Methamphetamine Activates Trace Amine Associated Receptor 1 to Regulate Astrocyte Excitatory Amino Acid Transporter-2 via Differential CREB Phosphorylation During HIV-Associated Neurocognitive Disorders

Methamphetamine (METH) use, referred to as methamphetamine use disorder (MUD), results in neurocognitive decline, a characteristic shared with HIV-associated neurocognitive disorders (HAND). MUD exacerbates HAND partly through glutamate dysregulation. Astrocyte excitatory amino acid transporter (EAAT)-2 is responsible for >90% of glutamate uptake from the synaptic environment and is significantly decreased with METH and HIV-1. Our previous work demonstrated astrocyte trace amine associated receptor (TAAR) 1 to be involved in EAAT-2 regulation. Astrocyte EAAT-2 is regulated at the transcriptional level by cAMP responsive element binding (CREB) protein and NF-κB, transcription factors activated by cAMP, calcium and IL-1β. Second messengers, cAMP and calcium, are triggered by TAAR1 activation, which is upregulated by IL-1β METH-mediated increases in these second messengers and signal transduction pathways have not been shown to directly decrease astrocyte EAAT-2. We propose CREB activation serves as a master regulator of EAAT-2 transcription, downstream of METH-induced TAAR1 activation. To investigate the temporal order of events culminating in CREB activation, genetically encoded calcium indicators, GCaMP6s, were used to visualize METH-induced calcium signaling in primary human astrocytes. RNA interference and pharmacological inhibitors targeting or blocking cAMP-dependent protein kinase A and calcium/calmodulin kinase II confirmed METH-induced regulation of EAAT-2 and resultant glutamate clearance. Furthermore, we investigated METH-mediated CREB phosphorylation at both serine 133 and 142, the co-activator and co-repressor forms, respectively. Overall, this work revealed METH-induced differential CREB phosphorylation is a critical regulator for EAAT-2 function and may thus serve as a mechanistic target for the attenuation of METH-induced excitotoxicity in the context of HAND.

We have previously shown METH-induced EAAT-2 downregulation may be partially due to the activation of astrocyte TAAR1 (7). Of the six functional human TAAR genes, TAAR1 is reported in multiple organs and within several CNS regions, including the prefrontal cortex (20). TAAR1 functions in the neuromodulation of biogenic amines and regulates subcortical monoaminergic transmission and NMDA receptor-mediated glutamate transmission. Thus, TAAR1 plays a critical role in cognitive processing (21)(22)(23). TAAR1 activation leads to increased secondary messenger, cAMP, and activation of protein kinase A and C (PKA/C) (24)(25)(26). Although the direct mechanism of PKC activation remains vague, PKC becomes activated by increased levels of intracellular calcium [(Ca +2 ) i ], which is documented to occur following METH exposure (27,28).
Glutamate dysregulation is a significant contributing factor to the neurotoxicity associated with METH abuse and HAND. Our previous data demonstrating METH increases intracellular cAMP via TAAR1 activation in astrocytes and subsequently regulates EAAT-2 and glutamate clearance levels, sets a strong basis for further investigations of METHinduced TAAR1 signaling in the transcriptional regulation of astrocyte EAAT-2. Therefore, in this study, we investigated the dichotomous outcomes of TAAR1-mediated activation of cAMP/PKA/pCREB Ser133 and [Ca +2 ] i /CaMKII/pCREB Ser142 . Our data revealed that differential CREB phosphorylation results in EAAT-2 regulation, indicating that tipping the balance of METH-induced signaling to favor cAMP/PKA/pCREB Ser133 serves as a promising countermeasure in reversing METH-induced EAAT-2 downregulation.

Isolation, Cultivation, and Activation of Primary Human Astrocytes
Human astrocytes were isolated from first and early second trimester electively aborted specimens as previously described (38,39). Tissues were procured in full compliance with the ethical guidelines of the National Institutes of Health, Universities of Washington and North Texas Health Science Center. Cell suspensions were centrifuged, washed, suspended in media, and plated at a density of 20 × 10 6 cells/150 cm 2 . Adherent astrocytes were treated with trypsin and cultured to enhance the purity of replicating astroglial cells. These astrocyte preparations were >99% pure as measured by immunocytochemistry staining for GFAP.  (49)(50)(51). Normal peripheral blood mononuclear cells (Nebraska Medicine, Apheresis Center, Omaha NE) were isolated and infected with HIV-1 JRFL as previously described (52). Culture supernatants were clarified by centrifugation at 10,000 g for 20 min and stored at −80 • C. The concentration of HIV-1 JRFL was determined by HIV-1 p24 ELISA (Cat#: XB-1000, Xpress Bio International). Viral stocks were diluted in ASM prior to primary human astrocytes treatment. Astrocytes are not actively infected with HIV-1. Untreated astrocytes were maintained in parallel as control.

