Effects of exogenous ghrelin administration and ghrelin receptor blockade, in combination with alcohol, on peripheral inflammatory markers in heavy-drinking individuals: Results from two human laboratory studies

Abstract

The ghrelin system has been garnering interest for its role in different neuropsychiatric disorders, including alcohol use disorder (AUD). Accordingly, targeting the ghrelin system is under investigation as a potential novel therapeutic approach. While alcohol provokes the immune system and inflammatory responses, ghrelin has potent immunomodulatory and anti-inflammatory properties. The present study aimed to shed light on the “crosstalk” between ghrelin and inflammation by examining the effects of exogenous ghrelin administration andghrelin receptor blockade on peripheral inflammatory markers in the context of two human laboratory studies with alcohol administration. Non-treatment-seeking, heavy-drinking individuals with alcohol dependence, the majority of whom were African American males, were enrolled. In the first randomized, crossover, double-blind, placebo-controlled human laboratory study, participants underwent two experimental paradigms – an intravenous alcohol self-administration (IV-ASA) and an intravenous alcohol clamp (IV-AC) – each consisting of two counterbalanced sessions (ghrelin, placebo). A loading dose of intravenous ghrelin (3 mcg/kg) or placebo, followed by a continuous ghrelin (16.9 ng/kg/min) or placebo infusion was administered. In the second dose-escalating, single-blind, placebo-controlled human laboratory phase 1b study, participants were dosed with an oral ghrelin receptor blocker (PF-5190457) and underwent an oral alcohol challenge. Repeated blood samples were collected, and plasma concentrations of the following inflammatory markers were measured: C-reactive protein (CRP), interleukin (IL)-6, IL-10, IL-18, and tumor necrosis factor alpha (TNF-α). During the IV-ASA experiment, significant drug × time interaction effects were observed for IL-6 (F_3,36 = 3.345, p = 0.030) and IL-10 (F_3,53.2 = 4.638, p = 0.006), indicating that ghrelin, compared to placebo, significantly reduced blood concentrations of the proinflammatory cytokine IL-6, while increasing blood concentrations of the anti-inflammatory cytokine IL-10. No significant drug × time interaction effects were observed during the IV-AC experiment, possibly because of its much shorter duration and/or smaller sample. Treatment with PF-5190457, compared to placebo, had no significant effect on the inflammatory markers investigated. In conclusion, a supraphysiologic pharmacological challenge with exogenous ghrelin in heavy-drinking individuals produced anti-inflammatory effects in the context of intravenous alcohol administration. On the contrary, ghrelin receptor blockade did not lead to any change in the inflammatory markers included in this study. Mechanistic studies are required to better understand the interaction between ghrelin, alcohol, and inflammatory processes.

1. Introduction

The existing literature on the role of neuropeptides in alcohol and other substance use disorders is rapidly growing. Several lines of evidence show that alterations in neuropeptide systems, resulting in, for example, changes in homeostatic levels, energy balance, reward, and/or emotional responses may contribute to the pathophysiology of addictive behaviors (Thiele, 2017). Accordingly, certain neuropeptide systems are being investigated as potential therapeutic targets for addictions. One such example is ghrelin, a peptide hormone that is increasingly being studied in relation to neuropsychiatric disorders, including alcohol use disorder (AUD) (Farokhnia et al., 2019; Morris et al., 2018; Koopmann et al., 2018).

Ghrelin is an orexigenic 28-amino acid peptide primarily produced by endocrine cells of the stomach and, to a lesser extent, of the intestines (Kojima et al., 1999; Kojima and Kangawa, 2005). A preproghrelin peptide, encoded by the GHRL gene on chromosome 3, is cleaved off to form the direct precursor of the ghrelin peptide, proghrelin (Gualillo et al., 2006). A portion of this proghrelin undergoes posttranslational modification, i.e. acylation via the ghrelin-O-acyl-transferase (GOAT) enzyme, before subsequent cleavage to form acyl-ghrelin. The remaining portion of ghrelin produced is released without this post-translational modification in the form of des-acyl ghrelin. Acyl-ghrelin is able to bind to a G-protein coupled receptor named growth hormone secretagogue receptor 1a (GHSR1a) to exert its physiological effects (Yang et al., 2008; Gahete et al., 2014; Yin et al., 2014). GHSR1a has high constitutive (ligand-independent) activity (Damian et al., 2012) and is expressed in various regions of the brain (e.g., hypothalamus, pituitary gland, ventral tegmental area, amygdala, and hippocampus), as well as peripheral organs (e.g., gut, pancreas, thyroid, heart, and adipose tissue) (Banks et al., 2002; Howick et al., 2017).

