Novel Splice-Site and Missense Variant of PNPLA6 in an Austrian Family Causing Spastic Paraplegia-39 with Cerebellar Oculomotor Disorder

Background The term hereditary spastic paraplegia comprises an ever-expanding array of neurological disorders with distinct aetiologies. Spastic paraplegia 39 is one of the many variants with additional features of other organs and neurological systems. We describe a large kindred with two hitherto undescribed mutations of PNPLA6 and a novel clinical phenotype as, in addition to spastic paraplegia, affected subjects suffered from a prominent cerebellar oculomotor dysfunction. Methods Three of ve genetically tested family members of a large kindred were affected by spastic gait and cerebellar oculomotor dysfunction. Clinical, imaging, laboratory and electrophysiological data were analyzed. Genetic analysis was done using next-generation sequencing. Segregation analyses were performed by Sanger sequencing. To assess the pathogenicity of genetic variants on the encoded protein, in silico assessments were carried out. Results Two hitherto unknown sequence variants in the PNPLA6 gene, a splice-site variant c.1635+3G>T and a missense variant c.3401A>T, p.(Asp1134Val), were detected. Compound-heterozygous siblings presented with lower limb spasticity and a marked cerebellar oculomotor disorder accompanied by moderate hypogonadotropic hypogonadism in the female. A paternal uncle being homozygous for the splice-site variant of PNPLA6 presented with increased lower limb reexes, an unstable gait and cerebellar oculomotor dysfunction. Treatment with 4-aminopyridin, a potassium channel blocker, lead to meaningful improvement of clinical symptoms. Conclusion PNPLA6 gene variants are associated with a broad phenotypic spectrum to which we add cerebellar oculomotor dysfunction. In our kindred, the full clinical manifestation only occurred in compound-heterozygous subjects indicating that biallelic pathogenic variants lead to more serious and earlier onset of symptoms. Our ndings emphasize the role of PNPLA6 in different neurodegenerative disorders.