RNA Extraction and Gene Expression Analyses
Astrocyte RNA was isolated 8 h post-treatment, as previously described (53), and mRNA levels were assayed by real-time polymerase chain reaction (PCR). TaqMan 5 ′ nuclease realtime PCR was performed using StepOnePlus detection system (Thermo Fisher Scientific, Carlsbad, CA). Commercially available TaqMan R Gene Expression Assays were used to measure EAAT-2 (Cat# Hs00188189_m1), TAAR1 (Cat# Hs00373229_s1), PKA (Cat# Hs00427274_m1), CaMKII (Cat# Hs00947041_m1), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Thermo Fisher Scientific; Cat# 4310884E) mRNA levels. GAPDH, a ubiquitously expressed housekeeping gene, was used as an internal normalizing control. The 25 µl reactions were carried out at 48 • C for 30 min, 95 • C for 10 min, followed by 40 cycles of 95 • C for 15 s and 60 • C for 1 min in 96-well-optical, real-time PCR plates. Transcripts were quantified by the comparative CT method and represented as fold-changes to respective controls.

Glutamate Clearance Assay
Primary human astrocytes were plated in 48-well-tissue culture plates at a density of 0.15 × 10 6 cells/well and allowed to recover for 24 h prior to treatment. Following 24 h of treatment, glutamate (400 µM), dissolved in phenol-free astrocyte medium was added into each well, and glutamate clearance was assayed at 10 h post-glutamate addition. The assay was performed and analyzed according to manufacturer's guidelines (Amplex Red Glutamic Acid/Glutamate Oxidase Assay Kit, Cat# A12221, Thermo Fisher Scientific, Carlsbad, CA). Following collection of glutamate supernatants, a colorimetric assay for measurement of metabolic activity was performed using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Cat# M2128, Sigma-Aldrich) (54). Briefly, five percent MTT reagent was added to astrocytes and incubated for 20-45 min at 37 • C. The MTT solution was removed, and crystals were dissolved in DMSO for 15 min with gentle agitation. Absorbance was assayed at 490 nm in a Spectromax M5 microplate reader (Molecular Devices, Sunnyvale, CA).

Kinase Assays
Intracellular PKA and CaMKII levels in astrocytes were measured using a commercially available homogenous, highthroughput screening method for measuring kinase activity, Kinase-Glo R Max Luminescent Assay, (Cat# V6073, Promega) and commercially available PKA Kinase Enzyme System (Cat# V4246, Promega) and CaMKIIα Kinase Enzyme System (Cat# V4018, Promega). cAMP-dependent PKA catalytic subunit α is a 40 kDa bovine recombinant enzyme expressed and purified from E. coli (Accession number NM_174584.2). PKA purity was 90% as defined from Promega quality control assays. Human recombinant CaMKIIα was expressed by baculovirus in sf9 insect cells using N-terminal GST tag (Accession number NM_171825, a Ser/Thr protein kinase and a member of the Ca +2 /calmodulin dependent protein kinase family). The specific activity of CaMKIIα was determined to be 960 nmol/min/mg as per assay protocol. Briefly, adherent monolayers of astrocytes cultured in 96-well-plates (50,000 cells/well) were treated (+/-METH, 500 µM) in PKA or CaMKIIα reaction buffer for 5, 15, or 30 min. Cells were directly activated in the tissue culture plate. Lysates were diluted to a final cell concentration of approximately 1,000 cells/µL in lysis buffer and transferred to a white opaque flat bottom 384-well-assay plate at approximately 3,000 cells/reaction. Intracellular PKA and CaMKIIα levels were assayed using GloMax 384 Microplate Luminometer with dual injectors (Promega).

