The Incentive Amplifying Effects of Nicotine: Roles in Alcohol Seeking and Consumption

Abstract

Nicotine has a unique profile among drugs of abuse. To the non-initiated user, nicotine has powerful aversive effects, while its relatively weak euphorigenic effects undergo rapid tolerance. Despite this, nicotine is heavily abused despite negative heath consequences and nicotine users have enormous difficulty quitting. Further, nicotine is one of the most commonly co-abused substances, in that it is often taken in combination with other drugs, in particular alcohol. One explanation of this polydrug use involving nicotine is that it has multiple appetitive and consummatory conditioning effects. For example, nicotine is a reinforcement enhancer in that it can potently increase the incentive value of other stimuli, including those surrounding drugs of abuse such as alcohol. In addition, nicotine also has a unique profile of neurobiological effects that alter regulation of alcohol intake and interoceptive status. This review discusses the psychological and biological mechanisms surrounding nicotine’s appetitive conditioning and consummatory effects, particularly its interactions with alcohol.

INTRODUCTION

Nicotine is frequently taken in combination with other drugs, especially alcohol. Nicotine and alcohol use co-morbidity has been well established among human users, (Weinberger, Funk, & Goodwin, 2016), with tobacco usage estimates as high as 90% among alcoholics (Batel, Pessione, Maitre, & Rueff, 1995; Burling & Ziff, 1988). Intensity of use between tobacco and alcohol are positively correlated (Istvan & Matarazzo, 1984), resulting in higher levels of individual intake. Further, attempts to quit smoking are more successful in individuals who are non-alcoholic, with at success rates as low as 7% of alcoholic smokers being able to successfully terminate use (DiFranza & Guerrera, 1990). In order to better characterize the relationship between alcohol and tobacco co-use, some lines of evidence have examined intake patterns during drinking episodes. For example, ecological studies have observed that smoking increases during episodes of drinking (Harrison, Hinson, & McKee, 2009). Subjective pleasure for drinking and smoking are positively correlated and are highest during peak intoxication (Piasecki et al., 2011; Piasecki, Wood, Shiffman, Sher, & Heath, 2012), suggesting that the intake of one substance may promote the use of the other when the two are concurrently available. However, tobacco and alcohol co-use research is currently in need of both human and animal research that better model the clinical condition, as well as how the two drugs interact (Van Skike et al., 2016).

Understanding nicotine’s psychological and behavioral effects are necessary to explain the underpinnings of pathological polydrug use. First, although the direct effects of nicotine on the central nervous system certainly contribute to the maintenance of tobacco use, the psychological process underlying nicotine’s appetitive effects are equally important, both in terms of nicotine’s mediation of tobacco use and the enhancement drug use from other drug classes. For example, the sensorimotor stimuli (“cues”) associated with smoking are important for the maintenance of tobacco use because of the conditioned effects resulting from thousands of nicotine pairings with orosensory tobacco cues (Rose, Behm, & Levin, 1993; Rose, Behm, Westman, & Johnson, 2000; Rose & Levin, 1991). The motivational response to these cues that surround tobacco use are thought to be amplified by nicotine via “reinforcement enhancement” or “incentive amplification” (Bevins & Palmatier, 2004; X. Liu, Palmatier, Caggiula, Donny, & Sved, 2007; Palmatier et al., 2007), and in humans this reinforcement enhancement can extend to enhance the response other drug and non-drug rewards (Perkins & Karelitz, 2013; Perkins, Karelitz, & Boldry, 2017, 2019). As we will discuss in this chapter, in the case of alcohol, nicotine can alter the response to alcohol-paired cues and contexts, in addition to altering alcohol’s reinforcing properties directly. These conditioning effects of nicotine first result in initiation of approach behaviors towards drug and non-drug cues and elicit conditioned motivational states, which can subsequently maintain both tobacco and alcohol use.

Second, intake of one substance appears to be directly linked to the other. Experimental studies in humans have sought to establish causal links using consummatory measures of intake and concepts of reward during tobacco and alcohol use. Under experimenter-controlled conditions, alcohol administration increases both the urge to smoke and subsequent smoking behavior (Epstein, Sher, Young, & King, 2007; A. C. King & Epstein, 2005; Mitchell, de Wit, & Zacny, 1995; Rose et al., 2004), suggesting alcohol potentiates both tobacco-directed appetitive and consummatory behaviors. The reciprocal has also been demonstrated under laboratory conditions, where subjects who smoked nicotinized cigarettes report increased alcohol-induced pleasure derived relative to those who smoked de-nicotinized cigarettes (Barrett, Tichauer, Leyton, & Pihl, 2006; E. M. Kouri, McCarthy, Faust, & Lukas, 2004). Thus, alcohol and tobacco are both taken together, show positively correlated usage, and appear to increase the craving and intake of each other bidirectionally (for review see Verplaetse & McKee, 2017).

This review will discuss two major domains of nicotine’s effects on drug-taking, with special attention to its interactions with alcohol. Here, we separate these domains into being defined as affecting either the 1) appetitive (conditioning, seeking, motivational, and approach) or 2) consummatory (intake and reward) components of alcohol-directed behavior. In this regard, preclinical animal models are widely employed given their utility in characterizing precise neurobiological processes, drug-drug interactions, and long-term changes that are involved across nicotine’s various actions (Koob, Kenneth Lloyd, & Mason, 2009).

APPETITIVE EFFECTS OF NICOTINE

Although smoked tobacco has been referred to as one of the most addictive and harmful drugs (Nutt, King, Saulsbury, & Blakemore, 2007), the evidence indicating that nicotine is reinforcing in the absence of tobacco smoke or other stimuli is scant and debated, (Dar & Frenk, 2004; Henningfield & Goldberg, 1983; Henningfield, Miyasato, & Jasinski, 1983; Perkins, Grobe, Caggiula, Wilson, & Stiller, 1997) (Perkins, 2004; Perkins et al., 1997; Rose, Behm, Westman, Bates, & Salley, 2003; Rose & Corrigall, 1997). Notably, studies demonstrating intravenous nicotine self-administration in human smokers have required the presence of auditory, visual, or olfactory smoke cues coincident with nicotine infusion (Harvey et al., 2004; Rose, Behm, Westman, & Bates, 2003). As noted throughout this chapter, these external cues are critical for nicotine’s reinforcing properties.

Cues influence behavior by acquiring incentive value

Reward-associated stimuli (“cues”) strongly influence may forms of motivated behavior. In most situations, cues are inherent to the rewards they are associated with (e.g., the smell of a freshly baked brownie, the taste of alcohol). However, in modern environments, and in laboratory preparations, cues can be spatially and temporally separated from their rewards (e.g., a neon alcohol advertisement, a deck of cards at a casino, an empty cigarette box). In these examples, cues have properties that fall into two categories: they predict the availability of rewards, and they can incentivize several behavioral responses. What is the incentive value of a cue and how is it determined? Generally, a cue that has acquired incentive value promotes some form of motivated response. Specifically, in the laboratory, cues can be said to have acquired incentive value if 1) it becomes attractive and individuals will approach it (e.g., during a Pavlovian conditioned approach paradigm) 2) individual will work for it (e.g., during a conditioned reinforcement test), or 3) it energizes an ongoing instrumental action (e.g., during drug self-administration or Pavlovian to instrumental transfer Milton & Everitt, 2010; Robinson, Yager, Cogan, & Saunders, 2014). In using these measures, several laboratories have demonstrated that the incentive value of cues are experimentally dissociable from just its predictive value (Robinson & Flagel, 2009; Robinson et al., 2014) and subserved by different neurobiological mechanisms (Flagel et al., 2011; Flagel, Watson, Robinson, & Akil, 2007; Saunders & Robinson, 2012). Further, there is substantial individual variability in the attribution of incentive value to reward cues (Meyer, Lovic, et al., 2012), which may be a vulnerability trait for addiction-related traits (Meyer, Ma, & Robinson, 2012; Saunders & Robinson, 2010, 2011).

