HIV-Induced Hyperactivity of Striatal Neurons Is Associated with Dysfunction of Voltage-Gated Calcium and Potassium Channels at Middle Age

Despite combination antiretroviral therapy, HIV-associated neurocognitive disorders (HAND) occur in ~50% of people living with HIV (PLWH), which are associated with dysfunction of the corticostriatal pathway. The mechanism by which HIV alters the neuronal activity in the striatum is unknown. The goal of this study is to reveal the dysfunction of striatal neurons in the context of neuroHIV during aging. Using patch-clamping electrophysiology, we evaluated the functional activity of medium spiny neurons (MSNs), including firing, Ca2+ spikes mediated by voltage-gated Ca2+ channels (VGCCs), and K+ channel-mediated membrane excitability, in brain slices containing the dorsal striatum (a.k.a. the caudate-putamen) from 12-month-old (12mo) HIV-1 transgenic (HIV-1 Tg) rats. We also assessed the protein expression of voltage-gated Cav1.2/Cav1.3 L-type Ca2+ channels (L-channels), NMDA receptors (NMDAR, NR2B subunit), and GABAA receptors (GABAARs, β2,3 subunit) in the striatum. We found that MSNs had significantly increased firing in 12mo HIV-1 Tg rats compared to age-matched non-Tg control rats. Unexpectedly, Ca2+ spikes were significantly reduced, while Kv channel activity was increased, in MSNs of HIV-1 Tg rats compared to non-Tg ones. The reduced Ca2+ spikes were associated with an abnormally increased expression of a shorter, less functional Cav1.2 L-channel form, while there was no significant change in the expression of NR2Bs or GABAARs. Collectively, the present study initially reveals neuroHIV-induced dysfunction of striatal MSNs in 12mo-old (middle) rats, which is uncoupled from VGCC upregulation and reduced Kv activity (that we previously identified in younger HIV-1 Tg rats). Notably, such striatal dysfunction is also associated with HIV-induced hyperactivity/neurotoxicity of glutamatergic pyramidal neurons in the medial prefrontal cortex (mPFC) that send excitatory input to the striatum (demonstrated in our previous studies). Whether such MSN dysfunction is mediated by alterations in the functional activity instead of the expression of NR2b/GABAAR (or other subtypes) requires further investigation.


Introduction
The success of combination antiretroviral therapy (cART) has transformed HIV/AIDS into a manageable chronic disease with HIV-Associated Neurocognitive Disorders (HAND) [1,2]. HAND are found in up to 50% of people living with HIV/AIDS (PLWH), and are expected to become more prevalent as the PLWH population ages [3,4]. However, the cause of HAND is not fully understood. A better understanding of the underlying mechanism(s) by which HIV induces neurological and neuropsychiatric dysfunction is desperately needed for developing effective treatment for HAND. with food and water available ad libitum at the Rush University Comparative Research Center. Rats were housed four to a cage when they weighed <200 g and two to a cage when weighed ≥200 g. Animal care and use procedures were conducted in accordance with NIH, USDA and institutional guidelines, and approved by the Institutional Animal Care and Use Committee at Rush University Medical Center. All studies were performed on 12mo-old rats (~250-300 g).
Tissue Preparation. Rats were anesthetized and euthanized for experiments at 12mo of age. Prior to transcardial perfusion, deep sedation was ensured by testing responsiveness to hind limb toe pinch and eye pokes. If any response was observed, additional dose was administered. Under deep sedation, the chest of a rat was cut open, and the blood vessels leading to the lower part of the body were blocked. Rats were transcardially perfused with ice-cold cutting solution (in mM: 248 sucrose, 2.9 KCl, 2 MgSO 4 , 1.25 NaH 2 PO 4 , 26 NaHCO 3 , 0.1 CaCl 2 , 10 glucose, 3 kynurenic acid, 1 ascorbic acid; pH = 7.4-7.5, 60 mL) to clear the brain tissue and to minimize the damage of neurons in the brain.
Statistical Analysis. All data were analyzed using t-tests or two-way repeated measures (rm) ANOVA followed by Sidak's post-hoc tests. Two-way rmANOVA was used to compare responses of striatal neurons at different current steps (e.g., action potential spike frequency, and V m alterations mediated by K v channels, n reflects neuron numbers). Action potential properties, calcium spike properties (n reflects neuron numbers), and protein levels (n reflects the number of rats) were compared using t-tests. The number of animals and neurons per group assessed in this study was determined from power analysis in combination with empirical data from our previously-published studies. Data were excluded from analysis if found to be an outlier (≥2× the standard deviation from the mean) or if experimental criteria were not met (described above in the electrophysiology section). Statistical significance was set at p ≤ 0.05.

