Neurodegenerative Diseases Associated with Mutations in SLC25A46 Neurodegenerative Diseases Associated with Mutations in SLC25A46

Neurodegenerative diseases present substantial clinical challenges. Their processes have been linked with various genetic causes, including mutations of genes encoding proteins associated with mitochondrial dynamics. Biallelic mutations in SLC25A46 have been identified as novel causes of a wide spectrum of neurological diseases with recessive inheritance, including optic atrophy, Charcot-Marie-Tooth neuropathy (CMT) type 2A neuropathy, Leigh syndrome, progressive myoclonic ataxia, and lethal congenital pontocerebellar hypoplasia. SLC25A46 (solute carrier family 25 member 46) is a mem- brane transit protein that is expressed in the mitochondrial outer membrane where it plays a major role in mitochondrial dynamics and cristae maintenance. This chapter presents recent findings on: (1) the clinical heterogeneity of SLC25A46-related neuropathies; (2) the SLC25A46 mutation spectrum and associated genotype-phenotype corre- lation; and (3) pathophysiological functions of SLC25A46 as characterized in cells and mouse models. A better understanding of the etiology of SLC25146 -linked diseases will elucidate therapeutic perspectives. optic atrophy, peripheral neuropathy, progressive ataxia, Leigh syndrome, and lethal congenital pontocerebellar hypoplasia. In SLC25A46-related neurodegenerative diseases, phenotype severity correlates strongly with the magnitude of SLC25A46 level deficit observed.


Introduction
Mitochondria have long been recognized as critical organelles for cellular energy generation. They produce $90% of neuronal adenosine triphosphate (ATP), which is continuously required for maintaining the complex morphology and specialized functions of neurons, including electrical excitability and synaptic transmission [1], and are regenerated continuously in postmitotic neurons through biogenesis. In addition to undergoing the dynamic processes of mitochondrial fission and fusion, mitochondria are transported bidirectionally within neurites, in which they are distributed purposefully, facilitating energy transmission over long distances to meet local demands and, when necessary, undergo controlled degradation by mitophagy [2,3]. Thus, mitochondrial dynamics play critical roles in neuronal homeostasis and survival.
Recent evidence suggests that abnormal mitochondrial dynamics may contribute to both familial and sporadic neurodegenerative diseases [4]. Most proteins related to mitochondrial dynamics are encoded by genes in the nucleus. Mutations in such nuclear-encoded genes can cause monogenic disorders in which mitochondrial dysfunction is unequivocally central to the pathogenesis of the disease. For example, mutations in MFN1/2 and OPA1 cause Charcot-Marie-Tooth neuropathy (CMT) type 2A [5][6][7] and autosomal dominant optic atrophy (ADOA) [8][9][10], respectively. A dominant negative allele of DRP1 was identified in a neonate with a lethal mitochondrial and peroxisomal fission defect associated with abnormal brain development, optic atrophy, and various other congenital anomalies [11]. Defects in proteins involved in axonal transport have also been identified in patients with CMT and related neuropathies [12][13][14]. In addition, there is increasing evidence linking mitochondrial dysfunction to neuronal loss in age-related neurodegenerative disorders, including Alzheimer's disease and Parkinson's disease [15,16].
Recent studies have implicated the 46th isoform of subfamily A of the solute carrier (SLC) family 25, termed SLC25A46, in mitochondrial dysfunction pathology. SLC25A46 is a mitochondrial outer membrane protein that was shown recently to be involved in mitochondrial dynamics, either playing a role in mitochondrial fission or serving as a regulator of mitofusin (MFN)1/2 oligomerization [17,18]. Disorders caused by recessive SLC25A46 mutations were defined recently as a new syndrome (introduced and elaborated in Section 3.1) that has a broad clinical spectrum of neurological phenotypes, including peripheral neuropathy, early-onset optic atrophy, cerebellar degeneration, and congenital pontocerebellar hypoplasia (PCH), with variable ages of onset and severities [17,[19][20][21][22][23][24][25][26]. In this chapter, we will focus on the phenotypic and genetic characteristics of SLC25A46-related neurological diseases and our current understanding of the pathophysiological mechanisms linking dysfunctional SLC25A46 to neurodegeneration.

