Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 17.
Published in final edited form as: Neuropharmacology. 2021 Mar 17;189:108527. doi: 10.1016/j.neuropharm.2021.108527

Repeated binge ethanol drinking enhances electrical activity of central amygdala corticotropin releasing factor neurons in vivo

Sonia Aroni 1, Rosa A M Marino 1, Kasey S Girven 1,2, James M Irving 1, Joseph F Cheer 1,2, Dennis R Sparta 1,2
PMCID: PMC8928451  NIHMSID: NIHMS1780919  PMID: 33741403

1. Introduction

Alcohol use disorder (AUD) is a progressive and persistent disease characterized by frequent cycles of binge/intoxication, withdrawal/negative affect and preoccupation/anticipation (“craving”) (Volkow et al., 2016). Alcohol-related health costs in the United States over 200 billion dollars annually. In particular, binge drinking accounts for 75% of the societal cost attributed to AUD (Sacks et al., 2015). Moreover, binge alcohol consumption is also linked to the progressive dysfunction of multiple organs and is considered a first step in the development of AUD (Bonomo et al., 2004). It is well known that behavioral consequences of binge drinking are mediated by long-term alterations in brain stress circuits residing within the “extended amygdala”, an interconnected macrostructure consisting of the central amygdala (CeA), substantia innominata, the nucleus accumbens shell and the bed nucleus of the stria terminalis (BNST) involved in the modulation of reward and emotion (Koob and Le Moal, 2008; Pleil et al., 2015; Rinker et al., 2017).

The CeA is the primary output nucleus of the amygdala and a critical regulator of stress and reward. It is comprised of predominantly GABAergic projection neurons, that also express many different neuropeptides including neuropeptide Y (NPY), dynorphin and corticotropin-releasing factor (CRF) (Rivier et al., 1983; Sah et al., 2003; Sah and Lopez De Armentia, 2003). Importantly CeA CRF neurotransmission has been implicated in regard to excessive binge-like ethanol intake. In particular, increased CRF immunoreactivity within the CeA is observed after excessive ethanol administration and withdrawal (Finn et al., 2007). Furthermore, administration of CRF antagonists within the CeA reduces excessive drinking in rodents (Finn et al., 2007; Funk et al., 2006; Lowery-Gionta et al., 2012). Moreover, de Guglielmo and coworkers show that optical inhibition of CeA CRF neurons reduced excessive ethanol drinking in dependent rats (de Guglielmo et al., 2019). The above-mentioned data provides strong, presumptive evidence that binge ethanol drinking can regulate CeA CRF activity. However, little is known about the precise mechanisms of how acute and repeated binge ethanol consumption engages and alters CeA CRF neuronal activity. For this reason, here, we used optogenetics coupled with in vivo electrophysiology to examine the activity of CeA CRF neurons during acute and chronic binge ethanol exposure. We found that CeA CRF neurons show heterogeneity in response to a lick of ethanol and higher electrical activity during ethanol consumption when compared to CeA non-CRF neurons. Additionally, one group of CRF neurons that were transiently excited immediately before a lick of ethanol (pre-lick activated) increased its activity after repeated ethanol consumption. Interestingly, this effect was not observed in a water and sucrose consumption drinking paradigm.

2. Materials and methods

2.1. Animals

Male and female Crhtm1(cre)Zjh mice (N = 12 and N = 5, respectively) expressing Cre recombinase under the corticotropin-releasing hormone gene (Crh) promotor were obtained from the Jackson Laboratory. They shall be referred to as corticotropin-releasing factor – cre (CRF-Cre) mice herein. All mice were heterozygous and generated by mating a male homozygous CRF-Cre mouse with a female wildtype C57. All mice were placed in a reverse light/dark cycle 3 weeks before any manipulation and maintained there until the end of the experimental procedures. All experiments and surgical procedures were conducted in accordance with the United States National Research Council Guide for the Care and Use of Laboratory Animals and were approved by the University of Maryland School of Medicine Institutional Animal Care and Use Committee (IACUC).

2.2. Ethanol consumption protocol

Mice (N = 13) underwent several weeks of a modified version of the drinking-in-the-dark (DID) protocol (Rhodes et al., 2007, 2005). Briefly, 3 hours into the dark phase of their light-cycle, mice were placed into a chamber with a water bottle filled with 20% (v/v) ethanol (in tap water) and food. Mice remained in the chamber for 4 hours and the number of licks and weight of ethanol consumed were analyzed. Mice were run for 4 consecutive days followed by 3 days of abstinence (Fig 1 C). Each mouse had at least 5 cycles of binge drinking (20 total binge drinking days over 5 weeks). Licks were quantified by the current passed when a mouse’s tongue completed the circuit between the metal floor grid and metal sipper of the ethanol bottles (Fig 1 C). In a separate cohort of mice (N = 4water/sucrose; mice were recorded first with water, then with sucrose) we used the DID procedure, but instead of ethanol, we examined water and 2% (v/v) sucrose consumption to analyze CRF activity during natural reward consumption. Mice were run for two DID cycles of 4 consecutive days with 3 days of abstinence using water only (2 weeks, 8 sessions). Then, during their third week of water sessions, when stable responding was observed, animals underwent recording during a 4 hour water DID session. Next, the day after, we proceed with sucrose sessions recorded only from ethanol naïve animals during their first exposure to sucrose in a 4 hour sucrose DID session.

Figure 1.

Figure 1.

Open in a new tab

Optical identification of CRF neurons in vivo. A) Scaled diagram of optical-fiber coupled microarray implanted into the central amygdala (CeA) of CRF-Cre mice (on the left horizontal section), and local CeA injection of AAV-DIO-Channelrhodopsin2-EYFP virus (on the right vertical section), whose expression is showed in the image (inset). B, top) Peristimulus time histogram and scatter plot showing identified CRF neurons. Units were classified as CRF only if they fired within 10 ms of the onset of a 4 ms-long light pulse (left), and the light-evoked waveforms had an R2 >0.9 compared to non-light evoked waveforms (right). B, bottom) The pie chart shows that out of 149 total units, 59 were identified as putative CRF neurons, 75 non-responsive to light, and a small population excited (4) or inhibited (11). Due to the low n for these light responses, we focused on CRF units vs non-light-responsive units, herein “Non-CRF”. C, top) Schematic diagram of the DID protocol (left) and the lickometer set-up comprising an analog-digital converter connected to the tip of the bottle and the cage metal floor. Single licks are detected as current transients as shown in the trace (right). C, bottom) Weekly average of ethanol licks in all mice that were run through our in vivo electrophysiology recording sessions (N = 13). D) Graphs show that overall, CRF units (n = 37) had a significant shift in their correlation values for firing rate vs cumulative licks from early to late sessions (early CRF (n = 20), late CRF (n = 17); whereas non-CRF units did not (early Non-CRF (n = 41), late Non-CRF (n = 34). K-S test; *p < 0.05. E) Electrophysiological characterization of CRF vs non-CRF neurons. Graphs showing firing rate, coefficient of variation, % of spikes in bursts (%SiB), burst duration, intraburst frequency, burst rate and # of spikes per burst. Top right, representative electrophysiological traces from Non-CRF (top) and CRF (bottom) neurons. M-W test; **p < 0.01, *p < 0.05.

Error bars are standard error of the mean (SEM). La = lateral amygdaloid nucleus. ec = external capsule. CeC = lateral capsular part of the central nucleus of the amygdala (CeA). CeL, CeM= lateral and medial subdivision of the CeA, respectively.

2.3. In vivo electrophysiology

2.3.1. Optical-fiber-coupled microelectrode arrays (optoarrays)

Custom-made microelectrode arrays (MEA) were obtained from Innovative Neurophysiology (Durham, NC). The MEAs had 16 × 35 μm tungsten electrodes in a U-shape around a central pore for optical fibers positioning (Fig 1 A). The MEA connector was offset from the center in a chair-like configuration to allow for the optical fiber connection.

