Skip to main content
NCBI home page
Journal of Cerebral Blood Flow & Metabolism logoLink to Journal of Cerebral Blood Flow & Metabolism
. 2012 May 23;32(9):1725–1736. doi: 10.1038/jcbfm.2012.66

Aspartoacylase supports oxidative energy metabolism during myelination

Jeremy S Francis 1,*, Louise Strande 1, Vladamir Markov 1, Paola Leone 1
PMCID: PMC3434629  PMID: 22617649

Abstract

The inherited leukodystrophy Canavan disease arises due to a loss of the ability to catabolize N-acetylaspartic acid (NAA) in the brain and constitutes a major point of focus for efforts to define NAA function. Accumulation of noncatabolized NAA is diagnostic for Canavan disease, but contrasts with the abnormally low NAA associated with compromised neuronal integrity in a broad spectrum of other clinical conditions. Experimental evidence for NAA function supports a role in white matter lipid synthesis, but does not explain how both elevated and lowered NAA can be associated with pathology in the brain. We have undertaken a systematic analysis of postnatal development in a mouse model of Canavan disease that delineates development and pathology by identifying markers of oxidative stress preceding oligodendrocyte loss and dysmyelination. These data suggest a role for NAA in the maintenance of metabolic integrity in oligodendrocytes that may be of relevance to the strong association between NAA and neuronal viability. N-acetylaspartic acid is proposed here to support lipid synthesis and energy metabolism via the provision of substrate for both cellular processes during early postnatal development.

Keywords: aspartoacylase, ATP, Canavan disease, NAA, oxidative stress

Introduction

N-acetylaspartic acid (NAA) is an abundant amino-acid derivative in the mammalian brain, and is synthesized from aspartate and acetyl coenzyme A (AcCoA) by an aspartate N-acetyltransferase (Patel and Clark, 1979). N-acetylaspartic acid undergoes a rapid increase in the brain soon after birth to reach concentrations of 5 to 10 mmol/L (Tallan, 1957; Tkac et al, 2003), with the vast majority concentrated in gray matter-rich regions (Tallan, 1957; Nalder and Cooper, 1972; Moffett et al, 1991; Simmons et al, 1991). N-acetylaspartic acid synthesis is dependent on mitochondrial integrity (Patel and Clark, 1979; Bates et al, 1996), and fluctuations in concentration can occur in parallel with changes in adenosine triphosphate (ATP) (Signoretti et al, 2001), suggesting an intimate relationship with metabolic energy. The juxtaposition of NAA and metabolic energy is tight enough to be able to quantify clinical recovery or remission (De Stefano et al, 1995), making the prominent NAA magnetic resonance imaging signal a robust clinical indicator of neuronal metabolic integrity (Clark, 1998).

While there is a large body of indirect clinical evidence in support of an association between NAA and metabolic energy, actual experimental evidence for function is limited to a role in lipid synthesis (D'Adamo et al, 1968; Chakraborty et al, 2001). The severely dysmyelinated phenotype of the inherited human pediatric leukodystrophy Canavan disease is an important foundation for this proposed role, as it is caused by the loss of function of the sole known NAA-catabolizing enzyme, aspartoacylase (ASPA; Kaul et al, 1993). A reduction in free acetate for lipid synthesis subsequent to loss of ASPA function is believed to contribute to disease etiology and could likely account for abnormalities in the lipid content of myelin (Madhavarao et al, 2005; Wang et al, 2009). In addition, mice that are null for the aspartate/glutamate transporter ACG1 have a drastic reduction in NAA and global hypomyelination (Ramos et al, 2011), suggesting that NAA is critical for normal myelination.

Fatty acid synthesis is one of the several processes in the brain that utilizes AcCoA. We hypothesized that the proposed augmentation of fatty acid synthesis by ASPA would support not only the AcCoA-dependent synthesis of fatty acids, but may also simultaneously support ATP synthesis by way of common substrate requirements. We present here a systematic analysis of postnatal development in the nur7 mouse model of Canavan disease (Traka et al, 2008) with the aim of identifying metabolic markers of oxidative stress as they relate to the presentation of Canavan disease-like pathology.

Materials and methods

Animals

All mice were maintained in house at the UMDNJ (University of Medicine and Dentistry of New Jersey) vivarium under an approved UMDNJ IACUC (Institutional Animal Care and Use Committee) protocol. The use of animals under this protocol followed all guidelines as set out under UMDNJ institutional animal use policy. Animals were housed with ad libitum access to food and water. The nur7 genotype has been described previously (Traka et al, 2008), and animals used in this study were generated from heterozygous founder mice (C57BL/6J background) purchased commercially (Jackson Laboratories, Bar Harbor, ME, USA). Individual animals homozygous for the nur7 mutation are referred to as ‘nur7' in the present study. Animals were genotyped using a customized small nucleotide polymorphism (SNP) assay developed in house (available upon request) using genomic DNA obtained by tail clip. For all analyses detailed here, tissue was obtained post mortem from animals euthanized with an intraperitoneal injection of sodium pentobarbital.

Immunohistochemistry

Immunohistochemistry was performed on paraformaldehyde-fixed tissue. Animals were deeply anesthetized and transcardially perfused with ice-cold 0.9% saline followed by 4% buffered paraformaldehyde. Perfused brains were removed and post fixed in 4% paraformaldehyde overnight, then cryoprotected in an ascending sucrose gradient (10%, 20%, and 30%). Cryoprotected brains were flash frozen and stored at −80 °C until processed. The entire forebrain was serially sectioned (40 μm) and maintained in order for sterological sampling. Immunohistochemistry for adenomatous polyposis coli (APC), myelin basic protein (MBP), neuronal nuclear antigen (NeuN), and malondialdehyde (MDA) was performed using commercially available antibodies (Millipore, Billerica, MA, USA). Free-floating sections were permeabilized in 0.1% triton-X in phosphate-buffered saline for 20 minutes, treated with 1% H2O2, then washed 3 × 5 minutes in 0.1% trition-X. Primary antibodies were diluted in immunobuffer (0.1% trition-X with 1% normal goat serum) and incubated with sections overnight at room temperature with gentle agitation. Following primary antibody and washes, sections were incubated for 2 hours at room temperature with biotinylated secondary antibody followed Avidin-peroxidase for 1 hour at room temperature. Positive cells were visualized with metal-enhanced DAB substrate (Thermo Scientific Pierce Protein Products, Rockford, IL, USA), and slide mounted in series. Mounted sections were dehydrated in an ascending ethanol series, cleared with xylene, and coverslipped.