SLC25 family and the discovery of SLC25A46
SLC25A46 belongs to the solute carrier family 25 (SLC25), a superfamily that contains 53 nuclear-encoded mitochondrial carrier proteins in humans [27]. SLC25 members are characterized by the presence of three tandem repeats of about 100 amino acids, each containing two transmembrane alpha helices linked by a large loop [28]. The mature carrier protein thus consists of six transmembrane helices that form an aqueous pore and have a highly conserved consensus sequence, P-X-[D/E]-X-X-[R/K], at the C-terminal ends of the three odd-numbered transmembrane alpha helices, whose charged residues form salt bridges that close the pore on the matrix side [29]. SLC25 proteins may shuttle a variety of solutes across the mitochondrial membrane to participate in various metabolic pathways [30]. Although common mechanisms of substrate translocation have been proposed, SLC25 members vary greatly in their size, the nature of substrates they transport, the modes of transport employed, and the driving forces they employ [30][31][32].
A number of genetic conditions associated with SLC25 mitochondrial transporters have been characterized biochemically and genetically [33]. SLC25 members mediate a variety of cellular functions, and mutations in SLC25 genes have been linked to various defects, such as carnitine/ acylcarnitine carrier deficiency (OMIM 212138), HHH syndrome (OMIM 238970), aspartate/ glutamate isoform 1 and 2 deficiencies (OMIM 612949, 603471, 605814), congenital Amish microcephaly (OMIM 607196), neuropathy with bilateral striatal necrosis (OMIM 613710), congenital sideroblastic anemia (OMIM 205950), neonatal epileptic encephalopathy (OMIM 609304), and citrate carrier deficiency (OMIM 190315) [33]. These disorders are characterized by specific metabolic dysfunctions related to the role of the particular carrier that has been affected. Most disease-related SLC25 members have been characterized in terms of substrate identification and associated metabolic pathways, with the exception of two orphan SLC25 members, namely SLC25A38 and SLC25A46 [33].
SLC25A46 was first mapped to chromosome 5 by genomic sequence analysis in 2006 [27]. Its location was further refined to chromosome 5q22.1 based on sequence alignment with NCBI's standard reference human assembly sequence, that is, the Genome Reference Consortium Human genome build 38. The largest transcript isoform of SLC25A46 contains eight exons, which encode a 418-amino acid protein. Quantitative real-time polymerase chain reaction (PCR) experiments in rodents have demonstrated variable expression of SLC25A46 in all tissues examined, with the highest levels occurring in the hindbrain, spinal cord, and coronal brain sections containing the corpus callosum, fornix, optic chiasm, thalamus, hypothalamus, midbrain, pons, and cerebellum, with particularly high levels in mouse embryo cerebellum [27].
Given the typical SLC25 molecular structure, the primary sequence of the SLC25A46 protein has been predicted to form six conserved transmembrane alpha helices, TM1-TM6, spanning a region between amino acids 100-418 ( Figure 1) [22]. However, the otherwise highly conserved P-X-(D/E)-X-X-(R/K) consensus sequence characteristic of SLC25 proteins is altered in SLC25A46. Moreover, the N-terminus of SLC25A46 is about five times longer than that of other members of the family ($100 vs. <20 amino acids). These unusual characteristics suggest that SLC25A46 is unlikely to have a conventional metabolite carrier function. Recently, studies have proposed that unlike most SLC25 members that are located in the inner mitochondrial membrane, SLC25A46 may be anchored to the outer mitochondrial membrane where it may act as a regulator of mitochondrial dynamics rather than as a substrate transporter.
