Estradiol-Induced Potentiation of Dopamine Release in Dorsal Striatum Following Amphetamine Administration Requires Estradiol Receptors and mGlu5

Abstract Estradiol potentiates behavioral sensitization to cocaine as well as self-administration of cocaine and other drugs of abuse in female rodents. Furthermore, stimulated dopamine (DA) in the dorsolateral striatum (DLS) is rapidly enhanced by estradiol, and it is hypothesized that this enhanced DA release mediates the more rapid escalation of drug taking seen in females, compared with males. The mechanisms mediating the effect of estradiol to enhance stimulated DA release were investigated in this study. Using in vivo microdialysis and high performance liquid chromatography coupled with electrochemical detection, we first examined the effect of estradiol on amphetamine-induced DA increase in the DLS of ovariectomized rats. We then tested whether the potentiation of this DA increase could be blocked by the estradiol receptor antagonist, ICI 182,780 (ICI), or an antagonist to the metabotropic glutamate receptor subtype 5 (mGlu5), 2-methyl-6-(phenylethynyl)pyridine (MPEP). There is evidence that estradiol receptors collaborate with mGlu5 within caveoli in DLS and mGlu5 is hypothesized to mediate many of the effects of estradiol in the addiction processes in females. Our data show that estradiol enhances the DA response to amphetamine. Either ICI or MPEP prevented the effect of estradiol to enhance DA release. Importantly, our results also showed that neither ICI or MPEP alone is able to influence the DA response to amphetamine when estradiol is not administrated, suggesting that ICI and MPEP act via estradiol receptors. Together, our findings demonstrate that estradiol potentiates amphetamine-stimulated DA release in the DLS and this effect requires both estradiol receptors and mGlu5.


Introduction
Women are more susceptible to drugs of abuse than men. They escalate faster from initial use to addiction, take more drugs when addicted, and have a harder time staying abstinent (Bobzean et al., 2014). This is mirrored in animal models, female rats acquire drug self-administration at a faster rate, are more motivated to take drugs, and respond stronger to drug cues during reinstatement (Becker 2016;Song et al., 2018).
It is suggested that these sex differences are regulated at least in part by estradiol (E2). Indeed, there is considerable evidence that shows the potentiating roles of E2 in cocaine self-administration, cocaine behavioral sensitization, and dopamine (DA) signaling in the nucleus accumbens (NAc) following cocaine administration (Hu and Becker, 2008). Despite this mounting evidence, how E2 enhances stimulated DA release or addiction-related behaviors are less well understood.
Many of the E2 effects involve intracellular estrogen receptors ER␣s and ER␤s (Foster 2012;Borrow and Handa, 2017). Recently, E2 was also shown to bind to a membrane G-protein-coupled receptor GPER-1 (Long et al., 2014). Depending on the types/locations of the receptors, the effects of E2 can range from minutes (nongenomic effects) to days (genomic effects;Ervin et al., 2015). In dorsal striatum (DS), a region that is critical for habitual drug taking behavior, E2 modulates behavior by acting on GABA medium spiny neurons (MSNs; Mermelstein et al., 1996) and by altering DA transmission indirectly through a presynaptic mechanism (Xiao and Becker, 1998;Schultz et al., 2009).
In the present study by using in vivo microdialysis and high-performance liquid chromatography (HPLC) coupled with electrochemical detection (ECD), we first examined effects of E2 on amphetamine (AMPH)-induced DA elevation in the striatum of female rats. We then tested whether the observed potentiated DA elevation could be blocked by an E2 receptor antagonist ICI 182 780 (ICI) or an antagonist to the metabotropic glutamate receptor subtype 5 (mGlu5), 2-methyl-6-(phenylethynyl)pyridine (MPEP) in the striatum as there is evidence that mGlu5 is required for many of the effects of E2 in addiction processes (Martinez et al., 2014(Martinez et al., , 2016.

