Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 May 4.
Published in final edited form as: Neuron. 2016 Apr 21;90(3):483–491. doi: 10.1016/j.neuron.2016.03.030

Microglial TNFα suppresses cocaine-induced plasticity and behavioral sensitization

Gil M Lewitus 1,1, Sarah C Konefal 1,1, Andrew D Greenhalgh 1, Horia Pribiag 1, Keanan Augereau 1, David Stellwagen 1,*
PMCID: PMC4860141  NIHMSID: NIHMS774316  PMID: 27112496

Summary

Repeated administration of cocaine results in the development of behavioral sensitization, accompanied by a decrease in excitatory synaptic strength in the nucleus accumbens (NAc) through an unknown mechanism. Furthermore, glial cells in the NAc are activated by drugs of abuse, but the contribution of glia to the development of addictive behaviors is unknown. Tumor necrosis factor alpha (TNFα), an inflammatory cytokine released by activated glia, can drive the internalization of synaptic AMPA receptors on striatal medium spiny neurons. Here we show that repeated administration of cocaine activates striatal microglia and induces TNFα production, which in turn depresses glutamatergic synaptic strength in the NAc core and limits the development of behavioral sensitization. Critically, following a period of abstinence, a weak TLR4 agonist can re-activate microglia, increase TNFα production, depress striatal synaptic strength and suppress cocaine-induced sensitization. Thus, cytokine signaling from microglia can regulate both the induction and expression of drug-induced behaviors.

Introduction

Changes in striatal processing, particularly in the NAc, are thought to be necessary for the maintenance of addictive behaviors, and repeated exposure to drugs of abuse leads to predictable changes in synaptic strength in the NAc (Luscher and Malenka, 2011). Drugs of abuse such as cocaine elevate dopamine levels in the striatum and ex-vivo treatment of striatal medium spiny neurons (MSNs) with D1 dopamine receptor agonists or with cocaine increases the phosphorylation and insertion of AMPA receptors (Chao et al., 2002; Mangiavacchi and Wolf, 2004; Snyder et al., 2000). However, repeated cocaine treatment in vivo (five days of non-contingent administration) results in an initial decrease in the AMPA/NMDA ratio on MSNs in the NAc, as measured 24hr after the last cocaine injection (Kourrich et al., 2007; Mameli et al., 2009). A period of abstinence results in a gradual elevation of AMPA/NMDA ratios and AMPAR surface expression (Boudreau and Wolf, 2005; Schumann and Yaka, 2009), although a challenge dose of cocaine will result in lowered ratios and surface receptor content (Boudreau et al., 2007; Kourrich et al., 2007; Thomas et al., 2001). Self-administration of cocaine also causes similar changes in the NAc, with cocaine exposure causing a loss of AMPA receptors and depressing synaptic strength on MSNs and extended abstinence resulting in synaptic strengthening and accumulation of surface AMPA receptors (Conrad et al., 2008; Ortinski et al., 2012; Schramm-Sapyta et al., 2006). This bidirectional plasticity suggests other factors, in addition to dopamine, contribute to the synaptic changes induced by drug exposure.

Recently, we have shown that TNFα drives internalization of AMPARs on MSNs, reducing corticostriatal synaptic strength, and reduces the aberrant changes in striatal circuit function induced by chronic blockade of D2 dopamine receptors (Lewitus et al., 2014). Glia are the main source of TNFα in the CNS, and both microglia (Sekine et al., 2008) and astrocytes (Bowers and Kalivas, 2003) are activated by psychostimulants. Further, glia have been suggested to regulate drug-induced behavior (Miguel-Hidalgo, 2009). Thus, glia through the release of TNFα could have a mitigating effect on the circuit changes induced by cocaine. Here we demonstrate that striatal microglia are activated by cocaine, and moderate the synaptic and behavioral changes induced by the repeated administration of cocaine.

Results

To determine the effect of in vivo cocaine exposure on TNFα levels in the NAc, we measured TNFα mRNA and protein levels in mice after i.p. injections of saline or cocaine. A single injection of cocaine had no effect on TNFα levels (measured 24hr post-injection), but 5 days of daily cocaine treatment (measured 24hr after the final injection) increased both TNFα mRNA and protein, compared to saline injected controls (Fig 1A–D). TNFα was no longer elevated following 10 days of abstinence from cocaine. To understand the impact of TNFα on synaptic function in the NAc, we measured AMPA/NMDA ratios on MSNs in the NAc core. Alteration in NAc core AMPA receptors are involved in the expression of behavioral sensitization to psychostimulants (Kalivas, 2009).

Figure 1. Cocaine increases TNFα levels in the nucleus accumbens, which causes synaptic depression on D1-MSNs and antagonizes cocaine-induced behavioral sensitization.

