Human Molecular Genetics Advance Access originally published online on December 1, 2005
Human Molecular Genetics 2006 15(2):233-250; doi:10.1093/hmg/ddi440
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mice deficient in the Rab5 guanine nucleotide exchange factor ALS2/alsin exhibit age-dependent neurological deficits and altered endosome trafficking
1Department of Molecular Neuroscience, The Institute of Medical Sciences and 2Department of Molecular Life Sciences, Tokai University School of Medicine, Isehara, Kanagawa 259-1193, Japan, 3Solution Oriented Research for Science and Technology, Japan Science and Technology Agency, Kawaguchi, Saitama 332-0012, Japan, 4Day Neuromuscular Research Laboratory, Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA 02129, USA, 5Center for Experimental Medicine, Institute of Medical Science, University of Tokyo, Tokyo 108-8639, Japan, 6Department of Neurology, SUNY Upstate Medical University, Syracuse, NY 13104, USA, 7The Jackson Laboratory, Bar Harbor, ME 04609, USA and 8Department of Paediatrics, Faculty of Medicine, University of Ottawa, Ontario, Canada K1H 8M5
* To whom correspondence should be addressed. Tel: +81 463915095; Fax: +81 463914993; Email: joh-e{at}nga.med.u-tokai.ac.jp
Received September 27, 2005; Accepted November 25, 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
ALS2/alsin is a member of guanine nucleotide exchange factors for the small GTPase Rab5 (Rab5GEFs), which act as modulators in endocytic pathway. Loss-of-function mutations in human ALS2 account for a number of juvenile recessive motor neuron diseases (MNDs). However, the normal physiological role of ALS2 in vivo and the molecular mechanisms underlying motor dysfunction are still unknown. To address these issues, we have generated mice homozygous for disruption of the Als2 gene. The Als2-null mice observed through 21 months of age demonstrated no obvious developmental, reproductive or motor abnormalities. However, immunohistochemical and electrophysiological analyses identified an age-dependent, slowly progressive loss of cerebellar Purkinje cells and disturbance of spinal motor neurons associated with astrocytosis and microglial cell activation, indicating a subclinical dysfunction of motor system in Als2-null mice. Further, quantitative epidermal growth factor (EGF)-uptake analysis identified significantly smaller-sized EGF-positive endosomes in Als2-null fibroblasts, suggesting an alteration of endosome/vesicle trafficking in the cells. Collectively, while loss of ALS2 does not produce a severe disease phenotype in mice, these Als2-null animals should provide a useful model with which to understand the interplay between endosomal dynamics and the long-term viability of large neurons such as Purkinje cells and spinal motor neurons.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
ALS2 was initially identified as a causative gene for a juvenile recessive form of amyotrophic lateral sclerosis (ALS), termed ALS2 (OMIM 205100 [OMIM] ), in a Tunisian kindred, and a rare juvenile recessive form of primary lateral sclerosis (PLS) (PLSJ; OMIM 606353 [OMIM] ) in both Kuwaiti and Saudi Arabian consanguineous families (1
,2). ALS2 is described as a spastic pseudobulbar syndrome with spastic paraplegia involving a loss of upper motor neurons (UMNs) and occasionally associated with several signs of lower motor neuron (LMN) defects (3), whereas PLSJ shows only UMN symptoms with no evidence of denervation (4). Recently, several independent homozygous ALS2 mutations have also been found in families segregating an infantile-onset ascending hereditary spastic paralysis (IAHSP) (5–7), a single family of a recessive complicated hereditary spastic paraplegia (HSP) (8) and a single family of ALS2 (9). Because ALS2, PLSJ and IAHSP/HSP are group of closely related MNDs (10–13), and loss-of-function ALS2 mutations account for a number of recessive MNDs, it is likely that the ALS2 gene product might play an important role in motor neurons.
ALS2 encodes a novel 184 kDa protein, termed ALS2 or alsin, comprising three predicted guanine nucleotide exchange factor (GEF) domains (1,2); that is, RCC1-like domain (RLD) (14), the Dbl homology and pleckstrin homology (DH/PH) domains (15) and a vacuolar protein sorting 9 (VPS9) domain (16–21). In addition, eight consecutive membrane occupation and recognition nexus (MORN) motifs are noted in the region between DH/PH and VPS9 domains (22,23). It has previously been demonstrated that ALS2 mediates the activation of Rab5 small GTPase via its Rab5-specific GEF activity that is associated with its C-terminal MORN/VPS9 domain (22,23) and that ALS2 modulates endosome/membrane trafficking in the cells (22–26). It has also been shown that ALS2 can stimulate Rac1 (25,27,28) and promote neurite growth in neuronal cultures (28). Moreover, overexpression of ALS2 protects cultured motor neuronal cells from toxicity induced by mutant Cu/Zn-superoxide dismutase 1 (SOD1) (27,29), suggesting a possible neuroprotective role for ALS2. However, the molecular mechanisms underlying motor neuron dysfunction and degeneration in ALS2-linked MNDs are still poorly understood. Common to 10 reported ALS2 mutations is the loss of the VPS9 domain either due to deletion (1,2,5,6,8,9) or nonsense (7) mutations in the coding exons or splicing site mutation (5,6). It is hypothesized that a perturbation of endosome and/or vesicle trafficking mediated by the ALS2-associated Rab5GEF activity underlies neuronal dysfunction and degeneration in the ALS2-linked MNDs (22,25).
