ALS2 (Alsin)

ALS2 is a rare, recessively inherited, juvenile-onset form of ALS. It is predominantly a UMN syndrome characterized by slowly progressive spasticity beginning in the lower limbs, with gradual extension upwards to involve the upper limbs and bulbar musculature. Truncation mutations in the ALS2 gene (coding for alsin) were found in kindreds from Tunisia and Kuwait, and also in a Saudi Arabian family with juvenile PLS. Mutations in alsin have also been found in several families with a syndrome described as infantile-onset ascending hereditary spastic paraplegia (IAHSP). Most mutations generate truncated alsin products, which are unstable leading to loss of function. Homozygous missense mutations have also been described, in association with alsin protein mislocalization, leading to loss of function. There are no published reports of the neuropathology of ALS2 patients.

The alsin gene is ubiquitously expressed in the CNS (especially cerebellum) and periphery (especially kidney). Alsin appears to protect cultured neurons from mutant SOD1-mediated toxicity ; its overexpression promotes hippocampal neurite outgrowth through Rac1 activation. Alsin may also have a role in preventing glutamate mediated excitotoxicity by appropriate expression of the Ca ^{2 +} impermeable Glur2 subunit of AMPA receptors.

The best characterized properties of alsin relate to its functions in endosomal dynamics. Present in the cytoplasm, alsin is relocalized to endosomes by Rac1. Alsin then acts as a guanine-nucleotide exchange factor (GEF) to activate the GTPase Rab5. Rab5 is important for endocytosis, endosome fusion and endosomal trafficking. Alsin knockout mice have relatively subtle motor phenotypes and pathological changes are limited to mild corticospinal tract degeneration, suggesting that ALS2 may be a distal axonopathy. However, in vitro studies of alsin null neurons and fibroblasts confirm reduced endosomal trafficking and altered endosomal fusion. Furthermore, the rare missense mutations of alsin appear to act through loss of function due to impairment of Rac1-mediated relocalization to endosomes.

ALS4 (Senataxin)

ALS4 is a rare, nonfatal, autosomal dominant, juvenile-onset dHMN characterized by limb weakness, severe muscle wasting and pyramidal tract signs including brisk reflexes and extensor plantar responses. Bulbar and respiratory muscles are spared and lifespan is unaffected. The mean age of onset is 17 years. Symptoms begin with difficulty walking followed by weakness and wasting of intrinsic hand and foot muscles. By the fourth to fifth decade, patients display significant proximal weakness and are often wheelchair-bound. By the sixth decade, many lose hand function entirely. Sensation remains normal both clinically and electrophysiologically, although pathological studies have shown dorsal column degeneration in addition to corticospinal tract degeneration and anterior horn cell loss.

ALS4 was first characterized in four kindreds from America and Europe and linked to a locus on Chr 9q. Dominant missense mutations in the SETX gene encoding senataxin were subsequently found. In all, as of 2013 only seven mutations in senataxin affecting 64 individuals were listed on the ALS Online Genetics Database (ALSoD). These affect the N-terminus and helicase domain. A dominant SETX mutation has also been found in a family with dominantly transmitted proximal SMA (ADSMA). Interestingly, mutations of SETX cause ataxia-oculomotor apraxia 2 (AOA2), although in these cases the mutations are recessive and mostly truncations. A toxic gain of senataxin function may be responsible for ALS4, while loss of function may lead to AOA2.

The full range of functions of senataxin remains to be determined. It seems clear that it has a putative DNA/RNA helicase domain and demonstrates punctate expression in the nucleus and diffuse expression in the cytoplasm. Senataxin is important in preventing oxidative damage to DNA, interacts with a variety of RNA processing proteins (including RNA polymerase II and SMN) and influences gene transcription and mRNA splicing. Senataxin also shows homology to another putative DNA/RNA helicase, immunoglobulin μ-binding protein 2 (IGHMBP2). Notably, recessive (probably loss of function) mutations of IGHMBP2 cause a childhood onset LMN disease, spinal muscular atrophy with respiratory distress type 1 (SMARD1).

Senataxin knockout mice have no motor phenotype but the observation that male animals are sterile and females are subfertile led to detailed studies demonstrating critical roles for senataxin in crossing over in meiosis, meiotic sex chromosome inactivation, and DNA damage response.

