Genetic mutation analysis of hereditary spastic paraplegia: A retrospective study.

Hereditary spastic paraplegias are heterogeneous disorders with diversified clinical manifestations, and genetic testing is important for the diagnosis and typing of hereditary spastic paraplegias.Gene panel sequencing containing 55 hereditary spastic paraplegias-related genes was performed to screen the pathogenic genes for hereditary spastic paraplegias. Sanger sequencing was adopted to validate if the family member carried the same pathogenic gene as the proband.Fifteen out of 53 patients carried mutation(s) in the screened hereditary spastic paraplegias-related genes. Among the 23 identified mutations, only one mutation had been previously reported as a pathogenic mutation. In the pedigree of case 6, the proband, his mother and uncle all carried the same novel deletion mutation (c.1459delA) at SPAST gene. Based on the pedigree, the disease was inherited in an AD pattern. In the pedigree of case 53, the family disease may be in an X-linked recessive inheritance pattern. The proband (case 53) carried two novel mutations in ALT1 gene and L1CAM gene (c.2511C>A), respectively. The L1CAM gene is the causative gene for the SPG1 X-linked recessive-hereditary spastic paraplegias.Our data confirm the genetic heterogeneity of hereditary spastic paraplegias, and SPG4/SPAST were the most frequent forms. The pathogenicity of the novel mutations is worth to be further investigated.

Introduction

Hereditary spastic paraplegias (HSPs) are clinically and genetically heterogeneous disorders characterized by progressive spasticity, extremities weakness and dorsal column impairment.[ Based on the clinical manifestations, HSPs are divided into pure and complicated forms. The inheritance patterns of HSP include autosomal dominant (AD), autosomal recessive (AR), X-linked recessive (XR), and mitochondrial maternally inheritances.[ An epidemiological study has shown that the global prevalences of AD-HSP and AR-HSP are roughly equal, ranging from 0.0 to 5.5 per 100,000 individuals.[ In Western countries, however, AD-HSPs accounts for about 70% to 80% of all HSPs.[ Although XR-HSP is extremely rare, its molecular genetic mechanism is relatively clear. Up to now, two subtypes of XR-HSPs have been identified, namely SPG1 and SPG2. These two subtypes usually manifest as complicated HSPs.[

The pure HSPs are mainly transmitted in an AD pattern, while the complicated HSPs are mostly transmitted in an AR pattern.[ At present, more than 70 genetic types of HSP have been reported.[ Common subtypes of AD-HSPs include SPG4, SPG3, SPG6, SPG8, SPG9, and SPG10 subtypes, while the common subtypes of AR-HSPs are SPG5, SPG7, and SPG11.[ Many AR-HSP patients are considered as sporadic cases due to their small pedigrees or unknown family history.[ HSP-related gene mutations are usually point mutations, commonly resulting in missense mutation, nonsense mutation, stop codon mutation, and splicing mutation.

Due to the high clinical heterogeneity, HSPs lack specificity of clinical manifestations, and are difficult to be differentially diagnosed from primary lateral sclerosis (PLS) and adrenal spinal cord neuropathy (AMN). In addition, there are many overlaps in the clinical manifestations among each HSP subtypes. Therefore, genetic testing is important for the diagnosis and typing of HSP.

Our previous study has reported the clinical manifestations of 56 HSP patients treated in our hospital.[ The purpose of this study was to further report the results of genetic test of these HSP patients.

Experimental procedures

Patients

A total of 53 patients diagnosed with HSP in our previous study underwent genetic test to identify the potential pathogenic mutation for HSP.[ To this end, 2 mL of peripheral blood were collected and subjected to extract its genomic DNA using QIAmp DNA Blood Mini kit according to manufacturer's protocol. This study was approved by the ethics committees of the Chinese PLA General Hospital. Written informed consent was obtained from each patient.

Exon sequencing for HSP-related genes

To screen the potential pathogenic genes for HSP, gene panel sequencing containing 55 HSP-related genes (GenCap Enrichment technologies, MyGenostics Inc, Beijing, China) was performed using HiSeq 2000 high-throughput sequencing system (Illumina Inc, USA). In addition, the following four genes were also included in the gene panel because they account for the inherited diseases with clinical manifestations similar to HSP: ABCD1 (adrenoleukodystrophy-related gene), MARS2 (hereditary spastic ataxias-related gene), ACADVL (very long-chain acyl-CoA dehydrogenase deficiency-related gene), and ETFDH (multiple acyl-CoA-dehydrogenase deficiency-related genes).

Bioinformatics analysis

After filtering the contaminant sequences using the Solexa QA package and cutadapt (https://cutadapt.readthedocs.io/en/stable/), the sequences were aligned to the reference genome by SOAPaligner (http://soap.genomics.org.cn/index.html) to obtain a unique alignment sequence. PCR duplicates were removed by Picard software (http://broadinstitute.github.io/picard/).

First, single nucleotide polymorphisms (SNPs) were identified using the SOAPsnp program (http://soap.genomics.org.cn/soapsnp.html), and were searched in 1000 Genomes Project database (www.1000genomes.org), followed by the Human Genome Mutation Database (http://www.biobase-international.com/product/hgmd HGMD) and the dbSNP database (http://www.ncbi.nlm.nih.gov/projects/SNP) to determine whether the variant has been reported as a pathogenic mutation.

Sanger sequencing validation of family members

After the candidate pathogenic genes mutations were obtained in the exon sequencing, Sanger sequencing was adopted to validate if the family member(s) of the patient carried the same candidate pathogenic gene mutations as the proband according to the standard protocols.

Results

Genetic test

Of the 53 patients underwent genetic testing, 9 cases had a family history of HSP, while 44 cases were sporadic patients. The gene panel testing showed that 15 out of 53 patients carried at least one mutation in the screened HSP-related genes, including 23 mutations (Table 1). According to the inheritance patterns of the mutated genes, there were 9 cases with mutated genes in AD inheritance pattern, including four SPG4 subtypes (case 4, 6, 23, 26 in Table 1), two SPG31 subtypes (case 9, 13), 1 case of SPG3 (case 53), SPG6 (case 20), and SPG10 (case 7), respectively. There were 6 cases with mutated genes in AR inheritance, including 3 cases of SPG11 (case 27, 30, 49), 1 case of SPG15 (case 39), SPG58 (case 14), and SPG69 (case 25), respectively. In addition, case 53 also carried a mutation in L1CAM gene, which is the causative gene for SPG1 HSP (XR inheritance pattern).[

Table 1

HSP patients found to carry mutations in HSP-related genes panel screening.

Among the 15 cases with mutation, there were 12 sporadic patients (case 4, 7, 9, 13, 14, 23, 25, 26, 27, 30, 39, 49) and 3 patients with a family history (case 6, 20, 53). The pedigree trees of case 6, 20, 53 were shown in Figure 1.

Figure 1

Pedigrees of three cases. The arrow marks the proband. (A) In the pedigree of case 6, the proband ((III1), his mother (II1) and uncle (II3) all carried the same novel deletion mutation (c.1459delA) at SPAST gene. (B) In the pedigree of case 20, both the proband (III7) and his brother (III1) carried a missense mutation at NIPA1 (c.8C>T). (C) In the pedigree of case 53, the family disease may be in an X-linked recessive (XR) inheritance pattern. The proband (case 53) carried two novel mutations in ALT1 gene and L1CAM gene (c.2511C>A), respectively.

Bioinformatic analysis

Bioinformatic analysis revealed that among the 23 identified mutations, one mutation at SPG11 gene (case no. 27) has been reported as a pathogenic mutation. The other cases of variants have not been reported as a pathogenic mutation in the literature, including one case of start codon mutation (case no. 13), frameshift mutation (case no. 26), nonsense mutation (case no. 30), splicing mutation (case no. 25), as well as 12 cases of missense mutations.

Among the 15 patients with mutation gene, Sanger sequencing was performed in the family members in 7 patients (case 4, 6, 7, 14, 20, 30, 39) to validate if their relatives carried the same mutation as the proband. The mother and uncle of case 6 both carried the same mutation (c.1459delA) in SPAST gene as the proband. The father of case 7 carried the same mutation (c.1816C>T) in KIF5A gene. The mother of the case 14 carried the same KIF1C gene mutation (c.2819G>A) as the proband. The brother of the case 20 patient carried the same mutation in NIPA1 gene (c.8C>T). The mothers of case 30 and case 39 also carried the same mutations in the SPG11 gene (c.6517C>T, Fig. 2) and ZFYVE26 gene (c.4529C>T) as the corresponding proband.

Figure 2

Sequence analysis of SPG11 detected a heterozygous mutation, c.6517C>T (circle).

Exclusion diagnosis of HSP

The genetic test data showed that there were 3 patients had no mutations in the HSP-related genes but had a point mutation in ABCD1 gene (case 1, 3, 36). The 3 cases were male, onset at 18 to 32 years old, without a family history of HSP. The mutation of case 3 was a frameshift mutation (c.1415_1416del), while the mutation of the case 36 was a nonsense mutation (c.1715C>A), both of which have been reported as a disease-causing mutation of adrenoleukodystrophy (ALD)–adrenomyeloneuropathy (AMN) in HGMD database. Thus, these two cases could be diagnosed with ALD–AMN. The mutation (c1166G>A) of case 1 was a missense mutation, which has also been reported as a pathogenic mutation. In addition, the patient had long-term fatigue and sexual dysfunction, therefore could be diagnosed with ALD.

Discussion

In this study, 28.3% (15/53) patients were found to carry mutation(s) in HSP-related genes. The results showed that the detected mutation rate was higher in patients with a family history than in the sporadic patients (33.3% [3/9] vs 27.3% [12/44]), which is consistent with a Norwegian epidemiological study (49.2% [32/65] vs 14.7% [5/34]).[ This phenomenon is due to the fact that the diagnosis of sporadic HSP is an exclusive diagnosis, therefore misdiagnosis in sporadic patients would decrease the rate of mutation detection in the following genetic testing. Thereby establishing a standardized diagnostic method for HSP is necessary for improving the accuracy of HSP diagnosis.

In this study, 3 of the 9 cases with positive family history were detected with HSP-related gene mutation (case 6, 20, 53). In the pedigree of case 6 (Fig. 1A), there are 3 patients in two consecutive generations, and all members of the first generation had already passed away. It can be seen from the pedigree that the disease was inherited in an AD pattern. The proband (III1, case 6) was found to carry a novel deletion mutation in SPAST gene (c.1459delA) which is predicted to induce a frameshift mutation. It is known that the mutation of SPAST gene is transited in an AD pattern. Moreover, Sanger sequencing confirmed that the family members II1 and II3 with similar manifestations to the proband also carried the same mutation in SPAST gene. Taken together, the frameshift mutation at the SPAST gene was likely to be the pathogenic mutation and may result in SPG4 AD-HSP. Based on patient's recall, the first generation of the pedigree had no symptoms. Nevertheless, according to the pedigree analysis, it is speculated that the first generation might have the same mutation but it manifested as incomplete penetrance. However, following protein function analysis is needed to validate if the mutation is the pathogenic mutation.

In the pedigree of case 20 (Fig. 1B), there are 4 affected persons in two consecutive generations. The parents of the proband (II1 and II2) had 6 children, and three of them were affected, with a prevalence of 50%. Because there were only two consecutive generations of patients, map, the inheritance pattern cannot be determined according to the pedigree. Both the proband (III7) and his brother (III1) carried a missense mutation at NIPA1 (c.8C>T). NIPA1 gene is the causative gene for SPG6 HSP,[ a rare type of AD-HSP. Currently, the pathogenic mechanism of SPG6 HSP remains unknown. SPG6 HSP can onset at any age, and progressed slowly. Most patients with SPG6 HSP manifests as pure HSP.[ However, it is not yet confirmed if the mutation (c.8C>T) is the causative mutation. The Sanger validation results of the family members III2, III3, III5, and III6 are needed for further assessment.

As for the pedigree of case 53 (Fig. 1C), there were only 3 patients (III3, III4, III6) in the third generation, and all of them were males. The mothers (II4, II6, II10) of the 3 patients were sisters. Based on the pedigree, therefore, it is speculated that the disease may be inherited in an XR pattern. Genetic testing showed that the proband (III6) carried two novel mutations in ALT1 gene (c.1408T>G) and L1CAM gene (c.2511C>A). The L1CAM gene is located on the X chromosome and is responsible for the SPG1 XR-HSP,[ while the ALT1 gene is associated with the SPG3 AD-HSP.[ There were no affected persons in the first and second generations, therefore the familial disease is less likely to be caused by ALT1 mutation. However, Sanger sequencing should be conducted in other family members for further validation.

