Below you will find the Research Strategy for an NIH K01 application submitted 2017 and 2018. Neither submission was funded. The experiments described under Aim 1 were published in Hippocampus [(2021) Jan;31(1):3-10. doi: 10.1002/hipo.23263. Epub 2020 Sep 18].
RESEARCH STRATEGY
BACKGROUND AND SIGNIFICANCE
Withdrawal from drugs of abuse directly impacts the brain’s stress and memory systems [12-14], which may underlie individual susceptibility to persistent drug seeking and stress-induced relapse. Preclinical studies demonstrate impaired fear memory processes in rodents withdrawn from alcohol, including abnormally heightened fear responses that are resilient to subsequent attenuation by extinction training [15-18]. Fear extinction and addiction share several features including sensitivity to contextual cues and stress-induced reinstatement of maladaptive behaviors [19-21], though the underlying neural circuits mediating, and sufficient to intervene with, impaired extinction following alcohol withdrawal have remained elusive. A mechanistic and circuit-level understanding of withdrawal-induced extinction impairments may lead to the development of effective treatments that facilitate abstinence and prevent relapse in alcohol and other drug disorders.
Alcohol is the most commonly abused drug world-wide. Alcohol Use Disorders (AUDs) is a diagnostic term referring to maladaptive behaviors associated with alcohol abuse and addiction – the continuation of alcohol consumption in the face of negative social, behavioral, and health outcomes (DSM-5). AUDs affect an estimated 29.1% of the US population at some time in their lives [22] and pose significant personal and economic costs; costs include lost productivity, healthcare, criminal justice, and quality of life, accounting for $249.0 billion to the US economy in 2010 [23, 24]. Behavioral and pharmacotherapies improve outcomes, yet even after formal treatment abstinence rates (one measure of treatment effectiveness) range from 25% to 43% [25] suggesting that more than half of individuals that obtain current treatments relapse within a year. Less than 15% of patients with probable AUDs receive treatment [26]. These sobering findings underlie the view that alcohol addiction is a chronically relapsing disorder [27]. Understanding the neurobehavioral mediators of relapse – specifically, context- and stress-induced reinstatement of maladaptive behaviors following withdrawal – will facilitate the development of effective treatments to prevent relapse in individuals with AUD and other drug disorders.
PRELIMINARY FINDINGS
Fear extinction in ethanol-withdrawn mice. Fear conditioning involves pairing a previously neutral conditional stimulus (CS; e.g., tone or context) with an aversive unconditional stimulus (US; e.g., mild foot shock). After CS-US pairings, the CS alone comes to elicit a conditioned response (CR; e.g., freezing). Continued presentations of the CS in the absence of the US leads to a reduction of the CR, a process of new learning (i.e., CS-no-US) referred to as extinction. Animal models of fear and extinction have provided insight into the neural basis of fear and anxiety. Importantly, extinction processes have been successfully translated to treat fear, anxiety, and addiction disorders in human populations [28]. Reductions in fear and stress memory through extinction could help mitigate cue- and stress-induced relapse following withdrawal from drugs, including alcohol. However, preclinical studies show impaired fear memory processes in rodents withdrawn from alcohol, including impaired extinction learning and exaggerated stress-induced reinstatement responses [15-18]. Understanding the neural mechanisms that maintain persistent fear responses and devising methods to facilitate extinction following ethanol withdrawal could lead to better treatment approaches for addiction.
| Figure 1a: Mice withdrawn from chronic ethanol (EtOH-WD) show impaired context extinction over multiple days compared to saline-treated controls (Sal). |
In our preliminary studies, mice received daily intraperitoneal injections of ethanol (1.5 g/kg; EtOH; n=6) or saline (Sal; n=6) for five consecutive days followed by a two-day alcohol-free period before behavioral testing (following methods described in [17]). EtOH and Sal mice were placed into a conditioning context (Ctx A) and, after two minutes, received five foot shocks (0.5 mA 0.5-sec; ISI=60s) during fear acquisition. In the first experiment, mice (male and female) were returned to Ctx A for four 12-min extinction sessions each 24-hours apart. As shown in Figure 1a, EtOH mice show significantly higher levels of context-mediated freezing across all four extinction days relative to controls, evidence of impaired contextual fear extinction. These preliminary results replicate the findings of others demonstrating marked impairments in fear extinction in rodents withdrawn from ethanol [15-18] and further demonstrate the persistence of this effect over multiple extinction sessions. We also show impaired extinction in both male and female mice withdrawn from ethanol, a first step in examining potential sex differences in this behavior (Figure 1b). These effects are not due to the acute effects of ethanol
| Figure 1b: Male and female mice withdrawn from EtOH show impaired fear extinction compared to SAL. Potential sex differences in extinction merit inclusion of both sexes. |
(which is no longer in the system) or acute effects of withdraw – they are a direct result of the neurobiological changes that occur following withdrawal from chronic alcohol. Further, fear memories formed prior to alcohol exposure and withdrawal also fail to extinguish [17], suggesting that withdrawal from ethanol augments the retrieval of fear memories and disrupts extinction learning. These fear extinction impairments complement and extend previous work on extinction using instrumental learning paradigms (e.g. rats trained to press a lever for an ethanol reward require more non-reinforced training sessions to reach extinction criteria compared to control rats [29].)