3. Clinical phenotypes of SLC25A46-related diseases 3.1. Hereditary motor and sensory neuropathy type VIB (HMSN6B) In 2015, recessive mutations in SLC25A46 in eight patients from four unrelated families of various ethnic origins were first reported. The proband phenotypes encompassed ADOA-like optic atrophy, CMT-like axonal peripheral neuropathy, and cerebellar atrophy with a variable age of onset and disease course ( Table 1) [17]. This new neurodegenerative syndrome is now defined as HMSN6B in OMIM (OMIM: 616505). ADOA and CMT type 2 are hereditary neurodegenerative disorders commonly caused by mutations in the mitochondrial fusion genes OPA1 and MFN2, respectively. However, both diseases lack a genetic diagnosis in up to 60% of patients due to genetic heterogeneity [34,35]. homogenous SLC25A46 mutations in an additional nine patients (age range, 7 days to 28 years) from five unrelated families who presented with neurological phenotypes similar to the core features of HMSN6B. Among these nine patients, eight had optic atrophy (the exception was a patient with an age of onset 23 years) and eight had cerebellar atrophy (the exception was a 28-year-old patient without remarkable cerebellar atrophy or axonal neuropathy) ( Table 1) [19,21,23,26]. Beyond the key clinical features of optic atrophy, peripheral neuropathy, and cerebellar atrophy, the presently documented population of 17 patients with HMSN6B (or an HMSN6B-like condition) exhibited other clinical symptoms sporadically, including ataxia, hypotonia, myoclonus, dysmetria, nystagmus, speech difficulties, abnormal brain imaging, and elevated lactic acid ( Table 1). The clinical manifestations, medical examination findings, and differential diagnoses for these patients were strongly suggestive of a progressive mitochondrial disorder.

SLC25A46-related PCH and Leigh syndrome
A recent study reported the identification of SLC25A46 loss-of-function mutations in four patients from two unrelated families with a diagnosis of severe congenital PCH, leading to very early mortality [22]. Then, two independent groups reported an additional seven patients from three unrelated families with severe PCH associated with truncating mutations of SLC25A46 (Table 1) [24,25].
PCH is a rare, heterogeneous group of prenatal onset neurodegenerative disorders, mainly (but not exclusively) affecting the cerebellum and pons. The current PCH classification scheme includes 10 distinct PCH subtypes defined by clinical features and genetic etiology. PCH1 is distinguished from the other PCH subtypes by its association with spinal muscular atrophy due to spinal motoneuron degeneration; it often leads to early death. All patients with obvious loss-of-function SLC25A46 genotypes in the literature suffered severe lethal congenital PCH, presenting with the phenotypic hallmarks of cerebellar and brainstem degeneration as well as spinal muscular atrophy, respiratory failure, early death, occasional optic nerve atrophy, and axonal neuropathy. Based on these features, SLC25A46-associated PCH could be classified as PCH1, and perhaps a new PCH1 subtype, PCH1D, clinically distinguished from other PCH1 subtypes (mutations in VRK1, EXOSC3, and EXOSC8 are associated with PCH1A, PCH1B, and PCH1C, respectively) [25]. However, the most severe clinical presentation associated with SLC25A46 mutations is probably not restricted to PCH. A homozygous SLC25A46 mutation that resulted in the complete absence of the protein was identified recently in a terminally ill child with progressive brain lesions consistent with those seen in Leigh syndrome (   Table 1). The identification and genetic diagnosis of additional cases in the future may reveal a SLC25A46 mutation pattern.

Genotype-phenotype correlation
A systemic genotype-phenotype analysis of all available cases indicates that phenotype severity correlates strongly with the magnitude of SLC25A46 protein level reduction caused by each Mutations are color coded as follows: red, nonsense and missense mutations that would be expected to destabilize the protein; blue, micro-deletions/insertions/duplications; orange, splice-site mutations; and black, regular missense mutations.

Pathophysiological function of SLC25A46
In the last 2 years, a series of experiments aiming at resolving the function of SLC25A46 and the pathogenesis of SLC25A46-associated diseases were conducted. Primary cultures of skin fibroblasts from HMSN6B patients (including PCH or Leigh syndrome patients) have been studied to investigate the pathophysiology of the diseases. Simultaneously, mutant SLC25A46 alleles or SLC25A46-targeted RNA-interference molecules were transfected into common cell lines (e.g., COS-7, HeLa, HEK293T, and HCT116) to study the consequences of the mutations or protein knock-down in a homogeneous genetic background. Finally, zebrafish, murine, and bovine models have been employed to examine the pathophysiological effects of SLC25A46 mutations in vivo ( Table 2).

Function of SLC25A46 in mitochondrial dynamics
In silico analysis led to the identification of a SLC25A46 homolog in yeast, namely Ugo1. The Ugo1 protein is a modified mitochondrial solute carrier expressed in the mitochondrial outer membrane that operates as a mitochondrial fusion factor and interacts physically with Mgm1 (homolog of human OPA1) and Fzo (homolog of human MFN2) [38]. A succession of studies in cells demonstrated consistently that, like Ugo1, MTCH1, and MTCH2, SLC25A46 also localizes to the mitochondrial outer membrane [17,20,22]. Note: mo represents month; w the week. Table 2.
