Background

Spinal muscular atrophy (SMA) is an autosomal recessive hereditary disease characterized by degeneration of spinal cord motor neurons, atrophy of skeletal muscles, and generalized weakness [1]. It affects 1 in 10,000 live births, and often leads to early death [2]. SMA manifests over a wide range of severity, affecting infants through adults. According to the onset time and severity of the disease, SMA is divided into 4 types (SMA1, SMA2, SMA3 and SMA4), and SMA 1, with onset before age 6 months; SMA 2, with onset between age 6 and 18 months; SMA 3, with onset in childhood after age 12 months; and SMA 4, with adult onset [3]. Nusinersen (trade name: Spinraza) is the only approved drug to treat spinal muscular atrophy, which is administered directly to the central nervous system using an intrathecal injection [4], but it’s mandatory to have a specific diagnosis in quick time. The genetic profile is essential for the quick and accurate diagnosis of SMA.

SMA is caused by homozygous disruption of the survival motor neuron 1 (SMN1) gene by deletion, conversion, or mutation [1]. Since SMA is one of the most common lethal genetic disorders, with a carrier frequency of 1 in 40 to 1 in 60, direct carrier dosage testing has been beneficial to many families [5]. About 96% SMA are caused by a homozygous deletion of SMN1 exon7, and the remaining 4% of cases are caused either by compound heterozygosity with a point mutation in one allele and a deletion in the other or by compound heterozygous point mutations in SMN1 [6]. SMN2 is a homologous gene of SMN1 and functions as a SMA modifier. In general, the copy number of SMN2 is substantial variation in SMA patients, and a high SMN2 copy number tends to a milder type [7]. Furthermore, more and more new genes or novel mutations have been reported to be related to the morbidity, severity, treatment and prognosis of SMA with the development of gene sequencing technology. A study revealed seven different mutations of SMN1, and among them, c.824G > C, and c.825-2A > T were described for the first time [8]. Another study found the NAIP copy number was inversely correlated with the clinical severity of SMA [9]. GTF2H2 and H4F5 have been proved to be associated with the onset and type of SMA [10]. It is important to understand the genetic characteristics of SMA pathology in order to better understand and diagnose it, so as to implement the nursing care of more and more affected patients.

The severity of SMA varies from infancy to adulthood, which is reflected in the clinical classification system of type 1–4. In addition, a recent study reported that the copy number of SMN genes may determine the pathological type of SMA patients. Cao et al. [11] found that in different copy numbers of SMN2, the distribution of type I and type 2/3 was significantly different (P < 0.001): when the copy number of SMN2 in parents was 1, 75% of SMA progenies were type 1 and 25% were type 2/3. However, when the parents carried three copies of SMN2, their SMA progenies were all 2/3 type. Therefore, it is very important to understand the pathological genetic characteristics of patients, which directly affects the diagnosis and treatment of patients.

To further improve our understanding of genetic variation in patients with SMA, in this study, multiplex ligation probe amplification (MLPA) was firstly used for preliminary diagnosis in 28 core families of suspected SMA patients, and then the complete gene sequence of SMN1 gene was detected by high-throughput sequencing to find more mutations in the 28 core families. Afterwards, 20 children diagnosed with SMA and 5 children diagnosed with non SMA were enrolled, and the whole exome screening of other related genes was performed to explore more genes and mutations involved in the onset of SMA.

Methods

Patients and samples

From December 2013 to May 2017, 28 probands of clinically suspected SMA, 15 males and 13 females, were accepted by our hospital (Tianjin Children’s Hospital, Tianjin, China) because of unstable walking, and their age ranging from 1 month to 12 years. The phenotypes of their parents were normal. 3–5 ml peripheral blood samples from the 83 enrolled cases (probands and their parents) were collected. All the subjects signed the informed consent forms for genetic testing routinely, and all procedures were in accordance with the ethical standards of the institutional and/or national research committee.

MLPA

Extracted Genomic DNA with the salting out method. The nucleic acid quantitative instrument NADO DROP 2000 (Thermo Fisher Scientific Inc., Waltham, USA) was utilized to determine the quality and quantity of the extracted DNA. MLPA was performed using a SALSA MLPA Kit P021 (MRC-Holland, Amsterdam, Netherlands) according to the manufacturer’s protocol. MLPA products were run on an ABI PRISM 3130 genetic analyzer (Applied Biosystems International Inc., California, USA). The evaluation criteria were based on the kit instructions: in normal individuals, SMN1 exon7 is 2 copies; the SMN1 gene of patients with homozygous deletion and carriers with heterozygous deletion is 0 copies and 1copy, respectively [12, 13].

