Novel SPAST deletion and reduced DPY30 expression in a Spastic Paraplegia type 4 kindred

Background The hereditary spastic paraplegias (HSPs) are pleiomorphic disorders of motor pathway and a large number of affected genes have been discovered. Yet, mutations in SPG4/SPAST represent the most frequent molecular etiology in autosomal dominant (AD) patients and sporadic cases. We describe a large, AD-HSP Sardinian family where 5 out of several living members harbored a novel deletion affecting also the 5′UTR of SPAST and resulting in reduced expression of DPY30, the gene located upstream SPAST in a head-to-head manner. Case presentation A 54-year-old woman manifested leg stiffness at age 39 and required a cane to walk at age 50. Neurological examination disclosed mild spasticity and weakness in the legs, hyperreflexia in all limbs, and bilateral Babinski sign. She also complained of urinary urgency, but no additional neurological symptoms or signs were detected at examination. The clinical examination of 24 additional relatives disclosed three further affected individuals, two men and one woman. In the four symptomatic patients the initial manifestations were walking abnormalities and leg stiffness with a mean age at onset (SD) of 46.75 (5.44) years (range 39–51). The mean disease duration was 13.2 (13.4) years (range 6–35), and it correlated well with clinical severity (SPRS score) (r = 0.975, p = 0.005). One patient was confined to bed and displayed knee and ankle contractures, another case needed a cane to walk, and two individuals were able to walk without aids. Interestingly, a patient had also had a miscarriage during her first pregnancy. Gene testing revealed an heterozygous deletion spanning from the 5′-UTR to intron 4 of SPAST in the affected individuals and in one clinically unaffected woman. In three affected patients, the deletion also determined low mRNA levels of SPAST and DPY30, a component of the Set1-like multiprotein histone methyltransferase complex located upstream, head-to-head with SPAST. Conclusion Together with data described in a Japanese family, our findings seem to suggest that genes close to spastin might be candidates in modulating the clinical phenotype. This report endorses future research on the role of neighboring genes as potential players in SPG4 disease variability.


Background
Heterogeneity is a key feature of the hereditary spastic paraplegias (HSPs). To date, autosomal, sex-linked, and cytoplasmic inheritance have been reported, an ample array of complicated phenotypes disclosed, more than 70 loci mapped, and roughly 50 disease genes cloned [1,2]. In common clinical practice, the gene-after-gene testing strategy allows definition of the molecular basis in about half of HSP cases, unless peculiar features emerge during examination or follow-up.
Mutations in SPAST/SPG4 encoding spastin represent the most common cause of autosomal dominant hereditary spastic paraplegias (AD-HSP) and also account for about 15% of sporadic cases [3,4]. In north-west Sardinia the relative frequency of HSP is higher than what is calculated in other Western European populations with an estimated crude prevalence of about 17.5/100,000 for AD-HSP [5]. As documented in several families and different populations [4,6], the SPG4 phenotype is usually pure and inter-and intra-familial variability of the clinical presentation are well established [7]. In some cases, single nucleotide polymorphisms in SPAST and variants in additional genes are invoked as modifiers of age at onset, disease course and severity [8,9].
We identified a novel SPAST mutation segregating in a Sardinian kindred (family IK).

Case presentation
The proband, a 54-year-old woman (IV-13) (Figure 1), manifested leg stiffness at age 39 and required a cane to walk at age 50. Neurological examination disclosed mild spasticity and weakness in the legs, enhanced deep tendon reflexes in all limbs, and bilateral Babinski sign. Case IV-13 also complained of urinary urgency, but no additional neurological symptoms or signs were detected at examination. The patient reported that her deceased father (III-03) and two uncles (III-02, III-04) had manifested similar gait abnormalities in adult age, although with different degree. Detailed clinical information and examinations were gathered from her living relatives.