In spite of this research on loss of senataxin function, which is clearly relevant to AOA2, the mechanisms by which dominant mutations cause ALS4 remain relatively unexplored. In vitro studies demonstrate a role for senataxin in neuritogenesis through fibroblast growth factor 8 and suggest that the dominant mutants lack this ability. Thus, haploinsufficiency is a possible mechanism of disease in ALS4.

Finally, caution should be taken in interpreting SETX sequencing results. Several variants described in the literature as ALS4-causing mutations may actually be benign polymorphisms, although it remains possible that they may contribute to disease risk or phenotype without being completely penetrant.

ALS5/SPG11 (Spatacsin)

ALS5, on chromosome 15q, was originally described in several kindreds from Tunisia and Germany with autosomal recessive jALS. Families were linked to chromosome 15q15.1-21.1. Patients develop progressive gait disturbance, presenting with mixed UMN and LMN features between 8 and 18 years of age. Dysarthria begins after 3 to 4 years and survival is generally 10–25 years from symptom onset. In parallel studies, a form of autosomal recessive hereditary spastic paraplegia (ARHSP) with thin corpus callosum (TCC), designated SPG11, was reported to be caused by mutations in a novel gene, labelled spatacsin. This is a transmembrane protein that is phosphorylated after DNA damage; its function is not well understood. Of interest, Orlacchio et al. collected a series of 25 families (from Italy, Japan, Turkey, Canada) with recessively inherited jALS; 10 of the 25 kindreds were found to have mutations in spatacsin . While we do not know if the gene defect in the original ALS5 families has been confirmed to be in spatacsin , the general premise now accepted is that the ALS5 gene is indeed spatacsin .

ALS6 (Fused in Sarcoma/Translocated in Liposarcoma)

This form of ALS most commonly arises in adults with a typical, if somewhat aggressive, phenotype. However, it is now well documented to arise in individuals in the second decade, usually with a devastatingly rapid course. These cases are all caused by mutations in the gene encoding FUS (fused in sarcoma), a multifunctional RNA- and DNA-binding protein of 525 amino acids. FUS is normally located almost exclusively within the nucleus but is observed to exit to the cytosol after exposure to some cellular stresses (e.g. hyperosmolar shock). Mutant FUS is observed to mislocalize to the cytosol where it sometimes participates in forming aggregates. Rapidly progressive jALS has been associated with several different FUS gene mutations; among the most aggressive is a particular dominant missense variant that substitutes lysine for proline at codon 525, thereby disrupting the nuclear localization signal. At autopsy these cases are associated with distinctive basophilic intranuclear aggregates.

ALS16 (Sigma-1 Intracellular Receptor) and SPG18 (ERLIN2)

A homozygous missense mutation (E102Q) in SIGMAR1 (encoding sigma nonopioid intracellular receptor 1, Sig-1 R) was comparatively recently reported in a consanguineous Saudi family with autosomal recessively inherited jALS. The phenotype was slowly progressive, with predominantly UMN involvement.

Sig-1 R regulates K ⁺ channels and is involved in Ca ²⁺ signaling through IP3R. It is a receptor for a variety of ligands including steroids, psychostimulants, and haloperidol. It also has chaperone activities in the ER, implicating the unfolded protein response in neurodegeneration. In vitro studies showed that the E102Q mutation abrogates the ability of SigR1 to suppress apoptosis induced by ER stress. The E102Q mutation occurs in a predicted transmembrane domain and subcellular fractionation suggests that the mutation causes a shift to lower density membrane fractions where mutant protein forms detergent-resistant complexes.

A role for SigR1 in sporadic ALS has also been suggested by pathological studies showing altered distribution of SigR1 at C-boutons (excitatory postsynaptic densities at spinal cord motor neurons) and cytoplasmic accumulation of SigR1.

The homozygous E102Q mutation is the only clearly pathogenic mutation yet described. Heterozygous 3′ untranslated region (UTR) variants described in familial ALS-FTLD have been identified but it remains unclear whether these are actually benign polymorphisms or whether they do contribute to disease risk. No animal studies of these mutations have yet been conducted, but a recent study in mice showed that the native Sig-1 R protein is located exclusively in motor neurons, and that knockout causes motility problems. Furthermore, the Sig-1 R agonist PRE-084 has recently been shown to be protective in both SOD1-G93A and wobbler models of ALS, supporting a loss of function as the mechanism by which Sig-1 R mutation may cause disease.

