Optogenetic investigation of neural mechanisms for alcohol use disorder

Abstract

Optogenetic techniques have been widely used in the study of neuropsychiatric diseases such as anxiety, depression, and drug addiction. Cell-type specific targeting of optogenetic tools to neurons has contributed to a tremendous understanding on the function of neural circuits for future treatment of neuropsychiatric disorders. Though optogenetics has been widely used in many research areas, the use of optogenetic tools to uncover and elucidate neural circuit mechanisms of alcohol’s actions in the brain are still developing. Here in this review article, we will provide a basic introduction on optogenetics and discuss how these optogenetic experimental approaches can be used in alcohol studies to reveal neural circuit mechanisms of alcohol’s actions in regions implicated in the development of alcohol addiction.

1. Introduction:

1.1. The concept of optogenetics

Optogenetics is a biological technique that employs genetically-modified, light-sensitive microbial ion channels or pumps expressed in cells of live tissue and animals. Neuroscientists have employed optogenetics to modulate the activity of specific populations of brain cells (Deisseroth et al., 2006). In 2005, Karl Deisseroth’s group first published the use of a rapidly gated light-sensitive cation channel found in algae, a type of channelrhodopsin (ChR) known as Channelrhodopsin-2 (ChR2), to achieve noninvasive, optogenetic control of neuronal activity with light at millisecond precision (Boyden et al., 2005). This remarkable finding had a transformative impact on the field of neuroscience. Since then, the refinement and development of optogenetic tools has propelled the neuroscience field to map out complex neural circuits or specific neurons that mediate behaviors.

Our brain neural networks are complicated. There are a myriad of neuronal subtypes with unclear neural circuit function (Deisseroth, 2015). In 1979, Francis Crick had first proposed the idea that in order to understand neural function better, “we need to find a method by which all neurons of just one type could be inactivated, leaving the others more or less unaltered to know how our brain works” (Deisseroth, 2015). There was an advancement of techniques that could generate transgenic mice and development of tools to spatially target specific brain regions using surgically implanted cannulae to pharmacologically manipulate receptors or ion channels; however, neither of those could modulate activation or inhibition of specific neuron subtypes with a spatiotemporal precise manner (Deisseroth, 2011). Later, scientists had developed various ways to achieve approaches of controlling neural activity by light (Lima and Miesenbock, 2005; Zemelman et al., 2002). Though great progress was made regarding the manipulation of activating neurons, precise targeting of subpopulations of neurons, as well as the ability of expression in deep layers of brain tissue remained difficult.

It was a long-sought goal in the neuroscience field to find a way to precisely control neuronal activity by using light. In the 1970s, microbiologists had found proteins called microbial rhodopsins, which were sensitive to light and could control membrane potential (Oesterhelt and Stoeckenius, 1973). In 2005, the Deisseroth group first expressed the light-sensitive channel ChR2 in neurons (Boyden et al., 2005). Since then, optogenetics has developed fast and been applied in different areas.

Rhodopsins are a key element in optogenetics. There are two types of rhodopsins in nature—microbial rhodopsins (Type-1 rhodopsins) and animal rhodopsins (Type-2 rhodopsins) (Ernst et al., 2014). The latter, Type-2 rhodopsins, belong to the G-protein coupled receptor (GPCR) family, which upon light illumination, catalyze GDP to GTP exchange in heterotrimeric G proteins within the cells. Type-1 rhodopsins are light-sensitive photo-receptors expressed in the membrane. Its structure is comprised of seven α-helices with retinal chromophores. Type-1 rhodopsins can be categorized into three light-gated forms: proton pumps, ion pumps and ion channels. In neuroscience, researchers have mostly employed Type-1 ion channel rhosopsins, such as ChRs, namely ChR2, and Type-1 ion pumps, such as halorhodopsins (NpHRs) (Fig. 1) (Deisseroth, 2015). These are two optogenetic tools that have opposing function. ChR2 absorbs blue light and activates neurons through the conductance of cations through a pore, while NpHRs inhibit neuronal activity after stimulation with green/yellow light which pumps in chloride (Fig. 2). Moreover, because there is little overlap in the absorption spectrums for stimulation of ChRs (470 nm, blue light) and NpHRs (590 nm, green/yellow light), they can both be used simultaneously in experiments (Zeng and Madisen, 2012; Zhang et al., 2007).

Figure 1. Rhodopsins found in nature.

Type-1 rhodopsins contain photo-receptors expressed in the membrane and can be categorized into three forms: proton pumps, ion pumps and ion channels. Type-2 rhodopsins are a type of G-protein coupled receptor that use light to activate G-proteins within the cells.

Figure 2. Channelrhodopsin-2 and Halorhodopsin are used to activate or inhibit cells.

Channelrhodopsin-2 (ChR2) is a blue-light activated cation channel that conducts cations through its pore to depolarize membrane potentials. Halorhodopsin is a chloride pump that is activated by green/yellow light to inhibit the cell through the inward movement of chloride ions.