SMN1 gene screening by high-throughput sequencing

The complete genome sequence of SMN1 gene was detected from above 83 enrolled cases via high-throughput sequencing. Firstly, high-molecular-weight genomic DNAs were extracted with the DNeasy Blood and Tissue kit (QIAGEN, Dusseldorf, Germany) and 10 μg genomic DNA was used for library generation according to the manufacturer’s recommendations. Secondly, the Illumina HiSeq-2500 platform (Illumina, California, USA) was used to conduct the sequencing reaction based on the manufacturer’s protocol. The average sequencing depth was 100×, and more than 96% regions was up to 20×. Thirdly, the raw data were compared with human genome (NCBI build 36, hg18), and marked repeated reads and filtered out low-quality data. Fourthly, the base quality of reads was calibrated again by the genome analysis toolkit (GATK) algorithm, and ultimately all the a single base of DNA (point mutation) or a loss of base pairs (deletion) were screened out with the GATK software V3.0 (Eli and Edythe L. Broad Institute, Massachusetts, USA) [14]. The QC30 of the raw data was more than 85% in all samples, and the allele frequency of all mutations was more than 20%.

The whole exome screening by high-throughput sequencing

Based on previous results, 20 SMA patients with homozygous deletion of SMN1 exon7 were selected from the 28 probands, and 5 non SMA children were as controls (2 children carried with heterozygous deletion of SMN1 exon7 and 3 children with 2 SMN1 exon7 copies). Whole exome high-throughput sequencing was performed by The Life Technologies SOLiD™ (version 3) technology with mate-pair chemistry. 20 μg high-molecular-weight genomic DNAs was used for library generation. The raw data were analyzed with GATK software. The average sequencing depth was 100×, and more than 96% regions was up to 20×. The QC30 of the raw data was more than 85% in all samples, and the allele frequency of all mutations was more than 40%.

Results

MLPA

MLPA results showed that 22 probands were with homozygous deletion of SMN1 exon7 (SMA patients), 3 probands carried with heterozygous deletion of SMN1 exon7 (carriers), and 3 probands with 2 SMN1 copies (normal individuals). However, the mother of a SMA patient had 3 SMN1 exon7 copies and the father was carrier, and the father of another SMA patient had 3 SMN1 exon7 copies and the mother was carrier. Here, we recorded the above 2 SMA patients as Proband-1 and Proband-2, respectively. Furthermore, 1 carrier, a 2-year old girl, had some clinical features that correspond to SMA, such as atrophy of skeletal muscles, generalized weakness and extensive neurogenic injury by electromyography. Here, the carrier was named as Proband-3. Besides, the father of Proband-3 was without heterozygous deletion of SMN1 exon7 and exon8, and the mother was a carrier.

Novel mutations of SMN1 in suspected SMA family

A total of 6 single nucleotide variants (SNVs) of SMN1 were identified in the 83 samples, and they were showed in Table 1. Three SNVs located in exon, 1 in UTR5 and 2 in intron; c.[84C > T], c.[271C > T], c.[−39A > G] and g.[70240639G > C] were firstly reported here; c.[84C > T] and c.[462A > G] were synonymous mutations, and c.[271C > T] were stop-gain; c.[271C > T] caused changes in encoded amino acids. Furthermore, c.[271C > T] was found in Proband-3 and her father, and c.[462A > G] was occurred in 17 SMA patients, 2 carrier and 2 normal individuals. In addition, 1 carrier was with c.[84C > T], 1 SMA patient with c.[−39A > G], and 1 SMA patient with g.[70240639G > C].

Table 1 Mutations in SMN1
Full size table

Novel mutations occurred in only SMA patients

Compared with control group, a total of 9102 mutation were selected out in SMA patients with homozygous deletion of SMN1 exon7. They were located in the exon region, and occurred only in SMA patients not in carriers and normal individuals. Among them, 2415 genes and some indefinite genes were included, and 8619 SNVs, 267 deletions and 216 inserts were contained. Here, the indefinite genes were removed, and the 30 most frequent mutations were showed in Table 2 (Frequency ≥ 50%). It was obvious from Table 2 that only the MCCC2 missense mutation c.[1001A > G] located on chromosome 5q13, which was the same location as SMN1 and SMN2. The others were unlinked to 5q13.

Table 2 The 30 most frequent mutations in SMA patients
Full size table

Discussion

Since both the clinical features and the mutation characteristics of SMA are obvious, the diagnosis of SMA is considered as relatively easy process. However, a recent review reported that the diagnostic process for SMA is not always simple, and there is usually a delay between the onset of clinical features and the diagnosis in all types of SMA [15]. At present, MLPA is the gold standard for clinical diagnosis of SMA. However, MLPA can only detect the deletion of SMN1 according to the gene copy number, not detect point mutations of SMN1. It is well known that about 4% of SMA patients bear one SMN1 copy with an intragenic mutation. Therefore, some SMA patients are inevitably misdiagnosed as carriers. In addition, the emergence of new therapies increases the need for early diagnosis of SMA, the approval of therapies and the neonatal screening programs urgently require a more detailed understanding of genetic variation [16]. In this study, Proband-3 was with one SMN1 copy and the SMN1 stop-gain mutation c.[271C > T], and the heterozygous deletion of SMN1 exon 7 was from her mother, and the SMN1 stop-gain mutation c.[271C > T] from her father. The SMN1 stop-gain mutation c.[271C > T] was never reported before, and it led to a amino acid change. Although MLPA results showed Proband-3 to be carrier, some SMA-related clinical features were occurred on her. Here, we suspected that Proband-3 might be a SMA patient caused by the heterozygous deletion of SMN1 exon 7 combined with the SMN1 stop-gain mutation c.[271C > T]. Simultaneously, c.[271C > T] might be involved in the onset of SMA. In addition, [2 + 0] genotype carriers are two SMN1 copies on one chromosome and with deletion of SMN1 on the other chromosome [17]. In this article, we found 2 patients (Proband-1 and Proband-2) whose one parent was carriers and the other parent with 3 SMN1 exon7 copies (Proband-1’s mother and Proband-2’s father). Based on our results we suspected that Proband-1’s mother and Proband-2’s father might be [2 + 0] genotype carriers.