Introduction
Hereditary spastic paraplegias (HSPs), also known as Strumpell-Lorrain disease, are a heterogeneous group of monogenetic neurodegenerative disorders where the unifying clinical presentation is that of progressive spasticity and paraplegia.(1) Inheritance follows autosomal-dominant (AD), autosomal-recessive (AR) or X-linked patterns. The classi cation is based on the affected genes. (2,3) So far more than 80 causal genes or loci have been detected and resulted in a spastic paraplegia gene (SPG) classi cation. (4,5) The histopathological feature of HSPs is a length-dependent axonal degeneration of motor and sensory bers within the corticospinal tract and the dorsal column.(6, 7) Spastic Paraplegia 39 (SPG39), an AR SPG, was rst described in 2008 by Rainier et al.(8) and is characterized by a childhood onset of progressive spasticity of the lower limbs and progressive weakness of the upper and lower limb muscles.
The phenotype of SPG39 corresponds to SPG20, an AR form of HSP with lower motor neuron involvement, and to organophosphorous compound-induced delayed neuropathy (OPIDN) with distal axonal degenerations.(8-10) SPG39 is caused by mutations in PNPLA6, a gene encoding neuropathy target esterase (NTE). Interestingly, mutations of PNPLA6 have also been found to cause other rare hereditary neurodegenerative diseases such as Gordon-Holmes, Boucher-Neuhauser, Laurence-Moon and Oliver-McFarlane syndrome. (11)(12)(13) Rainier et al.(8) reported two families in which affected subjects developed a childhood onset progressive spastic weakness of lower limbs and wasting of distal upper and lower limb muscles. In the consanguineous family, affected subjects carried a homozygous mutation in the PNPLA6 gene c.3034A > G, p.(M1012V) that was shown to alter the catalytic domain of the encoded protein NTE. Moreover, in the non-consanguineous family affected subjects were compound heterozygotes where one allele had a missense mutation c.2669G>A, p.(R890H) and the other allele had an insertion mutation c.2946_2947insCAGC, p.(S982fs1019). Both variants cause an impaired enzymatic function of NTE.
In 2014, Synofzik et al. (11) reported a family with HSP harboring the compound heterozygous variants c.787G>A, p.(Val263Ile) and c.2519G>A, p.(Gly840Glu) in the PNPLA6 gene. Affected subjects presented with spasticity, hyperre exia of lower extremities, positive Babinski's sign and a mild motor neuropathy.
Yoon et al. (14) presented a sole patient with a pure spastic paraplegia phenotype characterized by motor delay and gait abnormalities with falls. This patient was heterozygous for the variant c.2944_2947dup, p.(Arg983ArgfsX86) of the PNPLA6 gene.
We here describe a large kindred with two hitherto undescribed mutations in the PNPLA6 gene. Their clinical phenotype was novel as, in addition to spastic paraplegia, they had a marked cerebellar oculomotor dysfunction.
Case Presentation S1 (Subject 1), the index subject, was a 29-year-old male presenting with benign peripheral paroxysmal vertigo. Routine neurologic examination revealed marked gait disturbance for which the patient had never sought medical advice. He reported normal milestones during childhood and worked as a skilled specialist in manufacturing.
On examination he had grossly spastic gait with normal Romberg test and brisk muscle stretch re exes with sustained ankle clonus.
However, Babinski's sign was negative. His muscles were neither weak nor atrophic. Oculomotor testing revealed massive saccadic pursuit movements. There was coarse nystagmus evoked by upward and horizontal gaze. On downward gaze, however, the direction of the nystagmus was upward (to his forehead) (Additional le 2; Additional le 3). There was no nystagmus in primary position. Rebound nystagmus could easily be elicited. HIT (head impulse test) according to Halmagyi showed normal VOR (vestibular ocular re exes), but the suppression of the VOR by xation was markedly impaired. Optokinetic nystagmus was reduced. Despite these prominent ndings the patient had no subjective visual complaints. There was no ataxia of gait, the upper limbs or of speech. The remainder of the neurologic examination was inconspicuous. In particular there was no sensory, autonomic or cognitive dysfunction.
In S2 (subject 2), his 35-year-old sister, an o ce employee with good cognitive abilities, a marked gait disturbance had been present from early childhood with a very slowly progressive course. In terms of quality, her neurological ndings were identical to subject S1, however, somewhat more pronounced. At the age of around 18 years, she developed secondary amenorrhea.
The father, S5 (subject 5), of S1 and S2 had no complaints. On examination he had sustained ankle clonus without Babinski's sign. Gait and balance were normal and there was no oculomotor disturbance.
The same was true for the mother, S4 (subject 4), in whom ankle clonus was present but exhaustible. S3 (subject 3), the 71-year-old paternal uncle of S1 and S2, became symptomatic only in his early fourties with gait disturbance and vertigo on head movement. On examination there was gaze evoked nystagmus to both sides and upwards but not downwards, saccadic pursuit movements, moderately impaired VOR and impairment of xation-induced suppression of the VOR (corresponding with impaired reading on smooth sinusoid head movements).
Gait was clumsy but not spastic or atactic; Romberg test was normal. Muscle strength was normal without any atrophy, but knee and ankle jerks were markedly increased with sustained ankle clonus. Babinski's sign was negative. There were no sensory ndings.
The son of S3 had no complaints and was normal on thorough neurological examination.
In the extended family one female cousin in the paternal line of S1 and S2 had died from bulbar-onset amyotrophic lateral sclerosis at the age of 33, and another cousin had a well-founded diagnosis of multiple sclerosis. Genetic material was not available from these two subjects.

Paraclinical Findings
Routine blood tests were done in S1, S2 and S3 and did not yield any systematic abnormalities. Hormone status was compatible with hypogonadotropic hypogonadism in S1 and S2.
In the CSF, taken from S1, routine parameters as well as neuro lament light, total tau and phosphorylated tau were unrevealing.
In S1 and S3, standard MRI (Siemens Avanto; 1,5 Tesla) of the brain and the entire spinal cord was normal without cerebellar or spinal cord atrophy. MRI or other radiological examinations were not available from the other family members.
In VNG (videonystagmography) of S1 and S2 the most prominent nding was pathologic xation suppression of the VOR (Fig. 2). Nerve conduction studies, done in S1 and S2, were conspicuous only for increased F-latencies in the peroneal and tibial nerves of only S1. Needle EMG was normal. In S2, magnetic evoked potentials showed increased central motor conduction times to the distal legs while there were no responses in S1.
Ophthalmologic examination, done in S2, was normal including OCT (optical coherence tomography). Degree of relationship, genotype and clinical phenotype of subjects are given in Table 1 (more detailed information Additional le 2).

