Racing the clock: The role of circadian rhythmicity in addiction across the lifespan

Abstract

Although potent effects of psychoactive drugs on circadian rhythms were first described over 30 years ago, research into the reciprocal relationship between the reward system and the circadian system – and the impact of this relationship on addiction – has only become a focus in the last decade. Nonetheless, great progress has been made in that short time toward understanding how drugs of abuse impact the molecular and physiological circadian clocks, as well as how disruption of normal circadian rhythm biology may contribute to addiction and ameliorate the efficacy of treatments for addiction. In particular, data have emerged demonstrating that disrupted circadian rhythms, such as those observed in shift workers and adolescents, increase susceptibility to addiction. Furthermore, circadian rhythms and addiction impact one another longitudinally – specifically from adolescence to the elderly. In this review, the current understanding of how the circadian clock interacts with substances of abuse within the context of age-dependent changes in rhythmicity, including the potential existence of a drug-sensitive clock, the correlation between chronotype and addiction vulnerability, and the importance of rhythmicity in the mesocorticolimbic dopamine system, is discussed. The primary focus is on alcohol addiction, as the preponderance of research is in this area, with references to other addictions as warranted. The implications of clock-drug interactions for the treatment of addiction will also be reviewed, and the potential of therapeutics that reset the circadian rhythm will be highlighted.

Introduction

The circadian clock is an ancient, phylogenetically conserved, time-keeping mechanism that has evolved to allow organisms to adapt to the daily rotation of our planet on its axis. In humans, the master circadian oscillator is located in the suprachiasmatic nucleus (SCN) of the anterior hypothalamus (Inouye & Kawamura, 1979; Stetson & Watson-Whitmyre, 1976) and acts essentially like a musical conductor by directing the appropriate period and phase of physiological homeostasis, with significant modulation from other hypothalamic nuclei (Saper, Scammell, & Lu, 2005). This homeostasis occurs not only in individual tissues, but in the coordination of molecular, physiological, and behavioral rhythms within and between systems. When these rhythms are disrupted, the resultant discordance between systems can exacerbate, or even cause, a wide spectrum of diseases and disorders from metabolic to mood disorders and addiction.

At the molecular level, the circadian clock exists as a transcriptional and translational feedback loop (TTFL) in nucleated cells, which serves to maintain a period of approximately 24 h and is also conserved phylogenetically from cyanobacteria to humans. As this topic has been the subject of many excellent, comprehensive reviews (Baker, Loros, & Dunlap, 2012; Cohen & Golden, 2015; Hardin & Panda, 2013; Takahashi, 2017; Tataroglu & Emery, 2015), we will briefly describe the central mechanism with two focal areas: the core clock components that have been studied in addiction, and aging-related changes in rhythmicity. In mammals, the molecular clock is a network of clock genes that form a TTFL in which the transcription factors BMAL, CLOCK, and neural PAS domain-containing protein comprise the positive arm of the feedback loop and drive the expression of two families of negative arm proteins, PERIOD (PER) and CRYPTOCHROME. Mutation or knockout of these core clock genes attenuates or ablates circadian rhythmicity at both the molecular and physiological levels. Specifically, functional mutations of either Per1 or Per2 shorten the circadian period of activity in mice, whereas mutation of both Per1 and Per2 or knockout of BMAL1 causes arrhythmicity (Bunger et al., 2000; Zheng et al., 1999, Zheng et al., 2001). Additionally, post-translational modifications of core clock proteins are needed to maintain the normal period, phase, and amplitude of circadian rhythms. For example, the serine-threonine kinase casein kinase 1 (CK1) binds to and phosphorylates PERs to mark them for turnover. Thus, CK1 inhibitors lengthen the period of the clock by increasing the levels of PER (Meng et al., 2010).

