Journal of Genetic Disorders & Genetic Reports ISSN: 2327-5790

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Research Article, J Genet Disor Genet Rep Vol: 3 Issue: 1

Evidence that lithium Inhibits Export of N-Acetyl-L-Aspartate from Neurons: A Retrospective Study of Canavan Disease and Bipolar Disorder Patients

Morris H Baslow* and David N Guilfoyle
Nathan S Kline Institute for Psychiatric Research, 140 Old Orangeburg Road, Orangeburg, NY 10962, USA
Corresponding author : Morris H Baslow
Nathan S Kline Institute for Psychiatric Research, 140 Old Orangeburg Road, Orangeburg, NY 10962, USA
Tel: +1-845-398-5471; Fax: +1-845-398-5472
E-mail: [email protected]
Received: November 22, 2013 Accepted: December 26, 2013 Published: January 03, 2014
Citation: Baslow MH, Guilfoyle DN (2014) Evidence that lithium Inhibits Export of N-Acetyl-L-Aspartate from Neurons: A Retrospective Study of Canavan Disease and Bipolar Disorder Patients. J Genet Disor Genet Rep 3:1. doi:10.4172/2327-5790.1000110

Abstract

Evidence that lithium Inhibits Export of N-Acetyl-L-Aspartate from Neurons: A Retrospective Study of Canavan Disease and Bipolar Disorder Patients

Lithium (Li) is an effective treatment for human bipolar disorder (BD) but whose precise mechanism and site of action are unknown. N-acetyl-L-aspartic acid (NAA) is an amino acid synthesized by and maintained at high steady-state levels within neurons from where it is exported to extracellular fluid (ECF) upon depolarization. NAA is the only precursor for N-acetylaspartylglutamate (NAAG), a neurotransmitter synthesized by neurons and also exported to ECF upon depolarization. The physiological function of NAA is as yet unclear but its unique tri-cellular metabolism between neurons, oligodendrocytes (NAA) and astrocytes (NAAG) is vital for normal brain function. Canavan disease (CD) is a rare inborn error in metabolism of NAA where oligodendrocyte aspartoacylase (ASPA) is inactive and NAA cannot be hydrolyzed resulting in its buildup in brain ECF and excretion in urine.

Keywords: Aspartoacylase; Bipolar disorder; Canavan disease; Hypoacetylaspartia; Lithium; N-acetyl-L-aspartate; N-acetyl-Laspartylglutamate

Keywords

Aspartoacylase; Bipolar disorder; Canavan disease; Hypoacetylaspartia; Lithium; N-acetyl-L-aspartate; N-acetyl-Laspartylglutamate