Animals
Female Sprague-Dawley rats (weighting 200 -225 g at the beginning of each experiment; were obtained from Harlan Laboratories or Charles River Laboratories) and housed in groups of two or three per cage before cannula implantation and singly housed after cannula implantation, under a 14/10 h light/dark cycle. The rats were housed in a room maintained at a constant temperature of 20 -21°C, with phytoestrogen-free rodent chow (2014 Teklad Global, 14% protein rodent maintenance diet, Harlan rat chow; Harlan Teklad) and water available ad libitum. All procedures were performed according to the protocol approved by the Committee for Use and Care of Animals at the University and were in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

Ovariectomy
Approximately 1 week after arrival, all animals underwent bilateral ovariectomy (OVX). The OVXs were conducted using a dorsal approach under anesthesia of ϳ2% isoflurane/oxygen. The skin was opened with an incision ϳ1 cm long along the midline just below the ribs, and a small incision ϳ0.5 cm long was made through the muscle 1.5-2 cm lateral to the midline. The ovary was externalized with blunt forceps, and the tissue between the ovary and uterus was clamped with a hemostat. The ovary was removed, and the hemostat remained in place until there was no bleeding before being released. The uterus with associated tissue was then returned to the abdomen. The procedure was repeated on the other side, and the wound was closed with 9 mm wound clips. The wound clips were removed after 14 d of OVX. After 7 d of recovery, all animals underwent vaginal lavage testing daily for 10 consecutive days to confirm cessation of cycling.

Cannula implantation
Two to 3 weeks after OVX, all rats received buprenorphine (0.01 mg/kg, s.c.) or carprofen (5 mg/kg, s.c.) 30 -60 min ahead of the cannula implantation surgery. During the surgery, all rats were anesthetized with ketamine (60 mg/kg, i.p.) and dexmedetomidine (0.3 mg/kg, i.p.). Guide cannulae (matching for CMA/11 probes, CMA/ Microdialysis, or MAB 6 probes, SciPro; 4 mm membrane length) were inserted through the skull aimed at the striatum (AP ϩ0.20 mm, ML Ϯ3.00 mm, DV Ϫ1.50 mm) using standard stereotaxic techniques. The cannulae were held in place with acrylic polymer (Lang), which was secured to the brain with three to four stainless steel jewelry screws (Small Parts). A solid stylet was placed in each cannula when not in use, to keep the cannula patent. Animals were allowed to recover for at least 5 d before microdialysis. Starting 1 d after the surgery (both cannula implantation and OVX), rats were administered with carprofen (5 mg/ kg, s.c.) daily for 3 consecutive days and triple antibiotic was given when necessary on observation. All rats were observed at least once daily for 10 consecutive days to ensure their recovery.

Preparation for microdialysis
Animals were anesthetized with 3% isoflurane and maintained with 2% isoflurane during the procedure of removing the stylet and inserting a microdialysis probe into the brain through the guide cannula. Probes were placed into the brain 12-18 h in advance of the testing to allow sufficient time for the injury-related release associ-ated with probe implantation to subside. Animals were placed in the test chamber (31 ϫ 25 ϫ 25 cm) with continuous white noise. The microdialysis probes were attached to syringes mounted on the syringe pump, and a Ringer's solution (in mM:145 NaCl, 2.7 KCl, 1 MgSO 4 , 1.2 CaCl 2 , 1.55 Na 2 HPO 4 , 0.445 NaH 2 PO 4 , pH 7.3 at RT) was continuously pumped through the probe at 1.5 l/min during the first 30 -60 min after probe insertion. Then the pumping speed was reduced to 0.3 l/min until the next day. To prevent the microdialysis probe, which was secured to the animals' head, from being subjected to the torque created during the movement of animal, the rats was fitted with a custom-made harness, and the harness was attached to a swivel (liquid commutator 375/22 or 375/D/22, Instech Laboratories) by a flexible stainless steel cable. Rats were left overnight in the testing chamber with food and water ad libitum.