The widespread circadian control over biology has led to two general findings: 1) any change in circadian activity is likely to impact a number of systems and physiological functions, and 2) the circadian system offers a number of drug targets for pharmacotherapy in a variety of diseases. To the first point, knockout of BMAL1 is associated with obesity and metabolic dysfunction (Ferrell & Chiang, 2015), an early-aging phenotype that includes memory deficits and reduced hippocampal neurogenesis by 16 weeks of age (Ali et al., 2015), as well as reactive oxygen species-induced neurodegeneration (Musiek et al., 2013). Genetic polymorphisms in BMAL1 are also associated with a pattern of social drinking, whereas polymorphisms in the Per genes are associated with patterns of alcohol abuse (Dong et al., 2011; Wang et al., 2012) and knockout or mutation of the Per genes are associated with several addiction phenotypes in animal models (Abarca, Albrecht, & Spanagel, 2002; Agapito, Mian, Boyadjieva, & Sarkar, 2010; Brager, Stowie, Prosser, & Glass, 2013; Dong et al., 2011; Falcon & McClung, 2009; Gamsby et al., 2013; Perreau-Lenz, Zghoul, de Fonseca, Spanagel, & Bilbao, 2009; Spanagel et al., 2005; Wang et al., 2012; Zghoul et al., 2007). Therefore, as consumption and withdrawal from alcohol and drugs of abuse disrupt circadian rhythmicity (Currie, Clark, Rimac, & Malhotra, 2003; Danel, Cottencin, Tisserand, & Touitou, 2009; Danel, Libersa, & Touitou, 2001; Fonzi et al., 1994; Groh, Ehret, Peraino, Meinert, & Readey, 1988; Hasler, Bootzin, Cousins, Fridel, & Wenk, 2008; Imatoh, Nakazawa, Ohshima, Ishibashi, & Yokoyama, 1986; Kreek et al., 1983; Li et al., 2009; Pjrek et al., 2012), restoration of normal rhythmicity may be one avenue to reduce abuse and alleviate withdrawal symptoms. Nonetheless, several key questions remain to be investigated, while acknowledging the following caveats. Importantly, the large majority of work in this field has been done in animal models, which may not always reflect parallel changes to those occurring in the course of human addiction, and so the interpretation of the animal studies, including those presented in this review, must be considered cautiously in relation to clinical application. It is important to note the inherent difficulty in studying the endogenous rhythms of the SCN – to do so, all external cues that indicate daily cycles, like light, must be removed. In most clinical work, and some animal studies under rotating light:dark conditions, the changes in rhythmicity may result from changes in the SCN or in downstream pathways. Finally, little work has been done to elucidate the role of circadian desynchrony (CD) between internal time-keeping mechanisms and external time in addiction for elderly adults, and almost all work examining potential therapeutics that target the clock to reduce abuse or withdrawal symptoms has been done in clinical studies of young to middle-aged adults or in animal models of young to middle adulthood.

It is essential to examine the interactions of the circadian and reward mechanisms because these systems follow parallel developmental trajectories, with key transitions at the end of adolescence (around age 19) and in older adulthood (around age 65; Fig. 1). In the circadian system, the primary stimulus is light, conveyed via retinal melanopsin activation, which activates the retinohypothalamic tract to the SCN, followed by activation of the pineal gland and melatonin release as light shifts to dark. However, both the nighttime melatonin surge and average melatonin levels are highest in young children, and these measures decrease with age, rapidly declining in the first 20 years, then continuing to decline more slowly until older adulthood (at least age 70; Attanasio, Borrelli, & Gupta, 1985; Waldhauser et al., 1988). In animal models, the length of the active phase decreases gradually with age, with a greater effect in females, whereas the phase angle of entrainment plummets after adolescence, with greater drops only in the older male mice (Stowie & Glass, 2015). Individual chronotype, the innate preference to go to sleep and awaken at specific times of day, also varies with age in humans. Young children naturally fall asleep and awaken early (a so-called “early bird” or high morningness chronotype), but from childhood through adolescence, chronotypes shift to increasingly later bedtimes (a “night owl” or high eveningness chronotype). After hitting peak eveningness around age 20, chronotypes then reverse in trend and become increasingly early over the next 40 or more years (Roenneberg et al., 2004). Sleep architecture and quality also decline with age (specifically from middle age onward), largely due to impairments in circadian rhythmicity (Myers & Badia, 1995). One key finding in this area is that the evening chronotype is consistently associated with a greater risk of impulsive behavior, including alcohol and drug abuse (Hasler et al., 2017; Hasler, Sitnick, Shaw, & Forbes, 2013a; Kervran et al., 2015) and, interestingly, the prevalence of eveningness among adolescents correlates with the high rates of alcohol and drug abuse in this population.

Alcohol use disorders (AUDs) and substance use disorders (SUDs) remain a major issue in the United States despite a multitude of laws, campaigns, and preventative efforts meant to reduce them. Alcohol is by far the most commonly-abused drug across the lifespan. According to 2015 data from The National Survey on Drug Use and Health in the United States, almost half of the current 138 million alcohol users in the United States report problem (binge or heavy) alcohol consumption, and 15.7 million report an AUD. By contrast, only 64 million respondents reported past month tobacco use, 22 million reported past month marijuana use, and less than 13 million reported past month use of other illicit drugs, with a total of 5.1 million people reporting an SUD. An additional 2.7 million people reported an SUD co-occurring with an AUD (Administration, 2015). Most research into the causes of AUDs and SUDs has focused on adolescence and early adulthood, but with increasing life expectancy and greater independence throughout the geriatric period, addiction in elderly adults is a growing issue (Fig. 1; Simoni-Wastila & Yang, 2006).