Figure 1

Open in a new tab

(A) Diagram of the time points used for experiments: 24hr after a single injection of saline or cocaine (i.p. 15mg/kg), 24hr after 5 daily injections of saline or cocaine, and 10d after 5 daily injections of saline or cocaine. (B) Representative confocal projection images of NAc immunostained for Iba1 (top) and TNFα (bottom) from mice injected for 5d with saline or cocaine (scale bar = 20μm). (C) 5 daily injections of cocaine increases TNFα protein in the NAc. (D) 5 daily injections of cocaine increases TNFα mRNA in the ventral striatum. (E) Representative recording of EPSCs at −70 mV and +40 mV and mean AMPA/NMDA ratios from control slices and slices treated with 10 or 100ng/ml TNFα in D1 (red) and D2 (green) MSNs in the NAc core. AMPA/NMDA ratios were calculated using the peak amplitude at −70mV for AMPA and the amplitude at +40mV taken 40msec after the peak at −70mV. (F) Representative traces and mean AMPA/NMDA ratios from D1-MSNs in the NAc core, after 1d and 5d of cocaine or saline in WT or TNFα-KO mice. Ratios from mice injected with one or five daily doses of saline were not significantly different, and were combined. (G) Representative traces and mean AMPA/NMDA ratios from D1-MSNs in the NAc core from control slices or slices treated ex-vivo with TNFα. Treatment with 10ng/ml TNFα significantly reduced AMPA/NMDA ratios from TNFα−/− mice treated 24hr prior with cocaine. Ex-vivo treatment with 100ng/ml TNFα did not further decrease AMPA/NMDA ratios from WT mice previously exposed to 5 daily cocaine injections. (H) Mean locomotor activity in response to cocaine injections in TNFα−/− and WT mice, showing higher sensitization in TNFα−/− mice, that was maintained after abstinence (n = 12 WT, 17 TNFα−/− animals). (I) Blocking soluble TNFα signaling only during the sensitization protocol (DN-TNF sensi) with DN-TNF is sufficient to sustain the elevation of the cocaine response to the challenge dose on day 15, while blocking TNFα signaling during the withdrawal period (DN-TNF withd) had no effect on the response to the challenge dose after withdrawal. (n = 16 DN-TNF sensi, 8 DN-TNF withd, 12 Control). Results are expressed as mean±SEM, n’s (mice or cells) are given in bars. * p < 0.05, ** p < 0.01, *** p < 0.001

We have previously shown that TNFα drives internalization of AMPARs on MSNs in the dorsal striatum (Lewitus et al., 2014). As repeated cocaine administration primarily affects direct-pathway MSNs, we tested specific subpopulations of MSNs in the NAc core for their response to TNFα. Acute NAc slices were incubated with TNFα and whole-cell recording made from Drd1a-td Tomato (D1) positive and negative (D2) MSNs. A low dose of TNFα (10ng/ml) had no significant effect on either cell type. However, 100ng/ml TNFα significantly reduced the AMPA/NMDA ratio on D1-MSNs, with a non-significant reduction in ratios on D2-MSNs (Fig 1E). These results suggest that D1-MSNs are more sensitive to TNFα than D2-MSNs, although D2-MSNs may respond to a lesser degree.

Repeated non-contingent administration of cocaine results in lower AMPA/NMDA ratios of excitatory inputs onto the NAc specifically on D1-MSNs (Kim et al., 2011; Pascoli et al., 2012). To test whether this decrease in AMPA/NMDA ratios is due to increased TNFα expression, we evaluated AMPA/NMDA ratios in the NAc core after cocaine or saline administration in WT and TNFα−/− mice. As expected for WT mice, a single injection of cocaine did not significantly reduce AMPA/NMDA ratios on D1-MSNs as measured 1 day later, but 5 daily cocaine injections did (Fig 1F). Strikingly, in TNFα−/− mice, a single injection of cocaine significantly increased AMPA/NMDA ratios, which remained elevated after 5 days of cocaine treatment (Fig 1F). No significant differences were observed in D2-MSNs for either genotype (Fig S1A). These results suggest that the reduction in AMPA/NMDA ratios in D1-MSNs after repeated cocaine is due to increased TNFα in the NAc, and that cocaine itself increases synaptic strength. This is consistent with the exocytosis of AMPARs observed with direct stimulation of D1Rs on MSNs (Mangiavacchi and Wolf, 2004). Moreover, this result suggests that following cocaine treatment, D1-MSNs are more responsive to lower endogenous levels of TNFα, perhaps because newly inserted AMPARs are more labile, as has been seen at potentiated synapses in the amygdala (Clem and Huganir, 2010). To test the hypothesis that potentiated D1-MSNs are more sensitive to TNFα, we treated TNFα−/− mice with a single injection of cocaine and evaluated the effect of a low level of TNFα (10ng/ml) on AMPA/NMDA ratios. Although this level of TNFα had no significant effect on MSNs from WT untreated animals (Fig 1E), it significantly reduced AMPA/NMDA ratios on D1-MSNs from cocaine treated knockout animals (Fig 1G). This treatment had no effect on D2-MSNs (Fig S1B). To test whether cocaine-induced TNFα signaling occludes further synaptic depression by TNFα, we treated WT animals with cocaine for 5 days, and then treated striatal slices ex-vivo with TNFα (100ng/ml). The AMPA/NMDA ratio on D1-MSNs (already reduced compared to saline treated animals; Fig 1E) was not further reduced by treatment with TNFα (Fig 1G). This shows that the synaptic depression induced by repeated cocaine injections occludes the TNFα-mediated reduction in AMPA/NMDA ratio. Overall, these data suggest that repeated cocaine treatment elevates TNFα, which suppresses the synaptic changes directly induced by cocaine in the NAc core.