The ability to package these ChRs and NpHRs genetic vectors in viral vectors such as adeno-associated viruses (AAVs), allows researchers to attain spatial resolution for cell populations they wish to manipulate in the brain in vivo. Researchers can also genetically isolate specific cell populations to transfect these AAV-packaged rhodopsins by using the Cre-lox expression system. Restricting the expression of these rhodopsins using Cre-inducible lox sites flanking the gene encoding the rhodopsin achieves cell specificity when combining AAV and Cre transgenic animals. Most of the time, the virus will carry strong promoters like the elongation factor-1 α (EF1-α) promoter, two loxp sites as well as a target rhodopsin. For example, tyrosine hydroxylase (TH)-Cre transgenic mice have usually been widely used in isolating the dopaminergic circuit because the enzyme, Cre recombinase, is restricted to dopamine (DA) producing cells (TH positive neurons). After injecting an AAV virus, which carries a genetic sequence composed of: EF-1α promoter, two loxp sites in the opposite direction that are Cre-inducible, and a downstream target gene such ChR2 tagged with an enhanced yellow fluorescent protein (eYFP) into the ventral tegmental area (VTA) of TH-Cre mice, ChR2 will be specifically expressed in VTA TH-positive neurons, isolating the expression of ChR2 from TH-negative neurons, such as interneurons, in the VTA (Fig. 3). The implantation of a light-weight optic fiber above the VTA delivers light to the region of interest. This is an example in which neuroscientists can explore the function of genetically defined cell populations in a neural circuit (Fig. 3) (Chaudhury et al., 2013).

Figure 3. Combining viral-mediated gene transfer and transgenic mice to isolate specific neuronal populations.

One example of cell-specific transfection of ChR2 or NpHR is how researchers isolate dopamine producing neurons in mice. Tyrosine hydroxylase (TH)-Cre mice can be injected with adeno-associated viruses with ChR2 or NpHR under Cre recombinase control in neural substrates with TH to restrict ChR2 or NpHR to dopamine producing populations.

1.2. Applications of optogenetics

As discussed above, optogenetic techniques have been rapidly developing and its application to neuroscience research has grown exponentially in the past decade [4 papers in 2007 and 880 papers in 2017, searched by (optogenetic OR optogenetics) in PubMed]. It offers enormous potential for basic research by providing new opportunities to understand neural circuit function and control over behavior. Cell-type specificity, a typical feature employed in optogenetic research, selectively modulates the activity of genetically defined neuronal populations. This has allowed scientists to approach the elucidation of neural circuit connectivity in complex neural systems as well as regulating activity of specific neurons to monitor behaviors in free-moving animals (Stuber et al., 2012).

As we have already discussed, optogenetics generally refers to light-sensitive ion channels or pumps that enable precise spatial and temporal control of neuronal activity. There has also been the development of a variety of light-activated GPCRs (OptoXRs), which can modulate neural activity through G-protein signaling (Airan et al., 2009). Although optogenetics has had a remarkable impact in neuroscience, its application in vivo is relatively invasive due to surgeries and implantation of optic fibers that are often tethered to a light source. It is difficult to manipulate multiple brain regions with optogenetics at the same time because researchers are often limited by the amount of weight of optic fibers and space of the skull of the model organism. There has now been a rapid development of lightweight, implantable subdermal light sources on the skull that deliver light through a fiber directed to a region of interest (Park et al., 2015).

Another technique, which is similar to optogenetics in its ability to specifically target certain cell populations to alter neural activity, is chemogenetics. This tool also has been designed to control genetically defined neurons of freely-moving animals (Deisseroth, 2015; Roth, 2016). However, chemogenetics is less invasive to brain tissue, aside from the viral delivery during surgery. Chemogenetics employs a class of modified GPCRs that have been designed to respond to synthetic ligands or molecules, also known as designer receptors exclusively activated by designer drugs (DREADDs) (Armbruster et al., 2007; Roth, 2016). These DREADDs can be comprised of either stimulatory or inhibitory G-proteins that can modulate neural activity via modulation of downstream intracellular signaling cascades by the exposure to “designer drugs” (Boesmans et al., 2018; Sternson and Roth, 2014). Notably, a disadvantage of chemogenetics is the loss of temporal acuity in the modulation of neural activity and the potential for unwanted long lasting activation of second-messenger systems. Nevertheless, information garnered from well designed chemogenetic studies have increased researchers’ understanding of neural circuits involved in a variety of behaviors. We enthusiastically refer the reader to a comprehensive review on chemogenetics and alcohol research that is also published in this special issue: The use of chemogenetic approaches in alcohol use disorder research and treatment by Yifeng Cheng and Jun Wang.

