Introduction

Spinal muscular atrophy (SMA) is a common autosomal recessive neurodegenerative disorder with a prevalence of ∼1/10 000; it primarily presents with progressive muscular weakness and atrophy resulting from degeneration of the anterior horn cells in the spinal cord. Childhood onset SMA are usually classified into three types: SMA type I (MIM no. 253300), SMA type II (MIM no. 253550) and SMA type III (MIM no. 253400), by increasing age at onset and decreasing clinical severity.

The survival motor neuron 1 (SMN1) is the primary disease-causing gene. Over 90% of patients with SMA are homozygous for a deletion of SMN1, whereas 2–5% of patients are compound heterozygotes for an SMN1 deletion and an SMN1 intragenic mutation.1 The SMN2 gene, a nearly identical copy to SMN1, is considered a phenotypic modifier of SMA. SMN2 copy number is inversely correlated with disease severity.2, 3 However, this correlation is not absolute, and other genetic factors also appear to modify the SMA phenotype.

Plastin 3 (PLS3) has been identified as a gender-specific SMA modifier, first reported by Oprea et al.4 It was found that PLS3 was expressed at higher levels in asymptomatic female subjects with homozygous deletions in SMN1 when compared with their SMA-affected male siblings. Moreover, a study involving patients with SMA showed that PLS3 may ameliorate SMA phenotype in older postpubertal female patients.5

We analyzed PLS3 expression in 65 children with SMA and 59 healthy controls to determine whether the protective effect of PLS3 was also fulfilled in our cases.

Materials and methods

Subjects

In this study, 65 subjects with SMA and 59 healthy volunteers without neuromuscular symptoms were enrolled in the Department of Medical Genetics in the Capital Institute of Pediatrics (Beijing, China). All subjects with SMA were diagnosed by molecular analysis based on homozygous exon 7 deletion in the SMN1 gene. Informed consent was obtained from the minors’ parents and the study protocol was approved by the Ethical Committee of the Capital Institute of Pediatrics.

DNA extraction and SMN copy number

Genomic DNA was isolated from whole blood leukocytes by using the proteinase K phenol/chloroform method. The SMN1 and SMN2 copy numbers were determined by multiplex ligation-dependent probe amplification (MRC-Holland, Amsterdam, Netherlands), according to the manufacturer’s instruction.

RNA extraction and PLS3 expression level

Total RNA was extracted from peripheral blood by using RNA simple Total RNA Kit (Tiangen, Beijing, China). Total RNA (500 ng) was subjected to reverse transcription for 60 min at 37 °C using 3.75 mM of random primers (Sangon, Shanghai, China) and 2U of M-MLV reverse transcriptase (Invitrogen, Carlsbad, CA, USA) in a 40 μl volume. We performed real-time PCR to evaluate the expression level of PLS3. PLS3 cDNA was amplified using the primer pair: 5′-TGGCTACCACTCAGATTTCCAA-3′ and 5′-GAATCCGTTGCTGTTGAGATCA-3′. For PPIB (peptidylprolyl isomerase b) cDNA, an endogenous reference primer pair 5′-ATGATCCAGGGCGGAGACTT-3′ and 5′-GGAAGCGCTCACCGTAGATG-3′ was used. Subsequently, 10-fold dilutions of purified PCR products from a control fibroblast cell line were used to construct an external standard curve for PLS3 and PPIB. The relative standard curve method was used to quantify PLS3 expression. The 25 μl real-time PCR reaction included 1 × SYBR Green PCR mixture (Kangwei, Beijing, China), 5 μl of PLS3 cDNA and 0.25 pmol μl−1 of each primer. Real-time PCR conditions were 95 °C for 1 min as a denaturation step, and 40 cycles of 95 °C for 15 s and 64 °C for 1 min. The quantities of PLS3 and PPIB were determined by the appropriate standard curve. Each sample was run in triplicate.

Statistical analysis

The levels of PLS3 expression in different groups were compared using Student’s t-test or ANOVA when appropriate. Statistical analyses were performed by the software package SPSS 19.0 for Windows (SPSS Inc., Chicago, IL, USA). A value of P<0.05 was considered to be statistically significant.