Reconsolidation-extinction and reinstatement in ethanol-withdrawn mice. Fear extinction is thought to reflect new learning and not an erasure of the original memory – a process by which a new memory competes with and actively inhibits the original fear memory [19, 30]. Behavioral evidence for this involves the return of fear responses following extinction training through renewal or reinstatement. Fear renewal refers to increased fear responding when the CS is encountered in a context different from that in which it was extinguished; that is, extinction learning is context-dependent [31]. In fear reinstatement, exposure to the original US or stress returns fear responding to levels seen prior to extinction training. Although clinical applications of extinction processes for the treatment of stress, anxiety, and addiction disorders have shown promise [28, 32], context- and/or stress-induces renewal and reinstatement remain ongoing challenges to achieve permanent reductions in maladaptive behaviors [33].
| Figure 2: Mice were briefly exposed to the training context one hour prior to extinction training. Reconsolidation-extinction does not prevent fear reinstatement in mice withdrawn from chronic ethanol (EtOH) compared to saline-controls (Sal) |
| Reconsolidation-Extinction and Reinstatement |
Fear memories can be made labile and attenuated by presenting the CS in a brief reactivation session followed subsequently (~ 1-hr) by an extinction session [34]. Termed reconsolidation-extinction, this process facilitates extinction learning and mitigates renewal and reinstatement of fear. Reconsolidation-extinction may thus offer a behavioral means to facilitate the treatment of stress and addiction disorders in clinical patients, including alcohol disorders [34-38]. To determine if reconsolidation-extinction can facilitate extinction learning and reduce shock-induced reinstatement in EtOH mice, we conditioned female EtOH and Sal mice as previously stated, except mice were exposed to only 3 US presentations during acquisition to induce moderate fear learning. Following fear acquisition, mice received a 3-min exposure to the training context to reactivate the fear memory and initiate reconsolidation, and were then placed back into their home cages. One hour later, mice were given a 30-minute extinction session in Ctx A. Mice received three additional days of 30-min Ctx A extinction trials. During the initial five minutes of the first extinction session, the groups did not differ in context-mediated freezing (data not shown). However, average freezing across the first three 30-minute sessions revealed that EtOH mice showed higher levels of freezing compared to SAL mice. By Day 4, freezing levels between EtOH and SAL mice did not statistically differ (Figure 2).
Following the fourth day of extinction, mice received a brief foot shock immediately upon placement into a distinct context (Ctx B). Exposure to an immediate shock in a novel context does not promote contextual fear conditioning [6, 39], but is sufficient to reinstate the original fear memory, thus allowing for exposure to the US in the absence of additional context learning. EtOH mice increased their levels of freezing during the reinstatement test (Test); this increase was not seen in SAL mice. Although we intend to examine additional males and non-reconsolidated controls, these data suggest that fear memories in EtOH mice are more resistant to extinction and less susceptible to the reconsolidation effects demonstrated in Sal mice, which resonates with a previous report [40] [c.f., [35] for evidence of reconsolidation-extinction of EtOH mice in an instrumental conditioning design]. Together, preliminary experiments 1 and 2 suggest that fear extinction is impaired in EtOH mice and that the acquired fear memory is less susceptible to attenuation compared to controls when utilizing reconsolidation-extinction as a behavioral intervention.
| Figure 3. Genetically engineering hippocampus cells active during learning to express ChR2. A mouse is injected with a virus cocktail consisting of c-Fos-tTA and TRE-ChR2 into the hippocampus, followed by optic fiber implants. When off Dox, the formation of a memory induces the expression of tTA, which binds to TRE and drives the expression of ChR2, thereby labeling a population of activated cells (yellow). Modified from [45]. |
Activity-dependent and optogenetic approaches to fear reduction. Ensembles of neurons distributed throughout the brain are thought to encode and maintain specific memory [41, 42]. These neurons can be tagged during learning for subsequent identification and manipulation [43]. The hippocampus (HPC) in particular is pivotal for the encoding, storage, and retrieval of personally experienced, or episodic, memories [44]. Recently, our work has demonstrated that HPC cells in the dentate gyrus (DG) sub region of HPC that previously expressed the immediate early gene (IEG) c-Fos during learning are sufficient to activate the neuronal and behavioral expression of negative, neutral, and positive memory recall [45-47]. These cells also undergo plasticity-related changes during learning [48] and are necessary for the behavioral expression of memory recall [49, 50], thus corroborating their mnemonic nature and raising the possibility of modulating their activity to alter a variety of maladaptive behavioral states.
To activate memories in healthy and ethanol-withdrawn mice, we will utilize an activity-dependent and inducible system to tag HPC neurons active during memory formation [51]. This system leverages the activity-dependent nature of the c-Fos promoter and couples it to the tetracycline transactivator (tTA), which, when activated by neural activity, binds to the tetracycline response element (TRE) and thereby promotes transcription of the light-sensitive cation channel channelrhodopsin-2 (ChR2) in a doxycycline (Dox)-dependent manner (Figure 3). When Dox is removed from the animals’ diet, neural activity leads to c-Fos-promoter-driven ChR2 expression in a defined set of cells. When Dox is present, this process is inhibited, thus providing the ability to open and close windows for activity-dependent labeling and manipulation of HPC neurons [45].
| Figure 4. Chronic activation of a fear memory tagged in Ctx A. Extinction-like reductions in Ctx A freezing relative to Ctx B |
By utilizing these tools, my recent data demonstrate that repeated optical activation of hippocampus cell ensembles processing discrete fear memories is sufficient to induce context-specific extinction-like behavior (Figure 4; [52]). We infused AAV9-c-Fos-tTA and AAV9-TRE-ChR2-mCherry or AAV9-TRE-mCherry into the dorsal DG and implanted optical fibers above. After a 10-day recovery period, mice were taken off Dox and received four foot shocks (2-sec, 1.5 mA) in Ctx A. Following acquisition, mice were placed back on Dox to close the window for tagging active hippocampus cells. The following day, mice received a separate fear acquisition session in Ctx B. Over the next five days, mice were placed into a distinct context (Ctx C) and received repeated light-stimulation (473nm, 20 Hz) twice a day, 10 minutes per session. Mice were finally tested for contextual fear in both Ctx A and Ctx B. As shown in Figure 4, context-mediated freezing was significantly reduced in Ctx A relative to Ctx B, only in ChR2 mice. These data show a context-specific extinction-like reduction in freezing following chronic reactivation of DG cells labeled during fear acquisition in Context A. A primary goal of the current proposal is to utilize activity-dependent labeling and modulation of fear memoires with in vivo optogenetics to facilitate extinction and mitigate renewal and reinstatement in mice withdrawn from chronic alcohol.
