Plastin 3 Expression Does Not Modify Spinal Muscular Atrophy Severity in the ∆7 SMA Mouse

Spinal muscular atrophy is caused by loss of the SMN1 gene and retention of SMN2. The SMN2 copy number inversely correlates with phenotypic severity and is a modifier of disease outcome. The SMN2 gene essentially differs from SMN1 by a single nucleotide in exon 7 that modulates the incorporation of exon 7 into the final SMN transcript. The majority of the SMN2 transcripts lack exon 7 and this leads to a SMN protein that does not effectively oligomerize and is rapidly degraded. However the SMN2 gene does produce some full-length SMN and the SMN2 copy number along with how much full-length SMN the SMN2 gene makes correlates with severity of the SMA phenotype. However there are a number of discordant SMA siblings that have identical haplotypes and SMN2 copy number yet one has a milder form of SMA. It has been suggested that Plastin3 (PLS3) acts as a sex specific phenotypic modifier where increased expression of PLS3 modifies the SMA phenotype in females. To test the effect of PLS3 overexpression we have over expressed full-length PLS3 in SMA mice. To ensure no disruption of functionality or post-translational processing of PLS3 we did not place a tag on the protein. PLS3 protein was expressed under the Prion promoter as we have shown previously that SMN expression under this promoter can rescue SMA mice. High levels of PLS3 mRNA were expressed in motor neurons along with an increased level of PLS3 protein in total spinal cord, yet there was no significant beneficial effect on the phenotype of SMA mice. Specifically, neither survival nor the fundamental electrophysiological aspects of the neuromuscular junction were improved upon overexpression of PLS3 in neurons.

Spinal muscular atrophy is caused by loss of the SMN1 gene and retention of SMN2. The SMN2 copy number inversely correlates with phenotypic severity and is a modifier of disease outcome. The SMN2 gene essentially differs from SMN1 by a single nucleotide in exon 7 that modulates the incorporation of exon 7 into the final SMN transcript. The majority of the SMN2 transcripts lack exon 7 and this leads to a SMN protein that does not effectively oligomerize and is rapidly degraded. However the SMN2 gene does produce some fulllength SMN and the SMN2 copy number along with how much full-length SMN the SMN2 gene makes correlates with severity of the SMA phenotype. However there are a number of discordant SMA siblings that have identical haplotypes and SMN2 copy number yet one has a milder form of SMA. It has been suggested that Plastin3 (PLS3) acts as a sex specific phenotypic modifier where increased expression of PLS3 modifies the SMA phenotype in females. To test the effect of PLS3 overexpression we have over expressed full-length PLS3 in SMA mice. To ensure no disruption of functionality or post-translational processing of PLS3 we did not place a tag on the protein. PLS3 protein was expressed under the Prion promoter as we have shown previously that SMN expression under this promoter can rescue SMA mice. High levels of PLS3 mRNA were expressed in motor neurons along with an increased level of PLS3 protein in total spinal cord, yet there was no significant beneficial effect on the phenotype of SMA mice. Specifically, neither survival nor the fundamental electrophysiological aspects of the neuromuscular junction were improved upon overexpression of PLS3 in neurons.
Introduction female sib pair. In a second study no association of PLS3 expression was found in discordant female sib pairs [37]. In fact, PLS3 expression was slightly increased in the affected female sibling and not the asymptomatic individual. Recently, a tagged form of PLS3 protein has been investigated in SMA mice [37]. The authors report some mild benefits to the SMA phenotype under certain conditions. However, PLS3 expression is highly modulated at the protein level and the placement of a tag can affect both function and protein turnover [37]. We thus have investigated whether the overexpression of PLS3 without a tag can modify the SMA phenotype in mice.
In order to determine if PLS3 acts as a modifier of the SMA phenotype we generated transgenic mice expressing human PLS3 under control of the Prion (PrP) promoter. We have shown previously that the PrP:SMN transgene resulted in high expression of SMN in all neurons completely rescued the SMA phenotype in the mouse [38]. We proposed that if PLS3 is a SMA modifier then high expression of human PLS3 under control of this same promoter should alter survival and phenotype of Δ7 SMA mice. We found no increase in weight or survival of PrP:PLS3, Δ7 SMA mice in three different transgenic lines. Furthermore, we found no improvement in neuromuscular junction physiology in these PrP:PLS3 Δ7 SMA mice. Our results indicate that PLS3 is not a viable therapeutic target to modify the SMA phenotype in humans.

