Background: Methamphetamine induces neuronal N-acetyl-aspartate synthesis in preclinical studies. In a preliminary human proton magnetic resonance spectroscopic imaging investigation, we also observed that N-acetyl-aspartate+N-acetyl- aspartyl-glutamate in right inferior frontal cortex correlated with years of heavy methamphetamine abuse. In the same brain region, glutamate+glutamine is lower in methamphetamine users than in controls and is negatively correlated with depression. N-acetyl and glutamatergic neurochemistries therefore merit further investigation in methamphetamine abuse and the associated mood symptoms. Methods: Magnetic resonance spectroscopic imaging was used to measure N-acetyl-aspartate+N-acetyl-aspartyl- glutamate and glutamate+glutamine in bilateral inferior frontal cortex and insula, a neighboring perisylvian region affected by methamphetamine, of 45 abstinent methamphetamine-dependent and 45 healthy control participants. Regional neurometabolite levels were tested for group differences and associations with duration of heavy methamphetamine use, depressive symptoms, and state anxiety. Results: In right inferior frontal cortex, -acetyl-aspartate+ N -acetyl-aspartyl-glutamate corr N elated with years of heavy methamphetamine use (r = +0.45); glutamate+glutamine was lower in methamphetamine users than in controls (9.3%) and correlated negatively with depressive symptoms (r = -0.44). In left insula, -acetyl-aspartate+ N -acetyl-aspartyl-glutamate w N as 9.1% higher in methamphetamine users than controls. In right insula, glutamate+glutamine was 12.3% lower in methamphetamine users than controls and correlated negatively with depressive symptoms (r = -0.51) and state anxiety (r = -0.47). Conclusions: The inferior frontal cortex and insula show methamphetamine-related abnormalities, consistent with prior observations of increased cortical N-acetyl-aspartate in methamphetamine-exposed animal models and associations between cortical glutamate and mood in human methamphetamine users. Received: September 20, 2017; Revised: March 21, 2018; Accepted: May 15, 2018 © The Author(s) 2018. Published by Oxford University Press on behalf of CINP. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http:// creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, 1 provided the original work is properly cited. For commercial re-use, please contact firstname.lastname@example.org Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 2 | International Journal of Neuropsychopharmacology, 2018 Significance Statement Methamphetamine Use Disorder is a prevalent, treatment-resistant, public health problem. Treatment often fails during early abstinence from the drug, when many patients experience depression and anxiety. In the inferior frontal cortex and the insula, which influence negative emotions, we measured levels of N-acetyl-compounds and glutamate compounds using magnetic reso- nance spectroscopy in methamphetamine users who had been <2 weeks abstinent and in healthy controls. Methamphetamine users, especially those with more years of heavy use, had higher levels of N-acetyl-metabolites than controls. Those with more severe depression and anxiety had lower levels of glutamate metabolites. These findings extend observations from animals and in vitro human cell cultures that methamphetamine increases neuronal N-acetyl aspartate concentration in living humans. They also motivate the development of medications that operate on N-acetyl and glutamatergic systems to assist treatment during early abstinence from methamphetamine. Keywords: methamphetamine,N -acetyl-aspartate, glutamate, abstinence, depression Introduction Amphetamine-type stimulants, particularly methampheta- Such changes could involve tNAA. In our pilot study of meth- mine, are a rising class of abused drugs worldwide (UNODC, amphetamine users abstinent ≤2 weeks (O’Neill et al., 2010), 2017). With few treatment options and no approved medications tNAA in right inferior frontal cortex correlated positively with for Stimulant Use Disorders (Courtney and Ray, 2014), continued duration of heavy methamphetamine abuse. Here we report on investigation of the neurochemical correlates of methamphet- such effects in our full sample in inferior frontal cortex and in a amine abuse is warranted. The present study focused on the neighboring perisylvian brain region, the insula. neurometabolites N-acetyl-aspartate (NAA) and glutamate (Glu) Much preclinical evidence also links methamphetamine in the inferior frontal cortex and the insula, 2 perisylvian brain with glutamatergic systems of the brain (Kalivas, 2007 2009 , ). regions that show abnormalities in methamphetamine users. In vivo in humans, Glu is also measured with MRS, but since NAA production, or at least expression or activity of the the Glu spectrum overlaps with that of its derivative glutamine, enzyme that catalyzes NAA biosynthesis, is enhanced by meth- the two are often assayed as a combined entity, Glx. One study amphetamine in the ex vivo mouse nucleus accumbens (Niwa showed no differences between methamphetamine users and et al., 2007), in rat pheochromocytoma-12 cells (Niwa et al., controls in Glu or Glx in middle frontal cortex and anterior 2008), and in human SH-SY5Y neuroblastoma cells (Ariyannur middle cingulate cortex (Howells et al., 2014), but others found et al., 2010). The NAA synthase enzyme is denoted in the litera- differences in pregenual anterior cingulate cortex, anterior mid- ture by several names, including N-acetyltransferase N-acetyl- , dle cingulate cortex, posterior cingulate cortex, and precuneus transferase-8-like protein, and shati, among others. Increase in (Ernst and Chang, 2008; Crocker et al., 2014; O’Neill et al., 2014). the activity of this enzyme may represent an adaptive response In right inferior frontal cortex, Glx was lower in methampheta- to methamphetamine-induced elevation of extracellular dopa- mine users than controls in a sample that overlapped with that mine, as the enzyme appears to diminish methampheta- of the current study (O’Neill et al., 2014), and