INNOVATION
Recent viral targeting, chemogenetic, and optogenetic strategies are currently revealing cell type-specific and circuit level mediators of alcohol-dependence, alcohol-seeking behaviors, and alcohol associative learning [53-56]. One limitation of these studies is that they indiscriminately target whole populations of cell types or pathways, which may obscure learning-specific alterations induced by ethanol [57]. For example, optogenetic inactivation of the entire DG failed to affect the retrieval of a contextual fear memory, while the selective inactivation of 6% of DG granule cells active during learning impaired memory retrieval [45, 46]. One approach developed to manipulate learning-related neural ensembles in drug research, the Daun02 inactivation procedure, demonstrated that selective inhibition of neural ensembles, but not global inhibition of the infralimbic cortex, increased alcohol seeking during subsequent reinstatement sessions [58]. Likewise, Daun02 inactivation of the CeA neuronal ensemble during early withdrawal reduced alcohol drinking in dependent rats that lasted at least two weeks [59].
Our activity-dependent and inducible genetic tagging system allows for the incorporation of either excitatory (i.e., ChR2) or inhibitory (i.e., ArchT) opsins into neuronal ensembles active during discrete periods of learning, which can then be manipulated at later time-points [60]. Using these strategies, we recently report acute rescue of stress-induced depression-related behaviors and circuit-level pathways supporting this rescue [61]. The current proposal extends these findings by incorporating these genetic tools to assess and manipulate discrete circuit elements mediating impaired fear extinction and exaggerated reinstatement responses in an animal model of chronic alcohol withdrawal. Fear conditioning as a behavioral assay for drug-withdrawal related effects has several advantages. First, the neural circuits mediating fear conditioning, extinction, renewal, and reinstatement are well described [30, 31, 62, 63] and overlap with those affected by chronic ethanol and withdrawal [12, 27, 64, 65]. Second, fear extinction impairments and exaggerated reinstatement responses reported in ethanol withdrawn rodents [15-18, 40, 66, 67] recapitulate findings in instrumental paradigms [20, 29, 56, 68] that have traditionally been used to model drug self-administration and relapse behaviors. And third, fear conditioning can occur in a single training session, which allows for the identification and genetic tagging of discrete learning-related neural ensembles, a significant advantage over instrumental paradigms with protracted learning curves.
APPROACH
Specific Aim 1: Optogenetic-induced extinction-like behavior in ethanol-withdrawn mice.
The goal of Aim 1 is to determine if fear memories resistant to extinction in ethanol-withdrawn mice can be mitigated through optogenetic reactivation of a tagged fear memory in the DG and whether optogenetic manipulations reduce fear renewal or reinstatement. Lesion, drug inactivation, chemogenetic, and optogenetic studies all demonstrate that the dorsal hippocampus processes contextual information during fear acquisition, retrieval, and extinction [31, 45, 46, 69-71]. The dorsal hippocampus does not contain monosynaptic connections with the amygdala, yet optical activation of a DG fear memory reactivates cells encoding that fear memory in the amygdala [46]. We therefore speculate that chronic DG reactivation leads to functional changes in fear ensembles in the amygdala and other areas, like those seen following behavioral extinction paradigms [63, 72]. The experiments outlined in Aim 1 will extend our preliminary data demonstrating context-specific, extinction-like reductions in fear following chronic stimulation of a tagged fear memory in mice (Figure 4); we hypothesize that chronic reactivation of a fear memory will mitigate fear extinction impairments and exaggerated fear renewal and reinstatement in mice withdrawn from ethanol.
Experiment 1. A full description of methods can be found under General Procedures. We will conduct four separate experiments under this Aim. Briefly, mice will receive viral infusions of AAV9-cFos-tTA and AAV9-TRE-ChR2-eYFP (ChR2) or AAV9-TRE-eYFP (eYFP) and optic fiber implants targeted at the DG. Following recovery, mice will be taken off Dox diet two days prior to contextual fear acquisition in Context A (Ctx A). Mice will then be placed back on a Dox diet. For Exp.1A, mice will receive contextual fear acquisition in Ctx B the following day. Following acquisition, mice will undergo a four-day ethanol exposure paradigm (i.e., Drinking in the Dark paradigm [73]; EtOH) followed by a two-day ethanol-free period (withdrawal) or receive no ethanol (water control; Con) before behavioral testing. A previous study reported impaired fear extinction in ethanol-withdrawn mice when fear acquisition occurred before ethanol exposure and withdrawal [17]. This strategy will allow us to specifically tag hippocampus cells processing a fear memory prior to subsequent changes induced by ethanol withdrawal and offers a more ethologically valid paradigm to induce addiction- and withdrawal-related phenotypes.
Exp. 1A. A 2 (ChR2 vs. eYFP) x 2 (Con vs. EtOH) design will test the hypothesis that chronic activation of a tagged fear memory will produce context-specific extinction-like reductions in freezing in Con and EtOH mice. Following the withdrawal period, Con and EtOH mice will be placed into Ctx C and receive repeated light-stimulation (573nm, 20 Hz) in 10-min sessions, twice a day for five days, as aforementioned (Figure 4). Mice will then be tested for contextual fear in both Ctx A and Ctx B. We predict that Con and EtOH mice infused with eYFP will show high levels of freezing in both Ctx A and Ctx B, with a potential main effect of sex (lower levels of freezing in females compared to males [1]). As demonstrated in our preliminary results, we predict that Con mice infused with ChR2 will show reduced freezing in Ctx A relative to Ctx B, thus demonstrating context-specific extinction-like behavior. Importantly, our previous data suggesting that chronic reactivation of DG cells processing a fear memory is sufficient to lastingly attenuate fear responses; accordingly, we hypothesize that chronic activation of a fear memory will be sufficient to produce extinction-like reductions in Ctx A freezing in EtOH mice infused with ChR2. Together, these experiments provide a potential novel intervention for suppressing fear responses that are normally resistant to attenuation.