To delineate the normal physiological role of ALS2 and the impact of the loss of its function in vivo, we have generated mice homozygous for disruption of exon 3 of the mouse Als2 gene, and extensively characterized the resulting Als2-null mice. Observed through 21 months of age, these Als2-null mice reveal no obvious developmental or reproductive abnormalities. They also display no defects in motor performance. However, histological studies demonstrate that the Als2-null mice develop an age-dependent, slow loss of cerebellar Purkinje cells and evidence of subclinical dysfunction of spinal motor neurons. In addition, there is moderate astrocytosis and microglial cell activation. Cell cultural studies reveal no major abnormalities in dendrites and axons in the Als2-null neurons. However, there is a decrease in the size of epidermal growth factor (EGF)-positive endosomes/vesicles in the Als2-null fibroblasts. Taken together, our findings suggest that ALS2 is important in membrane trafficking, particularly in motor neurons and Purkinje cells, although loss of this protein does not result in a severe disease phenotype in mice. Thus, these Als2-null mice should provide insight into the interplay between membrane trafficking, endosomal dynamics and the long-term viability of large neurons such as Purkinje cells and spinal motor neurons.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Splicing and expression patterns for the mouse Als2 transcripts are different from those for human ALS2
In human, the short variant of the ALS2 transcript of
2.6 kb, resulting from an alternative splicing at the 5' donor site after exon 4, is rather ubiquitously expressed in various adult tissues including brain (1). Further, its expression is believed to play a role in the phenotypic variations, including ages at onset and the LMN involvement, as observed in the ALS2-linked MNDs, (1,30,31). In this study, northern blot analysis revealed that, in addition to the full-length transcript (Als2_L), the short variant of the mouse Als2 transcript of an 2.9 kb (Als2_S) was expressed both in liver and kidney, but was undetectable in other tissues examined including brain (Fig. 1A). BLAST searches of the GenBank/DDBJ/EMBL database and DNA sequence assembling in conjunction with an extensive reverse transcriptase (RT)–Polymerase chain reaction (PCR)-based cloning identified the Als2 short variant of 2955 nt with a single 2787 nt ORF encoding 928 amino acids (100 kDa) (GenBank accession no. BC031479) (32) and revealed that Als2_S was produced by alternative splicing at the 5' donor site after exon 13, resulting in a premature stop codon after 74 amino acid residues in intron 13 (Fig. 1B). Thus, the structure and expression pattern for the mouse Als2 short transcript are different from that for the human variant.
|
Als2 mutant mice are viable and appear to develop normally
To generate Als2-null mice, we constructed a targeting vector in which exon 3 of the Als2 gene was disrupted by inserting a stop codon, followed by the neomycin resistance gene transcribed under the control of the Pkg1 promoter (Fig. 2A), which allowed to disturb the normal expression of both Als2_L and Als2_S transcripts. Although the Als2 gene in a targeted allele can be transcribed by its own promoter, the protein translation is terminated after the first 14 amino acids; as a result, the peptide lacks all the functional domains for ALS2. Six of 14 homologous recombinant ES clones exhibiting the desired targeting event were selected and subjected to the generation of chimera mice. Two germline chimeras (clones 17C6 and 21B5) were identified by analyzing the F1 animals, which were produced by crossing each male chimera with female C57BL/6J mouse, using PCR (data not shown) and Southern blotting (Fig. 2B). F1 mice heterozygous for the Als2 mutation (Als2+/–) derived from clone 17C6 were interbred, generating F2 (n=237) and F3 (n=42) mice for the analysis. The homologous recombination event in F2/F3 animals was also confirmed by Southern blot analysis (Fig. 2C). Further, western blot analysis of various tissues using two independent anti-ALS2 antibodies, HPF1-680 and MPF1012-1651, demonstrated that the expression of ALS2, a product of Als2_L, was eliminated in the Als2 homozygous mutant mice, and also reduced by approximately half in the heterozygous mutant mice (HPF1-680, Fig. 2D; MPF1012-1651, Supplementary Material, Fig. S1). However, no band corresponding to the predicted mouse short ALS2 variant (
100 kDa) was observed using our anti-ALS2_RLD antibody (HPF1-680) (Fig. 2D). Collectively, the homozygous Als2 mutant animals created in this study can be considered to recapitulate the status of null-expression of ALS2 in ALS2-linked MND patients.
|
The Als2-null mice were viable and fertile with no evidences for motor abnormality as observed for at least 21 months of age. The mutant allele was transmitted in the expected Mendelian ratio of an autosomal gene [Als2–/– (homozygote), n=72 (25.8%), male (m)/female (f)=30/42; Als2+/– (heterozygote), n=128 (45.9%), m/f=71/57; Als2+/+ (wild-type), n=79 (28.3%), m/f=42/37]. Further, growth of both homozygous and heterozygous mice assessed by the changes in their body weight was not statistically different from that of their wild-type littermates, despite that a number of Als2-null female mice exhibited an excessive body weight (Fig. 3A). Survival data at the age of 92 weeks also revealed no statistical differences between the genotype groups, although a slightly lower rate of survival in homozygous mutants were observed (Fig. 3B).
|
Als2 mutant mice display no profound defects in motor performance
Repeated examination of the selected F2 mice (n=9 for each genotype group) on the accelerating rotarod test over a period of 81 weeks with a weekly testing frequency revealed that the rotarod performance among animals varied irrespective of their genotypes. Although the wild-type and Als2 heterozygous mice showed a tendency toward improved performance when compared with the Als2 homozygotes, this was not statistically significant (Fig. 3C). Further, preliminary analysis of the cage activities in these mice, measured by a SUPERMEX system with an infrared ray sensor monitor (Muromachi Kikai), also showed no statistically significant differences in the spontaneous motor activities between groups (data not shown).
Lack of ALS2 does not affect expression levels of proteins related to the cytoskeleton and membrane trafficking
To test whether lack of ALS2 expression affect a series of 33 cytoskeletal or membrane/vesicle associated proteins, including Rab5 and Rac1, both of which were known to bind to ALS2 (22,23,25,27), and ALS2CL, a novel ALS2 homologous protein (33), we conducted western blot analysis of extracts obtained from brain (cerebral cortex, cerebellum, medulla/pons and spinal cord), liver and kidney (8 and 24 weeks of age, Supplementary Material, Fig. S2; 72 weeks of age, Supplementary Material, Fig. S3). Among the proteins tested, the levels of early endosome antigen 1 (EEA1), a Rab5-effector downstream of ALS2 (22,34), of the 8- and 24-week-old homozygous and heterozygous mutant mice were slightly decreased in cerebral cortex, but not in cerebellum. Furthermore, the levels of neurofilament heavy chain were slightly increased in spinal cord of the 72-week-old homozygous and heterozygous mutant mice. However, none of other proteins, including Rab5, Rabaptin-5, Rab4, 
-tubulin, ß-tubulin, MAP2, -adaptin, amphiphysin, AP180, clathrin HC, Rac1, Rab3, Rab8, Rab11, ß-catenin, EGF, complexin II, Mint2, Munc-18, rabaphilin 3A, rSec8, SNAP25, synapsin I, synapsin IIa, synaprogyrin, synaptophysin, synaptotagmin, syntaxin 6, GRP78 and ALS2CL showed altered levels of expression.