Transfection of Astrocytes
Cultured human astrocytes were transfected with On-Target plus R small interfering RNA (siRNA, Dharmacon, Lafayette, CO) pools specific to PKAαβ (siPKAα Cat# L-004649 & siPKAβ, Cat# L−004650), CaMKII (siCaMKII, Cat# L-004942), non-targeting control siRNA pools (siCON, Cat# D-001810), and without siRNA (MOCK) or with pGP-CMV-GCaMP6s, deposited by Douglas Kim (RRID:Addgene_40753) using the Amaxa TM P3 primary cell 96-well-Nucleofector kit and shuttle attachment (Lonza, Walkersville, MD) according to the manufacturer's instructions. Briefly, 1.6 × 10 6 astrocytes were suspended in 20 µl nucleofector solution containing siCON, siPKAαβ, siCaMKII, (100 nM) or GCaMP6s (0.5 µg/1.6 × 10 6 cells) and transfected using shuttle protocol CL133. GCaMP6s transfection efficiency averaged approximately 80%. Multiple chambers from the same biological donors were assayed in a minimum of triplicate determinations. Areas were randomly chosen for confocal imaging from each chamber. Baseline green fluorescence suggested successful transfection. Background fluorescence was subtracted from the standardization well for each individual biological donor, to prevent saturation of fluorescence. Transfected cells were supplemented with astrocyte media and incubated for 30 min at 37 • C prior to plating. Cells were allowed to recover for 48 h prior to experimental use.

Confocal Analysis
MOCK-and GCaMP6s-transfected astrocytes were cultured on tissue culture treated µ-slides with a channel height of 0.4 mm (Cat# 86060, Ibidi, Madison, WI) at a density of 1 × 10 5 cells/chamber in astrocyte media. Prior to live cell imaging, astrocytes were briefly washed with PBS and supplemented with HBSS at 37 • C. Time lapse images were obtained every 500 ms, from astrocytes treated with METH (500 µM) or ionomycin (10 µM). Micrographs were obtained on a Carl Zeiss LSM (Jena, Germany). Objective used was 20× Plan-Apochromat, 0.8NA, 0.55 mmWD. PMT photo detection was used with an excitation of 450-490 nm and emission of 593-668 nm. Histogram analysis were performed using ImageJ software; Version: 2.0.0-rc-41/1.5d (Fiji ImageJ Software, the National Institutes of Health, Bethesda, MD) (55). Histogram was generated from fluorescence units obtained at selected time points.
[Ca +2 ] i was calculated using a Fura-2 calcium calibration standard curve (Thermo Fisher Scientific, Cat# F6774). Basal [Ca +2 ] i measurements were taken following stabilization period of 5 min prior to METH administration, and peak [Ca +2 ] i were measured following METH treatment. Ratiometric data were collected from cells that were alternately illuminated with 340-and 380-nm wavelengths using xenon light source (Lumen200PRO, Prior Scientific, Rockland, MD). The emitted light was captured at 520 nm wavelength using a CCD camera (Hamamatsu camera controller C10600, Hamamatsu Photonics KK, Hamamatsu, Japan). Pixel data were binned (2x2), and images were captured every 3 s. Data were collected and analyzed using commercially available software (Slidebook 5.0, Intelligent Imaging Innovations, Denver, CO).

Statistical Analyses
Statistical analyses were performed using GraphPad Prism (Version 8.4.0, RRID:SCR_002798) with one-way analysis of variance (ANOVA) and Tukey's post-test for multiple comparisons. Linear regression and correlation analysis were performed using Prism with a two-tailed, Pearson correlation coefficient set at 95% confidence interval. P ≤ 0.05 were considered statistically significant, and data represent means ± standard error of the mean (SEM).

METH Transiently Increases Intracellular Calcium in GCaMP6s-Transfected Astrocytes
Reports suggest that METH results in increased [Ca +2 ] i in neurons (27,59). To determine METH-induced activation of [Ca +2 ] i stores, primary human astrocytes were transfected with a GCaMP6s plasmid (60). Baseline fluorescence was measured in the absence of external stimuli and plotted as 0 FLU (Figures 3A,D), demonstrating undetectable [Ca +2 ] i , i.e., no fluorescence ( Figure 3A). There was an approximate 80% transfection efficiency for GCaMP6s transfection in primary human astrocytes. Although low fluorescence was detectable following 12 s of METH treatment (Figure 3B), a robust increase in fluorescence was visualized at 100 s ( Figure 3C). FLU represents fluorescence from imaged astrocyte over time ( Figure 3D). These data support the innovative use of GCaMP6s-transfected astrocytes in visualizing METH-induced increases of intracellular calcium via live cell imaging. To further support METH-mediated increases of [Ca +2 ] i , astrocytes were treated with Fura-2-AM and stimulated with METH, and subsequent calcium levels were quantified (Figures 3E,F) (Figure 3F, * * * p < 0.001). These data confirm that increases of [Ca +2 ] i are sufficient to activate downstream signaling cascades and phosphorylation of substrates.