Glx correlated mine-associated behavioral effects such as hyperlocomotion, negatively with severity of depressive symptoms. Depression sensitization, and conditioned place preference (Niwa et al., and other mood disorders can contribute to relapse in meth- 2007; Ariyannur et al., 2013). These observations and addi- amphetamine abuse (Glasner-Edwards et al., 2008 2009 , , 2010). tional rodent studies intimate a more direct role for N-acetyl- We found no previous MRS reports of Glu or Glx in the insula compounds in higher brain functions than had previously been of methamphetamine users. Glu interacts with NAA in multi- envisioned (Ariyannur et al., 2013) and suggest novel pharmaco- ple ways. NAAG is biosynthesized from NAA and Glu (Becker logical pathways to management of methamphetamine abuse. et al., 2010; Collard et al., 2010) and is biodecomposed into NAA In human brain, NAA is measured with H MRS. It is com- and Glu (Robinson et al., 1987). NAAG colocalizes with Glu in monly assayed alongside N-acetyl-aspartyl-glutamate (NAAG), synaptic vesicles (Neale et al., 2000) and is an antagonist of and the combined NAA+NAAG signal is abbreviated tNAA (total the N-methyl-D-aspartate glutamate receptor (Sekiguchi et al., N-acetyl-compounds). Contrasting with the aforementioned 1989). Therefore, it is of interest to investigate MRS tNAA and preclinical findings, MRS has indicated below-normal tNAA Glx together. in multiple brain regions of methamphetamine users (Ernst Previous studies suggested that the insula and the infe- et al., 2000; Nordahl et al., 2002, 2005; Salo et al., 2007, 2011a; rior frontal gyrus show metabolical and structural abnormali- Sung et al., 2007; Sailasuta et al., 2010). Because NAA and NAAG ties in chronic methamphetamine users (London et al., 2004; are abundant in neurons but not in other cells (Simmons et al., Thompson et al., 2004; Tabibnia et al., 2011; Morales et al., 2012), 1991; Urenjak et al., 1992), impaired neuronal metabolic activity and neuroimaging endpoints in these regions are associated in methamphetamine users is often offered as an explanation with affective symptoms. The insula participates in emotion of low tNAA. The relevant studies, however, overwhelmingly processing (Medford and Critchley, 2010Goerlic ; h-Dobre et al., sampled methamphetamine users in mid- to long-term absti- 2014), and anxiety covaries negatively with glucose metabolism nence from the drug and not during active abuse or early absti- in left insula of methamphetamine users (London et al., 2004). nence. Longitudinal investigations, including studies of cerebral The right inferior frontal gyrus has a special role in response glucose metabolism (Berman et al., 2008) and MRS (Ernst and inhibition (Aron et al., 2014), which is impaired in methamphet- Chang, 2008; Salo et al., 2011a), have shown that cortical metabo- amine users (Monterosso et al., 2005), and in inhibitory control lism changes over the course of methamphetamine abstinence. more generally (Tabibnia et al., 2011). Therefore, the insula and Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 Tang et al. | 3 the inferior frontal cortex were chosen for this planned analy- Corporation). The methamphetamine users were inpatients at sis of tNAA and Glx in relation to methamphetamine use and the UCLA General Clinical Research Center, where they main- symptoms of depression and anxiety. tained abstinence from all drugs of abuse other than nicotine (in cigarettes) and caffeine (in beverages), as verified by urine toxi- cology. Control participants visited the research site on different Methods days for psychological testing and imaging. The Beck Depression Inventory (BDI) and the Spielberger State-Trait Anxiety Inventory Research Participants (STAI Y1 score) were administered within 1 week of admission to assess depressive symptoms and state anxiety, respectively. All procedures were approved by the University of California Los Angeles Office for the Protection of Research Subjects. Participants (18–55 years old), who were recruited using Internet Magnetic Resonance Acquisition and newspaper advertisements, provided written informed con- As described in O’Neill et al. (2014), magnetic resonance spectro- sent. Forty-five individuals who met criteria for methampheta- scopic imaging (MRSI) using point-resolved spectroscopy (PRESS; mine dependence per DSM-IV (Table 1) and were not seeking repetition-time/echo-time = 1500/30 ms, voxels 11× 11 × 9 mm , treatment agreed to remain abstinent from methamphetamine 16 × 16 phase-encoding steps in-plane, 11:04-minute runt- (verified by random urine screening) while residing at a research ime including water-suppressed and non-water-suppressed unit for 2 weeks. They were compared with 45 healthy controls acquisitions) of bilateral inferior frontal cortex and insula and without history of substance abuse (apart from tobacco and/or whole-brain structural MRI (MPRAGE,× 1 1 × 1 mm voxels) were light use of marijuana or alcohol, defined as ≤1 joint per week or acquired at 1.5 T (Siemens Sonata, standard quadrature head ≤10 drinks of liquor or the equivalent of beer or wine per week) coil). MRSI was acquired from 2 sagittal-oblique 2D slices ori- or current use, as indicated by urine toxicology. The sample ented parallel to the left, respectively, right temple, set ~2 cm included here shared 42 methamphetamine and 23 control sub- deep into the brain, rotated parallel to the Sylvian fissure, and jects from a previous study (O’Neill et al., 2014) in which all the positioned to straddle this fissure dorsoventrally. Subject move- controls were smokers; the control group in the present report ment that could have occurred between the MPRAGE and MRSI included almost equal numbers of smokers and nonsmokers. scans was detected by comparing head position on the MPRAGE Methamphetamine dependence and absence of other psy- to that on a structural scan acquired after MRSI; scans were chiatric disorders were established using the Structured Clinical then repeated as needed. The MRSI prescription sampled infe- Interview for DSM-IV Axis