Depressing synaptic strength in the NAc can reduce behavioral sensitization to cocaine (Pascoli et al., 2012). Behavioral sensitization is a simple model of drug-induced behavioral change, which measures the progressive increase in locomotor response to psychostimulants. TNFα−/− mice displayed an increased initial locomotor response to cocaine and increased sensitization, compared to WT mice (Fig 1H). This is similar to what has been observed with methamphetamine sensitization in TNFα−/− mice (Nakajima et al., 2004). To exclude compensatory mechanisms resulting from the absence of TNFα during development, we pharmacologically blocked the soluble form of TNFα in WT mice using XENP1595 (a dominant negative variant of TNFα (DN-TNF)). WT mice were administered DN-TNF either during the 5 days of conditioning (to block TNFα signaling during acquisition) or during the abstinence period starting immediately after the last cocaine injection (to test the role of TNFα in the maintenance of the behavior). Blocking TNFα signaling during acquisition was sufficient to increase sensitization as well as maintain the elevated response on the challenge day, while blocking TNFα signaling during the 10d period of abstinence had no effect on the response to the challenge dose (Fig 1I). These results suggest that TNFα is active during acquisition but not during drug abstinence, consistent with the increased TNFα expression observed during that period (Fig 1C, D). Further, the increased sensitivity observed in TNFα−/− mice on the first day of cocaine is not due to an acute loss of TNFα signaling and is likely unrelated to the increase in sensitization.

In other brain regions, TNFα that regulates synaptic function is produced by glia (Stellwagen and Malenka, 2006). To assess the source of TNFα regulating sensitization, we utilized a Cre-loxP system to selectively delete TNFα from microglia (CX3CR1-Cre; Fig S2A) and astrocytes (GFAP-Cre; Fig S2B). Mice that lack microglial TNFα showed significantly higher sensitization to cocaine from the second cocaine injection that was maintained through the period of abstinence (Fig 2A). Conversely, mice that lack astrocytic TNFα did not display a significant change in sensitization compared with littermate controls (Fig 2B). These results suggest that microglia are important for the adaptive TNFα response to repeated cocaine administration. To verify this, we isolated microglia from the striatum of cocaine and saline treated animals by magnetic bead sorting (Fig S2C; (Butovsky et al., 2014)), and compared TNFα mRNA in the microglial and non-microglial fractions. Microglia contained the vast majority of TNFα mRNA, showing over a 20-fold enrichment compared with the other striatal cell types (Fig 2C). Further, the TNFα mRNA was increased by cocaine treatment specifically in microglia cells (Fig 2D) and not other cell types (Fig S2D).

Figure 2. Microglia are activated by cocaine and release TNFα to antagonize cocaine-induced behavioral sensitization.

Figure 2

Open in a new tab

(A) Mean locomotor activity in response to cocaine in mice that lack microglial TNFα (CX3CR1-Cre+; TNFαflox/flox) and littermate controls (CX3CR1-cre negative or TNFα+/flox). The elevation was sustained for a final test dose of cocaine (n = 16 control; 12 microglia deletion). (B) Mean locomotor activity in response to cocaine injections in mice that lack astrocytic TNFα (GFAP-Cre+; TNFαflox/flox and littermate controls (GFAP-cre−; TNFαflox/flox). GFAP-Cre mice had normal sensitization and response on the final test day (n = 25 per condition). (C) Purified microglia (CD11b+ fraction of cells) from whole striatum tissue express significantly more TNFα mRNA compared to other cell types (CD11b− fraction) (5 mice pooled per group in each experiment). (D) 5 daily injections of cocaine increases TNFα mRNA in striatal microglia (5 mice pooled per group per experiment). (E) Representative confocal projection images of Iba1-labeled microglia in the NAc 24hr after a single cocaine injection, 5 daily injections or 10d withdrawal (scale bar = 20μm). (F) Semi-quantitative analysis of Iba1 immunoreactivity in the NAc, normalized to the mean saline intensity for each time point. (G) Quantification of microglia cell body size (μm2) measured by Iba1 immunoreactivity and normalized to the mean saline value for each time point. (n = cells from 4 animals). (H) Representative examples of microglia processes, after 5d of cocaine or saline. (I) Total length of microglia processes is decreased after 5 days cocaine by 20%, but is not significantly altered after withdrawal (n = microglia from 4 animals). Results are expressed as mean±SEM, n’s (experiments, mice or microglia) are given in bars. * p < 0.05, ** p < 0.01