Introduction

N-acetyl-L-aspartate (NAA) is an amino acid present in brain whose physiological function is unknown but whose highly compartmentalized and unusual tri-cellular metabolism between neurons, astrocytes and oligodendrocytes is vital for normal brain function [1]. As such it has been described as the brain’s “operating system” [2]. NAA is synthesized by neurons from acetyl co-enzyme A (AcCoA) with the acetyl group derived from on-going oxidation of glucose (Glc), and from L-aspartate (Asp). NAA is one of the most abundant amino acids present in brain at about 10 mM, most of which at any given moment is stored in neurons that constitute about 50% of brain volume. At about 20 mM it is the most abundant amino acid present in neurons. NAA is not a neurotransmitter but it is the only known precursor for the neurotransmitter N-acetyl-Laspartylglutamate (NAAG), an adduct of NAA and glutamate (Glu) that is also synthesized by and is present in neurons at about 1 mM. The importance of the NAA metabolic system is attested to by the fact that about 1 in 40 Glc molecules taken up by neurons as a source of energy is diverted in order to synthesize an NAA molecule [3].
Although synthesized by neurons and maintained at high levels, NAA and NAAG cannot generally be catabolized by neurons. Nevertheless, they are dynamic and turn over in gray matter about every 14-16 h. Upon electric stimulation of rat brain slices at 15 Hz and 20 mA for 3 min both are released at an increased 3-fold rate to extracellular fluid (ECF) [4]. NAA is targeted to oligodendrocytes where it is rapidly hydrolyzed by aspartoacylase (ASPA), an enzyme highly expressed in these cells [5]. NAAG is targeted to the Glu metabotropic receptor 3 (GRM3) on the astrocyte surface where the Glu moiety is cleaved by NAAG peptidase liberating NAA to ECF and activating astrocytes to release second messengers to vascular endothelial cells signaling a need for additional energy. These pathways are illustrated in Figure 1.
Figure 1: Tri-cellular metabolism of NAA and NAAG.
While the precise function of the release and hydrolysis of NAA is as yet unclear, its unique and complex intercellular metabolism is vital for normal brain activity as evidenced by two human inborn errors in its metabolism, Canavan disease (CD) [6,7], and hypoacetylaspartia (HA) [8,9]. CD is a rare early-onset genetic spongiform leukodystrophy where ASPA is inactive so that NAA cannot be hydrolyzed. As a result, NAA builds up in brain ECF to a level approximately that of neurons as measured using magnetic resonance spectroscopy (MRS) within large vacuolar spaces in a case of CD [10] and all additional NAA synthesized and exported to ECF is balanced by its transport via the vascular system and excretion in urine [11]. The buildup of NAA in brain ECF in CD is associated with phenotypes that include splitting of extant myelin sheath layers at their inter-period lines, extensive vacuolization in white matter and failure of normal myelin formation. In a singular case of HA, another enzyme in the tri-cellular metabolism of NAA is affected. In HA, NAA synthase is inactive [9] so that NAA as well as its adduct NAAG are absent from neurons. In both CD and HA some neurons are clearly spared but there are profound disturbances in neuron signaling and brain function as evidenced by major motor and cognitive failures.
Lithium (Li) was first tested in a rat model of CD and showed that brain NAA levels returned partially towards normal levels within days [12]. However, it could not be determined from that study just how Li achieved this reduction. Possible mechanisms included (1) decreasing NAA in the neuron compartment, (2) blocking its release to ECF, or (3) increasing its rate of transport into the vascular system. Based on this animal model finding, several human CD cases have been treated with Li and similar results on NAA levels have been observed [11,13,14]. In these human cases it was found that there was a reduction of NAA level in some areas of the brain and where tested, also in urine. Along with the reduction in NAA, there was also a degree of normalization of both myelin formation and brain function. Li is an effective treatment for human bipolar disorder (BD) but whose precise mechanisms and sites of action are unknown [15]. In this retrospective analysis, results of the effect of Li on NAA levels in children with CD are compared to Li effects on NAA levels in cases of children with BD as well as in normal individuals.

Results of Literature Review

The use of Li to treat CD has provided a unique opportunity to evaluate a possible mechanism by which it may affect CD pathology and also may regulate brain activity in BD. The results of these studies are presented in Table 1. While the data base for CD is limited and the protocols and Li dosages vary, the consistency and magnitude of the changes in NAA observed warrant attention.
Table 1: Effects of Li on NAA levels in cases of Canavan disease and bipolar disorder.
A summary of the retrospective analysis of the results of these studies in CD and BD children on brain NAA levels and an evaluation of the singular case of HA indicate the following:
Li treatment in CD decreased brain NAA levels in both the short and the long-term
In Table 1, it is shown that the effects of Li treatment are consistent in CD cases and uniformly result in a marked and significant reduction in brain NAA. This is demonstrated in both CD children and in the rat model of CD. The rat CD model results also demonstrate that this reduction can occur within as short a period as 4 days after initiating treatment and the human results indicate that the effects of Li treatment on NAA levels are also long-term and can be observed for at least a year.
Li treatment in CD decreased NAA content in urine
Within the small sample of 8 CD children in three different studies, in only one case was a child reported to be tested for the presence of NAA in urine both before and after treatment. However, this case is important in that it clearly demonstrated that Li treatment reduced the NAA content in urine, a telltale phenotype in this disease, by a large and significant amount.
Li treatment in BD did not alter brain NAA levels
The measurement of brain NAA in the selected example of 11 normal children and in 11 untreated cases of BD in Table 1 indicated no significant differences in brain NAA content between these cohorts. In addition, Li treatment in the 11 cases of BD used as pre-treatment controls showed that brain NAA levels were not significantly different from either untreated BD or normal brain NAA levels.
The presence of NAA or NAAG in brain is not necessary for neuron survival or myelination but is required for normal brain function
The singular case of HA where NAA synthase is inactive and NAA and NAAG are absent indicates that these substances are not required for neuron survival or for neuron signaling. The formation of myelin in this unique individual also indicates that neither NAA nor ASPA activity are needed by oligodendrocytes in order to myelinate neuron axons. However, this case, like those of CD where ASPA is inactive again demonstrates the vital importance of an intact NAA tri-cellular metabolic system for maintenance of normal brain signaling and function.