Brain and spinal cord of the Als2 mutant mice are histologically normal
ALS2 immunostaining was observed in the cerebellum, brainstem and spinal motor neurons of wild-type but not Als2-null mice (Fig. 4A–I). ALS2 is highly expressed in the granular and Purkinje layers of wild-type mice (Fig. 4C and F) and colocalizes with some, but not all calbindin immunopositive Purkinje cells (Fig. 4J). High magnification of Purkinje and spinal cord demonstrate cytoplasmic ALS2 immunostaining (Fig. 4F' and H'), with a dense localization to perinuclear membranous compartments in spinal motor neurons (Fig. 4H'). Low level ALS2 expression was also detected in CA2 of the hippocampus and motor cortex (data not shown). A comparable pattern of immunostaining was obtained on brain sections from wild-type mice using an independent anti-ALS2 antibody (HPP1024) (22) (Supplementary Material, Figs S4–S6). Double immunostaining with anti-ALS2 and anti-calbindin confirmed that ALS2 is expressed in subpopulations of Purkinje and surrounding cells in wild-type animals (Fig. 4J and J'; Supplementary Material, Figs 5 and 6). Hematoxylin and eosin (H&E) morphological staining for mid-brain (Fig. 4M and O), cerebellum (Fig. 4N and P) and spinal cord (Fig. 4S–V) demonstrated no gross abnormality and neuronal lesion phenotype in Als2-null mice at 7 months of age.
|
|
|
Cerebellar Purkinje cells are significantly decreased in the aged Als2-null mice
Immunohistochemial analysis revealed that the density of calbindin positive Purkinje cells in cerebellum of aged Als2-null mice (18 months of age) was significantly decreased (Fig. 4L). A higher-magnification analysis of the Purkinje cell layers confirmed this finding (Fig. 4Q and R). To determine whether loss of ALS2 expression grossly affects viability and the spatial pattern of Purkinje cells, an estimation of Purkinje cell number was conducted by counting calbindin immunopositve Purkinje cells in the cerebellum of Als2-null and wild-type mice using the BIOQUANT system (Fig. 5). The results showed that, at 7 months of age, there was no significant difference either of number or of area of Purkinje cells between wild-type and Als2-null mice (Fig. 5A and B). However, consistent with the results of immunohistochemistry, a 22.9% decrease in the number of Purkinje cells was observed in Als2-null mice (n=3) at 18 months of age. Thus, there were 25.74±1.56 SE Purkinje cells per 1 mm length of Purkinje layer in Als2-null mice compared to 33.36±1.34 SE cells in age-matched wild-type mice (n=3, Fig. 5A) (P=0.023, ANOVA). In addition, there was a 30.5% reduction in cell soma size of Purkinje cells in Als2-null mice (n=3), with a mean soma area of 135.3±5.1 SE µm2 compared with 194.6±13.1 SE µm2 soma size in wild-type mice (n=3, Fig. 5B) (P=0.013, ANOVA). Representative images of the calbindin immunostaining for the cerebellum of wild-type (Fig. 5C and C') and Als2-null (Fig. 5D and D') mice at 18 months of age were shown. Together, the results are indicative of a slow progressive loss of cerebellar Purkinje cells in Als2-null mice.
Astrocytosis and microglial activation are progressively enhanced in the Als2 mutant mice brain and spinal cord
In comparison with wild-type, age-matched control mice, brain and spinal cord of Als2-null mice at 7 and 18 months of age revealed significant progressive increases in the intensities of immunostaining for GFAP, CD68 and CD11b, markers for astrocytes, activated microglia and macrophages, respectively (Fig. 6). This strongly suggests that there is an increase in astrogliosis and activation of inflammatory responses in Als2-null mice. Immunostaining for the neuronal markers SMI32 was normal in Als2-null mutants both at 7 and 18 months of age (Fig. 6A–D).
Als2-null mice show evidences for motor unit remodeling
To determine whether there is a functional impairment in LMN of Als2-null mice, we have conducted motor unit number estimation (MUNE) analysis. The mean of estimated total number of viable axons (motor units) in the distal hind limb, determined by MUNE, is significantly reduced from 274.5 (±29.1) in wild-type mice (n=5) to 208.5 (±50.8) in Als2-null mutant mice (P=0.022; n=4) at 12 months of age (Fig. 7A). Concurrently, the single motor unit potential (SMUP) increases from 0.156 mV (±0.032) in wild-type mice to 0.178 mV (±0.037) in Als2-null mutant mice (Fig. 7B), reflecting an increase in the response of a single motor unit with decreasing MUNE. A similar pattern was observed in a single Als2-null mutant at 20 months of age, with mean MUNE and SMUP values of 191.0 (±0.76) and 0.225 (±0.165), respectively, compared with 285.3 (±0.76) and 0.145 (±0.023) in age-matched wild-type animals (n=3). This decrease in MUNE and concurrent increase in SMUP amplitude is consistent with remodeling of the architecture of the motor unit in Als2-null mice, with cycles of motor neuron degeneration leading to denervation followed by reinnervation. This process reduces the number of motor units (MUNE) but increases the mean size of the remaining motor units (SMUP).
|
Ventral motor axons are significantly decreased in Als2-null mice
In the Als2-null mice, quantitative histological analysis of the numbers of axons in the L4 ventral root failed to demonstrate any change in the numbers of the large motor neurons (>5 µm; the alpha motor neuron category) or small motor neurons (in the gamma motor neuron category with axon diameters between 1.4 and 4.5 µm) at 7 months of age (n=3 per group) (Fig. 7C and D). However, analysis of Als2-null mutants at 18 months of age (n=3) showed a significant decrease in the numbers of axons of all sizes (Fig. 7E and F) in particular small motor neurons with axon diameters between 1.5–4.5 µm. These morphometric findings suggest that there has been a progressive loss of motor axons of Als2-null animals.