Clinical phenotypes associated with mutant SLC25A46 animal models.
Knock down of SLC25A46 in various cell lines by different research groups caused mitochondrial hyperfusion and abnormal cristae architecture visualized with fluorescent staining and electron microscopy [17,20,22]. In concordance, in an ultrastructural study of a SLC25A46 knock-out mouse model, we observed enlarged mitochondria with swollen cristae in Purkinje cell (PC) dendrites and sciatic nerves ( Table 2) [39]. Hyperfused mitochondria consequent to SLC25A46 loss was unexpected because loss of Ugo1 function usually results in mitochondrial fission; however, it should be noted that strikingly similar cristae architecture abnormalities from loss of function are common to both genes [38,40,41]. Interestingly, in SLC25A46 mutant Purkinje cell bodies, ring-shaped or C-shaped mitochondria (a rarely reported morphology) were more commonly observed than hyperfused mitochondria [39,42]. Furthermore, mitochondria were found to have an abnormal distribution and impaired movement within mutant Purkinje cells in a primary culture of mouse cerebellar cells [39]. These findings confirm that SLC25A46 plays an important role in the regulation of mitochondrial dynamics, including mitochondrial fusion/fission, distribution, and movement, as well as the maintenance of cristae architecture. Regarding the molecular actions of SLC25A46 in the balance of mitochondrial dynamics, recent research findings present three possible explanations: (1) SLC25A46 may act as an independent pro-fission factor; (2) SLC25A46 may serve as a regulator by interacting with mitochondrial fusion machinery, such as through an association with MFN1/2 oligomerization; and (3) SLC25A46 may regulate mitochondrial dynamics through its functions in lipid transfer between the endoplasmic reticulum (ER) and mitochondria.
In an inter-institution collaborative exploratory study employing immunoprecipitation assays and mass spectrometry analysis, there was no evidence of SLC25A46 interacting with MFN2 or OPA1 in HEK293T cells, but rather SLC25A46 was observed forming a complex with mitofilin that was independent of MFN2 [17]. Furthermore, overexpression of wild-type SLC25A46 protein led to mitochondrial fragmentation and disruption of the mitochondrial network. Thus, SLC25A46 was proposed to act as a pro-fission factor [17]. In contrast, two subsequent studies using similar immunoprecipitation approaches in patient fibroblasts and two cell lines (HEK293T with stable wild-type SLC25A46 expression and LAN5 neuronal cells) showed SLC25A46 interactions with proteins involved in fission and fusion, including MFN1/ 2 and OPA1, as well as with components of the MICOS (mitochondrial contact site and cristae organizing system) complex ( Figure 1) [18,20]. Moreover, decreased expression of SLC25A46 resulted in increased stability and oligomerization of MFN1 and MFN2 in association with mitochondria, thus promoting mitochondrial hyperfusion [18]. In SLC25A46 knock-out mice, two independent mass spectrometry studies yielded opposite results regarding the interaction between SLC25A46 and common dynamic proteins [42,43]. Hence, although it seems reasonable that SLC25A46 would have interaction relationships with MFN1/MFN2 and the MICOS complex similar to those of Ugo1, further studies are needed to resolve its molecular mechanisms given the current conflicting results in the literature.
MFN2 tethers the ER to the mitochondrial network, suggesting that the ER may have a physical relationship with the mitochondrial network [44]. SLC25A46 has also been shown to interact with all nine components of the endoplasmic reticulum membrane complex (EMC) [18,20], an ER protein complex recently shown to be necessary for phospholipid transfer from the ER to mitochondria in yeast. Most mitochondrial phospholipid species were altered dramatically by the loss of SLC25A46, indicating that SLC25A46 provides direct coupling of lipid flux between the ER and mitochondria at outer-inner mitochondrial membrane contacts [20]. Meanwhile, endoplasmic reticulum chaperone BiP (a.k.a. 78 kDa glucose-regulated protein), which acts at the ER-mitochondria interface under stress conditions and is considered as a major regulator of the ER, was down-regulated in SLC25A46 knock-out mice [43]. These findings support the notion that the facilitation of lipid flux at ER contact sites may be a primary function of SLC25A46.