Implantable optical fibers were fashioned in-house as previously described (Sparta et al., 2012). Briefly, 0.48 NA 200/230 μm optical fiber was stripped and affixed with a ceramic ferrule (235 μm ID, as above) and polished. Percent transmittance (%T) was calculated using an optical power meter (PM20A, ThorLabs) and fibers with < 75 %T were discarded. Optical fibers were attached to the MEA approximately 500 μm (300–800 μm) dorsal to the tips of the electrodes, at a slight angle toward the electrodes, using dental cement (Industrial Grade Grip Cement, powder #675571, liquid #675572, Dentsply, York, OA).

2.3.2. Survival surgeries

At 8–12 weeks of age, CRF-Cre mice (N = 17) were anesthetized with isoflurane via an E-Z Anesthesia brand vaporizer (E-Z Anesthesia, Palmer, PA). A craniotomy above the CeA was centered at −1.1 mm A/P, +/−2.95 mm L/M (from bregma) and was extended into a concentric square +/−0.25 mm A/P and L/M. An adeno-associated viral vector (serotype 5) containing a double inverted open reading frame with the Channelrhodopsin2 (ChR2) ChR2(h134R)-EYFP construct (AAV-EF1a-DIO-hChR2(H134R)-EYFP-WRE-Pa) was purchased from the University of North Carolina’s Vector Core, with permission from Karl Deisseroth. The virus was infused into the CeA via a Hamilton 7001 1-μL syringe (Hamilton Company, Reno, NV), which was inserted into a Micro4 microinfusion pump (UMC4 controller, UMP3 pump; World Precision Instruments, Sarasota, FL). The needle was slowly lowered −4.85 mm D/V (from brain surface) into the brain and 500 nl of the virus were infused at 100 nl/min. After 5 minutes, the needle was slowly removed and the optoarray was implanted. Optoarray was secured to the skull using anchor screws (MF-5182 bone screws; Bioanalytical Systems, Inc, West Lafayette, IN), with the ground wire wrapped around one screw and inserted 2 mm into the cerebellum. The array was then encased in a dental cement headcap.

Following surgery, the mice were singly housed and recovered for two weeks before starting the drinking protocol. At the completion of the experiment, mice were deeply anesthetized, and we passed high current through the headcap to check for placements. Then, they were euthanized, and brains were processed for histology.

2.3.3. Recording hardware

In vivo electrophysiology recordings were performed using an Omniplex-DHP system from Plexon, Inc (Dallas, TX), with a motorized commutator from NeuroTek (Toronto, Ontario, Canada). The wideband signal was acquired at 40 kHz, with an analog high pass filter of 7.5 Hz, and a 7.5 kHz low pass filter before the signal was digitized at the headstage. The continuous spike data were extracted from the wideband signal by applying a 4 pole 500 Hz butterworth high pass filter and a 4 pole 3000 Hz butterworth low pass filter. Common median referencing was used for all recording sessions.

2.3.4. Optical identification of genetically-defined neurons (phototagging)

Optoarray mice were tethered to a headstage and mini patch cord with ferrules on both ends plugged onto the fiber implant of the optoarray and bound to the headstage cables. The other end of the mini patch cord remained unplugged for the duration of the 4 hour drinking session. After the DID session, the mini patch chord attached to the mouse was then connected to the laser-attached patch chord and the phototagging procedure began. Phototagging was done in accordance with previous reports (Cohen et al., 2012). Briefly, at the end of the recording session, mice received a series of stimulation trains containing 20 pulses (4 ms pulse-length) at 1, 5, 10, and 20 Hz in a randomized order. Light intensity was adjusted per mouse based upon the measured transmittance of the implanted fiber, such that 20 mW of 473 nm light were delivered into the brain.

We defined units as putative CRF neurons only when all criteria for phototagging were fulfilled: a significant increase in firing rate within 10 ms after the beginning of a light pulse, when compared to the 10 ms prior to the event (as identified by a Wilcoxon signed rank test on the binned spike/sec data); the light-evoked waveforms had to have a cross-correlational coefficient R2 > 0.9 when compared to the naturally occurring waveforms (Cohen et al., 2012). Units were classified as light-excited, light-inhibited or no response, using Wilcoxon signed rank tests on time bins before and after the event (Fig 1B). We classified any unit that did not fulfill these requirements as non CRF, however this does not mean that the neuron was non CRF based on many factors such as light permanence, viral penetration, etc. We recorded from 59 putative CRF neurons and 75 putative non CRF units. We also observed 4 neurons that were excited beyond our 10 ms threshold and 11 units that were inhibited by light. Due to the low number of these neurons, and the fact that they may be influenced by a polysynaptic connection with CRF units, they were excluded from our analyses.

2.3.5. Lick response type classification

To classify the encoding pattern for each neuron in relation to licking, we calculated peri-event raster plots and histograms for all units, from −100 ms to +100 ms, using 5 ms bins, with licks centered at time 0 s. We analyzed the average firing rates in 50 ms periods before and after the lick event (Baseline: −100 ms to −50 ms; Pre-Lick: −50 ms to 0 ms; Post-Lick:0 ms to +50 ms). We performed Wilcoxon rank-sign tests on these firing rates and classified the response types based on which of these comparisons were significantly different and the sign of the difference between averages. We identified four types of lick-responses: no response (NR), lick-inhibited (I), pre-lick activated (P), and lick-excited (E; Fig 2 A; see Table 1 for detailed classification criteria). For all neurons, spike data was aligned to every lick. Using these criteria, we found 27 CRF-NR, 2 CRF-I, 24 CRF-P, and 6 CRF-E units (Fig 2 B). Due to the low number of neurons in the CRF-E and CRF-I groups, our main comparisons were only between the CRF-P and CRF-NR groups.

Figure 2.

Figure 2.

Open in a new tab

CRF Neurons encode licking-behavior. A) Perievent Raster plots for representative units for each lick-response type. Units were classified into 4 lick-response types (no response, inhibited, pre-lick activated, and excited), based upon changes in firing rates during 3 time periods: baseline (−100 ms to −50 ms before licks) vs. pre-lick (−50 ms to 0 ms), post-lick (0 ms to +50 ms). Wilcoxon signed-rank tests were performed on pairs of these time periods to determine if there were significant changes in firing rates from baseline to pre-lick and from pre-lick to post-lick. B) The pie chart indicates that out of our 59 CRF units, 27 showed no response (CRF-NR), 2 were lick-inhibited (CRF-I), 24 were pre-lick activated (CRF-P), and 6 were lick-excited (CRF-E). Due to the low number of CRF-E and CRF-I units, we focused on the two major response classes, CRF-NR and CRF-P. C) Graph show electrophysiological parameters such as firing rate, coefficient of variation, %SiB, burst duration, intraburst frequency, burst rate, and # of spikes per burst, of CRF-P, CRF-NR, and non-CRF NR (included as a control) neurons. CRF-P units had a higher firing rate, %SiB, burst duration, burst rate, and # of spikes per burst than CRF-NR and non-CRF-NR units. Furthermore, CRF-NR cells showed a lower coefficient of variation when compared to non-CRF-NR. K-W test; ***p < 0.001, **p < 0.01, *p < 0.05.

Table 1.

Lick classification criteria.

TYPE Pre-Lick - Baseline Post-Lick - Pre-Lick
No Response (NR) No change No change
Inhibited (I) No change Sig. decrease
Pre-lick activated (P) Sig. increase Sig. decrease
Excited (E) No change Sig. increase
Open in a new tab

2.3.6. Bouts analysis

A bout of ethanol was defined as at least 20 licks, with an interlick-interval shorter than 60 s. (Barkley-Levenson and Crabbe, 2015).

2.3.7. Burst-firing

We identified the beginning of a burst as a series of at least two spikes with an interspike-interval less than 50 ms and its end when the interspike-interval exceeded 100 ms. Based upon these parameters, we analized the % of spikes in bursts (%SiB), burst duration, intraburst frequency, burst rate, and # of spikes per burst. Furthermore, we measured the coefficient of variation (CV), as index of firing regularity.

