|Home | About | Journals | Submit | Contact Us | Français|
The mitochondrial protein SLC25A46 has been recently identified as a novel pathogenic cause in a wide spectrum of neurological diseases, including inherited optic atrophy, Charcot-Marie-Tooth type 2, Leigh syndrome, progressive myoclonic ataxia and lethal congenital pontocerebellar hypoplasia. SLC25A46 is an outer membrane protein, member of the Solute Carrier 25 (SLC25) family of nuclear genes encoding mitochondrial carriers, with a role in mitochondrial dynamics and cristae maintenance. Here we identified a loss-of-function mutation in the Slc25a46 gene that causes lethal neuropathology in mice. Mutant mice manifest the main clinical features identified in patients, including ataxia, optic atrophy and cerebellar hypoplasia, which were completely rescued by expression of the human ortholog. Histopathological analysis revealed previously unseen lesions, most notably disrupted cytoarchitecture in the cerebellum and retina and prominent abnormalities in the neuromuscular junction. A distinct lymphoid phenotype was also evident. Our mutant mice provide a valid model for understanding the mechanistic basis of the complex SLC25A46-mediated pathologies, as well as for screening potential therapeutic interventions.
Neurodegenerative diseases encompass a wide range of clinical conditions remaining incurable while the genetic and molecular basis of most of these conditions remains unknown due to their heterogeneity and the lack of animal models. Mutations in nuclear genes encoding mitochondrial proteins have recently emerged as novel causalities in diseases affecting the nervous system. SLC25A46, a novel mitochondrial protein, has recently been identified as a pathogenic target in a wide spectrum of rare genetic neurological diseases, including optic atrophy, Charcot-Marie-Tooth type 2, Leigh syndrome, progressive myoclonic ataxia and lethal congenital pontocerebellar hypoplasia. Following a genetic approach, we identified a novel neurological mouse model caused by a functional mutation in the Slc25a46 gene. Our SLC25A46 mutant mice constitute the first genetic animal model for this disease and represent an ideal tool for elucidating the underlying molecular mechanisms. The aim of this study was to characterize the complex phenotype displayed by the mutant mice in order to understand the mechanistic basis of the SLC25A46-mediated pathologies.
Mitochondria are highly dynamic organelles with a central role in a plethora of cellular processes, including ATP synthesis through oxidative phosphorylation (OXPHOS), metabolism, apoptosis, and reactive oxygen species generation, with important implications for neurodegenerative diseases. Recent discoveries show that mitochondrial shape and bioenergetics are tightly linked. Mitochondria have a characteristic structure as they are surrounded by two membranes: the inner membrane (IM) and the outer membrane (OM). The IM is further subdivided into the inner boundary membrane (IBM), which is adjacent to the OM and the cristae that protrude into the matrix. IM proteins involved in mitochondrial fusion and protein translocation are preferentially located in the IBM, whereas proteins involved in oxidative phosphorylation are enriched in the cristae membrane. A variety of proteins regulate cristae biogenesis, with optic atrophy 1 (OPA1) and the mitochondrial contact site (MICOS) complex being master regulators in this process . The OM separates the mitochondria from the cytosol, yet it allows the passage of metabolites through the voltage-dependent anion channel (VDAC) and nuclear-encoded proteins through the translocase of the outer membrane (TOM) complex [2,3]. Mitochondria fuse and divide dynamically in response to changes in the cellular environment through family members of dynamin-related GTPases. Dynamin-related protein 1 (Drp1) controls mitochondrial fission while mitofusins (MFN) in the outer membrane and OPA1 in the inner membrane regulate mitochondrial fusion. Interestingly, mutations in MFN2 are associated with Charcot-Marie-Tooth type 2 (CMT2)  while mutations in OPA1 are the major cause of dominant optic atrophy [5,6]. Recent findings also demonstrate that OPA1 controls mitochondrial dynamics by sensing changes in nutrient availability through mitochondrial solute carriers .
Solute Carrier 25 (SLC25) is a large family of more than 50 nuclear-encoded transporters embedded mainly in the inner mitochondrial membrane that shuttle a variety of substrates participating in numerous metabolic pathways . SLC25 members vary greatly in their size, the nature of transported substrates and the modes of transport, whereas common mechanisms of substrate translocation have been proposed [9,10]. To date, numerous SLC25-related rare metabolic diseases have been identified, caused by mutations in 13 genes encoding SLC25 proteins with various symptomatology, depending on the affected metabolic pathways and their significance in specific tissues . Even though the majority of these disease-related SLC25 members have been biochemically characterized through identification of their substrates and correlated metabolic pathways, there are yet two orphan SLC25 members, SLC25A38 and SLC25A46, causing disorders in humans whose substrates have not been identified . SLC25A46 has recently been identified as a cause in heterogeneous clinical conditions, including optic atrophy, axonal peripheral neuropathy Charcot-Marie-Tooth type 2 (CMT2) , an optic atrophy spectrum disorder , Leigh syndrome , progressive myoclonic ataxia with optic atrophy and neuropathy  and lethal congenital pontocerebellar hypoplasia . It has been proposed that human SLC25A46 is a mitochondrial outer membrane protein with a role in mitochondrial dynamics and cristae maintenance that interacts with MFN2, OPA1 and members of the MICOS complex [12,14]. However, the effects of its ablation in animal models in vivo have not been addressed yet. Thus, the unavailability of genetic animal models has compromised the in-depth dissection of the related pathology and its correlation with the human disorders.
In the present study, following a forward genetics approach, we identified a loss-of-function mutation in the Slc25a46 gene that causes lethal neuropathology in mice. Mutant mice manifest the main clinical features described in patients, including ataxia, optic atrophy and cerebellar hypoplasia as well as additional previously unidentified neuropathological alterations, while they were rescued by expression of the human ortholog. The aim of this study was to characterize the complex phenotype displayed by the mutant mice in order to understand the mechanistic basis of the heterogeneous SLC25A46-mediated pathologies.
Following a forward genetics approach through N-ethyl-N-nitrosourea (ENU) random mutagenesis [17,18], we have recently identified a novel mouse model of autosomal recessive neurological phenotype characterized by ataxia, unsteady locomotion, tonic-clonic seizures, limb-clasping during tail suspension, reduced muscle strength, and growth retardation with premature lethality by the age of 3 months (Fig 1A–1D, S1 Video). Because of the prominent neurological symptoms, this phenotype was designated as ataxic (atc). The first symptoms were noticeable at the second postnatal week with full penetrance in both sexes. From the third week, atc/atc mice stopped gaining weight (Fig 1A), their muscle strength began to deteriorate (Fig 1B) and by the fifth week of age the majority of atc/atc mice had died (Fig 1C). Heterozygous mice (atc/+) were indistinguishable from their wild-type (WT) littermates throughout life. In order to exclude the possibility that atc/atc mice succumb to starvation due to difficulties in accessing food caused by weakness and tremor, mashed wet food pellets were added inside the cage. No gross histopathological lesions such as hepatic steatosis were observed in atc/atc mice (S1 Fig). Apart from the neurological phenotype, atc/atc mice displayed severe thymic and splenic hypoplasia (Fig 1E–1N). Thymi of 4-week-old atc/atc mice had approximately 30% of the WT littermate weight and flow cytometric analysis revealed significant reduced cellularity in all four thymic subpopulations, CD4+CD8-, CD4-CD8+, CD4+CD8+, and CD4-CD8- (Fig 1E–1I). Atc/atc mice also developed severe splenic hypoplasia (25% of WT weight) with a significant decrease in the absolute number of T lymphocytes (CD4+ and CD8+), B lymphocytes (B220+) and myeloid cells (CD11b+, Gr1+) (Fig 1J–1N). To examine whether the immune system is associated to the neurological phenotype we generated atc/atc mice lacking mature T and B lymphocytes upon genomic deletion of RAG-2 (recombination activating gene) gene that is necessary for immunoglobulin and T-cell receptor gene rearrangements . Clinical evaluation of the double mutant mice showed that the neurological symptoms, survival, body weight, and muscle strength were not reversed (Fig 1O and 1P), implying that the lymphoid and neurological symptoms are distinct.