Firing of Striatal MSNs Is Abnormally Increased in 12mo HIV-1 Tg Rats
We determined the excitability of MSNs in 12mo HIV-1 Tg and non-Tg rats. This age of rats is equivalent to 30-year-old humans [26]. We assessed the characteristics of action potentials and firing frequency. Age-matched non-Tg rats were used as the control for each individual assessment in the present study. The majority of striatal MSNs had a firing/spike response evoked by rheobase ≤300 picoamperes (pA, 76.5%, 13/17 neurons in 10 non-Tg rats and 74.1%, 20/27 neurons in 14 HIV-1 Tg rats). In these MSNs ( Figure 1A,B), we found a significant increase in the spike number in HIV-1 Tg rats compared to non-Tg rats (two-way rmANOVA: HIV: F (1,31) = 5.092, p = 0.0312; current, F (15,465) = 115, p < 0.0001; interaction, F (15,465) = 3.019, p = 0.0001). The passive and active membrane properties were similar in striatal neurons regardless of genotype, but there was a trend of RMP depolarization in the context of neuroHIV (p = 0.074, Table 1A). A small portion of MSNs (~25%) from 12mo non-Tg rats or HIV-1 Tg rats (which might have been affected by aging or injured after chronic exposure to HIV in vivo), was much less responsive to V m depolarization, and therefore, were excluded from data analysis.  showing that striatal MSNs from HIV-1 Tg rats are hyperexcitable compared to those from non-Tg rats. These data are from the majority of neurons, which displayed a spike response to rheobase ≤300 pA, from non-Tg and HIV-1 Tg rats, respectively. Significance is denoted as *: p ≤ 0.05, **: p ≤ 0.01, and ***: p ≤ 0.001.  showing that striatal MSNs from HIV-1 Tg rats are hyperexcitable compared to those from non-Tg rats. These data are from the majority of neurons, which displayed a spike response to rheobase ≤300 pA, from non-Tg and HIV-1 Tg rats, respectively. Significance is denoted as *: p ≤ 0.05, **: p ≤ 0.01, and ***: p ≤ 0.001.

Activity of VGCCs Is Reduced in Striatal MSNs of 12mo HIV-1 Tg Rats
To determine if and how VGCCs contribute to the increased striatal MSN firing in HIV-1 Tg rats, we assessed VGCC activity under the conditions in which Na + channel-, K + channel-, NMDAR/AMPAR-, and GABA A R-mediated responses, were blocked. Unexpectedly, we found that MSNs in HIV-1 Tg rats (n = 8 in 4 rats) had a significantly-reduced Ca 2+ spike duration ( Figure 2B; t (18) = 2.918, p = 0.0092) and area ( Figure 2C; t (18) = 2.666, p = 0.0158; reflecting decreased Ca 2+ influx) compared to MSNs in non-Tg rats (n = 12 in 10 rats, Figure 2A-C). Likewise, striatal MSNs in HIV-1 Tg rats required a significantly greater rheobase to evoke a Ca 2+ spike compared to those in non-Tg rats (t (18) = 2.303, p = 0.0335; Table 1B), further indicating that VGCCs in striatal MSNs of HIV-1 Tg rats were less active than those in non-Tg rats. Other Ca 2+ spike properties were not significantly different in Tg and non-Tg rats, although there was a trend of reduced half-amplitude duration (t (13) = 1.872, p = 0.0838) in striatal MSNs from HIV-1 Tg rats compared to non-Tg rats (Table 1B). Such a trend could suggest alterations in the gating mechanism (e.g., the opening, closing, and/or inactivation) of VGCCs that might be responsible for the reduced Ca 2+ influx. 0.0335; Table 1B), further indicating that VGCCs in striatal MSNs of HIV-1 Tg rats were less active than those in non-Tg rats. Other Ca 2+ spike properties were not significantly different in Tg and non-Tg rats, although there was a trend of reduced half-amplitude duration (t(13) = 1.872, p = 0.0838) in striatal MSNs from HIV-1 Tg rats compared to non-Tg rats (Table 1B). Such a trend could suggest alterations in the gating mechanism (e.g., the opening, closing, and/or inactivation) of VGCCs that might be responsible for the reduced Ca 2+ influx. There are different subtypes of VGCCs [28], including the L-channel. In the brain, Lchannels consist of a high voltage-activated (HVA-Ca v 1.2 α1c) and a low voltage-activated (LVA-Ca v 1.3 α1d) form [29]. We demonstrate