SMN1 and SMN2 present on chromosome 5q13, and of the 5q13-linked SMA patients, 96.4% show homozygous absence of SMN1 exons 7 and 8 or exon 7 only, whereas 3.6% present a compound heterozygosity with a subtle mutation on one chromosome and a deletion/gene conversion on the other chromosome [6]. Due to the complexity of 5q chromosome structure, the mechanism of SMA has not been fully elucidated. For example, the rare variations in SMN2 have been described by several studies [18,19,20], and some scholars believe that the variations in SMN2 locus, such as the deletion of adjacent NAIP1 gene, will affect or even change the severity of SMA [21, 22]. With the development of bioinformatics, more and more mutations have been discovered in SMN1, and some of them possess significant clinical implications. Yamamoto et al. [23] revealed 4 intragenic mutations (p.Ala2Val, p.Trp92Ser, p.Thr274TyrfsX32 and p.Tyr277Cys), and location of the mutations were associated with the clinical severity of SMA. Ronchi et al. [24] described a novel SMN1 mutation that affected the donor splice site of exon 7 and resulted in an unusually severe SMA phenotype with rapid fatal outcome in an Italian infant. In this article, we found 6 SMN1 SNVs in 28 core families of suspected SMA patients, including 4 novel mutations c.[84C > T], c.[271C > T], c.[−39A > G] and g.[70240639G > C], which had never been previously reported. Besides, we identified more mutations combined with homozygous absence of SMN1 exons7 (Table 2). The 3 most frequent mutations were the insertion mutation c.[−41_-40insCTCT] in SPTA1 exon1 (rs111674514), the SNV c.[1001A > G] in FUT5 exon2 (rs778984), and the SNV c.[−117A > G] in MCCC2 exon1 (rs11746722). SPTA1 encodes the human erythroid alpha-spectrin, which is an actin crosslinking and molecular scaffold protein that links the plasma membrane to the actin cytoskeleton, and functions in the determination of cell shape, arrangement of transmembrane proteins, and organization of organelles [25, 26]. Mutations in SPTA1 can lead to a variety of hereditary red blood cell disorders, including elliptocytosis type 2, pyropoikilocytosis, and spherocytic hemolytic anemia [27, 28]. FUT5 encodes alpha1,3-fucosyltransferase in human, the down-regulation of FUT5 reduces the expression of sialyl-Lewis antigens and the adhesion and binding capacities of gastric cancer cells [29]. Gene transfer of alpha1,3-fucosyltransferase increased tumor growth of the PC-3 human prostate cancer cell line through enhanced adhesion to prostatic stromal cells [30]. Methylcrotonyl CoA carboxylase β (MCCβ) is encoded by MCCC2, and point mutations and deletion events in MCC2 can lead to MCC deficiency [31]. MCC deficiency is a rare autosomal recessive genetic disorder whose clinical presentations range from benign to profound metabolic acidosis and death in infancy, which is has something in common with SMA in some ways. MCCC2 locates on chromosome 5q13, which was the same as SMN1. Some studies indicated that SMN1 was the causative gene, and other genes on 5q13 region acted as modifier gene (such as SMN2, NAIP and GTF2H2), which were associated with disease severity [32]. The mutations rs111674514 in SPTA1, rs778984 in FUT5 and rs11746722 in MCCC2 have been identified previously, but the clinical significance remains uncertain. In this article, we found they were widely prevalent in SMA patients, and almost nonexistent in non-patients. Therefore, it suggested they might be involved in the morbidity of SMA.

One limitation of this study is the failure to use healthy children as control. The reason is that the children involving in this study are younger (0 to 12 years old), many parents are reluctant to let their healthy children participate in the study because of insufficient understanding of genetic testing. Another objective reason is that the cost of sequencing is high, so parents of healthy children are unwilling to pay for it. However, in terms of the rationality of the entire study, we really should add healthy children as control to make our research more complete, and that is what we will improve in the follow-up related studies. In contrast, using non-SMA patients with similar clinical characteristics of SMA as a control could exclude some genetic mutations that may lead to similar clinical characteristics of children with SMA, so as to make the object of this study more accurate. Using this population as a control can exclude other mutations that may cause dyskinesia, and only retain mutations unique to SMA patients.

Conclusions

We found more mutations in both SMN1 and other genes, and some of them were associated with the onset of SMA, such as the SMN1 stop-gain mutation c.[271C > T], the SPTA1 insertion mutation c.[−41_-40insCTCT], the FUT5 SNV c.[1001A > G], and the MCCC2 SNV c.[−117A > G].