To map the causal mutation we performed genetic linkage analysis using an initial set of 68 Simple Sequence Length Polymorphic markers (SSLPs) in 20 atc/atc and 20 not affected F2 mice, and identified clear linkage with chromosome 18 with a LOD score of 12 (Fig 1Q). Fine mapping by analyzing 37 additional F2 mice and screening with 2 Single Nucleotide Polymorphisms (SNPs) located in between the two extreme SSLPs, narrowed down the mutant locus to a genetic region of 8.84 Mbp between D18Mit22 and rs30151517 (S1 Table). Selected candidate gene sequencing, identified a C to T transition at the coding region of the Slc25a46 gene, introducing a premature termination codon and resulting in a truncated protein of 95 amino acids instead of the 418 amino acids composing WT SLC25A46 (Fig 1R). SLC25A46 is a member of the Solute Carrier 25 (SLC25) family of nuclear genes encoding mitochondrial carriers that shuttle a variety of substrates across the inner mitochondrial membrane . 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. As the primary sequence of the SLC25A46 protein indicates the presence of six conserved transmembrane regions spanning a region between amino acids 100 to 418 , the identified nonsense mutation presumably leads to a loss-of-function allele. Functional point mutations in the SLC25A46 gene have recently been reported in heterogeneous neurological disorders [12,13,14,15,16]. Our mutant mice emerge as the first genetic mouse model for understanding the complex pathophysiological role of the SLC25A46 protein in vivo.
The majority of SLC25 members are embedded in the inner mitochondrial membrane. The mitochondrial localization of mouse SLC25A46 was confirmed by immunoblot analysis in WT central nervous system (CNS) tissues (Fig 2A) and by fluorescence microscopy upon expression of a chimeric SLC25A46-EGFP protein in the N2A mouse neuroblastoma cell line (Fig 2B). Similarly to human SLC25A46, the mouse ortholog localizes to the outer mitochondrial membrane, as shown upon swelling and proteinase K protection assays in mitochondria isolated from the cerebellum of WT mice (Fig 2C) or cell lines (S2 Fig). Analysis of SLC25A46 expression profile showed prominent expression in the cerebrum, cerebellum and spinal cord (Fig 2D and 2E), and complete absence of the full-length 46 kD protein in mitochondria from atc/atc mice (Fig 2F). However, the mRNA expression levels of Slc25a46 remained unaffected in atc/atc mice (S3 Fig), indicating that the truncating mutation does not lead to nonsense-mediated mRNA decay. According to the "50 nucleotide rule" this premature termination codon is not predicted to subject to decay, as it is located 38 nucleotides apart from the next exon-exon junction . Double immunofluorescence labelling for SLC25A46 and the alpha subunit of the mitochondrial enzyme pyruvate dehydrogenase E1 (PDH-E1a) revealed high expression of SLC25A46 in mitochondria of WT cerebellar Purkinje cells (PCs), while a diffuse cellular distribution was identified in atc/atc PC somata (Fig 2G), similarly to N2A cells expressing the mutant SLC25A46-EGFP protein (Fig 2B). These results demonstrate that the mutant SLC25A46 protein is not embedded in the mitochondrial membrane, thus being ensued with a compromised function.
Because of the prominent SLC25A46 expression in the nervous system of WT mice and the ataxic phenotype of mutant animals, we examined the brain for possible histological alterations. Macroscopic examination showed no overt abnormalities in the gross anatomy of atc/atc brain. However, histological staining revealed a selective 25% reduction in the size of the cerebellum (S4 Fig) with prominently reduced thickness of the molecular layer (ML) (Fig 3A and 3B). Although PC numbers and alignment were not affected, PC dendrites were severely stunted (Fig 3C–3F). PCs are the main output neurons of the cerebellar cortex receiving signals that modulate their activity through multiple synapses on their elaborate dendritic arbors . Parallel fibers (the axons of granule neurons) extending from the internal granule layer (IGL) of the cerebellar cortex and climbing fibers originating from the inferior olivary nucleus form excitatory glutamatergic synapses on PC dendrites . Additionally, mossy fibers coming from pre-cerebellar nuclei indirectly target PCs, via formation of synapses with granule cell dendrites within the IGL . To elucidate the basis of cerebellar hypoplasia and PC impairment, we examined the distribution of glutamatergic synapses between PCs and parallel or climbing fibers in the ML, as well as the distribution of mossy fiber synapses in the IGL, which indirectly target PCs via granule cells, using antibodies against the vesicular glutamate transporters vGlut1 and vGlut2. Our analysis showed a significant reduction in vGlut1 and vGlut2 immunoreactivity both in the ML and the IGL, indicating a paucity of glutamatergic synapses in atc/atc mice (Fig 3G–3M). Interestingly, a reduction in vGlut1 and vGlut2 immunoreactivity was visible at postnatal day 9 when no obvious defects in PC morphology were discernible as yet (S5 Fig). This feature, which is compatible with the ponto-cerebellar hypoplasia observed in the human disease , suggests that the stunted PC dendritic growth observed later could be linked to or exacerbated by the compromised glutamatergic input. An alternative hypothesis is that the reduced glutamatergic input is a consequence rather than a cause of the underdeveloped PC dendritic arborization, possibly due to primary impaired mitochondrial dynamics.
To examine the dendritic ultrastructure of PCs we employed electron microscopy (EM) in ultrathin sections of the cerebellum of 4-week-old atc/atc and WT mice. Unlike the well-preserved appearance of PC dendrites in the molecular layer of the cerebellum in WT mice (Fig 3Ni), several degenerated PC dendrites were identified in atc/atc mice as judged by their disorganized cytoskeleton, often containing remnants of mitochondria and other organelles (Fig 3Nii and 3Niii and inset Fig 3Niii’). Additionally, numerous atypical mitochondria with cytoplasmic inclusions were found both in the soma and dendrites of atc/atc PCs (Fig 3Nv and 3Nvi) that were not present in the WT (Fig 3Ni and 3Niv). Structural configurations containing concentric stacks of flattened cisternae that surrounded clusters of mitochondria were frequently observed in atc/atc (Fig 3Nvii and 3Nviii) but not in WT mice. Surprisingly, other alterations associated with mitochondrial ultrastructural organization or distribution, were not noted.
Apart from aberrations in cerebellar cytoarchitecture, we observed significant astrogliosis and microgliosis in the spinal cord of mutant mice, indicating high levels of neuroinflammation, as estimated by the increase in GFAP+ astrocytes and resting (Iba1+/CD68-) or activated (Iba1-/CD68+) microglia by comparison with WT mice (S6 Fig). A reduction in the size of the spinal cord was also evident in atc/atc mice without any differences in neuronal density, motor neuron appearance or the profile of cholinergic nerve terminals (S7 Fig).
As the converging phenotype of SLC25A46-mediated pathologies is optic atrophy, we further investigated whether atc/atc mice develop histological optic lesions. Retinal ganglion cells (RGCs) receive visual information from photoreceptors located in the retina, which are transmitted to the brain via their long axons converging to form the optic nerve. Although RGC numbers and alignment were not affected, RGC dendrites were clearly underdeveloped as observed using the somatodendritic marker MAP2 (Fig 4A and 4A’). Additionally the Pax6+ amacrine cells in the inner nuclear layer (INL) and the ganglion cell layer (GCL) as well as the GAD65+ GABAergic amacrine cells, both forming synapses with RGCs , were significantly reduced (Fig 4B–4D). Further, aberrations in RGC axons were noted in the retina with reduced expression of neurofilament (NF) protein and a disorganized appearance within the optic nerve head of mutant mice, indicative of optic atrophy  (Fig 4E and 4F). This was confirmed in semi-thin cross-sections of the atc/atc optic nerve where morphometric analysis demonstrated a 30% reduction in the number of myelinated axons (Fig 4G and 4H).