that L-channel function and expression are abnormally increased in mPFC of HIV-1 Tg rats; associated with mPFC neuronal hyperactivity [19,20,22]. Here, we assessed L-channel protein levels in the caudate-putamen ( Figure 2). We found that the levels of full-length (~250 kDa) HVA-Ca v 1.2 and LVA-Ca v 1.3 were not statistically different in the caudate-putamen of Tg (n = 8) and non-Tg rats (n = 7, Figure 2D-F). However, the expression of a shorter, and likely less functional [30], 150 kDa Ca v 1.2 form was significantly higher in the caudate-putamen of HIV-1 Tg rats (n = 7) compared to non-Tg rats (n = 7; 150 kDa: t (12) = 2.022, p = 0.0479; 150 kDa/250 kDa: t (12) = 3.533, p = 0.0305; Figure 2D,G,H). This dysregulation also induced a significant increase in the ratio of 150 kDa vs. 250 kDa Ca v 1.2 form ( Figure 2H). These data suggest that the reduced voltage-sensitive Ca 2+ influx among striatal MSNs in HIV-1 Tg rats results from an abnormal increase of less functional L-channels due to increased expression of the shorter, 150 kDa Ca v 1.2 form.

Expression of NMDAR and GABA A R Is Unaltered in the Striatum of 12mo Rats by neuroHIV
Our data show that Ca 2+ influx was reduced in striatal MSNs of HIV-1 Tg rats but under blockade of NMDARs and GABA A Rs. Thus, it is possible that alterations in excitatory or inhibitory synaptic activity could facilitate VGCC activity, leading to increased firing. To examine the potential influence of NMDARs and GABA A Rs, we assessed the protein levels of NMDAR NR2B subunits and GABA A R β 2,3 subunits in the caudate-putamen (Figure 3). We chose to assess the NR2B subunit because it is functionally involved in excitotoxicity in neurons [31,32]; while we chose to assess the β 2,3 subunit of GABA A R because the β 2 and β 3 subunits, but not β 1 , are detected in the rat striatum [33]. However, we did not find any significant difference in NR2B (n = 7 HIV-1 Tg vs. n = 8 non-Tg, Figure 3B), or GABA A Rβ 2,3 (n = 8 HIV-1 Tg vs. n = 7 non-Tg, Figure 3C) protein in the caudate-putamen between HIV-1 Tg and non-Tg rats. These data suggest that striatal hyperexcitability was not mediated by increased NR2B or decreased GABA A R β 2,3 . But it remains to be determined if (i) NMDAR activity is increased; (ii) GABA A R activity is reduced, and (iii) changes in the expression of other NMDAR subunits or AMPARs exist, in 12mo HIV-1 Tg rats.  For GABAAR staining and respective actin staining, membranes were first probed for GABAAR protein, then stripped and re-probed for actin, For the representative NMDAR bands, the samples were run on the same gel, but were not side-by-side; i.e., the bands shown were cropped from two different areas of the same membrane. Supplementary Figure  S2 shows the full blots for the representative samples. (B,C) Bar graphs showing relative levels of NMDAR2B (B) and GABAARβ2,3 (C) normalized to β-actin protein.

Kv Channel Activity Is Significantly Increased, While Activity of Kir and K2P Channels Is Not Significantly Altered, in Striatal MSNs in the Context of neuroHIV
Excitability (Vm) of striatal MSNs is also mediated by various subtypes of K + channels. Thus, we first assessed the activity of Kv channels that conduct K + efflux. We found that Kv activity (activated by Vm depolarization) was significantly increased, indicated by enhanced downward Vm traces ( Figure 4A (Figure 4A,B). This increased Kv activity could diminish VGCC activity, rather than increase, MSN firing. Then we also assessed the ac- For GABA A R staining and respective actin staining, membranes were first probed for GABA A R protein, then stripped and re-probed for actin, For the representative NMDAR bands, the samples were run on the same gel, but were not side-by-side; i.e., the bands shown were cropped from two different areas of the same membrane. Supplementary Figure S2 shows the full blots for the representative samples. (B,C) Bar graphs showing relative levels of NMDAR2B (B) and GABA A Rβ 2,3 (C) normalized to β-actin protein.