| Figure 5. Experimental design for Exp. 1B. Reconsolidation-extinction utilizing a brief optogenetic activation of tagged fear memory followed one-hour later by a 30-min extinction session. |
Exp. 1B. A 2 (ChR2 vs. eYFP) x 2 (Con vs. EtOH) design will test the hypothesis that a brief optogenetic reactivation of a tagged fear memory one hour prior to extinction training will facilitate extinction learning in both Con and EtOH mice. Following withdrawal, Con and EtOH mice will then be placed into Ctx B and receive light-stimulation during a 3-minute reactivation session followed one hour later by a 30-minute extinction session in Ctx A (i.e., reconsolidation-extinction; behaviorally demonstrated to facilitate extinction and mitigate reinstatement and renewal [34]) over four days (Figure 5). Mice will then be tested for extinction retention. In mice injected with eYFP, we expect to see higher levels of freezing in EtOH mice relative to Con during the context test, in agreement with extinction impairments in EtOH mice previously demonstrated (see Figures 1 & 2). We hypothesize that briefly reactivating a fear memory for Context A prior to extinction will facilitate extinction in both Con and EtOH mice, with lower levels of freezing during testing relative to their eYFP counterparts. This result would be in agreement with a previous study reporting facilitated extinction of instrumental responding to ethanol-associated cues following a reconsolidation-extinction paradigm [35]. Alternatively, our optogenetic reconsolidation-extinction procedure may recapitulate the effects of behavioral reconsolidation-extinction shown in our preliminary data (i.e., Figure 2), with higher levels of freezing in EtOH mice during testing regardless of viral infusion. A previous study reports that reconsolidation manipulations during fear conditioning are less effective in rodents withdrawn from ethanol compared to controls [40].
Experiments 1A and 1B provide two potential optogenetic methods to facilitate fear extinction in ethanol withdrawn mice: one method relies on chronic reactivation of a fear memory to induce extinction-like changes in behavior whereas the other relies on brief optogenetic reactivation prior to extinction to facilitate extinction learning. Although we hypothesize that both methods will result in reduced fear responding in ETOH mice, only the latter technique is hypothesized to mitigate context renewal and reinstatement, two mediators of drug seeking behavior following extinction (in animal models) or abstinence (in humans).
Experiments 1C and 1D will examine the effects of optogenetic reconsolidation-extinction on context renewal and shock-induced reinstatement in Con and EtOH mice. First, all mice will receive a viral infusion and optical implants, fear acquisition and optical reconsolidation-extinction as described above. Following this design, separate groups of mice will undergo testing for fear renewal (Exp. 1C) or reinstatement (Exp. 1D).
| Figure 6. Fear memory tagged in Ctx A. Following EtOH treatment, mice receive four days of a brief reactivation session in Ctx C followed 1hr later by a 30-min extinction session in Ctx A. Mice will then be tested for extinction retention (Ctx A) or renewal (Ctx B). |
Exp. 1C. A 2 (ChR2 vs. eYFP) x (Con vs. EtOH) design will test the hypothesis that optogenetic reconsolidation-extinction will facilitate extinction learning and mitigate context-dependent fear renewal in Con and EtOH mice. Following reconsolidation-extinction, EtOH and Con mice will be tested for tone-induced freezing in both Ctx A (i.e., extinction testing) and Ctx B (i.e., fear renewal), with the order of testing counterbalanced (Figure 6). In mice infused with eYFP, both EtOH and Con mice are expected to show fear renewal, with lower levels of CS induced freezing in Ctx A relative to Ctx B. Holmes et al. [15] report a non-significant trend towards higher freezing levels in ethanol-withdrawn mice in a fear renewal context and we anticipate similar fear responses in EtOH and Con mice in the current design. Fear renewal is expected to be absent in Con mice infused with ChR2, with low levels of CS freezing in both contexts in agreement with mitigated fear renewal following standard reconsolidation-extinction techniques [34]. Finally, we hypothesize that optogenetic reactivation of a fear memory prior to extinction will facilitate extinction and mitigate fear renewal in EtOH mice infused with ChR2, like the behavior predicted for Con mice infused with ChR2.
| Figure 7. Following tagging in Ctx A, EtOH treatment and withdrawal, and optical reconsolidation-extinction, mice receive either an immediate shock or not in Ctx B. Mice will then be tested for fear reinstatement in Ctx A. |
Exp. 1D. A 2 (ChR2 vs. eYFP) x 2 (Con vs. EtOH) x 2 (Shock vs. No Shock) design will test the hypothesis that optical reconsolidation-extinction will facilitate extinction learning and mitigate shock-induced reinstatement in both Con and EtOH mice. To elicit reinstatement, mice will be placed into Ctx B and receive either no shock or an immediate shock upon placement (Figure 7). Mice will then be tested for CS-mediated freezing in Ctx A. We will calculate differences scores (the % freezing during the reinstatement test minus the % freezing during the last 5 minutes of CS extinction); reinstatement can thereby be defined as positive values significantly higher than zero (where zero denotes no difference). We expect significant extinction impairments and an exaggerated shock-induced reinstatement in EtOH relative to Con mice infused with eYFP, consistent with previous reports [16]. Con mice infused with ChR2 are expected to show facilitated extinction and attenuated reinstatement (i.e. lower levels of CS freezing) relative to eYFP counterparts, consistent with mitigated reinstatement following reconsolidation-extinction [34]. If optogenetic reactivation of a fear memory prior to extinction training facilitates extinction learning in EtOH mice (as predicted in Exp. 1B), we hypothesize that EtOH mice infused with ChR2 will show a blunted reinstatement effect relative to their eYFP counterparts. Together, studies in Experiments 1C and 1D will determine if optical activation of a previously tagged fear memory prior to extinction training will facilitate extinction and mitigate fear renewal and reinstatement in both control and EtOH mice.