Als2-null mice show evidence of fiber redistribution in skeletal muscle
To investigate the effect of ALS2 loss on skeletal muscle integrity, the morphology of muscle fibers and fiber types was examined in gastrocnemious and quadriceps muscles. No major difference was observed in H&E or acetylcholinesterase histochemistry stain of transverse gastrocnemius muscle of wild-type or Als2-null mice at 7 and 20 months of age (data not shown), although examination of thoracic muscle from Als2 mutant mice occasionally showed some isolated, angular atrophic muscle fibers and regions of dense pyknotic nuclear clumping and central nuclei suggestive of denervation (data not shown). Interestingly, myosin-ATPase (pH 4.3) staining revealed a slight redistribution and fiber grouping of dark (Type I, slow) myofibrils in the gastrocnemius muscle from the Als2-null mice at 7 months, with a severely abnormal pattern of fiber distribution at 20 months of age compared with wild-type (Fig. 7G–J). Again, these histochemical data are consonant with the electrophysiological analysis, suggesting the occurrence of some denervation of scattered motor units followed by reinnervation.
Neuromuscular junctions of Als2-null mice are morphologically abnormal
We have also performed a qualitative analysis of the neuromuscular junctions (NMJs) by staining post-synaptic acetylcholine receptors with fluorescently labeled, -bungarotoxin (BTX). These studies showed that there were fewer NMJs detected by BTX staining of gastrocnemius muscle in Als2-null mice at 12 (data not shown) and 20 months of age when compared with wild-type controls (Fig. 7K and L). Interestingly, in 20-month-old Als2-null mice, the post-synaptic folding within the NMJ appeared to be less complex, of smaller size and somewhat more globular conformation (Fig. 7N) when compared with normal endplate formation in age-matched wild-type muscle (Fig. 7M). Again, these findings are consistent with a chronic, slowly progressive impairment of function of the distal motor terminal and altered integrity of the NMJ.
Primary neuronal cells derived from Als2-null mutant mice grow and differentiate normally
Previous studies using primary neuronal cultures have demonstrated that ALS2 is localized within small punctate structures throughout the cells (25), suggesting that it functions in endosomes, possibly mediating vesicle trafficking in neurons (22). As the rearrangement of cytoskeletons and membrane trafficking is thought to play major roles for the neuronal differentiation/polarization (35), we have investigated the role of ALS2 in growth and maturation of neuronal cells in detail and the impacts of its functional loss in neurons. Primary hippocampal neuronal cultures from E18 Als2-null mice and their wild-type littermates on a mixed genetic background (F2) were prepared and maintained. The results showed that both cultured neurons were normally differentiated (stages 3–4/DIV 4.5) (Fig. 8A, upper panels) and displayed the fully elaborated MAP2-positive neurites (dendrites) at a late stage (stage 5/DIV21) (Fig. 8A, lower panels). Further, branching numbers of dendrites were also normal in Als2-null neurons (data not shown). The results suggest that ALS2 is dispensable for the neurite/dendrite formation in hippocampal neurons.
|
Next, to examine whether loss of ALS2 affects the formation of axon in neurons, quantitative analysis of the axonal sprouting was performed using primary granule neurons. Granule neuronal cultures were established from cerebellum of P6 homozygous mutant (n=3), heterozygous mutant (n=11) and wild-type (n=4) animals, which were produced by intercrossing the fourth-backcrossed generations (N4). Representative images of the primary granule cells (4 h after plating) were shown (Fig. 8B). Approximately 40% of the granule cells showed a sprouting phenotype, but no significant differences in their frequencies were observed among the different genotype groups (Fig. 8C).
Endosome dynamics are slightly affected by the Als2 mutation in fibroblasts
To further investigate the effects of ALS2 absence on cellular function, particularly receptor-mediated endocytosis in detail, we have prepared primary fibroblasts from new-born Als2–/– and Als2+/+ littermates, which were produced by intercrossing the N4 backcrossed heterozygous mutant mice. Fibroblasts were exposed to Alexa Fluor-488 labeled EGF for 10 min, allowing the internalization of EGF via receptor-mediated endocytosis, and then analyzed at 10, 30 and 60 min time points. Internalized EGF forms a punctate pattern of vesicles, representing EGF-positive endosome compartments. No gross abnormality was observed at all time points analyzed both in Als2-null and wild-type cells. These results imply that endocytosis of the EGF receptor per se does not require ALS2 (Fig. 9A). However, a quantitative analysis of the fluorescence intensities of the EGF-labeled endosomes/vesicles revealed that frequency of the vesicles with stronger fluorescent signals, thus larger in size, was significantly lower at a 10 min point in the Als2-null mutants than in the wild-type cells (Fig. 9B and C, P=0.045 by t-test). Notably, signal intensities in the wild-type cells were gradually decreased thereafter, whereas those in Als2-null mutants rather increased with highest at a 30 min point (Fig. 9C). The results imply that trafficking and fusion of EGF-positive endosomes/vesicles in fibroblasts were significantly delayed by the lack of ALS2. Thus, ALS2 might control efficiency of vesicles/endosomes trafficking and fusion in the cells.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Thus far, 10 independent homozygous ALS2 mutations have been reported, which include a single-nucleotide deletion in exon 3 of the ALS2 gene that was originally found in Tunisian ALS2 patients (1
,2), and nine additional independent mutations in families segregating ALS2, PLSJ and IAHSP/HSP (1,2,5–9). Interestingly, there are recognizable phenotypic differences between ALS2 and PLSJ/IAHSP, in which ALS2 patients develop a spastic pseudobulbar syndrome with spastic paraplegia involving a loss of UMN and occasionally associated with several signs of LMN defects (3,9), whereas those with PLSJ or IAHSP/HSP shows only UMN symptoms with no evidence of denervation (4,6). Although the molecular basis underlying such phenotypic differences is still obscure, it is tempting to speculate that the expression of a short variant of the ALS2 gene is believed to play a role in the phenotypic variations (1,30,31). We reasoned that generating mice homozygous for disruption of exon 3 of the mouse Als2 gene resembling the Tunisian ALS2 mutation would delineate disease pathogenesis for MNDs involving loss of both UMN and LMN.