Finally, the nodes of Ranvier appeared more elongated in myelinated fibers of the atc/atc optic nerve (Fig 4I and 4J) and cerebellum (S4D and S4E Fig), suggesting disorganization of nodal structures and alterations in action potential propagation by saltatory conduction [26,27], in the absence of gross myelination defects (S4A–S4C Fig).
Neuromuscular junction alterations were identified in the diaphragm of atc/atc mice (Fig 5). A significant increase in the number and density of endplates was detected, resulting in a narrower endplate band along the costal parts of the muscle (Fig 5A–5C). Quantification of the mean endplate area and volume revealed a significantly reduced size of individual endplates in atc/atc mice (Fig 5E and 5F). A number of structural and functional alterations occur upon mammalian NMJ development during the first postnatal weeks [28,29]. At birth, NMJs are innervated by multiple motor axonal inputs while excess branches withdraw progressively through the dynamic, activity-dependent process of synapse elimination, so that by 2 weeks after birth the majority of mouse NMJs become innervated by a single motor axon . At the same time, the post-synaptic area transitions from a plaque-like structure to a complex pretzel-shaped domain with multiple-folds and perforations, formed as acetylcholine receptors (AChRs) migrate to closely appose to the pre-synaptic input . Analysis of endplate classification into 5 characteristic stages of maturation according to their shape indicated a shift towards immature pretzel endplates (Fig 5G). Additional abnormal features included endplate poly-innervation, increased axonal ramification within the endplate, axon terminal protrusions extending beyond the endplate boundaries and bifurcation of single axons into two endplates (Fig 5H and 5I). Retarded NMJ maturation and improper innervation are early hallmarks of CMT2D  and in a vital muscle such as the diaphragm, these features are indicative of a functional deficit, which could cause or contribute to early lethality . On the other hand, analysis of semi-thin cross sections of the sciatic nerve for myelination (g-ratio), axon diameter and number of myelinated or non-myelinated axons (Remak bundles) did not yield any differences between atc/atc and WT mice (S8 Fig). However, fine alterations in the ultrastructure of peripheral nerves cannot be excluded at present evidence.
To confirm that the nonsense mutation identified in the Slc25a46 gene is causal for the ataxic phenotype, genetic rescue experiments were performed by introducing a human SLC25A46 transgene (TghuSLC25A46) into the Slc25a46atc/atc background (Fig 6, S9 Fig). As the human and the mouse SLC25A46 proteins exhibit 89% identity, they are expected to have common function. To achieve a physiological expression pattern of the human transgene in the mouse, a 202.7 kb genomic fragment containing the whole human SLC25A46 gene (Fig 6A) was isolated from a Bacterial Artificial Chromosome (BAC) clone and was used for pronuclear microinjections. This BAC fragment did not contain any coding regions of genes other than SLC25A46, ensuring exclusive production of human SLC25A46 protein. Four transgenic founders were generated carrying either one copy (Tg1255, Tg1332, and Tg1358) or two copies (Tg1351) of the transgene, as shown by quantitative real-time PCR (qPCR) (Fig 6B), all of which were fertile and appeared healthy, exhibiting no obvious neurological symptoms. From these lines, Tg1332 and Tg1351 (Fig 6B) were selected for crosses with atc/atc mice. The expression of the human SLC25A46 protein in the Slc25a46atc/atc background, which lacks the endogenous full-length SLC25A46 protein, followed a physiologically relevant pattern in various tissues tested, including the cerebrum, cerebellum, spinal cord, muscle, heart, thymus, spleen and liver (Fig 6C, S9A Fig) as compared to the mouse SLC25A46 (Fig 2E). The clinical phenotype of these mice was fully reversed as they had normal life span, normal body weight and grip strength compared to their WT and Slc25a46atc/atc littermates (Fig 6D–6F, S9B and S9C Fig). To confirm the phenotypic reversal at the cellular level, we performed immunofluorescence labeling on cerebellum mid-sagittal sections of one-year-old mice with antibodies against calbindin, vGlut1 and vGlut2. The profile of calbindin expression in the double mutant mice, TghuSLC25A46/Slc25a46atc/atc, was indistinguishable from that in their WT littermates (Fig 6G). Moreover, the compromised dendritic arborization of PCs in atc/atc mice was fully restored as well as the thickness of the molecular layer. The distribution of glutamatergic synapses, as defined by the expression of vGlut1 and vGlut2, also appeared normal in the “rescued” transgenic mice, both in the molecular and the granular layer (Fig 6H). In conclusion, phenotypic analysis of these mice showed complete reversal of the clinical features and the cerebellar histopathological abnormalities displayed by atc/atc mice, thus providing genetic confirmation for the causality of the mutation identified in the Slc25a46 gene.
Recently, SLC25A46 has been shown to interact with proteins that regulate cristae maintenance, including MFN2, OPA1 and the MICOS complex . To investigate whether in vivo ablation of SLC25A46 affects the levels of such proteins in the cerebellum, we followed a comparative proteomic approach using LC-MS/MS and label free quantitation in mitochondrial enriched protein extracts isolated from 4-week-old atc/atc mice and WT littermates (Table 1, S2 Table). As expected, SLC25A46 was identified in the CNS of WT mice and was absent in the mitochondrial extracts from atc/atc cerebellum. The depletion of SLC25A46 resulted in significantly altered levels of mitochondrial proteins critically involved in fusion and fission, including OPA1, MFN2, and mitochondrial fission factor (MFF). As regards the levels of the MICOS members Mitofilin (IMMT), CHCHD3 and CHCHD6, they were significantly increased in the absence of SLC25A46. Upregulated levels were also noted for the MICOS interacting protein SAMM50, the outer membrane proteins VDAC1 and VDAC2, members of the translocases of the outer (TOMM22, TOMM40 and TOMM70) and the inner (TIMM22 and TIM44) membrane complexes. Collectively, our data indicate that ablation of SLC25A46 affects remarkably the protein levels of mitochondrial proteins involved in mitochondrial dynamics and cristae maintenance in the cerebellum, a tissue mostly affected in atc/atc mice.
Following a forward genetics approach through ENU-mediated random mutagenesis, we identified a novel mouse model of severe autosomal recessive neurological disease characterized by ataxia, tonic-clonic seizures, unsteady locomotion, lymphoid abnormalities and growth retardation with early lethality. Through genetic linkage analysis, we identified a nonsense point mutation in the coding region of the Slc25a46 gene that introduces a premature stop codon and results in a loss-of-function truncated protein. All aspects of this lethal neurological phenotype were rescued through the expression of the human ortholog in transgenic mice, providing a genetic confirmation for the causality of the identified mutation. This is the first mouse model of human SLC25A46-mediated pathologies that have recently been reported by five independent research groups with heterogeneous clinical manifestations including optic atrophy, axonal peripheral neuropathy Charcot-Marie-Tooth type 2 (CMT2) , an optic atrophy spectrum disorder , Leigh syndrome , progressive myoclonic ataxia with optic atrophy and neuropathy  and lethal congenital pontocerebellar hypoplasia . Mutant mice manifest the main clinical features identified in patients, including ataxia, optic atrophy and cerebellar hypoplasia while histopathological analysis revealed previously unseen lesions, most notably disrupted cytoarchitecture in the cerebellum and retina, neuroinflammation in the spinal cord, prominent abnormalities in the neuromuscular junction, as well as a distinct lymphoid phenotype.