2.3.8. Data processing and analysis

Plexon OfflineSorter 4 was used to process and separate identified units from each recording. Identified neurons recorded from the same electrode were considered different if the waveforms were significantly separated in 3D principal component space as tested by multivariate ANOVA (p < 0.05), with L-ratios < 0.05 for each cluster. Units were classified based upon their responses to light stimulation, and licks.

2.3.9. Outlier time bin removal and replacement

In order to eliminate any large fluctuations in noise that may have contaminated individual time bins out of the 4 hour recordings, we first calculated raw firing rate in 5 minute bins. We then used the Matlab function filloutliers to identify and replace outlier time bins, as identified using a sliding window 6 time-bins-wide (30 mins) to calculate the median, and any bins that were more than 3 scaled deviations away from the median were identified as outliers and substituted with a linear interpolation of the time bins surrounding the outlier. This was done prior to the baseline subtracted firing rate Z-score calculation. The same outlier time bins were also used to remove outlier time bins for burst firing metrics.

2.4. Statistics

All data were presented as means ± SEM and analyzed using GraphPad Prism 7 and Matlab 2017. We collapsed male and female data because we did not see any statistical difference among groups. Comparisons between CRF and non-CRF units were analyzed using Mann-Whitney U tests (M-W test). Comparisons between 3 unit types were analyzed using Kruskal-Wallis tests (K-W) (Benhamou and Cohen, 2014) with post-hoc Dunn’s multiple comparison tests. All comparisons across hours were done using two-way repeated measures ANOVA with post-hoc Sidak’s multiple comparison tests for between-group comparisons and post-hoc Tukey’s multiple comparison tests. Comparisons between early and late ethanol session correlations were analyzed using Kolmogorov-Smirnov tests (K-S test).

3. Results

3.1. Identification and electrophysiological profile of CRF vs non-CRF neurons

We recorded electrical activity of CeA neurons that were classified as putative CRF units only if they fired within 10 ms of the onset of a 4 ms-long light pulse, and the light-evoked waveforms had an R2 > 0.9 compared to non-light evoked waveforms (Fig 1 A, B, S1). Using these criteria, we analyzed and recorded from 149 total neurons, of which 59 were classified as putative CRF, 75 non-light-responsive, 4 light-excited (we observed spiking approx. 10 ms after optical stimulation, hence we did not include these cells as CRF positive based on our criteria), and 11 light-inhibited (Fig 1 B). Herein, we refer to non-light-responsive units as non-CRF neurons and optically identified units as CRF neurons. To ascertain ethanol consumption, we recorded the total number of ethanol licks as our metric while each mouse went through in vivo electrophysiological recordings, every DID day and found that CRF-Cre mice (N = 13) showed a modest increase in ethanol licks over each week, leveling off at approximately 600 licks from binge drinking week 2 – 5 (Fig 1 C). Since the tether for in vivo electrophysiology and optogenetics could impede ethanol licking, in a separate cohort of untethered CRF-Cre mice (N = 12), we analyzed ethanol licks during a 3 week DID protocol. We found that CRF-Cre mice steadily escalated their licks for ethanol over repeated DID sessions compared to water (Fig S2 A). Importantly, the averaged total number of ethanol licks were equivalent to mice run in our in vivo electrophysiology setup (Fig 1 C).

We next determined whether CeA CRF neuronal activity was correlated with ethanol licks by comparing CRF to non-CRF neurons during early ethanol sessions (session # 1–8) versus late sessions (session # 17+). We found that CRF units significantly increased the strength of their correlation in late sessions compared to early sessions (K-S test; D = 0.4804, p = 0.0134) compared to non-CRF neurons (K-S test; D = 0.2267, p = 0.2951; Fig 1 D). We also correlated the activity of CRF units to every ethanol lick but did not find any significant correlation during the entire ethanol session as well as during the last hour (Fig S2 C, D).

Furthermore, throughout all of our ethanol recordings, we observed that CRF neurons (n = 59) had significantly higher firing rates (M-W test; U = 1589, p = 0.005), burst rates (M-W test; U = 1770, p = 0.0469), and smaller coefficients of variation (an index of firing regularity; M-W test; U = 1488, p = 0.0011), compared to non-CRF units (n = 75) with no differences in percentage of spikes in bursts (%SiB; M-W test; U = 2151, p = 0.7827), burst duration (M-W test; U = 2063, p = 0.4901), intraburst frequency (M-W test; U = 1829, p = 0.0859), and # of spikes per burst (M-W test; U = 2199, p = 0.9510; Fig 1 E).

3.2. CeA-CRF neurons have distinct responses to a lick of ethanol

To determine whether CeA CRF neurons encoded ethanol consumption, we analyzed peri-event histograms for each unit time-locked to individual licks (Fig 2 A; see also General Methods for lick-response-type classification). We identified four types of CRF units in response to an ethanol lick: no response (CRF-NR; n = 27), lick-inhibited (CRF-I; n = 2), pre-lick activated (CRF-P; n = 24), and lick-excited (CRF-E; n = 6; Fig 2 A, B). Due to the small number of neurons within both the CRF-I and CRF-E groups, we focused our analyses on the two most prevalent types, CRF-P and CRF-NR, which represented 86.4% of all CRF units recorded (Fig 2 B). Moreover, we analyzed the response to an ethanol lick for non-CRF neurons and we found the same pattern of responses (non-CRF-NR; n = 47; non-CRF-I; n = 5; non-CRF-P; n = 14; non-CRF-E; n = 9; Fig S2 B) but with a lower % of non-CRF-P neurons when compared to CRF-P cells (18.67% and 40.68%, respectively). We also included non-CRF non-lick responsive (non-CRF-NR; n = 47) as a control group. Our analysis revealed that CRF-P units showed a significantly higher firing rate (K-W test; H = 16.43, p = 0.0003), %SiB (K-W; H = 12.55, p = 0.0019), burst duration (K-W test; H = 14.09, p = 0.0009), burst rate (K-W test; H = 16.80, p = 0.0002), and # of spikes per burst (K-W test; H = 11.45, p = 0.0033), compared to both CRF-NR or non-CRF-NR units, suggesting that CRF-P cells were more active during ethanol consumption compared to the other groups. Additionally, CRF-NR neurons exhibited a lower coefficient of variation when compared to non-CRF-NR units (K-W test; H = 9.492, p = 0.0087) with no differences among groups in intraburst frequency (K-W test; H = 5.880, p = 0.0529; Fig 2 C).

3.3. CRF-P neurons increase firing activity throughout drinking sessions

Due to the heterogeneity of CeA CRF neuronal activity in response to a lick of ethanol during the entire DID paradigm (Fig 2), we further analyzed the temporal firing properties of every group during each 4 hour binge ethanol session in 5 min bins. We compared the change in firing rate within-sessions between all units using Z-scores normalized to the first 30 minutes as our baseline (Fig 3 A). We found a significant difference among CRF types, with CRF-P neurons displaying higher firing rates over time during the session compared to CRF-NR cells (two way ANOVA; main effect of CRF type: F(1, 49) = 7.957, p = 0.0069; hour: F(3, 147) = 2.545, p = 0.0583; CRF type by hour interaction: F(3, 147) = 3.924, p = 0.0099; Fig 3 B, top panel), that was significantly different during hour 3 (Sidak’s; p = 0.078) and 4 (Sidak’s; p = 0.002; Fig 3 B, top right). Moreover, CRF-P units steadily increased their firing rate over the session that become significant at hours 3 and 4 (Hour 3: Tukey’s; p=0.0324; Hour 4: Tukey’s; p=0.003; Fig 3 B, top panel). This effect was not observed in the CRF-NR group (Fig 3 B, top panel). Next, we analyzed the %SiB over time for both CRF-P and CRF-NP groups (Fig 3 B, bottom panel). We observed that CRF-P neurons exhibited higher burst activity compared to CRF-NP cells (two way ANOVA; main effect of CRF type: F(1, 49) = 21.60, p < 0.0001; hour: F(3, 147) = 1.775, p = 0.1545; CRF type by hour interaction: F(3, 147) = 0.2058, p = 0.8922; Fig 3 B, bottom right). However, within each group we did not observe any change in %SiB over time (Fig 3 B, bottom panel). Thus, CRF-P neurons change firing rate dynamically during ethanol sessions, while bursting activity remains consistently higher throughout ethanol sessions.