Our gene targeting strategy has been designed to disrupt the expression not only of the full-length Als2 transcript but also a newly identified 2.9 kb short variant of Als2 (Als2_S), which is otherwise expressed predominantly in liver and kidney. Southern and Western blot analyses clearly demonstrated a complete loss of the functional full-length ALS2 protein in Als2–/– mice. We also identified a loss of expression of Als2_S in the Als2–/– mice (data not shown), while there is no explicit evidence for its expression at the protein level even in the wild-type animals (Fig. 3D). Taken together, our Als2 knockout mice represent a genuine Als2-null lacking the expression of the functional ALS2 protein.
To our surprise, the Als2-null mice and the Als2 heterozygotes have demonstrated normal growth, reproductivity, survival and motor performance. Biochemical and histological examinations also revealed no profound abnormalities in the brain and spinal cord of the Als2-null mice. Further, cell cultural studies showed no abnormal growth and differentiation of dendrites and axons in Als2-null primary neurons. Most recently, Cai et al. (36) reported that similar Als2-null mice do not demonstrate major motor deficits, but have a moderate, age-dependent impairment in motor coordination and motor learning, a higher level of anxiety response, increased body weight and increased susceptibility to oxidative stress. Although our results did not reach a statistical significance, tendencies in the decreased levels of motor coordination and increased body weight in female Als2-null mice are consistent with their results. At this stage, we cannot formally exclude the possibility that the phenotypic variations observed in Als2-null mice are due to a strain effect, because most of the analyses were conducted using F2 mice on a mixed genetic background, in which knockout mice show a flanking gene effect with the 129/Ola genetic context than wild-type littermates (37). To address this further, we are characterizing the Als2-null mice generated by backcrossing 10 generations with either C57BL/6J or FVB/N mice. Nonetheless, both data suggest that the Als2 gene per se is dispensable for the normal growth and development at least in mice, in stark contrast to the importance of ALS2 in humans.
Our detailed investigation using immunohistochemical and electrophysiological techniques allowed us to detect some abnormalities in Als2-null mice that were not documented by Cai et al. (36). First, Als2-null mice develop an age-dependent, slowly progressive loss of cerebellar Purkinje cells. Secondly, these mice also develop late-life, subclinical deficits of spinal motor neurons, evidenced by muscle fiber denervation followed by reinnervation and a reduction in numbers of ventral motor axons. Thirdly, Als2-null mice show increased astrogliosis and activation of inflammatory responses in brain and spinal cord. Lastly, in cultured Als2-null fibroblasts, trafficking and fusion of the internalized vesicles/membrane compartments are decreased. Collectively, these data document that the Als2 null mutation causes age-dependent, subclinical dysfunction of motor system and intracellular membrane trafficking in mice.
It is notable that progressive loss of Purkinje cells is observed in Als2-null mice, implying that long-term absence of ALS2 expression in Purkinje cells and/or the surrounding environment disrupts normal homeostasis of these cells. In this study, we showed that ALS2 is expressed in the granular and Purkinje layers of the cerebellum in wild-type mice. We also found that ALS2 expression colocalizes with some, but not all calbindin immunopositive Purkinje cells, suggesting that ALS2 is expressed at least in subpopulations of Purkinje cells and surrounding cells. This is consistent with our previous studies that used in situ hybridization to document that Als2 mRNA is strongly expressed in subpopulation of Purkinje cells (1). Devon et al. (31) demonstrated ALS2 expression in molecular and granular layers of the cerebellum, but not in Purkinje cells. Although it is conceivable that differences in ALS2 expression in Purkinje cells could reflect differences in our methodologies, the reason for this discrepancy remains elusive. In any case, a careful assessment of the cerebellar pathology, which has not been clinically implicated in ALS/MNDs, should be warranted not only in animal models but also in human ALS2/PLSJ/IAHSP cases.
MUNE is a highly sensitive method to estimate axon loss in diseases affecting the lower motor system (38). Notably, using this technique in conjunction with the assessment of ventral root, NMJ, and skeletal muscles, we detected abnormalities in the architecture of motor units in aged Als2-null mice. Our results demonstrating a decrease in MUNE and concurrent increase in SMUPs associated with abnormal NMJs, reduced ventral motor axons and muscle fiber-type grouping, all support the occurrence of fiber redistribution (denervation followed by reinnervation) and mild distal axonopathy in Als2-null mice. Recent studies have shown that abnormalities in the NMJ are an early finding in mice overexpressing mutant human SOD1 (39,40). Further, a decrease in the motor unit number in the distal hind limb was also evident before behavioral abnormalities appeared in the SOD1-transgenic mice (41), suggesting that a distal axonopathy is present in an early, pre-clinical phase in those ALS mice. Thus, the chronic, slowly progressing denervation and neurogenic atrophy observed in our Als2-null mice may also be an early and pre-clinical abnormality of the motor system dysfunction. Our observation of increased astrogliosis and activation of inflammatory responses without motor neuron death of spinal cord in Als2-null mice is consistent with this concept.