1.1. Ghrelin and alcohol

A well-documented function of the ghrelin system is regulation of appetite and food intake, mainly through the hypothalamic arcuate nucleus (De Vriese et al., 2010; Currie et al., 2005; Wren et al., 2001; Wren et al., 2000; Perelló and Zigman, 2012; Davis et al., 2012). In addition, ghrelin has been increasingly explored and implicated in different reward-seeking (Stievenard et al., 2017; Schellekens et al., 2013; Mason et al., 2014; Vengeliene, 2013) and stress-related behaviors (Chuang et al., 2011; Currie et al., 2012; Spencer et al., 2015; 2012). The ghrelin system’s interaction with neurobiological circuits such as the cholinergic-dopaminergic pathway suggests involvement in drug-seeking and addictive behaviors, which are influenced by positive (reward) and negative (stress) reinforcement mechanisms (Jerlhag et al., 2012). Evidence from both animal and human studies demonstrate that the ghrelin system is involved in the regulation of alcohol seeking, consumption, and other related outcomes (Zallar et al., 2017).

While early preclinical studies indicated no effect on alcohol intake following exogenous ghrelin administration (Lyons et al., 2008; Schneider et al., 2007), successive studies indicated the contrary (Zallar et al., 2017). Jerlhag and colleagues reported that while central ghrelin administration enhanced alcohol consumption and reward in mice, suppression of GHSR1a, both genetically and pharmacologically, attenuated this effect (Jerlhag et al., 2009). The same group also noted increased GHSR1a gene expression in different brain regions of high alcohol consuming rats, compared to low alcohol consuming rats (Landgren et al., 2011). Additional rodent studies have reported that ghrelin administration leads to enhanced alcohol intake, alcohol conditioned place preference (CPP), alcohol-induced locomotor activity, and accumbal dopamine release (Jerlhag et al., 2009; Cepko et al., 2014; Davis et al., 2012; Jerlhag et al., 2011; Jerlhag, 2008), while pharmacological antagonism of the GHSR1a suppresses these outcomes (Jerlhag et al., 2009; Landgren et al., 2011; Davis et al., 2012; Suchankova et al., 2013; Stevenson et al., 2016; Stevenson et al., 2015; Bahi et al., 2013; Kaur and Ryabinin, 2010). In some cases, however, these pharmacological effects were not of long duration, possibly due to sensitization (Gomez et al., 2015; Gomez and Ryabinin, 2014; Suchankova et al., 2016). Genetic manipulation of the ghrelin system has yielded similar results, i.e., knocking out the ghrelin peptide or receptor gene in rodents suppresses alcohol intake and other alcohol-related outcomes (Jerlhag et al., 2011; Bahi et al., 2013; Zallar et al., 2019; Zallar et al., 2019).