I Disorders. Heavy methampheta- rior frontal cortex, insula, and superior temporal cortex. MRSI mine use was defined as using methamphetamine 3 times/wk was also acquired from an axial-oblique (genu-splenium paral- or having a 2-day binge each week. Smoker status was verified lel) slice at the level of the basal ganglia and extending from the by ≥10 ppm carbon monoxide in expired air (MicroSmokerlyzer; caudal edge of the corpus callosum anteriorly past the corpus Bedfont Scientific Ltd) and the presence of urinary cotinine callosum splenium posteriorly (Figure 1). These regions were (≥200 ng/mL by Accutest NicAlert strips; JANT Pharmacal sampled at a nominal voxel size of 1.1 cc. The effective voxel size, however, is larger, because the voxel is modestly smeared Table 1. Characteristics of Research Participants by spatial apodization and the point-spread function of MRSI. The effective voxel size is not trivial to calculate, but using fac- Methamphetamine tors obtained for other MRSI techniques (Théberge et al., 2005; (n = 45) Control (n = 45) P Posse et al., 2007) we estimate it at 1.3 to 1.4 cc. Sex .99 Female 24 23 Magnetic Resonance Post-Processing Male 21 22 Age, y 33.0 ± 9.3 32.9 ± 8.6 .94 MRSI was post-processed as in O’Neill et al. (2014) with minor Education, y 11.8 ± 2.3 13.5 ± 1.8 <.0005 update in methods for the insula. Briefly, MR spectra were fit Depression, BDI Score 15.6 ± 13.0 2.3 ± 2.8 <.0005 automatically with LCModel yielding levels of tNAA and Glx Anxiety, STAI Y1 Score 33.1 ± 9.8 27.0 ± 7.4 .002 normed to the unsuppressed water resonance and expressed Cigarette smoking in Institutional Units (IU). Each MPRAGE was segregated in Smokers, n 41 23 <.0005 ICBM152 space into gray matter, white matter, and CSF subvol- Pack-years tobacco 11.7 ± 11.4 9.7 ± 7.9 .42 umes using SPM (http://www.fil.ion.ucl.ac.uk/spm/software/). (smokers only) The subvolumes were brought back into native space using the Fagerström Score 3.4 ± 2.1 3.6 ± 2.1 .78 inverse of the MPRAGE-to-ICBM152 transform. Inferior frontal (smokers only) cortex MRSI voxels were selected manually as in O’Neill et al. Methamphetamine use (2014). For the insula, MRSI voxel selection was automated with Duration of use, y 11.1 ± 7.8 0 ± 0 — the help of binary masks of left and right insula generated from Duration of heavy 6.7 ± 6.4 0 ± 0 — the MPRAGE by FreeSurfer (https://surfer.nmr.mgh.harvard. use, y edu/) in ICBM152 space. Again, masks were restored to native Current use, g/wk 2.0 ± 1.4 0 ± 0 — space with the inverse transform. Then, for both inferior frontal Marijuana use cortex and insula, using custom-written software, the MPRAGE, Marijuana users, n 27 4 <.0005 Current use, d/month 5.5 ± 9.2 0.2 ± 0.8 <.0005 tissue subvolumes, and masks were aligned with each MRSI slice; the tissue-composition of each MRSI voxel was calculated; 2 MRSI voxels were retained with >50% overlap with the target Shown are numbers of participants or group means ± SDs. P values are for Χ brain region (exclusive of CSF); and neurometabolite levels were (sex, number of smokers, number of marijuana users) or 2-way independent t corrected for voxel CSF content. Typically for all 4 brain regions tests (all other variables), unless otherwise indicated for comparisons of the full and all 3 subject groups, 1 to 4 voxels per region passed all methamphetamine to the full control sample. Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 4 | International Journal of Neuropsychopharmacology, 2018 Figure 1. (A) sagittal-oblique T1-weighted MRI of the brain showing sample coregistration of FreeSurfer right insula volume-of-interest (yellow-red) with superim- posed right perisylvian H magnetic resonance spectroscopic imaging (MRSI) point-resolved spectroscopy (PRESS) excitation volume (“slice,” white rectangle). A simi- lar sagittal-oblique PRESS slice sampled the left insula. (B) Overlap of the right insula volume with a third, axial-oblique PRESS slice. The opposite side of this slice cosampled the left insula. Each slice was 9 mm thick and consisted of a rectangular in-plane array of 11 × 11 mm voxels (green grids). (C) Sample right perisylvian MRSI spectrum showing high-amplitude, low-noise, well-separated peaks for the target neurometabolites N-acetyl-aspartate+N-acetyl-aspartyl-glutamate (tNAA) and glutamate+glutamine (Glx). quality-control criteria and were included in the subject’s a -ver comparing methamphetamine users with control smokers age for that region. Thus, differences in voxel sampling were and control nonsmokers separately using independent t tests. unlikely to have affected results and—given the strict quality- Within-group relationships between regional metabolites and control criteria, the voxel size relative to region volume, and the years of heavy methamphetamine abuse, BDI Score, and STAI substantial local tissue content of voxels—the spectrum of even Y1 Score were examined with Pearson correlation, Bonferroni- one voxel was deemed sufficient to represent a region. corrected for multiple comparisons. Significant findings were retested as Spearman correlations, partialling out demographic, clinical, or tissue content variables that may have significantly Correction for Differences in MRSI Voxel affected the metabolite level in question. Spearman rather than Tissue-Composition Pearson partial correlation was chosen, since the former is more The group-mean volume fraction of gray matter, white matter, conservative, less biased by outliers, and does not a priori pre- and CSF in MRSI voxels did not differ significantly between the sume linearly related and normally distributed variables. Based 2 samples, except in the right insula, where the volume fraction on our pilot study (O’Neill et al., 2010), we hypothesized that of CSF was slightly (12% vs 14%) but significantly lower in meth- tNAA would be higher in the methamphetamine-dependent amphetamine users than in controls (P = .030, independent test). t vs control subjects and/or would be correlated positively with MRSI neurometabolite levels were corrected for voxel CSF content years of heavy methamphetamine