Discussion

The use of Li to treat CD, a rare inborn error in metabolism of NAA, has provided an unusual insight into a possible mechanism by which Li therapy may be beneficial in CD, and also may affect and regulate brain activity in BD.
NAA exhibits both static and dynamic functions within the brain
Under normal conditions, NAA exhibits both static and dynamic functions even though the specific physiological roles of these functions are as yet unclear. Within the neuron soma NAA is maintained at high static levels varying by only a few % when measured in whole brain or brain regions over long periods of time including weeks, months and years. NAA is present in neurons at an estimated average concentration of about 20 mM (20 mOsmol/L) based on a neuronal compartment in brain of about 50%, and therefore serves as an important intracellular osmolyte in these cells. However, NAA in neurons is also dynamic and turns over every 14- 16 h or about 6%/h. The steady-state level of NAA in neuron soma is apparently maintained by a homeostatic mechanism which balances stimulation-induced perturbations in its rate of efflux (E) into ECF with its rate of synthesis (S) by neurons. In gray matter the average rate of synthesis is 0.55 μmol/g/h [3] and under average conditions its rates of synthesis, efflux and hydrolysis (H) by ASPA are equal so that S=E=H. Thus, if (E) increases due to an increase in stimulation, (S) must also increase. Under normal conditions (H) capacity is always much higher than (E) so that there is no buildup of NAA in ECF.
The average rate of NAA synthesis in human brain has been measured using a labeled 1-13C-Glc precursor [16]. The dynamic changes in the rate of NAA synthesis have been measured in normal human brain by functional MRS (fMRS) using high levels of visual stimulation over a 10 min period to increase (E) and perturb the system. This resulted in a many-fold increase in NAA (S) which returned NAA levels to the steady-state condition within a following 10 min rest period [17-19]. During intense visual stimulation, the apparent rate of NAA efflux increased 14.2-fold associated with a transient 13.1% decrease in the NAA signal, and during a subsequent 10 min recovery period the apparent rate of NAA synthesis increased 13.3-fold restoring the resting state NAA signal [18]. While the average rate of NAA synthesis in resting brain gray matter has been measured at 0.55 μmol/g/h [16], there is no comparable direct synthesis data currently available for the maximum rate of synthesis of NAA in brain that is possible under stimulatory conditions.
In CD the dynamic function of NAA is affected
In CD, ASPA is inactive so that (H) is zero. As a result, neurons continue to synthesize and export NAA upon stimulation so that that it builds up in brain ECF [10]. Since NAA in ECF cannot be hydrolyzed in CD, the major pathway for its removal from brain is by entering the blood from where it is filtered out by the kidneys, concentrated in urine and periodically excreted. It is the buildup of NAA in brain ECF to high osmotic levels that is considered to be the basic cause of CD leukodystrophic pathology since its complete absence in HA does not show a similar pathology.
How Li appears to interact with the NAA metabolic system
The original observation in CD rats and the subsequent studies in CD children showed that Li treatment rapidly reduced brain NAA toward normal levels. Based on the dynamics of the NAA system, this could be accomplished in several ways. Li could possibly decrease the level of NAA in the neuron compartment; it could block the efflux of NAA into ECF, or it could increase the rate of NAA transport into the vascular system. The several case studies of CD and BD children analyzed in this report have provided some answers.
Li does not decrease normal steady-state levels of NAA in brain: If Li decreased the static level of NAA in the neuron compartment; one would expect to see a corresponding reduction in brain NAA. The treatment of the 11 cases of BD children with Li shown in Table 1 indicates that Li had little or no effect on brain NAA levels and indeed the NAA MRS peak is reported to be the metabolite least affected by Li [20]. Similar studies involving larger BD cohorts treated with Li for periods of up to 5 years have also shown no significant effect on brain NAA levels [21,22]. In addition, Li tested in 12 normal individuals ages 8-52 for 4 weeks did not show any significant changes in brain NAA levels [23]. Finally, in two recent reviews of the biology and mechanisms of action of Li, no evidence of any significant decreases in brain NAA levels were reported [15,24]. However, there is one study involving 28 adolescents in a depressive episode where Li treatment was reported to significantly decrease NAA levels by about 7.5% in the medial ventral prefrontal area after 42 days although this effect was not observed in the adjacent left and right ventral lateral prefrontal cortices [25]. A further comparison with 10 healthy controls indicated that the “depressed” group had a significantly elevated mean baseline NAA in their prefrontal cortices suggesting that the effect of Li in these cases was to normalize rather than decrease static brain NAA levels. In another case it was observed that a BD cohort had lower than normal NAA levels and that Li treatment raised their NAA levels to the normal range [28]. Taken together, the weight of evidence in these studies and reviews indicates that Li treatment does not decrease the normal static level of NAA in the brain neuronal compartment and therefore cannot account for the large decreases in brain NAA levels observed in Li treated CD patients.
Li appears to inhibit the release of NAA from neurons: The consistent reduction in brain NAA, towards normal levels in 8 cases of CD children treated with Li for varying periods of up to 1 year, and the findings that Li treatment in most BD children and in normal individuals did not alter normal static levels of brain NAA, indicate that only the ECF compartment is involved in the elevation of brain NAA in CD. Therefore, it is reasoned that the change in brain NAA levels in response to Li treatment in CD is likely due to its reduction in the ECF compartment. ECF comprises about 20% of brain volume and its NAA content is normally very low at 80-100 μM [13]. In CD, its NAA content is elevated and is in the high mM range as measured by MRS [10]. If NAA were reduced by Li treatment to its normal level only in this compartment, the drop in brain NAA would be expected to be about 20%. This value is consistent with the results of Li treatment in the CD cases. A drop of greater than 20% would indicate an effect of Li on NAA in the neuronal compartment. Since Li use in CD lowers brain NAA levels by no more than 20% even after a year of treatment, and Li in most BD studies show no significant decrease in brain NAA levels, these studies complement one another and support the hypothesis that Li treatment in CD lowers brain NAA levels by inhibiting NAA efflux from neurons into the ECF compartment. This hypothesis as well as hypotheses that Li might facilitate the passage of NAA at the blood-brain barrier (BBB) into the blood or might increase the rate of renal excretion of NAA have previously been suggested based on results of Li treatment in a case of CD [11].
Li treatment in CD does not increase NAA permeability at the BBB: The last possibility for the observed change in brain NAA levels noted in CD upon treatment with Li is that the rate of removal of NAA out of brain ECF at the BBB and into the blood might be increased thus lowering ECF NAA levels as well as overall brain NAA levels. If this were the case, there would be no change in NAA excreted in urine since the rates of synthesis and efflux of NAA from neurons into ECF would not be changed. Thus, all NAA generated and exported each day would still have to be removed using the alternative blood-urine pathway. While NAA levels in urine of 6 treated CD patients were initially reported to be “high” [13], unfortunately there is only one published report where NAA in urine was actually measured before and after Li treatment. However, this case demonstrated that the level of NAA in urine was not maintained at its initial high level, but was significantly decreased by 80% after treatment. This finding when coupled with the reduction of brain NAA of no more than 20% upon Li treatment in CD, and no change in static brain NAA levels upon Li treatment in most cases of BD supports the notion that Li acts in CD by inhibiting the release of NAA into the ECF compartment and not by increasing its permeability across the BBB.

Conclusions

While brain levels of NAA can be very stable varying by only a few % over long periods of time, there are a number of factors including use of drugs that can alter these steady-state levels over periods of time ranging from minutes to years [29]. In one animal study, Li administration IP in normal rats resulted in a 9% decrease in NAA over a period of 14 days [30]. In another rat study, oral administration of Li resulted in a non-significant decline in NAA of 3% over a period of 15 days [31], but in mice, oral administration of Li for up to 33 days showed no significant changes in the NAA to creatine ratio [32]. Although the results of these animal studies using Li were mixed, they influenced our initial decision to test whether Li could reduce NAA levels in a rat model of CD [12].
In humans, most studies of the effect of Li on NAA in BD patients and in normal individuals failed to show any decrease in steady-state brain NAA levels [20-23]. However, it is now clear from studies of CD that Li interacts very strongly with the dynamics of the NAA metabolic system [11-14]. The original test of Li as a treatment in an animal model of CD where NAA could not be hydrolyzed showed a marked and significant decrease in the level of brain NAA after only four days [12]. Based on this finding, three possibilities were proposed: (1) that the level of NAA in neurons was decreased, (2) that the efflux of NAA into ECF was blocked, and (3) that the rate of transport of NAA into the vascular system was increased. Analysis of the effects of Li treatment in human CD and BD cases as well as in normal individuals has now provided a tentative answer. From these results, it is concluded that the action of Li in brain in CD is associated with blocking the release of NAA at the neuron plasma membrane- ECF boundary during depolarization. The site of this proposed action is illustrated in Figure 2.
Figure 2: Proposed site of interaction of Li with NAA intercellular dynamics.
Li treatment is considered to be a safe and effective agent for treatment of BD and is recommended as a first-line treatment for this disorder [15]. While Li is known to affect many metabolites and metabolic systems in the brain, the results of this retrospective analysis based on studies of CD show that Li interacts directly with the critically important NAA metabolic system in brain. This leads to the novel hypothesis that a previously unrecognized site of action of Li is on the depolarization-induced export of NAA by neurons, linking for the first time the effects of Li treatment with the dynamics of the unique tri-cellular NAA metabolic system. Thus, based on the analysis of Li use in cases of CD where a specific enzyme is inactive, a clue to its pharmacological action in BD appears to have been uncovered that would not have been predictable or readily apparent in other studies. While most studies of brain NAA, including the effects of drugs, focus on changes in its static level as a measure of neuron density or viability, the observation that Li has little or no effect on static levels of NAA but strongly affects NAA dynamics opens a potential new area for understanding relationships between other commonly used CNS drugs and brain function.

Future Studies

This report and hypothesis presented is based on a limited number of cases of CD patients. In addition, the possible role of Li on NAA metabolism was deduced from the outcome of Li treatment on only certain aspects of NAA metabolism including its normal levels in brain and its normal absence in urine. However, several studies to test the hypothesis that Li inhibits the release of NAA can now be envisioned. One study using an animal model to measure NAA turnover [33] would be to determine if the rate of NAA synthesis is changed by Li treatment. If Li reduces the efflux of NAA and S=E, it should be possible to demonstrate that the rate of NAA synthesis is also reduced. A second method using fMRS would be to test if visual stimulation in Li treated BD patients’ produces as significant a loss of NAA signal as in normal individuals where release of NAA is not inhibited and it is rapidly hydrolyzed by ASPA. Lastly, it is not known just how NAAG fits into the above hypothesis since it is also released by stimulated neurons to ECF [4]. However, since NAAG release and its hydrolysis activates astrocytes to send second messengers to the vascular system, a measure of the blood oxygenation level-dependent (BOLD) response of Li treated BD individuals using magnetic resonance imaging may be informative.

References

Track Your Manuscript

Media Partners

Associations