Stereology

The optical fractionator was used to sample processed sections for APC, MBP, NeuN, and APC-negative oligodendrocytes (k=8). All antigens were scored simultaneously in individual brains. For the generation of estimates of APC-negative cells, sections were simultaneously stained for APC, NeuN, and glial fibrillary acidic protein (GFAP) and counterstained with 0.1% cresyl violet to identify nuclei that were negative for all three antigens. Negative nuclei that presented with a darkly cresyl-stained nucleoplasm and prominent nucleoli. Sampling of individual sections was performed in two discrete anatomical regions of each section to generate estimates in the cortex and the subcortical white matter of the corpus callosum and external capsule. Optical dissectors within regions of interest were viewed at the 100 × -oil objective on an upright microscope fitted with an analog camera. Scoring for antigens was performed using Stereologer software (Stereologer Resource Center, Chester, MD, USA). Statistical significance between groups was calculated using a two-tailed Student's t-test.

High Performance Liquid Chromatography

The high performance liquid chromatography (HPLC) analysis of NAA, AcCoA malonyl coenzyme A (MalCoA), ATP, reduced glutathione (GSH), oxidized GSH (GSSG), and MDA was performed using previously published methodology (Lazzarino et al, 2003). Flash frozen forebrains were homogenized in an acetonitrile: 10 mmol/L K2HPO4 precipitation solution (3:1 v/v), extracted with chloroform and stored at −80 °C. In all, 25 μL of each sample was run on a Thermo Scientific HPLC system equipped with a Surveyor PDA plus UV detector, a Hypersil BDS-C18 column (5 μm particle size; 25 cm × 4.9 mm), and analyzed with ChromQuest software (Thermo Scientific). Run conditions were as follows: 100% Buffer A 25 minutes, 80% Buffer A 10 minutes, 55% Buffer A 11 minutes, 40% Buffer A 11 minutes, 35% Buffer A 10 minutes, 25% Buffer A 15 minutes, 100% Buffer B 35 minutes, 100% Buffer A 25 minutes (Buffer A: 12 mmol/L Bu4NOH, 10 mmol/L K2HPO4, 0.125% CH3OH pH 7.00; Buffer B: 2.8 mmol/L Bu4NOH, 100 mmol/L K2HPO4, 30% CH3OH pH 5.50). Target metabolites in samples were quantified against standard curves. All comparisons were made using age-matched (to the day) samples.

Primary Oligodendrocyte Cultures and Glucose Deprivation

Primary oligodendrocyte cultures were utilized to test the ability of NAA to provide substrate for ATP synthesis in the absence of glucose. Oligodendrocyte progenitors were generated from postnatal days 1 to 2 mouse brains as previously described (Francis et al, 2011). The cortices of freshly isolated brains were mechanically dissociated to a single cell suspension and A2B5-positive cells isolated by immunomagnetic sorting. Sorted cells were then maintained as suspension cultures in defined media (Francis et al, 2011) containing 10 ng/mL each of platelet derived growth factor (PDGF) and basic fibroblast growth factor (bFGF). Suspension cultures were passaged two times then plated at a density of 2 × 105 cells/mm2 on ploy-D-lysine coated 60 × 15 mm2 dishes. At the time of plating, cells were transduced with recombinant adeno-associated virus (rAAV) vectors containing either a green fluorescent protein (GFP) (rAAV–GFP) or wild-type human ASPA (rAAV-ASPA) expression cassette at a multiplicity of infection of 500:1 (i.e., 500 viral genomes for each cell present). Cells were maintained for 2 days with PDGF and bFGF (10 ng/mL each), and then a further 2 days with just bFGF (10 ng/mL). On the fifth DIV, cells were subject to glucose deprivation to starve cells of substrate for ATP synthesis. Media was replaced by a balanced salt solution (166 mmol/L NaCl, 5.4 mmol/L KCl, 0.8 mmol/L MgCl2, 1 mmol/L NaH2PO4, 1.8 mmol/L CaCl2, 26.2 mmol/L NaHCO3) supplemented with 20 mmol/L D-glucose for 1 hour. After 1 hour, fresh balanced salt solution with glucose was replaced with fresh glucose-free solution, and cells maintained for 3 hours before being provided with fresh balanced salt solution containing 20 mmol/L D-glucose. In all, 12 hours after glucose deprivation, cells were harvested for HPLC analysis of NAA and ATP. HPLC samples were generated from cells by first scraping plates with a rubber policeman and gentle aspiration with media. Cell suspensions were then pelleted by centrifugation and resuspended in 3 × pellet weight of fresh ice-cold precipitation solution. Resuspended cells were homogenized in precipitation solution by sonication and centrifuged at 12,500 × g for 15 minutes. Supernatant was removed, and remaining pellet sonicated in a further 300 μL precipitation solution and recentrifuged. The supernatant from the second spin was added to that from the first, and extracted with 2 × volume of HPLC-grade chloroform.

Adeno-Associated Virus Production

Recombinant adeno-associated virus was produced as described previously (Francis et al, 2011). Vectors were AAV1/2 chimeras and purified by heparin affinity chromatography. Vector titers were expressed as viral genomes per mL, and this figure was used to determine the multiplicity of infection for transduction of primary oligodendrocyte cultures.