SLC25A46, a highly conserved protein among various species, is a member of the Solute Carrier 25 (SLC25) family of nuclear genes encoding mitochondrial carriers, characterized by the presence of a tripartite structure, six conserved transmembrane regions and the three-fold repeated signature motif P-X-[DE]-X-X-[RK]. Similar to other SLC25 members with transport activity, the latter motif displays modifications in SLC25A46 making it difficult to predict transport of putative substrates [9,10]. This could be associated with the fact that SLC25A46 is a mitochondrial outer membrane protein both in humans [12,14] and mice according to our results. On the other hand, a role for SLC25A46 in mitochondrial dynamics and cristae maintenance has been recently proposed based on the identification of hyperfused mitochondria with abnormal cristae architecture upon SLC25A46 elimination [12,14]. We could not detect such ultrastructural mitochondrial aberrations in the cerebellar cortex of atc/atc mice, although degenerated PC dendrites were clearly identified often containing atypical mitochondria with cytoplasmic inclusions. Nevertheless, it is possible that the recurring presence of mitochondrial clusters surrounded by concentric stacks of flattened cisternae presumably corresponding to endoplasmic reticulum (ER) membranes, may be associated with disturbances in ER-controlled mitochondrial dynamics .
The extended N-terminus of SLC25A46 (~ 100 amino acids) could account for recognition functions and interactions with identified partners, including OPA1, MFN2 and members of the MICOS complex . Through its unique features, SLC25A46 seems to be part of a larger interactome. Towards this direction we examined whether in vivo loss of SLC25A46 affected the levels of suggested interactors in the cerebellum, a tissue most affected in mutant mice, using a proteomic approach. We demonstrated that loss of SLC25A46 resulted in significantly altered levels of mitochondrial proteins critically involved in fusion and fission including OPA1, MFN2, and mitochondrial fission factor (MFF). Moreover, increased levels were identified for MICOS members, the MICOS interacting protein outer membrane protein SAMM50, VDAC proteins, as well as members of the TOM and TIM complexes. Collectively, our results show that loss of SLC25A46 affects remarkably the mitochondrial interactome in the cerebellum of atc/atc mice.
In summary, the atc/atc mice displaying the main features identified in patients including ataxia, premature death, cerebellar hypoplasia and optic atrophy constitute a valid model for mechanistic studies and further investigation of the associated human pathology. Towards this scope, formerly unseen histopathological lesions were identified in lymphoid tissues, the cerebellum, retina, spinal cord and NMJ, facilitating the analysis of the complex clinical phenotype caused by SLC25A46 mutations. Further work should elucidate the role of SLC25A46 in mitochondrial metabolism and associated neuropathology.
DBA/2J mice were purchased from the Jackson Laboratories and C57BL/6J mice from the Charles River Laboratories. Mice were maintained and bred under specific pathogen-free conditions in the animal facility of B.S.R.C. (Biomedical Sciences Research Center) ‘Alexander Fleming’. Mice were sacrificed via CO2 inhalation followed by cervical dislocation or as otherwise described below.
All animal procedures were carried out in strict compliance with the European Directive 2010/63/EU and the Greek National Law 56/2013 for Use of Laboratory Animals, according to the Federation of European Laboratory Animal Science Association recommendations for euthanasia and the National Institutes of Health Guide for Care and Use of Laboratory Animals and in accordance to the Hellenic License for Animal Experimentation at the B.S.R.C. ‘Alexander Fleming’ (Licence Protocol Number 1169/08.03.2012) issued by the Veterinary Department of Athens Prefecture after protocol approval by the Institutional Animal Ethical Committee of B.S.R.C. ‘Alexander Fleming’ (Protocol Number 259/13.02.2012).
Mutagenesis and breeding schemes were performed at B.S.R.C. ‘Alexander Fleming’ as previously described [17,34]. Briefly, three weekly doses of 100 mg/kg ENU (Sigma-Aldrich) were administered intraperitoneally into C57BL/6Jx129S6 male mice (G0) which were subsequently crossed to WT C57BL/6Jx129S6 females to produce G1 males. Backcross breeding schemes were set up between G1 males and G2 females to conduct a recessive phenotypic screen for ataxia in G3 progeny. Mutant mice were backcrossed at least for 10 generations on the C57BL/6J inbred strain to obtain congenic mice carrying the Slc25a46 mutation for phenotypic characterization.
The Grip Strength Test Meter, Model GS3 (Bioseb) was used. Mice were allowed to stand with four paws on the metal grid and were gently pulled backwards by the tail until they were off the grid. The mean from three repeats was calculated for every measurement.
Thymi and spleens were ground on size-40 metallic mesh discs (Sigma-Aldrich) and cell suspensions were isolated after filtering through 100-μm sheet. Thymocytes were resuspended in PBS and counted using a hematocytometer. Splenocytes were resuspended in Gey’s solution, filtered, pelleted and counted. 3x106 thymocytes were plated in 96 V-Bottom well plates (Costar) and stained with antibodies against CD4-Alexa 700, CD8-Alexa 647, CD44-FITC, CD25-PE (Biolegend, U.S.). 106 splenocytes were plated and stained with antibodies against CD4-Alexa 700, CD8-Alexa 647, B220-PerCP, CD11b-PE and Gr1-FITC (Biolegend, U.S.). BD FACSCanto II Flow Cytometer was used for running the samples and results were analyzed with DIVA software.
Liver parts were isolated from WT and atc/atc mice, mounted on O.C.T. compound (WVR) and snap frozen in liquid N2. The samples were let at -80°C O/N. Sections of 10 μm thickness were collected using Leica CM1950 cryotome and processed for staining according to Roy Ellis Staining Protocol. Images were acquired using a Nikon Eclipse fluorescence microscope on 20x magnification. A WT mouse was subjected to 24h starvation as a control of lipid accumulation .
Heterozygous +/atc mice were outcrossed to DBA/2J mice to generate F1 mice that were further intercrossed and produced F2 progeny used in genetic linkage analysis. Genomic DNA samples from F2 mice, both ataxic and not affected, were screened for Simple Sequence Length Polymorphisms (SSLPs) on 4% high resolution agarose gels (Sigma-Aldrich) and Single Nucleotide Polymorphisms (SNPs) by pyrosequencing (Pyromark ID, Biotage AB). Results were statistically analyzed using the qtl library of R (The R Foundation for Statistical Computing, version 3.2.2) as previously described . For sequencing purposes, PCR products were extracted from agarose gel and sent for analysis at MWG Biotech AG.
For Slc25a46 genotyping the following primers were used for pyrosequencing on tail DNA: F, 5’-CTC CCT GTT CCA TAT ATA TCA GAA TTT G-3’; R, 5’-AGA TTA CCT TGC AAG TCC AAT ACC-3’ and sequencing primer 5’-TCT TAT TCC TTT TAT TTT AG-3’.
Tissue mitochondria isolation was performed via differential centrifugation. Mitochondria Isolation Buffer (MIB) (210 mM mannitol, 70 mM sucrose, 5 mM HEPES, 1 mM EGTA, 0.5% BSA, pH 7.2) supplemented with proteinase inhibitors cocktail (Roche) was used to homogenize approximately 100 mg of tissue in a 1 ml glass mortar and pestle, manually, on ice. The homogenate was centrifuged (1000 g, 5 min, 4°C), the supernatant was kept on ice and the pellet was resuspended in MIB and centrifuged again (1000 g, 5 min, 4°C). The two supernatants were mixed and centrifuged (12000 g, 10 min, 4°C) twice to acquire the mitochondrial pellet and the cytoplasmic fraction. The mitochondrial pellet was washed with 500 μl MIB and after centrifugation was resuspended in a minimal volume of MIB. Protein concentration was determined using the Bradford method (Bio-Rad Protein Assay).