Ethics statement
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the University Laboratory Animal Resources at The Ohio State University and Wright State University. Our protocol was approved by The Ohio State University Institutional Animal Care and Use Committee (IACUC), Office of Responsible Research Practices, under Permit Number 2008A0089. Anesthesia was administered with Isoflurane according to our animal protocol. Carbon Dioxide followed by cervical dislocation for secondary means of confirmation was used for euthanasia according to our approved protocol.

Generation of PLS3 expressing transgenes
Human PLS3 cDNA (Clone ID 6064540, Open Biosystems) was end filled and cloned into the Prion (PrP) vector that contains the mouse prion promoter and exon 1, intron 1, part of exon 2 [34]. Human PLS3 cDNA was directionally cloned between the end filled KpnI and XhoI/SalI sites in PrP vector exon 2. The resulting construct was sequenced, linearized by digestion with PvuI, gel purified and dialyzed. The PrP:PLS3 plasmid was transfected into MN1 cells using Lipofectamine 2000 Transfection Reagent according to the manufacture's instructions (Invitrogen). After expression PLS3 was confirmed by Western blot the construct was injected into fertilized FVB/N Δ7 (JAX 5025) mouse oocytes to generate transgenic mice. The PrP:PLS3 transgene was detected with PrP exon 2 FP 5`GGACTCGTGAGTATATTTCAG and PLS3 RP 5`GAAGGTCTTGGCAATATCACTACT. Three founder mice, named PLS-14, PLS-39 and PLS-46, were bred to Δ7 mice (SMN2 +/+ ; Smn +/-: Δ7SMN +/+ ). PLS-14 female founder would not breed therefore an ovary transfer was performed into a FVB/N female and then progeny were bred to Δ7 mice. The SMN2 transgene and mouse knockout allele were detected as previously described [38,39]. PLS-39 and PLS-46 lines were bred to homozygosity, PLS-14 is not homozygous viable. Each transgenic line conformed to Mendelian autosomal patterns of inheritance thus multiple transgenic insertion sites were not detected. Homozygosity of the transgene was determined by qPCR on genomic tail DNA.

Weight and survival measurements
Mice were housed and fed at no more than 5 per cage according to our IACUC approved animal protocol and the Standard Operating Procedure for SMA mice SMA_M.2.2.003 (Treat-NMD. eu). Weight and survival analysis was performed as previously described for the Δ7 line [40,41]. A similar number of male and female mice were observed and weighed at minimum once per day from the date of birth (P0) until death or day 21 (P21). Any change in behavior, appearance or survival was noted. Mice where humanely euthanized when they achieved exclusion criteria including the inability of neonatal SMA mice to go to the mother (homing) to suckle and loss of greater than 20% of maximum weight that particular animal achieved according to our IACUC approved animal protocol. Required steps were taken to minimize suffering of the mice including administration of systemic analgesia (Motrin) in the water bottle at 100mg/5ml providing a dose of approximately 30 mg/kg when needed. All animals were grouped according to genotype. Kaplan-Meier survival curves and mean weights were graphed with SigmaPlot.
Expression of PLS3 in Brain and Spinal cord of PrP:PLS3 SMA mice RNA was isolated from brain and total spinal cord at P12 using TRIzol reagent (Invitrogen), purified with the RNeasy kit (Qiagen) and converted to cDNA as previously described [38]. Reactions were run on the ABI 7300 Real-Time PCR System. Relative human and mouse PLS3 levels were determined by normalizing to mouse cyclophilin (PIPB) expression. Three technical replicates and five to seven biological replicates were performed for each sample. All three transgenic lines as well as a non-transgenic control were tested.
Expression of PLS3 in LCM isolated motor neurons of PrP:PLS3 SMA mice Motor neurons were collected from fresh frozen spinal cord sections on a Zeiss Palm Robo 3 Laser Capture Microdissection System. Motor neurons were located based on size and location in the anterior horn after Nissl staining for contrast. RNA was isolated with the Ambion RNAaqueous Micro kit (AM1931) and aRNA was generated with the Arcturus PicoPure RNA Isolation Kit (ABI KIT0204). Droplet generation and reader analysis were performed on the QX200 (Bio-Rad). 15,000 to 18,000 droplets containing cDNA, primers, probe, 2x ddPCR SuperMix for Probes, and droplet generation oil were generated and amplified. Primers used to detect cDNA include: Prion:PLS3 transgene, FP: 5`CGGATCAGCAGACCGATTCT, RP:5`GCACCTCGGAATC TTTGCA,probe:FAM-ATCGGTGGCAGGACT-MGB; mouse Pls3, FP:5`ccgcgactccctatg aatctt (mouse specific), RP:5`gagttcatcaagctcatctttgga, Probe:FAM-ACATGGA TGAGATGGC-MGB; Mouse cyclophilin, FP:5`GTCAACCCCACCGTGTTCTT, RP: 5`TTGGA ACTTTGTCTGCAAACA, PROBE: VIC-CTTGGGCCGCGTCT-MGB. A sufficient number of positive and negative droplets were read by the QX200 reader and quantified using the QuantaSoft software (Bio-Rad). The concentration of transcripts was determined using Poisson statistical distributions and relative human or mouse plastin levels were determined by normalizing to mouse cyclophilin expression. Two technical replicates (for a total of >20,000 droplet PCR reactions) and three biological replicates were performed for each sample. All three transgenic lines as well as a non-transgenic control were tested.

Protein expression of PLS3 spinal cord
Brain and spinal cord tissues were harvested from 3 PrP:PLS3 male mice at P10 for each transgenic line and non-transgenic controls. Protein isolation and western blots were performed as previously described [42]. The antibody used to detect PLS3 (1:250, GenTEX, 103323) is not specific for human PLS3 thus the total amount of mouse and human PLS3 protein was detected. Accurate size detection of the PLS3 protein was confirmed by detecting PrP:PLS3 protein isolated from transfected MN-1 cells. Blots were incubated with anti-rabbit Fab fragment HRP (1:10,000, Jackson ImmunoResarch, 111-035-006). Three concentrations of the same protein sample were loaded on the gel (90μg, 45μg and 25μg). Blots were probed with mouse anti beta-tubulin mAb (1:10,000, Abcam Ab7291) to measure protein loading and developed using the ECL system as described by the manufacturer (GE Healthcare Life Sciences). Blots where scanned and quantified as described (http://lukemiller.org/index.php/ 2010/11/analyzing-gels-and-western-blots-with-image-j/) and the area under each peak determined with ImageJ software. Statistical analysis was performed with SigmaPlot. All samples collected for RNA and protein analysis were from male mice.

Zebrafish axon correction
Zebrafish embryos were maintained at~28.5°C and staged by hours post fertilization (hpf) [43]. Transgenic Tg(mnx1:0.6hsp70:GFP)os26 [44] embryos expressing GFP in their motor axons (referred to as Tg(mnx1:GFP) embryos) were used for smn morpholino (MO) and human PLS3 mRNA injections. The antisense smn MO was described previously by McWhorter et al. [33]. One cell-stage embryos were injected with 9ng of smn MO with or without 250 pg of synthetic human PLS3 mRNA.
To generate mRNA, human PLS3 in pCMV.sport 6 vector was subcloned into pCS2+ vector and linearized with NotI. Capped RNA was generated using the Sp6 mMESSAGE mMA-CHINE kit (Ambion, Austin, TX) following the manufacturer's instructions.
To visualize motor axons, 28 hpf Tg(mnx1:GFP) embryos were anesthetized with tricaine (160 μg/ml) and fixed overnight at 4°C in 4% formaldehyde/PBS. After removing from fix, embryos were mounted on glass coverslips for observation under a Zeiss Axioplan microscope, scored [45] and imaged on a Leica confocal microscope. Ten motor axons were scored per animal and animals were designated as containing severe, moderate, mild, or no defects based on criteria in Carrel et al., [45]. Three separate experiments were performed and for each condition (control, smn MO and smn MO + PLS3 RNA), n was between 19-24 embryos. Data was plotted as mean ± SEM for the three experiments and Mann-Whitney non-parametric rank test was used to test significance.

Statistical analysis
Quantitative data are expressed as mean ± SEM. Values for number of animals are given in Results and the figure legends. Kaplan-Meier survival curves were generated with SigmaPlot and statistical significance was determined using the log-rank test. The Holm-Sidak method was used for all pairwise multiple comparisons. Significance of weight data were determined with one-way ANOVA and by the Compare Growth Curve function found in the R-Package (Statmod). Specific tests for qPCR, ddPCR and western blot analysis are as described in the figure legends. Values of p<0.05 were considered significant.

Functionality of PLS3 coding sequence used to generate transgenic mice
To test whether the PLS3 cDNA used in generating our transgenic lines encoded functional PLS3 protein, we tested PLS3 mRNA in zebrafish. Decreasing smn transiently in zebrafish embryos using an smn morpholino has been shown to result in motor axon defects [33]. Furthermore, injecting human PLS3 RNA into these smn morphants rescued the axonal defects [36,49]. Therefore, we tested the PLS3 sequence used to construct our transgene in this same assay and found that it was able to significantly rescue the smn morphant motor axon defects (Fig 2). This finding demonstrates that the PLS3 coding sequence used to generate our transgenic mice produced functional PLS3 protein. A construct containing the mouse Prion (PrP) promoter, exon 1, intron 1 and part of exon 2 was fused to the human PLS3 cDNA. This same promoter was used previously to express SMN in neurons [38]. Arrows indicate the location of PrP exon1 forward primer and PLS3 reverse primer used to specifically amplify PLS3 transcripts produced by this transgene. Expression of PLS3 in brain and spinal cord tissue To determine expression of Prp:PLS3 in the brain and spinal cord we used quantitative qRT-PCR. Endogenous mouse Pls expression (Fig 3A and 3B) and human PLS3 expression (Fig 3C and 3D) was measured by quantitative RT-qPCR in the brain (Fig 3A and 3C) and in total spinal cord (Fig 3B and 3D) tissue at P10 for each transgenic line (PLS-14, PLS-39 and PLS-46). We found that the expression of the PrP:PLS3 transgene was nearly 100 fold increased over endogenous mouse Pls3 levels in both brain and spinal cord in all three transgenic lines. (n = 5-7 for each transgenic line and tissue). We used primers located in PrP exon 2 and PLS3 exon 1 to specifically detect the transgenic expression of human PLS3 from PrP:PLS3. Mouse Pls was specifically amplified using a forward primer that was unique to mouse Pls. The highest

Expression of PLS3 in motor neurons
To ensure that PrP:PLS3 was indeed expressed in neurons we sectioned P10 lumbar spinal cord (L3-L5) from each transgenic line. The motor neurons were isolated by laser capture microdissection (LCM) and RNA was extracted. One round of aRNA amplification (Arcturus) was followed by quantitative PCR using droplet digital PCR (ddPCR, Bio-Rad). The amount of PLS3 expression detected in the motor neuron is more than 100x greater than the endogenous mouse Pls3 expression for each transgenic line examined (Fig 4A). The level of mouse Pls3 expression for each transgenic line was no different from that of a non-transgenic age matched control (Fig 4B).

Total Plastin protein expression in spinal cord tissue
We next examined the expression of total PLS3 protein. The Plastin antibody used was first tested to ensure that it reacted with PLS3 protein at the correct size (~70kD). MN-1 cells were transiently transfected with the PrP:PLS3 construct and the approximately 70kd PLS3 protein was detected by Western blot on transfected cells. In the case of transfected MN-1 cells a marked increase in total PLS3 expression was observed at the protein level. Western blot analysis of total spinal cord from P10 mice expressing PLS3 is shown in Fig 5. Despite a 100-fold increase in PLS3 mRNA expression, the increase in total PLS3 protein levels was only 2 fold when compared to non-transgenic animals. There was a significant increase in PLS3 protein in the PLS-14 line (p<0.005) (Fig 5A). This result is similar to the findings of Ackerman et al in which the PLS3 transgene was tagged and therefore more easily detected [50]. However the levels indicated for total PLS3 and mouse Pls3 combined are similar to Figure 7 in Ackerman et al. [50]. Thus PLS3 mRNA is dramatically increased in the spinal cord but post-translational regulation mechanisms present in the mouse limit the level of PLS3 protein that can be obtained. Any regulation of the PLS3 protein is important to consider in determining if PLS3 expression alters SMA.

Effect of PLS3 expression on SMA phenotype
We measured the weight of PrP:PLS3 mice to determine if overexpression of PLS3 in neurons increased the weight of Δ7 SMA the mouse. We found that the weight is not increased in three PrP:PLS3 transgenic lines in the Δ7 SMA mouse (Fig 6). The weight of each PrP:PLS3 transgenic line in the presence and absence of mouse Smn was measured daily until weaning at 21 days of age. The three PrP:PLS3 transgenic lines (PLS-14 +/-, SMN2 +/+ ; Smn -/-; Δ7SMN +/+ , To determine if survival of the Δ7 SMA mouse is improved upon overexpression of PLS3 in neurons we monitored survival. Survival is not increased in three PrP:PLS3 transgenic lines in the Δ7 SMA mouse (Fig 7).

Electrophysiology of PLS3 SMA mice
We and others have previously shown that early stages of SMA disease pathogenesis are characterized by functional abnormalities of the neuromuscular junction (NMJ) [46,[51][52][53]. It has recently been reported that PLS3 expression rescues function of the NMJ in mice with SMA [50]. In order to determine whether expression of PLS3 rescues functional NMJ abnormalities we examined the physiology of NMJs in the tibialis anterior muscle of mice on P10 to P13 as previously described [46,51]. Each PrP:PLS3; Δ7SMA mouse was compared to an age-matched littermate that was studied on the same day. The most dramatic abnormality in SMA is a 60% reduction in endplate current (EPC) amplitude, which is determined by both the number of synaptic vesicles released following nerve stimulation (quantal content) and the amplitude of the muscle response to the transmitter released from a single vesicle (quantal amplitude) [46,51]. When plastin SMA mice were compared to their unaffected littermates they had a 60% reduction in EPC amplitude (p <0.05, Fig 7) that was very similar to the reduction we found previously in the same line of SMA mice at P10-P14 [46,51]. The reduction in endplate current amplitude was due to both reduction in quantal content and quantal amplitude (Fig 8) with the magnitude of reduction of both parameters similar to what we found previously in SMA mice at P10-P14 [51].
Previously, we found an increase in MEPC and EPC time constants that was likely due to prolonged expression of embryonic acetylcholine receptors (AChRs) [46,51]. A similar increase in EPC time constant was present in PrP:PLS3 SMA mice (p <0.05, Fig 8). Finally, we and others previously found that a reduction in the probability of synaptic vesicle release as shown by increased facilitation during repetitive stimulation was a likely contributor to reduced quantal content in SMA NMJs [46,51,53]. A similar increase in facilitation was present in PrP:PLS3 SMA mice relative to control littermates (p <0.05, Fig 8).

Discussion
The overexpression of PLS3 has been suggested to modify the SMA phenotype [36]. In particular, it has been suggested to act as a female specific modifier thus overexpression of PLS3 would only alter female SMA patients. This finding was reported after the identification of higher PLS3 expression in lymphoblasts of the less severely affected SMA individual of siblings with identical haplotypes but variant phenotype. However, an increase in PLS3 expression does not occur in all haploidentical cases. Thus it was reported that PLS3 is a female specific modifier of SMA phenotype that is not always penetrant [36]. A close examination of the 6 pedigrees studied reveals that in all but one example (family #34), the severely affected case was male [35,36]. Thus low PLS3 expression would not be predicted to have any impact on these individuals anyway. In essence, the initial evidence for PLS3 modifying the SMA phenotype comes down to one family where the two mildly affected female patient showed a one fold (BW279) and 1.6 fold (BW283) increase in PLS3 transcript compared to their more severe sister (BW280) [36]. In a separate study female patients with the more severe phenotype showed high PLS3 expression compared to their less severe female siblings [54]. Thus it does not seem PLS3 expression can always modify SMA in females and it is unclear why this modification would be partially penetrant and female specific. Lastly no clear insight into how increased PLS3 expression occurs has been presented. Does it occur due to an alteration of regulatory sequence at the PLS3 locus, alteration of methylation at the PLS3 locus alteration of a transregulator of PLS3 expression or altered escape from X inactivation at that loci. In the latter case it can be noted that PLS3 has been reported to undergo X inactivation [55] and it is hard to see how this would specifically give rise to high PLS3 expression in certain individuals. Furthermore males do show high PLS3 expression but modification is not reported to occur in this case (family #800, individual LN421 found in Oprea, et al.) [36]. Regardless it is important to address the issue of how PLS3 is activated as well as why PLS3 expression is only believed to operate in certain individuals.
In a subsequent study, Stratigopoulos et al. [56] found no difference in PLS3 expression in 47 female SMA patients when all ages, SMA types or SMN2 copy number were compared. An inverse correlation between PLS3 expression and SMA severity was only identified when females were grouped by age (pre and post pubescence). No change in PLS3 expression was identified in males grouped by SMA severity or SMN2 copy number. PLS3 levels were found to be 50% lower in older males. Finally, expression of PLS3 did not correlate with the functional measures of CMAP or MUNE in males or females [56].
Recently, Yanyan et al. found higher levels of PLS3 in type 3 female SMA children (over the age of 3) compared to type 2 female children [57]. Yet the level of PLS3 expression was always higher in females than in males and correlated positively with SMN2 copy number. The level of PLS3 was higher in SMA patients 3 to 12 years of age compared to healthy controls. Thus it is suggested that PLS3 may be playing some compensatory role in SMA, however levels of PLS3 were highest in healthy controls under age of 3. PLS3 is unlikely to be useful as a biomarker due to the alteration of expression in blood with patient age and sex [37].
Although PLS3 mRNA is elevated in lymphoblasts at the mRNA level there are differences in the reports of protein expression. Opera et al. [36] reported altered PLS3 protein levels whereas Bernal et al. [54] found that PLS3 protein levels where not detectable in lymphoblasts and not significantly altered in fibroblasts of patients that had different mRNA levels of PLS3. Similarly, in our study we observed only a 2 fold increase in PLS3 protein in transgenic mice heavily overexpressing (up to 300 fold higher) PLS3 mRNA indicating the likely occurrence of posttranslational regulation of PLS3 expression. Hao et al. [58] have reported in zebrafish that reduction of SMN resulted in reduced PLS3 protein levels whereas in the mouse Ackerman et al. [50] found that SMN levels did not alter PLS3 levels. Other studies have shown that PLS3 levels are increased under various conditions. For instance cisplatin-resistant human bladder, prostatic, and head and neck cancer cell lines express high levels of PLS3 when compared to cisplatin-sensitive cells [59]. High PLS3 levels have also been found in Sezary Syndrome patients and this was associated with loss of CD26. In addition, PLS3 positive cells showed hypomethylation of the PLS3 CpG island at sites 95-99 [60,61]. Interestingly, the polymorphism SNP PLS3 rs871773 T allele is associated with a higher protein expression of the PLS3 gene in colon cancer and an increased risk of recurrence of colon cancer [62]. If PLS3 does alter severity of SMA, defining the role of both PLS3 rs871773 and the hypomethylation of sites 95-99 is important as it gives a mechanism of PLS3 activation and may even result in a DNA marker that could be followed in patient material. However, this does not explain why increased PLS3 expression only modifies female SMA patients and is often non-penetrant. Indeed our results show no marked alteration of SMA phenotype in mice with a 100-fold increase in mRNA expression of PLS3. The studies we present here do not support a role for PLS3 in SMA. Moreover, the lack of penetrance in modifying the phenotype in males, as well as certain female cases, is difficult to reconcile.