Results

Oligodendrocyte Loss in the nur7 Brain Correlates with Spatial Patterns of Dysmyelination In Situ

The nur7 mouse has a point mutation in exon 4 of the aspa gene and presents a progressive pathology from 2 to 3 weeks of age onwards that includes oligodendrocyte loss and spongiform degeneration (Traka et al, 2008), which coincides with normal developmental increases in ASPA activity (Bhakoo et al, 2001) and expression (Kirmani et al, 2003). Nur7 mice do not generate any detectable ASPA protein and present with chronically elevated NAA. In the present study, levels of NAA in the nur7 brain became elevated from 2 weeks of age onwards (Figure 1A), coincident with the appearance of ASPA-positive oligodendrocytes in wild-type brains (Figure 1B). We next employed design-based stereology to generate spatiotemporal estimates of late stage oligodendrocyte numbers at 2, 4, and 8 weeks of age to highlight possible oligodendrocyte loss relative to the accumulation of NAA. Large tracts of subcortical white matter remain unaffected until relatively late in life in the nur7 mouse (Traka et al, 2008), and so two general regions of the brain were distinguished for sampling purposes, namely, the neuron-rich neocortex (cortex; Figure 2A) and white matter-rich areas of the corpus callosum and internal capsule (collectively termed subcortical white matter; Figure 2B). At 2 weeks of age, the nur7 cortex contained an average of 3.1 × 105 oligodendrocytes per hemisphere, which was not significantly different from wild-type controls (P=0.995; Figure 2A). Similarly, in subcortical white matter tracts, no significant differences in oligodendrocyte numbers were seen at 2 weeks (Figure 2B). At 4 and 8 weeks of age, however, a distinct pattern of abnormalities in nur7 oligodendrocyte numbers was manifest in the form of significant reductions in the cortex at 4 and 8 weeks (1.9- and 1.5-fold, respectively). In contrast to the cortex, nur7 subcortical white matter tracts manifest increased numbers of oligodendrocytes, which at 8 weeks was statistically significant (P=0.01; Figure 2B). Overall, total numbers of oligodendrocytes (cortex+subcortical white matter) did not differ at 2 or 8 weeks between the two groups, but at 4 weeks there was a significant 1.4-fold reduction (Figure 2C; P=0.002), suggesting that the period between 2 and 4 weeks is a critical period of development with respect to ASPA function. Cell loss in nur7 brains appeared to be oligodendrocyte-specific, as counts of NeuN-positive cortical neurons revealed no differences between the two genotypes at any age examined (Figure 2D). Adenomatous polyposis coli-negative oligodendrocytes were quantified at 2 and 4 weeks of age in the cortex and subcortical white matter to gain insight as to whether the decrease in nur7 cortical oligodendrocytes is due actual cell loss or to arrested development. Serial sections were simultaneously stained with APC, NeuN, and GFAP to eliminate late stage oligodendrocytes, neurons, and astrocytes, respectively. Sections were then counterstained with cresyl violet to identify ‘free' nuclei that presented with oligodendrocyte-like nuclear morphology (Figure 3A). At 2 weeks of age, no significant differences in APC-negative oligodendrocytes were apparent between the two genotypes in either region of anatomy (Figures 3B and 3C). At 4 weeks of age, there were significant increases in APC-negative oligodendrocytes in both cortical (+1.3-fold) and subcortical white matter regions (+1.6-fold). The 1.6-fold increase in subcortical white matter APC-negative oligodendrocytes likely accounts for the 1.2-fold increase in APC-positive cells (Figure 2B). The 1.3-fold increase in cortical APC-negative oligodendrocytes translates to ∼47,839 cells per hemisphere, while the 1.9-fold decrease in APC-positive cells in the same region represents a difference of ∼290, 225. Even when factoring for the ∼32,370 extra APC-negative cells in 4 week nur7 subcortical white matter (Figure 3B), it is unlikely that arrested development alone can account for cortical oligodendrocyte discrepancies in the nur7 brain.

Figure 1.

Figure 1

Open in a new tab

(A) N-acetylaspartic acid (NAA) concentration in Nur7 and wild-type brains as assayed by HPLC at the indicated ages. Elevated NAA becomes significant in nur7 brains from 2 weeks of age onwards. (B) Low levels of both adenomatous polyposis coli (APC) (green) and aspartoacylase (ASPA) (red) are present in the 1-week wild-type brain. (C) Increases in APC and ASPA are seen at 2 weeks of age, with the majority of ASPA cells colabeling with APC (APC+ASPA). Representative immunflourescent images of the border of the external capsule (EC) and somatosensory cortex (Ctx) in wild-type brains shown.

Figure 2.

Figure 2

Open in a new tab

Oligodendrocyte counts in the cortex (A) and subcortical white matter (B) in wild-type and nur7 brains. Systematic sampling of the entire forebrain was performed to generate estimates of adenomatous polyposis coli (APC)-positive cells in each. Significant differences (P<0.05) brains are indicated with asterisk. Subcortical white matter oligodendrocyte numbers appear unaffected at 2 weeks of age, but increase in number from 4 to 8weeks. By 8 weeks, subcortical white matter oligodendrocytes are significantly increased in number in nur7 brains (asterisk in A). Combined numbers of subcortical white matter and cortical oligodendrocytes (‘Total') are shown in (C) with 4 weeks the only age showing a significant difference. Counts of neuronal nuclear antigen (NeuN)-positive cells (D) revealed no difference between wild-type and nur7 cortical neuron numbers at any age examined.

Figure 3.

Figure 3

Open in a new tab

Adenomatous polyposis coli (APC)-negative counts. Individual sections were simultaneously stained for APC, neuronal nuclear antigen (NeuN), and GFAP, and counterstained with cresyl violet. Nuclei manifesting darkly stained nucleoplasm and absent for all three antigens were scored as ‘APC-negative.' (A) A representative high power image of the external capsule of a 4-week nur7 brain. Brown cell bodies are APC-positive and a single cresyl-positive/APC-negative nucleus is highlighted (red arrowhead). Counts were made of APC-negative nuclei in the cortex (B) and subcortical white matter (C) using the same unbiased stereological methodology as for APC-positive cells (Figure 2).