Microdialysis
Sample collection was initiated the next morning, and all samples were collected in the light phase during 08: 00 -12:30. All dialysates were briefly stored on ice in the dark, and then manually injected into the HPLC-ECD system for measuring DA concentration in dialysates during 08:00 -15:00 of the same day. Dialysate was collected into vials mounted just above the harness assembly. Drugs and hormones of interest were administered systemically (i.p. or s.c.) or intrastriatally via the microdialysis probe (reverse dialysis). For delivering E2, ICI, or MPEP via reverse dialysis method, drugs were first dissolved in pure UPS grade ethanol as 1000ϫ (or above) stock solution; then, at use, they were further freshly diluted in Ringer's solution and manually filtered via 0.2 m syringe filters. With reverse dialysis, the drug of interest passes through the membrane of a microdialysis probe and diffuses into the striatum down a concentration gradient. Based on the in vitro results, we estimate that the effi-ciency of drug delivery with infusion method is 3-10% (data not shown). Thus, the effective concentration in the brain is considerably lower than the concentration in the probe. Thirty-sixty minutes before the first sample collection, the pumping speed was increased to 1.5 l/min. Each dialysate sample was collected for 10 min. Baseline samples were collected for 30 min. When drugs were delivered via reverse dialysis, five samples were collected after the solutions were changed and the last three samples were used as the new baseline (it took ϳ20 min for a new solution to reach equilibrium in the system). All rats in all experiments received an AMPH injection during microdialysis (2.5 mg/kg in saline, i.p.) and 10 min samples were collected for the following 2 h (12 samples). Figure 1 shows treatment details during the microdialysis sample collection in each experiment. Briefly, all rats were infused with Ringer's solution for determining baseline DA and then treated with one or two pretreatments before AMPH injections. Specifically, in Experiment 1, rats were randomly assigned to 1 of 4 groups: (1) E2 Group (n ϭ 7), rats were infused with Ringer's solution with E2 in it (1 ng/ml E2; first dissolved in 100% ethanol and then diluted in Ringer's solution, ethanol final concentration 0.02%); (2) Estradiol benzoate (EB) Group (n ϭ 8), rats were treated with a subcutaneous injection of EB (5 g in 0.1 ml peanut oil); (3-4) Control Groups, rats received either a subcutaneous injection of peanut oil (0.1 ml per rat; n ϭ 6) or 0.02% ethanol in Ringer's solution (vehicle for E2; n ϭ 7). Rats that were treated with peanut oil or ethanol in Ringer's solution did not significantly differ from each other and were combined in the analyses. There were two groups in Experiment 2: the ICI Group (n ϭ 9) was infused with Ringer's solution with ICI in it (2.32 g/ml ICI, which is an equimolar concentration to E2 1 ng/ml; New Research first dissolved in 100% ethanol and then diluted in Ringer's solution; ethanol final concentration was 0.1%). The rats then received a subcutaneous EB injection following the ICI treatment. Control Group (n ϭ 8) received Ringer's solution with 0.1% ethanol (vehicle for ICI), followed by an EB administration. Experiment 3 also had two groups: E2 ϩ MPEP Rats (n ϭ 9) received E2 via reverse dialysis as described above and an intraperitoneal MPEP injection (10 mg/kg). Control rats (n ϭ 9) received E2 via reverse dialysis and an intraperitoneal saline injection. In Experiment 4, rats were assigned into 1 of the 4 groups: ICI group (n ϭ 7), where ICI dissolved in Ringer's solution was administered via reverse dialysis; MPEP group (n ϭ 6), where MPEP was injected systemically as above; and two control groups where rat received intraperitoneal saline (n ϭ 4) or ICI vehicle (n ϭ 4). The two control groups were combined because of similar levels of baseline DA as well as DA concentrations following AMPH injections. All rats were injected with AMPH following these pretreatments and dialysate samples from the DLS were collected every 10 min for 2 consecutive hours.