In recent years, Ken Berglund’s laboratory has developed tools called Luminopsins (LMOs), which are composed of fusion proteins of a light-emitting luciferase and a light-sensing opsin. LMOs are new tools that help to achieve combined optogenetics and chemogenetics manipulation (Berglund et al., 2016; Deisseroth, 2015) which make possible precise and temporally acute neuronal activation, as well as chronic and less invasive control of entire populations throughout the brain through the same molecules. The tool box for the development of improved activation of LMOs is currently under development. Additionally, the development of magnetogenetics is a potential new technology for the manipulation of neural circuits (Chen et al., 2015; Huang et al., 2010). Magnetogenetics uses the co-expression of either thermo-sensitive or mechano-sensitive modified TRPV ion channels in conjunction with ferritin nano-particles to gate ionic currents with strong magnetic fields (Nimpf and Keays, 2017). Some disadvantages include heat production and the need to build magnetic coils modified for in vivo modulation. The exact biophysical mechanism of ferritin nano-particle activation of these channels has been disputed and the need for further replication by other investigators is needed (Meister, 2016).

Similar to the combined use of optogenetics and chemogenetics, other useful techniques such as fast scan cyclic voltammetry (FSCV), electrophysiology, as well as in vivo calcium imaging, are also being used together with optogenetics and applied in various studies on neurological and neuropsychiatric disorders (Barker et al., 2017; Calipari et al., 2017; Juarez et al., 2017; Krishnan et al., 2007; Melchior and Jones, 2017). The combined use of techniques is enabling scientists to have a better understanding of the functions in complex neural systems.

The alcohol neuroscience field aims to understand the mechanisms of alcohol’s action on the brain throughout the development of alcohol addiction in order to tackle the devastating problem alcohol abuse has on society. Though alcohol is commonly consumed throughout the world, the development of pathological alterations across precise neural circuits or cell types that control different stages of alcohol abuse still needs to be elucidated. Optogenetic tools have developed fast during the past decade. Although optogenetics hasn’t been as widely used as it is in other fields of neuroscience, the advantage of specific cell-type regulation will make it play a critical role in future alcohol studies. Yet, there already have been some works applying these tools. Here we will give a brief introduction on those works, as well as the applications that have played an important role in mapping out specific neural circuits or cell types in different brain regions that are essential to alcohol drinking behaviors.

2. The ventral tegmental area

The VTA is a key region for the processing of stimuli salience, association and value of stimuli (Russo and Nestler, 2013). The VTA is comprised of functionally heterogeneous DA neurons and inhibitory GABAergic interneurons (Juarez and Han, 2016; Juarez et al., 2017; Morales and Margolis, 2017; Walsh and Han, 2014). These DA neurons project to neural substrates of the mesocorticolimbic circuit that are involved in encoding the salient and associative cues for stimuli, executive control and emotion (Juarez and Han, 2016; Russo and Nestler, 2013). VTA DA neurons engage in temporally specific firing patterns to encode aspects of stimuli. Accurately timed transitions between slow, irregular tonic firing to rapid short bursts of firing of VTA DA neurons are important for associative learning and reinforcement of stimuli that in turn regulates an organism’s future behavioral responses for survival (Schultz, 2016).

Dysfunction of fine-tuned VTA signaling has been implicated in many neuropsychiatric disorders, including alcohol use disorder (Koob and Volkow, 2010). Drugs of abuse, such as alcohol, pathologically alter DA signaling to lead to aberrant neural communication of the mesocorticolimbic circuit (Juarez and Han, 2016; Koob and Volkow, 2010). Early research has demonstrated that acute ethanol (EtOH) increases DA neuron activity, including phasic activity, and DA concentrations in downstream neural substrates, implicating DA neurons in the EtOH reward (Brodie et al., 1999; Di Chiara and Imperato, 1985). EtOH has been shown to directly alter DA neuron activity via intrinsic actions of the neuron as well as by altering the strength of extrinsic excitatory and inhibitory inputs onto the neuron (Brodie et al., 1999), (Theile et al., 2008). Interestingly, many studies have reported subpopulations of VTA DA neurons that failed to respond to the excitatory effects or plasticity-inducing effects of EtOH, which have been outlined in our recent review (Juarez and Han, 2016). Additionally, there are incongruent reports of sensitization and tolerance to EtOH following chronic administration (Brodie, 2002; Okamoto et al., 2006). This suggests that EtOH may have distinct actions on subpopulations of DA neurons in the VTA; however isolating subpopulations of DA neurons for in depth circuit investigations to identify causal mechanisms between dopaminergic activity and alcohol drinking behaviors was difficult.

Recently, with the use of advanced circuit-dissecting optogenetic, electrophysiological and immunohistochemical techniques, researchers have been elucidating the incredible functional diversity of VTA DA neurons (Chaudhury et al., 2013; Friedman et al., 2014; Juarez et al., 2017; Lammel et al., 2008; Lammel et al., 2012). Understanding functional diversity could explain how the VTA mediates a constellation of behaviors. Investigators have discovered that sub-populations of projecting VTA DA neurons receive distinct inputs from cortical and subcortical regions. This specific input-output organization sorts the type of information the sub-populations of VTA DA neurons process and regulate (Lammel et al., 2012). Furthermore, the VTA has been shown to send GABAergic and glutamatergic projections to regions of the mesocorticolimbic system, and projecting DA neurons themselves have been shown to co-release GABA and glutamate (Morales and Margolis, 2017).