As oligodendrocytes are the myelin producing cells of the brain, reduced numbers would be expected to impact on myelination. Estimates of the combined length of cortical MBP-positive fibers were generated to provide an index of developmental myelination in wild-type and nur7 brains (Figure 4). As was the case with oligodendrocyte numbers, no difference in cortical MBP-fiber length between the two genetic backgrounds was evident at 2 weeks of age (Figure 4A). At 4 weeks however, there was a significant 1.4-fold reduction in fiber length in nur7 brains (P=0.0002), which persisted at 8 weeks, suggesting a relationship with oligodendrocyte abnormalities (Figure 2A). The volume of the corpus callosum delineated by MBP-positive fibers was then quantified to explore the relationship between oligodendrocyte numbers and myelination in subcortical white matter. Again, in contrast to the cortex, the MBP-positive volume of the nur7 corpus callosum appeared unaffected at all ages examined, and actually manifest an increased volume at 8 weeks of age (Figure 4B). The degree of myelination in both the cortex and subcortical white matter therefore appears proportional to oligodendrocyte numbers in each region.

Figure 4.

Figure 4

Open in a new tab

Developmental myelination. Stereology was used to quantify both myelin basic protein (MBP)-positive fiber length in the cortex (A), and the MBP-positive volume of the corpus callosum (B). Significant reductions (asterisk; P<0.05) in cortical MBP-positive fiber length were seen in nur7 animals at 4 and 8 weeks of age. No differences in MBP volume in the corpus callosum was evident at 2 or 4 weeks of age, but a significant increase in volume was seen in 8 week Nur7 brains. Representative images of MBP immunohistochemistry in the forebrain of wild-type and nur7 animals at 2 (C) 4 (D) and 8 (F) weeks. At higher magnification, the density of cortical MBP fibers is comparable in wild-type and nur7 brains at 2 weeks (E) but at 8 weeks is significantly reduced in nur7 brains (G).

Oxidative Stress Precedes Oligodendrocyte Loss in the nur7 Brain

Developmental myelination requires a sharp increase in lipid synthesis (Muse et al, 2001), and NAA is implicated in the de novo synthesis of long chain fatty acids from AcCoA, MalCoA, and NADPH through ASPA-supplied acetate for AcCoA synthesis (Mheta and Namboodiri, 1995). However, a genetically engineered reduction in myelin cerebroside synthesis does not exacerbate the phenotype of nur7 mice (Traka et al, 2008), suggesting that reduced lipid synthesis is not solely responsible for gross pathology. Given the strong association between fluctuations in NAA and neuronal oxidative metabolism, we hypothesized that the loss of ASPA function would impact negatively not only on developmental myelination, but on metabolic energy also. The brains of nur7 mice were analyzed at 1, 2, 4, and 8 weeks of age of age for AcCoA, MalCoA, and ATP relative to age-matched wild-type controls (Figure 5) to provide metabolic indices of both fatty acid synthesis and tricarboxylic acid cycle integrity. At 2 weeks of age, significant reductions in both AcCoA (3.1-fold; P=0.008) and MalCoA (1.6-fold; P=0.0001) were seen in nur7 brains, consistent with the loss of ASPA activity and downstream effects on lipid synthesis (Bourre et al, 1977). Reductions in coenzyme A derivatives were accompanied by a significant reduction in ATP at 2 weeks of age in the absence of any significant discrepancies in oligodendrocyte numbers (Figure 2), suggesting that oxidative stress may be an early feature of pathology in the nur7 brain. Consistent with this, lowered reduced (GSH) and increased oxidized GSH (GSSG) were evident from 2 weeks of age onwards in nur7 brains (Figure 6). This apparent oxidative stress response became progressively greater with age, and by 8 weeks of age, nur7 brains had a nearly threefold decrease in the ratio of GSH to GSSG.

Figure 5.

Figure 5

Open in a new tab

Reduction in Coenzyme A derivatives mirrors reductions in adenosine triphosphate (ATP). aspartate and acetyl coenzyme A (AcCoA), malonyl coenzyme A (MalCoA), and ATP were assayed in the brains of nur7 and wild-type mice at the indicated ages. From 2 weeks onwards, significant reductions in all three were evident in nur7 brains. For each age, n=4 to 5 per genotype. Concentration expressed as μmol per wet weight of tissue.

Figure 6.

Figure 6

Open in a new tab

Reduced and oxidized glutathione (GSH) levels in the brain. Nur7 brains show a simultaneous decrease in reduced GSH and increase in oxidized GSH (GSSG). The ratio of GSH to GSSG (GSH/GSSG) decreases steadily from 2 weeks onwards. GSH expressed in μmol per wet weight of tissue, GSSG as nmol per wet weight of tissue.

Levels of the lipoxidation end product MDA were significantly increased in nur7 brains at 2 and 4 weeks of age in nur7 brains (Figure 7B), and immunohistochemical evidence of increased MDA was prominent in vacuolated areas of the 4-week cortex (Figure 7C), suggesting increased lipid peroxidation associated with reduced antioxidant capacity. While increased on average in the 8-week nur7 brain, increases in MDA were not statistically significant at this age. Thus, the period between 2 and 4 weeks is notable for a reduction in AcCoA and ATP and the onset of oxidative stress (Figures 5 and 6) coincident with a decrease in cortical oligodendrocytes (Figure 2A). Another mouse model of Canavan disease has recently been reported to manifest an increase in apoptotic markers coincident with the onset of myelin abnormalities, between 15 and 30 days of age (Mattan et al, 2010), and processing serial sections of nur7 and wild-type brains at 3 weeks of age for TUNEL staining revealed 2.5 times more TUNEL-positive cells in the nur7 cortex as compared with wild-type controls (P=0.004; Figure 7E). However, while an increase in cortical apoptotic cells was evident, there were no differences in TUNEL-positive cells within subcortical white matter. This distinction between cortical and subcortical regions therefore mirrors the pattern of oligodendrocyte loss seen after 2 weeks (Figure 2), and suggests that the death of oligodendrocyte lineage cells is associated with oxidative stress in the nur7 brain.