abuse in one or more perisyl- to adjust for any such differences. Voxel white matter content was vian regions. The criterion for statistical significance was P < .05, trendwise (P = .088) lower (3% vs 4%) while gray matter content whereby in cases with Bonferroni-correction the raw P value was slightly higher (86% vs 84%; P = .052) in methamphetamine was multiplied by the number of multiple comparisons. users; therefore, gray and white matter content were included as covariates in the relevant analyses to adjust for such differences. Results Statistical Analyses Sample Characteristics MRSI voxels within the PRESS excitation volume with ≥50 vol- The methamphetamine users and controls were well matched ume% of each target region were retained. Spectra with linewidth for sex and age (Table 1 ), although the methamphetamine users >0.10 ppm, signal-to-noise ratio <5, or not passing operator inspec- had approximately 2 fewer years of education than controls on tion were discarded, as were metabolite signals with Cramer-Rao average (P < .0005) and most smoked cigarettes, whereas about Lower Bounds >20%. Metabolite data passing the foregoing quality- one-half of the controls were smokers. Lifetime tobacco expo- control criteria were averaged together for each subject to obtain a sure (pack-years) and current tobacco dependence (Fagerström representative value for each metabolite region combination (e.g., Score) did not differ significantly between smokers in the two tNAA in left insula). These quality control measures resulted in groups. Although both groups reported light marijuana use, the data from different numbers of participants contributing to the methamphetamine users reported more (P < .0005); therefore, various analyses across regions and metabolites (Table 2). marijuana use was included as a covariate in between-group Group differences in mean clinical endpoints were exam- comparisons. The groups differed on self-reports of mood. The ined with independent t tests. Between-group differences mean BDI score was ~6× higher (P < .0005, independent t test), in regional metabolite levels were tested using ANCOVA, and state anxiety was 23% higher (P = .002) in the methampheta- Bonferroni-corrected for multiple comparisons. Covariates mine users than controls. included age, sex, tobacco-smoking status, marijuana use sta- tus, and white mater and gray matter content. Given potential Regional Neurometabolite Levels effects of tobacco use on neuroimaging endpoints, particularly in the insula (Morales et al., 2014 Naqvi et ; al., 2014), significant Across the various comparisons, the number of subjects per group between-group differences in metabolite levels were retested with data passing quality control varied from 31 to 44 (Tabl ). e 2 Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 Tang et al. | 5 In left insula, tNAA was 9.1% higher in the methamphetamine (Bonferroni-corrected P = .004, ANCOVA). Although they were in than the control sample (Bonferroni-corrected P = .0064, ANCOVA no identifiable way atypical of the sample and none had a tNAA covarying sex, age, smoking status, marijuana status, and gray level ≥4 SDs above the mean of the remaining sample (a com- and white matter). Of the 31 subjects with passing data quality in monly used criterion for outliers), group-mean tNAA was still this region, 23 (74.2%) had tNAA values greater than the control higher for methamphetamine than for control subjects at trend mean. This difference remained significant when the metham- level (P = .086) when the 3 methamphetamine subjects with high- phetamine users were compared separately with the smoking est tNAA in this region were removed on an exploratory basis. (8.8% higher; P = .019, independent t test) and nonsmoking (10.0% When the full samples were compared using nonparametric sta- higher; P = .038) controls (Figure 2). The difference in tNAA, com- tistics (less sensitive to outliers), tNAA was significantly higher paring the methamphetamine and total control samples, also for methamphetamine than for control subjects (P = .023). There remained significant when BDI score was added as a covariate were no significant group differences in tNAA in other regions. Table 2. Levels of tNAA and Glx in Perisylvian Cortices Left Inferior Frontal Cortex Right Inferior Frontal Cortex Methamphetamine Control Methamphetamine Control tNAA 8.2 ± 1.2 8.1 ± 1.0 8.2 ± 1.0 8.2 ± 0.9 Participants 41 44 37 41 Glx 12.4 ± 2.2 12.8 ± 2.2 12.7 ± 2.5* 14.0 ± 2.4 Left insula Right insula Methamphetamine Control Methamphetamine Control tNAA 8.4 ± 1.3** 7.7 ± 0.9 7.8 ± 1.0 8.1 ± 2.1 Participants 31 31 37 37 Glx 9.8 ± 2.7 9.3 ± 2.0 9.3 ± 2.0* 10.6 ± 3.5 Indicated are numbers of participants with usable MRSI data for each metabolite in each target brain region. tNAA and Glx levels are group means ± SD in IU , corrected for voxel CSF content. Significant between-group comparisons in bold (*P < .05, **P< .01 Bonferroni-corrected for multiple comparisons, ANCOVA covarying for age, sex, tobacco-smoking sta- tus, marijuana use status, and white and gray matter content). Figure 2. (A) CSF-corrected levels of N-acetyl-aspartate+N-acetyl-aspartyl-glutamate (tNAA) (NAA+NAA in Institutional G) Units (IU) in left insula of methamphetamine users (MA), control smokers, and control nonsmokers. Horizontal bars denote group means. < .05 for differ *P ence between the methamphetamine users and controls by independent t test. (B) Associations of tNAA in right inferior frontal cortex vs years of heavy methamphetamine use. P = .032, Pearson with Bonferroni correction. P = .047, (r = +0.34, Spearman) when effects of age, BDI Score, and pack-years of smoking were partialled out. (C–D) Negative associations of right insula glutamatergic compounds with symptoms of depression and anxiety in methamphetamine users. Y axes indicate CSF-corrected levels of Glx (Glu+ glutamine) in Institutional Units (IU). Depressive symptoms were self-reported using the Beck Depression Inventory (BDI), and state anxiety was self-reported using the Spielberger State-Trait Anxiety Inventory. P < *.05, ***.0005, Spearman with Bonferroni correction, partialling sex, age, smoking status, marijuana status, and gray and white matter. Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 6 | International Journal of Neuropsychopharmacology, 2018 Glx was significantly lower in the right inferior cortex (9.3% include stimulation of NAA synthase by methamphetamine as lower; Bonferroni-corrected P = .004, ANCOVA covarying sex, in preclinical models (Niwa et al., 2007 2008 , ; Ariyannur et al., age, smoking status, marijuana status, and gray and white mat- 2010), deceleration of NAA degradation, and (possibly unto- ter) and in the right insula (12.3% lower; Bonferroni-corrected ward) neuroplastic increase in neuronal mass or metabolic P = .0004) in methamphetamine users than in controls (Table ). 2 activity, and others. Noninvasive in vivo human proton MRS There were no significant between-group differences in Glx in is not capable of distinguishing among these various mecha- other regions. nisms, which are also not mutually exclusive. The tNAA effect There was a significant main effect of sex on Glx in left was small, comprising an approximately 10% group-mean dif- insula (P = .048), whereby Glx was 13.4% higher in males than ference (left insula) and a 0.07 IU/y increment (right inferior in females. As indicated, groups were well balanced for sex frontal cortex) and was not statistically significant in all regions (Table 1) and sex was included as a covariate in statistical analy- tested. Although significant effects on the order of 5% to 15% ses. There were no significant main effects of marijuana use sta- are common in MRS studies of brain N-acetyl- (and glutamater - tus on regional metabolites. gic) compounds in methamphetamine abuse (Ernst et al., 2000; Nordahl et al. 2002Salo et ; al., 20072011a , , 2011b; Sailasuta et al., 2010; Cloak et al., 2011; Crocker et al., 2014; Howells et al., 2014), Correlations of tNAA with Duration of the subtlety of the present tNAA effects may reflect superim- Methamphetamine Abuse position of tNAA elevation from methamphetamine on tNAA In right inferior frontal cortex, tNAA correlated positively with declines due to aging (Aoki et al., 2012), chronic cigarette smok- years of heavy methamphetamine abuse (r = +0.43, Bonferroni- ing (Durazzo et al., 2016), or other factors. That tNAA correlated corrected P = .032, Pearson; Figure 2). This association remained with duration of heavy (as opposed to any) methamphetamine significant when partialling age, BDI Score, and pack-years of abuse is consistent with in vitro observations that methamphet- smoking (r = +0.39, P = .047, Spearman). tNAA also correlated amine significantly increased NAA only at higher methamphet- positively with years of any methamphetamine abuse, but this amine doses (Ariyannur et al., 2010). Overall, these results are result did not survive Bonferroni correction. There were no sig- consistent with a gradual rise in tNAA with continuing metham- nificant correlations of tNAA with duration of heavy metham- phetamine exposure, perhaps as an adaptation to excess dopa- phetamine abuse in other regions or correlations of Glx with mine signaling (Niwa et al., 2007). It is recommended that tNAA duration of heavy methamphetamine abuse in any region. in early abstinence from methamphetamine be reinvestigated for verification in further samples and at high-field strength. Prior studies of methamphetamine users documented Correlations of Neurometabolite Levels with below-normal tNAA in some brain regions (Ernst et al., 2000; Symptoms of Depression and Anxiety Nordahl et al., 2002, 2005; Salo et al., 2007, 2011a; Sung et al., There were no significant associations of tNAA with BDI Score 2007; Sailasuta et al., 2010). Rather than inferior frontal cortex or STAI Y1 Score. Glx was negatively correlated with BDI Score and insula, as sampled here, these studies interrogated other in right inferior frontal gyrus (r = -0.51, Bonferroni-corrected brain regions (pregenual anterior and anterior middle cingulate P = .0052, Spearman, partialling sex, age, smoking status, mari- cortices, prefrontal white matter, caudate-putamen, occipital juana status, and gray and white matter) and right insula cortex, and dorsal posterior cingulate cortex) than those exam- (r = -0.44, Bonferroni-corrected P = .0004, partialling sex, age, ined here. Moreover, participants in those studies had been smoking status, marijuana status, and gray and white matter). abstinent from methamphetamine for more than the 2-week In right insula, Glx was additionally significantly negatively maximum in this study. Methamphetamine effects on tNAA correlated with STAI Y1 Score (r = -0.47, Bonferroni-corrected may vary with brain region or duration of abstinence. In early P = .024, Spearman, partialling sex, age, smoking status, mari- abstinence, tNAA elevations in compensation to dopaminergic juana status, and gray and white matter). hyperactivation due to methamphetamine may predominate in certain regions over tNAA decreases from neuronal loss or impairment. In mid-term abstinence, the former effect may Discussion fade leading the latter effects to predominate. Finally, in late This study yielded 3 major findings: (1) tNAA in methampheta- abstinence rise of tNAA may be observed with recovery of func- mine users was moderately elevated in the left insula and in tion by some neurons (Salo et al., 2011a). Although, if metham- the right inferior frontal cortex tNAA was positively correlated phetamine enhances NAA production, one might anticipate a with years of heavy methamphetamine exposure; (2) in the right decrease in NAA production during abstinence due to removal inferior frontal cortex and right insula, Glx was lower in meth- of the enhancing factor; it may take time, especially after years amphetamine users; and (3) in those regions, Glx was negatively of methamphetamine abuse as in our sample, for such decrease correlated with severity of depression and anxiety symptoms to become detectable. Hence, present tNAA observations may, in in methamphetamine