In the context of ER stress it is worth noting the recent discovery of recessive mutations in ERLIN2, an endoplasmic reticulum protein that is associated with lipid rafts, in an inbred family with juvenile PLS. Patients were normal at birth but developed impaired crawling at 8 months of age, toe walking at two years with progressively impaired mobility such that they were bedridden by just 15 years of age. In addition to severe generalized spasticity by end stage, patients also developed severe dysarthria and pseudobulbar palsy, skeletal deformities including kyphoscoliosis, saccadic eye movements, and moderate distal limb wasting. Brain PET scan showed hypometabolism in the temporal and parietal lobes. Erlin2 is similar to Sig-1 R in that it is a component of ER lipid rafts. Splice junction mutation of the ERLIN2 gene results in nonsense-mediated decay of aberrantly spliced transcripts causing a loss of function. It will be interesting to determine whether Erlin2 and Sig-1 R interact and how they contribute to distinct clinical phenotypes.

jALS Linked to Chromosome 6p25 and 21q22

A family from Utah has been reported in which four of six children from a consanguineous first cousin marriage had jALS. The phenotype was characterized by prominent upper motor neuron signs as well as amyotrophy of limb and bulbar musculature. Atypical features included mild ptosis, gynecomastia in one patient, and mild distal impairment of vibration sensation. Onset was between 4–10 years of age with slow progression to loss of ambulation after 10 years. Homozygosity mapping identified linkage to two loci (6p25 and 21q22) but no disease-causing mutation has yet been identified.

Brown-Vialetto-Van Laere Syndrome

The Brown-Vialetto-Van Laere (BVVL) and Fazio Londe (FL) syndromes are pontobulbar palsies usually of juvenile onset. The conditions appear identical, except that BVVL syndrome cases also have hearing loss. The occurrence of both phenotypes within the same kindreds demonstrate that BVVL and FL syndromes are one disease (hereafter referred to as BVVLS). The condition is extremely rare—a 2012 literature review found only 101 reported cases. It is both fascinating and clinically quite significant that some inherited cases of BVVLS are eminently curable, emphasizing the importance of a molecular diagnosis (see below).

The clinical presentation of BVVLS is usually with bulbar palsy (92%), sensorineural hearing loss (81%), facial weakness (77%), and respiratory compromise (64%). The age of onset varies widely from infancy to adulthood, with around 80% of cases having onset before 18 years of age with a mean age of onset of 8 years. Early motor milestones are reported to be normal. Two thirds of patients are female. Around half of teenage-onset patients have muscle weakness and UMN signs. Around a quarter of patients may have oculomotor palsies (cranial nerves III, IV, and VI). Less commonly involved are the trigeminal and optic nerves and the diaphragm. Around 20% may have limb LMN signs. Ataxia and tremor are seen in 10% of cases, while behavioral changes are seen in 4%. Disease progression can vary markedly between and within families. Survival varies from a few months to 32 years (mean survival 5 years) from presentation and is worst in those with early onset. Death is usually from respiratory insufficiency.

Clues to the genetic underpinnings of some cases of BVVLS came from the apparent autosomal recessive pattern of inheritance seen in consanguineous families of Arabic and Pakistani origin. Using linkage analysis (autozygosity mapping) and candidate gene sequencing, homozygous and compound heterozygous mutations in C20orf54 ( SLC52A3 ) were identified in these cases. SLC52A3 encodes an intestinal riboflavin transporter. These disease-linked mutations impair riboflavin absorption, as evidenced by low serum flavin and acylcarnitine. Amazingly, striking clinical improvement was seen in these patients following oral or intravenous riboflavin supplementation, with improvement seen from a few days to several months after initiation of therapy. Some patients were able to come off mechanical ventilation. Of note, mutations in related riboflavin transporter genes SLC52A2 and SLC52A1 have now also been detected in other BVVLS cases, with similar responses to riboflavin. Thus, vitamin B2 deficiency seems to be the common pathogenic mechanism in these BVVLS cases. Riboflavin (vitamin B2) is an essential component of the nucleic acid cofactors FAD (flavin adenine dinucleotide) and FMN (flavin mononucleotide) and accordingly is important in numerous cellular functions including electron transport and a wide range of reactions including synthesis of niacin and folate.

Ubiquilin1

A missense mutation of UBQLN1 has also been described in one individual with the BVVLS phenotype whose motor neuron symptoms began before the age of 10 years. This mutation may impair the ability of ubiquilin1 to carry out its normal function of shuttling ubiquitinated proteins for degradation through the proteasome or autophagosomes.