DISCUSSION
This study uncovers critical signaling pathways in astrocytes mediated via METH-induced TAAR1 activation. Importantly, we identify CREB as a master regulator of astrocyte EAAT-2 following METH treatment. We demonstrate METH activates canonical cAMP/PKA/pCREB Ser133 and [Ca +2 ] i /CaMKII/pCREB Ser142 signaling in primary human astrocytes, downstream of TAAR1. TAAR1 is established to trigger an accumulation of intracellular cAMP (24), regulating expression, localization and function of monoamine transporters via phosphorylation of PKA and PKC (61)(62)(63)(64). Activation of kinases such as PKA, results in phosphorylation of substrates including CREB. CREB phosphorylation traditionally involves transcriptional activation of CREB binding genes (65). Recruitment of CREB to the EAAT-2 promoter suggests increased promoter activity and EAAT-2 upregulation (29,66). This study identifies that METH phosphorylates CREB at serine 133; however, EAAT-2 transcription decreases, indicating a differential role for CREB in EAAT-2 regulation, downstream of TAAR1. We tested the hypothesis that TAAR1-mediated signaling, following METH stimulation, dually triggers the activating and dominant repressing forms of CREB, thus dictating EAAT-2 downregulation. We demonstrate astrocyte TAAR1 levels increase following IL-1β and HIV-1 treatment, suggesting a broader role for astrocyte TAAR1 during neuroinflammation. Taken together, our studies reveal a delicate balance between METH-induced activation of cAMP/PKA/pCREB Ser133 and [Ca +2 ] i /CaMKII/pCREB Ser142 signaling in the regulation of astrocyte EAAT-2, which tips the downstream balance of EAAT-2 function from glutamate uptake to excitotoxicity.
In this study we confirm that following METH treatments, intracellular cAMP increases, activating PKA, and phosphorylating CREB at serine 133, yet results in astrocyte EAAT-2 downregulation. Kim et al. showed that exogenous application of dibutyl cAMP increases EAAT-2 transcription supporting our hypothesis that selectively activating cAMP/PKA/pCREB Ser133 pathway is enough to prevent METH-mediated EAAT-2 downregulation (66). Decreases in astrocyte EAAT-2, subsequent to METH treatment, is not prevented with PKA downregulation, indicating METH responses in astrocytes are likely activating transcriptional repressors. Alternatively, the abundance and distribution of cAMP-regulated guanine nucleotide exchange factors (Epac/cAMP-GEF) warrants further studies, since activation of this exchange protein may converge synergistically with PKA, or mediate increases of [Ca +2 ] i to regulate other biological functions [reviewed in (67)]. Nonetheless, we demonstrate TAAR1 activation increases [Ca +2 ] i in primary human astrocytes in parallel to cAMP. Several lines of evidence suggest CaMKs, downstream of increased [Ca +2 ] i , activate CREB, and result in transcriptional activation and/or repression . Statistical analyses were performed using GraphPad Prism V6.0 with One-way ANOVA and Tukey's post-test for multiple comparisons. P ≤ 0.05 were considered statistically significant, and data represent ± SEM. Representative donors chosen from multiple astrocyte donors were tested; each was analyzed in a minimum of triplicate determinations (**p < 0.01, ***p < 0.001). (68). This balance is mediated via phosphorylation of serine 133 and 142 (36,37). Phosphorylation of CREB at serine 142 acts as a dominant negative regulator of pCREB Ser133induced transcriptional activation, despite significantly higher amounts of pCREB at serine 133 vs. serine 142 (37,68). Efficient binding of pCREB Ser133 to promoter elements continues; however, pCREB Ser142 prevents CBP dimerization, thus inhibiting CREB-supported transcription (68). We hypothesize that the absence of the dominant repressor, pCREB Ser142 , would permit transcriptional activation of EAAT-2. Consistent with increased EAAT-2 transcription following dibutyl cAMP application that selectively activates PKA/pCREB Ser133 , we propose that forfeiting CaMKII/pCREB Ser142 -induced transcriptional repression, while . CaMKII activity was measured in astrocytes stimulated with IL-1β (20 ng/mL, hatched bars) for 30 min (E). Primary human astrocytes were treated with IL-1β and HIV-1 (blue bars), alone and in combination. Protein