users. Thus, chronic methamphetamine some aspects, reflect “overhang” from the chronic state of active users exhibit modest increments in regional cortical tNAA lev- abuse rather than acute changes precipitated by abstinence. els, similar to those observed in ex vivo and in vitro systems. Inflammatory processes may influence MRS neurometabo- Results also support previous findings associating the mood lite levels (Chang et al., 2013) and the evolution of brain metabo- symptoms of early methamphetamine abstinence with gluta- lism during recovery from methamphetamine abuse (Volkow matergic metabolism. et al., 2001; Berman et al., 2008). We previously suggested that Mean tNAA was greater than the control mean in one brain higher cortical tNAA in early abstinence from methampheta- region of methamphetamine abusers, and tNAA correlated mine could be due to local invasion of microglia as part of the with duration of heavy methamphetamine exposure in another inflammatory response (O’Neill et al., 2010), since activated region. These effects remained significant after accounting for microglia contain high concentrations of NAAG (Passani et al., age, tobacco smoking, MRSI voxel tissue-composition, and BDI 1998). NAA itself, moreover, acts as an antiinflammatory agent Score, and for multiple comparisons. Possible explanations for reactive astroglia (Rael et al., 2004). Hence, an inflammatory Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 Tang et al. | 7 response in early abstinence could initiate a transient rise in notions of a possible glutamatergic basis for the mood symp- tNAA levels. Longitudinal cortical MRS measurements may help toms of early abstinence from methamphetamine. answer such questions. Findings also included Glx approximately 10% lower in right Acknowledgments inferior frontal cortex and insula in methamphetamine users and that Glx in these regions was negatively correlated with This work was supported by the National Institute on Drug Abuse depression and anxiety. The findings in right inferior frontal (P20DA022539 to E.D.L., R01DA020726 to E.D.L., R03DA20512 cortex largely replicated those in O’Neill et al. (2014), which had to J.O.N., and R21DA023192 to J.O.N.); the National Center for essentially the same methamphetamine sample. However, the Research Resources (MOIRR00865 from the UCLA General control sample in the present analysis was now as large as the Clinical Research Center); and endowments from the Thomas methamphetamine sample, rather than being about one-half P. and Katherine K. Pike Chair in Addiction Studies (E.D.L.) and the size. The present study extends findings to the neighbor - the Marjorie M. Greene Trust. ing right insula, where a negative relation of Glx with anxiety was also found. As discussed in O’Neill et al. (2014), depression Statement of Interest appears to be associated with glutamatergic systems and with the right forebrain. Glutamatergic agents might offer promise None. in combatting the often severe depressive symptoms of early abstinence from methamphetamine (Zorick et al., 2011). Present References findings also motivate exploration of MRS Glx as a potential bio- marker of relapse in methamphetamine abuse in systematic Aoki Y, et al. (2012) Absence of age-related prefrontal NAA clinical trials. change in adults with autism spectrum disorders. Transl Overall, the current findings add to evidence pointing to Psychiatry 2:e178. (especially right) inferior frontal cortex and insula as sites of Ariyannur PS, Moffett JR, Manickam P, Pattabiraman N, Arun P, sequelae of methamphetamine abuse (London et al., 2004; Nitta A, Nabeshima T, Madhavarao CN, Namboodiri AM (2010) Thompson et al., 2004; Tabibnia et al., 2011; Morales et al., 2012; Methamphetamine-induced neuronal protein NAT8L is the O’Neill et al., 2014Okita et ; al., 2016). That the major findings were NAA biosynthetic enzyme: implications for specialized acetyl hemispherically lateralized is not surprising, as functional later - coenzyme A metabolism in the CNS. Brain Res 1335:1–13. alization within perisylvian cortices is well known. Prominent Ariyannur PS, Arun P, Barry ES, Andrews-Shigaki B, Bosomtwi A, examples include lateralization of language (Ojemann, 1991), Tang H, Selwyn R, Grunberg NE, Moffett JR, Namboodiri AM mood symptoms following cerebrovascular insult (Starkstein (2013) Do reductions in brain N-acetylaspartate levels con- et al., 1989; Morris et al., 1996), and motor response inhibition tribute to the etiology of some neuropsychiatric disorders? J (Aron et al., 2014). In fact, a possible alternative explanation Neurosci Res 91:934–942. of our Glx findings in right inferior frontal cortex and insula Aron AR, Robbins TW, Poldrack RA (2014) Inhibition and the (Table 2) is that controls have a higher Glx on the right side, per - right inferior frontal cortex: one decade on. Trends Cogn Sci haps reflecting functional lateralization that does not pertain in 18:177–185. the methamphetamine subjects. We have previously observed Arun P, Madhavarao CN, Moffett JR, Namboodiri AM (2008) effects on glucose metabolism lateralized to the left insula in Antipsychotic drugs increase N-acetylaspartate and methamphetamine abuse (London et al., 2004). It remains a N-acetylaspartylglutamate in SH-SY5Y human neuroblas- challenge of neuroimaging research to delineate more fully the toma cells. J Neurochem 106:1669–1680. course, character, and extent of methamphetamine-associated Becker I, Lodder J, Gieselmann V, Eckhardt M (2010) Molecular neurophysiological effects on these regions and their possible characterization of N-acetylaspartylglutamate synthetase. J role in therapeutic response. Biol Chem 285:29156–29164. This investigation had limitations including MRS acquisition Berman SM, Voytek B, Mandelkern MA, Hassid D, Isaacson A, at 1.5 T. Studies at 3 T would yield larger and better separated Monterosso J, Miotto K, Ling W, London ED (2008) Changes metabolite signals. Acquisition at ultrahigh-field (e.g., 7 T) would in regional cerebral metabolism during early abstinence further enable separation of the NAA and