Madras Motor Neuron Disease

This rare form of jALS seen predominantly in South India was first described in 1970 in the city of Madras (now known by its precolonial name of Chennai). Madras motor neuron disease (MMND) has also rarely been reported in Thailand, Korea, and Turkey. There is phenotypic overlap with BVVLS though a recent study has shown no mutations in the riboflavin transporter genes linked with BVVLS. A genetic cause has yet to be identified but is a likely contributor given that around a quarter of patients have a family history of MMND (familial MMND [FMMND]). Incidence in males and females appears equal in a case series of 116 patients from Southern India seen over 36 years (1971–2007). Patients with MMND typically present at a mean age of 16 years (range 1–39 years); deafness often heralds the onset.

The features that set MMND apart from typical ALS are a high incidence of hearing loss (seen in 90% of patients, and the initial presenting feature in 50%) and optic atrophy (25% of MMND patients, 100% of MMND variant [MMNDV] patients). The overall picture is of predominantly LMN disease with striking wasting, particularly of proximal and distal upper limbs, and lower cranial nerve involvement causing profound facial weakness and bulbar palsy. A minority of patients (6%) were observed to have schizophrenia. UMN signs are seen in most patients, more commonly in the lower limbs, with an extensor plantar reflex in 62% of patients. Survival is of the order of 30 years from onset, though females notably have a shorter survival of around 20 years.

EMG findings are typical of ALS with active and chronic denervation and reinnervation findings in most cases. Muscle biopsy demonstrates neurogenic atrophy with no evidence of mitochondrial myopathy. Brain imaging is normal. Visual evoked potentials in those with MMNDV are consistent with optic atrophy.

No mutations in any known ALS genes have so far been found in MMND, including the riboflavin transporter genes implicated in BVVLS. Although consanguinity is not clearly apparent in the majority of familial cases, most of these families demonstrate an autosomal recessive pattern of inheritance and the parents of these families tend to come from the same regional location in south India, suggesting that a closed gene pool influenced by unknown environmental factors could underpin the disease.

Berardinellie-Seip Congenital Lipodystrophy Type 2 Gene

Recessive loss of function mutations in the Berardinellie-Seip Congenital Lipodystrophy type 2 ( BSCL2 ) gene were originally found to cause congenital generalized lipodystrophy type 2. Dominant mutations in BSCL2 were subsequently identified in several Austrian kindreds demonstrating a variety of motor neuron phenotypes, including dHMN with pyramidal features (dHMN5), Silver syndrome (late childhood onset of hand weakness and limb spasticity, in association with pes cavus, also designated SPG17) and HSP. BSCL2 mutations are among the commonest known causes of the dHMN phenotype. The mean age of onset is typically 15 years with a range of 2–40 years. LMN features tend to start in the hands with wasting of the intrinsic palmar muscles (e.g. thenar eminence). Most patients eventually develop lower limb features including foot deformities. Reflexes may be brisk. Some sensory involvement is seen in the form of reduced vibration sense in the lower limbs. Progression is very slow over decades. Nerve conduction studies essentially show a motor axonal neuropathy.

BSCL2 encodes seipin, an integral ER membrane protein. While the originally defined cases of BSCL2 mutations were recessively inherited, it is now appreciated that some dHMN-associated mutations are dominantly transmitted. These appear to enhance glycosylation and are more prone to aggregation. Activation of the unfolded protein response has also been seen in cells expressing mutant seipin in vitro . Transgenic mice expressing mutant seipin show enhanced activation of the autophagy pathway. These observations implicate impaired protein homeostasis as a key component of the pathogenesis of dHMN5. As an interesting point, it has also recently been shown through exome sequencing that one of the original CMTX3 families linked to chromosome X harbors a mutation in BSCL2. Determining how BSCL2 mutations can result in such diverse phenotypes (from lipodystrophy to dHMN to CMT) will likely yield important therapeutic targets.

DCTN1

The G59S dominant missense mutation in the DCTN1 (Dynactin) gene is associated with a rare form of dHMN with vocal cord paresis characterized by progressive facial and arm weakness, with later leg weakness. Onset may be in early adulthood, but UMN features are generally not seen, meaning that this is not a true jALS. However, other missense mutations in DCTN1 are actually associated with a typical ALS phenotype, with disease duration of 4 to over 9 years, and also a frontotemporal dementia (FTD) phenotype. However, the onset in these typical ALS cases is from age 48–64 years, indicating that this is not a juvenile disease. DCTN1 encodes the p150 glued subunit of dynactin, which is involved in retrograde axonal transport.