Results

To determine whether PLS3 expression was associated with SMA phenotypes in children, we examined the level of PLS3 in 59 healthy controls and 65 subjects with SMA types I, II and III. The characteristics of subjects were summarized in Table 1. Consistent with findings of the previous study, expression of PLS3 was very low in peripheral blood.4

Table 1 Characteristics of subjects and PLS3 expression
Full size table

Interestingly, we observed a relationship between age and PLS3 expression in controls. Stratification by age showed that younger children (⩽3 years) had >1.75-fold higher levels of PLS3 expression than did older child cohorts. However, PLS3 expression was comparable among the groups of preschool-aged children (∼3–6 years), school-aged children (∼6–12 years) and adolescents (>12 years) (Figure 1). It was surmised that children showed higher expression at earlier stages of development, which then decreased with age and gradually stabilized. In addition, the level of PLS3 was 2-fold and 1.7-fold higher in SMA groups of preschool-aged children (∼3–6 years) and school-aged children (∼6–12 years) than in controls, respectively (P=0.001 and P=0.01, respectively) (Table 2).

Figure 1
figure 1

PLS3 expression in controls. The younger children (⩽3 years) had higher levels of PLS3 expression than did in ∼3–6-year group (P=0.000), ∼6–12-year group (P=0.001) and >12 year (P=0.000), respectively. Date are plotted as (mean±s.d.) × 1000, and the number of each group (n) is shown within each bar.

Full size image
Table 2 PLS3 expression in controls and SMA patients
Full size table

In the absence of age or gender stratification, no differences in expression of PLS3 were observed among SMA clinical types (P>0.05). A cutoff of 3 years was chosen to avoid bias because of the amount of PLS3 expression, which significantly decreased after this age. When stratified by age (⩽3 vs >3 years) and gender, differences in PLS3 expression by SMA type were observed in female subjects alone. However, male subjects in both age groups showed no difference in PLS3 expression when examined by SMA type (P>0.05). In older female subjects (>3 years), PLS3 expression was 55.6 and 56.7% lower in age-matched controls (mean age±s.d., year: 6.24±2.31) and type II patients (5.82±0.77) than that in type III subjects (7.17±2.80), respectively (Figure 2; P=0.001 and P=0.011, respectively), with comparable age distribution (P=0.338). Moreover, in older female patients with type III, 10 of them could walk unaided but 3 could not. Although no correlation was found between PLS3 expression and phenotype in older female SMA type III patients (P=0.261), there was a trend that PLS3 expression of patients who could walk independently (mean±s.d., 3.75±1.5) was higher than those of patients who could not (2.68±0.4).

Figure 2
figure 2

PLS3 expression in female controls and patients. In older female subjects (>3 year), PLS3 expression was higher in type III patients than that in type II patients (P=0.011) and controls (P=0.001), respectively. In older female controls, the age distribution were from >3 year to ⩽10 year in order to match the age of the control group and SMA groups. Date are plotted as (mean±s.d.) × 1000, and the number of each group (n) is shown within each bar.

Full size image

Among the 65 subjects with a SMN1 homozygous deletion, the distributions of SMN2 copy number were as follows: 50 subjects with three copies (76.9%), 14 subjects with two copies (21.5%) and 1 subject with four copies (1.5%). No difference was observed in PLS3 expression divided by SMN2 copy number (Table 1). To eliminate the modification of SMN2 to SMA phenotypes, we analyzed the PLS3 expression in the subjects carrying the same SMN2 copy number. In the 22 female subjects with three copies of SMN2, PLS3 expression was 62.6% lower in the type II subjects than in the type III subjects (Figure 3; P=0.008). However, in the six female patients with two copies of SMN2, five were type I and only 1 was type II. Thus, stratification analysis could not be applied. Moreover, this difference in PLS3 expression was not observed in male patients with three or two copies of SMN2.

Figure 3
figure 3

PLS3 expression in SMA patients with three SMN2 copies. In female patients with three SMN2 copies, PLS3 expression was lower in the type II than in the type III (P=0.008). The age distribution (mean±s.d., year) of the female groups were comparable (4.54±1.87 in type II, 6.47±3.00 in type III, P=0.125). Date are plotted as (mean±s.d.) × 1000, and the number of each group (n) is shown within each bar.