NAAG components of from chronic methamphetamine abuse. Molec Psychiatry the tNAA signal. In addition, cross-sectional, rather than longi- 13:897–908. tudinal, data were used. As is common in studies of stimulant Chang L, Munsaka SM, Kraft-Terry S, Ernst T (2013) Magnetic abuse, marijuana use was much more prevalent in the meth- resonance spectroscopy to assess neuroinflammation and amphetamine than in the control sample, representing a poten- neuropathic pain. J Neuroimmune Pharmacol 8:576–593. tial confound. Marijuana use status, however, did not affect Cloak CC, Alicata D, Chang L, Andrews-Shigaki B, Ernst T (2011) regional metabolite endpoints and was included as a statistical Age and sex effects levels of choline compounds in the anter - covariate. These weaknesses are balanced by certain strengths, ior cingulate cortex of adolescent methamphetamine users. including: high spatial resolution (nominally approximately 1.1 Drug Alcohol Depend 119:207–215. cc, effectively approximately 1.3–1.4 cc) of MRSI; a rarely used Collard F, Stroobant V, Lamosa P, Kapanda CN, Lambert DM, sagittal-oblique MRSI prescription to access inferior frontal - cor Muccioli GG, Poupaert JH, Opperdoes F, Van Schaftingen E tex and insula; water-referenced neurometabolite levels rather (2010) Molecular identification of N-acetylaspartylglutamate than ratios to creatine+phosphocreatine; determination of MRSI synthase and beta-citrylglutamate synthase. J Biol Chem voxel tissue composition and its use in quantitation; a well-char - 285:29826–29833. acterized methamphetamine user sample in recent abstinence Courtney KE, Ray LA (2014) Methamphetamine: an update on from drug; samples fairly well matched for sex; and considera- epidemiology, pharmacology, clinical phenomenology, and tion of tobacco smoking. The tNAA findings represent a possible treatment literature. Drug Alcohol Depend 143:11–21. human in vivo expression of methamphetamine effects previ- Crocker CE, Bernier DC, Hanstock CC, Lakusta B, Purdon SE, Seres ously observed only preclinically, and the Glx findings reinforce P, Tibbo PG (2014) Prefrontal glutamate levels differentiate Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 8 | International Journal of Neuropsychopharmacology, 2018 early phase schizophrenia and methamphetamine addiction: Morris PL, Robinson RG, Raphael B, Hopwood MJ (1996) Lesion a (1)H MRS study at 3tesla. Schizophr Res 157:231–237. location and poststroke depression. J Neuropsychiatry Clin Durazzo TC, Meyerhoff DJ, Mon A, Abé C, Gazdzinski S, Murray Neurosci 8:399–403. DE (2016) Chronic cigarette smoking in healthy middle- Naqvi NH, Gaznick N, Tranel D, Bechara A (2014) The insula: a aged individuals is associated with decreased regional brain critical neural substrate for craving and drug seeking under N-acetylaspartate and glutamate levels. Biol Psychiatry conflict and risk. Ann N Y Acad Sci 1316:53–70. 79:481–488. Neale JH, Bzdega T, Wroblewska B (2000) Ernst T, Chang L (2008) Adaptation of brain glutamate plus glu- N-acetylaspartylglutamate: the most abundant peptide tamine during abstinence from chronic methamphetamine neurotransmitter in the mammalian central nervous system. use. J Neuroimmune Pharmacol 3:165–172. J Neurochem 75:443–452. Ernst T, Chang L, Leonido-Yee M, Speck O (2000) Evidence for Niwa M, Nitta A, Mizoguchi H, Ito Y, Noda Y, Nagai T, Nabeshima long-term neurotoxicity associated with methamphetamine T (2007) A novel molecule “shati” is involved in methamphet- abuse: A 1H MRS study. Neurology 54:1344–1349. amine-induced hyperlocomotion, sensitization, and condi- Glasner-Edwards S, Marinelli-Casey P, Hillhouse M, Ang A, tioned place preference. J Neurosci 27:7604–7615. Mooney LJ, Rawson R, the Methamphetamine Treatment Niwa M, Nitta A, Cen X, Kitaichi K, Ozaki N, Yamada K, Nabeshima Project Corporate Authors (2009) Depression among meth- T (2008) A novel molecule ‘shati’ increases dopamine uptake amphetamine users association with outcomes from the via the induction of tumor necrosis factor-alpha in pheochro- Methamphetamine Treatment Project at 3-year follow-up. mocytoma-12 cells. J Neurochem 107:1697–1708. Am J Addictions 17:99–102. Nordahl TE, Salo R, Possin K, Gibson DR, Flynn N, Leamon M, Glasner-Edwards S, Mooney LJ, Marinelli-Casey P, Hillhouse M, Galloway GP, Pfefferbaum A, Spielman DM, Adalsteinsson E, Ang A, Rawson R, the Methamphetamine Treatment Project Sullivan EV (2002) Low N-acetyl-aspartate and high choline Corporate Authors (2008) Identifying methamphetamine in the anterior cingulum of recently abstinent methamphet- users at risk for major depressive disorder: findings from the amine-dependent subjects: a preliminary proton MRS study. methamphetamine treatment project at three-year follow- Magnetic resonance spectroscopy. Psychiatry Res 116:43–52. up. Drug Alcohol Rev 29:12–20. Nordahl TE, Salo R, Natsuaki Y, Galloway GP, Waters C, Moore Glasner-Edwards S, Mooney LJ, Marinelli-Casey P, Hillhouse M, CD, Kile S, Buonocore MH (2005) Methamphetamine users in Ang A, Rawson RA, Methamphetamine Treatment Project sustained abstinence: a proton magnetic resonance spectros- Corporate Authors (2010) Psychopathology in methampheta- copy study. Arch Gen Psychiatry 62:444–452. mine-dependent adults 3 years after treatment. Drug Alcohol Ojemann GA (1991) Cortical organization of language. J Neurosci Rev 29:12–20. 11:2281–2287. Goerlich-Dobre KS, Bruce L, Martens S, Aleman A, Hooker CI Okita K, Ghahremani DG, Payer DE, Robertson CL, Mandelkern (2014) Distinct associations of insula and cingulate volume MA, London ED (2016) Relationship of alexithymia ratings to with the cognitive and affective dimensions of alexithymia. dopamine D2-type receptors in anterior cingulate and insula Neuropsychologia 53:284–292. of healthy control subjects but not methamphetamine- Howells FM, Uhlmann A, Temmingh H, Sinclair H, Meintjes E, dependent individuals. Int J Neuropsychopharm 19:1–8. Wilson D, Stein DJ (2014) (1)H-magnetic resonance spec- O’Neill J, Thomas S, Hudkins M, Dean A, Tobias MC, London ED troscopy ((1)H-MRS) in methamphetamine dependence (2010) MRSI and a model of metabolic dysfunction in human and methamphetamine induced psychosis. Schizophr Res methamphetamine dependence. Paper presented at: NIDA 153:122–128. Translational Research in Methamphetamine Addiction Kalivas PW (2007) Cocaine and amphetamine-like psycho- Conference; July 21, 2010; Pray, MT. stimulants: neurocircuitry and glutamate neuroplasticity. O’Neill J, Tobias MC, Hudkins M, London ED (2014) Glutamatergic Dialogues Clin Neurosci 9:389–397. neurometabolites during early abstinence from chronic Kalivas PW (2009) The glutamate homeostasis hypothesis of methamphetamine abuse. Int J Neuropsychopharmacol addiction. Nat Rev Neurosci 10:561–572. 