Nicotine imbues cues with incentive value via multiple mechanisms

Evidence from Pavlovian paradigms

Nicotine increases the incentive value of cues by each of these three measures, albeit via different biopsychological mechanisms. For example, our laboratory has shown that nicotine-paired cues are attractive in a Pavlovian paradigm (Fig. 1). To demonstrate this, rats were equipped with an intravenous (i.v.) catheter and were then presented with a cue (an 8-second presentation of a lever), that predicted the subsequent delivery of 0.03 mg/kg intravenous nicotine infusions (Fig. 1A). These rats learned to approach the lever cue, compared to rats that received unpaired cue/nicotine presentations (Fig. 1B). This demonstration of approach behavior shows that nicotine can imbue its associated cues with incentive value through Pavlovian mechanisms. However, paired rats did not show differences from unpaired rats when working for the cue itself via conditioned reinforcement (Fig. 1C), although a similar study demonstrated that a cue paired with nicotine did acquire this additional reinforcing value (Yager & Robinson, 2015). Thus, within strictly associative Pavlovian conditioning paradigms, nicotine-paired cues can acquire incentive value. As discussed later, the incentive value acquired through these associative processes also plays an important role in nicotine self-administration in animal models as well as smoking behavior in humans.

Figure 1: A nicotine-predictive cue elicits approach in a Pavlovian conditioning paradigm.

A) Rats were equipped with an intravenous catheter and trained to associate an 8-s presentation of a retractable lever with a 0.03 mg/kg infusion of nicotine over 8 trials on a VI-15m schedule for 14 days. B) Rats approach the nicotine-paired cue to a greater degree than animals presented with the lever and nicotine in an unpaired fashion. C) On a subsequent test for conditioned reinforcement, rats nose-poked for 3-s presentations of the conditioned stimulus in the absence of nicotine during a 40-min test, but there were no differences in paired and unpaired groups.

Evidence from operant paradigms

The incentive properties of cues are particularly important in the study of addiction because they strongly enhance the maintenance of drug-taking and relapse into drug addiction. For example, in a seminal study by Caggiula (Caggiula et al., 2001), rats were trained to press a lever at high rates to obtain i.v. infusions of nicotine that were paired with visual cues. When the cues were removed, the number of acquired nicotine infusions was immediately and substantially reduced, despite nicotine still being freely available. Reintroduction of the cues immediately restored responding for nicotine. Nicotine cues are therefore critical for motivating drug-taking behavior during ongoing nicotine self-administration and is another example of the cues controlling behavior via their incentive properties. Similarly, nicotine-associated cues enhance nicotine-directed behavior by increasing the motivation to obtain the drug (Chaudhri et al., 2007), slow the extinction of operant responding for nicotine when nicotine is no longer available, and reinstate nicotine seeking after extinction (Feltenstein, Ghee, & See, 2012; LeSage, Burroughs, Dufek, Keyler, & Pentel, 2004; Versaggi, King, & Meyer, 2016).

In a series of follow-up studies on Caggiula’s initial study, Palmatier et al. demonstrated that cues need not be directly associated with nicotine to produce these effects. In a particularly elegant experiment (Palmatier et al., 2006), four groups of rats were trained to respond for nicotine and/or a visual cue. One group lever-pressed for a visual stimulus (VS), one group pressed only for 0.03 mg/kg nicotine, and a third group pressed for VS-nicotine combination on a single lever. Crucially, the fourth group could concurrently press for the VS and nicotine on separate levers. The nicotine-only group responded the least, with responding lower than the VS-only group. The VS/nicotine single lever group responded at high levels, as had been shown many times. However, the separate nicotine-VS lever group responded for nicotine at similarly low levels as the nicotine group, but responded at high levels for the VS, in fact as high as the VS/nicotine single-lever group. This demonstrates a crucial finding: nicotine self-administration is primarily driven not by its own reinforcing effects, but by its ability to enhance the incentive value of other reinforcers. Subsequent studies further characterized this phenomenon, having very important implications regarding nicotine’s effects on cues 1) nicotine enhances the reinforcing value of a stimulus even when nicotine is given non-contingently at the beginning of the session (Chaudhri, Caggiula, Donny, Booth, et al., 2006; Chaudhri et al., 2007; Chaudhri, Caggiula, Donny, Palmatier, et al., 2006; X. Liu et al., 2007; Palmatier et al., 2007; Perkins et al., 2019; Versaggi et al., 2016) 2) the incentive amplification induced by nicotine depends on the initial reinforcing value of the stimulus (Palmatier et al., 2007), and 3) nicotine can enhance the incentive value of stimuli in Pavlovian and operant paradigms (Palmatier et al., 2013; Stringfield, Boettiger, & Robinson, 2018; Versaggi et al., 2016). Thus, nicotine enhances the incentive and reinforcing properties of stimuli across a variety of reward conditioning protocols, consequently imbuing those stimuli with the enhanced capacity to instigate motivated behavior or maintain responding.

In humans, nicotine-paired cues clearly instigate motivated behavior as well. Nicotine cues can induce nicotine craving (Carter & Tiffany, 1999), evoke conditioned physiological responses (Carter & Tiffany, 1999; Erblich, Bovbjerg, & Sloan, 2011; Winkler et al., 2011) maintain smoking behavior independently of nicotine intake (Donny, Houtsmuller, & Stitzer, 2007), and are necessary for the subjective positive experiences induced by nicotine (Alia-Klein et al., 2007; Rose et al., 2000). Thus, smoking, nicotine self-administration, and other smoking-associated behaviors are principally controlled by nicotine’s interactions with reward-associated cues. This incentive amplification may explain why the initiation and maintenance of smoking behavior is often highest in the presence of other rewarding situations, such as social interaction (Nguyen & Zhu, 2009). However, the isolation of the nicotine’s incentive amplifying effects have been only measured directly in a few studies, with mixed results (Addicott, Oliver, & Joseph McClernon, 2017; Barr, Pizzagalli, Culhane, Goff, & Evins, 2008; Perkins, Grottenthaler, & Wilson, 2009; Wignall & de Wit, 2011). Specifically, one study (Barr et al., 2008); found that nicotine, delivered via transdermal patch to non-smokers, biased responding toward reward stimuli, while another (Perkins et al., 2009), found that intranasal nicotine or smoking had no effect on an operant task in which non-dependent smokers responded for various non-drug rewards on an escalating fixed-ratio (FR) (FR11, FR12, etc.) schedule of responding. However, expended effort is only one measure of a reward’s incentive value, and as described above, nicotine may amplify the incentive properties of other reward (including alcohol) through multiple mechanisms (Robinson et al., 2014).

Nicotine amplifies motivation for alcohol by interacting with alcohol-associated cues.

As described above, nicotine’s conditioned effects are not limited to nicotine cues, because nicotine similarly increases responding for cues associated with other reinforcers (Caggiula et al., 2009), including alcohol (Le, Wang, Harding, Juzytsch, & Shaham, 2003b). Thus, one mechanism for the comorbidity of alcohol and nicotine use (Weinberger et al., 2016; Weinberger, Platt, Jiang, & Goodwin, 2015) is that nicotine enhances the reinforcing properties of alcohol-associated cues separately from the pharmacological interactions between these drugs. Consistent with this, rodent studies show that nicotine produces a high-approach, low-avoidance alcohol phenotype in response to alcohol associated cues. For example, nicotine enhances the approach response evoked by an alcohol-predictive cue in Pavlovian paradigms (H. Angelyn, Loney, & Meyer, in press; Loney, Angelyn, Cleary, & Meyer, 2019; Maddux & Chaudhri, 2017; Srey, Maddux, & Chaudhri, 2015). We have shown nicotine enhanced the response to a cue that predicted access to a retractable sipper bottle containing ethanol, without affecting the response to a similar, non-predictive stimulus. In other experiments, nicotine reduced the conditioned taste (but not place) aversion induced by ethanol (Loney et al., 2019; Loney, Pautassi, Kapadia, & Meyer, 2018), indicating that nicotine alters the motivational effects of alcohol-associated cues in some, but not all, Pavlovian paradigms.

Alcohol self-administration models assess the effects of motivation and intake

Alcohol self-administration involves both the preparatory responding and motivation to gain access to alcohol, as well as the subsequent consumption of alcohol following its delivery. In our laboratory, we use a procedure in which the appetitive and consummatory phases of ethanol self-administration can be dissociated [Fig. 2 and 3; see also Schier, Dilly, and Gonzales (2013); Simms, Bito-Onon, Chatterjee, and Bartlett (2010)]. First, rats are given 24-hour access to 20% ethanol or water, three times a week for four weeks. This produces escalation in ethanol intake (Fig. 2A). Then, rats make responses that are reinforced with a 30-second presentation of a bottle containing 20% ethanol (Fig. 2B). A cue-light is illuminated with the bottle presentation. Using contact lickometers, we can dissociate the appetitive (nose-poking or lever-pressing) and consummatory (licking and consumption) aspects of ethanol drinking (Fig 2C). In addition, the motivational properties of ethanol reinforcement can be assessed using a progressive ratio (PR) test, in which work requirement to access alcohol sipper becomes increasingly difficult. The PR test is used as a standard measure of ethanol reinforcement with minimal influence of post-ingestive factors. A shown in Fig. 2, the alcohol exposure provided in the home cage produces increases in both the appetitive (Fig. 2D) and consummatory measures (Fig. 2C) of alcohol-directed behavior. Using this design, it is therefore possible to specifically observe nicotine’s effects on appetitive alcohol-directed behavior, separately from its effects on regulating intake.

Figure 2: Home-cage ethanol exposure promotes operant ethanol self-administration.

A) 20% alcohol v/v is given every other day in the home-cage for 4 weeks, resulting in escalated intake and increased preference. Then, B) rats are trained to operantly respond for alcohol by nosepoking for 30s presentations of an alcohol sipper. During self-administration, consummatory behavior is measured using contact lickometers to measure C) alcohol intake, and also D) appetitive nosepoking behavior during tests such as progressive ratio.

Figure 3: Nicotine enhanced operant responding for alcohol.

A) Rats were tested for both alcohol-directed responding and intake in daily 1-hour sessions under saline or 0.4 mg/kg injections of nicotine. B) Nicotine enhances responding for alcohol in the presence of cues relative to saline treated rats, however C) this did not result in heightened intake.

Nicotine enhances responding for alcohol in the presence of cues

Being able to distinguish between the appetitive and consummatory consequences of nicotine is important, particularly in order to dissociate the mechanisms by which nicotine modifies behavior. Factors regulating intake will be discussed in detail below, however generally speaking, nicotine reduces alcohol intake when it is given acutely, immediately prior to testing (Le, Corrigall, Harding, Juzytsch, & Li, 2000; Sharpe & Samson, 2002). In addition, chronic nicotine also increases responding for alcohol in the presence of cues (Le et al., 2003b), even when delivered continuously over a 24-hour period via osmotic pump (Clark, Lindgren, Brooks, Watson, & Little, 2001a). Using our method of alcohol self-administration to identify appetitive-specific effects of nicotine (Fig. 3A), we have found that nicotine enhances responding for presentations of an alcohol sipper (Fig. 3B) in the presence of alcohol-paired cues without affecting subsequent intake (Fig. 3C). These data fit with the general body of work suggesting nicotine can enhance responding for other reward-paired stimuli, crucially those paired with alcohol. These alcohol cues presented alongside the alcohol reinforcement can acquire incentive value, as measured by their ability to cause the reinstatement of an extinguished response (Lamparelli & Meyer, 2018). Nicotine enhances this effect and may be one reason why alcohol abstinence is hard to maintain in smokers (C. L. Smith et al., 2020).

Occasions setters and discriminative stimuli

In the experimental designs described in this chapter, cue presentation was contingent on the rats’ response. However, cues can be presented independent of the subject’s response, for example acting as “occasion-setters” to signal drug-availability or acting as a discriminative stimulus (S^d). For example, nicotine can serve as a S^d to occasion motivated responding (Randall, Cannady, & Besheer, 2016; Troisi, Dooley, & Craig; Wooters, Bevins, & Bardo, 2009). In an example using rats, we tested the effect of nicotine injections in a discriminative stimulus (S^d)-modulated alcohol self-administration and reinstatement procedure. Here, rats nose-poked for presentations of an alcohol bottle in the presence of an “alcohol-available” S^d (illumination of a houselight; Fig. 4A). We found that this S^d effectively modulated alcohol self-administration compared to a separate alcohol “non-available” DS (S^delta), and crucially nicotine specifically enhanced appetitive responding only in the presence of the S^d during self-administration (Fig. 4B). Further, the enhancement was also observed during a reinstatement test when the alcohol bottle and its associated cues were removed (H Angelyn & Meyer, 2019). Since these cues were not response contingent, the term “reinforcement enhancement” in reference to nicotine’s effects is insufficient. Instead, because nicotine’s immediate effects on regulating appetitive behavior through cues therefore also extend to other salient stimuli that signal reward availability such as occasion setters and discriminative stimuli, nicotine is better described as acting more generally as an amplifier of the incentive value of cues Bevins and Palmatier (2004), including alcohol cues. Further this incentive amplification can apply to complex contextual stimuli, which have powerful control over not only nicotine seeking (Conklin, 2006; McClernon et al., 2016), but alcohol seeking as well (Zipori et al., 2017).

Figure 4: Nicotine enhances responding for alcohol in the presence of a discriminative stimulus signaling alcohol availability.

A) Rats were trained to respond for alcohol in 7.5 min “alcohol-available” (S^d) periods, where active nose-pokes resulted in presentation of an alcohol sipper solution. Each 7.5 min Sd period was followed by a 7.5 min “alcohol non-available” period signaled by a separate “S^delta” stimulus. B) Rats specifically responded for alcohol during the Sd period, and Sd-induced responding specifically was enhanced by nicotine.

Collectively, these findings indicate that nicotine promotes appetitive alcohol responding by amplifying the incentive value of alcohol-associated cues. This is consistent with Cagiulla’s ideas regarding nicotine’s effects on reinforcement (Caggiula et al., 2009), with the modification that nicotine amplifies the incentive value of cues generally, whether they are response-contingent or not. However, one aspect of alcohol and nicotine co-use that may be unique to the combination of these drugs is the effect of nicotine on alcohol taste cues. While oral cues can instigate alcohol seeking (Knight et al., 2016), higher doses of alcohol can condition an aversion to alcohol-associated tastes (e.g., Anderson, Varlinskaya, & Spear, 2010). Nicotine interferes with this conditioned taste aversion to alcohol-associated flavors (Loney & Meyer, 2019; Loney et al., 2018) (Bienkowski, Piasecki, Koros, Stefanski, & Kostowski, 1998; Kunin, Smith, & Amit, 1999; Loney et al., 2019; Rinker et al., 2011), which is likely due to an alteration of the interoceptive properties of alcohol by nicotine (discussed more later), and not to a general effect of nicotine on learning, because nicotine does not alter a taste aversion induced by the nausea-producing drug lithium chloride. In this manner, in addition to increasing the incentive value of environmental alcohol cues, nicotine may also facilitate alcohol seeking by altering the aversive interoceptive effects of higher doses of alcohol.

CONSUMMATORY EFFECTS OF NICOTINE

Thus far we have discussed the appetitive effects of nicotine, crucially through increasing the salience of drug paired cues and instigating seeking and approach behavior and enhancing reinforcement for alcohol in the presence of cues (Le, Wang, Harding, Juzytsch, & Shaham, 2003a). During ongoing drug self-administration, in addition to the presence of discrete alcohol and nicotine cues, drug-directed behavior is influenced by lifetime experience with both substances, as well as the individual’s own ability to titrate their alcohol and nicotine intake. As discussed, nicotine’s profile of effects envelopes both appetitive and consummatory processes. including nicotine’s enhancement of responding for alcohol during self-administration and reinstatement (Le et al., 2003a), its promotion of intake in a time-dependent manner (Le et al., 2000), and its effects when given chronically at higher doses (Bito-Onon, Simms, Chatterjee, Holgate, & Bartlett, 2011). However, the relationship between intake and responding is not always straightforward. In one study using rats (Le et al., 2010), alcohol and nicotine were made available together in one group, while other groups self-administered only one drug. Alcohol self-administration occurred at the expense of nicotine, but later the extinction of responding for alcohol was delayed in rats that had nicotine concurrently available, indicating separable intake and operate response consequences of nicotine. Priming with either nicotine or alcohol during a subsequent reinstatement test increased responding for both drugs. Thus, this study showed that self-administered nicotine, although itself decreased by alcohol, increased responding for self-administered alcohol. Thus, segregation of the appetitive and consequential intake effects of nicotine is crucial to fully understand the interaction between drugs.

Several studies have used responding for alcohol and intake interchangeably as endpoints for nicotine’s effects on alcohol-directed behavior. Importantly however, the effects of nicotine are highly dependent on when it is given, for how long, and the exact behaviors being measured and when. Our studies show that although nicotine increases alcohol’s appetitive effects in standard self-administration paradigms, it either does not change or reduces alcohol intake acutely. Instead, nicotine-induced changes related to the ingestive, consummatory aspects of alcohol and are likely due to nicotine’s chronic effects, which is discussed in this section.

Nicotine’s temporal factors regulating alcohol intake

Acute effects on intake

Alcohol intake involves the reinforcing properties of alcohol during and following drinking, as well as processes involving of fluid intake and regulation (Samson, Slawecki, Sharpe, & Chappell, 1998). Naturally, given its high face validity, quantified alcohol intake is the one of the most frequently reported methods for operationalizing alcohol-directed behavior. However, the timing, dosing, and experience with nicotine may differentially alter alcohol intake, and thus with the wealth of methods to measure alcohol-directed behavior under nicotine, each measure and experimental design may target different psychological processes.

Perhaps the simplest way to model the interaction between nicotine and alcohol in rodents is to look at how consumption of one substance changes while under the influence of the other. The simplicity of this design is likely one of the reasons why this approach has historically been one of the most widely employed. For example, a number of early studies from Samson and colleagues have demonstrated that alcohol intake in rats is acutely suppressed when measured following subcutaneous injections of nicotine (Hendrickson, Zhao-Shea, Pang, Gardner, & Tapper, 2010; Nadal & Samson, 1999; Sharpe & Samson, 2002). Nicotine is anorexigenic (Jo, Talmage, & Role, 2002), which likely suppresses alcohol intake during early phases of nicotine exposure. Home-cage alcohol intake is suppressed specifically via nicotinic acetylcholine receptors (Dyr, Koros, Bienkowski, & Kostowski, 1999) including the α4 subtype (Hendrickson et al., 2010). Importantly, these studies occurred in the absence of a dedicated testing environment and alcohol-paired CSs, thus more directly reflecting the consummatory phase of alcohol intake.

Protracted effects on intake

In humans, experience and history with nicotine undoubtably contributes to the longer term behavioral and neurobiological changes, with age of smoking onset predicting alcohol use problems later (Chen et al., 2002; Grant, 1998; Jensen et al., 2003). Thus, in the laboratory, intake-specific effects of nicotine are similarly influence by an individual’s history and timing of exposure, particularly when experienced outside the testing environment. For example, alcohol-experienced rats show a suppression of intake when nicotine is given immediately prior to a relapse to alcohol intake test, but instead showed heightened intake when nicotine is was given during a withdrawal period on the days prior (Alen, Gomez, Gonzalez-Cuevas, Navarro, & Lopez-Moreno, 2009; William M. Doyon et al., 2013). Indeed, other designs have similarly demonstrated nicotine can later enhance alcohol intake when it is first given in the absence of alcohol (Hauser, Getachew, et al., 2012; Lopez-Moreno et al., 2004), and during adolescence (Kemppainen, Hyytia, & Kiianmaa, 2009). Importantly, these data suggest that longer-term intake-specific effects may be divorced from the other more immediate effects, where intake is heightened when it is temporally separated from the drinking event (Alen et al., 2009). Thus, procedures that generally employ longer nicotine exposure periods across days better facilitate alcohol intake in the home cage (Blomqvist, Ericson, Johnson, Engel, & Soderpalm, 1996; Olausson, Ericson, Lof, Engel, & Soderpalm, 2001; B. R. Smith, Horan, Gaskin, & Amit, 1999). Taken together, it is likely that the intake specific processes surrounding nicotine are a result longer term exposure periods, which eventually come to interact with the various appetitive and conditioning properties in various drug-taking settings. Indeed, nicotine delivered continuously by osmotic minipump can also promote the acquisition and maintenance of alcohol self-administration (Clark, Lindgren, Brooks, Watson, & Little, 2001b), perhaps better reflecting the chronic nature of tobacco use.

Behavioral models of co-use

A brief summary of these data suggests that, at least acutely, experimenter-delivered nicotine has suppressive effects on both alcohol intake and seeking when it is given in close temporal proximity but can enhance it when given chronically or separately from the testing session. Despite the great utility of studies that utilize non-contingent injections, controlled site-specific infusions, and other highly experimentally controlled methods, individuals can titrate their own tobacco and alcohol use under normal conditions. It stands to reason rodent studies employing dual self-administration protocols may model key components of voluntary human usage, although very few groups have tried this. Marshall and colleagues originally reported that, when nicotine and alcohol solutions were presented together, there were no changes in consumption relative to when they were presented alone (Marshall, Dadmarz, Hofford, Gottheil, & Vogel, 2003). Other attempts to model co-use using operant conditioning instead of free intake were later attempted, but either nicotine self-administration decreased in the presence of alcohol (Le et al., 2010), or the concurrent availability of both decreased the intake of both (Funk, Lo, Coen, & Le, 2016).

It is likely that the specific parameters surrounding drug-availability, timing, training, and presence of cues are important for these models. Other variations to this procedure have been deployed, but with mixed results. Orally self-administering rats demonstrate preference for an alcohol only solution over alcohol and nicotine (Hauser, Katner, et al., 2012), but show heightened intake for a cocktail of intravenous nicotine and alcohol over either by itself (Karatayev et al., 2015), likely reflecting factors surrounding relative aversion and reward due to route of administration. As discussed previously, nicotine has been shown to enhance intake when taken sufficiently in advance of alcohol, and in one instance during daily alternating access to alcohol and nicotine, nicotine enhanced alcohol intake when nicotine was taken earlier in the day (Le, Funk, Lo, & Coen, 2014), suggesting that that timing of drug-taking is important both under experimenter controlled and voluntary conditions.

Because nicotine enhances appetitive responding for alcohol in the presence of cues (see Fig. 3), to assess the immediate effect of nicotine availability on free-alcohol consumption in the absence of cues, we used a procedure in which rats were allowed to freely self-administer nicotine in the presence of an alcohol sipper (Fig. 5) that required no appetitive responding. In contrast to Le and collegues, we found that i.v. nicotine was readily self-administered (Fig. 5A), but at the expense of alcohol (Fig. 5B), crucially in rats that had an established history of home cage drinking (adapted from: (Simms et al., 2010; Simms et al., 2008)). Upon nicotine removal, responding for nicotine rapidly decreased (Fig. 5C), and alcohol consumption increased to saline levels (Fig. 5D). Recently, Chandler and colleagues have sought to try and optimize i.v. nicotine and alcohol intakes under co-use conditions, with the primary goal of evaluating pharmacological manipulations (naltrexone and varenicline) directed at reducing intake of one or both substances. Specifically, they demonstrated that alcohol consumption could escalate under nicotine, but specifically only when the fixed-ratio response requirement was sufficiently high (Chandler, Maggio, Peng, Nixon, & Bardo, 2020).

Figure 5: Alcohol and nicotine continuous access.

Rats were trained to lever press for nicotine or saline in the presence of a continuously available 20% alcohol sipper. Nicotine responses are shown on the left panels, and alcohol intake is shown in the right panels. A) Rats lever pressed for nicotine more than saline controls. However, B) rats with nicotine available consumed less alcohol. When nicotine was removed, C) responding on the nicotine lever decreased, and D) alcohol intake increased quickly to control levels.

These data together suggest that 1) alcohol and nicotine can both be readily self-administered, but only when 2) either the response requirement for nicotine is sufficiently high to discourage high levels of intake, or 3) nicotine-availability is sufficiently separated from alcohol to avoid the immediate alcohol suppressing effects. Further, separation of both the motivational and intake-specific effects of nicotine are especially crucial, given the differential effect of nicotine on both under various conditions.

The future utility of validating co-use models may provide important insights into the unique neuroadaptive changes surrounding motivation and regulation of intake when the subject can regulate their own consumption of both nicotine and alcohol. This may be especially useful because pharmacological treatments differ in efficacy among individuals with both tobacco and alcohol use histories, and the use of well-designed animal models of co-use may help reconcile currently existing disparities with behavioral outcomes and treatment mechanisms between sets of models, making them more generalizable (Motschman et al., 2016). Future models of co-use will thus likely be most successful when they are able to exploit the availability of cues, the long-term potentiating effects of nicotine, and dissociate both the appetitive/motivational components of drug-seeking from intake.

NEUROBIOLOGICAL UNDERPINNINGS OF NICOTINE’S REINFORCING AND INCENTIVE AMPLIFYING EFFECTS

As discussed, nicotine produces a range of appetitive and consummatory effects on alcohol. The neurobiological processes underling tobacco and alcohol co-use are complicated and involve drug actions across a range of neurotransmitter systems and brain areas. Each of these systems may engage different psychological and behavioral processes, producing an ensemble of changes working in concert to modify alcohol and nicotine directed behavior across the lifetime of the individual. It is therefore difficult to disentangle the direct effects of nicotine on alcohol-directed behavior from the synergistic effect of both taken in conjunction, although some lines of evidence have been useful in dissociating the various drug- and neuroanatomical specific effects involved in co-use.

Neurobiological substrates for nicotine and alcohol: mesolimbic dopamine

The involvement of the mesolimbic dopamine system in drug reinforcement and learning has been well documented and is probably the most widely examined neural system to date in the development of drug addiction across a variety of compounds. In addition to its involvement in the cued response to drugs, the reinforcing properties of most drugs of abuse show strong convergent evidence for the involvement in the limbic system (Everitt & Robbins, 2005, 2013; Nestler, 2005; Robinson & Berridge, 1993, 2000, 2008), and are necessary for the maintenance of drug-taking across different drug-classes such as the psychomotor stimulants, opioids, nicotine, and alcohol. It is thus reasonable to predict that the reinforcing and motivational properties of alcohol and its associated cues are directly modified by nicotine through convergent neuroadaptations in dopaminergic circuitry.

Functionally, as a key structure in the mesolimbic dopamine system, the ventral tegmental area (VTA) has been characterized for its role encoding the motivational value of rewards and their associated cues (D’Ardenne, McClure, Nystrom, & Cohen, 2008; Wolfram Schultz, 1997; W. Schultz, 1998, 2015), principally through phasic bursts of firing (Tsai et al., 2009) which can instigate reward seeking behavior (Adamantidis et al., 2011). The firing of dopamine-containing cell bodies in the VTA results in dopaminergic overflow in downstream forebrain regions such as the nucleus accumbens (NAcc) (Sombers, Beyene, Carelli, & Wightman, 2009) that contrast tonic single-spike firing (Hyland, Reynolds, Hay, Perk, & Miller, 2002). Thus, these transient increases in NAcc dopamine release become linked to reward cues and initiation of reward seeing behavior (Owesson-White, Cheer, Beyene, Carelli, & Wightman, 2008; Roitman, Stuber, Phillips, Wightman, & Carelli, 2004). These actions likely underlie the conditioned, appetitive effects of drugs such as nicotine following drug-use.

The direct, immediate actions of addictive drugs within the mesolimbic dopamine system are therefore recognized to drive abnormal motivational learning when compared to conventional reinforcers (Di Chiara, 1998; Di Chiara et al., 1999; Pierce & Kumaresan, 2006). Like many drugs of abuse, nicotine and alcohol involve actions on mesolimbic dopamine, principally by increasingly dopamine release in the NAcc through various mechanisms. Although here we discuss convergent nicotine and alcohol reinforcement mechanisms, an exhaustive review of the complex cellular and neurobiological mechanisms of alcohol and nicotine reinforcement is beyond the scope of this review. These systems have recently been reviewed in excellent detail elsewhere (e.g., for a recent review see Morel, Montgomery, & Han, 2019).

Mechanisms of nicotine reinforcement

Nicotine binds to the various subtypes of the nicotinic acetylcholine receptor (NAchR) that are neuroanatomically distributed (Gotti et al., 2009), and for example can stimulate dopamine release via VTA and NAcc actions (Ferrari, Le Novere, Picciotto, Changeux, & Zoli, 2002). In addition to being expressed at multiple nodes in the midbrain, dorsal, and ventral striatum, it is further expressed in the habenular-peduncular pathway, amygdala, hippocampus, and cortical areas (Tuesta, Fowler, & Kenny, 2011). Furthermore, nicotinic receptor subtype distribution varies anatomically as well (for a full review, see: (Fowler, Arends, & Kenny, 2008), although the α4, α7, and β2 have received a great deal of addition given their location in the VTA (Charpantier, Barneoud, Moser, Besnard, & Sgard, 1998), involvement in nicotine self-administration (Brunzell & McIntosh, 2012; Laviolette & van der Kooy, 2003), and involvement in reinforcement within in the VTA (Pons et al., 2008). Furthermore, β2-containing nicotinic receptors stimulate dopamine release from the VTA (Mameli-Engvall et al., 2006), and systemic antagonism of α4β2 reduces the conditioned reinforcing properties of nicotine (L. Liu, Zhao-Shea, McIntosh, Gardner, & Tapper, 2012; X. Liu et al., 2007). The appraisal of the β2-containing nicotinic receptor in drug related behaviors, and its proximity to the VTA-NAcc pathway have made this an attractive target for nicotine actions on reinforcement.

Mechanisms of alcohol reinforcement

Alcohol-elicited dopamine release is a key process in reinforcement (Gonzales, Job, & Doyon, 2004), and alcohol is largely been identified as reinforcing in the VTA (Gatto, McBride, Murphy, Lumeng, & Li, 1994; Rodd et al., 2004), stimulating VTA dopamine neurons directly (Brodie, Pesold, & Appel, 1999) (Appel, Liu, McElvain, & Brodie, 2003) resulting in downstream DA release in areas including the NAcc (Di Chiara & Imperato, 1985) to promote self-administration (Rassnick, Pulvirenti, & Koob, 1992). However, the VTA is a relatively large and heterogeneous structure, with separate neuroanatomical regions distinguished by various afferent targets (Swanson, 1982; Walsh & Han, 2014) and cytoarchitectural distinctions (Sanchez-Catalan, Kaufling, Georges, Veinante, & Barrot, 2014). The multiple projections of the VTA innervate the forebrain (Swanson, 1982) including areas crucial for drug reinforcement such as the NAcc. The various regions of the VTA have been simplified into dimensions based on neuroanatomical location, for example as either anterior (aVTA) or posterior (pVTA), both of which are differentially sensitive to different drugs of reinforcement (Sanchez-Catalan et al., 2014). Alcohol-stimulated dopamine release, particularly in the pVTA, innervates a variety of forebrain regions including the ventral pallidum (VP), medial prefrontal cortex (mPFC), NAcc core (Ding, Ingraham, Rodd, & McBride, 2015), and NAcc (Boileau et al., 2003; Rassnick et al., 1992; Weiss, Lorang, Bloom, & Koob, 1993) which are understood to drive the reinforcing effects of alcohol. At least in the pVTA, repeated exposure to alcohol via microinjections appears to sensitize the release of NAcc shell (NAccSh) DA (Ding, Rodd, Engleman, & McBride, 2009).

Mechanisms of nicotine-alcohol interactions

Nicotine and alcohol are thought to produce complex interactions on reinforcement through their direct actions on each other within the mesolimbic system, which have also been previously discussed in several other excellent reviews (Doyon, Thomas, Ostroumov, Dong, & Dani, 2013; Hendrickson, Guildford, & Tapper, 2013). Passive administration of nicotine and alcohol together produces and additive effect of dopamine release in the NAcc (Y. Tizabi, Bai, Copeland, & Taylor, 2007) which likely involves interactions within the VTA (Yousef Tizabi, Copeland, Louis, & Taylor, 2002). Specifically, the reinforcing value of lower doses of nicotine and alcohol together are enhanced in the pVTA (Truitt et al., 2015), as both taken at subthreshold doses together intracranially, resulting in robust brain-derived neurotrophic factor (BDNF) upregulation and drug-elicited dopamine and glutamate release in the NAccSh (Waeiss, Knight, Engleman, Hauser, & Rodd, 2020). The alcohol-enhanced firing rate of VTA neurons is further exacerbated by nicotine (Clark & Little, 2004) but is concentration-dependent. Regardless, at least locally, nicotine appears to sensitize local dopamine release in the NAccSh to the subsequent effects of alcohol (Ding et al., 2012). These studies together would support an enhancement of the reinforcing and motivational properties of combined nicotine and alcohol, although they largely model acute effects, rather than neuroadaptations following prolonged consumption and drug history. Further, these studies largely focus on localized effects under intracranial manipulations.

Protracted changes in nicotine-alcohol interactions

As discussed previously, nicotine appears to better enhance voluntary alcohol intake when it is given in exposure windows either developmentally or temporally dissociated from the period of alcohol consumption. In contrast to the previously described studies, subchronic nicotine can actually blunt alcohol-induced DA release in the NAccSh (Lopez-Moreno et al., 2008). The effect of nicotine pre-exposure can persist long after it is metabolized, attenuating the alcohol-induced release of NAcc DA and VTA firing, while producing concurrent increases in alcohol consumption (W. M. Doyon, Y. Dong, et al., 2013). In addition, although nicotine can acutely increase VTA firing, NAchRs are rapidly desensitized, producing acute tolerance to nicotine’s effects (Pidoplichko, DeBiasi, Williams, & Dani, 1997). The repeated combination of desensitization (Grady, Wageman, Patzlaff, & Marks, 2012) and NAchR upregulation (Fasoli et al., 2016; Staley et al., 2006) specifically on VTA gamma aminobutyric acid (GABA) inhibitory interneurons (Nashmi et al., 2007) may result in acceleration of DA-ergic blunting. These pre-exposures specifically in development appear to amplify inhibitory synaptic GABA mechanisms (Thomas et al., 2018) resulting in increased alcohol consumption.

Paradoxically, despite chronically induced decreases in alcohol induced DA release following nicotine, intake is amplified. Indeed, blunted DA function has thought to be a consequence of alcohol dependence (Martinez et al., 2005; Volkow et al., 1996). Interestingly, phasic firing appears to be a crucial feature surrounding the response to rewards and associated cues, as tonic stimulation of the VTA blunts alcohol intake (Bass et al., 2013). In light of this, it is likely that other mechanisms other than simply DA surges in the NAcc, such as changes in inhibitory GABA function (Chester & Cunningham, 2002) that can regulate consumption (Nie, Rewal, Gill, Ron, & Janak, 2011) may underlie persistent changes following nicotine histories. There is some evidence to suggest that the GABA-alcohol interaction may be especially important, given GABAergic inhibitory currents in the cerebellum are related to variation in alcohol consumption (Kaplan, Mohr, & Rossi, 2013; Richardson & Rossi, 2017).

Additional work separating cue and intake-specific effects under different drug histories will be important in elucidating the specific neuroadaptations following concurrent nicotine and alcohol use. It is, however, likely that the varieties of effects produced by nicotine in the mesolimbic system underlie features that alter cue conditioning and intake, and likely do so differently following prolonged experience with the drug. Further, nicotine-driven escalations in alcohol intake may involve other non-mesolimbic actions of nicotine.

Neurobiological mechanisms of nicotine amplification of the incentive value of cues

Because nicotine has weak reinforcing properties when administered alone, most studies examining nicotine reinforcement either present nicotine in combination with cues or measure the response to the cues themselves. Human studies using fMRI have indicated the appetitive activation (i.e., craving) activated by nicotine cues (typically presented on a video screen) are associated with activation of limbic brain areas described in reinforcement, including the NAcc (ventral striatum), amygdala, and areas of the cortex, including the prefrontal, cingulate, insular, and orbitofrontal cortices (Brody et al., 2002; McClernon, Hiott, Huettel, & Rose, 2005; Seungdamrong & Yasunaga, 1975; Smolka et al., 2006; Wilson, Sayette, Delgado, & Fiez, 2005). In one study, smokers were presented cues after smoking cigarettes, thus minimizing the effects of nicotine withdrawal. Not only was robust cue-induced craving observed, but robust activation of the ventral striatum, amygdala, orbitofrontal cortex, hippocampus, medial thalamus, and left insula were observed as well (Franklin et al., 2007). However, no non-nicotine sessions were included in the design, so it is unclear whether the neural cue responsivity was due to the acquired incentive properties of the cues (through Pavlovian mechanisms), or whether the brain activation was due to more generalized non-contingent incentive amplifying properties of nicotine. Another study measured the neural response to money-predictive (or neutral) cues after oral nicotine or placebo, and on some trials the outcome was unexpected (Addicott et al., 2017). Nicotine amplified the neural response to all cues in this task, including money-predictive cues, neutral cues, and those that preceded unexpected outcomes, particularly in the anterior insula/inferior frontal gyrus. This was also associated with a decrease in cue-induced activity in the dorsal striatum. Together, these data suggest a key role for the insular cortex (discussed in detail below) in nicotine’s ability to amplify the incentive value of cues.

Several studies, especially those in rodent models, indicate that nicotine self-administration is dependent on the mesolimbic dopamine system (Corrigall, Coen, & Adamson, 1994), which is likely regulated by brainstem areas that can influence the rewarding and aversive properties of nicotine (Laviolette, Alexson, & van der Kooy, 2002). However, in most of these investigations, nicotine is presented in combination with an associated cue, and thus the primary reinforcing effects of nicotine cannot be dissociated from its incentive amplifying effects. However, one study used a Pavlovian conditioned approach paradigm to demonstrate that dopamine receptors are required for the enhancement of the sign-tracking response by nicotine (Palmatier, Kellicut, Brianna Sheppard, Brown, & Robinson, 2014). This finding, in combination with others showing that sign-tracking requires dopamine neurotransmission (Danna & Elmer, 2010; Flagel et al., 2011), indicates that nicotine likely amplifies the incentive value of reward-associated stimuli via dopaminergic neurotransmission. Also, several studies have measured the role dopamine and related brain areas during cue-induced reinstatement, which largely support a role for the mesocorticolimbic circuity described above (Caggiula et al., 2001). In addition, transition to nicotine self-administration is accompanied by a reduction in the aversive properties of nicotine (and alcohol), which is associated with changes in the lateral habenula and peduncular nuclei (Baldwin, Alanis, & Salas, 2011; Fowler & Kenny, 2012; Glover, McDougle, Siegel, Jhou, & Chandler, 2016; Pang et al., 2016; Wolfman et al., 2018; Zhao-Shea et al., 2015).

At least some of the appetitive effects of nicotine are similar to other drug categories, and may be explained by actions within the mesocorticolimbic dopamine system which is important in coupling incentive salience to reward-paired cues (Berridge, 2012). In humans for example, DA-stimulating drugs such as d-amphetamine can enhance the positive response to stimuli independently of its euphorigenic effects (Wardle & de Wit, 2012). Amphetamine likely shares mechanistic overlap with nicotine, for example by enhancing the conditioned reinforcing value of CSs via its DA-promoting actions in the NAcc (Parkinson, Olmstead, Burns, Robbins, & Everitt, 1999; Taylor & Robbins, 1984, 1986; Wyvell & Berridge, 2000) and enhancing sign-tracking in the dorsolateral quadrant of the neostriatum (DLS) (DiFeliceantonio & Berridge, 2016). It is likely that these various pharmacological actions may converge simultaneously, as nicotine and amphetamine given in conjunction can further enhance VS responding (McNealy, Ramsay, Barrett, & Bevins, 2021), perhaps via their combined DA actions in the NAcc which is involved in conditioned reinforcement (Wolterink et al., 1993) and sign-tracking (Fraser & Janak, 2017). Besides the psychomotor stimulants, other drug categories including intra-NAcc agonists targeting opioid receptors can also enhance incentive salience measures such as pavlovian instrumental transfer (Peciña & Berridge, 2013) and conditioned reinforcement (Phillips, Robbins, & Everitt, 1994). Thus, it seems highly likely that there are DA-modulated actions on both reinforcement and incentive salience in the mesolimbic forebrain that can be recapitulated by different drug classes.

The neurobiological basis for the effects of nicotine on alcohol cue responsivity, however, is less well-studied and may reflect unique drug-drug interactions. The combination of these drugs can serve as a complex discriminative stimulus (Ford, Davis, McCracken, & Grant, 2013; Ford, McCracken, Davis, Ryabinin, & Grant, 2012; Randall et al., 2016; Troisi et al., 2013) that involves the NAcc and prefrontal cortex, (Chatterjee & Bartlett, 2010; Randall, McElligott, & Besheer), and alcohol cues activate similar circuitry. One study (A. King, McNamara, Angstadt, & Phan, 2010) showed that the neural response to smoking cues in the NAcc is enhanced by alcohol. The converse study has not been done, but one study found similar activation in the insular and cingulate cortices during separately presented alcohol and nicotine cues in co-users of these drugs (J. Liu, Claus, Calhoun, & Hutchison, 2014). A role for the insular cortex in nicotine’s interoceptive effects has been implicated in humans as well (Naqvi & Bechara, 2009, 2010), and in animals the insular cortex has been implicated in nicotine and alcohol self-administration (Forget, Pushparaj, & Le Foll, 2010; Jaramillo, Randall, et al., 2018; Jaramillo, Van Voorhies, Randall, & Besheer, 2018), as well as the enhancement of opioid self-administration by nicotine. Together, these data from humans and animals suggest that, whereas the reinforcing effects of nicotine, alcohol, and their associated cues are dependent on the activation of mesocorticolimbic circuitry, the incentive amplification properties of nicotine, and thereby its effects on alcohol-associated cues, are largely due to its actions within the insular cortex.

Neurobiological processes regulating intake: interoception and the insular cortex

Although initiation and regulation of alcohol intake is likely driven by concepts of reinforcement and conditioned motivation, quantity of intake is also likely to driven by post-ingestive feedback, including relative interoceptive state of intoxication, and the post-ingestive rewarding effects of alcohol. Interoception refers to perception of internal state (Craig, 2010), which can refer to both internal processes antecedent to drug-taking (Paulus & Stewart, 2014), as well perception of state when the drug is onboard. Nicotine can alter interoception, for example, by blunting the subjective effects of other drugs. In humans for example, transdermal nicotine patches reduce subjective intoxication for alcohol (McKee, O’Malley, Shi, Mase, & Krishnan-Sarin, 2008) and cocaine (Elena M. Kouri, Stull, & Lukas, 2001), and both nicotine and alcohol produce tolerance to the somatic effects of alcohol (Collins, Wilkins, Slobe, Cao, & Bullock, 1996). In rodents, experience with nicotine in periadolescence and adulthood can also reduce the subsequent aversive properties of alcohol (Rinker et al., 2011), as well as interfere with the conditioned taste aversion induced by ethanol, cocaine, and morphine in adult rats (Loney & Meyer, 2019) which is largely dependent on the interoceptive aversive properties of these drugs.

Insular Cortex

The insular cortex (IC) has received a great deal of attention for its role in interoception, and integration of limbic and sensory information (Gogolla, 2017), including properties surrounding drug conditioning (Arguello et al., 2017; Thompson, Bevins, & Murray, 2019) which likely explains its involvement in response to nicotine and alcohol cues. The interoceptive subjective effect of drugs is also important for regulation of intake and dose control, and loss of IC function can both enhance or reduce intake in rodents (Rotge et al., 2017) depending on whether the loss of function occurs prior to or following drug-experience. Nicotine and the IC have been of special interest, especially following several lines of evidence indicating damage following stroke in smokers can interrupt smoking (Naqvi, Rudrauf, Damasio, & Bechara, 2007), reduce perception of withdrawal (Abdolahi et al., 2015), and predict continued cessation of tobacco use following stroke (Gaznick, Tranel, McNutt, & Bechara, 2014; Suner-Soler et al., 2012).