Using optogenetics, investigators are now exploring in more detail how VTA DA neurons regulate the reinforcing properties of EtOH, which can lead to a better understanding on the progression of pathological consumption of alcohol and factors that regulate alcohol relapse. The combination of behavioral paradigms with face or construct validity for excessive and/or compulsive alcohol consumption in concert with advanced methods to monitor and manipulate neural activity has allowed the alcohol field to unravel the intricacies underlying EtOH’s actions on the reward system. For example, in an effort to determine how DA release in the nucleus accumbens (NAc) from projections originating in the VTA regulates alcohol drinking, Bass and colleagues transfected DA neurons in the VTA of alcohol drinking rats with ChR2 (Bass et al., 2013). During a drinking task, the investigators then stimulated DA neurons of the VTA with either a tonic (5 Hz) stimulation pattern to increase dopamine tone in downstream targets or a phasic (50 Hz) stimulation to induce large transient bursts of DA release in downstream targets. They discovered that while phasic DA activation did not alter alcohol consumption during the drinking task, tonic dopamine stimulation did induce a reduction in EtOH consumption.

In a different animal model of alcohol consumption, the Han lab has sought to understand how VTA DA neurons regulate individual drinking behaviors. Using a continuous access, two-bottle choice alcohol-drinking model, we are able to generate low and high alcohol drinking populations within a genetically identical mouse line, similar to other research groups (Juarez et al., 2017). This model provides investigators with the unique opportunity to determine the neurophysiological and epigenetic mechanisms of individual alcohol drinking behavior, independent of diverse genetic backgrounds that may result in behavioral predispositions. Using anesthetized in vivo electrophysiology, Juarez and colleagues discovered that after chronic alcohol drinking, low alcohol drinking mice had an increase in VTA DA neuron firing rate and phasic activity, a phenomenon that was not observed in high alcohol drinking mice.

In order to determine if this increase in phasic activity of VTA DA neurons of low alcohol drinking mice was causal to lower alcohol-drinking behaviors in a continuous access, two-bottle choice alcohol drinking paradigm, these investigators turned to cell-specific expression of excitatory ChR2 in VTA DA neurons. It was discovered that 15 minutes of 20 Hz phasic activation of VTA DA neurons in high alcohol drinking mice reduced alcohol-drinking behaviors for up to 72 hours after the stimulation. Furthermore, using circuit-dissecting electrophysiology and optogenetics, it was found that the increase in activity was specific to the NAc projecting VTA subpopulation of neurons, and not those that project to the medial prefrontal cortex (mPFC). Specifically activating the NAc projecting VTA subpopulation reduced alcohol-drinking behaviors for up to 48 hours (Juarez et al., 2017).

Interestingly, in this model, neither 0.5 Hz tonic or 5 Hz tonic stimulation of VTA DA neurons reduced alcohol drinking behaviors following stimulation, although alcohol drinking behaviors were not monitored during the optical stimulation, which could explain the seemingly disparate results between this study and the study by Bass and colleagues (Bass et al., 2013; Juarez et al., 2017). Furthermore, the phasic induced reduction of alcohol drinking behaviors was specific to mice that were classified as high alcohol drinkers, because no changes in alcohol preference or consumption were observed in EtOH naïve mice (Juarez et al., 2017). This suggests a pattern-specific, projection-specific, and context-specific role of VTA dopaminergic regulation of individual alcohol drinking behaviors. The findings from these two groups highlight how important it is that as the alcohol field embarks on a more prevalent use of optogenetic tools to elucidate behaviors, we must interpret results across research groups carefully.

Optogenetically increasing activity of VTA DA neurons has been shown to induce long-lasting molecular and neurophysiological effects that underlie behavioral changes in response to stress (Chaudhury et al., 2013; Friedman et al., 2014; Walsh et al., 2014). Optogenetic manipulations have also been used as a proxy for replicating the long-lasting behavioral, morphological and molecular effects of ketamine (Fuchikami et al., 2015). When using optogenetics, if a long-lasting behavioral effect is observed, it is important to further determine what the molecular mediators of these long lasting alcohol drinking reductions are. This could help the field identify novel pharmacotherapeutic targets for the treatment of excessive alcohol consumption. In our own studies that parsed out low and high alcohol drinking groups, we observed differences in hyperpolarization-activated cyclic nucleotide-gated channel (HCN, conducts inhibitory I_h) function between the drinking groups. High alcohol drinking mice displayed lower I_h while low alcohol drinking mice had a potentiated I_h (Juarez et al., 2017). We hypothesized that these differences in an inhibitory current may be the cause of the neurophysiological differences we observed in our alcohol drinking mice. It would be interesting to determine if the long-lasting optogenetic experiments that reduced high alcohol drinking behaviors altered HCN channel function. Future directions to continue these studies include a more in depth investigation into these effects. These results and our hypothesis also underline how investigators should approach results that have long-lasting effects.