18:1–9. London ED, Simon S, Berman SM, Mandelkern MA, Lichtman AM, Passani L, Elkabes S, Coyle JT (1998) Evidence for the presence of Bramen J, Shinn AK, Miotto K, Learn J, Dong Y, Matochik JA, N-acetylaspartylglutamate in cultured oligodendrocytes and Kurian V, Newton T, Woods R, Rawson R, Ling W (2004) Mood LPS activated microglia. Brain Res 794:143–145. disturbances and regional cerebral metabolic abnormalities Posse S, Otazo R, Caprihan A, Bustillo J, Chen H, Henry PG, in recently abstinent methamphetamine abusers. Arch Gen Marjanska M, Gasparovic C, Zuo C, Magnotta V, Mueller B, Psychiatry 61:73–84. Mullins P, Renshaw P, Ugurbil K, Lim KO, Alger JR (2007) Proton Medford N, Critchley HD (2010) Conjoint activity of anterior insu- echo-planar spectroscopic imaging of J-coupled resonances lar and anterior cingulate cortex: awareness and response. in human brain at 3 and 4 tesla. Magn Reson Med 58:236–244. Brain Struct Funct 214:535–549. Rael LT, Thomas GW, Bar-Or R, Craun ML, Bar-Or D (2004) An Monterosso JR, Aron AR, Cordova X, Xu J, London ED (2005) anti-inflammatory role for N-acetyl aspartate in stimu- Deficits in response inhibition associated with chronic meth- lated human astroglial cells. Biochem Biophys Res Commun amphetamine abuse. Drug Alcohol Depend 79:273–277. 319:847–853. Morales A, Lee B, Hellemann G, O’Neill J, London ED (2012) Gray- Robinson MB, Blakely RD, Couto R, Coyle JT (1987) Hydrolysis matter volume in methamphetamine dependence: cigarette of the brain dipeptide N-acetyl-L-aspartyl-L-glutamate. smoking and changes with abstinence from methampheta- Identification and characterization of a novel N-acetylated mine. Drug Alcohol Depend 125:230–238. alpha-linked acidic dipeptidase activity from rat brain. J Biol Morales AM, Ghahremani DG, Kohno M, London ED (2014) Chem 262:14498–14506. Cigarette exposure, dependence and craving are Sailasuta N, Abulseoud O, Hernandez M, Haghani P, Ross BD related to insula thickness in young adult smokers. (2010) Metabolic abnormalities in abstinent methampheta- Neuropsychopharmacology 39:1816–1822. mine dependent subjects. Subst Abuse 2010:9–20. Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018 Tang et al. | 9 Salo R, Nordahl TE, Natsuaki Y, Leamon MH, Galloway GP, Waters proton magnetic resonance spectroscopy study. Drug Alcohol C, Moore CD, Buonocore MH (2007) Attentional control and Depend 88:28–35. brain metabolite levels in methamphetamine abusers. Biol Tabibnia G, Monterosso JR, Baicy K, Aron AR, Poldrack RA, Psychiatry 61:1272–1280. Chakrapani S, Lee B, London ED (2011) Different forms of Salo R, Buonocore MH, Leamon M, Natsuaki Y, Waters C, Moore self-control share a neurocognitive substrate. J Neurosci CD, Galloway GP, Nordahl TE (2011a) Extended findings of 31:4805–4810. brain metabolite normalization in MA-dependent subjects Théberge J, Menon RS, Williamson PC, Drost DJ (2005) across sustained abstinence: a proton MRS study. Drug Implementation issues of multivoxel STEAM-localized 1H Alcohol Depend 113:133–138. spectroscopy. Magn Reson Med 53:713–718. Salo R, Nordahl TE, Buonocore MH, Natsuaki YT, Moore CD, Thompson PM, Hayashi K, Simon SL, Geaga JA, Hong MS, Sui Y, Waters C, Leamon MH (2011b) Spatial inhibition and the Lee JY, Toga AW, Ling W, London ED (2004) Structural abnor - visual cortex: a magnetic resonance spectroscopy imaging malities in the brains of human subjects who use metham- study. Neuropsychologia 49:830–838. phetamine. J Neurosci 24:6028–6036. Sekiguchi M, Okamoto K, Sakai Y (1989) Low-concentration United Nations Office on Drugs and Crime (NODC) (2017) World N-acetylaspartylglutamate suppresses the climbing fiber Drug Report 2017, United Nations Publication, Vienna. response of purkinje cells in guinea pig cerebellar slices and Available at https://www.unodc.org/wdr2017/index.html. the responses to excitatory amino acids of Xenopus laevis Urenjak J, Williams SR, Gadian DG, Noble M (1992) Specific oocytes injected with cerebellar mRNA. Brain Res 482:87–96. expression of N-acetylaspartate in neurons, oligodendrocyte- Simmons ML, Frondoza CG, Coyle JT (1991) Immunocytochemical type-2 astrocyte progenitors, and immature oligodendrocytes localization of N-acetyl-aspartate with monoclonal antibod- in vitro. J Neurochem 59:55–61. ies. Neuroscience 45:37–45. Volkow ND, Chang L, Wang GJ, Fowler JS, Franceschi D, Sedler Starkstein SE, Robinson RG, Honig MA, Parikh RM, Joselyn J, Price MJ, Gatley SJ, Hitzemann R, Ding YS, Wong C, Logan J (2001) TR (1989) Mood changes after right-hemisphere lesions. Br J Higher cortical and lower subcortical metabolism in detoxi- Psychiatry 155:79–85. fied methamphetamine abusers. Am J Psychiatry 158:383–389. Sung YH, Cho SC, Hwang J, Kim SJ, Kim H, Bae S, Kim N, Chang KH, Zorick T, Sugar C, Hellemann G, Shoptaw S, London ED (2011) Daniels M, Renshaw PF, Lyoo IK (2007) Relationship between Poor response to sertraline in methamphetamine depend- N-acetyl-aspartate in gray and white matter of abstinent ence is associated with sustained craving for methampheta- methamphetamine abusers and their history of drug abuse: a mine. Drug Alcohol Depend 118:500–503. Downloaded from https://academic.oup.com/ijnp/advance-article-abstract/doi/10.1093/ijnp/pyy042/5000087 by guest on 13 July 2018
International Journal of Neuropsychopharmacology – Oxford University Press
Published: May 21, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera