Estradiol enhances ethanol stimulation of ventral tegmental area dopamine neuron firing without limiting ethanol inhibition onto those neurons

Females can progress to alcohol and other substance use disorders more quickly than males. The ovarian hormone 17β-estradiol (E2) contributes to sex differences observed in drug use and abuse and may be a principal driver of these differences. However, it is not entirely clear how E2 acts to affect processing of ethanol reward, and several brain regions and mechanisms are implicated. We sought to clarify the role of E2 in modulating the response of ventral tegmental area dopamine neurons to ethanol. To this end, we recorded spontaneous action potentials and inhibitory post synaptic currents from dopaminergic neurons in acute horizontal brain slices from ovariectomized (OVX) dopamine neuron reporter mice (Pitx3-eGFP) treated with either vehicle (VEH) or E2. On the basis of prior work, we hypothesized that E2 administration would cause dopamine cells from OVX+E2 animals to show a more substantial ethanol-induced increase in firing rate compared to control animals. Our data confirmed that ethanol stimulation of the firing rate of dopamine neurons from OVX+E2 mice was greater than that of OVX+VEH animals. Further, we hypothesized that the firing rate increase would be accompanied by a concomitant decrease in ethanol stimulated inhibition onto those same neurons. We found that although ethanol caused the expected increase in GABAA receptor-mediated synaptic inhibition in both groups, there was no difference in this response between OVX+E2 and OVX+VEH animals. Our findings lend additional support for the ability of E2 to enhance ventral tegmental area dopamine neuron responses to ethanol and suggest that this effect is not mediated by an E2-elicited suppression of synaptic inhibition.

152 Moreover, it was demonstrated that GABA A receptor antagonism potentiated ethanol stimulation 153 of firing [33] or attenuated ethanol inhibition of firing [32], indicating that a concurrent increase in 154 GABA release elicited by ethanol can limit the excitation of, or inhibit, dopamine neuron firing 155 activity. Therefore, we speculated that E2 might facilitate ethanol stimulation of firing via MISS 156 activity that suppresses GABA transmission onto dopamine neurons.

157
In summary, prior work indicates that E2 can act to promote drinking behavior in females 158 via facilitation of dopamine neuron responses to ethanol. In general, it is known that several 159 processes converge onto dopamine neurons and combine to mediate ethanol's effect on firing 160 rate, but the specific mechanism(s) by which E2 acts to facilitate ethanol excitation of firing are 161 not known. Here, we focus on suspected MISS elicited by E2, which we hypothesized would 162 suppress GABA signaling and thereby disinhibit dopamine neurons. We used ovariectomy (OVX) 163 followed by injections of E2 or vehicle (VEH), rather than using freely cycling animals, in order to 164 directly probe the role of estradiol. Firing rate and synaptic GABA responses to ethanol were 165 recorded from VTA dopamine neurons using ex vivo brain slice electrophysiology. Although we 166 replicated the prior finding of E2-potentiatation of ethanol stimulation of VTA dopamine neuron 167 firing activity, we did not observe any effects of E2 on GABA transmission in these neurons.

277
For confirmation of the nature of the spontaneous IPSCs (sIPSC), two recordings were 278 performed in which picrotoxin (50 M, Sigma) was added at the end of the wash period. In both 279 analyzed cells, the application of picrotoxin eliminated ≥99% of sIPSCs, indicating that these 280 events are predominantly GABA A -R mediated.

282
All voltage and current responses were acquired using CV-7B headstage with MultiClamp 283 700B amplifier, filtered at 1 kHz (firing rate) or 2 kHz (IPSCs), and digitized at 10 kHz via a Digidata 284 1440A interface board using pClamp 10 (Axon Instruments). All recordings were spontaneous 285 events, whether in cell-attached mode for firing rate experiments, or whole cell for the IPSC 286 experiments. IPSCs were monitored with a −60 mV holding potential. For the firing rate recordings, 287 after a loose seal (5-25 M) was formed, 10 minutes of stable baseline was acquired, at which 288 point the first concentration of ethanol was washed on to the slice. The ethanol-containing 289 solutions were freshly prepared in the recording aCSF using 95% ethanol (Pharmco-Aaper) to 290 final concentrations of 40, 80, and 120 mM. After the baseline period, each concentration was 291 successively applied for 6 minutes, after which a wash period of at least 6 and up to 20 minutes 292 was recorded. The IPSC recordings were performed as described above, except that after forming 293 a tight seal (> 1 G), a current clamp recording of 2 to 3 minutes was acquired (to monitor the 294 firing rate) after which the cell was broken into by application of slight negative pressure.
295 Immediately after break in, I h was measured by administering a 1.5 s hyperpolarizing pulse from 296 −60 mV to −120 mV and measuring the amplitude of the sag current at steady state. Following, 297 the baseline spontaneous IPSC recording began and proceeded as with the firing rate, with the 298 wash period consisting of at least 6 minutes. Spontaneous acquisition was performed in sweeps 299 of 9.375 s repeated every 10 s, with a 0.625 s period for membrane test between sweeps. Once 300 any ethanol had been applied to the slice, no additional cells in that slice were used for recordings.
301 For all whole cell recordings, series resistance was monitored throughout the experiment and we 302 rejected those cells for which the variability was greater than 30%, measuring from the time at 303 which ethanol reached the bath. Events were detected using either the Threshold Search (AP 304 firing) or Template Search (sIPSCs) features in ClampFit (Axon Instruments); sIPSC events 305 smaller than 10 pA in amplitude were excluded from analysis. Search results (event times, and 306 peak amplitudes for sIPSCs) were transferred into Microsoft Excel and examined for quality 307 control by inspecting anomalous interevent intervals to identify and remove spurious events 308 before being processed in MATLAB (Mathworks) to produce average frequency and amplitude 309 per sweep. These were then used to derive the one-minute bin averaged values. The 310 experimenter was blinded to the treatment that the animal received from injections throughout the 311 acquisition and analysis in 90% of cases; there were 3 animals total who comprised part of the 312 final analysis (2 E2 and 1 VEH) whose treatments were known to the experimenter throughout 313 the experimentation and analysis.

315
Baseline comparisons were planned via t-test. Data were tested for normality. In some 316 cases, the baseline data were non-normally distributed, or the variances were not equal, and non-317 parametric tests or Welch's t-test were used, and are indicated, where appropriate. Group 318 summary data show mean ± SEM for normally distributed data; median and interquartile ranges 319 for non-normally distributed data. Time course data were expressed as percent change from the We first verified that we could replicate the effect of E2 described in [20] given several 335 differences in our electrophysiology methodology and the line of mice used. As described in more 336 detail in the Materials and methods section, Pitx3-eGFP mice underwent an OVX surgery, 337 followed by a recovery period of 1-4 weeks. For two days prior to the recording day, mice received 338 daily injections of E2 (0.2 µg) or VEH, with a larger dose (1 µg) of E2 administered on the third 339 day (Fig 1A). Beginning 60-75 minutes after the third injection of E2 or VEH, brain slices 340 containing the VTA were prepared for electrophysiological recordings and action potential firing

369
Next, we tested our hypothesis that the firing rate differences described above were the 370 result of a different GABA response to ethanol in cells from E2-treated mice compared to VEH-371 treated. To examine this question, we performed whole-cell recordings in VTA dopamine neurons 372 and measured spontaneous GABA A receptor-mediated IPSCs (Fig 3A). The experimental design 373 was similar to that of the firing rate experiment described above, but with added steps as the 381 When analyzing IPSCs, as with the firing rate, there were no significant differences between 382 neurons from E2-and VEH-treated animals at baseline (Fig 3 B  396 This indicates an increased inhibitory drive in response to ethanol, regardless of in vivo treatment 397 condition, although this effect on GABA signaling appears to be mainly driven via the increased 398 frequency. Therefore, we did not observe evidence that E2 acts to potentiate the dopamine 399 neuron firing rate response to ethanol through modulation of the response of GABA afferents to 400 ethanol.

528
Finally, while the present work provides further support for the idea that E2 may promote 529 drinking behavior in females by influencing dopamine neuron responses to ethanol, it remains to 530 be determined what roles E2 and estrogen receptors play in male mice. Vandegrift et al. [23] 531 found that the same intra-VTA knockout of ERα that limited female ethanol drinking did not affect 532 drinking by male mice, but it is not known whether E2 or estrogen receptors modulate dopamine 533 neuron responses to ethanol in males, or whether E2 promotes drinking in males. Estradiol has 534 in fact been shown to induce potentiation of hippocampal glutamatergic synapses in both sexes, 535 albeit with different molecular signaling requirements [83]. Oberlander and Woolley [86] found 536 that although different ER subtypes were engaged in males compared to females, both sexes 537 experienced similar effects of E2 to increase pre-and post-synaptic glutamatergic signaling in the 538 hippocampus. Notwithstanding differences in gonadal synthesis and circulating levels in the blood 539 when compared to females, males can locally synthesize E2 in the brain from testosterone via 540 aromatase [87,88]. This raises the possibility that E2 could influence VTA dopamine neuron 541 excitation by ethanol and promote drinking in male mice, but via an ER-independent mechanism. 542 Recently, more researchers are starting to use the four-core genotype mice to probe genetic and 543 gonadal components independently [89], which may allow for a clearer understanding of how 544 circulating hormone levels interact with structural and organizational differences to express the 545 full complement of sex differences. The increased use of this model may also help to illuminate 546 how E2 and ERs function differently in males and females and sometimes, as indicated above, 547 use different mechanisms to achieve the same end result.

548
Disentangling the mechanisms that underlie mesolimbic responses to ethanol and 549 understanding how they may work together to produce behavioral output is a major task of the 550 alcohol field. Our data did not support our hypothesis regarding the underlying mechanisms that 551 could explain the increased excitation by ethanol following E2 treatment of OVX females, but