Full size image

Discussion

SMA is a debilitating fatal neuromuscular disease in children, characterized by the degeneration of the anterior horn cells of the spinal cord due to reduced expression of the SMN protein. PLS3 is an isoform of a family of actin-binding proteins involved in the stability and organization of F-actin filaments. Two studies in humans have also shown a correlation between the amount of PLS3 and SMA severity in female subjects.4, 5 Further studies in SMA cell and zebrafish models showed that high PLS3 levels can increase the amount of F-actin in motor neurons, resulting in improved axon outgrowth.4 Therefore, PLS3 should be considered as a protective SMA modifier.

To determine whether PLS3 caused a phenotype modification in children with SMA, we analyzed the level of PLS3 expression in peripheral blood among 65 subjects with SMA type I, II and III and 59 healthy controls. We found that PLS3 was expressed at very low levels in blood cells, which was consistent with other studies.4, 5 However, PLS3 was expressed at higher levels in preschool-aged and school-aged subjects with SMA than in age-matched healthy controls, suggesting that a relatively higher expression of PLS3 in subjects with SMA may act as a compensatory mechanism in humans. Moreover, levels of PLS3 expression in blood may be associated with age, because we found that children showed higher expression levels at early stages of development (⩽3 years), and these levels decreased with age and gradually stabilized.

The modification of PLS3 in subjects with SMA was first reported by Oprea et al.4 This study showed that asymptomatic female subjects with SMN1 deletion exhibited significantly higher expression levels of PLS3 than did their SMA-affected counterparts. Moreover, female subjects with higher PLS3 expression sporadically presented with milder phenotypes, which indicated that PLS3 is a gender-specific modifier of SMA. Another study performed by Stratigopoulos et al.5 showed an inverse correlation between PLS3 expression in blood and SMA severity in older postpubertal female patients (mean age, 20 years). In younger prepubertal female patients (mean age, 5 years), PLS3 expression levels were higher in type III than in type II female subjects with SMA only. In our study, we found that in the group of children older than 3 years (mean±s.d., 6.40±2.41 years), type III female subjects showed higher expression levels of PLS3 than did type II female subjects. However, no difference was observed in the group of children below 3 years of age. Notably, the SMA subjects that we focused on were all children and the overall age was relatively lower than that of subjects examined by Stratigopoulos et al.5 In our study, only one postpubertal female subject was included. However, in the study by Stratigopoulos et al.,5 postpubertal female subjects accounted for 38% of all female subjects. Moreover, they found that PLS3 expression levels were higher in younger patients, regardless of gender. Because expression of PLS3 may be associated with age, this expressed difference among clinical types in younger subjects may be covered, especially in earlier-onset type I patients.

Further, we analyzed the amount of PLS3 in the subjects with the same number of SMN2 copies in order to rule out the potential impact of SMN2 dosage. The data showed that in the presence of three copies of SMN2, expression of PLS3 was higher in type III female subjects than in type II female subjects. Notably, in our study, there was a trend that PLS3 expression of patients who were able to walk unaided was higher than those of patients who were unable to walk unaided, although no significant difference of PLS3 expression was observed between these two groups. It will be necessary to verify the correlation between PLS3 expression and phenotype in a lager sample. In a word, our data suggested that PLS3 may function as an age and/or gender-specific phenotype modifier in SMA.

A potential association between PLS3 and SMN was observed in a zebrafish SMA model described by Hao le et al.6 However, more recent data showed that increased PLS3 expression did not affect SMN levels, as was previously shown in humans, mice and zebrafish, suggesting that PLS3 exerts its modifying effects independently of SMN.7 PLS3 plays an important role in neuromuscular junction development, maturation, integrity and function by delaying axon pruning, resulting in amelioration of the SMA phenotype.

In view of the gender-specific protection of PLS3 in humans and in combination with the data demonstrating that over-expressed PLS3 in both zebrafish and mice could rescue the neuromuscular defects and corresponding movement phenotypes caused by low SMN levels, PLS3 is suggested as a putative target for SMA therapy.6, 7 Nevertheless, other modified factors, such as zinc finger protein,8 methylation status9 and experimental variables, should also be considered in any final analysis, for their potential to function cooperatively to affect the final phenotype in a patient. An investigation of these modifiers will help in understanding the mechanism underlying SMA pathology.