Moreover, there is a great opportunity in the field now to understand how dopaminergic sub-populations that project to other regions of the mesocorticolimbic system regulate alcohol preference and consumption throughout the progression of alcohol addiction. Finally, more studies are needed to elucidate how specific inputs onto VTA DA neurons regulate alcohol induced plasticity mechanisms. Understanding how diverse cortical and subcortical regions alter information processing of VTA DA neurons is important for determining how EtOH alters the mesocorticolimbic system. Optogenetics is providing the field with such opportunities to conduct these advanced investigations.

3. The nucleus accumbens

The NAc has been an interest in addiction research for years. The NAc is a major neural substrate target of DA from the VTA (Russo and Nestler, 2013). In regards to alcohol, the NAc has been implicated in the reinforcing properties of low doses of EtOH (Koob, 1992; Koob and Volkow, 2010). EtOH was shown to increase extracellular DA release in the NAc (Di Chiara and Imperato, 1986; Imperato and Di Chiara, 1986). Blocking DA signaling through the infusion of a DA receptor antagonist into the NAc reduces operant self-administration of EtOH (Rassnick et al., 1992; Samson et al., 1991). However, there have been conflicting reports of the exact role the NAc plays in EtOH reward and reinforcement. There have been mixed results from early studies that attempted to lesion DA signaling in the NAc with 6-hydroxydopamine (6-OHDA) and observe alcohol consumption behaviors (Brown and Amit, 1977; Myers and Quarfordt, 1991; Richardson and Novakovski, 1978). While this could be due to differences in strains of animals investigated, the size of the lesioned area and the model of alcohol consumption employed, another potential reason for such conflicting reports may be due to the structural and cellular diversity of the NAc itself. Here, we briefly describe the heterogeneity of the NAc, and then delve into how optogenetics has allowed researchers to discretely parse out the many functions of the NAc and its role in alcohol reward and reinforcement.

The NAc is a ventral structure of the basal ganglia that is comprised mostly of projecting GABAergic medium spiny neurons (MSNs). Similar to those MSNs in the dorsal striatum, these neurons express GPCRs that bind DA to stimulate two main classes of G-proteins: the stimulatory G_s/α-coupled D1-like receptors (D1 and D5) and the inhibitory G_io/α-coupled D2-like receptors (D2, D3 and D4). Overall, MSNs of the NAc are investigated as two subpopulations, those that express DA receptor D1 (D1DR), which are believed to be excitatory, and those that express DA receptor D1 (D2DR), which are believed to be inhibitory. However, the NAc does have MSNs that either coexpress D1 and D2 or express other types of DA receptors, but for the purpose of this review we will limit our discussion to D1DR or D2DR expressing MSNs (Hikida et al., 2016; Lobo et al., 2010).

Changes in D2DR MSN activity of the NAc have been implicated in the progression of alcohol addiction. Raclopride, a selective D2DR antagonist was shown to reduce lever responding for EtOH in rats (Samson et al., 1991). Moreover, overexpressing D2DR in the NAc of rats reduces EtOH self-administration behaviors (Thanos et al., 2001). In humans, researchers have used positron emission tomography (PET) imaging, to further elucidate D2DR function across the brain. Interestingly, researchers observed a possible protective effect of increased D2DR expression in non-addicted members from alcoholic families (Volkow et al., 2006). A DA-deficit hypothesis to explain the progression of alcohol addiction suggests that the restoration of DA signaling could be an effective therapeutic treatment (Koob and Volkow, 2010; Volkow et al., 2007). Along with projecting MSNs, the NAc also contains two main types of interneurons, GABAergic interneurons and choline acetyltransferase ChAT) expressing interneurons (Lenz and Lobo, 2013). These two types of interneurons are important for modulating the activity of projecting MSNs as well as dopamine release from dopamine terminals from the VTA. Optogenetic tools have helped elucidate the roles these interneuron populations play in modulating NAc activity.

Previously, it was believed that D1DR MSNs selectively projected to output regions of the basal ganglia, such as the ventral midbrain, via the direct pathway and D2DR MSNs indirectly signaled to these output regions via the ventral pallidum. Yet, recently, researchers have been able to combine transgenic mice and circuit-dissecting optogenetic techniques to characterize these projections. One such study sought to determine the selectivity of D1DR and D2DR MSNs to the ventral midbrain or the ventral pallidum. These researchers discovered that while the ventral midbrain exclusively receives synaptic input from D1DR MSNs of the NAc, the ventral pallidum receives an almost equal contribution of GABAergc synaptic input from both D2DR MSNs and D1DR MSNs (Kupchik et al., 2015). The use of optogenetics to further delineate the projection patterns of these two projecting MSN types continues to advance the field, and our understanding, of the function of the NAc.