Figure 3.

Figure 3.

Open in a new tab

CRF-P Neurons increase firing activity during ethanol sessions, with heterogeneous sub-types. A) Normalized Firing Rate Z-Scores were calculated using the first 30 mins as the baseline period for measuring the mean and std used to calculate Z-scores for the full-session. Z-scores are shown in color, with each horizontal line is one unit’s activity for the 4-hour drinking session. Units are grouped by lick-response type and then ordered from top-bottom by rate change (hour 4 – hour 1). B) Left: Average normalized firing rates (top) and %SiB (bottom) calculated in 5 min bins. Right: Hourly averages of the normalized firing rates (top) and %SiB (bottom) used for statistical analysis. Right, top: CRF-P units had a higher firing rate vs CRF-NR units (two way ANOVA; **p < 0.01). Post-hoc tests show that CRF-P units increased throughout the session with significantly higher rates by hours 3 and 4 (Tukey’s tests; ααp < 0.01, αp < 0.05), whereas CRF-NR did not change across hours. Right, bottom: CRF-P units also had higher % of spikes in bursts but did not change over the session. (two way ANOVA; ****p < 0.0001). C) Left: The average normalized firing of each CRF sub-types: CRF-NR(Δ+) (n=8), CRF-NR(Δ−) (n=19), CRF-P(Δ+) (n=9), CRF-P(Δ−) (n=14). Right: Pie charts and bar graphs show how units were sorted by change in firing rate (hour 4 – hour 1). CRF-NR and CRF-P were separated into two statistically different subtypes that increased (Δ+) and decreased (Δ−) firing rate respectively. Two way ANOVA; ****p < 0.0001, ***p < 0.001, *p < 0.05. D) Left, top: CRF-P(Δ−) and CRF-P(Δ+) firing rates were significantly for hours 3 and 4 (two way ANOVA; ****p < 0.0001). CRF-P(Δ+) increased throughout the sessions with hours 1 distinct from 3 and 4 (Tukey’s test; ¤¤p < 0.01, ¤p < 0.05), but CRF-P(Δ−) did not change. Left, bottom: CRF-P subtypes had a significantly different change in %SiB for hours 2–4 (two way ANOVA; ****p < 0.0001) with a significant hour by sub-type interaction CRF-P(Δ+) changed from hour 1 vs hour 3 (Tukey’s test; €p < 0.05), whereas CRF-P(Δ−) changed from hour 1 vs 2–4 (Tukey’s test; $ $ $p < 0.001, $ $p < 0.01, $p < 0.05). Right, top: CRF-NR(Δ+) and CRF-NR(Δ−) were significantly different for hours 2–4 (two way ANOVA; ****p < 0.0001, **p < 0.01). CRF-NR(Δ+) increased firing rate by hour 4 vs hours 1–3 (Tukey’s test; ####p < 0.0001). CRF-NR(Δ−) had lower firing rates in hours 2–4 vs hour 1 (Tukey’s test; ‡‡‡‡p < 0.0001, ‡‡p < 0.01). Right, bottom: While CRF-NR subtypes did not have a significant main effect of sub-type or hour there was a significant sub-type by hour interaction). Post-hoc tests show that CRF-NR(Δ−) units decreased %SiB from hour 1 to hours 3 and 4 (Tukey’s test; §§§§p < 0.0001, §p < 0.05) and CRF-NR(Δ+) units increased from hours 1 and 2 to hour 4 (Tukey’s test; †p < 0.05).

3.4. CRF lick-response types show heterogeneous changes in firing activity

When we analyzed the change in firing rates from hour 1 to hour 4 (Δ-rate) in CRF-P and CRF-NR groups, we found heterogeneity within populations based on this change (Fig 3 A–C). Hence, we split each CRF-NR and CRF-P groups into two sub-populations: CRF-NR(Δ+) (n = 8), CRF-P(Δ+) (n = 10; neurons that exhibited a significant increase in activity over time during the ethanol sessions) and CRF-NR(Δ−; n = 19), and CRF-P(Δ−) (n = 14; units that exhibited a significant decrease in activity over time during the ethanol sessions; two way ANOVA; ethanol response: F(1, 46) = 7.019, p = 0.0110; CRF type: F(1, 46) = 55.06, p < 0.0001; Fig 3 C). Within groups both CRF-NR(Δ)+ and CRF-P(Δ+) neuronal activity was significantly increased compared to CRF-NR(Δ−) (Sidak’s test; p = 0.0003) and CRF-P(Δ−) groups (Sidak’s test; p < 0.0001) respectively. Additionally, CRF-P(Δ+) units were significantly more active than CRF-NR(Δ+) cells overall (Sidak’s test; p = 0.0213; Fig 3 C, right). However, there was no significant difference between the firing rate of CRF-P(Δ−) and CRF-NR(Δ−) (Fig 3 C, right).

Next, we compared changes in activity between CRF-P(Δ+) and CRF-P(Δ−) and CRF-NR (Δ+) and CRF-NR(Δ−) units during each hour of the binge ethanol session, using average normalized firing rate and %SiB (Fig 3 D, E). We found significant differences in firing activity between CRF-P(Δ+) and CRF-P(Δ−) cells (two way ANOVA; CRF-P type: F(1,22) = 32.86, p < 0.0001; hour: F(3,66) = 6.777, p = 0.0005; interaction CRF-P type x hour F(3,66) = 10.65, p < 0.0001) at hours 3 and 4 (Sidak’s p < 0.0001) (Fig 3 D, top). Furthermore, within groups, we observed that CRF-P(Δ+) neurons increased steadily their firing rate that was robustly higher during hours 3 (Tukey’s p = 0.0002) and 4 (Tukey’s test; p < 0.0001), an effect which was not observed in the CRF-P(Δ−) group (Fig 3 D, top). Although CRF-P(Δ+) units dynamically changed frequency of discharge throughout the session, they showed only modest and inconsistent variations in %SiB. However, there was a significant difference in %SiB between both the CRF-P(Δ+) and CRF-P(Δ−) groups (two way ANOVA CRF-P type: F(1,22) = 22.73, p < 0.0001; hour: F(3,66) = 1.991, p = 0.1239; interaction CRF type x hour F(3,66) = 11.98, p < 0.0001) at hours 2 (Sidak’s p = 0.0105), 3 and 4 (Sidak’s p < 0.0001; Fig 3 D, bottom). Within the CRF-P(Δ+) group, only hour 3 was significantly higher compared to the first hour for %SiB (Tukey’s p = 0.0179), whereas hour 2 (Tukey’s p = 0.0127), 3 (Tukey’s p = 0.0044), and 4 (Tukey’s p < 0.0001) were significantly lower from hour 1 in the CRF-P(Δ−) group. Compared to CRF-P(Δ+) neurons, we only observed slight increases in cell activity within the CRF-NR(Δ+) group. We also found differences in frequency of discharge between CRF-NR(Δ+) and CRF-NR(Δ−) groups (two way ANOVA CRF-NR type: F(1, 26) = 34.41, p < 0.0001; hour: F(3,78) = 2.75, p = 0.0483; CRF NR type x hour interaction: F(3, 78) = 11.82, p < 0.0001) that was significant at hours 2 (Sidak’s p = 0.0012), 3 (Sidak’s p = 0.0067), and 4 (Sidak’s p<0.0001; Fig 3 E, top). Within the CRF-NR(Δ+) group we observed that only at hour 4 the %SiB was significantly higher than hour 1 (Tukey’s p < 0.0001), whereas at hour 2 (Tukey’s p = 0.0068), 3 (Tukey’s p = 0.0046), and 4 (Tukey’s p<0.0001) the %SiB significantly decreased compared to hour 1 in CRF-NR(Δ−) units, although this effect was merely modest (Fig 3 E, top). Finally, we only found mild changes in %SiB between the CRF-NR groups (two way ANOVA; CRF-NR subtype: (F(1, 25) = 3.071, p = 0.0920; hour: F(3, 75) = 0.9743, p = 0.4095; CRF-NR subtype x hour interaction: F(3, 75) = 7.535, p = 0.0002) that became significant between the groups at hour 4 (Sidak’s p = 0.0018). Within the CRF-NR(Δ+) group, only at hour 4 the %SiB was significantly higher compared to hour 1 (Tukey’s p = 0.0490), whereas at hour 3 (Tukey’s p = 0.0166) and 4 (Tukey’s p = 0.0008) the %SiB significantly decreased from hour 1 in the CRF-NR(Δ−) group (Fig 3 E, bottom).

Therefore, while CRF-P(Δ+) sub-group showed a substantial rise in their electrical activity over the course of ethanol drinking sessions, accompanied by a modest increase in burst firing, CRF-P(Δ−) neurons did not change their overall firing rate across the session, but significantly decreased their bursting activity (Fig 3 D). Conversely, CRF-NR(Δ+) units had slightly higher frequency of discharge and burst firing throughout the ethanol sessions, while CRF-NR(Δ−) showed a steady decrease in their firing and bursting activity (Fig 3 E).

3.5. CRF-P neurons increase firing and bursting activity over repeated drinking sessions

Finally, we examined whether the firing activity of CRF neurons changed over repeated sessions of ethanol consumption. When we compared CRF lick-response types, we found that CRF-P units had much higher firing rates in later sessions compared to CRF-NR cells (two way ANOVA; main effect of ethanol session: F(1, 38) = 11.34, p = 0.0017; main effect of CRF type: F(1, 38) = 13.85, p = 0.0006; ethanol Session x CRF type interaction: F(1,38)= 9.967, p=0.0031; Fig 4 A, left). Furthermore, CRF-P neurons showed significantly higher frequency of discharge in late sessions, when compared to early sessions (Tukey’s, p = 0.0005; Fig 4 A, left). On the other hand, CRF-NR cells did not display any adaptive change in their firing rates in later drinking sessions (Fig 4 A, left). A further analysis of the frequency of discharge across the 4 hour session, revealed also that CRF-P neurons in late sessions show a higher firing rate from hour 1 to hour 4, when compared to CFR-P units during early sessions, and both CRF-NR groups during early and late sessions (two way ANOVA; main effect of hours: F(3, 114) = 2.489, p = 0.0639; main effect of CRF type: F(3, 38) = 11.31, p < 0.0001; hour x CRF type interaction: F(9,114)= 4.484, p<0.0001; Fig 4 A, right). We also examined the %SiB in early vs late sessions and found similar results, with only CRF-P cells showing a substantial increase in burst firing during late sessions (two way ANOVA; main effect of ethanol session: F(1, 38) = 7.138, p= 0.011; main effect of CRF type: F(1, 38) = 12.44, p = 0.0011; hour x CRF type interaction: F(1,38)= 8.808, p= 0.0052; Fig 4 B, left), while CRF-NR units displayed no changes (Fig 4 B, left). According to firing rate results over the 4 hour sessions, CRF-P neurons in late sessions show an increased %SiB from hour 2 to hour 4, when compared to CFR-P units during early sessions, and both CRF-NR groups during early and late sessions (two way ANOVA; main effect of hours: F(3, 114) = 1.279, p = 0.2849; main effect of CRF type: F(3, 38) = 8.447, p = 0.0002; hour x CRF type interaction: F(9,114)= 0.9542, 0.4817) (Fig 4 B, right).

Figure 4.

Figure 4.

Open in a new tab

Changes in CRF activity over repeated ethanol sessions. A) Left, CRF-P increased raw firing rates after repeated ethanol sessions (two way ANOVA; main effect of ethanol session: ***p = 0.0005), which was absent in CRF-NR units (p = 0.9986). B) Left, Similarly, CRF-P units showed a significant increase in the %SiB after repeated ethanol sessions (two way ANOVA; **p = 0.0029) and CRF-NR did not (p = 0.9961). A, B) Right, CRF-P neurons in late sessions show a higher frequency of discharge from hour 1 to hour 4 (two way ANOVA; ****p < 0.0001), and an increase in %SiB from hour 2 to hour 4 (two way ANOVA; **p < 0.01) when compared to CFR-P during early sessions, and both CRF-NR during early and late sessions. C) The pie charts indicate that out of 71 non-CRF units (top panel, left), at the beginning of the lick bouts, 52 showed no response (non-CRF-NR), 4 were bout-inhibited (non-CRF-I), 7 were pre-bout activated (non-CRF-P), and 8 were bout-excited (non-CRF-E). Similarly, at the end of the bouts, out of 71 non-CRF neurons (top panel, right), 57 were non responsive (non-CRF-NR), 1 was bout-inhibited (non-CRF-I), 7 were pre-bout activated (non-CRF-P), and 6 were bout-excited (non-CRF-E). On the other hand, out of 52 CRF units (bottom panel, left), at the beginning of the lick bouts, 30 showed no response (CRF-NR), 5 were bout-inhibited (CRF-I), 3 were pre-bout activated (CRF-P), and 14 were bout-excited (CRF-E). On the contrary, at the end of the bouts, out of 52 CRF neurons (bottom panel, right), 39 were non responsive (CRF-NR), 2 were bout-inhibited (CRF-I), 7 were pre-bout activated (non-CRF-P), and 4 were bout-excited (CRF-E). The number of non-CRF and CRF neurons differs from the previous data set because one session had no lick bouts.

These results suggest that CRF-P neurons go through plastic changes over repeated exposure to ethanol that induce an increase in their tonic and phasic electrical activity, while CRF-NR units are not affected.

3.6. CRF neurons show higher sensitivity to ethanol lick bouts

Then, we analyzed the macrostructure of the ethanol drinking, focusing on the response of CRF and non-CRF neurons to ethanol lick bouts (at the beginning and at the end of the bouts), and we classified every single neuron based upon the time-locked response to bouts. A bout was defined as at least 20 consecutive licks, with an interlick-interval shorter than 60 s. Interestingly, we observed a different distribution of responses to a bout compared to a lick (Fig 4 C). Moreover, CeA CRF neurons appear to show a higher % of neurons with a time-locked response during the beginning of a bout (E+P+I: 42.31%) than the end (E+P+I: 25%). CeA CRF neurons exhibited also more time-locked responses at the beginning of a bout (42.31%) when compared to putative non-CRF neurons (26.76%).

3.7. CRF neurons do not encode for natural reward

In a final set of experiments, we recorded and analyzed CeA CRF and non-CRF neuronal activity in a sub cohort of CRF-Cre mice (N = 4) in response to water and 2% sucrose consumption in a limited access model identical to the DID procedures. In both our water and sucrose experiments there were no differences between CRF and non-CRF neurons firing rate (M-W testwater; U = 225, p = 0.1914; M-W testsucrose; U = 283, p = 0.2833) and %SiB (M-W testwater; U = 286, p = 0.9438; M-W testsucrose; U = 313, p = 0.5390; Fig 5 A, C). Furthermore, when we analyzed the response to licks of water and sucrose, in contrast to our ethanol data, we did not find any increase in firing rate (K-W testwater; H = 0.5198, p = 0.7711; K-W testsucrose; H = 0.2677, p = 0.8747) and bursting activity (K-W testwater; H = 0.4918, p = 0.7820; K-W testsucrose; H = 1.245, p = 0.5367) for CeA unit classified as CRF-P (Fig 5 B, D). These data indicate that CeA CRF neurons respond differentially to ethanol and natural rewards.