After subcellular fractionation, 20 mg of mitochondria were mixed with denaturating buffer (10% SDS, 50% Glycerol, 60 mM Tris/HCl pH 6.8, 100 mM DTT, 0.1%, Bromophenol blue, 5% 2-Mercaptoethanol) and separated by SDS-PAGE (12% gel, 0.1% SDS) at 120 V for 3 h at room temperature. Proteins were transferred to 0.45-μm nitrocellulose membrane (Santa Cruz Biotechnology) performing Wet Electroblotting (Tank Transfer, Bio-Rad) at 110 V for 1 h, on ice. The membranes were blocked with 5% BSA (Sigma-Aldrich) in PBS/Tween-20 (for SLC25A46, Translocase of Outer Mitochondrial Membrane 40/TOMM40, SLC25A5) or 5% skimmed milk in TBS (for Cytochrome c Oxidase Subunit IV/COX IV, Glyceraldehyde 3-Phosphate Dehydrogenase/GAPDH, Mitofilin, Glucose Regulated Protein/GRP75, Prohibitin) for 2 h at room temperature. Membranes were then incubated with antibodies against SLC25A46 (goat polyclonal; 1:2500 in 5% BSA in PBS/Tween-20; Santa Cruz Biotechnology), COX IV (rabbit polyclonal; 1:4000 in 5% skimmed milk in TBS; Novus Biologicals), GAPDH (mouse polyclonal; 1:4000 in 5% skimmed milk in TBS; Santa Cruz Biotechnology), TOMM40 (rabbit polyclonal; 1:500 in 5% BSA in PBS/Tween-20; Proteintech), Mitofilin(rabbit polyclonal; 1:500 in 5% skimmed milk in TBS; Proteintech), SLC25A5 (rabbit polyclonal; 1:500 in 5% BSA in PBS/Tween-20; Proteintech), GRP75 (rabbit polyclonal; 1:2000 in 5% skimmed milk in TBS; Santa Cruz Biotechnology) or Prohibitin (rabbit polyclonal; 1:600 in 5% skimmed milk in TBS; NeoMarkers), overnight at 4°C. Membranes were washed with PBS/Tween-20 or TBS and incubated with the HRP-conjugated secondary antibody according to primary antibody’s host species for 1 h at room temperature. Membranes were then washed with PBS/Tween-20 or TBS and protein bands were visualized using the Molecular Imager ChemiDoc XRS+ system (Bio-Rad). Results were analyzed through the Image Lab software (Bio-Rad).
Neuro2A (N2A) cells grown on glass coverslips were transiently transfected using the calcium phosphate (CaCl2) method  either with a control pEGFP (Clonetech) plasmid, or SLC25A46-pEGFP plasmids containing mouse Slc25a46 cDNAs, WT or mutant, fused to the C-terminal of the EGFP cDNA. 48 h post-transfection, cells were stained with 150 nM MitoTracker Red (Molecular Probes) in cell culture media for 20 min at 37°C. Samples were washed with PBS/methanol 1:1, fixed in 4% PFA for 20 min at room temperature and permeabilized with methanol for 10 min. Pictures were taken with a Nikon Eclipse fluorescence microscope and analyzed using ImageJ software.
50 mg of mitochondria were freshly isolated from WT cerebellum and resuspended either in isotonic buffer (250 mM sucrose, 1 mM EDTA, 10 mM Tris/HCl pH 7.6) or hypotonic buffer (1 mM EDTA, 10 mM Tris/HCl pH 7.6) on ice, as previously described . After 45 min incubation, mitochondria were pelleted by centrifugation (12.000 g, 5 min, 4°C). Samples were treated with 50 mg/ml proteinase K (PK, +) for 30 min on ice and proteinase K was inactivated by the addition of 2 mM PMSF. Mitochondria were pelleted by centrifugation and all samples were analyzed through immunoblotting.
Total RNA was extracted from various tissues by the guanidinium thiocyanate phenol-chloroform method according to the manufacturer’s instructions using Tri-Reagent (MRC Inc.). Concentration and quality of RNA was determined with a NanoDrop ND-1000 spectrophotometer. DNase I (Sigma-Aldrich) treatment, for removal of residual DNA, was followed by cDNA synthesis using 2 mg of total RNA, 0.5 mg of oligo d(T) 18 primer (New England Biolabs) and M-MLV reverse transcriptase (Sigma-Aldrich). qPCR was performed at 58°C with a primer pair for Slc25a46 gene (see ‘Transgenic copy number determination’) and a primer pair for B2M gene F, 5’-TTC TGG TGC TTG TCT CAC TGA-3’ and R, 5’-CAG TAT GTT CGG CTT CCC ATT C-3’ used as normalization control. Standard curves were generated for both genes, CT values were determined and data were analyzed by the comparative CT method (Bio-Rad RelQuant). For each experiment, two technical and three biological replicas were used.
Human BAC clone RP11-671K7 (ImaGenes, GmbH) was digested with NotI and products were analyzed by Pulsed Field Electrophoresis. A 202.693 bp fragment containing the human Slc25a46 gene was isolated upon electrophoresis in a 4% low melting agarose gel and digestion with β-agarase (New England Biolabs). This fragment was microinjected into fertilized (C57BL/6J x CBA/J) F2 oocytes as previously described [18,38,39]. Microinjections and embryo implantations were carried out by Transgenics & Gene Targeting Facility at B.S.R.C. 'Alexander Fleming'. Screening of transgenic mice was performed by PCR analysis on genomic DNA with primers that hybridize a sequence of the human but not the mouse gene as follows: F, 5'-AAT CAC GTG CTC CGA AGA CT-3'; R, 5'-TAA CCC CTC ATC CCT GTG TC-3'.
Determination of transgene copy number was performed by quantitative real-time PCR (qPCR) using tail DNA from TghuSlc25a46 mice and WT littermates. For the reaction SsoFast EVA Green Supermix (Bio-Rad) and a pair of primers specific for both mouse and human Slc25a46 gene were used: F, 5'-TCT GAC GTT ATA CTT TAC CC-3'; R, 5'-CAG TCT CTC ATT CCC TCA TA-3'. The nuclear gene HuR was used as normalization control for each sample: F, 5’-AGG ACA CAG CTT GGG CTA CG-3’ and R, 5’-CGT TCA GTG TGC TGA TTG CT-3’. Standard curves were generated for both primer pairs and CT values were determined and analyzed by the comparative CT method. The values of both Slc25a46 and HuR genes were extrapolated from their respective standard curves for each transgenic line. qPCR was performed using on a Rotor-Gene 6000 RT-PCR machine (Corbett Life Science).