Materials And Methods
The study was approved by the local ethics committee, and subjects provided written informed consent (Declaration of Helsinki).
An Illumina Enrichment Kit (Panel-ID 109980) was used to capture the genes under investigation. The generated library was sequenced on a MiSeq sequencer (Illumina). The generation of the analyzed "bam" or "FastQ" les was carried out with the MiSeqReporter Software (v.2.6.2, Illumina, reference genome: hg19).
Variants with a frequency of the rarer allele of > 1% in relevant databases (EXAC, dbSNP, 1000G), synonymous variants and variants that are not within the coding region, outside of the canonical splice sites, as well as variants which are not bi-directionally covered were not considered.
The following genes were analyzed using the Illumina Enrichment Kit: Sanger sequencing con rmed heterozygosity for each of the detected variants in patients S2 and S1 (Additional le 1A, Additional le 1B). Segregation analysis revealed compound-heterozygosity and indicated AR inheritance. S4 showed heterozygosity for the missense variant c.3401A>T, p.(Asp1134Val) assuming her as heterozygous carrier of the possibly disease-causing sequence alteration in PNPLA6 gene, whereas S5 and S3 were homozygous carriers of the splice-site variant c.1635+3G>T. Partial family pedigree is given in Figure 1. (Fig. 1) Taken together, the clinical phenotype of siblings S1 and S2, who shared the same mutations in PNPLA6 gene (c.1635+3G>T/c.3401A>T, p.Asp1134Val) as a compound heterozygosity, consisted of lower limb spastic paraparesis and a marked cerebellar oculomotor disorder accompanied by hypogonadotropic hypogonadism. In contrast, their father, a homozygous carrier of the PNPLA6 variant c.1635+3G>T, was free of complaints but had sustained ankle clonus. His older brother S3 with the same homozygous mutation had gait disturbance with increased stretch re exes and an oculomotor dysfunction of the cerebellar type. The mother S4, a heterozygous carrier of c.3401A>T, p.(Asp1134Val) was free of complaints but she had a somewhat brisk ankle re ex without sustained clonus. Oculomotor function in both parents was normal. 4-aminopyridine 10mg t.i.d (ter in die) was given to subjects S1, S2 and S3 who requested symptomatic treatment. All reported meaningful improvement in dexterity and steadiness of gait. Improvement in cerebellar oculomotor function was documented in S2 where VNG revealed coarse saccadic eye movements on testing the xation-induced suppression of the VOR when the patient was off the drug and visible improvement one hour after 10mg 4-Aminopyridin (Fig. 2).