Figure 5.

Figure 5.

Open in a new tab

CRF neurons do not encode for natural reward. A) Left: The pie chart shows that out of 51 total units recorded during water sessions, 20 were identified as putative CRF neurons, 29 non-responsive to light, and a small population inhibited (2; N = 4 mice). Right: Electrophysiological characterization of CRF vs non-CRF neurons. Graphs showing firing rate (top), and %SiB (bottom). M-W test; p > 0.05. B) Left: The pie chart indicates that out of our 20 CRF units, 2 were lick-excited (CRF-E), 6 were pre-lick activated (CRF-P), and 12 showed no response (CRF-NR). Right: Graph show electrophysiological parameters such as firing rate and %SiB of CRF-NR, CRF-P, and non-CRF NR (included as a control) neurons. K-W test; p > 0.05. C) Left: The pie chart shows that out of 72 total units recorded during sucrose sessions, 12 were identified as putative CRF neurons, 59 non-responsive to light, and a small population excited (1; N = 4 mice). Right: Electrophysiological characterization of CRF vs non-CRF neurons. Graphs showing firing rate (top), and %SiB (bottom) M-W test; p > 0.05. D) Left: The pie chart indicates that out of our 12 CRF units, 1 was lick-excited (CRF-E), 1 was lick-inhibited (CRF-I), 3 were pre-lick activated (CRF-P), and 7 showed no response (CRF-NR). Right: Graph show electrophysiological parameters such as firing rate and %SiB of CRF-NR, CRF-P, and non-CRF NR (included as a control) neurons. K-W test; p > 0.05.

4. Discussion

The CeA is identified as a critical brain region involved in mediating excessive binge ethanol drinking (Cozzoli et al., 2014; De Guglielmo et al., 2016; Lee et al., 2015; Olney et al., 2017). Particularly, histochemical and electrophysiological evidence suggests that binge drinking engages the CeA CRF system (Gilpin et al., 2012; Gilpin and Roberto, 2012; Karanikas et al., 2013; Lowery-Gionta et al., 2012). However, the functional dynamics of CeA CRF activity during these behaviors have remained elusive. Here, we used optogenetic strategies coupled with in vivo electrophysiology to identify and record from CeA CRF neurons during acute and repeated binge drinking sessions in mice (Fig 1). Since CeA CRF neurons cannot be identified by waveform characteristics in traditional electrophysiology experiments, we identified them by measuring spike latency after photostimulation in vivo (Fig 1 B). Although all photostimulations occurred at the end of each ethanol session to avoid unintended excitation during the behavioral experiments (Cohen et al., 2012), it is also possible that our phototagging stimulation procedure could induce plasticity in CeA CRF neurons. Therefore, we utilized a brief photostimulation protocol (20 pulses 4 ms length, using 4 random frequencies: 1, 5, 10, and 20 Hz) to limit the chances of plasticity. It is also important to note that a unit that exhibited no response to optical stimulation does not mean that was not a CRF unit. Factors such as light penetrance through brain tissue as well as viral transduction play a role in the ability to optically phototag CRF cells. However, using our latency cutoff time of 10 ms should have minimized our chances of recording non CRF neurons due to downstream trans-synaptic activation. Approximately 39% of all CeA units recorded were identified as putative CRF cells. First, we compared identified CRF neurons to other recorded non-light responsive cells (which we termed non-CRF) in the CeA from our entire binge drinking recording sessions. We found that CeA CRF neurons exhibit increased firing and bursting activity compared to non-light responsive CeA units (Fig 1 E). Since, it is hypothesized that multiple ethanol exposure, as well as withdrawal, engages CeA CRF neurons (de Guglielmo et al., 2019), our result may be due to our repeated binge drinking procedure. However, one caveat is that we did not record during the 3-day abstinence period during the DID procedures. Due to CeA CRF neurons involvement in ethanol withdrawal (Zorrilla et al., 2014), we cannot rule out that the changes we observed are influenced by ethanol withdrawal.

Since CeA CRF cells project to multiple brain areas including the BNST, ventral tegmental area (VTA), periaqueductal grey, and substantia innominata, it was not surprising that we identified a heterogeneous population within our CeA CRF neurons in regard to ethanol consumption. Therefore, we subdivided the CeA CRF population based on their firing rate in response to a lick of ethanol. The vast majority of CeA CRF units were either pre-lick activated (P) or exhibited no effect (NR) with only smaller numbers within the inhibition (I) and excitation (E) groups (Fig 2 A, B). Next, we examined plastic changes of CRF-P, -NR, and non-CRF cells within the CeA after binge ethanol drinking. CRF-P neurons were more excitable and exhibited higher bursting activity than CRF-NR and non-CRF units (Fig 2 C). We speculate that the increased phasic activity observed could be due to the increased probability of CRF peptide release from CeA neurons after excessive and/or repeated binge ethanol consumption (Deussing and Chen, 2018). Although we did not observe a significant correlation between CRF cell activity and ethanol consumption in this model (Fig S2 C, D), the increased burst firing may be representative of pulsatile peptide release, since it is believed that individual spikes do not elicit peptide release. Future studies are necessary to determine whether peptide release can be observed with phasic high frequency stimulation patterns.

As described above, CRF-P neurons exhibited significant increases in firing rates when compared to CRF-NR neurons. However, with further analyses we observed a biphasic distribution in the CRF-P and the CRF-NR cells during our ethanol sessions (Fig 3 B). In both groups (CRF-P and CRF-NR) we found that one subpopulation significantly increased their frequency (CRF-P(Δ+) and CRF-NR(Δ+)) as the mouse consumed more ethanol, while the other group exhibited a decrease (CRF-P(Δ−) and CRF-NR(Δ−)). This is an important point since, paradoxically, human studies have shown that global administration of a CRF1 antagonist did not reduce ethanol craving in ethanol-addicted patients (Kwako et al., 2015; Schwandt et al., 2016). Additionally, a recent study has shown that serotonin has biphasic effects on GABA neurotransmission of CeA CRF during ethanol withdrawal (Khom et al., 2020). Thus, the targeting of CeA CRF microcircuits may have more therapeutic value. One drawback with these analyses was that all data were grouped together regardless of ethanol session. So consequently, our next set of analyses examined how acute and chronic binge ethanol consumption may modify the CeA CRF system.

To do this, we examined the activity of CRF-P and CRF-NR neurons during acute and repeated binge ethanol sessions (Fig 4 A, B). Studies reveal that repeated ethanol vapor administration results in excessive drinking, which is blocked by CeA CRF antagonists (Finn et al., 2007; Funk et al., 2006; Koob and Kreek, 2007). Moreover, chronic intermittent ethanol vapor leads to an upregulation of CRF and CRF1 receptor mRNA (Hansson et al., 2015). These data provide strong presumptive evidence that chronic ethanol administration can hijack the CeA CRF system leading to greater maladaptive neuronal plasticity. Critically, we observed that only CRF-P neurons increased their firing rate and bursting activity during binge drinking cycles. This effect was not observed in CRF (NR) cells. Importantly, our data show that CeA CRF units did not exhibit either an increased electrical activity or a strong transient pre-lick activation in our water and sucrose control groups (Fig 5). However, we did not record after repeated exposure to multiple sucrose and water sessions.