This review examines how age-dependent changes in rhythmicity may affect the development and treatment of addiction because consumption patterns of many drugs of abuse, especially alcohol, shift during the same transitional periods in which circadian chronotypes tend to shift. For example, adolescents are more likely to binge drink and use marijuana than other age groups, and the level of alcohol and/or marijuana use correlates with their preference to stay up later at night and wake up later in the day (Hasler et al., 2017). On the opposite end of the spectrum, elderly adults may abuse alcohol or prescription drugs to help them compensate for disrupted circadian sleep patterns (Foley et al., 1995; Wang et al., 2015).

Adolescence (age 10–19) and young adulthood (age 20–25), represent a critical period of social, behavioral, and biological development, and often a time of intense peer and academic pressure (Larson & Asmussen, 1991). The prefrontal cortex (PFC) is still developing (Spear, 2000), and inhibitory control over the limbic system is attenuated; this, combined with a high sensitivity to reward (Ernst & Fudge, 2009), increases high-risk substance abuse during adolescence. Although rates of adolescent alcohol intake have decreased slightly over the past two decades, almost one-half of teenagers report consuming alcohol in the last month, with minimal differences between the sexes, and AUDs make up ~80% of teen substance use disorders (Administration, 2014). Three-quarters of the estimated $224 billion in annual alcohol-related costs are related to binge drinking (Bouchery, Harwood, Sacks, Simon, & Brewer, 2011; https://www.cdc.gov/features/alcoholconsumption/), and adolescents and young adults show the highest rates of binge drinking (Bouchery et al., 2011; Naimi et al., 2003). Rates of adolescent tobacco use and cocaine abuse have dropped by at least 20% over the last decade, whereas rates of both marijuana and heroin abuse have increased. However, rates of marijuana and heroin use disorders remain low in adults, suggesting that most use tapers off without resulting in an adult SUD (Administration, 2015).

Data suggest that a staggering 45% of those individuals who begin drinking in adolescence will meet the criteria for alcohol dependence later in life (Hasin, Stinson, Ogburn, & Grant, 2007; NIAAA, 2010). The National Epidemiological Survey on Alcohol and Related Conditions, the broadest of its kind, reported that approximately 30% of adults are high-risk drinkers (NIAAA, 2010), although only 5% self-report an AUD (Administration, 2015). By contrast, past-month marijuana, prescription drug, and illicit drug use drops by at least 50% in the transition from young adulthood to adulthood, and rates of SUDs, across all illicit substances, drop by 15% to 70% between young adulthood and adulthood (Administration, 2015).

On the opposite end of the spectrum, older adults have demonstrated increased rates of alcohol and drug abuse over the past decade; one reason that they may abuse alcohol or prescription drugs is to help them compensate for disrupted circadian sleep patterns (Foley et al., 1995; Wang et al., 2015). Further, alcohol and substance use in the elderly has only recently been recognized as a topic in need of greater research attention. One issue complicating research is that there is no set definition of the elderly period, as there is with the adolescent period. The most common operating definition is anyone over age 65, but this ranges as high as age 75, and varies within specific populations (Singh & Bajorek, 2014). With greater longevity comes a rapidly-expanding geriatric population, and with the greater independence of many elderly adults come more opportunities for alcohol and prescription drug abuse (Koechl, Unger, & Fischer, 2012). Recent studies have demonstrated that women with little social support and a history of substance abuse or affective disorders are at especially high risk of developing new issues with alcohol and substance abuse as they age (Simoni-Wastila & Yang, 2006). Alcohol remains the most-commonly abused substance, with 40% of individuals over 65 reporting past month drinking, and 10% reporting problem drinking, whereas only 1% report past month illicit substance use (Administration, 2010). However, even these rare instances of illicit substance abuse and lower levels of alcohol and drug consumption represent a crisis for a population with increased sensitivity to alcohol and other drugs, co-existing health concerns, multiple pharmaceutical prescriptions, and the likelihood of using alcohol and drugs to self-medicate for pain or distress. To the latter, addiction may develop even in old age, as anxiety, loneliness or deteriorating health drive moderate use into abuse, especially in the case of addictive mediations prescribed to vulnerable patients. For example, one of the most commonly prescribed drug classes in the elderly, benzodiazepines (Markota, Rummans, Bostwick, & Lapid, 2016), which reduce anxiety and aid in sleep initiation, are highly addictive and contraindicated for many elderly adults.