The bidirectional relationship between the ghrelin system and alcohol use has also been confirmed in human research. Case-control studies suggest that AUD patients have lower ghrelin levels when actively drinking, but have higher ghrelin levels when abstinent from alcohol (Addolorato et al., 2006; Badaoui et al., 2008; de Timary et al., 2017; Kim et al., 2013; Koopmann et al., 2012; Kraus et al., 2005; Leggio et al., 2012). Furthermore, an acute decrease in peripheral ghrelin concentrations following oral or intravenous alcohol administration has been observed in non-dependent social drinkers (Calissendorff et al., 2005; Calissendorff et al., 2006; Calissendorff et al., 2012; Zimmermann et al., 2007; Leggio et al., 2013; Ralevski et al., 2017). Observational studies comparing peripheral ghrelin concentrations between AUD individuals and controls are controversial, with some studies showing higher (Kraus et al., 2005; Kim et al., 2005; Wurst et al., 2007), while others showing lower, ghrelin levels in those with AUD (Addolorato et al., 2006; Badaoui et al., 2008; de Timary et al., 2017). These inconsistent findings may be explained, at least in part, by the different methodologies used across these studies (e.g., length of alcohol use, timing of blood collection, and characteristics of participants). Moreover, the form of ghrelin represented in these reports varies, with some studies evaluating the relationship between alcohol and total (acyl + des-acyl) ghrelin, and others specifically measuring acyl-ghrelin. Interestingly, peripheral concentrations of endogenous ghrelin in humans have been found to be positively associated with alcohol craving, reward sensitivity, impulsivity, subjective response to alcohol, and cue-induced brain activity (Leggio et al., 2012; Ralevski et al., 2017; Wurst et al., 2007; Akkişi Kumsar and Dilbaz, 2015; Koopmann et al., 2019; Ralevski et al., 2018). In addition, human laboratory studies conducted by our group indicate that intravenous administration of exogenous ghrelin increases alcohol craving (Leggio et al., 2014) and alcohol self-administration (Farokhnia et al., 2018), and modulates brain activity in response to reward anticipation (Farokhnia et al., 2018). Accordingly, the ghrelin system has begun to be investigated as a possible target to develop novel pharmacotherapies for AUD (Farokhnia et al., 2019). For example, we recently completed a set of experiments showing the safety of a GHSR1a blocker (PF-5190457), co-administered with alcohol, in both rodents and humans (Lee et al., 2020).

1.2. Alcohol and inflammation

Several lines of evidence suggest that alcohol use and dependence closely interact with the body’s immune system and its capacity for immune response (Coleman et al., 2018; Crews and Vetreno, 2014). While it is generally accepted that alcohol is a potent immunomodulator, the specific effects on the inflammatory cascade depend on a variety of factors such as acute versus chronic alcohol use. For example, alcohol consumption acutely suppresses immune cell activation, whereas chronic ethanol consumption results in enhanced proinflammatory cytokines production (Vetreno and Crews, 2014; Neupane, 2016). Markers of inflammation are broadly classified into proinflammatory (e.g., interleukin (IL)-1β, IL-6, IL-18, C-reactive protein (CRP), IL-18, tumor necrosis factor alpha (TNF-α)) (Slaats et al., 2016; Scheller et al., 2011) or anti-inflammatory (e.g., IL-10) (Siqueira Mietto et al., 2015). To date, there is no consensus on specific inflammatory markers that may be influenced as a result of AUD, and more so whether and how acute alcohol administration may affect these inflammatory pathways in social and/or chronic heavy alcohol drinkers. In most cases, elevated levels of the proinflammatory markers have been reported in AUD individuals (Crews and Vetreno, 2014; Vetreno and Crews, 2014; Leclercq et al., 2017; Míguez et al., 2012; Laso et al., 2007; Donnadieu-Rigole et al., 2016; Costello et al., 2013; Bell et al., 2017; Imhof et al., 2001; Kawaratani et al., 2013; González-Reimers, 2014; Portelli et al., 2019).

While immune response is critical for normal physiological functioning, chronic systemic inflammation may have detrimental consequences and has been shown to be involved in the pathophysiology of different physical and mental disorders (Vetreno and Crews, 2014; Bell et al., 2017; Bishehsari et al., 2017; Kosmidou et al., 2019; Urman et al., 2018; O’Brien et al., 2014; Esmailbeig and Ghaderi, 2017; Wohleb et al., 2016; Corsi-Zuelli et al., 2017; Crews et al., 2017; Hofford et al., 2019). As a chemical, alcohol provokes the immune system and activates microglial cells; the consequent inflammatory responses (e.g., cytokine production), in turn, contribute to the development and progression of AUD by triggering, for example, synaptic remodeling, alteration in neurotransmission, and changes in neuronal structure and function (Lacagnina et al., 2016; Cui et al., 2014). A close link has been also observed between inflammatory pathways and bio-behavioral disorders comorbid with AUD, e.g., depression and anxiety. Of note, some studies suggest that blood concentrations of proinflammatory markers positively correlate with levels of anxiety, depression and alcohol craving in AUD individuals (Leclercq et al., 2012; Leclercq et al., 2014). Accordingly, anti-inflammatory medications have started to be studied as potential novel treatments for psychiatric disorders, including alcohol and other substance use disorders (Jacobsen et al., 2016; Ray, 2014; Banks et al., 2018; Miller et al., 2017; Farokhnia et al., 2020). In a recent human laboratory study, for instance, Ray and colleagues found that the immunomodulator ibudilast, compared to placebo, significantly reduced alcohol craving and improved ratings of mood following exposure to stress and alcohol cues (Ray et al., 2017).