DA concentration measurement by HPLC
DA concentration was assayed using a HPLC-ECD system described in (Hu and Becker, 2003). In brief, dialysate samples were separated on an ESA (ESA Biosciences) HPLC column (HR-80 ϫ 3.2, 3 m particle size, 80 mm length) at 40°C, with a mobile phase consisting of the following (in mM): 75 NaH 2 PO 4 , 0.2 EDTA, 1.4 OSA (1octanesul fonic acid sodium salt monohydrate; Fluka, catalog #74882) and 17% methanol in HPLC water, pH 4.7). Flow rate through the column was set to 0.7 ml/min. Dopamine was quantified using a coulometric detector (Coulochem II, ESA) equipped with a high-sensitivity analytical cell containing dual coulometric working electrodes (ESA, model #5014B). The detector settings were as follows: Detector 1, Ϫ150 mV; Detector 2, ϩ100 mV; and guard cell ϩ300 mV. Output from Detector 2 was used for dopamine quantification. The retention time of DA was ϳ2.5 min.

Histology
Four to 7 d following completion of microdialysis, animals received an overdose of anesthesia and were sacrificed. Their brains were prepared for histologic analysis using standard techniques for frozen sections and cresyl violet staining was used to determine the location of the microdialysis probes. Only data from the rats where probes were located inside the DLS are reported here. Two rats were excluded because of the probes going too ventral and six more rats were also excluded because of probe damage or sickness.

Statistical analyses
We used software SPSS V24 in all data analyses. Data were expressed in mean Ϯ SEM. The percentage increase from baseline of each rat was used to assess DA response to AMPH in each 10 min sample. Baseline was determined by the mean of all samples before AMPH injections because no difference in DA concentrations was found in these samples (data not shown). Mixed-design repeated-measure ANOVA was used to examine treatment effect (e.g., ICI vs vehicle) among groups and the effect of time on DA concentrations within each group. We focused our analyses a priori on the first four samples collected following AMPH to catch patterns of peak DA concentrations in each condition. When significant effects of treatment were found, one-way ANOVA or t test was used to determine whether there was a significant difference in the each of the four samples post-AMPH among treatment group(s) and the control group. A priori planned contrast post hoc analysis was used to examine differences among Ͼ2 groups. Two data points in the first four samples post-AMPH of all rats (from 2 separate rats) were missing because of technical issues and were replaced by the average of the data points right before and after. In cases when assumptions for parametric tests were not met, nonparametric tests (e.g., Mann-Whitney U and Kruskal-Wallis tests) were used.