K v Channel Activity Is Significantly Increased, While Activity of K ir and K 2P Channels Is Not Significantly Altered, in Striatal MSNs in the Context of neuroHIV
Excitability (V m ) of striatal MSNs is also mediated by various subtypes of K + channels. Thus, we first assessed the activity of K v channels that conduct K + efflux. We found that K v  (Figure 4A,B). This increased K v activity could diminish VGCC activity, rather than increase, MSN firing. Then we also assessed the activity of inwardly rectifying K + (K ir ) channels, which were activated by V m hyperpolarization to conduct K + influx. There was no significant difference in K ir activity between non-Tg (n = 8 MSNs/6 rats) and HIV-1 Tg rats (n = 10 MSNs/8 rats. HIV: F (1,15) = 0.5037, p = 0.4881; current: F (15,240) = 233.8, p < 0.001; interaction: F (15,240) = 1.295, p = 0.2056, Figure 4C,D). Finally, we assessed the K 2P channel activity that mediates the resting membrane potential (RMP). Given that RMP was maintained and regulated predominantly by outflowing K + currents through activating K 2P channels in neurons at the resting status, we also determined if HIV altered the RMP of striatal MSNs by disturbing the function of K 2P channels. We found that there was a trend in which RMP of striatal MSNs was slightly, but not significantly depolarized, in the context of neuroHIV at 12mo of age (n = 12 MSNs/8 rats) compared to age-matched non-Tg control rats (n = 9 MSNs/7 rats; non-Tg vs. HIV-1 Tg: −78.5 ± 1.3 mV vs. −75.1 ± 1.2 mV; t-test: t 19 = 0.266, p = 0.074; Table 1A).   Collectively, neuroHIV-induced changes in voltage-gated membrane ion channels, as well as unaffected protein levels of ionotropic NMDARs and GABA A Rs among striatal MSNs in 12mo rats are summarized in Table 2.

Discussion
In the present study, we initially examined how neuroHIV disrupts striatal MSN activity to define potential mechanisms that could be targeted for the treatment of HAND. We demonstrate that neuroHIV abnormally increases MSN excitability in 12mo-old HIV-1 Tg rats compared to age-matched non-Tg rats. Surprisingly, this increase was associated with reduced VGCC function and enhanced VGKC activity, rather than the opposite conditions expected. Concurrently, we also found that the protein expression of a shorter, and less functional form of Ca v 1.2 L-channels was significantly increased in the dorsal striatum of HIV-1 Tg rats, while the protein expression of NMDAR (NR2B subunit) and GABA A R (β2,3 subunits) was unchanged. These novel findings indicate that the hyperexcitability of striatal MSNs in the context of neuroHIV cannot be attributed to VGCC upregulation or reduced VGKC activity (found in younger HIV-1 Tg rats); but suggest that it could be mediated by increased activity of NMDARs (e.g., opening probability/frequency/time without overexpression), in response to enhanced glutamate input from the hyperactive mPFC [19,20,22,23], reduced glutamate uptake by astrocytes, or decreased GABA A R activity in the HIV-altered striatum. Future study is needed to define this potential mechanism during aging.
One major finding of the current study is that MSNs have significantly-increased firing in HIV-1 Tg rats. To our knowledge, this is the first study in the field that reveals hyperexcitability of striatal MSNs from 12mo-old rats in the context of neuroHIV. The characteristics of action potentials (Table 1A) that provide a glimpse into the function of membrane ion channels that are critically involved in generating firing suggest that the mechanism underlying this MSN hyperactivity does not include increased function of voltage-gated Na + /Ca 2+ channels, or reduced K v channel activity. Moreover, the unchanged resting membrane potential, input resistance, afterhyperpolarization, and inward rectification, which were mediated by K 2P channels, Ca 2+ -activated K + channels, and K ir channels, respectively (Table 1A), also exclude their role in causing such MSN hyperactivity in 12mo-old rats in the context of neuroHIV.