Resting microglia continuously survey the healthy brain and respond to a variety of activation signals by undergoing progressive morphological and functional changes (Kettenmann et al., 2011). Using Iba1, we labeled microglia in adult mice, 24 hours after a single cocaine injection, 24hrs after 5 days of daily cocaine injections, or after 10d of drug abstinence. Although the number of microglia in the NAc did not change at any time point (Fig S2E), Iba1 intensity was increased in microglia by 5 days of cocaine, and after a period of abstinence (Fig 2E–F). Microglia cell body size was increased by 5d of cocaine (Fig 2G), accompanied by a decrease in process length (Fig 2H–I). These changes in microglia morphology are consistent with an activated phenotype. In contrast, we did not observe any activation of astrocytes, as judged by the area or intensity of GFAP expression (Fig S2F–G). These data strongly suggest that microglia are the source of the cocaine-induced upregulation of TNFα production in the striatum observed during sensitization.

Cocaine could activate microglia directly by binding the sigma receptor (Navarro et al., 2010) or the Toll-like receptor 4 (TLR4) (Northcutt et al., 2015), or indirectly through the elevation of dopamine. To test this, we treated microglia cultures with dopamine or cocaine for 3hr. Treatment of microglia with 0.1μM dopamine, but not 1μM cocaine, significantly increased TNFα mRNA (Fig 3A–B). This concentration of dopamine is reflective of the concentration found in the NAc in vivo following cocaine administration in rats (Hooks et al., 1992). The same treatment applied to cultured astrocytes had no effect on TNFα mRNA (Fig S3A). Multiple dopamine receptors are expressed on microglia (Kettenmann et al., 2011). Stimulating microglia with the D2-like agonist quinpirole increased TNFα mRNA, while the D1 agonist SKF-38393 and specific D3 agonist pramipexole had no effect (Fig 3C). This suggests that dopamine increases TNFα production in microglia through D2 receptors. Microglia transcriptional profiles change substantially during both development and the culturing process (Butovsky et al., 2014). We therefore cultured microglia from adult animals, in a manner that preserves an in vivo transcriptional profile (Butovsky et al., 2014). We saw a similar response to dopamine and lack of response to cocaine (Fig S3B–C). To directly test the response of microglia in vivo, we treated mice with quinpirole (0.5 mg/kg, i.p.; 24 and 1hr prior to harvest) and isolated microglia. Quinpirole treatment increased the TNFα mRNA in striatal microglia cells (Fig 3D), but not in other striatal cell types (Fig S3D) nor in cortical microglia (Fig S3E), indicating that D2-agonism specifically increases microglial TNFα in the striatum in vivo. Co-administrating cocaine with the D2 antagonist L741,626 for 5d reduced the cocaine-induced increase in microglial TNFα production (Fig 3E). Finally, to verify that D2-like receptor activation was required for the TNFα-dependent decrease in AMPA/NMDA ratio observed on D1-MSNs, we treated animals for 5d with cocaine and the D2 antagonist. Blocking D2 receptors (and thus preventing the activation of microglia) resulted in a large cocaine-induced increase in AMPA/NMDA ratio on D1-MSNs, similar to TNFα−/− animals (Fig 3F). Overall, this suggests that cocaine elevates dopamine levels, which act on D1 receptors on direct pathway MSNs to increase synaptic strength, and simultaneously activates microglia through D2 receptors and temporarily increases TNFα production.

Figure 3. Dopamine increases TNFα mRNA in microglia through D2 receptors.

Figure 3

Open in a new tab

(A) Primary rat microglia cultures were treated with vehicle or dopamine for 3h (n = biological replicates from 4 independent cultures). (B) Treatment with cocaine (1μm, 3h) did not alter TNFα mRNA levels in microglia cultures (n = replicates from 3 cultures). (C) Normalized change in TNFα mRNA in primary rat microglia cultures treated for 3h with vehicle (Control), D1-receptor agonist (SKF-38393, 1μm), D2-receptor agonist (quinpirole, 1μm) or D3-agonist (pramipexole, 1μm). (n = replicates from 6 cultures). (D) Quinpirole (i.p. 0.5mg/kg; 24hr and 1hr before harvesting) significantly increases TNFα mRNA in microglia isolated from striatal tissue, compared to saline treatment (n = experiments, 4 mice pooled per group in each experiment). (E) Co-administration of the D2-receptor antagonist L741,626 (i.p. 3mg/kg; 15min before cocaine) with daily cocaine injections over 5 days significantly decreases TNFα mRNA in ventral striatum tissue in adult mice. (F) Co-administration of L741,626 with cocaine results in an increase in AMPA/NMDA ratio on D1-MSNs compared with mice treated with cocaine alone. Results are expressed as mean±SEM, n’s (experiments, mice or cells) are given in bars. * p < 0.05, ** p < 0.01, *** p < 0.001