Alternative Outcomes and Future Directions. The experiments in Aim 1 provide a highly novel intervention strategy for mitigating fear responses in mice by examining if chronic or brief reactivation of cells processing a discrete fear memory facilitates extinction in both control and EtOH mice. A histological examination of brain tissue following the final behavior tests will allow us to examine the overlap between cells active during fear acquisition (eYFP-positive cells) and those active during the final test (c-Fos-positive cells). In addition, these experiments provide significant training in the design, implementation, analysis, and interpretation of optogenetic experiments (a key benefit to the K01 mechanism) that will allow me to apply these techniques to answer questions about the studies outlined below.
Our working hypothesis is that chronic activation of a fear memory or a brief reactivation followed by extinction will facilitate extinction learning in both Con and EtOH mice. The results of these studies may suggest that optical manipulations of fear memories in EtOH mice are insufficient to reduce fear behavior, as seen under standard behavioral protocols [15-17]. If our optical manipulations fail to reduce fear responses in EtOH mice, we will examine additional strategies. For example, our viral strategy allows for the labeling of neurons active during discrete experiences. By opening a tagging window during ethanol withdrawal, we could target cells that undergo withdrawal-related changes and later inhibit these cells during fear learning and/or extinction. A recent preprint reports inhibition of CRF+ cells in the central amygdala or their projections suppresses activation of neural ensembles during withdrawal and decreases alcohol drinking in dependent rats [74]. Viral targeting of cells active during withdrawal will further our understanding of the relationship between withdrawal-induced changes to these neurons and behavioral outcomes. Withdrawal from a number of drugs of abuse -including alcohol, nicotine, heroin, cocaine – cause significant changes in the brain’s stress system [27]. To make our findings translatable across a variety of drugs of abuse, we will examine potential fear extinction impairments and utilize our optogenetic approaches to mitigate those impairments following withdrawal from other drugs (e.g., cocaine). We hypothesize that common mechanisms link withdrawal-induced behavioral effects across drug classes and therefore a common therapy may be broadly efficacious.
Specific Aim 2: Neural circuits mediating fear renewal and reinstatement in ethanol-withdrawn mice. Specific Aim 2 will complement, but not rely, on the findings of Aim 1 by testing the hypothesis that inhibiting basal nucleus amygdala (BA) terminals in the bed nucleus of the stria terminalis (BNST) will mitigate reinstatement, while inhibition of BA terminals in the ventral CA1 (vCA1) will mitigate renewal following extinction training in EtOH mice. This aim also serves as a launching pad for my independent career in a tenure-track position following the K01 period, which Drs. Ramirez and Eichenbaum have formally agreed in terms of permitting my findings to form my unique scientific niche to develop thereafter. The BA contains distinct populations of neurons that mediate fear acquisition and extinction [63, 75-77], with intra-BA inhibitory circuits gating the expression of fear versus extinction through its projections to the medial central amygdala, the main amygdala output nucleus [63, 72, 75]. BA neurons project to the BNST [78, 79] and the vCA1 [77], where they may modulate fear reinstatement and renewal, respectively. The BNST modulates stress and anxiety responses [80] and plays a significant role in stress-induced reinstatement of fear and drug-seeking behaviors [21, 81-83]. Likewise, vCA1 processes contextual cues in fear conditioning and mediates context-dependent fear renewal following extinction through reciprocal connections with the amygdala [70, 71]; inactivation of the vCA1 or ventral subiculum prevents context-induced reinstatement of drug seeking behavior following extinction [84, 85]. Given the central role of the BA in fear acquisition and extinction, in gating expression of fear through the CeA, and its direct projections to the BNST and vCA1, we hypothesize that blocking BA terminals originating from BA tagged ensembles, which project to the BNST or the vCA1, will reduce shock-induced reinstatement and fear renewal, respectively, in mice withdrawn from chronic alcohol.
A full description of methods can be found under General Procedures. We will conduct four separate experiments under this Aim. Briefly, EtOH and Con mice will be treated following the design of Aim 1. In Experiments 2A and 2B, we will infuse the retrograde tracer cholera toxin subunit-B (CTB)-Alexa555 into the anterodorsal BNST and CTB-Alexa647 into the vCA1 mice. Following a five-day recovery, mice will undergo ethanol treatment. Following a two-day withdrawal period, EtOH and Con mice will undergo tone fear conditioning, extinction training (4 sessions), reinstatement training, and reinstatement testing, with each session 24-hr apart (Exp. 2A). During reinstatement training, half of the mice will receive an immediate foot shock upon placement into a novel context (reinstatement) while the other half will not receive a foot shock (no reinstatement). Ninety minutes following reinstatement testing, when c-Fos expression peaks, mice will be sacrificed, brains collected and processed for c-Fos. A 2 (EtOH vs. Con) x 2 (Reinstatement vs. No Reinstatement) design will test the hypothesis that BA neurons projecting to the BNST are preferentially activated during reinstatement testing. We further hypothesize a relative increase in the number of co-labeled BAàBNST projecting neurons in EtOH compared to Con mice, in line with a predicted increased reinstatement effect [16]. We expect distinct populations of BA neurons to project to the BNST and to the vCA1, as separate populations of BA neurons project to distinct regions of the medial prefrontal cortex and the vCA1 [77].
Exp. 2B. EtOH and Con mice will undergo tone fear conditioning and extinction training as above. During an extinction retention test, half of the mice will be tested in Ctx A (extinction test) while the other half will be tested in Ctx B (renewal). A 2 (EtOH vs. Con) x 2 (Extinction vs. Renewal) design will test the hypothesis that BA neurons projecting to the vCA1 are preferentially activated during context renewal. Previous research demonstrated a non-significant trend for heightened context-dependent fear renewal in EtOH mice compared to control mice [15]. Others have demonstrated significant increases in context-mediated renewal of lever-pressing for alcohol following extinction [20, 68, 86]. Together, these studies suggest that contextual changes are likely to renew fear responses following extinction; we hypothesize that EtOH mice will show an exaggerated renewal effect relative to Con and a corresponding increase in the number of double-labeled BAàvCA1 projection neurons following renewal.