To provide more temporal specificity we also analyzed the distribution of ethanol lick responses by each week (data not shown). However, we did not observe any pre-lick activated CeA CRF neurons during week 1. There are a few possibilities for this finding: 1. Pre-lick activated CeA CRF neurons may only come online after multiple ethanol sessions. It is well known that alcohol use disorder, as well as addiction from other drugs of abuse, is characterized by a three stages cycle: binge/intoxication, withdrawal/negative affect, and preoccupation/anticipation. Importantly, the negative emotional affect arises from dysregulation of brain reward and stress systems modulated by CRF (Koob, 2015). Accordingly, we only observed pre-lick activated neurons after repeated cycles of DID. Furthermore, this neuronal population increased its tonic and phasic electrical activity over repeated exposure to ethanol as we showed when compared early and late ethanol sessions (Fig 4 A, B). These results suggest that repeated exposure to ethanol might induce adaptations in the CeA CRF microcircuit that become clear only in pre-lick activated CRF neurons. 2. Alternatively, pre-lick activated cells may come online due to learning mechanisms. Therefore, to further elucidate these mechanisms, future studies will examine CeA CRF neuronal activity during ethanol exposure and learning paradigms, using in vivo whole cell calcium imaging. This technique may provide more information on the plastic changes of individual CeA CRF neurons activity over time as in vivo electrophysiology cannot reliably examine the activity from the same neuron over days. Nevertheless, these findings indicate that ethanol may “hijack” or plastically alter CeA CRF neuronal activity.

We also examined the firing properties of CeA CRF neurons in response to a bout of ethanol licking. We defined an ethanol drinking bout as at least 20 licks with an interlick interval shorter than 60 s. Interestingly, we observed a different distribution of populations in response to a bout compared to a lick (Fig 4 C). CeA CRF neurons appear to exhibit a higher % of neurons with a time-locked response at the beginning of a bout (E+P+I; 42.31%) than at the end (25%). Furthermore, CeA CRF neurons also showed more time-locked responses at the beginning of a bout (42.31%) when compared to putative non-CRF neurons (26.76%). These data suggest that CeA CRF neurons may be more responsive during the initiation of ethanol consumption.

A recent study showed sex-specific differences in CRF-mediated excitation/inhibition of CeA neurons (Agoglia et al., 2020). However, in our study we pooled male and female data together since we only observed slight non-significant sex differences in ethanol consumption. Finally, human studies have shown that administration of CRF antagonists does not alter cravings for ethanol in patients from AUD (Kwako et al., 2015; Schwandt et al., 2016). These data indicate that the CRF system may not be a target for pharmacological interventions of AUD. However, it is important to note, that these studies looked at global inactivation. Nevertheless, we found that the CRF microcircuitry in the CeA is diverse, complex, and heterogeneously affected by the duration of ethanol exposure. Moreover, since CRF is coexpressed with other neurotransmitters and neuropeptides, it could be possible that more selective targeting may be more efficacious.

In conclusion, we identified a diverse role of CeA CRF neurons in response to binge drinking. These data make a strong case that genetically identical neurons within a structure are heterogeneous and complex. It bears the caveat that optogenetic and/or pharmacological manipulation of identical groups of neurons may provide a false positive in regard to behavior. Future studies should examine the role of circuits and activity-dependent groups of neurons for a more dissected analysis of behaviorally relevant patterns of activity.

Supplementary Material

S1

Figure Supplemental 1. Optoarrays placement. The Atlas plates containing the CeA show that the optoarrays were all located inside the CeA (red dots, N = 17). BLA = basolateral amygdaloid nucleus, anterior part. CeC = lateral capsular part of the central nucleus of the amygdala (CeA). CeL, CeM= lateral and medial subdivision of the CeA, respectively.

S2

Figure Supplemental 2. No correlation between the average firing rate of CRF neurons and the total number of ethanol licks. A) Graph of representative number of ethanol licks over repeated ethanol sessions in a different cohort of mice without implanted optoarray (N = 12), compared to water licks. B) The pie chart indicates that out of our 75 non-CRF units, 47 showed no response (non-CRF-NR), 5 were lick-inhibited (non-CRF-I), 14 were pre-lick activated (non-CRF-P), and 9 were lick-excited (non-CRF-E). C) Correlation between the average firing rate of all CRF neurons (n = 59) to the corresponding animal’s total number of ethanol licks. We found no significant correlation between the total number of licks versus the average firing rate across the whole 4 hour session (r = 0.006066, p = 0.9636). D) Correlation between the average firing rate of all CRF neurons (n = 59) in only the 4th hour of the DID paradigm to the corresponding animal’s total number of ethanol licks. We found no significant correlation between total number of licks and the 4th hour average firing rate (r = 0.03102, p = 0.8156).

Acknowledgments:

We thank Dr. Brian Mathur and Antonio Figueiredo for discussion. We thank Colin Maehler, Houman Qadir, and Soham Roy for their technical assistance.

Funding and Disclosure:

This work was supported by NIH/NIAAA grant R01 AA 027516 R00 AA021417 (D. R. S.), NIH/NIDA grant F31 DA 047050 (K. S. G.), NIH/NIDA grant R01 DA 022340, R01 DA 045639 (J. F. C.). The authors declare no competing interests