Our data suggests that the activation of microglia limits the cocaine-induced changes to NAc circuitry, but this activation occurs only during a narrow window following cocaine exposure. Because depotentiation of MSNs reduces cocaine-induced behavioral sensitization (Pascoli et al., 2012), we tested if re-activation of microglia to increase TNFα could depress NAc synapses and suppress sensitization. To do this, we utilized monophosphoryl lipid A (MPLA), a detoxified variant of LPS (Casella and Mitchell, 2008). MPLA is a weak TLR4 agonist that does not induce extensive neuroinflammation or sickness behavior (Michaud et al., 2013). We first verified that MPLA activates microglia in the NAc, by injecting 10μg MPLA IP after 10 days of abstinence from cocaine. MPLA treatment significantly increased Iba1 intensity within 4hr compared to saline treated controls (Fig 4A) and was associated with an increase in striatal TNFα expression, at both 4hr and 24hr after injection (Fig 4B). We next tested if MPLA would depress synaptic strength in the NAc core. After 10d of abstinence from cocaine, mice were injected with MPLA and evaluated 24hr later for AMPA/NMDA ratios on D1-MSNs. MPLA treatment significantly reduced AMPA/NMDA ratio in D1-MSNs compared to saline treated controls (Fig 4C). Because artificially reducing synaptic strength in the NAc can reduce behavioral sensitization (Pascoli et al., 2012), these data suggest that MPLA might suppress drug-induced behaviors.

Figure 4. MPLA activates microglia in the nucleus accumbens and decreases behavioral sensitization to cocaine via TNFα.

Figure 4

Open in a new tab

(A) Representative confocal projection images of Iba1 immunostaining in the NAc, after MPLA (10μg) or saline injection in mice after 10d withdrawal. Scale bar 20μm. Semi-quantification of immunoreactivity reveals that Iba1 intensity was significantly increased 24hr after a single MPLA injection. (B) MPLA (10μg) significantly increases TNFα mRNA in the ventral striatum at 4hr and 24hr. (C) Representative traces and AMPA/NMDA ratios from D1-MSNs in the NAc core 24hr after MPLA (10μg) or saline injection in mice after 10 days withdrawal. (D) MPLA does not reduce the initial locomotor response to cocaine. Mice were given 7 daily saline injections, then after 9d of abstinence given an injection of saline or MPLA (10μg), followed the next day by a challenge dose of cocaine. MPLA did not alter the response to the challenge dose. This response was lower than the sensitized response in control animals given cocaine during training (n = 8 sal/sal, 7 sal/MPLA, 6 coc/sal). (E) MPLA did reduce the sensitized response to cocaine. After withdrawal, WT mice were injected with MPLA (10μg or 50μg) or saline, and tested 24hr later with a challenge dose of cocaine (n = 20 for control, 12 for 10μg MPLA, and 10 for 50μg MPLA). (F) MPLA treatment had no effect on sensitization in TNFα−/− mice, as MPLA (10μg) did not reduce the response to a challenge dose of cocaine in TNFα−/− mice (n = 10 saline, 11 MPLA). Results are expressed as mean±SEM. * p < 0.05, ** p < 0.01, *** p < 0.001

We first established that MPLA does not alter basal locomotion (Fig S4A). Further, MPLA (10μg; 24hr prior to testing) did not decrease the locomotor response to an initial dose of cocaine, as tested in saline-treated animals given an initial dose of cocaine at the challenge time point (Fig 4D). We then tested sensitized mice by injecting 10μg or 50μg MPLA or saline, 24hr prior to the challenge dose of cocaine. Mice treated with MPLA had significantly reduced locomotor response to the cocaine challenge in a dose dependent manner (Fig 4E). This suggests that MPLA reduces sensitization, rather than the locomotor response to cocaine. The effects of MPLA also did not appear to be due to an increase in sensitivity to cocaine, as we observed no increase in stereotypic behaviors in MPLA treated mice (Fig S4B–D). Moreover, MPLA had no effect on sensitization in TNFα−/− mice (Fig 4F), which suggests that MPLA acts through an elevation of TNFα and not other cytokines. However, this effect is temporary, as MPLA had little impact on sensitization when tested 4d after injection (Fig S4E). These observations suggest that even after a prolonged period of abstinence from cocaine, increasing TNFα can be effective in reducing the behavioral response to cocaine although it does not revert the system to the pre-sensitized state.

Discussion

Chronic cocaine administration produces long term neuroadaptions of glutamatergic signaling in the NAc that contribute to addiction-related changes in drug sensitivity and craving. Here we show that microglia in the NAc are transiently activated following cocaine administration, and act to down-regulate AMPARs on MSNs through TNFα signaling. Importantly, this limits the development of behavioral sensitization. Following a period of abstinence, MPLA can re-activate microglia and decrease both synaptic strength in the NAc and locomotor sensitization to cocaine. Our results suggest that microglia have an adaptive role in the response to cocaine and their modulation could be an effective avenue of treatment. Alterations in the NAc core may be particularly important for the expression of locomotor sensitization (Kalivas, 2009). Infusion of AMPA into the NAc core enhances locomotion in animals exposed to cocaine 2–3 weeks earlier (Bell and Kalivas, 1996), while AMPAR antagonists administered into the NAc core prevent the expression of sensitization (Bell et al., 2000; Pierce et al., 1996). The increase in NAc synaptic strength has been hypothesized to correlate with the development of craving (Conrad et al., 2008), and reducing synaptic strength reduces cue-induced self-administration (Wisor et al., 2011). This suggests that increasing TNFα with mild TLR4 activation may help blunt craving or incentive sensitization