A corollary hypothesis suggested by these in vivo data is that loss of ALS2 is not rapidly detrimental but nonetheless mediates a long-term adverse effect on cellular physiology. Indeed, our cellular studies on the receptor-mediated endocytosis and trafficking of the internalized vesicles revealed that ALS2 is not required for either endocytosis or endosome trafficking. Nonetheless, a quantitative assessment of the fluorescence intensities of the EGF-labeled endosomes/vesicles demonstrated a significant decrease in its intensities at a 10 min point after the EGF internalization, implying that ALS2 might modulate an early phase of trafficking and fusion of the internalized vesicles and/or endosomal membrane compartments. The results support the concept that, at least in fibroblasts, ALS2 mediates modulatory functions on endosome dynamics via the activation of Rab5 (22,23,25). In addition to such endosome-related ALS2 function, a recent study has also revealed that loss of ALS2 predisposes neurons to oxidative stress (36), suggesting a possible neuroprotective role for ALS2. Further investigations on the effect of ALS2 loss in neuronal cells and a relationship between endosome dynamics and oxidative stress are in progress.
Given the subtle role of ALS2 in cellular physiology in mice, why do humans with ALS2 mutations develop such severe phenotypes? Three explanations are proposed. First, the differences in the architecture of corticomotoneuronal systems may be critical. In rodents, it is generally believed that there is no direct synaptic connection between UMN and LMN in the spinal cord (42), which might be associated with the lack of overt motor phenotypes in Als2-null mice. Against this is the observation that, irrespective of their connectivity, neither UMN nor LMN in Als2-null mice showed overt degenerating phenotypes. A second possibility is the longer length of human motor axons. A recent pathological study of postmortem brain and spinal cord of patients with HSP has suggested that axonal loss in the corticospinal tract is length-dependent, but not size-selective (43). Further, recent evidence supports the view that axonal transport plays a crucial role in the maintenance of longer motor neurons in human (44,45) and mice (46–51). Because human motor axons are markedly longer than those in the mouse, it is conceivable that human motor neurons are more susceptible to defects in axonal trafficking caused by loss of ALS2. Thirdly, a specific compensatory mechanism or redundant gene expression could alleviate the disease phenotypes in mice. Alternatively, human may have a unique mechanism aggravating the disease when ALS2 is absent. We demonstrate here that there are differences in the expression pattern and sequence structure for short alternative splicing variants between human ALS2 and mouse Als2 genes. In addition, it has been reported that human and mouse ALS2CL proteins, novel ALS2 homologs, possess slightly different biochemical and enzymatic properties (33). Thus, these homologs and/or variants as well as yet unidentified factors including the ALS2-interacting proteins may explain the divergences of phenotypes in mice and humans devoid of ALS2.
In conclusion, our findings suggest that ALS2 is important in membrane trafficking, particularly in motor neurons and Purkinje cells. Our Als2-null mice might provide a unique resource to understand the interplay between membrane trafficking, endosomal dynamics and the long-term viability of large neurons such as Purkinje cells and spinal motor neurons. Ultimately, understanding these complex phenomena will provide insights into the molecular pathogenesis of MNDs arising from inactivating mutations in the ALS2 gene.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Genomic analysis of the mouse Als2 gene and its expression
To characterize the potential variants for the mouse Als2 gene, we conducted BLAST searches on the public databases. Cloning of the identified mouse short variants for the Als2 gene was performed by a RT–PCR-based method. Expression and tissue distribution of the mouse Als2 mRNA were characterized by northern blot analysis. Briefly, mouse adult tissue Multiple tissue northern (MTN) blot (BD Biosciences) was hybridized with the [32P]dCTP-labeled mouse Als2 or mouse glyceraldehydes 3-phosphate dehydrogenase (Gapdh) cDNA in PerfectHyb hybridization solution (Toyobo) at 68°C. Membranes were washed with 0.1xSSC containing 1% sodium dodecyl sulfate (SDS) at 65°C and exposed to X-ray film (BioMax; Kodak).
Construction of the Als2 targeting vector
A 
FIXII 129SV/J genomic DNA library (Stratagene) was screened with the 5' portion (1–595 nt) of the mouse Als2 cDNA (GenBank accession no. AB053307). Eleven independent genomic clones covering 29 kb genomic region containing exons 2–9 of the Als2 gene were obtained and characterized by restriction enzyme mapping and DNA sequencing. A targeting vector was constructed by replacing a 39 bp fragment flanked by two BamHI sites within the third exon of the Als2 gene with the neomycin resistant gene (neo), as a positive selection marker, under the control of the phosphoglycerate kinase (PGK)-1 promoter. Further, the diphtheria toxin A (DT-A) gene, as a negative selection marker, was also used in this targeting vector. Briefly, the 5'-long SpeI/BamHI fragment (9537 bp) containing a portion of intron 1, exon 2, entire intron 2 and the 5' half of exon 3 was cloned into the SpeI/FbaI sites of the modified DT-A vector in which the NotI(mutated)–NruI–SpeI–FbaI–TGA(stop codon)–Cfr9I–NotI linker was inserted at the original NotI site of the pMC1DTpA vector. The resulting 5'-fragment vector was digested with Cfr9I/NotI and linearized. Next, the 3'-short BamHI/MluI fragment (1206 bp) containing 3' half of exon 3 and intron 3 was cloned into the BglII/MluI sites of the modified neo-cassette vector in which the XhoI–BglII–MluI–NotI–XhoI(mutated) linker was introduced into the original XhoI site of the pKJ2(X+) vector, and the resulting plasmid DNA was digested with Cfr9I/NotI to release the DNA fragment comprising the neo cassette connecting with the 3'-short fragment at the 3' end. This Cfr9I/NotI fragment was ligated to the 5'-long linearized Cfr9I/NotI fragment and circularized. Finally, the generated plasmid DNA was digested with NruI and linearized, generating the targeting vector for the third exon of mouse Als2 gene, which resembled the mutation found in Tunisian ALS2 patients (1,2).