In Experiments 2C and 2D, we will infuse AAV9-c-Fos-tTA and TRE-ArchT-eYFP (ArchT) or AAV9-c-Fos-tTA and TRE-eYFP (eYFP) into the BA of ETOH and Con mice and implant optic fibers above the BNST (Exp. 2C) or the vCA1 (Exp. 2D). Mice will be taken off Dox for two days prior to tone fear acquisition in Ctx A to label cells with the inhibitory opsin ArchT or eYFP in an activity-dependent manner. Mice will then be placed back on Dox to close the tagging window. Mice will undergo ethanol exposure and a two-day withdrawal period as previously stated before undergoing four consecutive days of tone fear extinction (Ctx A).
| Figure 8. Light inactivation of BLA cells expressing ArchT during reinstatement testing reduces shock-induced reinstatement relative to eYFP controls. Insert: BLA, green=ArchT+, blue=DAPI, red=c-Fos. |
Following the last extinction session, mice will be placed into a novel context (Ctx B) and receive an immediate foot shock (Exp. 2C). The following day, mice will be placed back into Ctx A and undergo reinstatement testing under light illumination targeted at the BNST. In a 2 (ArchT vs. eYFP) x 2 (EtOH vs. Con) design, we will test the hypothesis that inactivation of BAàBNST terminals labeled during fear acquisition will mitigate fear reinstatement following extinction in EtOH and Con mice. As hypothesized above, we expect impaired extinction and heightened reinstatement in EtOH mice compared to Con mice injected with eYFP (see preliminary data). Our preliminary data suggest that inactivating tagged cells in the BLA during reinstatement testing reduces shock-induced reinstatement following extinction (Figure 8). We hypothesize that shock-induced reinstatement may activate the original fear ensemble and, through their projections to the BNST, drive reinstatement following extinction. By inhibiting BAàBNST projections, we expect to mitigate shock-induced reinstatement of fear following extinction in both EtOH and Con mice injected with ArchT relative to eYFP mice.
Exp. 2D. EtOH and Con mice will be tested for CS responses in either Ctx A (extinction test) or Ctx B (renewal) under light illumination targeted to the vCA1 one day following the last extinction session. In a 2 (ArchT vs. eYFP) x 2 (EtOH vs. Con) x 2 (Extinction vs. Renewal) design, we will test the hypothesis that inactivation of BAàvCA1 terminals labeled during fear acquisition will mitigate fear renewal following extinction in EtOH and Con mice. EtOH mice injected with eYFP will show impaired fear extinction (i.e., higher levels of freezing in Ctx A) and exaggerated renewal (i.e., higher levels of freezing in Ctx B) relative to Con mice injected with eYFP. We hypothesize that inhibiting BAàvCA1 terminals will block context-mediated fear renewal in both EtOH and Con mice injected with ArchT (i.e., similar levels of freezing in both Ctx A and Ctx B) relative to EtOH and Con mice injected with eYFP. Exps 2C and 2D predict a double dissociation, with BAàBNST mediating reinstatement, but not renewal and BAàvCA1 mediating renewal, but not reinstatement.
Alternative Outcomes and Future Directions. In Experiment 2, we hypothesize that fear reinstatement will preferentially activate BA neurons projecting to the BNST and fear renewal will preferentially activate BA neurons projecting to the vCA1. These hypotheses would be supported by an increase in c-Fos overlap in BNST or vCA1 projecting BA cells during reinstatement or renewal testing, respectively. Alternatively, reinstatement training, may preferentially activate BAàBNST cells. To examine this possibility, we will treat mice as above but sacrifice them 90-min following reinstatement training and examine c-Fos overlap with BAàBNST cells. However, in support of our hypothesis, a previous study reported that vCA1àCeA is necessary for context-dependent renewal [71]. Anatomically defined neurons within the BLA preferentially process appetitive and aversive stimuli [87]. Examination of BLA during initial drug exposure (appetitive) and following withdrawal (aversive) may demonstrate key functional changes that occur during the transition to addiction [13].
GENERAL PROCEDURES
Subjects. Wildtype male and female C56BL/6J mice (2-3 months of age; Jackson Laboratory) will be housed in groups of five same-sex mice per cage at the Center for Life Sciences and Engineering animal vivarium (Boston University). The animal facilities (vivarium and behavioral testing rooms) are maintained on a 12:12-hour light cycle (lights on at 1900). Mice will be placed on a diet containing 40 mg/kg doxycycline (Dox) for a minimum of one week before receiving surgery at age 12-16 weeks, with access to food and water ad libitum. Post-operation, mice will be individually housed and allowed to recover at least ten days. All mice will be taken off Dox (and given standard lab chow ad libitum) for an undisturbed 42 h to open a time window of activity-dependent labelling. With our virus cocktail (see below), the promoter of c-Fos – an immediate early gene used as a marker of recent neural activity – is engineered to drive the expression of the tetracycline transactivator (tTA), which in its protein form binds to the tetracycline response element (TRE). Subsequently, the activated TRE leads to the transcription of the light-responsive channelrhodopsin-2 (ChR2), archaerhodopsin (ArchT), or the non-light responsive enhanced yellow florescent protein (eYFP). Importantly, the transcription of ChR2, ArchT, or eYFP only occurs in the absence of Dox from the animal’s diet, thus permitting inducible expression of rhodopsin in correspondingly active cells. Immediately following behavioral tagging, mice will be placed back on Dox for the remainder of the study. Studies will include a minimum of 8 animal per sex per group to examine potential sex effects. All subjects will be treated in accordance with National Institutes of Health guidelines following protocols approved by the Institutional Animal Care and Use Committee at Boston University.