Additionally, The NAc is often divided into structural subtypes, the NAc core and the NAc shell, which is often divided into the medial and lateral shell. These subregions contain both D1DR and D2DR MSNs and are thought to mediate different aspects of reward, reinforcement and motivation. These subregions also receive distinct dopaminergic innervation from the VTA and are thought to integrate appetitive and aversive stimuli (Al-Hasani et al., 2015; Lammel et al., 2012). Some studies have found that lesioning the core or shell has effects on a reduction of alcohol consumption (Cassataro et al., 2014; Dhaher et al., 2009). Yet, interpretation of important alcohol research listed earlier that used lesions may have been limited because of inability to consistently restrict lesions to subregions. Optogenetics can allow researchers to refine their investigations and more closely determine how these subregions are modulated by EtOH.

The NAc receives major dopaminergic innervation from the VTA. In the past, researchers investigated whether DA signaling in the NAc via FSCV with electrical stimulation is affected by acute and chronic EtOH exposure. Recently, however, the limitations of using electrical stimulation to induce DA release have been elucidated with the help of optogenetics. Electrical stimulation not only recruits the activation of dopaminergic terminals, but it can activate other cell types that, unless using advanced pharmacological techniques to isolate DA terminals, may affect DA release (McElligott, 2015).

Optogenetic manipulations of a variety of inputs into the NAc have revealed the diversity of function in alcohol related studies. Using a combination of FSCV recordings and optogenetic stimulation of DA release in acute slices of the NAc core or medial shell, investigators have discovered subregion specific modulation of dopamine signaling following chronic intermittent ethanol exposure (CIE). CIE did not cause changes in DA release in the core or the shell, but it did increase DA uptake rate from the synapse specifically in the medial shell (Melchior and Jones, 2017).

The NAc also receives glutamatergic input from a variety of cortical and subcortical regions of the brain, including the prefrontal cortex (PFC), the hippocampus (HPC), the amygdala, and the thalamus (Bagot et al., 2015; Christoffel et al., 2015; Stuber et al., 2011) (Ji et al., 2015). Each of these upstream neural substrates mediate specific behaviors, thus it may have been hypothesized that not all glutamatergic input into the NAc relays the same information. Specific optogenetic manipulation of these regions to investigate synaptic strength (long term potentiation (LTP) or long term depression (LTD)) has revealed the diverse function of these inputs (Ma et al., 2018).

A classic study demonstrated that glutamate signaling via NMDARs in the NAc is important for the reinforcing properties of EtOH (Rassnick et al., 1992). Yet the mechanism of how EtOH influences differential glutamatergic inputs into regions of the NAc had been difficult to resolve. In regards to aversion-resistant EtOH consumption, a new form of NMDAR adaptation was discovered, where hyperpolarization active NMDARs were found in PFC→NAc core synapses and insula→NAc core, but not amygdala→NAc core synapses (Seif et al., 2013). In an optogenetic investigation into how glutamatergic plasticity in the NAc mediates aversion-resistant alcohol consumption, researchers discovered that modulation of glutamatergic activity with D-serine, a coagonist for NMDARs, infused into the NAc core, but not the dorsal striatum, was sufficient to reduce aversion-resistant alcohol consumption (Seif et al., 2015).

Additionally, using patch clamp electrophysiology combined with ChR2 expression in glutamatergic terminals from the PFC, HPC or amygdala to determine how the weight of synaptic strength of diverse glutamatergic inputs is altered after acute EtOH exposure, Ji and colleagues discovered differential synaptic strengthening on MSNs of the NAc (Ji et al., 2015). Acute EtOH bath application from brain slices of EtOH naïve mice blocked time-dependent LTP but enhanced LTD in amygdala→NAc synapses specifically, while no differences were observed in HPC→NAc or PFC→NAc synapses. Interestingly, in a later study, these researchers also discovered that EtOH experience can modulate this strengthening of synapses. Acute EtOH bath application on NAc slices from mice that had undergone the drink-in-the-dark binge alcohol drinking paradigm induced LTP occurring at all glutamatergic synapses (Ji et al., 2017).

As researchers continue to define the function of these NAC subregions using circuit dissecting optogenetic tools, the field will continue to grow and redevelop its hypothesis of how EtOH regulates specific parts of the development of alcohol addiction.

4. Amygdala

The amygdala is another critical brain region for many aspects of alcoholism. As we had already known, alcoholism is a chronically relapsing disorder that is composed of three stages: binge/intoxication, withdrawal/negative affect, and preoccupation/anticipation (Koob and Volkow, 2010). The amygdala is considered an important element in mediating these stages at the anatomical level. In the binge stage, the mesocorticolimbic DA system plays a key role in regulating the reinforcing effects of alcohol drinking behaviors. Related neuronal circuits are VTA→NAc and VTA→extended amygdala. A variety of neurotransmitters including GABA, opioids, DA, glutamate, neuropeptide Y and glucocorticoids of the hypothalamus-pituitary-adrenal (HPA) axis are involved in this reinforcing alcohol behavior (Koob and Le Moal, 2001; Kranzler et al., 2006). When approaching the withdrawal stage, the extended amygdala is a focal point for this stage, also functioning as a regulator of motivational behavior. It is hypothesized that in the withdrawal stage, there is decreased activity in reward neurotransmitter systems and a recruitment of “anti-reward” systems, such as those that release corticotrophin releasing factor (CRF). The preoccupation/relapse stage is the last stage of the addiction cycle that involves impulsive and compulsive behaviors. However, the changes of neurotransmitter systems are different from the previous two stages. The reward and anti-reward systems are both reactivated during chronic relapse (Kranzler et al., 2006). Based on previous works and hypotheses, numerous studies have focused on the mechanisms and functions of the amygdala in manipulating different alcohol behaviors.