Further, our results explain a perplexing feature of chronic cocaine administration – that synaptic strength on D1-MSNs decreases initially and then slowly increases during withdrawal. While the formation and subsequent unsilencing of silent synapses likely contributes (Huang et al., 2009) to the changes in AMPA/NMDA ratios, our results suggest an additional mechanism is involved. Our data support the idea that dopamine does, as predicted by in vitro results, increase AMPA/NMDA ratios while simultaneously activating microglia to release TNFα. This TNFα release causes the decreased AMPA/NMDA ratios observed following repeated administration of cocaine. This suppression is temporary as microglia slowly de-activate during abstinence from cocaine, revealing the underlying dopamine-induced potentiation. A challenge dose would reactivate the microglia, increase TNFα release, and again suppress AMPA/NMDA ratios as observed (Boudreau et al., 2007; Thomas et al., 2001). Taken together, our data suggest that TNFα has an adaptive role in regulating glutamatergic transmission in the NAc when circuit homeostasis is perturbed, akin to the role of TNFα in homeostatic synaptic plasticity (Stellwagen and Malenka, 2006).

Moreover, our data suggest that microglia are adaptive regulators of striatal function. This is not to suggest that astrocytes do not regulate striatal function, including the response to drugs; merely that astrocytes do not supply the TNFα that opposes the circuit and behavioral changes induced by cocaine. CX3CR1 is expressed in other cell types (including macrophages and a small number of neurons), and a TLR4 agonist like MPLA will act on astrocytes and other cell types. Therefore we cannot exclude the contribution of other cell types to the TNFα response, but it is difficult to argue that microglia are not the major source of the response. Microglia express a variety of neurotransmitters, neuropeptides, and immune receptors and have the capacity to rapidly respond to physiological changes in the brain (Kettenmann et al., 2011). Our results support the idea that moderate microglia activation has a role in a homeostatic-type response to significant deviations from the basal state (Kierdorf and Prinz, 2013) and has a similar beneficial response in reestablishing homeostasis following stress (Kreisel et al., 2014). Hence, augmenting the microglial response, through TLR4 or other means, might be a useful approach to treat addiction, provided it only moderately activates the microglia. MPLA, a weak TLR4 agonist, has been shown to significantly improve cognitive function in a mouse model of neurodegeneration (Michaud et al., 2013). We found that MPLA can acutely reduce behavioral sensitization after prolonged abstinence from cocaine. If MPLA is found to similarly diminish reinstatement, it would suggest that MPLA could reduce the motivation to acquire drugs and be used to prevent relapse, a significant problem in the treatment of addiction.

Experimental Procedures

All experiments were approved by the Canadian Council for Animal Care and the Montreal General Hospital Facility Animal Care Committee. Experimental procedures and details of statistics are available in Supplemental Experimental Procedures.

Supplementary Material

supplement

Acknowledgments

We thank Alexandre Trottier, JP Clement, Kevin Jepson and Joseph Lee for expert technical assistance, and David Szymkowski (Xencor) for the donation of XENP1595. This work was supported by NIDA, CIHR, and NSERC (DS), the Ronald Peter Griggs Fellowship from ALS Canada (GML), and the Neuroinflammation Training Program (SCK).