lysates were collected 30 min post-treatment and immunoblotted for total CREB, pCREB Ser133 , pCREB Ser142 , and GAPDH (F,G). Dividing lines on blots represented in (F,G) represent cut sections from the same blot. Representative western blots are shown in (F,G). Densitometry analyses were performed to quantify band intensities on multiple immunoblots and represented as fold changes to control ± SEM, in [(F,G), n = 3]. Statistical analyses were performed using GraphPad Prism V6.0 with one-way ANOVA and Tukey's post-test for multiple comparisons. P ≤ 0.05 were considered statistically significant, and data represent means ± SEM. This figure depicts representative donors chosen from multiple astrocyte donors that were tested; each was analyzed in a minimum of triplicate determinations. Data are shown as cumulative fold changes. n represents individual biological replicates. Molecular weight markers are identified on each western blot (MW) (*p < 0.05, **p < 0.01, ***p < 0.001).
Recently Kumar et al. demonstrated that METH exposure decreased pCaMKII levels in several brain regions of HIV tat transgenic mice (69). These changes correlated with decreased working and spatial memory, neurotrophin levels, and decreased synaptodendritic integrity. Autophosphorylation of CaMKII at threonine 286 is required for kinase activity (70). Both HIV and METH have been shown to reduce pCaMKII in rodents and SIV in rhesus macaques (71,72). These changes have been associated with neuronal responses and have not been evaluated in astrocytes. Here we show that METH significantly reduced CaMKII activity at 5 min and increased activity by 30 min in human astrocytes. Further, METH, HIV-1, and IL-1β increased pCREB Ser142 phosphorylation, which was associated with total CaMKII levels and activity. Together, this suggests that CaMKII activity may be regulated differently in neurons and astrocytes by METH and HIV.
TAAR1, with EPPTB, may be enough to prevent METHinduced EAAT-2 downregulation; yet, METH abuse during CNS inflammation and HIV-1 poses greater threat due to their potential to increase TAAR1 levels/activity and crosstalk between PKA, CaMKII and NF-κB. For instance, the rate of NF-κB translocation into the nucleus is regulated by PKA-induced phosphorylation of NF-κB/Rel complexes, downstream of cAMP (73). Interestingly, NF-κB activation of target genes is optimized by interaction of the RelA subunit with CREB co-activators, CBP and p300 (74,75). Furthermore, the region of pCREB Ser133 that interacts with CBP also interacts with RelA, thereby inhibiting NF-κB activity (76). This serves as compelling evidence for CREB and NF-κB mechanistic crosstalk, ultimately influencing how they regulate transcription of target genes in astrocytes. IL-1β and HIV-1 initiate similar signaling mechanisms in astrocytes as METH, independent of TAAR1 activation. The additive effects of IL-1β and HIV-1 on pCREB Ser133 vs. inhibitory effects on pCREB Ser142 need to be further investigated and may be FIGURE 8 | METH-induced TAAR1 signaling regulates astrocyte EAAT-2. METH likely activates different signaling pathways in astrocytes that are exacerbated by HIV relevant inflammatory cytokine, IL-1β and/or HIV-1. Intracellular signal transduction pathways lead to CREB phosphorylation, thereby differentially regulating EAAT-2. We have previously demonstrated METH-induced activation of astrocyte TAAR1 increases intracellular cAMP and regulates EAAT-2. In this manuscript, TAAR1 upregulation and increased activity negatively correlate to astrocyte EAAT-2 and impair glutamate clearance capabilities as demonstrated by (Figure 1). Investigation of signal transduction pathways revealed METH-induced TAAR1 activation leads to cAMP/PKA/pCREB Ser133 ( Figure 2) and [Ca +2 ] i /CaMKII/pCREB Ser142 (Figures 3,  4). As METH activates TAAR1, antagonism with TAAR1 selective antagonist, EPPTB, prevents METH-induced increases of intracellular cAMP and [Ca +2 ] i , blocking METH-induced phosphorylation of CREB at both serine 133 and 142 ( Figure 5). Additionally, exogenous treatment of IL-1β and HIV-1 dually activate cAMP/PKA/pCREB Ser133 and [Ca +2 ] i /CaMKII/pCREB Ser142 suggesting similar mechanisms mediating EAAT-2 downregulation ( Figure 6). Extrinsic regulation of signaling factors