Materials and methods
Total genomic DNA was purified by peripheral blood with standard methodologies. Analysis of common AD genes associated with HSP used traditional Sanger sequencing and the BigDye 3.1 Chemistry, as reported [5]. Search for copy number variation and gene deletion/duplication adopted reported array-comparative genomic hybridization (aCGH) and multiple ligation-dependent probe amplication (MLPA) methodologies [10].
Twenty-four additional relatives were examined and sampled with written informed consent in family IK. Using oligonucleotide primers (5′-3′) SPASTPrimer F, SPASTPrimer R1 and SPASTPrimer R2 that were designed flanking the mutation deletion breakpoints (sequences available upon request), we PCR-amplified genomic DNA to generate a single 955-bp fragment in wild-type individuals. The presence of the heterozygous SPAST deletion produced an additional fragment of 303bp ( Figure 2). To test the effects on mRNA expression of the mutation deletion in SPAST, total blood RNA was extracted using a micro-scale total RNA separator kit (Ambion INC., Austin, TX). For standard gene expression experiments, the mRNA transcript levels were determined by qPCR runs in an ABI7500Fast system (Applied Biosystems, Foster City, CA) using the TaqMan Universal PCR Protocol, and human SPAST (Hs00368084_m1, Applied Biosystems), SLC30A6 (Hs01071782_m1), and DPY30 (Hs00261491_m1) as probes. GAPDH (Hs99999905_m1, Applied Biosystems) was used for endogenous normalization, and expressions were determined using the comparative Ct method [11]. Values were normalized in reference to the average control value obtained from three agematched normal control subjects. Statistical analyses used unpaired two-tailed Student-test (significance was set at p <0.05).
Using a rapid PCR-based method to quickly genotype individuals in the family, and rule out the mutation in 500 ethnically-matched control chromosomes, we identified the new mutation in a total of five individuals, including the yet asymptomatic IV-25 ( Figure 1). In all, global disease severity was assessed using the Spastic Paraplegia Rating Scale (SPRS) [12]. Table 1 summarizes clinical data in carriers of the novel SPAST mutation deletion. In the four symptomatic patients the initial manifestations were walking abnormalities and leg stiffness with a mean age at onset (SD) of 46.75 (5.44) years (range 39-51). The mean disease duration was 13.2 (13.4) years (range 6-35). Pearson's correlation coefficient indicated a positive correlation between disease duration and clinical severity (SPRS score) (r = 0.975, p = 0.005). One patient (III-04) was confined to bed and displayed knee and ankle contractures, another needed a cane to walk, and two individuals were able to walk without aids. Interestingly, subject IV-07 had also had a miscarriage during her first pregnancy. At this point, we cannot state if IV-25 will develop any clinical manifestation in the near future or if she is a true not penetrant carrier of the mutation, since her present age is slightly younger than the mean age of onset of motor disturbances in the family. Reduced or no penetrance has repeatedly been described in spastin-related HSP kindred [1,3], including cases from Sardinia [10].
In three affected patients, we observed that the deletion also determined low mRNA levels of SPAST and DPY30, a component of the Set1-like multiprotein histone methyltransferase complex located upstream, headto-head with SPAST [15] (Figure 3).

Discussion
This report is the third description of a deletion characterizing the 5′-UTR of SPAST. Interestingly, the mutation also affected even more upstream sequences likely regulating SPAST and the neighbor DPY30 gene. Previously, a deletion was found in a Japanese family with clinical features and disease duration highly similar to our cases [16]. Also, six men and four women in a further Japanese kindred harbored a 70 kb deletion involving exons 1 to 4 of SPAST and also exons 1 to 3 of DPY30 [17]. Those patients had on average a teenage onset and a slowly progressive course leading to wheelchair in four, and use of a walking stick in one case. Other clinical features in that family were mild cognitive impairment and slight peripheral neuropathy. We believe that the presence of the spastin deletion might well explain a tendency towards less severe walking disability in family IK and in the Japanese kindred as described before in other families [3,9]. It is, however, intriguing that all affected Japanese women experienced miscarriages of unknown etiology similar to subject IV-07 in our kindred. Whether the shared feature of birth interruption relates to a similar defect in DPY30 is still questionable.
DPY30 is yet to be fully characterized but it seems to be essential for neural fate of embryonic cells, cellcycling and cellular proliferation [15,18]. In nematodes, null mutations in dpy-30 cause XX-specific lethality and the gene is required for normal development of XO males [19]. As it orthologue, it can be speculated that human DPY30 is also implicated in brain development and infertility [20] and when mutated might lead to miscarriages. The hypothetical function of the gene, however, cannot clearly explain how a low DPY30/mRNA expression (probably because of a position effect involving long-range gene regulatory elements) in the IK family with a milder phenotype agrees with the reduced mRNA levels (because of partial gene deletion) in the Japanese family with an earlier onset and apparently more severe neurological features [17]. We have no clear-cut explanation for this apparent "clinical riddle" other than raising the possibility of additional modifiers in the two families, maybe related to the different ethnic  origin. Yet, alike the recent identification of a small deletion of SLC30A6the gene flanking SPAST 3′UTRin another Italian SPG4 family [10], this report endorses future research on the role of neighboring genes as potential players in SPG4 disease variability.

Conclusion
We describe an AD-HSP Sardinian family where 5 out of several living members harbored a novel deletion affecting also the 5′UTR of SPAST and resulting in reduced expression of DPY30, the gene upstream SPAST in a head-to-head manner. If the presence of the spastin deletion might well explain a tendency towards less severe walking disability in our family, it is intriguing that a patient in our kindred experienced a miscarriage of unknown etiology similar to all affected women in a Japanese family harboring a SPAST and DPY30 deletion. This report encourages future research on the role of neighboring genes as potential players in SPG4 disease variability.

Consent
Written informed consent was obtained from the patients in the family for publication of this Case report.