Discussion
Among the more than 80 different genetic causes of HSP, SPG39, which is caused by alterations in the gene PNPLA6, is likely to be one of the rarest forms. While PNPLA6 has been implicated in a variety of disorders, among which the leading manifestations are often extraneural, spastic paraplegia as the leading or even sole manifestation has been described in only three case reports. PNPLA6 causes different neurodegenerative motor diseases with a wide spectrum of symptoms. Rainier et al.(8) postulated that complex HSP, i.e. further clinical features in addition to spasticity such as distal muscle wasting, is the most common phenotype of PNPLA6 mutations. However, Synofzik et al. (11) showed that besides complex HSP which is "the peak of the iceberg", PNPLA6 can cause a spectrum of neurodegenerative disorders with four main clinical features: ataxia, chorioretinal dystrophy, spasticity and hypogonadism. So far cerebellar oculomotor dysfunction has not been associated with PNPLA6 mutations.
The family reported here has a novel clinical phenotype due to novel compound heterozygous mutations in PNPLA6. Apart from lower limb spasticity with spastic gait, the most salient feature was marked oculomotor dysfunction comprising saccadic pursuit movements, gaze evoked nystagmus, rebound nystagmus and impaired suppression of the VOR by xation. Most peculiar, however, was an intermittent variant of downbeat nystagmus where on downward gaze there was a nystagmus beating upwards (to the forehead). While these ndings are typical for cerebellar dysfunction other cerebellar features such as gait ataxia and ataxia of upper limbs or speech were conspicuously absent. As the sole non-neurologic feature, patient S1 & S2 had low concentrations of blood gonadotropic hormones, indicating mild hypogonadotropic hypogonadism. Taken together, the organ systems affected to the extent of becoming clinically apparent were the cortical motor neurons, cerebellar neurons or connections and, to a much lesser extent, the reproductive endocrine system. Hufnagel et al. (13) found individual tissues to express different levels of NTE activity. Thus, in our patients, NTE activity was likely most prominently reduced in cortical motor and cerebellar neurons.
Pathogenic mutations of PNPLA6 hitherto described all affect genetic sequences that code for the catalytic domain of NTE and that are highly conserved across species. There are a variety of PNPLA6 phenotypes with speci c combinations of clinical features such as spastic ataxia (spasticity & ataxia), Gordon Holmes syndrome (spasticity, primarily gait ataxia & hypogonadism), Boucher-Neuhäuser syndrome (ataxia, hypogonadism & chorioretinal dystrophy) and pure HSP (spasticity). (11) With an upper motor neuron disorder, isolated oculomotor cerebellar dysfunction and a limited degree of hypogonadism we add a novel phenotype to the spectrum of PNPLA6associated disorders.
In our subjects S1 and S2 with compound heterozygosity the onset was earlier and the clinical symptoms more severe than in their father S5 and uncle S3 who both were homozygous for the splice-site variant c.1635+3G>T. This suggests a more deleterious impact of the missense variant c.3401A>T than of the splice-site variant c.1635+3G>T. Although S3 and S5 carried the same novel splice-site variant the older brother S3, with cerebellar dysfunction and spasticity, was clinically more affected than his younger brother S5 who only had brisk re exes. This nding corresponds to previous ndings of Synofzik et al. (11) who found weak genotype-phenotype correlations. Although S3 was 13 years older than S5 he had become symptomatic already at the age of 40, when he was 18 years younger than S5 who still had neither a cerebellar nor gait disturbance. A further possible explanation for the discrepancy between genotype and phenotype in S3 and S5 might be epigenetic factors in uencing the splicing process of PNPLA6 gene. Our ndings mirror those of previous studies where the clinical phenotype was found to be neither dependent on the location within the gene nor the type of the pathogenic variant. (11,13,16) Hence, a genotype-phenotype correlation between our novel PNPLA6 variants and clinical phenotypes cannot be postulated.
Subjects with NTE mutations frequently suffer from progressive upper-more than lower-motor neuron disease leading to the assumption that NTE plays a major role in the integrity of motor axons.(8) Amyotrophic lateral sclerosis is a similar progressive degenerative motor neuron disorder. Rainier et al.(8) considered mutations of the NTE to contribute to the pathogenesis of amyotrophic lateral sclerosis. Interestingly, a cousin of subject S1 and S2 from the paternal line died of bulbar-onset amyotrophic lateral sclerosis at the age of 33. This raises speculations that her amyotrophic lateral sclerosis was based on a PNPLA6 mutation; however, she did not undergo genetic testing.
4-aminopyridine, a potassium channel blocker that is being used in downbeat nystagmus syndrome, a similar disorder of cerebellar nuclei and connections, proved to be clinically bene cial in our patients.

Conclusion
To summarize, our observations expand the already wide spectrum of clinical phenotypes reported with alterations in the PNPLA6 gene. The hitherto undescribed (compound heterozygous) genotype was associated with a novel phenotype comprising prominent dysfunction of cerebellar oculomotor structures and the upper motor neuron system as well as some degree of hypogonadotropic hypogonadism.  Partial pedigree of PNPLA6 family with novel genotype and phenotype. The pedigree structure of the PNPLA6 family; Affected subjects are shown as lled symbols. Compound-heterozygous subjects (S1, S2) presented with the novel genotype and phenotype consisting of a marked cerebellar oculomotor dysfunction, lower limb spasticity and a hypogonadotropic hypogonadism are shown in black. Grey symbols present subjects S3 and S5 suffering from a milder phenotype.

Figure 2
Suppression by xation of the VOR before and after 4-aminopyridine in VNG. Suppression by xation of the VOR, before and after 4aminopyridine. The seated subject is repeatedly turned around his vertical axis with sinusoidal movements while xating a visual target.
The visual target is moving in parallel with his body and head so that, in order to continuously xate the target, the subject needs to keep his eyes in the primary position. Without 4-aminopyridine a high-amplitude drift with corresponding nystagmus is seen in both (horizontal) directions. 60 minutes after 10mg 4-aminopyridine the amplitude of drift and nystagmus is signi cantly reduced. The angular velocity and amplitude of rotation were kept the same in the two examinations.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.