Results
Experiment 1: as can be seen in Figure 2, E2 delivered via reverse dialysis directly into the DLS or EB subcuta-Figure 2 E2 in the DLS or EB, subcutaneously, potentiates DA release following AMPH injections. E2 was dissolved in Ringer's solution and was infused into the DLS via reverse microdialysis. ‫ء‬indicates a significant difference between rats treated with E2 or EB and control rats.
New Research neously significantly enhanced AMPH-induced striatal DA release relative to the control group. Repeated-measures test showed there were a significant effect of treatment (F (2,25) ϭ 4.659, p ϭ 0.019) and a significant interaction effect of Treatment ϫ Time (F (6,75) ϭ 3.640, p ϭ 0.003) in the DLS DA concentrations of the first four samples following AMPH injections. Planned post hoc comparison tests showed there were significant effects of E2 and EB compared with controls (p ϭ 0.009 and p ϭ 0.048, respectively). To better understand the time course on the differentiated elevation of peak DA levels among the three groups, one-way ANOVA tests (and a priori planned post hoc comparisons) were used to compare each of the four DA concentrations across conditions. Significant differences were found between the E2-and EB-treated rats and the control rats in the DA concentrations shortly after AMPH injections (Table 1).
Experiment 2: as shown in Figure 3, ICI significantly decreased EB-induced enhancement in the DA release in the DLS after an intraperitoneal injection of AMPH. A significant main effect of Treatment was found in the measured DA concentrations in the DLS (F (1,15) ϭ 7.360, p ϭ 0.016). There was also a significant interaction be-tween Treatment ϫ Time (repeated-measures: F (3,45) ϭ 4.045, p ϭ 0.010). Mann-Whitney U tests showed all the four samples collected after right after AMPH injections differed in DA concentrations for ICI-treated versus vehicle-treated rats (assumptions for parametric t tests were not met, so nonparametric tests were used; Table 1).
Experiment 3: as shown in Figure 4, MPEP also significantly decreased EB-induced DA potentiation in the DLS post-AMPH treatment. There were a significant main effect of Treatment in the DA concentrations (repeatedmeasures: F (1,16) ϭ 5.895, p ϭ 0.027) as well as a significant interaction of Treatment ϫ Time in the DA concentrations (repeated-measures: F (3,48) ϭ 6.031, p ϭ 0.001). Independent t tests showed in nearly all samples after AMPH injections there were significant differences between rats treated with MPEP versus those with saline in DA concentrations (Table 1).
Experiment 4: as shown in Figure 5, MPEP or ICI did not influence DA release post-AMPH when E2 was not administered in OVX rats, unlike what was seen in Experiments 2 and 3. The increase in DA concentrations post-AMPH administration did not differ in MPEP or ICI-treated rats versus controls. There was a significant effect of Time  Lastly, there was no effect of EB, E2, ICI, or MPEP on the DA release in the DLS prior to AMPH injections across all experiments (data not shown). This lack of difference indicates the above agents do not affect DA release independent of AMPH.

Discussion
The present study showed E2 enhances DA release in the DLS following AMPH administration. This enhancing effect of E2 is mediated by E2 receptors and mGlu5 receptors as blocking E2 receptors in the DLS by ICI or intraperitoneal injections of mGlu5 receptor antagonist MPEP inhibits the E2-induced DA elevation in DLS. We also showed that ICI and MPEP are not able to influence DA levels in the DLS when E2 is not administered in OVX rats.
There is mounting evidence that E2 has been implicated in addictive behavior. E2 enhances ethanol reward in female mice (Hilderbrand and Lasek, 2018). E2 is even found to increase choice of cocaine over food in male rats as observed in females (Bagley et al., 2017). Our data support the enhancing effect of E2 on reward and thus the notion that it exacerbates addictive behavior, as it increases dopamine levels in response to AMPH challenge. Interestingly, there is considerable evidence that estradiol reduces food intake in female rats (Yu et al., 2008;Butera et al., 2010;Santollo et al., 2010;Santollo and Daniels, 2015;Butler et al., 2018;but see Boswell et al., 2006;Butera et al., 2010). The mechanisms underlying the apparent differences in the roles of E2 in motivated behaviors are less well understood, but it could be that E2 acts in different brain regions to modulate different types of rewards (e.g., drug addiction vs food reward).
The ability of E2 in influencing addiction or reward may be because of its action in the midbrain dopamine reward system. Mice treated with E2 or ER␤ agonists showed increased conditioned place preference for cocaine, whereas specific knockdown of the ER␤ gene decreased cocaine conditioned place preference (Satta et al., 2018). Another study shows E2 acts on ventral tegmental area to increase the sensitivity of dopamine neurons to ethanol Figure 4 MPEP reduces E2-induced DA potentiation following AMPH injections. ‫ء‬indicates a significant difference between rats treated with MPEP and control rats.