HSPB1

Dominant or recessive missense mutations in Heat Shock Protein 27 ( HSPB1 ) (encoding HSP27, a chaperone) are associated with dHMN in families of European ancestry. Onset is between 21 and 54 years. Distal lower leg involvement is seen early with arms involved 5–10 years after onset. Brisk reflexes are often noted.

Other dHMN Loci

Autosomal dominant dHMN with UMN features has also been linked to three other genetic loci. Firstly, an Australian family has been linked to Chr 7q34–36. Age of onset is 10 years (range 3–40 years) with initial onset in the lower limbs causing impaired gait. As well as wasting, there is increased lower limb tone and extensor plantar responses in around half of patients. There may be slight sensory symptoms. Progression is slow. Axonal loss is seen on biopsy.

Secondly, an Italian family with dominantly inherited dHMN with pyramidal features has been linked to chromosome 4q34.3-q35.2. Onset is from age 25 to 40 years with spastic gait, brisk reflexes, distal leg wasting, and weakness. Again, progression is slow. Nerve conduction studies demonstrate axonal neuropathy.

Thirdly, autosomal recessive Jerash type dHMN (seen in Jordanians) has been linked to chromosome 9p21.1-p12. Onset is age 6–10 years. Weakness is distal at onset and seen initially in the legs and then in the upper limbs. There is wasting in the hands and feet. Reflexes tend to be brisk at the knees and absent at the ankles with extensor plantars.

Juvenile-onset HSP with Lower Motor Neuron Involvement

Some early onset HSP variants demonstrate distal amyotrophy or lower motor neuropathy and may therefore be described as jALS. The HSP variants associated with peripheral neuropathy are HSP types SPG2, 3 A, 5, 6, 7, 10, 25, 27, 30, 31, 55, and 56; other related genes in which a usual HSP phenotype is coupled with LMN dysfunction are SPOAN, Cct5 (epsilon subunit mutation), and mitochondrial ATP6 gene mutation. Variants associated with distal amyotrophy are HSP types 5, 10, 14, 15, 17, 20, 26, 30, 38, 39, 41, 43, 55, and 57. Distal amyotrophy is also rarely seen in HSP types 3A and 4. A full description of all of these entities is beyond the scope of this review. However, we offer here a description of the better-understood variants.

SPG3A (Atlastin)

SPG3A (atlastin) is essentially an uncomplicated HSP though patients can sometimes develop an axonal motor neuropathy and/or distal muscle wasting. Sensory neuropathy has also been described and interestingly SPG3A is allelic with hereditary sensory neuropathy I. Symptoms usually begin in childhood (and may not progress), with a mean age of onset of four years (range from one year to the seventh decade). Adolescent and adult onset cases tend to progress insidiously. Nonetheless, wheelchair dependency is rare. SPG3A is caused by dominant mutations in ATL1 , which encodes atlastin. Over 40 ATL1 mutations have been described, mostly missense. Atlastin is most highly expressed in the CNS, especially layer V of the cortex, which is where the primary motor neurons reside. Atlastin is a dynamin-like GTPase that mediates endoplasmic reticulum (ER) membrane fusion and interacts with a number of other proteins mutated in HSP, including spastin, REEP1, and NIPA1 (all of which may be associated with LMN degeneration). Atlastin has also been shown to be required for Golgi apparatus morphogenesis.

Zebrafish atlastin has been shown to control motility and spinal motor neuron architecture through inhibition of bone morphogenetic (BMP) signaling. A study in Drosophila has recently shown that atlastin interacts with valosin containing protein (VCP), a multifunctional AAA-ATPase with roles in the ubiquitin proteasome system, membrane fusion, and DNA repair. This is particularly interesting given that VCP mutations are a rare cause of ALS.

SPG4 (Spastin)

SPG4 (spastin) is the most common autosomal dominant form of HSP, accounting for around 40% of cases. The SPG4 gene encodes the protein spastin, which is important for microtubule severing and thus critical for axonal organelle transport. Mutant spastin lacks the normal microtubule severing capacity, and its binding to wild-type spastin may inhibit the normal allele, thus having a dominant negative influence on wild type spastin.