Viral constructs. pAAV9-TRE-ChR2-eYFP, pAAV9-TRE-ArchT-eGFP, and pAAV9-TRE-eYFP were constructed as previously described [45, 46, 88]. Our double-virus system to label active cells has previously been described [60]. pAAV9-c-Fos-tTA is combined with pAAV9-TRE-ChR2-eYFP, pAAV9-TRE-ArchT-eFGP, or pAAV9-TRE-eYFP prior to injection.
Retrograde labeling. Retrograde labeling with cholera toxin subunit B (CTB) has been previously described [89]. Briefly, we will unilaterally (right hemisphere) inject 0.2 μL of CTB-Alexa555 or CTB-Alexa647 into the BNST or vCA1 (see below) to label neurons projecting from the BA.
Stereotaxic injection and optical fiber implant. Stereotaxic injections and optical fiber implants will follow methods previously reported [45, 46, 61] with modifications noted. All surgeries will be performed under stereotaxic guidance and subsequent coordinates are given relative to bregma. Mice will be anesthetized with 1-3% isoflurane and mounted into a stereotaxic frame (Kopf Instruments, Tujunga, CA, USA) that provides a constant and adjustable flow of isoflurane. After cleaning the scalp, an incision will be made to expose the skull. Bilateral craniotomies involve drilling two windows through the skull above the injection site using a 0.5 mm diameter drill bit. Coordinates for dorsal dentate gyrus (DG) will be -2.2 mm anteroposterior (AP), ±1.3 mediolateral (ML) and -2.0 dorsoventral (DV) [46]; for the basomedial nucleus of the amygdala (BA) will be -1.7 AP, ±3.4 ML, and -4.2 DV [77]; for the bed nucleus of the stria terminalis (BNST) will be +0.2 AP, ±1.0 ML, and -4.3 DV [79]; and for the ventral CA1 (vCA1) will be -3.5 AP, ±3.5 ML, and -3.7 DV [71]. All mice will be injected with a volume of 0.2 μl of AAV9 or CTB per site at a control rate of 0.6 μl min-1 using a mineral oil-filled 33-gage beveled needle attached to a 10 μl Hamilton microsyringe (701LT; Hamilton) in a microsyringe pump (UMP3; WPI). The needle will remain at the target site for five minutes post-injection before removal. For DG targets, a bilateral optic fiber implant (200 μm core diameter; Doric Lenses) will be lowered above the injection site (-1.6 DV). For the optical inhibition of BA à BNST or BA à vCA1, optic monofiber implants will be lowered above the injection sites (-3.9 and -3.3 DV, respectively). Jeweler screws will be secured into the skull to act as anchors. Layers of adhesive cement (C&B Metabond) followed by dental cement (A-M Systems) will be spread over the surgical site. Mice will receive postoperative injections of 1.5 mg kg-1 analgesics (intraperitoneally). Mice are placed on a heating pad throughout the procedure and during recovery. Histological assessment will verify viral targeting and fiber placement. Data from off-target injections will not be included in analyses.
Optogenetic Methods. Optic fiber implants will be plugged into a patch cord connected to appropriate lasers (473 nm blue laser) controlled by automated software (Doric Lenses). For chronic stimulation studies, mice will be placed into a novel context and receive a 10-min session with laser activation (15 ms pulse width, 20 Hz) over morning and afternoon session for five consecutive days. For reactivation studies, mice will receive 3-mins of optical stimulation using the aforementioned criteria. For inactivation studies, mice will be placed into the appropriate context and receive green light (520nm, constant pulse).
Ethanol Exposure. Mice will be exposed to ethanol using the drinking in the dark (DID) paradigm, an established animal model of human binge alcohol consumption previous described [53, 73, 90]. Briefly, standard water bottles are replaced by ones containing a 20% (vol/vol) ethanol solution three hours into the dark cycle for two hours on days 1-3 and for four hours on day 4. C57Bl6 mice readily consume alcohol at this concentration and achieve blood ethanol concentrations (BEC) greater than 100 mg/dl [73]. Mice will undergo two days without access to ethanol solution (withdrawal period) prior to behavioral testing. Control mice will not receive access to ethanol solution. Water and ethanol solution bottles and body weights will be weighed daily. Studies reporting significant fear extinction impairments following chronic ethanol withdrawal have used a variety of ethanol exposure paradigms [15, 16, 18] and report BECs ranging from 80-200 mg/dl, within the rang achievable through DID. To determine BEC, blood will be collected from a small tail clip 1.5 hours following the last ethanol exposure, plasma collected, and analyzed using an Analox GL5 Analyzer (Analox Instruments, Luneburg, MA). During the withdrawal period, mice will be monitored and scored for signs of ethanol withdrawal using methods previously reported [55, 91]. Briefly, irritability to touch, abnormal gait, ventromedial limb retraction, body tremors, and tail rigidity will be scored on a 0-2 scale: 0=no sign, 1=moderate, and 2=sever. The sum of scores will be used as a measure of withdrawal severity.
Fear Conditioning Chambers. Fear conditioning will occur in one of four mouse conditioning chambers (Coulbourn Instruments, Whitehall, PA, USA) with metal-panel side walls, Plexiglas front and rear walls, a house light that provides ambient illumination, and a stainless-steel grid floor composed of 16 grid bars. The grid floor is connected to a precision animal shocker (Coulbourn Instruments, Whitehall, PA, USA) set to deliver a 2-sec 0.6 mA foot shock unconditioned stimulus (US). A speaker on one of the metal walls can deliver tone conditioning stimuli (CS). A ceiling-mounted video camera records activity and feeds into a computer running FreezeFrame3 software (Actimetrics, Wilmette, IL, USA). The software controls stimuli presentations and records videos from four chambers simultaneously. The program determines movement as changes in pixel luminance over a set period. Freezing is defined as a bout of 0.75-sec or longer without changes in pixel luminance and will be verified by an experimenter blind to treatment groups. Modifications to the conditioning chambers will create distinct context. Context alterations will include changes to spatial, olfactory, tactile, and lighting cues. The chambers will be cleaned prior to animal placement.