Figure 5
Neither ICI or MPEP influences DA release in the DLS when estradiol was not administered. These rats were ovariectomized and were not given EB injections or E2 infusions. New Research (Vandegrift et al., 2017). E2 in the MPOA also increases DA release in the NAc in response to cocaine (Tobiansky et al., 2016). Our finding showed that E2 in the DLS potentiates dopamine release following AMPH injections. DLS plays a critical role in addictive behavior in both rodent and human studies. In humans, damage to dorsal striatum alleviates addiction to alcohol and nicotine (Muskens et al., 2012). In rodent studies, it has been suggested that dorsal medial striatum and NAc are crucial in the initial acquisition of the reward and then DLS and NAc begin to take over when the behavior becomes more addiction-like. Together, it is possible that E2 acts on different regions to convergently modulate addictive behavior.
Both ER␣ and ER␤ have been reported in the E2 modulation of addictive behavior. The ER␣ agonist (propylpyrazole triol) and the ER␤ agonist (diarylpropionitrile), independently increased choice on the high-reward tested in an operant chamber (Uban et al., 2012). These effects were most pronounced 24 h after administration suggesting genomic action of the receptors. Effects of E2 via its action on membrane receptors have been debated (Govind and Thampan, 2003) and there is increasing evidence showing rapid effects of E2 that are likely via non-genomic receptors (Revankar et al., 2005;Micevych et al., 2015;Paletta et al., 2018;Yoest et al., 2018). E2 is found to exert its effects via acting on G-protein-coupled estrogen receptors (GPER-1), as well as ER␣ and ER␤ receptors, to rapidly facilitate short-term memory in female mice . Our finding in the present study showed that E2 rapidly potentates dopamine release following AMPH treatment in the DLS. It will be important to further investigate the roles of each receptor type/location in these effects.
Several studies have shown that mGlu5 is involved in the effects of E2 in the regulation of behavior and physiology (Grove-Strawser et al., 2010;Peterson et al., 2015;Al-Sweidi et al., 2016). E2 is reported to mediate dendritic spine plasticity in the NAc through activation of mGlu5, evaluated via DiI labeling and confocal microscopy . The results of Peterson et al. (2015) suggest E2's role in mediating neuronal plasticity in the NAc via mGlu5 is important for E2's effect in drug addiction. Another study shows E2 facilitates cocaine selfadministration in OVX rats and mGlu5 activation is essential for this effect (Martinez et al., 2016). The study also demonstrates direct activation of mGlu5 is insufficient to mimic the effect of E2 in cocaine self-administration, suggesting that E2 receptors possibly need to be activated simultaneously to have the effect. Together, these findings are consistent with the results of the present study that both E2 receptors and mGlu5 s are necessary for E2's potentiation in DA release in DLS. It will be important to extend these results by examining the involvement of mGlu5 in other E2-mediated behaviors.
Although it is clear that both E2 receptors and mGlu5 are required for the estradiol evoked DA release from the DA terminals, our study does not show whether estradiol directly acts on or whether the two receptors are on the DA neurons. In fact, studies suggest estradiol activates E2 receptors coupled with mGlu5S on MSNs, which then modulates the release of GABA to influence DA terminals (Schultz et al., 2009). E2 receptors can be anchored to plasma membrane via caveolin protein, which then allow them to functionally couple with mGluRs (Yoest et al., 2018). The authors propose that E2 and mGlu receptors collaboratively act on MSNs in the DLS to modulate DA release from DA neuronal terminals. It is also possible that E2 acts on other interneurons (such as cholinergic neurons) to modulate DA release in the DLS, or influences glutamate release on cortical afferents.
Our data demonstrated marked increase of DA release in DLS following AMPH injections. This effect has been reported both in vivo and in vitro in our previous studies (Becker and Ramirez, 1981;Xiao and Becker, 1998;Becker and Rudick, 1999). Because of unknown vendor/ batch effects, different magnitudes of overall increase in DA concentrations following AMPH administration were observed in Experiments 1 and 3 (rats from Harlan Laboratories) than in Experiments 2 and 4 (rats from Charles River Laboratories).

Conclusion
The present study demonstrates that E2 directly potentiates the AMPH-induced increase in DA in the DLS. The effects of E2 are mediated by E2 receptors and can be blocked by an mGlu5 antagonist. These results provide important information on the neural mechanism through which E2 may contribute to sex differences in behaviors such as, but not limited to, addictive behavior. Our data also suggest that targeting mGlu receptors could be a potential treatment for E2 related disorders in female individuals.