SPG20 (Spartin)

SPG20 is an autosomal recessive, complicated HSP caused by mutations in spartin. The phenotype is called Troyer syndrome. Spartin has some similarity at the N-terminus to spastin and is homologous to proteins involved in the morphology and trafficking of endosomes. It localizes to mitochondria and may have roles in endocytosis, vesicle trafficking, or mitogenic activity. Spartin also inhibits BMP signaling and interacts with a number of proteins involved in ER stress, ubiquitination, and the nucleolar protein nucleolin.

SPG39 (Patatin-like Phospholipase Domain-containing Protein 6, Also Known as Neuropathy Target Esterase)

Rare autosomal recessive families with a Troyer-like phenotype but without developmental, cognitive, skeletal, extrapyramidal or cerebellar features have also been described. Screening for spartin mutations was negative in these kindreds, but linkage and candidate gene studies identified homozygous and compound heterozygous mutations in the gene encoding neuropathy target esterase (NTE). Onset is in childhood with progressive lower limb spasticity and weakness, and distal upper and lower limb wasting and weakness. Electrophysiology demonstrates motor axonal neuropathy and MRI scans identify spinal cord atrophy.

NTE is a phospholipase B localized to ER membranes, which may reduce cytotoxic phosphatidylcholine. Its activity may be important in preventing toxic organophosphorus compound induced delayed neurodegeneration (OPIDN). It may also function to regulate cyclic AMP-dependent protein kinase. The offending SPG39 mutations in NTE appear to result in a loss of function, either by affecting the catalytic domain or by causing a frameshift and protein truncation. It can be argued that the NTE mutations provide an intriguing link to the organophosphate poisoning hypothesis of ALS in Gulf War veterans, consistent with the view that rare, ALS-specific paraoxonase variants (PON1, 2, and 3) contribute to ALS risk.

Spastic Paraplegia, Optic Atrophy, and Neuropathy

Spastic Paraplegia, Optic Atrophy, and Neuropathy (SPOAN) is a complicated autosomal recessive HSP with juvenile onset described in a large inbred Brazilian family. The earliest noted abnormality is congenital optic atrophy, followed in infancy by the onset of progressive spastic paraplegia. Motor and sensory axonal neuropathy develops in late childhood/early adolescence causing significant weakness in grip strength and reduced sensation. Dysarthria starts in the third decade of life. Other features are an exaggerated startle response, progressive joint contractures, and spinal deformities. Patients are wheelchair dependent by age 15. No mental dysfunction was described and brain MRI is normal. Although the family was convincingly linked to a genetic locus at chromosome 11q13, no mutation has yet been identified.

SPG57 (Tyrosine-kinase Fused Gene)

SPG57 was recently described in two Indian brothers in their teens. These boys had identical clinical histories with normal early milestones in the first year, but difficulty walking due to spasticity, visual impairment due to optic atrophy, and peripheral muscle wasting in all limbs, indicating lower motor neuron degeneration. Though there were no sensory symptoms or signs, nerve conduction studies demonstrated sensory axonal neuropathy. SPG57 is caused by a homozygous missense mutation, c316C>T substituting cysteine for arginine at codon 106 in the Trk-fused gene (TFG).

TFG is a highly conserved regulator of protein secretion in the endoplasmic reticulum. It normally functions as an oligomer, but the disease-linked mutation is in a coiled-coil domain and disrupts the ability of TFG to oligomerize. Interestingly, a heterozygous TFG mutation (substituting leucine for proline at codon 285) causes autosomal dominant hereditary motor and sensory neuropathy (HMSN) with proximal dominant involvement. The P285L mutation is in the C-terminal P/Q rich region. The implication is that this mutation may differ from that associated with SPG57 by virtue of an undetermined toxic gain of function effect. Intriguingly, histological studies demonstrate TDP-43 positive skein-like inclusions similar to those found in ALS. TFG positive inclusions are also seen, though these do not always coaggregate with TDP-43 and TFG inclusions were not seen in sporadic ALS cases.

SPG31 (Receptor Expression-enhancing Protein 1)

Mutations in the REEP1 gene (encoding (receptor expression-enhancing protein 1)) cause diverse phenotypes including HSP at one extreme or predominant LMN axonopathy on the other. This phenotypic diversity is reminiscent of BSCL2/dHMN. REEP1 is involved in ER dynamics, functioning in vesicle transport, ER formation, and interaction with the cytoskeleton. It remains unclear why REEP1 mutants so selectively influence viability and function of motor neurons, although it is striking that other ER proteins (above) and proteins that influence vesicle transport are also implicated in motor neuron disease.