The IC contains NAchRs available to nicotine, that normally receive input from the basal forebrain (Toyoda, 2019), specifically including β2-NAchRs that depress pyramidal neurons in layer V, likely via GABAergic action (Sato, Kawano, Yin, Kato, & Toyoda, 2017; Toyoda, 2018). Thus, convergent evidence for the IC in integration of interoceptive status via the nicotinic system appears likely. Crucially, both systemic nicotine and intra-IC nicotine produce rightward shifts in morphine conditioned place preference and aversion (Loney, King, & Meyer, 2021), suggesting that dose perception is blunted via the IC for both the rewarding and aversive properties of the drug under nicotine. While other studies have shown the importance of the IC specifically in drug-conditioning (Geddes, Han, Baldwin, Norgren, & Grigson, 2008; Li, Zhu, Meng, Li, & Sui, 2013), very few have demonstrated disruptions specifically as rightward dose-response shifts, such that perception of dose is less sensitive under both appetitive and aversive conditioning. This raises the interesting possibility that escalation of alcohol intake under nicotine may also be driven by interoceptive factors surrounding the perception of intoxication, in addition to changes in motivational strength involving neuroadaptive changes in mesolimbic dopamine. To date however, escalation of drug-intake under control of IC NAchRs has yet to be established but raises an interesting possibility for dysregulated intake control.

CONCLUSION

In the current review, we discuss the various nicotine effects on drug-seeking and drug-taking behavior, ranging from the antecedent “appetitive” drug-seeking and approach behaviors, to the actual consummatory intake of drug and the resulting internal consequences. For alcohol in particular, the complex characterization of appetitive and consummatory processes have come a long way since early separation of the terms (Freedland, Sharpe, Samson, & Porrino, 2001; Samson, Czachowski, Chappell, & Legg, 2003; Sharpe & Samson, 2001) to include a robust ensemble of conditioning and reinforcing effects. Here, we argue that the behavioral effects of nicotine can be conceptualized as domains that influence appetitive conditioning (cues, context, and motivation), as well as consummatory behavior (intake and interoception), summarized in Fig. 6. By increasing the incentive value of both nicotine and alcohol paired stimuli, facilitating approach behaviors and resulting in activation of conditioned motivational states, nicotine can lead to behaviors that increase the likelihood of alcohol consumption. Then, nicotine can alter the reinforcing and intake properties of alcohol directly by both interfacing with the mesolimbic dopamine system and blunting at least some of the interoceptive properties of alcohol to promote consumption. Relatively few studies have directly tried to dissociate the primary versus reinforcement enhancing effects of nicotine, which highlight the need for additional work that directly assesses modulation alcohol seeking in the presence of cues separately from the reinforcing properties of alcohol consumption directly.

Figure 6: Schematic of how nicotine’s effects on the appetitive and consummatory processes of alcohol.

Nicotine can enhance appetitive responding and approach towards alcohol in the presence of 1a) alcohol-paired occasions setters and contexts, and 1b) discrete alcohol paired conditioned stimuli. Following approach and initiation of alcohol consumption, nicotine also can 2a) increase alcohol consumption following chronic treatment, although acute nicotine and decrease intake in some situations. Finally, 2b) nicotine is suggested to further enhance alcohol intake by blunting the post-ingestive consequences of alcohol, for example by blocking the perception of dose and intoxication.

Impact for treatment and cessation strategy

Development and validation of pharmacotherapies used for nicotine, alcohol, and codependence often involves using well validated rodent models with an understanding of the behavior concepts described in this review. Preclinical and clinical research do not always directly map onto each other. For example, although varenicline is used in treating alcohol and nicotine dependence (Bold et al., 2019; Erwin & Slaton, 2014; Hurt et al., 2018; McKee et al., 2009), preclinical data suggest that varenicline shows difficulty in reducing combined alcohol and nicotine intake (Funk et al., 2016; Maggio et al., 2018; Waeiss et al., 2019). Naltrexone has also been thought to be more effective in heavy drinkers that smoke (Fucito et al., 2012; A. King, Cao, Vanier, & Wilcox, 2009), although rodent models have had difficulty demonstrating reduced nicotine intake (Le et al., 2014; Maggio et al., 2018).

Importantly, this pattern of results show that it worth considering the behavioral model and outcome measure being used. For example, preclinical pharmacotheraputics often focus on drugs that reduce intake. Given what we know about the appetitive and motivational conditioning effects however, consideration of measures that reflect likelihood of initiating drug-use and motivation to work for drug, rather than expression free intake, may make better sense from a prevention perspective. Medications, mechanisms, and the generalization of varenicline’s effects in even in clinical data may also complicate the general pattern of findings (Motschman et al., 2016). As clinical models continue to improve the characterization of data in individuals and research groups (Gorelick & McPherson, 2015), and as animal models continue to refine observations to the various processes surrounding drug actions, well-defined behavioral measures will hopefully yield more convergent mechanistic data.

Differences in drug histories on treatment outcome

Considering the multitude of biobehavioral effects of nicotine, clinical outcomes and research may benefit by prioritizing the treatment of tobacco use disorder, an approach that has not directly been examined using preclinical modelling. Indeed, smoking is negatively related to treatment and clinical outcome for quitting drinking (Chiappetta, Garcia-Rodriguez, Jin, Secades-Villa, & Blanco, 2014; Weinberger et al., 2015). Instead, a “multimorbid” treatment approach that addresses both nicotine and other comorbid drug use may yield better outcomes (MacLean, Sofuoglu, & Rosenheck, 2018). In individuals undergoing treatment for alcohol use, smoking cessation does not appear to interfere with success maintaining alcohol abstinence (Gulliver, Kamholz, & Helstrom, 2006), which may be crucial maximizing treatment outcomes (Yokoyama et al., 2014) particularly in maximizing long-term success rates (Tsoh, Chi, Mertens, & Weisner, 2011). Approaches to encourage smoking cessation using contingency management may benefit from providing intermittent alternate reinforcement (Chudzynski, Roll, McPherson, Cameron, & Howell, 2015; Packer, Howell, McPherson, & Roll, 2012). Use of urinalysis to verify abstinence may be better predictor of substance use outcome (McPherson, Packer, Cameron, Howell, & Roll, 2014), and some recent evidence suggests that contingency management for both treatment of alcohol and tobacco use together can improve urinalysis outcomes for both (Orr et al., 2018). Preclinical models may eventually benefit from co-use models that examine behavioral outcomes following nicotine removal, or examining alternative reinforcement (Venniro et al., 2018).

Final remarks

The variety of approaches to studying drug-directed behavior under nicotine is remarkable, with the design of each study assessing different behavioral consequences of the drug. While these behavioral effects involve dissociable processes and multiple biological substrates, we suggest that it is useful to dissociate the various properties of nicotine into these various components of drug-directed behavior, specifically because it helps prevent the conflation of processes leading to, and resulting from, bouts of drug use.

Abbreviations:

aVTA

anterior ventral tegmental area

CS

conditioned stimulus

DLS

dorsolateral quadrant of the neostriatum

GABA

gamma aminobutyric acid

i.v.

intravenous

IC

insular cortex

DA

dopamine

FR

fixed ratio

mPFC

medial prefrontal cortex

NAcc

nucleus accumbens

NAccSh

nucleus accumbens shell

NAchR

nicotinic acetylcholine receptor

PR

progressive-ratio

pVTA

posterior ventral tegmental area

VTA

ventral tegmental area

VP

ventral pallidum

VS

visual stimulus