In the current study, we also found that VGCC function was uncoupled from abnormallyincreased firing in the context of neuroHIV. In fact, VGCC activity was reduced (instead of increased) in striatal MSNs of 12mo-old HIV-1 Tg rats, which was the opposite of an expectation if VGCC overactivation contributed to this MSN hyperexcitability. This HIV-induced reduction of VGCC function in striatal MSNs differs remarkably from what we previously demonstrated in mPFC pyramidal neurons [20], which could be a compensatory response to dampen the impact of excessive excitatory inputs from the hyperactive mPFC, where the functional activity of glutamatergic pyramidal neurons is significantly increased by HIV [19,20,22], or a consequence of overactivation-caused VGCC inactivation in the context of neuroHIV [22,34,35]. Thus, further study is also needed to determine if this seemingly compensatory reduction, or over-activation-induced inactivation in Ca 2+ influx mediated by dysfunctional Ca v 1.2 L-channels (see below), is related to cognitive deficits induced by HIV, and if so, whether such Ca 2+ dysregulation can be intervened pharmacologically.
VGCCs are known to undergo activity-dependent regulation [28,36], which is crucial because disturbed neuronal Ca 2+ homeostasis can cause dysregulation, damage, and even death of neurons, by disturbing excitability, signal transduction, and gene expression pathways [6,37,38]. Different forms of activity-dependent regulation have been identified, which operate over a range of time frames [28,36]. Such regulatory processes include proteolytic cleavage [30,36,39] and channel internalization [36,40]. Here, we found the significantly-increased expression of a shorter (150 kDa) form of the Ca v 1.2 L-channel in the caudate-putamen of HIV-1 Tg rats. Expression of this shorter 150 kDa Ca v 1.2 form could result from cleavage of the full-length L-channel. In fact, proteolytic processing of full-length Ca v 1.2 to a 150 kDa cleavage product is reported to produce partially (i.e., less) functional L-channels, with distinct biophysical characteristics that alter the properties of full-length L-channels [30]. This is in line with our findings of reduced Ca 2+ influx via VGCCs. It is also possible that the shorter 150 kDa Ca v 1.2 form found here results from increased expression of an alternatively-spliced Ca v 1.2 channel. Further studies are needed to determine whether this 150 kDa Ca v 1.2 form is a product of proteolytic processing or alternative splicing.
Other forms of activity-dependent negative regulation, which could also contribute to the reduced VGCC function, include channel internalization [40], the formation of a potent autoinhibitor produced by proteolytic C-terminal cleavage [39], and/or calmodulinbased/Ca 2+ -dependent inactivation [28,36]. Finally, further VGCC regulation by VGCC auxiliary subunits (α 2 δ, β, and γ subunits), which modulate functional expression and trafficking of VGCCs to the membrane surface of neurons, could also contribute to the reduced Ca 2+ influx [41,42]. Whether activity and/or expression of these L-channel subunits are altered in the context of neuroHIV requires further investigation.
It is also important to note that VGCCs consist of a number of different channel subtypes [28]. Although we have focused on L-channels, other VGCCs, including the P/Q-, N-, R-, and/or T-type Ca 2+ channels [28] could also contribute to the reduced Ca 2+ influx found in HIV-affected striatal MSNs. For instance, HIV alters the function of P/Qchannels in the nucleus accumbens [43] and suprachiasmatic nucleus [44], and N-channels are involved in neuropathic pain [45], a symptom that is prevalent in HIV. Additionally, findings from our previously studies also suggest that HIV alters the function/expression of LVA-Ca 2+ channels (i.e., T-type and/or Ca v 1.3, L-type) [19,20,22,23]. The impact of neuroHIV on neuronal Ca 2+ dysregulation due to alterations in non-L-channels also is an interesting topic for future studies.
Alterations in excitatory (and/or inhibitory) neurotransmission are the likely cause of this MSN hyperexcitability in 12mo HIV-1 Tg rats. They could stem from differing sources including dysregulation of presynaptic activity, local astrocyte activity, and NMDAR or GABA A R expression and/or function in striatal MSNs. We previously demonstrated that mPFC pyramidal neurons, which project excitatory inputs to neurons in the caudateputamen [12], are hyperactive in the context of neuroAIDS modeled in HIV-1 Tg rats regardless of age [19,20,22]. Enhanced glutamate input from which could aberrantly enhance MSN excitability by promoting the activity of NMDARs without overexpression. Additionally, HIV impairs glutamate uptake by astrocytes via disrupting β-catenin signaling, increasing local levels of extracellular glutamate [15,46,47], and therefore, facilitating neuron firing. HIV also downregulates GABAergic systems [48,49], which could also enhance striatal MSN excitability. Finally, although NR2B upregulation and/or GABA A R downregulation could be additional mechanisms of altered neurotransmission, our current data exclude the likelihood of their involvement. However, because here we only initially evaluated the protein expression of the NR2B subunit of NMDARs and the β2,3 subunit of GABA A Rs, further study is needed to determine if the activity of NR2Bs and GABA A Rs; specifically, their opening probability, opening frequency, and/or opening time, or other subtypes of glutamate and GABA receptors, play a key role in this MSN dysfunction.