Mice 4-weeks old or at postnatal day 9 (P9), were sacrificed by exposure to isoflurane and transcardially perfused with fixative (4% paraformaldehyde in phosphate-buffered saline). Brain, eye bulbs with attached the optic nerve, spinal cord and dorsal root ganglia were dissected out and post-fixed overnight at 4°C in the same fixative. Tissues were cryoprotected in 30% w/v sucrose solution in PBS, for 2 days at 4°C, embedded in O.C.T. compound (VWR Chemicals) and frozen at -80°C. A series of 4–5 sagittal sections, 20-μm thick, through the midline of the cerebellar hemispheres were collected and stored at -20°C until processed for immunohistochemistry. Similarly, eye bulbs with the optic nerve were cut in sagittal sections at the level of the optic nerve head. Spinal cords were cryosectioned transversely at the lower thoracic/lumbar level. The diaphragm, after dissection, was separated into the two costal muscle parts that were further processed as whole mount preparations, as previously described . Cryosections were thawed and subjected to antigen retrieval in 10 mM sodium citrate solution, pH 6, followed by 1 h blocking of non-specific sites with 5% normal donkey serum (NDS), simultaneously with permeabilization using 0.2% v/v Triton X-100 in PBS. Primary antibodies diluted in 2.5% NDS in PBS were applied overnight at 4°C, followed by incubation with appropriate secondary antibodies for 2 h at room temperature. Primary antibodies used were: goat polyclonal anti-SLC25A46 (N-13) (1:50; Santa Cruz Biotechnology); rabbit polyclonal anti-Calbindin (1:1000; Chemicon); mouse monoclonal anti-pyruvate dehydrogenase E1 alpha subunit (PDHE1a; 1:50; Santa Cruz Biotechnology); mouse monoclonal anti-vesicular glutamate transporter 1 (vGlut1; 1:500; Millipore); guinea pig polyclonal anti-vesicular glutamate transporter 2 (vGlut2; 1:10000; Millipore); mouse monoclonal anti-beta III tubulin (TUJ1; 1/200; Covance); rabbit polyclonal anti- 2΄,3-cyclic-nucleotide 3΄-phosphodiesterase (CNPase; 1:200; Cell Signaling); rabbit polyclonal anti-contactin associated protein (CASPR 1:2000; gift from Prof. D. Karagogeos at the Institute of Molecular Biology and Biotechnology, Crete, Greece); mouse monoclonal anti-voltage gated sodium channels (Navclone K58/35; 1:500; Sigma-Aldrich); mouse-monoclonal anti-neurofilament (NF; 1:100; Sigma-Aldrich) rabbit polyclonal anti-glial fibrillary acidic protein (GFAP; 1:500; Dako); rabbit polyclonal anti-ionized calcium-binding adapter molecule 1 (Iba-1; 1:400, Cell Signaling); rat anti-CD68 (1:100; Serotec); mouse anti-Neuronal Nuclei (NeuN; 1:500; Chemicon-Millipore); goat anti-choline acetyl-transferase (ChAT; 1:300; Millipore, gift from Dr Zagoraiou); goat anti-vesicular acetylcholine transporter (vAchT; 1:2000; Millipore, gift from Dr Zagoraiou); mouse monoclonal anti-microtubule-associated protein 2 (MAP2; 1:200; Millipore); rabbit polyclonal anti-glutamic acid decarboxylase 65 (GAD65; 1:100; Cell Signaling); mouse monoclonal anti-Pax6 (Pax6; 1:100; Developmental Studies Hybridoma Bank). Secondary antibodies (all from Molecular Probes) used for immunofluorescence were conjugated to Alexa Fluor 488 or 546 and cell nuclei were counterstained with TO-PRO-3 (1:1000; Molecular Probes). FITC-conjugated alpha-bungatotoxin (a-btx; 1:200; Molecular Probes) binding on acetylcholine receptors (AChRs) was used to label the endplates on diaphragm samples. Prolong Gold antifade curing mountant (Molecular Probes) was used for mounting. Images were acquired using Leica TCS SP and Leica TCS-SP5II confocal microscopes.
Cryosections were thawed, washed with phosphate-buffered saline (PBS) for 5 min, and then transferred in Toluidine blue solution (0.5% w/v in PBS) (Sigma-Aldrich) for 20 sec. Excessive staining was washed off in PBS for as long as necessary, sections were dehydrated through successive dilutions of ethanol (from 70% to 100%), followed by 2 x 10 min washes with Histolene clearing reagent (CellPath). Samples were mounted with DPX mounting medium (Sigma-Aldrich) and observed under a Zeiss-Axiophot up-right light microscope. The average thickness of ML and average cerebellum area from 5 mice per genotype, were calculated.
For Golgi staining , cerebella of adult mice (4-week-old) were dissected out and fixed in 3.7% PFA/1% methanol in PBS overnight at 4°C. Tissue was then incubated for 3 days in 3% potassium bichromate in the dark at room temperature with daily changes. Cerebella were transferred in 2% silver nitrate, incubated overnight at room temperature in the dark and cut in 50-μm sagittal sections using a vibrating microtome (Leica VT1000E). Sections were air-dried and mounted in DPX mounting medium (Sigma-Aldrich). The length of each Purkinje cell (distance from opposite end of the cell soma to the tip of the longest dendrite) was measured for at least 20 cells in a mid-sagittal section, for 3 mice per genotype in digital images obtained under a Zeiss-Axiophot up-right light microscope using a 63x objective.
The distribution of glutamatergic synapses between PCs and parallel or climbing fibers in the ML, as well as the distribution of mossy fiber synapses in the IGL, was examined using antibodies against the vesicular glutamate transporters vGlut1 and vGlut2. For evaluation of the expression levels of vGlut1 and vGlut2, fluorescence intensity was measured as pixel intensity on single channel stacks of confocal images, acquired under the same settings, as previously described [42,43]. The laser power was set to the lowest level that would be adequate for high signal/noise ratio acquisition and would not bleach the specimen. Gain and offset were set to constant levels optimized so as to avoid over- and under- exposure acquisition. Each series in the confocal stack was averaged 3 times and the step size was 0.5 μm while image resolution was 1024 x 1024 pixels. Image processing for quantification of fluorescence intensity was performed using the ImageJ software, after selecting the region of interest within the internal granule layer (IGL) and the molecular layer (ML), on single channel stacks using the free-hand selection tool and setting the threshold at a value that was kept constant throughout. Measurements were performed on each single image of the confocal stack and then added up and normalized to the corresponding area. Analysis was performed on postnatal day 9 and 4-week-old cerebella, and 3 animals were analyzed per genotype and per age tested.
On confocal images of sagittal retinal sections immunostained for Pax6, fields of the same size (110503 sq. microns) were sampled and the number of Pax6+ cells was counted in ImageJ, separately for the GCL and INL. Values from 3 fields were averaged per mouse (n = 3 mice per genotype).
For estimation of the nodal length of myelinated fibers in the optic nerve and in the white matter of the cerebellum, images from double immunostained samples for the voltage-gated sodium channels Nav that cluster in this region to facilitate saltatory conduction and contactin-associated protein (CASPR) that marks the paranodes, were acquired by sequential scanning under the confocal microscope using 63x objective and 3x zoom, with the step-size set to 0.5 μm and 1024 x 1024 pixels resolution. In the ImageJ software the length of the Nav positive area flanked by CASPR positive paranodes was quantified. Values were pooled from 3 mice per genotype for each tissue.
Analysis of the number, density, area, volume and maturity of endplates was performed as follows. Endplate density was estimated on confocal images of FITC-a-btx labeled endplates that covered the total area of the endplate band on each costal part of the muscle, acquired using 10x objective and 2x zoom. Using the surface module of Imaris software (v.8.2.0) the endplates were recognized as individual objects and their number was divided by the corresponding tissue volume calculated as (image area) x (image depth) for each image and added up for the total number of images per sample. To estimate the mean size of endplates, confocal images of individual endplates were acquired using 63x objective and 2x zoom and (a) the maximum projection area of each endplate was calculated using the ImageJ software after thresholding the a-btx positive area to obtain a binary image as that shown below the corresponding graph in Fig 3g; (b) the volume of each endplate was calculated using the surface module of Imaris software by obtaining 3D images as shown below the corresponding graph in Fig 3h. For both (a) and (b) values were obtained from 150–350 endplates pooled from 3 mice per genotype, as specified in the figure legend. For quantification of their maturation level the endplates were categorized into 5 distinct developmental stages and the stage frequency patterns of the two genotypes were compared . Images below the corresponding graph in Fig 3i are representative of each endplate maturation stage, classified as follows. P: Plaque; Endplates appearing as an oval or round plaque of contiguous staining for AChRs. SP: Plaque with a single or multiple small perforations with a diameter smaller than 1/5th of the endplate diameter. LP: Plaque with a single or multiple large perforations with a diameter greater than 1/5th of the endplate diameter. Single perforations might reach the outer border of the endplate, but strip-like structures typical of pretzels, are not found. IP: Immature pretzel, in which strip-like structures appear. The strips of the nascent pretzels exhibit a varying width. MP: Mature pretzel, in which the strips of the pretzel exhibit a homogeneous width. Approximately 200 endplates were characterized per sample (n = 4).