Studies implicating OPA1 and the MICOS complex in the maintenance of cristae architecture are compelling, but it is unclear how they may interact [45]. The observation that SLC25A46 interacts with OPA1 and MIC60, the major MICOS organizer, provides a molecular link that may integrate their functions in modulating cristae architecture [18,20].
It now seems likely that SLC25A46 may possess multiple homeostatic functions in mitochondrial dynamics. Further studies are expected to reveal more refined details of the pathophysiological functions of SLC25A46, such as which domain interacts with dynamic proteins and which domain recognizes and communicates with the ER.

Consequences of SLC25A46 dysfunction on mitochondrial metabolism
Disorganization of cristae leads to disruption of the assembly of the respiratory supercomplexes that mediate oxidative phosphorylation, which reduces the activity of their components (i.e., respiratory complexes I-V) and, thus, diminishes respiration efficiency [46,47]. Mitochondrial metabolism is disrupted in both patients with mutant SLC25A46 alleles and animal models of disrupted SLC25A46. Fibroblasts from patients harboring mutant SLC25A46 showed varying extents of decreased oxygen consumption rate and glycolytic shift (decreased oxygen consumption rate-to-extracellular acidification rate ratio), consistent with the increased lactate levels indicated by magnetic resonance spectroscopy (MRS) [17]. Moreover, siRNA-mediated suppression of SLC25A46 in control fibroblasts phenocopied the basal oxygen consumption defect observed in the cells of patients with mutant SLC25A46, confirming a specific function of SLC25A46 in respiration [20]. Given the specificity of each tissue and each cell type in responding to physiological stresses and mutations, mitochondria isolated from the cerebellum of SLC25A46 knock-out mouse displayed a remarkable decrease in the activities of respiratory complexes I-IV, and thus a dramatically reduced ATP production potential [39]. Taken together, these observations support the hypothesis of an energetic defect being a principal cause of SLC25A46-related neurological disease.

SLC25A46 dysfunction in pathology
Mitochondrial pathobiology has long been linked to the pathogenesis of neurodegenerative diseases, in part because neurons are highly dependent upon mitochondrial metabolism. Autopsy on a pair of deceased siblings who died due to SLC25A46 mutations showed a very small cerebellum, as well as degenerative changes in the inferior olive nucleus. Histology of the cervical spinal cord illustrated loss and ongoing degeneration of spinal motoneurons in the medial anterior horn at all levels from the cervical to the lumbar region. Rectus femoris muscle biopsies showed severely atrophied muscle fibers of all types, without type grouping [48].
For now, most anatomical and histological analyses for SLC25A46-related diseases are based on animal models. Four SLC25A46 mutant mammalian models have been discovered or designed, including a calf model and 3 mouse models with different genetic backgrounds ( Table 2). Turning calf syndrome, a neurodegenerative disease, is characterized during massive inbreeding in cattle. Genetic study determines that this disease resulted from a single substitution in the coding region of the SLC25A46 [43]. All affected calves manifested early onset of ataxia, especially of hind limbs, and paraparesis (2-6 weeks old). Despite symptomatic care, nervous symptoms progressed over the next months, leading to repetitive falls and ultimately resulting in permanent recumbency and inevitably euthanasia. Characteristic degenerative microscopic lesions were found in both gray matter (brainstem lateral vestibular nuclei and spinal cord thoracic nuclei) and white matter (dorsolateral and ventromedial funiculi of the spinal cord) regions of the central nervous system, as well as demyelination of some peripheral nerves [49]. Electron microscopy confirmed this neuropathy phenotype, revealing discrete demyelinating lesions and some enlarged nodes of Ranvier.
Three SLC25A46 knock-out mouse models with different genetic backgrounds, including FVB/ N, C57BL/6J, and B6D2, were generated, respectively ( Table 2). In spite of various mutation positions and sizes, three mouse lines displayed very similar phenotypes, including growth delay, progressive ataxia, optic atrophy, short life span, which recapitulated the pathological state in human. Further histopathologic studies have shown tissue-and cell-specific lesions in both the central nervous system and peripheral nervous system ( Table 2).