References

  1. Agoglia AE, Tella J, Herman MA, 2020. Sex differences in corticotropin releasing factor peptide regulation of inhibitory control and excitability in central amygdala corticotropin releasing factor receptor 1-neurons. Neuropharmacology. 10.1016/j.neuropharm.2020.108296 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barkley-Levenson AM, Crabbe JC, 2015. Distinct ethanol drinking microstructures in two replicate lines of mice selected for drinking to intoxication. Genes, Brain Behav. 10.1111/gbb.12225 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Benhamou L, Cohen D, 2014. Electrophysiological characterization of entopeduncular nucleus neurons in anesthetized and freely moving rats. Front. Syst. Neurosci. 10.3389/fnsys.2014.00007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bonomo YA, Bowes G, Coffey C, Carlin JB, Patton GC, 2004. Teenage drinking and the onset of alcohol dependence: A cohort study over seven years. Addiction. 10.1111/j.1360-0443.2004.00846.x [DOI] [PubMed] [Google Scholar]
  5. Cohen JY, Haesler S, Vong L, Lowell BB, Uchida N, 2012. Neuron-type-specific signals for reward and punishment in the ventral tegmental area. Nature 482, 85–88. 10.1038/nature10754 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cozzoli DK, Courson J, Wroten MG, Greentree DI, Lum EN, Campbell RR, Thompson AB, Maliniak D, Worley PF, Jonquieres G, Klugmann M, Finn DA, Szumlinski KK, 2014. Binge alcohol drinking by mice requires intact group1 metabotropic glutamate receptor signaling within the Central nucleus of the Amygdale. Neuropsychopharmacology. 10.1038/npp.2013.214 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. De Guglielmo G, Crawford E, Kim S, Vendruscolo LF, Hope BT, Brennan M, Cole M, Koob GF, George O, 2016. Recruitment of a neuronal ensemble in the central nucleus of the Amygdala is required for alcohol dependence. J. Neurosci. 10.1523/JNEUROSCI.1395-16.2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. de Guglielmo G, Kallupi M, Pomrenze MB, Crawford E, Simpson S, Schweitzer P, Koob GF, Messing RO, George O, 2019. Inactivation of a CRF-dependent amygdalofugal pathway reverses addiction-like behaviors in alcohol-dependent rats. Nat. Commun. 10.1038/s41467-019-09183-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Deussing JM, Chen A, 2018. The corticotropin-releasing factor family: Physiology of the stress response. Physiol. Rev. 10.1152/physrev.00042.2017 [DOI] [PubMed] [Google Scholar]
  10. Finn DA, Snelling C, Fretwell AM, Tanchuck MA, Underwood L, Cole M, Crabbe JC, Roberts AJ, 2007. Increased drinking during withdrawal from intermittent ethanol exposure is blocked by the CRF receptor antagonist D-Phe-CRF(12–41). Alcohol. Clin. Exp. Res. 10.1111/j.1530-0277.2007.00379.x [DOI] [PubMed] [Google Scholar]
  11. Funk CK, O’Dell LE, Crawford EF, Koob GF, 2006. Corticotropin-releasing factor within the central nucleus of the amygdala mediates enhanced ethanol self-administration in withdrawn, ethanol-dependent rats. J. Neurosci. 10.1523/JNEUROSCI.3096-06.2006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gilpin NW, Karanikas CA, Richardson HN, 2012. Adolescent binge drinking leads to changes in alcohol drinking, anxiety, and amygdalar corticotropin releasing factor cells in adulthood in male rats. PLoS One. 10.1371/journal.pone.0031466 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gilpin NW, Roberto M, 2012. Neuropeptide modulation of central amygdala neuroplasticity is a key mediator of alcohol dependence. Neurosci. Biobehav. Rev. 10.1016/j.neubiorev.2011.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hansson AC, Spanagel R, Bilbao A, 2015. Chronic intermittent ethanol exposure in mice leads to an Up-Regulation of CRH/CRHR1 signaling. Alcohol. Clin. Exp. Res. 10.1111/acer.12686 [DOI] [PubMed] [Google Scholar]
  15. Karanikas CA, Lu YL, Richardson HN, 2013. Adolescent drinking targets corticotropin-releasing factor peptide-labeled cells in the central amygdala of male and female rats. Neuroscience. 10.1016/j.neuroscience.2013.04.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Khom S, Wolfe SA, Patel RR, Kirson D, Hedges DM, Varodayan FP, Bajo M, Roberto M, 2020. Alcohol dependence and withdrawal impair serotonergic regulation of GABA transmission in the rat central nucleus of the amygdala. J. Neurosci. 10.1523/JNEUROSCI.0733-20.2020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Koob G, Kreek MJ, 2007. Stress, dysregulation of drug reward pathways, and the transition to drug dependence. Am. J. Psychiatry. 10.1176/appi.ajp.2007.05030503 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Koob GF, 2015. The dark side of emotion: The addiction perspective. Eur. J. Pharmacol. 10.1016/j.ejphar.2014.11.044 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Koob GF, Le Moal M, 2008. Review. Neurobiological mechanisms for opponent motivational processes in addiction. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 363, 3113–3123. 10.1098/rstb.2008.0094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kwako LE, Spagnolo PA, Schwandt ML, Thorsell A, George DT, Momenan R, Rio DE, Huestis M, Anizan S, Concheiro M, Sinha R, Heilig M, 2015. The Corticotropin Releasing Hormone-1 (CRH1) Receptor Antagonist Pexacerfont in Alcohol Dependence: A Randomized Controlled Experimental Medicine Study. Neuropsychopharmacology. 10.1038/npp.2014.306 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lee KM, Coehlo M, McGregor HA, Waltermire RS, Szumlinski KK, 2015. Binge alcohol drinking elicits persistent negative affect in mice. Behav. Brain Res. 10.1016/j.bbr.2015.05.055 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lowery-Gionta EG, Navarro M, Li C, Pleil KE, Rinker JA, Cox BR, Sprow GM, Kash TL, Thiele TE, 2012. Corticotropin releasing factor signaling in the central amygdala is recruited during binge-like ethanol consumption in C57BL/6J mice. J. Neurosci. 10.1523/JNEUROSCI.6256-11.2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Olney JJ, Navarro M, Thiele TE, 2017. The Role of Orexin Signaling in the Ventral Tegmental Area and Central Amygdala in Modulating Binge-Like Ethanol Drinking Behavior. Alcohol. Clin. Exp. Res. 10.1111/acer.13336 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Pleil KE, Rinker JA, Lowery-Gionta EG, Mazzone CM, McCall NM, Kendra AM, Olson DP, Lowell BB, Grant KA, Thiele TE, Kash TL, 2015. NPY signaling inhibits extended amygdala CRF neurons to suppress binge alcohol drinking. Nat. Neurosci. 10.1038/nn.3972 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rhodes JS, Best K, Belknap JK, Finn DA, Crabbe JC, 2005. Evaluation of a simple model of ethanol drinking to intoxication in C57BL/6J mice. Physiol. Behav. 10.1016/j.physbeh.2004.10.007 [DOI] [PubMed] [Google Scholar]
  26. Rhodes JS, Ford MM, Yu CH, Brown LL, Finn DA, Garland T, Crabbe JC, 2007. Mouse inbred strain differences in ethanol drinking to intoxication. Genes, Brain Behav. 10.1111/j.1601-183X.2006.00210.x [DOI] [PubMed] [Google Scholar]
  27. Rinker JA, Marshall SA, Mazzone CM, Lowery-Gionta EG, Gulati V, Pleil KE, Kash TL, Navarro M, Thiele TE, 2017. Extended Amygdala to Ventral Tegmental Area Corticotropin-Releasing Factor Circuit Controls Binge Ethanol Intake. Biol. Psychiatry. 10.1016/j.biopsych.2016.02.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Rivier J, Spiess J, Vale W, 1983. Characterization of rat hypothalamic corticotropin-releasing factor. Proc. Natl. Acad. Sci. U. S. A. 10.1073/pnas.80.15.4851 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Sacks JJ, Gonzales KR, Bouchery EE, Tomedi LE, Brewer RD, 2015. 2010 National and State Costs of Excessive Alcohol Consumption. Am. J. Prev. Med. 10.1016/j.amepre.2015.05.031 [DOI] [PubMed] [Google Scholar]
  30. Sah P, Faber ESL, De Armentia ML, Power J, 2003. The amygdaloid complex: Anatomy and physiology. Physiol. Rev. 10.1152/physrev.00002.2003 [DOI] [PubMed] [Google Scholar]
  31. Sah P, Lopez De Armentia M, 2003. Excitatory synaptic transmission in the lateral and central amygdala, in: Annals of the New York Academy of Sciences. 10.1111/j.1749-6632.2003.tb07072.x [DOI] [PubMed] [Google Scholar]
  32. Schwandt ML, Cortes CR, Kwako LE, George DT, Momenan R, Sinha R, Grigoriadis DE, Pich EM, Leggio L, Heilig M, 2016. The CRF1 antagonist verucerfont in anxious alcohol-dependent women: Translation of neuroendocrine, but not of anti-craving effects. Neuropsychopharmacology. 10.1038/npp.2016.61 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sparta DR, Stamatakis AM, Phillips JL, Hovelsø N, van Zessen R, Stuber GD, 2012. Construction of implantable optical fibers for long-term optogenetic manipulation of neural circuits. Nat. Protoc. 7, 12–23. 10.1038/nprot.2011.413 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Volkow ND, Koob GF, McLellan AT, 2016. Neurobiologic advances from the brain disease model of addiction. N. Engl. J. Med. 10.1056/NEJMra1511480 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Zorrilla EP, Logrip ML, Koob GF, 2014. Corticotropin releasing factor: A key role in the neurobiology of addiction. Front. Neuroendocrinol. 10.1016/j.yfrne.2014.01.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

S1

Figure Supplemental 1. Optoarrays placement. The Atlas plates containing the CeA show that the optoarrays were all located inside the CeA (red dots, N = 17). BLA = basolateral amygdaloid nucleus, anterior part. CeC = lateral capsular part of the central nucleus of the amygdala (CeA). CeL, CeM= lateral and medial subdivision of the CeA, respectively.

NIHMS1780919-supplement-S1.tif (96.3MB, tif)
S2

Figure Supplemental 2. No correlation between the average firing rate of CRF neurons and the total number of ethanol licks. A) Graph of representative number of ethanol licks over repeated ethanol sessions in a different cohort of mice without implanted optoarray (N = 12), compared to water licks. B) The pie chart indicates that out of our 75 non-CRF units, 47 showed no response (non-CRF-NR), 5 were lick-inhibited (non-CRF-I), 14 were pre-lick activated (non-CRF-P), and 9 were lick-excited (non-CRF-E). C) Correlation between the average firing rate of all CRF neurons (n = 59) to the corresponding animal’s total number of ethanol licks. We found no significant correlation between the total number of licks versus the average firing rate across the whole 4 hour session (r = 0.006066, p = 0.9636). D) Correlation between the average firing rate of all CRF neurons (n = 59) in only the 4th hour of the DID paradigm to the corresponding animal’s total number of ethanol licks. We found no significant correlation between total number of licks and the 4th hour average firing rate (r = 0.03102, p = 0.8156).

NIHMS1780919-supplement-S2.tif (96.3MB, tif)

RESOURCES