Footnotes

Contributions

All authors did experiments and analyzed data; GML and DS conceived the project; GML, SCK, DS wrote the paper.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Bell K, Duffy P, Kalivas PW. Context-specific enhancement of glutamate transmission by cocaine. Neuropsychopharmacology. 2000;23:335–344. doi: 10.1016/S0893-133X(00)00100-7. [DOI] [PubMed] [Google Scholar]
  2. Bell K, Kalivas PW. Context-specific cross-sensitization between systemic cocaine and intra-accumbens AMPA infusion in the rat. Psychopharmacology (Berl) 1996;127:377–383. doi: 10.1007/s002130050101. [DOI] [PubMed] [Google Scholar]
  3. Boudreau AC, Reimers JM, Milovanovic M, Wolf ME. Cell surface AMPA receptors in the rat nucleus accumbens increase during cocaine withdrawal but internalize after cocaine challenge in association with altered activation of mitogen-activated protein kinases. Journal of Neuroscience. 2007;27:10621–10635. doi: 10.1523/JNEUROSCI.2163-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boudreau AC, Wolf ME. Behavioral sensitization to cocaine is associated with increased AMPA receptor surface expression in the nucleus accumbens. Journal of Neuroscience. 2005;25:9144–9151. doi: 10.1523/JNEUROSCI.2252-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bowers MS, Kalivas PW. Forebrain astroglial plasticity is induced following withdrawal from repeated cocaine administration. Eur J Neurosci. 2003;17:1273–1278. doi: 10.1046/j.1460-9568.2003.02537.x. [DOI] [PubMed] [Google Scholar]
  6. Butovsky O, Jedrychowski MP, Moore CS, Cialic R, Lanser AJ, Gabriely G, Koeglsperger T, Dake B, Wu PM, Doykan CE, et al. Identification of a unique TGF-beta-dependent molecular and functional signature in microglia. Nat Neurosci. 2014;17:131–143. doi: 10.1038/nn.3599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Casella CR, Mitchell TC. Putting endotoxin to work for us: monophosphoryl lipid A as a safe and effective vaccine adjuvant. Cellular and molecular life sciences: CMLS. 2008;65:3231–3240. doi: 10.1007/s00018-008-8228-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chao SZ, Lu W, Lee HK, Huganir RL, Wolf ME. D(1) dopamine receptor stimulation increases GluR1 phosphorylation in postnatal nucleus accumbens cultures. Journal of Neurochemistry. 2002;81:984–992. doi: 10.1046/j.1471-4159.2002.00877.x. [DOI] [PubMed] [Google Scholar]
  9. Clem RL, Huganir RL. Calcium-permeable AMPA receptor dynamics mediate fear memory erasure. Science. 2010;330:1108–1112. doi: 10.1126/science.1195298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Conrad KL, Tseng KY, Uejima JL, Reimers JM, Heng LJ, Shaham Y, Marinelli M, Wolf ME. Formation of accumbens GluR2-lacking AMPA receptors mediates incubation of cocaine craving. Nature. 2008;454:118–121. doi: 10.1038/nature06995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hooks MS, Colvin AC, Juncos JL, Justice JB., Jr Individual differences in basal and cocaine-stimulated extracellular dopamine in the nucleus accumbens using quantitative microdialysis. Brain Res. 1992;587:306–312. doi: 10.1016/0006-8993(92)91012-4. [DOI] [PubMed] [Google Scholar]
  12. Huang YH, Lin Y, Mu P, Lee BR, Brown TE, Wayman G, Marie H, Liu W, Yan Z, Sorg BA, et al. In vivo cocaine experience generates silent synapses. Neuron. 2009;63:40–47. doi: 10.1016/j.neuron.2009.06.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kalivas PW. The glutamate homeostasis hypothesis of addiction. Nat Rev Neurosci. 2009;10:561–572. doi: 10.1038/nrn2515. [DOI] [PubMed] [Google Scholar]
  14. Kettenmann H, Hanisch UK, Noda M, Verkhratsky A. Physiology of microglia. Physiol Rev. 2011;91:461–553. doi: 10.1152/physrev.00011.2010. [DOI] [PubMed] [Google Scholar]
  15. Kierdorf K, Prinz M. Factors regulating microglia activation. Frontiers in cellular neuroscience. 2013;7:44. doi: 10.3389/fncel.2013.00044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kim J, Park BH, Lee JH, Park SK, Kim JH. Cell type-specific alterations in the nucleus accumbens by repeated exposures to cocaine. Biol Psychiatry. 2011;69:1026–1034. doi: 10.1016/j.biopsych.2011.01.013. [DOI] [PubMed] [Google Scholar]
  17. Kourrich S, Rothwell PE, Klug JR, Thomas MJ. Cocaine experience controls bidirectional synaptic plasticity in the nucleus accumbens. Journal of Neuroscience. 2007;27:7921–7928. doi: 10.1523/JNEUROSCI.1859-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kreisel T, Frank MG, Licht T, Reshef R, Ben-Menachem-Zidon O, Baratta MV, Maier SF, Yirmiya R. Dynamic microglial alterations underlie stress-induced depressive-like behavior and suppressed neurogenesis. Mol Psychiatry. 2014;19:699–709. doi: 10.1038/mp.2013.155. [DOI] [PubMed] [Google Scholar]