Several approaches combining the use of transgenic animals, complete ablation of neural substrates through lesion techniques, pharmacological targeting of receptors or ion channels, immunohistochemical mapping of molecular targets of interest, and electrophysiology have been applied to investigations to further understand the function of complex neural circuits involving the amygdala. For example, there was a study geared towards understanding whether the central extended amygdala, composed of the central amygdala (CeA) and the bed nucleus of the stria terminalis lateral posterior (BNST-LP) was associated with increased alcohol consumption when undergoing CIE (Dhaher et al., 2008). These investigators electrolytically lesioned either the region of CeA or BNSTLP to see if any behavioral changes occurred in a model of EtOH drinking that combined a two hour, two-bottle choice test followed by CIE and then another two hour, two-bottle choice test. Overall, this study found that while lesioning the BNSTLP can reduce baseline EtOH consumption and lesioning the CeA can reduce baseline EtOH preference, neither can prevent the CIE induced increase in alcohol drinking behaviors.

In another study, experiments were conducted to explore the role of Group II metabotropic glutamate receptors (mGluRs, specifically mGluR2/3) in the HPC and amygdalar regions had in attenuating stress- and cue-induced EtOH-seeking behavior (Zhao et al., 2006). By employing pharmacological tools, they selectively activated mGlu2/3Rs with agonists. Attenuated EtOH-seeking behaviors were observed in conditions of both drug stress cues. By immunohistochemically mapping the induction of an immediate early gene generated protein, c-fos, across different brain regions, researchers can identify brain regions involved in regulating responses of alcohol related behaviors. Similar methods were also used in other studies (Cannady et al., 2011).

Other works have focused on identifying the neurophysiological adaptions induced with alcohol exposure using electrophysiology. One study found that L-type voltage-gated calcium channels (LTCCs) were involved in EtOH’s ability to increase neuron firing of the CeA in EtOH naïve rats (Varodayan et al., 2017). However, in an EtOH dependent model, this LTCC-based mechanism of EtOH’s ability to increase firing was disrupted. They found that in alcohol dependent animals, the CRF receptor type 1 (CRFR-1) was now responsible for neuronal responses to EtOH in the CeA and also escalated alcohol intake of EtOH-dependent rats. However, some other studies achieved the approaches of verifying studies on both in vitro and in vivo levels by combining microdialysis techniques with electrophysiology (Roberto et al., 2004).

As we described above, a lot of work has been focused on the amygdala. Though in recent years, remarkable progress has been made by using a variety of these techniques, some technical limitations still exist. First, most of the previous amygdala alcohol literature has observed neuronal excitability differences that manifest following alcohol exposure when compared to EtOH naïve mice. However, many studies had not been able to determine whether these neural differences were innate individual differences or if they are alterations that occur across the development of alcohol related pathological behaviors. Second, few studies have applied tools to dissect out the functional neuronal circuits of the amygdala in freely behaving animals with the temporal precision that optogenetics can lend. Third, a body of evidence has shown that different neuronal subtypes of amygdala may have different functions on manipulating alcohol behavior (Silberman and Winder, 2015). Previous techniques such as transgenic animals, pharmacology and electrophysiology may still have limitations on achieving the approaches of precisely examining the function of neuronal subpopulations. However, combining these techniques with optogenetics can enable researchers to resolve those limitations.

One study sought to understand the specific neuronal circuit function of BLA→NAc shell in a Pavlovian cue-induced alcohol seeking behaviors. They discovered that optogenetic activation of this neural circuit during the presentation of the conditioned cue disrupts alcohol seeking (Millan et al., 2017). In other fields of neuroscience, progress has been made in understanding how subcircuits of the amygdala mediate anxiety-like behaviors. Optogenetics has provided researchers with a tool to dissect the subnuclei of the amygdala, including the basolateral amygdala, the central amygdala lateral portion and the central amygdala medial portion (Tye et al., 2011). Another study used optogenetics and in vivo electrophysiology to explore the mechanisms and functions of two interneuron subtypes of the amygdala, the genetically distinct the parvalbumin and somatostatin populations, and their role in fear learning (Wolff et al., 2014). These interesting studies demonstrate the advantages of applying optogenetic tools in manipulating neuronal circuits specifically.