Together, the present study suggests that the functional activity of functional NMDARs is increased in response to enhanced glutamate inputs from the mPFC, and that may also be associated with reduced reuptake of glutamate by astrocytes in the striatum. Further, the activity of functional GABA A R could also be reduced, without changes in their expression.
Regardless of the exact source of enhanced excitatory neurotransmission that increases striatal MSN excitability in the context of neuroHIV, sustained NMDAR activation has been shown to result in internalization and degradation of Ca v 1.2 L-channels in rat cortical neurons [50], as well as reduced Ca 2+ currents via L-channels in rat ventricular myocytes [51]. Our novel findings suggest that chronically-enhanced excitatory neurotransmission initiates negative feedback to alter the function and expression of Ca v 1.2 L-channels (and likely other non-L-type VGCCs). Continued glutamate stimulation could also induce the internalization of glutamate receptors [52][53][54]. In combination with L-channel internalization [50], they could ultimately reduce Ca 2+ influx to protect neurons from excitotoxicity-induced injury or death [6,28,36,37,50], at least for a limited period of time. Accordingly, we did not observe any evident neuron death in the caudate-putamen of HIV-1 Tg rats based on NeuN protein assessments. Further studies that assess these processes in older HIV-1 Tg rats (≥24mo) would be informative regarding HIV-induced excitotoxicity during aging.
In summary, our novel findings from this study initially reveal striatal neuronal dysfunction in the mesocorticostriatal pathway of aging rats in the context of neuroHIV, which is crucial in regulating neuronal/astrocytic function and neurocognition disturbed by HIV. Together, these findings, in combination with our previous studies, suggest that HIVinduced mPFC neuronal overactivation causes abnormally-enhanced glutamate output, which induces hyperactivity of striatal MSNs, and which is associated with dysfunction of L-channels and K v channels. However, these findings unexpectedly differ from VGCC upregulation and reduced VGKC activity we found in glutamatergic mPFC pyramidal neurons of youth rats in the context of neuroHIV. Further, because rat age at 12mo is equivalent to humans at middle age (~30-yr-old) [26], such neuronal and L-channel/K v channel dysfunction could be exacerbated during further aging. Advancing our understanding of how HIV-1 disrupts the most vulnerable and susceptible brain regions that regulate neurocognition, and whether chronic exposure to HIV induces channelopathies in the brain, will aid in defining therapeutic targets for HAND.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/membranes12080737/s1, Figure S1: The full molecular weight (MW) ranges of the membranes probed for Cav1.2 and Cav1.3 are shown for the representative samples shown in Figure 2D of the main text. All membranes were cut in half to facilitate multiple probing for proteins of interest. (A) The full MW ranges of membranes probed for Cav1.2 (high MW portion) and actin (low MW portion) are shown. For the representative Cav1.2 bands, the samples were run on the same gel, but were inverted compared with the order shown; ie, the bands were rearranged to achieve the desired order for presentation. (B) The full MW ranges of membranes probed for Cav1.3 (high MW portion) and actin (low MW portion) are shown; Figure S2: The full molecular weight (MW) ranges of the membranes probed for NMDAR2B and GABAARβ2,3 are shown for the representative samples shown in Figure 3A of the main text. All membranes were cut in half to facilitate multiple probing for proteins of interest. (A) The full MW ranges of membranes probed for NMDAR (high MW portion) and actin (low MW portion) are shown. For the representative NMDAR bands, the samples were run on the same gel, but were not side-by-side; ie, the bands shown were cropped from two different areas of the same membrane. (B) The full MW ranges of membranes probed for GABAARβ2,3 and actin (both low MW portions) are shown. The membrane was first probed for GABAARβ2,3, then stripped and re-probed for actin. Funding: This research was funded by National Institutes of Health, grant number R01 NS084817, R01 DA044552 and R01 NS084817-03S1 to X-TH; and R01 DA033966, R01 NS060632, and MH122241 to LA.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: All the analyzed data related to this study are provided here. Further discussion and question can be directed to the corresponding author.