On confocal images of 3 transverse sections per animal at the lower thoracic/lumbar spinal cord immunostained for GFAP, 4 fields of the same size (29675 sq. microns) per section were randomly selected in respective areas of the gray matter and the number of GFAP+ cells counted in ImageJ was expressed as the percentage of total TO-PRO-3 counterstained nuclei. Values from 12 fields were averaged per mouse (n = 3 mice per genotype).
On confocal images of 3 transverse sections per animal at the lower thoracic/lumbar spinal cord double immunostained for Iba1 which marks both resting and activated microglia/macrophages and CD68 which marks activated microglia/macrophages, the number of Iba1+ cells and the number of activated Iba1+/CD68+ cells (also exhibiting phagocytic morphology) was counted per total section area calculated in ImageJ using the free-hand selection tool. Values from 3 sections were averaged per mouse (n = 3 mice per genotype).
On confocal images of 3 transverse sections per animal at the lower thoracic/lumbar spinal cord immunostained for NeuN which marks all neuronal cells, the total number of positive cells was counted and normalized per total section area in Image J. Values from 3 sections were averaged per mouse (n = 3 mice per genotype).
For conventional electron microscopy, mice were perfused with 2.5% glutaraldehyde—4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. Cerebella, optic nerves and sciatic nerves were removed and fixed overnight at 4°C in the same fixative. Tissues were dissected into 1-mm cubes and post-fixed with 1% osmium tetroxide for 1h at 4°C. They were then passed through a graded series of ethanol, followed by propylene oxide (PO), infiltrated gradually in a mixture of Epon/Araldite resins diluted in PO and then embedded in fresh epoxy resin mixture. Finally, the specimens were allowed to polymerize at 60°C for 24h. Semithin sections (1-μm thick) were stained with toluidine blue. Ultrathin sections were cut with a Diatome diamond knife at a thickness of 65nm on a Leica EM UC7 ultramicrotome (Leica Microsystems, Vienna, Austria), were mounted onto 300 mesh copper grids and stained with uranyl acetate and lead citrate. Sections were examined with a Philips 420 transmission electron microscope at an acceleration voltage of 40 kV and images were acquired with a Megaview G2 CCD camera (Olympus SIS, Münster, Germany).
Toluidine blue-stained 1-μm semi-thin cross sections of optic and sciatic nerves were observed under an upright Zeiss Axiophot microscope and images were captured using a Leica DC300 camera. For the optic nerve, the numbers of myelinated axons were counted per field in four randomly sampled fields per section, using ImageJ, and the total number of fibers was calculated for the total area of the cross section. For the sciatic nerve, the total numbers of myelinated axons or Remak-bundles of non-myelinated axons were counted per cross-section, using ImageJ. The diameters of myelinated axons and axons plus myelin were calculated by hand-tracing and measuring the corresponding circumferences, using ImageJ. The g-ratio as an index of myelination, was calculated for each fiber by dividing the inner diameter of the axon to the total diameter of the myelinated fiber. A minimum of 100 axons were measured in each nerve and values were averaged from 3–4 nerves per genotype.
Crude mitochondrial fraction was isolated as described above.
Purified mitochondria were resuspended in 150 μl lysis buffer containing 100 mM Tris-HCl, pH 7.6, 4% SDS and freshly made 100 mM DTT. Samples were incubated for 3 min at 95°C, followed by 20 minute incubation in a sonication water bath in order to shear the DNA. Finally, the samples were centrifuged at 17000xg for 30 min at 4°C and the supernatants were transferred to new tubes.
The protein extracts were processed according to the Filter Aided Sample Preparation (FASP) protocol  using spin filter devices with 10kDa cutoff (Sartorius, VN01H02). 150 μl lysate were diluted in 8 M Urea/100 mM Tris-HCl pH 8.5, the filters were extensively washed with the urea solution, treated with 10 mg/ml iodoacetamide in the urea solution and incubated for 30 min in the dark for the alkylation of cysteines. The proteins on the top of the filters were washed three times with 50 mM ammonium bicarbonate and finally the proteins were digested adding 1 μg trypsin/LysC mix in 80 μl 50 mM ammonium bicarbonate solution (Mass spec grade, Promega) and incubated over night at 37°C. The peptides were eluted by centrifugation, followed by speed-vac-assisted solvent removal, reconstitution in 0.1% formic acid, 2% acetonitrile in water and transferring to LC-MS glass sample vials. Peptide concentration was determined by nanodrop absorbance measurement at 280 nm.
2.5 μg peptides were pre-concentrated with a flow of 3 μl/min for 10 min using a C18 trap column (Acclaim PepMap100, 100μmx2cm, Thermo Scientific) and then loaded onto a 50 cm C18 column (75 μm ID, particle size 2 μm, 100Å, Acclaim PepMap RSLC, Thermo Scientific). The binary pumps of the HPLC (RSLCnano, Thermo Scientific) consisted of solution A (2% v/v ACN in 0.1% v/v formic acid) and solution B (80% ACN in 0.1% formic acid). The peptides were separated using a linear gradient of 4% to 40% B in 450 min at a flow rate of 300 nl/min. The column was placed in an oven operating at 35°C.
The purified peptides were analyzed by HPLC MS/MS coupled to an LTQ Orbitrap XL Mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) equipped with a nanospray source. Full scan MS spectra were acquired in the orbitrap (m/z 300–1600) in profile mode and data-dependent acquisition with the resolution set to 60,000 at m/z 400 and automatic gain control target at 106 ions. The six most intense ions were sequentially isolated for collision-induced (CID) MS/MS fragmentation and detection in the linear ion trap. Dynamic exclusion was set to 1 min and activated for 90 sec. Ions with single charge states were excluded. Lockmass of m/z 445.120025 was used for internal calibration. Xcalibur (Thermo Scientific) was used to control the system and acquire the raw files.
The raw files were analyzed using MaxQuant (version 184.108.40.206) [45,46] using the complete Uniprot proteome of Mus musculus (version of 7/3/2016) and a common contaminants database by the Andromeda search engine. Search parameters were strict trypsin specificity, allowing up to two missed cleavages. Oxidation of methionines and N-terminal acetylation were set as variable modifications. Cysteine carbamidomethylation was used as a fixed modification. The protein and peptide false discovery rate (FDR) was set to 1% for both proteins and peptides with a minimum length of seven amino acids that was determined by searching a reverse database. Protein abundance was calculated on the basis of the normalized spectral protein intensity as label free quantitation (LFQ intensity) enabling the “match between runs” option (set at 0.7 min). LFQ was performed with a minimum ratio count of 2. The statistical analysis was performed using Perseus (version 220.127.116.11). Proteins identified as contaminants, “reverse” and “only identified by site” were filtered out. The LFQ intensities were transformed to logarithmic (log2(x)). The replicas (four biological and three technical) were grouped together. The protein groups were filtered to obtain at least eight valid values across the twelve replicates and filtered so at least two peptides were present per valid value. The zero intensities were imputed—replaced by normal distribution, assuming that the corresponding protein has low abundance in the sample (imputation criteria: width 0.1 and down shift 1.8). A total of 2216 label free quantified proteins were subjected to statistical analysis and GO analysis, in which 525 were mitochondrial (based on GO cellular component). Two-sided Student’s t-test was performed for the comparison of the WT versus atc/atc mouse mitochondrial proteomes using p-value for truncation (p-value threshold of 0.05).
Two tailed unpaired Student’s t-test was performed using Prism 5 (GraphPad Software). p-value were symbolized as follows: *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.
Oil Red O staining of 4-week old male liver cryosections. (A) 24h starvation in WT mice (control) leads to accumulation of triglycerides in liver (steatosis) as lack of phosphatidylcholine synthesis limits the export of excess triglyceride from liver in lipoproteins, (B) WT and (C) atc/atc mice with ad libitum food access. Neutral triglycerides and lipids were stained with Oil Red O (red). The nuclei were stained with haematoxylin (purple). Scale bar: 150 μm.