Generation of the Als2 knockout mice
The linearized targeting vector was electroporated into E14.1 ES cells originated from 129/Ola strain, followed by the selection in G418 (52). Targeted clones were screened by PCR and Southern blot hybridization as described below. Among 507 G418-resistant E14.1 clones, 14 homologous recombinant clones (2.8%) were identified. After confirming the normal chromosome numbers and structure by karyotyping, six selected ES clones were subjected to the generation of chimera mice by the aggregation method using C57BL/6J blastocysts as the recipients (52). The resulting male chimeras were further mated with C57BL/6J female mice for germline transmission. Two ES clones (clones 17C6 and 21B5) gave germline chimeras, and we analyzed the knockout mice derived from clone 17C6. The heterozygous mice (F1 mice) were interbred to obtain wild-type, heterozygous and homozygous littermates (F2); the following generations (F3 and F4) were in a mixed 129Ola/C57BL6J (50%/50%) genetic background. Independently, the F1-heterozygous mice were also backcrossed to the C57BL/6J strain mice for four generations, and resulting heterozygous mutants (N4 mice) were interbred to obtain wild-type, heterozygous and homozygous littermates for the use of the primary cultured cells. The genotypes of the mice were determined by PCR and Southern blot analysis of genomic DNA obtained from the tails as below. Body weight of all animals was measured from 8 weeks of age and monthly (every 4 weeks) thereafter. Mice were allowed to freely access to food and water and housed at an ambient temperature of 23°C and at a 12 h light/dark cycle. All animal experiments were performed in accordance with the guidelines of the institutional committee on Animal Care and Use, and with the safety and ethical guidelines for gene manipulation experiments approved by the local institutional committee.
Polymerase chain reaction
Two pairs of primers allowing to specifically detect the mutant allele were designed as follows: neo-L1: 5'-ATCAGGATGATCTGGACGAAGAGC-3'/mCR6out-R1: 5'-ACCTTCAAAGACTCAACTCAGAAGCCG-3' (2.4 kb) and neo-L2: 5'-TACCCGTGATATTGCTGAAGAGCTTG-3'/mCR6outR2: 5'-GTCCTGAGACAAAAGTCCTGCTATGCC-3' (2.2 kb). Approximately 100 ng of genomic DNA prepared from ES clones or tail tissues was subjected to the PCR amplification using KOD-PLUS-DNA polymerase (Toyobo) with 2 min of pre-denaturation at 94°C, followed by 10 cycles of 15 s at 94°C, 30 s at 62°C and 10 min at 68°C, and additional 35 cycles of 15 s at 94°C, 30 s at 62°C and 10 min with extending 10 s every cycle at 68°C. Another set of primers: mCR6ex03L1: 5'-AACCCTCCCACCATGTACCC-3'/mCR6ex03R1: 5'-CCATTAGCATCGCTGTCCTG-3' was designed to amplify the entire exon 3 and its flanking intronic sequences. Genomic DNA was amplified by LA Taq DNA polymerase (Takara) with a condition of 1 min of 94°C, followed by 35 cycles of 5 s at 98°C and 7 min at 68°C. The wild-type and mutant alleles gave rise to PCR-fragments of 0.6 and 2.2 kbp, respectively.
Southern blot analysis
Two independent probes, mCR6_probe1 (3' external to the targeting vector) and mCR6_probe3 (5' external to the targeting vector), were prepared by PCR amplification using the primer sets as follows: mCR6_probe1L: 5'-TTTCATCTCATATCATGGCGTATGG-3'/mCR6_probe1R: 5'-TCTCTGCTGAGTGACAATGCCAG-3' (product; 627 bp), and mCR6probe3-L: 5'-TTCTCGCTCTCTCTAAATAAATGTTG-3'/mCR6probe3-R: 5'-GTTTCCTTTTTGAAGTTTGAGTTATAATTTG-3' (682 bp). In addition, the neo cassette was also utilized as a probe. Genomic DNA samples prepared from positive ES clones or tail tissues were digested with either BglII or KpnI, separated by electrophoresis, and blotted onto nylon membranes (Hybond; Amersham Biosciences). The BglII blot was hybridized with either mCR6_probe1, detecting an 4 kb of the wild-type and/or an 5.6 kb of the mutant alleles, or neo cassette, detecting only an 5.6 kb of the mutant allele. The KpnI blot was hybridized with mCR6_probe3, allowing to detect an 14.3 kb of the wild-type and/or an 15.9 kb of the mutant alleles.
Antibodies
Three independent anti-ALS2 rabbit polyclonal antibodies, HPF1-680 (22), HPP1024 (22) and MPF1012–1651 (23), were used in this study. Anti-ALS2CL rabbit polyclonal antibody, CLHPF560–953, was newly generated by immunizing rabbits with the recombinant C-terminal fragment (amino acids 560–953) of human ALS2CL (33) and affinity-purified using an antigen-coupled sepharose column. Other antibodies used for western blot analysis were listed in Supplementary Material (see legend of Supplementary Material, Fig. S2). Antibodies used for immunohistochemical and immunocytochemical studies included rabbit polyclonal anti-MAP2 antibody (1:1000; CHEMICON), mouse monoclonal anti-ßIII-tubulin (Tuji-1) antibody (Upstate), mouse monoclonal anti-GFAP antibody (1:500; CHEMICON), rat polyclonal anti-CD68 antibody (1:500; Serotec), rat polyclonal anti-CD11b antibody (1:500; Serotec), mouse monoclonal anti-SMI32 antibody (1:10 000; Sternberger Monoclonals) and mouse monoclonal anti-calbindin antibody (1:1000 for fluorescence, 1:5000 for histochemistry, Sigma).
Western blot analysis
Whole-tissue extracts were prepared from fresh mouse tissues by homogenizing in lysis buffer (25 mM Tris–HCl; pH 7.5, 5 mM MgCl2, 50 mM NaCl, 1 mM dithiothreitol, 5% (w/v) sucrose, 1% (w/v) IGEPAL CA-630, protease inhibitor cocktail (Roche, Mannheim, Germany), 1 mM phenylmethylsulfonyl fluoride), denatured in Laemmli's SDS sample buffer, subjected to SDS–PAGE and transferred onto a polyvinylidene difluoride membrane (Bio-Rad). The membranes were blocked with 10% skim milk in TBST (20 mM Tris–HCl; pH 7.5, 150 mM NaCl, 0.1% Tween-20) overnight at 4°C and incubated with the indicated primary antibody for 2 h and with horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG sheep secondary antibody (Amersham Biosciences). Signals were visualized by the ECL PLUS system (Amersham Biosciences) and BioMax X-ray films (Kodak).