  19. Lewitus GM, Pribiag H, Duseja R, St-Hilaire M, Stellwagen D. An Adaptive Role of TNFalpha in the Regulation of Striatal Synapses. J Neurosci. 2014;34:6146–6155. doi: 10.1523/JNEUROSCI.3481-13.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Luscher C, Malenka RC. Drug-Evoked Synaptic Plasticity in Addiction: From Molecular Changes to Circuit Remodeling. Neuron. 2011;69:650–663. doi: 10.1016/j.neuron.2011.01.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mameli M, Halbout B, Creton C, Engblom D, Parkitna JR, Spanagel R, Luscher C. Cocaine-evoked synaptic plasticity: persistence in the VTA triggers adaptations in the NAc. Nature Neuroscience. 2009;12:1036–U1108. doi: 10.1038/nn.2367. [DOI] [PubMed] [Google Scholar]
  22. Mangiavacchi S, Wolf ME. D1 dopamine receptor stimulation increases the rate of AMPA receptor insertion onto the surface of cultured nucleus accumbens neurons through a pathway dependent on protein kinase A. Journal of Neurochemistry. 2004;88:1261–1271. doi: 10.1046/j.1471-4159.2003.02248.x. [DOI] [PubMed] [Google Scholar]
  23. Michaud JP, Halle M, Lampron A, Theriault P, Prefontaine P, Filali M, Tribout-Jover P, Lanteigne AM, Jodoin R, Cluff C, et al. Toll-like receptor 4 stimulation with the detoxified ligand monophosphoryl lipid A improves Alzheimer’s disease-related pathology. Proc Natl Acad Sci U S A. 2013;110:1941–1946. doi: 10.1073/pnas.1215165110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Miguel-Hidalgo JJ. The role of glial cells in drug abuse. Current drug abuse reviews. 2009;2:72–82. [PubMed] [Google Scholar]
  25. Nakajima A, Yamada K, Nagai T, Uchiyama T, Miyamoto Y, Mamiya T, He J, Nitta A, Mizuno M, Tran MH, et al. Role of tumor necrosis factor-alpha in methamphetamine-induced drug dependence and neurotoxicity. Journal of Neuroscience. 2004;24:2212–2225. doi: 10.1523/JNEUROSCI.4847-03.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Navarro G, Moreno E, Aymerich M, Marcellino D, McCormick PJ, Mallol J, Cortes A, Casado V, Canela EI, Ortiz J, et al. Direct involvement of sigma-1 receptors in the dopamine D(1) receptor-mediated effects of cocaine. Proceedings of the National Academy of Sciences of the United States of America. 2010;107:18676–18681. doi: 10.1073/pnas.1008911107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Northcutt AL, Hutchinson MR, Wang X, Baratta MV, Hiranita T, Cochran TA, Pomrenze MB, Galer EL, Kopajtic TA, Li CM, et al. DAT isn’t all that: cocaine reward and reinforcement require Toll-like receptor 4 signaling. Mol Psychiatry. 2015;20:1525–1537. doi: 10.1038/mp.2014.177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ortinski PI, Vassoler FM, Carlson GC, Pierce RC. Temporally dependent changes in cocaine-induced synaptic plasticity in the nucleus accumbens shell are reversed by D1-like dopamine receptor stimulation. Neuropsychopharmacology. 2012;37:1671–1682. doi: 10.1038/npp.2012.12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pascoli V, Turiault M, Luscher C. Reversal of cocaine-evoked synaptic potentiation resets drug-induced adaptive behaviour. Nature. 2012;481:71–75. doi: 10.1038/nature10709. [DOI] [PubMed] [Google Scholar]
  30. Pierce RC, Bell K, Duffy P, Kalivas PW. Repeated cocaine augments excitatory amino acid transmission in the nucleus accumbens only in rats having developed behavioral sensitization. J Neurosci. 1996;16:1550–1560. doi: 10.1523/JNEUROSCI.16-04-01550.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Schramm-Sapyta NL, Olsen CM, Winder DG. Cocaine self-administration reduces excitatory responses in the mouse nucleus accumbens shell. Neuropsychopharmacology. 2006;31:1444–1451. doi: 10.1038/sj.npp.1300918. [DOI] [PubMed] [Google Scholar]
  32. Schumann J, Yaka R. Prolonged withdrawal from repeated noncontingent cocaine exposure increases NMDA receptor expression and ERK activity in the nucleus accumbens. J Neurosci. 2009;29:6955–6963. doi: 10.1523/JNEUROSCI.1329-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sekine Y, Ouchi Y, Sugihara G, Takei N, Yoshikawa E, Nakamura K, Iwata Y, Tsuchiya KJ, Suda S, Suzuki K, et al. Methamphetamine causes microglial activation in the brains of human abusers. J Neurosci. 2008;28:5756–5761. doi: 10.1523/JNEUROSCI.1179-08.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Snyder GL, Allen PB, Fienberg AA, Valle CG, Huganir RL, Nairn AC, Greengard P. Regulation of phosphorylation of the GluR1 AMPA receptor in the neostriatum by dopamine and psychostimulants in vivo. J Neurosci. 2000;20:4480–4488. doi: 10.1523/JNEUROSCI.20-12-04480.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Stellwagen D, Malenka RC. Synaptic scaling mediated by glial TNF-alpha. Nature. 2006;440:1054–1059. doi: 10.1038/nature04671. [DOI] [PubMed] [Google Scholar]
  36. Thomas MJ, Beurrier C, Bonci A, Malenka RC. Long-term depression in the nucleus accumbens: a neural correlate of behavioral sensitization to cocaine. Nat Neurosci. 2001;4:1217–1223. doi: 10.1038/nn757. [DOI] [PubMed] [Google Scholar]
  37. Wisor JP, Schmidt MA, Clegern WC. Cerebral microglia mediate sleep/wake and neuroinflammatory effects of methamphetamine. Brain, behavior, and immunity. 2011;25:767–776. doi: 10.1016/j.bbi.2011.02.002. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplement
NIHMS774316-supplement.pdf (2.6MB, pdf)

RESOURCES