Previously, overexpression of PLS3-V5, which contains an amino terminal tag, was reported to improve the Taiwanese model of SMA [50,63] but only very slightly and under specific conditions [50]. No improvement of survival of the Taiwanese 2 copy SMN2 mice was seen on a C57BL/6 background with overexpression of PLS3-V5 and only marginal improvement of muscle fiber size and connectivity of the NMJ. In a F1 mixed background of FVB/N and C57BL/6 the mean survival rate was increased by 2 days and the maximum survival was not increased [50]. In our experience with the Δ7 SMA mice this kind of survival increase is not significant and can vary between tests. In essence, the differences in survival between the current study and that of Ackerman et al. are minimal and we suggest there is no significant improvement in survival of Δ7 SMA mice with overexpression of PLS3. Alternatively, the modest increase in survival of 2 days in the mixed background Taiwanese SMA mouse model could be due to a neuroprotective effect of PLS3. A slight increase in survival was also observed in the Taiwanese SMA mice upon administration of neuroprotective factors IGF-1 [64], cardiotrophin-1 [65], and Bcl-xL [66].
In our study as well as that of Ackerman et al., electrophysiology studies of the function of the NMJ were performed. Ackerman et al. reported a small improvement in the time constant of the endplate potential and quantal content [50]. However this was only on a mixed background and is unlikely to have a major impact on NMJ function. We did not find evidence to suggest significant improvement in either parameter. There is no evidence to suggest that expression of PLS3 improved any of the pre-and postsynaptic physiologic deficits at the neuromuscular junction in our study of Δ7 SMA mice. One difference between our study and the previous study is that we used voltage clamp of muscle fibers to directly measure synaptic currents whereas the study by Ackermann et al. measured endplate potentials. Endplate potentials can be affected by changes in muscle fiber properties (fiber size and specific membrane resistance) that are unrelated to synaptic function. These differences might account for the difference in findings relating to time constant, however, it seems unlikely that a difference in muscle fiber property could account for the difference between the two studies on quantal content. We cannot rule out that overexpression of PLS3 has a very modest effect on synaptic function that would be picked up with study of more mice. Driving expression of human PLS3 in motor neurons rescued the NMJ defects and motor function in zygotic zebrafish smn mutants suggesting that under low conditions of SMN, PLS3 can indeed benefit vertebrate motor neurons [58] In conclusion, we have shown in the Δ7 mouse model of SMA no beneficial effects of PLS3 overexpression in the neuron. This is also consistent with the study of Bowerman et al. [67] using a milder model of SMA where loss of Profilin results in increased PLS3 expression but no modification of the SMA phenotype [67]. A puzzling feature of all the reports of PLS3 modification is that the effect is proposed to be sex-specific and partially-penetrant. To date, this hypothesis has not been replicated in any animal studies and is not explained by PLS3's location on the X chromosome because a transgene on an autosome will not be subjected to inactivation. It is clear that there are males in the population that express PLS3 but this does not modify the SMA phenotype in humans. Why would this be the case? We suggest that modifiers of SMA need to be revisited in the human population in discordant sibling pairs. A genetic modifier that has a DNA change or a solid mechanistic base as to why altered plastin expression occurs can be studied in these individuals. In this case it would be preferable to study haploidentical pairs of discordant type 1 and type 2, or type 2 and type 3 siblings. The genetic modifier will not be present in any severe SMA type 1 patient therefore type 1 patient DNA can be used to exclude false modifiers. Indeed there are SNPs and methylation changes associated with altered PLS3 expression that could be investigated in SMA. Currently PLS3 can be viewed as a candidate modifier where an understanding of mechanism of activation, and DNA changes associated with increased expression is not understood, nor why modification only occurs under certain circumstances. However an alternative explanation is that PLS3 is in fact not the critical modifier of SMA phenotype. Thus studies that remain open to the possibility of defining alternate modifiers in SMA are of critical importance.