including PKA and CaMKII not only reduce activation of subsequent signaling but also regulate METH-mediated decreases in astrocyte EAAT-2 ( Figure 7). Together, our data suggest that pCREB Ser142 acts as a dominant repressor of CREB transcriptional activation. Therefore, therapeutically targeting and inhibiting the [Ca +2 ] i /CaMKII/pCREB Ser142 signal transduction pathways have broader implications in the context of METH abuse and neuroinflammation. a promising mechanistic intervention to prevent glutamate excitotoxicity during neuroinflammatory disorders. Targeting downstream of CaMKII, to prevent pCREB Ser142, has larger implications for all genes with CREB promoter elements in astrocytes including inflammatory mediators, oxidative stress genes and growth factors (77)(78)(79)(80)(81).
Non-functional mutations in mouse TAAR1 affect METH intake, hypothermia and conditioned taste aversion (82,83). Humans possess many TAAR1 variants with sensitivity to ligands (84)(85)(86)(87). However, the effects of TAAR1 variants on signaling cascades and risk for MUD, HAND, and neuropsychological disease will need to be evaluated. Additionally, there is an apparent dichotomy for TAAR1-associated regulation in neurons vs. astrocytes. TAAR1 function appears to be critical for proper neuronal function (20)(21)(22)88), while increased activity in astrocytes may be detrimental to brain health (27,41,89). Targeting TAAR1 in the CNS in a cell specific manner may be difficult; however, these findings open the door for personalized medical interventions for these disorders.
In this study we investigated the duality of METH-induced signaling pathways leading to EAAT-2 transcriptional repression, thereby exacerbating HIV-1-induced decreases in EAAT-2. Figure 8 illustrates signal transduction pathways mediated via TAAR1, METH, HIV-1, and IL-1β in astrocyte EAAT-2 downregulation. We propose that triggering the preferential activation of cAMP/PKA/pCREB Ser133 while inhibiting [Ca +2 ] i /CaMKII/pCREB Ser142 signaling would alleviate excitotoxic insult via upregulation of astrocyte EAAT-2 transcription. Additionally, counteracting IL-1β and HIV-1 induced upregulation of TAAR1 would reduce METH effects in astrocytes. Transcriptional regulation of astrocyte TAAR1 via NF-κB warrants further investigation. In addition, mutation of the four NF-κB sites within the EAAT-2 promoter will elucidate on the correlation identified between TAAR1 and EAAT-2 levels and function following IL-1β treatment. Altogether, our studies shed light on tripartite signaling of PKA, CaMKII, and NF-κB involved in TAAR1 and EAAT-2 regulation, METH abuse, and HAND.
Taken together, we have identified crucial mechanistic pathways involved in METH-induced astrocyte neurotoxicity in the context of HAND. Our data provide strong evidence to support the notion that manipulation of signal transduction pathways to favor cAMP/PKA/pCREB Ser133 and to abolish [Ca +2 ] i /CaMKII/pCREB Ser142 is a promising strategy for restoring astrocyte EAAT-2 function in the context of METH abuse and HAND.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by North Texas Regional IRB. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

AUTHOR CONTRIBUTIONS
AG, IC, and KB conceived, designed and coordinated the study, and wrote the manuscript. All authors reviewed the results and approved the final version of the manuscript.

FUNDING
This work was supported by R01DA039789 from NIDA to AG and then Dr. Michael Mathis, and F31DA037832 from NIDA to IC. We appreciate the assistance of the Laboratory of Developmental Biology for providing us with human brain tissues, which is supported by NIH Award Number 5R24HD0008836 from the Eunice Kennedy Shriver National Institute of Child Health & Human Development.

ACKNOWLEDGMENTS
We acknowledge the National Institute of Drug Abuse Drug Supply Program for the methamphetamine supplied in these studies. Ms. Lenore Price and Ms. Satomi Stacy assisted with editing of the manuscript. Ms. Lin Tang provided consistent primary human astrocyte cultures. Dr. Thomas Cunningham and his lab group graciously assisted in the acquisition of calcium imaging and quantification of intracellular calcium concentrations. Ms. I-Fen Chang provided technical assistance for all confocal imaging.