50 mg of mitochondria were incubated in isotonic buffer (Sw, -) or hypotonic buffer (Sw, +) in the absence (PK, -) or presence (PK, +) of proteinase K. Samples were subsequently analyzed by immunoblotting with antibodies against SLC25A46, the outer membrane protein TOMM40, the intermembrane space protein Mitofilin, the integral inner membrane protein SLC25A5 and the matrix protein GRP75.
qPCR analysis in Brain (Cerebrum), Cerebellum and Spinal Cord from WT and atc/atc mice. Specific primers designed to hybridize to the 8th exon of Slc25a46 gene were used to confirm the absence of mRNA decay in atc/atc mice.
(A,C) Confocal images of sagittal cerebellar sections immunostained for the neuronal marker beta-3 tubulin (TUJ1, red) and the myelin protein CNPase (green). Nuclei are counterstained with TO-PRO-3 (blue). (B) Quantification of mid-sagittal cerebellum area (n = 4, p = 0,0029). Scale bar (a), 400 μm; (c), 40 μm. (D) Double immunostaining for the paranodal marker CASPR and voltage-gated sodium channels (Nav) clustered at the nodes of Ranvier in myelinated fibers of the cerebellar cortex. Scale bar = 2 μm. (E) Quantification of the nodal length based on Nav-channel immunostaining in atc/atc mice and WT littermates (n = 485 WT and 587 atc/atc nodes pooled from 3 mice per genotype, p<0.00001). Statistical significance was determined by two-tailed unpaired Student’s t-test. Data represent mean values ± SEM.
(A) Immunofluorescence labeling of P9 cerebellar mid-sagittal sections with SLC25A46, reveals diffused expression in the calbindin-positive PC somata of atc/atc mice. No differences were noted in the PC dendritic tree at this age. (B) Immunostaining of sagittal cerebellar sections reveals a significant decrease in the expression of vGlut1, both in the ML and the IGL of atc/atc mice. The insets are shown at higher power magnification. (C) Quantification of fluorescence intensity in the ML (p = 0.0028) and (D) in the IGL (p = 0.06) (n = 3 mice per genotype). (E) Immunostaining of sagittal cerebellar sections shows significant reduction in the expression of vGlut2, both in the ML and the IGL of atc/atc mice. (F) Quantification of fluorescence intensity in ML (p = 0.078) and (G) in the IGL (p = 0.0008) (n = 3 mice per genotype). Statistical significance was determined by two-tailed unpaired Student’s t-test. Data represent mean values ± SEM Scale bars, 40 μm.
Confocal immunofluorescence images of transverse sections of the lower thoracic/lumbar spinal cord of WT and atc/atc littermates. (A) GFAP labeling (white) at low and the insets at high power micrographs. (B) Quantification of the percentage of GFAP+ cells showing a statistically significant increase in atc/atc (n = 3 mice per genotype, p = 0.019). (C) Confocal images of double immunofluorescence against CD68 (activated microglia/macrophages; green in merged images) and Iba-1 (total microglia/macrophages; red in merged images) in spinal cord sections of WT and atc/atc mice, at low and the insets at high power magnification. Scale bars, 100 μm and 50 μm for low and high power images, respectively. TO-PRO-3 (blue) shows nuclear counterstaining. (D) Quantification revealed a significant increase in the density of Iba1+/CD68- cells/ mm2 (resting microglia/macrophages, p = 0.002) and Iba1+/CD68+ cells/ mm2 (activated microglia/macrophages, p = 0.0284) in atc/atc mice as compared to the WT (n = 3 mice per genotype). * p≤0.05; ** p≤0.01; *** p≤0.001. Statistical significance was determined by two-tailed unpaired Student’s t-test. Data represent mean values ± SEM.
(A) Toluidine blue staining of spinal cord transverse sections at the lower thoracic/lumbar level; (B) Quantification of the area of equivalent, toluidine blue stained spinal cord sections verified the smaller size of the spinal cord as expected due to the reduced growth of atc/atc mice as compared to the WT (n = 3 mice per genotype, p = 0.041). Scale bars, 100 μm for A and C (low magnification); 40 μm for E and F. * p≤0.05. Statistical significance was determined by two-tailed unpaired Student’s t-test. Data represent mean values ± SEM Confocal immunofluorescence images of transverse sections of the lower thoracic/lumbar spinal cord of WT and atc/atc littermates. (C) NeuN labeling of neuronal cells (white) and TO-PRO-3 counter stain of nuclei (blue). (D) Quantification of the number of NeuN+ neurons per area of the spinal cord section (n = 3 mice per genotype, p = 0.415). (E) Confocal images of immunofluorescence against ChAT which labels the motor neurons that reside in the ventral horns of the spinal cord. (F) Double immunofluorescence against vAchT (white) and NeuN (blue) visualizes the cholinergic synapses on the motor neurons of the spinal cord ventral horns.
(A) Representative toluidine blue stained semithin cross sections of 4-week-old WT and atc/atc mice. Scale bar: 5 μm. (B) Scatter plot of g-ratio as a function of axon diameter indicating a similar distribution of values among WT and atc/atc fibers (n = 689 myelinated WT fibers and n = 905 myelinated atc/atc fibers pooled from 3 nerves per genotype). (C) Histogram representing mean values ± SEM of calculated g-ratios (n = 3 nerves per genotype; p = 0.621). Quantification of (D) the mean diameter of myelinated axons (n = 3 nerves per genotype; p = 0.711), (E) the number of myelinated axons per nerve (n = 3 nerves per genotype; p = 0.582) and (F) the number of Remak bundles of non-myelinated small diameter axons (n = 3 nerves per genotype; p = 0.908). Statistical significance was determined by two-tailed unpaired Student’s t-test. Data represent mean values ± SEM.
(A) Expression pattern of human SLC25A46 on mitochondrial extracts from cerebrum (B), cerebellum (C), spinal cord (SC), muscle (M), heart (H), thymus (T), spleen (S) and liver (L) tissues of TghuSLC25A46/Slc25a46atc/atc (Tg1332-atc/atc) mice through immunoblotting with antibodies against SLC25A46 and the mitochondrial protein GRP75. Complete rescue of (B) body weight gain and (C) muscle weakness in atc/atc mice expressing human SLC25A46 (Tg1332-atc/atc). All mice used were sex matched littermates(n = 6 per group).
SSLPs and SNPs are shown in the left column and respective chromosome 18 position (in bp) are shown in the right column. B, H and D denote homozygosity for mutagenesis strain, heterozygosity and homozygosity for the mapping strain, respectively. Bottom row shows numbers of individual mice. Red and black font denotes ataxic and not affected mice, respectively. All ataxic mice screened were found homozygous for markers D18Mit225, D18Mit60 and D18Mit158. Four ataxic mice were found heterozygous for D18Mit22 and one mouse was found heterozygous for rs30151517 thus defining the chromosomal position where the mutant locus lies. The two candidate region defining markers and their respective positions are shown in bold.
We would like to thank Professor George Kollias and his lab for collaborations on the ENU screens and genotyping facilities. We also thank Professor A. Plaitakis for fruitful discussions. Bioimaging was performed at the Light Microscopy Unit of the Hellenic Pasteur Institute. We also thank the InfrafrontierGR infrastructure (co-funded by the European Regional Development Fund and Greek NSRF 2007–2013) for mouse hosting and phenotyping facilities.
This research was partially financed by the European Union (European Social Fund-ESF) and Greek national funds through the Operational Program "Education and Lifelong Learning" of the National Strategic Reference Framework (NSRF) Research Funding Programs THALIS-UOC "Mitochondrial dysfunction in neurodegenerative diseases" (grant code 377226) to RM and ED, ARISTEIA II Grant "DnaJmito" (grant code 4676) to ED and the Hellenic General Secretariat for Research and Technology Grant SYNERGASIA Noiseplus 09SΥΝ-21-969 to RM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
All relevant data are within the paper and its Supporting Information files.