Although macroscopic examination showed no overt abnormalities in the gross anatomy of mutant brain, histological staining revealed markedly reduced cerebellums, with Purkinje cells (PCs) that had stunted dendrites and were reduced in number. Degeneration (evidenced by Fluoro-Jade C dye) was selectively present in mutant PCs [39]. Examination via electron microscopy (EM) revealed that degenerated PC dendrites exhibited disorganized cytoskeleton, often containing remnants of mitochondria and other organelles. Numerous atypical mitochondria with cytoplasmic inclusions were found both in the soma and dendrites of PCs. In addition, a significant reduction in vGlut1 and vGlut2 immunoreactivity both in PCs and molecular layer indicated a paucity of glutamatergic synapses in mutant mice [42]. Apart from PCs, degenerative signals were also aggregated in the vestibular nucleus of brainstem, deep cerebellar nuclei, the striatum, the corpus callosum, and the spinal cord, but not in other parts of the brain [39]. The neurodegeneration was associated with astrogliosis and microgliosis in the cerebellum and spinal cord, indicating high levels of neuroinflammation [39,42]. These observations suggest that although SLC25A46 mRNA is transcribed in most neural tissues, SLC25A46 may have a tissue-or cell-specific function. Interestingly, mice with conditional knock-out MFN2 or DRP1 mutations showed a similar phenomenon, with major changes in cerebellar Purkinje cells, whereas granule cells seemed to be spared [50]. It is not yet known why specific neurons are selectively affected by mitochondrial dysfunction. It could be that a large neuronal size makes particular neurons susceptible; Purkinje cells are one of the largest neuron types in the brain with long axons and extremely extensive dendritic branches. Their complex architecture requires the transport and distribution of highly organized organelles, including mitochondria.
Aged SLC25A46 mutant mice displayed enhanced hind limb clasping reflex and muscle atrophy, suggesting potential peripheral neuropathy. Acquiring compound muscle action potentials (CMAPs) reduced in mutant sciatic nerve measured by electromyography (EMG) in vivo [39]. Mutant peripheral nerves exhibited sporadic degenerative lesions with local macrophages containing lipid debris and signs of demyelination. In addition, the size of individual endplates in mutant mice was significantly reduced. Retarded neuromuscular junction maturation and improper innervation, early hallmarks of CMT2D, were also documented in mutant muscle [42]. All of these alterations are indicative of peripheral neuropathy.
Optical coherence tomography (OCT) scanning on retina for live mice revealed that although the optic discs were grossly normal in terms of retinal appearance, retinas were thinner in aged SLC25A46 mutant mice [39]. Further quantitative measurements indicated that ganglion cell complex (GCC) thicknesses, which includes the nerve fiber layer (NFL), ganglion cell layer (GCL), and inner plexiform layer (IPL), were significantly reduced in adult mutant mice.
Retinal and these reductions were associated with retinal ganglion cell loss and atypically small optic nerve axons with reduced neurofilament expression, as well as some axons that exhibited signs of degeneration and demyelination [39,42]. Pax6+ and GAD65+ GABAergic amacrine cells-both of which form synapses with retinal ganglion cells-were also significantly reduced. These pathological changes are in line with the phenotypic features of ADOA.
Ultrastructural studies revealed dysmorphic mitochondria in both the central and peripheral nervous systems. Numerous enlarged and round mitochondria with abnormal cristae were found in Purkinje cell dendrites, while ring-or C-shaped mitochondria were commonly observed in soma. Peripheral nerve axons also had abnormal round, fused, and aggregated mitochondria in myelinated and non-myelinated fibers [39,43].
Given the degeneration in long peripheral axons and distal optic nerves of SLC25A46 knockout animal models, the aforementioned findings support the idea that neurons with long axons or complicated dendrites are more sensitive to abnormal mitochondrial dynamics. Similar to the findings in mutant Purkinje cells, this sensitivity could also be due to the impaired transport of hyperfused mitochondria along axons and dendrites, probably due to their abnormal size and/or reduced ATP availability in the distal portions of long axons secondary to mitochondrial dysfunction. Further studies are needed to clarify this point.

Conclusion
SLC25A46 plays a critical role in mitochondrial dynamics and the maintenance of mitochondrial cristae, which are particularly important in neurodevelopment and neurodegeneration. Loss of SLC25A46 function causes a wide spectrum of neurodegenerative diseases, including optic atrophy, peripheral neuropathy, progressive ataxia, Leigh syndrome, and lethal congenital pontocerebellar hypoplasia. In SLC25A46-related neurodegenerative diseases, phenotype severity correlates strongly with the magnitude of SLC25A46 level deficit observed.