5. Habenula

The habenula is a small brain structure that has been increasingly studied due to its role in regulating both emotional behaviors (such as anxiety, reward, depression as well as drug withdrawal) and as a regulator of neuromodulatory systems, including the DA and serotonin systems. It is located posterior-dorsal-medial to the thalamus and is divided into the lateral habenula (LHb) and the medial habenula (MHb) (Hikosaka, 2010). The LHb receives inputs from limbic regions and the basal ganglia. The LHb then projects to the interpeduncular nucleus (IPN), rostromedial tegmental nucleus, substantia nigra pars compacta and the VTA. On the other hand, the septum is the main input to the MHb and its outputs mostly project to the IPN (Fowler et al., 2011; Lecca et al., 2014; Molas et al., 2017). Based on its inputs, it is hypothesized that the habenula is involved in relaying or encoding negative or aversive stimuli. That means when involving a reward, habenula activity will be generally diminished, and if the stimuli is negative or aversive, habenula activity will be increased (Velasquez et al., 2014).

In the alcohol field, there is already a body of work that has described the vital role of the habenula, especially the LHb, in encoding rewarding and aversive aspects of stimuli. Studies in the past have used various approaches such as pharmacology, lesions of the LHb region and/or deep brain stimulation. Electrophysiological recordings were used to investigate whether neuronal activity had altered in vitro, in combination with animal behavioral tests to explore the functional role of the habenula in alcohol withdrawal or relapse behaviors (Kang et al., 2018). Though these works have revealed a crucial role of the habenula in alcohol drinking behaviors, the field is now primed to combine findings from other researchers elucidating subcircuits of the LHb with alcohol paradigms to determine the mechanisms the habenula plays in encoding different drinking behaviors. By applying viral-mediated tracing, and optogenetic manipulations and synaptic electrophysiology methods, it was discovered that the LHb→VTA circuit and the laterodorsal tegmental nucleus→VTA circuit are responsible for two opposing behaviors, aversive and appetitive respectively (Lammel et al., 2012). This finding added to the field of knowledge that the habenula has a role in manipulating DA activities. To reach such precisely neuronal targeting, optogenetics offers great opportunities and support in experimental studies.

2. Conclusion

In this review, we have discussed recent literature that has employed cutting edge optogenetic techniques to further understand neural circuits that are involved in alcohol use disorder. The advantages of cell specificity and circuit dissection in the use of optogenetics has helped the field understand neuronal populations that underlie alcohol consumption and those that are neurophysiologically modulated by alcohol exposure. The field of optogenetics and alcohol research is rapidly expanding. Further targeted investigations focused on the hypothesis of how the progression of alcohol use disorder is a neural circuit dysfunction involving the recruitment of distinct mesolimbic (binge/intoxication), amygdalar (withdrawal) and cortical (craving and executive function) circuits would be improved with the employment of optogenetics. Additionally, optogenetics could help the field delineate the neural circuits involved in the co-morbidity of alcohol use disorder with other disorders, such as anxiety or major depressive disorders.

In the future, it would also be beneficial to determine if the same neural circuits that modulate alcohol consumption and seeking are generalized to other appetitive consummatory stimuli, such as sucrose. Investigations have already begun to delve into this question (Millan et al., 2017) (Mikhailova et al., 2016). These studies have identified mesolimbic and limbic-striatal circuits that regulate alcohol and sucrose seeking. Additionally, an interesting phenomenon in some optogenetic studies is a long-lasting effect on behavior following the stimulation. The identification of molecular mediators of these long-lasting effects could help the field further understand how neuronal stimulation in a context-specific way regulates behavior.

As optogenetic tools continue to develop to provide more precise and less invasive use, the field of neuroscience will no doubt continue to decipher the function of specific neural circuits in behavior. Alcohol researchers can benefit from implementing these tools to answer questions about how cell populations in the brain contribute to alcohol addiction.

Abbreviations:

ChR

Channelrhodopsin

ChR2

Channelrhodopsin-2

GPCR

G-protein coupled receptor

NpHRs

halorhodopsins

EF1-α

elongation factor-1 α

TH

tyrosine hydroxylase

DA

dopamine

eYFP

enhanced yellow fluorescent protein

VTA

ventral tegmental area

DREADDs

designer receptors exclusively activated by designer drugs

LMOs

Lumiopsins

FSCV

fast scan cyclic voltammetry

EtOH

ethanol

NAc

nucleus accumbens

mPFC

medial prefrontal cortex

6-OHDA

6-hydroxydopamine

MSNs

medium spiny neurons

D1DR

dopamine receptor D1

D2DR

dopamine receptor D1

ChAT

choline acetyltransferase

CIE

chronic intermittent ethanol exposure

PFC

the prefrontal cortex

HPC

the hippocampus

LTP

long term potentiation

LTD

long term depression

HPA

hypothalamus-pituitary- adrenal

CRF

corticotrophin releasing factor

CeA

central amygdala

BNST-LP

bed nucleus of the stria terminalis lateral posterior

mGluR

metabotropic glutamate receptor

LTCCs

L-type voltage-gated calcium channels

CRFR-1

CRF receptor type 1

LHb

lateral habenula

MHb

medial habenula

IPN

interpeduncular nucleus