pathogenic mutations that cause SPG4; it also provides information for use in genetic counseling. Furthermore, specific pathogenic mutations in the SPAST gene were analyzed and the findings suggested that the loss of AAA domain function is a crucial component of SPG4 pathogenesis.