In this study, 6 cases with a family history had no HSP-related gene mutation detected in the genetic testing. One possible explanation is that there are still some HSP-related genes which have not been found. In order to screen out the causative gene, whole exome or whole genome sequencing should be further performed. Another explanation is that some patients may have other hereditary disease but were misdiagnosed as HSP. For example, one of the six cases had onset after 30 years old and a short course of the disease. The clinical manifestations were simple spasmodic paraplegia of both lower extremities. Therefore, AD-PLS still could not be ruled out for this patient although it is extremely rare.

Among the 44 sporadic cases, 9 cases were detected with HSP-related gene mutations, of cases 27 have been reported as a pathogenic mutation.[ Sporadic cases may be due to incomplete penetrance, unknown family history, AR/XR inheritance pattern, and small pedigree. Generally, single gene mutation in sporadic patients is difficult to be proven as a pathogenic mutation, especially the missense mutations; it is difficult to predict whether its protein product is functional. The previous study showed that about 10% to 20% of sporadic HSP patients are found to carry the mutation at SPAST gene (SPG4 subtype).[ In the 5 non-sporadic patients with AD inheritance pattern, only one case (1/5 = 20%) was found to carry SPAST mutation. As for the 44 sporadic patients, there were only 3 cases with a mutation at SPAST gene (3/44 = 6.8%). The mutation rate was far lower in our study as compared with a previous report.[ The discrepancy might be attributed to the small sample size of this study or there may be geographic variation in the mutation rate of the SPAST gene.

REEP1 gene mutation is responsible for the SPG31 HSP. It has been suggested that the pathogenic mechanism of SPG31 HSP is haploinsufficiency of the REEP1 gene.[ SPG31 accounts for about 2% to 8% of AD-HSP, and the SPG31 patients usually have onset before 20 years old. The majority of SPG31 are manifested as pure HSP, and some SPG31 patients with complicated HSP have white matter signal abnormalities.[ In this study, 2 patients were found to carry REEP1 gene mutation. One of the mutations occurred at the start codon (case 13), which may be more likely to be a pathogenic mutation. The other case is a missense mutation (case 9), and its pathogenicity is still unclear and further protein function prediction is needed.

KIF5A is the causative gene for SPG10 HSP,[ and the pathogenic mutations in KIF5A are mainly missense mutation.[ SPG10 accounts for 2% to 10% of AD-HSP and sporadic patients, the age of onset is often >35 years. Clinically, SPG10 patients usually manifest as pure HSP.[ One of our patient (case 7) carried a missense mutation (c.1816C>T) at KIF5A gene, but which has not been reported as a pathogenic mutation. Further protein functional prediction is necessary. In this study, three patients had homozygous (case 27) or compound heterozygous (case 30, 49) mutations in the SPG11 gene. The homozygous mutation (c.4306C>T, nonsense mutation) in the case 27 has been reported as a pathogenic mutation.[ However, the pathogenicity of the other two cases are unclear and needed to be further investigated. SPG11 gene encodes the spatacsin protein.[ SPG11 subtype accounts for about 20% of AD-HSP, and the majority of HSP with thinning corpus callosum (TCC) are associated with SPG11 gene mutations.[ HSP-TCC patients usually have mental retardation, and other common manifestations include sensor and motor axonal damage, cerebellar symptoms, extrapyramidal symptoms, and retinitis pigmentosa. The symptoms of SPG11 are typically more severe as compared with other subtypes. On average, two-thirds of patients become wheelchair-bound 15 years after disease onset.[

In summary, we reported the genetic testing results of our 53 HSP patients. Except for the three known pathogenic mutations (case 4, 12, 27), the pathogenicity of other novel mutations in HSP-related is worth to be further investigated in the following study, especially the c.1459delA mutation in SPAST gene (case 6) and the c.2511C>A mutation at L1CAM gene (case 53). Our report is helpful for the understanding of mutations in the HSP-related genes.

Author contributions

Conceptualization: XuSheng Huang, Fang Cui.

Data curation: Fang Cui, Mao Li, JianYong Li.

Formal analysis: JianYong Li, LiuQing Sun.

Funding acquisition: XuSheng Huang.

Investigation: LiuQing Sun.

Resources: Jie Qiao.

Software: Bo Sun.

Visualization: SiYu Chen.

Writing – original draft: Fang Cui.

Writing – review & editing: Fang Cui, XuSheng Huang.