Fear Conditioning Behaviors. Mice will be handled for two minutes each over four days to acclimate them to the experimenters. Mice will be taken off Dox diet for 48-hrs prior to fear acquisition and will be placed back on Dox diet following acquisition. Contextual fear conditioning. Contextual fear acquisition will occur in Ctx A or B, with the order of acquisition counterbalanced. Briefly, mice will be placed into the conditioning context for a 300-sec acquisition session, including a 120-sec baseline period followed by three 0.5 mA, 2-sec foot shock USs (interstimulus interval [ISI] equals 60-sec). Context extinction will involve a 30-min stimulus-free session in the conditioning context. Contextual fear testing will involve a 300-sec stimulus-free session in the acquisition context. Tone fear conditioning. Tone fear conditioning will follow modified methods previously reported [76]. Briefly, tone fear acquisition will occur in Ctx A. Following a 120-sec baseline period, mice will receive three tone CSs (30-sec, 6 kHz, 75-80 db) that overlap and co-terminate with a 0.5 mA, 2-sec foot shock US (ISI 90-sec) for a total session length of 480-sec. Mice will undergo tone fear extinction in Ctx A, involving 15 30-sec CS presentations (ISI 60-sec). Tone extinction retention (Ctx A) or fear renewal (Ctx B) will involve five 30-sec CS presentations (ISI 60-sec) following a 120-sec baseline period. Reinstatement. For reinstatement training, mice will receive a 0.5 mA, 2-sec foot shock immediately upon placement into Ctx D and will be removed following a 60-sec post-shock period. Reinstatement testing will involve placement into the conditioning context for a 300-sec testing session (context reinstatement testing) or involve five 30-sec CS presentations (ISI 60-sec) following a 120-sec baseline period (tone reinstatement testing).
Immunohistochemistry. Immunohistochemistry follows protocols previously reported [45, 46, 61]. Mice will be anesthetized with 3% isoflurane and perfused transcardially with cold (4° C) phosphate-buffered saline (PBS) followed by 4% paraformaldehyde (PFA) in PBS. Brains will be extracted and stored overnight in PFA at 4°C. 50 μm coronal sections will be collected in serial order using a vibratome and collected in cold PBS. Immunostaining will involve washing sections in PBS with 0.2% triton (PBST) for 10-minutes (x3). Sections will be blocked for 1 hr at RT in PBST and 5% normal goat serum (NGS) on a shaker. Sections will then be transferred to wells containing primary antibodies (1:5000 rabbit anti-c-Fos [SySy]; 1:1000 chicken anti-GFP [Invitrogen]) and allowed to incubate on a shaker for 24-hrs at 4°C. Sections will then undergo three 10-min washes in PBST, followed by 1-hr incubation with secondary antibody (1:200 Alexa 555 anti-rabbit [Invitrogen]; 1:200 Alexa 488 anti-chicken [Invitrogen]). Following three additional 10-min washes in PBST, sections will be mounted onto micro slides (VWR International, LLC). Vectashield Hart Set Mounting Medium with DAPI (Vector Laboratories, Inc) will be applied to sections and slides will be cover slipped and allowed to dry overnight.
Cell counting. The number of eYFP, eGFP, and/or c-Fos immunoreactive neurons in the DG and BA will be quantified bilaterally as the average number of cells per brain region across 3-5 coronal sections (spaced 160 μm apart) per mouse. Sections that show expression in the targeted area will be selected for counting. Confocal fluorescent images will be acquired on a Leica TCS SP2 AOBS canning laser microscope using a 20X/0.70 NA oil immersion objective.
Data Analysis. Fear conditioning data are collected using FreezeFrame3 software (Actimetrics, Wilmette IL) with the bout length set at 0.75-sec and the freezing threshold initially set as described in the program instructions. Freezing is defined as changes in pixel luminance falling below a threshold. An experimenter blind to treatment groups will adjust the threshold so that freezing behavior involves the absence of all movement except those needed for respiration as previously described [1, 5-7]. Freezing behavior is scored as the percentage of time spent freezing during a given bout of time. Acquisition data will include freezing measures of baseline and training trials (ISI); extinction sessions will consist of bouts of 5-min bins. For testing trials, a single 5-min freezing measure will be taken. A repeated-measures ANOVA will examine changes in freezing over bouts for acquisition and extinction trials; between-subject factors will include virus (e.g., ChR2 vs. eYFP), treatment condition (EtOH vs. Con), and sex (M vs. F) and the within-subject factor of freezing bins. Interactions will be examined using a Newman-Keuls posthoc test. Data will be imported and analyzed using Statistica software. Data are expressed in means ± SEM. Significance is set at p<0.05.
TIMELINE AND FEASIBILITY
Table 1 summarizes experiments outlined in Aims 1 and 2 and provides estimates of the total number of mice and time to completion. Groups are outlined in the individual experiments; the number of mice needed for each is listed at 12, which will provide an n=6 per group per sex to determine potential sex effects. Experiments with 8 groups will use only males with an n=8 per group. The total number of mice needed to complete Aims 1 and 2 include additional mice to examine alternative outcomes and future directions and assumes a potential loss of 25% due to improper targeting of viruses or optic fibers, issues with surgery (including loss of head stages), or other potential issues. Estimates assume running cohorts of 16 mice at 4 weeks each, with 3 cohorts per experiment. Aim 1 experiments are expected to require 16 months to complete, with an additional 4 months for alternative outcomes and manuscript preparation. Aim 2 experiments are expected to require 12 months to complete, with an additional 4 months to explore alternative outcomes, future directions, and manuscript preparation.
Nathen Murawski 