Figure 7.

Figure 7

Open in a new tab

Increased markers of lipid peroxidation and apoptosis in the nur7 brain. (A) Wild-type and nur7 cresyl violet-stained sections at 4 weeks of age, showing severe vacuolation in mutant tissue. (B) Increased levels of malondialdehyde (MDA) at 2 and 4 weeks of age in the nur7 brain (red asterisk) accompany the onset of vacuolation. MDA expressed in μmol per wet weight of tissue. (C) Within cortical vacuoles, dense areas of MDA-positive puncta are visible. Shown are representative sections from the cortex of a wild-type and nur7 brain at 4 weeks. (D) TUNEL-positive cells (brown) were sampled using design-based stereology within both the cortex and subcortical white matter of 3-week wild-type and nur7 brains. (E) Significant increases in apoptotic cells were seen within nur7 cortices, but not subcortical white matter.

Aspartoacylase Activity Supports Adenosine Triphosphate Synthesis in Oligodendrocytes During Hypoglycemia In Vitro

The results presented here suggest that the loss of ASPA function during the early stages of postnatal development significantly compromises oligodendrocyte oxidative integrity. As AcCoA is a common substrate for both myelination and energy production, it is possible that the provision of acetate via the catabolism of NAA may support ATP synthesis in myelinating oligodendrocytes. Primary oligodendrocyte cultures were prepared from postnatal day 2 mouse brains, and transduced with rAAV vectors containing either aspa or gfp expression cassettes (Figure 8A) before glucose deprivation for 3 hours to determine if ASPA is capable of supporting ATP synthesis. Culture media was supplemented with 2 mmol/L NAA to provide extraphysiological levels of substrate to compensate for low endogenous NAA. Increased ASPA activity in aspa-overexpressing oligodendrocytes was confirmed by a significant reduction in NAA relative to gfp-expressing controls (Figure 8B). When ATP was assayed in the two insulted experimental groups, aspa-expressing cells presented a significant increase in ATP (Figure 8C; P=0.01). Thus, while glucose-deprived gfp-expressing cells had ATP levels that were 46% of nonglucose-deprived controls levels, aspa-expressing glucose-deprived cells presented with ATP levels that were 69% of normal controls, indicating the partial rescue of an oxidative deficit by ASPA.

Figure 8.

Figure 8

Open in a new tab

Partial rescue of energetic deficit in glucose-deprived primary oligodendrocytes. Cells were transduced with recombinant adeno-associated virus (rAAV) vectors for either GFP of aspartoacylase (ASPA) before insult. (A) O4-positive oligodendrocytes (red) colabeled with ASPA (Blue) in AAV-ASPA-transduced cells only. After 6DIV, cells were subject to 3 hours of zero glucose in the presence of 2 mmol/L N-acetylaspartic acid (NAA). In all, 12 hours after glucose deprivation, levels of NAA (B) and adenosine triphosphate (ATP) (C) were assayed by HPLC. NAA and ATP expressed as μmol and nmol per mg of cellular protein, respectively.

Discussion

N-acetylaspartic acid is associated with two broadly distinct pathologies that are defined by elevated and reduced concentration, respectively. Current experimental evidence for NAA function accounts for only compromised lipid synthesis in Canavan disease, and the relevance of this mechanism to lowered NAA and neuronal metabolic integrity is not apparent. The present study provides a possible link between NAA-associated pathology in neurons and oligodendrocytes by demonstrating compromised oligodendrocyte metabolic integrity during developmental myelination in the nur7 mouse.

Aspartate and AcCoA is at the intersection of a number of key biochemical processes in the brain, including ATP production and fatty acid synthesis. Dysfunction of the latter is implicated in the pathogenesis of Canavan disease (Madhavarao et al, 2005), with this clinical correlate constituting a focus for most efforts to define NAA function. Ascribing the severe Canavan phenotype to compromised lipid synthesis dictates uniform severity in all cases where access to free acetate via the catabolism of NAA is prevented, but an important clinical exception to this assumption suggest a more complex disease etiology. Recently, reports of an atypical Canavan phenotype associated with a specific point mutation in aspa have surfaced (Janson et al, 2006; Vellinov et al, 2008). This atypical phenotype is characterized by mild developmental delay without typically severe degeneration and dysmyelination. As this specific mutation generates a protein with no detectable catabolic activity (Janson et al, 2006), the associated atypical phenotype does not conform to a mechanism centered wholly on a reduction in free acetate (Madhavarao et al, 2005) as the cause of the typically severe Canavan phenotype. This important clinical exception would not exist if NAA functioned solely as an acetate source for developmental myelination. Recent experimental evidence generated in the nur7 mouse suggests that Canavan disease involves more than just defective lipid synthesis in that histological and behavioral markers of phenotype are unaffected by the genetic downregulation of cerebroside synthesis on the nur7 background (Traka et al, 2008). Thus, while NAA is demonstrably involved in lipid synthesis in the brain (Burri et al, 1991), Canavan disease is not necessarily the result of reduced lipid synthesis alone.

The brain relies overwhelmingly on glucose for metabolic energy and myelination places an acute demand on finite metabolic reserves during development. In the mouse, the bulk of myelination takes place during a relatively short developmental window during the first 2 to 4 weeks of life (Horrocks, 1973). This period of nur7 development was notable in the present study for the accumulation of NAA and simultaneous reduction in ATP. This association of accumulating NAA with lowered ATP stands in contrast to a number of neuronal pathologies that manifest lowered ATP associated with reduced NAA (De Stefano et al, 1995), but does imply that the NAA metabolic cycle is associated with metabolic energy in both neurons and oligodendrocytes.

Nur7 development after 2 weeks is characterized by a reduction in cortical oligodendrocytes, specifically (Figure 2), and a prominent feature of the present study is the demonstration of a striking difference in pathology between white and gray matter. Although oligodendrocyte loss was manifest in cortical gray matter, subcortical white matter remained relatively unaffected. Differences in oligodendrocyte numbers also correlate with differing degrees of myelination in specific regions of the brain (Figure 4). This surprising anatomical delineation extends the original examination of pathology in the nur7 mouse (Traka et al, 2008) and bears similarities to aralar knockout mice, which have a drastic reduction in NAA and hypomyelination that is more pronounced in gray matter than white matter (Ramos et al, 2011). The pattern of oligodendrocyte loss and hypomyelination presented here suggests that NAA is more important for gray matter oligodendrocytes than it is for those in white matter tracts. Whether this pattern is due to increased substrate competition between neurons and oligodendrocytes remains to be determined, and this anatomical heterogeneity will place some restriction on pathological modeling in vitro. It must be noted, however, that the present study does not comprehensively distinguish between APC-positive cells (Bhat et al, 1996) and earlier stages in the oligodendrocyte lineage, and increased subcortical white matter oligodendrocytes in nur7 brains (Figure 2B) may reflect an increased turn over of progenitors, as previously suggested (Mattan et al, 2010). The present study used a combination of antigen exclusion and nuclear morphology to address numbers of APC-negative oligodendrocytes specifically (Figure 3), and while significant increases in these cells were manifest in both cortical and subcortical white matter regions, the absolute numbers of such cells falls well short of the deficit in APC-positive oligodendrocytes. The unbiased design based sampling methodology employed here permits the generation of accurate, absolute estimates of number, and the 4-week cortical deficit of 290,000 APC-positive cells per hemisphere in nur7 mice is only reduced by ∼80,000 when APC-negative cells from both the cortex and subcortical white matter are subtracted. Overall, this translates to over 400,000 ‘unaccounted for' oligodendrocytes in the nur7 cortex, and so cell death (Figure 7A) must be considered a prominent feature of pathology in this model system. The paucity of late stage APC-positive cells clearly impacts negatively on myelination, and should be considered when performing not only analyses of myelin lipid content (Madhavarao et al, 2005), but also when assessing potential targets for therapeutic intervention (Segel et al, 2011). Although ASPA has been reported in nonoligodendroglial cells (Madhavarao et al, 2004), in our experience, the vast majority of ASPA-positive cells in the rodent brain colocalize with the APC antigen (Francis et al, 2011), and the transcriptional and activity profile of the enzyme coincides with peak periods of myelination (Kirmani et al, 2003; Bhakoo et al, 2001). While the present study is unable to discount the contribution of nonoligodendroglial cells to ASPA function, the tight correlation between APC-positive cell numbers and in situ markers of myelination (Figure 3), along with a lack of evidence of neuronal loss (Figure 2D), suggests that the oligodendroglial compartment is of primary relevance when discussing Canavan disease pathology.

The present study points to a potential point of substrate competition between myelination and oxidative energy that would be both consistent with and broader than currently posited pathological models of Canavan disease. An acute reduction in AcCoA in the nur7 brain (Figure 5A) appears to have negative consequences for both fatty acid synthesis, as reflected in low MalCoA (Figure 5B), and metabolic energy, as reflected in reduced ATP (Figure 5C). Given that oligodendrocytes appear to be refractive to damage by NAA (Kolodziejczyk et al, 2009), it is unlikely that the accumulation of noncatabolized NAA is toxic, and emphasis must therefore be placed on NAA synthesis and catabolism, rather than on absolute quantities of NAA per se. N-acetylaspartic acid synthesis requires AcCoA and an aspartate donor (Patel and Clark, 1979), and the strong correlation between NAA, metabolic energy and AcCoA in oxidatively stressed neurons (Vagnozzi et al, 2007) suggests that this process is energy intensive. Parallels to other AcCoA-intensive synthetic processes in neurons that draw on oxidative metabolic reserves (Browning and Schulman, 1968) can be drawn, and NAA synthesis likely places a demand on energetic reserves in the brain. By extension, viewing NAA catabolism as energy yielding in its capacity to provide acetate would define the complete metabolic cycle as an energy shuttle for the movement of substrate between neurons and oligodendrocytes. The loss of ASPA function in this multicompartmental context would compromise both fatty acid synthesis and energy production by way of a reduction in a common substrate. The manifest ability of ASPA to support ATP synthesis in oxidatively stressed oligodendrocytes in the present study (Figure 8) would seem to be consistent with such a role and suggests some potential for the promotion of oligodendrocyte viability under conditions of oxidative stress (Back et al, 2002). As the majority of fatty acid synthesis for myelination is microsomal (Bourre et al, 1977) and ASPA activity appears to be cytoplasmic (D'Adamo et al, 1973), the uncoupling of fatty acid synthesis and ATP production would require the shuttling of acetate into the mitochondria. Unfortunately, relatively little is known about the metabolism of oligodendrocytes with respect to preferred energy substrates, and future studies addressing carbon flow will be required to determine the significance of the contribution of NAA to oligodendrocytic energy production, particularly during developmental myelination.

The authors declare no conflict of interest.

Footnotes

This study was funded by Jacobs Cure (http://jacobscure.org/) and the Foundation of UMDNJ.

References

  1. Back SA, Han BH, Luo NL, Chricton CA, Xanthoudakis S, Tam J, Arvin KL, Holtzman DM. Selective vulnerability of late oligodendrocyte progenitors to hypoxia-ischemia. J Neurosci. 2002;22:455–463. doi: 10.1523/JNEUROSCI.22-02-00455.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bates TE, Strangward M, Keelan J, Davey GP, Munro PM, Clark JB. Inhibition of N-acetylaspartate production: implications for 1H MRS studies in vivo. Neuroreport. 1996;7:1397–1400. [PubMed] [Google Scholar]
  3. Bhakoo KK, Craig TJ, Styles P. Developmental and regional distribution of aspartoacylase in rat brain tissue. J Neurochem. 2001;79:211–220. doi: 10.1046/j.1471-4159.2001.00561.x. [DOI] [PubMed] [Google Scholar]
  4. Bhat RV, Axt KJ, Fosnaugh JS, Johnson KA, Hill DE, Kinzler KW, Baraban JM. Expression of the APC tumor suppressor protein in oligodendroglial. Glia. 1996;17:169–174. doi: 10.1002/(SICI)1098-1136(199606)17:2<169::AID-GLIA8>3.0.CO;2-Y. [DOI] [PubMed] [Google Scholar]
  5. Bourre JM, Paturneau-Jouas MY, Daudu OL, Baumann NA. Lignoceric acid biosynthesis in the developing brain. Activities of mitochondrial acetyl-CoA-dependent synthesis and microsomal malonyl-CoA chain-elongating system in relation to myelination. Comparison between normal mouse and dysmyelinating mutants (quaking and jimpy) Eur J Biochem. 1977;72:41–47. doi: 10.1111/j.1432-1033.1977.tb11222.x. [DOI] [PubMed] [Google Scholar]
  6. Browning ET, Schulman MP. (14C) Acetylcholine synthesis by cortex slices of rat brain. J Neurochem. 1968;15:1391–1405. doi: 10.1111/j.1471-4159.1968.tb05921.x. [DOI] [PubMed] [Google Scholar]
  7. Burri R, Steffen C, Herschkowitz N. N-acetyl-L-aspartate is a major source of acetyl groups for lipid synthesis during rat brain development. Dev Neurosci. 1991;13:403–411. doi: 10.1159/000112191. [DOI] [PubMed] [Google Scholar]
  8. Chakraborty G, Mekala P, Yahya D, Wu G, Ledeen RW. Intraneuronal N-acetylaspartate supplies acetyl groups for myelin lipid synthesis: evidence for myelin-associated aspartoacylase. J Neurochem. 2001;78:736–745. doi: 10.1046/j.1471-4159.2001.00456.x. [DOI] [PubMed] [Google Scholar]
  9. Clark JB. N-acetyl aspartate: a marker for neuronal loss of mitochondrial dysfunction. Dev Neurosci. 1998;20:271–276. doi: 10.1159/000017321. [DOI] [PubMed] [Google Scholar]
  10. D'Adamo AF, Gidez LI, Yatsu FM. Acetyl transport mechanisms. Involvement of N-acetyl aspartic acid in de novo fatty acid biosynthesis in the developing rat brain. Exp Brain Res. 1968;5:267–273. doi: 10.1007/BF00235902. [DOI] [PubMed] [Google Scholar]
  11. D'Adamo AF, Jr, Smith JC, Woiler C. The occurrence of N-acetylaspartate amidohydrolase (aminoacylase II) in the developing rat. J Neurochem. 1973;20:1275–1278. doi: 10.1111/j.1471-4159.1973.tb00097.x. [DOI] [PubMed] [Google Scholar]
  12. De Stefano N, Matthews PM, Arnold DL. Reversible decreases in N-acetylaspartate after acute brain injury. Magn Reson Med. 1995;34:721–727. doi: 10.1002/mrm.1910340511. [DOI] [PubMed] [Google Scholar]
  13. Francis JS, Strande L, Pu A, Leone P. Endogenous aspartoacylase exoresion is responsive to glutamatergic activity in vitro and in vivo. Glia. 2011;59:1435–1446. doi: 10.1002/glia.21187. [DOI] [PubMed] [Google Scholar]
  14. Horrocks LA. Composition and metabolism of myelin phosphoglycerides during maturation and ageing. Prog Brain Res. 1973;40:383–395. doi: 10.1016/S0079-6123(08)60701-3. [DOI] [PubMed] [Google Scholar]
  15. Janson CG, Kolodny EH, Zeng BJ, Raghavan S, Pastores G, Torres P, Assadi M, McPhee S, Goldfarb O, Saslow B, Wang DJ, Shera D, Leone P. Mild-onset presentation of Canavan disease associated with novel G212A point mutation in aspartoacylase gene. An Neurol. 2006;59:428–431. doi: 10.1002/ana.20787. [DOI] [PubMed] [Google Scholar]
  16. Kaul R, Gao GP, Balamurugan K, Matalon R. Cloning of the human aspartoacylase cDNA and a common missense mutation in Canavan disease. Nat Genet. 1993;5:118–123. doi: 10.1038/ng1093-118. [DOI] [PubMed] [Google Scholar]
  17. Kirmani BF, Jacobowitz DM, Namboodiri MA. Developmental increase of aspartoacylase parallels CNS myelination. Brain Res Dev Brain Res. 2003;140:105–115. doi: 10.1016/s0165-3806(02)00592-8. [DOI] [PubMed] [Google Scholar]
  18. Kolodziejczyk K, Hamilton NB, Wade A, Karadottir R, Attwell D. The effect of N-acetyl-aspartyl-glutamate and N-acetyl-aspartate on white matter oligodendrocytes. Brain. 2009;132:1496–1508. doi: 10.1093/brain/awp087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lazzarino G, Amorini AM, Fazzina G, Vagnozzi R, Signoretti S, Donzelli S, Di Stasio E, Giardina B, Tavazzi B. Single-sample preparation for simultaneous cellular redox and energy state determination. Anal Biochem. 2003;322:51–59. doi: 10.1016/j.ab.2003.07.013. [DOI] [PubMed] [Google Scholar]
  20. Madhavarao CN, Moffett JR, Moore RA, Viola RE, Namboodiri MA, Jacobowitz DM. Immunohistochemical localization of aspartoacylase in the rat central nervous system. J Comp Neurol. 2004;472:318–329. doi: 10.1002/cne.20080. [DOI] [PubMed] [Google Scholar]
  21. Madhavarao CN, Arun P, Moffett JR, Szucs S, Matalon R, Garbern J, Hristova D, Johnson A, Jiang W, Namboodiri MA. Defective N-acetylaspartate catabolism reduces brain acetate levels and myelin lipid synthesis in Canavans disease. Proc Natl Acad Sci USA. 2005;102:5221–5226. doi: 10.1073/pnas.0409184102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Mattan NS, Ghiani CA, Lloyd M, Matalon R, Bok D, Cassacia P, de Vellis J. Aspartoacylase deficiency affects early postnatal development of oligodendrocytes and myelination. Neurobiol of Dis. 2010;40:432–443. doi: 10.1016/j.nbd.2010.07.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mheta V, Namboodiri MA. N-acetylaspartate as an acetyl source in the nervous system. Brain Res Mol Brain Res. 1995;31:151–157. doi: 10.1016/0169-328x(95)00044-s. [DOI] [PubMed] [Google Scholar]
  24. Moffett JR, Namboodirir MA, Cangro CB, Neale JH. Immunohistochemical localization of N-acetylaspartate in rat brain. Neuroreport. 1991;2:131–134. doi: 10.1097/00001756-199103000-00005. [DOI] [PubMed] [Google Scholar]
  25. Muse ED, Jurevics H, Toews AD, Matsushima GK, Morell P. Parameters related to lipid metabolism in as markers of myelination in mouse brain. J Neurochem. 2001;76:77–86. doi: 10.1046/j.1471-4159.2001.00015.x. [DOI] [PubMed] [Google Scholar]
  26. Nalder JV, Cooper JR. Metabolism of the aspatyl moiety of N-acetyl-L-aspartic acid in the rat brain. J Neurochem. 1972;19:2091–2105. doi: 10.1111/j.1471-4159.1972.tb05119.x. [DOI] [PubMed] [Google Scholar]
  27. Patel TB, Clark JB. Synthesis of N-acetyl-L-aspartate by rat brain mitochondria and its involvement in mitochondrial/cytosolic carbon transport. Biochem J. 1979;184:539–546. doi: 10.1042/bj1840539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ramos M, Pardo B, Liorente-Folch, Saheki T, del Arco A, Satrustegui J. Deficiency of the mitochondrial transporter of aspartate/glutamate aralar/ACG1 causes hypomyelination and neuronal deficits in mouse brain. J Neurosci Res. 2011;89:2008–2017. doi: 10.1002/jnr.22639. [DOI] [PubMed] [Google Scholar]
  29. Segel R, Anikster Y, Zevin S, Steinberg A, Gahl WA, Fisher D, Staretz-Chacham O, Zimran A, Altarescu G. A safety trial of high dose glyceryl triacetate for Canavan disease. Mol Genet Metab. 2011;103:203–206. doi: 10.1016/j.ymgme.2011.03.012. [DOI] [PubMed] [Google Scholar]
  30. Signoretti S, Marmarou A, Tavazzi B, Lazzarino G, Beaumont A, Vagnozi R. N-Acetylaspartate reduction as a measure of injury severity and mitochondrial dysfunction following diffuse traumatic brain injury. J Neurotrauma. 2001;18:977–991. doi: 10.1089/08977150152693683. [DOI] [PubMed] [Google Scholar]
  31. Simmons ML, Frondoza CG, Coyle JT. Immunocytochemical localization of N-acetyl-aspartate with monoclonal antibodies. Neuroscience. 1991;45:37–45. doi: 10.1016/0306-4522(91)90101-s. [DOI] [PubMed] [Google Scholar]
  32. Tallan HH. Studies on the distribution of N-Acetyl-L-aspartic acid in the brain. J Biol Chem. 1957;224:41–45. [PubMed] [Google Scholar]
  33. Tkac I, Rao R, Georgieff MK, Gruetter R. Developmental and regional changes in the neurochemical profile of the rat brain determined by in vivo 1H NMR spectroscopy. Magn Reson Med. 2003;50:24–32. doi: 10.1002/mrm.10497. [DOI] [PubMed] [Google Scholar]
  34. Traka M, Wollmann RL, Cerda SR, Dugas J, Barres BA, Popko B. Nur7 is a nonsense mutation in the mouse aspartoacylase gene that causes spongy degeneration of the CNS. J Neurosci. 2008;28:11537–11549. doi: 10.1523/JNEUROSCI.1490-08.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Vagnozzi R, Tavazzi B, Signoretti S, Amorini AM, Belle A, Ciamatti M, Delfini R, Di Pietro V, Finocchiaro A, Lazzarino G. Temporal window of metabolic brain vulnerability to conscussions: mitochondrial related impairment-part I. Neurosurgery. 2007;61:379–388. doi: 10.1227/01.NEU.0000280002.41696.D8. [DOI] [PubMed] [Google Scholar]
  36. Vellinov M, Zellers N, Styles J, Wisniewski K. Homozygosity for mutation G212A of the gene for aspartoacylase is associated with atypical form of Canavan disease. Clin Genet. 2008;73:288–289. doi: 10.1111/j.1399-0004.2007.00934.x. [DOI] [PubMed] [Google Scholar]
  37. Wang J, Leone P, Wu G, Francis JS, Li H, Jain MR, Serikawa T, Ledeen RW. Myelin lipid abnormalities in the aspartoacylase-deficient tremor rat. Neurochem Res. 2009;34:138–148. doi: 10.1007/s11064-008-9726-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Cerebral Blood Flow & Metabolism are provided here courtesy of SAGE Publications

RESOURCES