Rotarod tests
Motor performance, coordination and balance were evaluated with the rotarod apparatus (MK-660A; Muromachi Kikai) using an accelerating mode. Groups of age-matched F2 homozygous mutant (Als2–/–, n=9; four males and five females), heterozygous mutant (Als2+/–, n=9; four males and five females) and wild-type (Als2+/+, n=9; four males and five females) mice on a mixed 129Ola/C57BL6J genetic background were placed on the accelerating rod at a starting speed of 0 r.p.m., reaching a final speed of 80 r.p.m. in 1 min. Each mouse was given five trials per day during the light-cycle; the test was performed once a week starting from 8 to 81 weeks of age. The maximum speeds at which mice fall off from the rotarod were scored.
Histological and immunohistochemical analyses
The F2-homozygous mutant (Als2–/–), heterozygous mutant (Als2+/–) and wild-type (Als2+/+) mice on a mixed 129Ola/C57BL6J genetic background (7–18 months of age) were used for histological and immunohistochemical studies. Animals were deeply anesthetized with 4% halothane and transcardially perfused with 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB) (pH 7.5). Brain, spinal cord and other organs were removed and post-fixed for at least 24 h in 4% PFA followed by cryoprotection in 30% sucrose in 0.1 M PB (pH 7.5) for 72 h and processed either for frozen tissue sectioning or for paraffin embedding. For fluorescent immunohistochemistry, 10 µm frozen sections were cut on a cryostat, and brain and spinal cord sections were incubated in phosphate-buffered saline (PBS, pH 7.6) with 3% normal goat serum and 0.3% Triton X-100 for 1 h at room temperature (Rtemp). For double immunostaining, sections were incubated with primary antibodies in PBS containing 0.01% Triton X-100 overnight at 4°C. Sections were incubated with CY3- or FITC-conjugated secondary antibodies (1:200 anti-mouse, Vector Laboratories; 1:300 anti-rabbit and anti-rat antibodies, Jackson ImmunoResearch Laboratories) for 3 h at Rtemp. TSA-amplification (Molecular probes) was used for fluorescent double labeling with anti-ALS2 and calbindin according to manufacturer's instructions. Formalin-fixed mouse brain samples were embedded in paraffin and sectioned at 7 mm thickness for immunohistochemistry for ALS2 using the affinity-purified anti-ALS2_RLD polyclonal antibody (HPF1-680) as described previously (22) with the following modifications: sections were incubated with 20 µg/ml trypsin (Zymed Laboratories) in PBS for 10 min at 37°C and washed in PBS prior to processing for normal immunohistochemistry. Sections were processed for DAB-immunostaining following a standard protocol using Vectorstain Elite ABC kit (Vector Laboratories) according to manufacturer's instructions and visualized using Nickel-enhanced DAB labeling as previously described (53). For identification of NMJs, fresh gastrocnemius muscle was harvested, post-fixed in 4% PFA for 24 h followed by 30% sucrose in 0.1 M PB for 24 h and sectioned longitudinally at 30 µm using a cryostat. Sections were incubated in PBS containing 0.2% Triton X-100, 5% normal goat serum, 2.5% bovine serum albumin (BSA) for 1 h at Rtemp followed by PBS containing BTX conjugated to Alexa Fluor-594 (1:500; Molecular Probes) and anti-SMI32 (1:500) overnight at 4°C. Sections were washed and processed for fluorescent immunostaining using standard procedures. Controls for all immunostaining were performed simultaneously by omitting the primary antibody. Sections were cover slipped using either DPX mountants (Bdh Chemicals) or VectorShield (Vector Laboratories) and analyzed under a Nikon fluorescent microscope equipped with a SPOT camera. All images presented are representative of at least n=3–4 animals examined in each group at each time point.
Morphological analysis
Brain was removed and embedded in paraffin. Mid-sagittal and coronal brain sections (7 µm) were stained with H&E for general pathological observation. Fresh muscle (gastrocnemius and quadriceps) was harvested and flash frozen on 2-methylbutane cooled in liquid nitrogen. Transverse 10 µm sections were processed for H&E, acetylcholinesterase and myosin ATPase (pH 4.3, 4.6 and 9.4) histochemistry according to the standard procedures.
Quantitative assessment of Purkinje cell numbers by BIOQUANT analysis
To estimate the effect of ALS2 inactivation on the number of Purkinje cells, Purkinje cells were counted in age- and region-matched sagittal brain sections from wild-type control and Als2-null mice. Half brains were serially sectioned (15 µm) from the midline. The number of Purkinje cells in every 25th sagittal section throughout the entire hemi-cerebellum was determined, as described previously (54,55). The number of Purkinje cells was counted as those identified to be positively immunostained for anti-calbindin antibody along the Purkinje cell layer. Purkinje cell profiles were counted as a function of the length of the Purkinje layer within the same section. Area measurements for Purkinje cell soma were recorded from 100 cells (selected using random sampling methodology) per section counted. All sections were processed simultaneously in an identical manner, and the Purkinje cell counts were performed by the same person (S.C.B.) with the genetic identity of the animals masked, using a light microscope equipped with BIOQUANT Image analysis software (R&M Biometrics).
Motor unit number
MUNE was performed using a modification of the incremental stimulation method (38,41,56). Under inhalation anesthesia, needle stimulation electrodes were inserted close to the sciatic nerve in the proximal leg and threshold for stimulation minimized by small movements of the needle. Recordings were made using a circumferential surface recording electrode around the distal hind limb. A maximum response reflecting activation of all viable motor axons was obtained. Following this, repeated stimuli were applied at very low intensities, slowly increasing the intensity until a single all-or-none response was obtained, reflecting the lowest threshold single motor unit. Intensity was slowly increased until 10 clearly defined increments were obtained, reflecting the summation of the first 10 motor units stimulated. Digital sub