1.3. Ghrelin and inflammation

The presence of a close interaction between the ghrelin system and inflammatory pathways has been extensively reported. As an example, ghrelin was found to inhibit microglial activation and pro-inflammatory cytokine release in preclinical models of Parkinson’s disease, Alzheimer’s disease, epileptic seizures, ischemia, and traumatic brain injury (Lee et al., 2014; Bayliss and Andrews, 2013; Jeong et al., ; Lee et al., 2010; Cheyuo et al., 2011; Jiao et al., 2017). Peripheral ghrelin administration in a rodent model of inflammatory pain led to analgesic effects through elevation of IL-10 and transforming growth factor beta (TGF-β) levels (Azizzadeh et al., 2017), while in a rodent model of colitis, ghrelin administration reduced the expression of several proinflammatory cytokines (Gonzalez-Rey et al., 2006). Ghrelin’s known anti-inflammatory effects appear to be mediated through GHSR1a (Pereira et al., 2017), but the exact mechanism is not fully understood. For instance, microglial cells are known to express GHSR1a; activation of these receptors by ghrelin blocks microglial activation and suppresses the expression of proinflammatory cytokines (Moon et al., 2009). Collectively, previous studies provide ample evidence supporting a role for ghrelin as an anti-inflammatory and immunomodulatory hormone (Dixit and Taub, 2005; Baatar et al., 2011; Prodam and Filigheddu, 2014).

In order to better understand the “crosstalk” between ghrelin and inflammation, the present study aimed to investigate the effects of exogenous ghrelin administration and ghrelin receptor blockade on peripheral inflammatory markers in the context of experimental alcohol administration procedures. Specifically, we explored whether administration of intravenous acyl-ghrelin or an oral GHSR1a blocker may affect blood concentrations of pro-inflammatory (CRP, IL-6, IL-18, TNF-α) and anti-inflammatory (IL-10) markers in heavy-drinking alcohol-dependent individuals who received concomitant intravenous or oral alcohol.

2. Methods

2.1. Study setting and participants

This was a secondary investigation based on two previously published studies (Farokhnia et al., 2018; Lee et al., 2020). The first study examined the effects of exogenous ghrelin administration on intravenous (IV) alcohol self-administration and brain functional activity (Farokhnia et al., 2018). The second study investigated the safety and pharmacokinetic profile of a novel GHSR1a blocker, PF-5190457, when co-administered with oral alcohol (Lee et al., 2020). Candidates for each study were screened through a phone interview followed by an in-person screening visit; a comprehensive medical/psychiatric evaluation was conducted, and data required to assess eligibility were collected (for the list of inclusionary and exclusionary criteria, see Appendices 1 and 2). Eligible individuals were enrolled in the respective studies after providing written informed consent. Briefly, participants of both studies were non-treatment-seeking, heavy-drinking (>15 and >20 standard drinks per week for females and males, respectively) individuals with no clinically significant medical or psychiatry comorbidities. Both studies were conducted at the National Institutes of Health (NIH) Clinical Center in Bethesda, Maryland, and their inpatient setting allowed for a strictly controlled human laboratory design, where parameters such as diet, alcohol intake, and smoking breaks were closely monitored and standardized (see Figs. 1 and 2 and Appendices 3 and 4). Protocols were approved by the NIH Addictions Institutional Review Board, registered at ClinicalTrials.gov (NCT01779024 and NCT02039349), and conducted under Investigational New Drug applications following review by the Food and Drug Administration. Alcohol administration procedures were performed in accordance with the National Institute on Alcohol Abuse and Alcoholism (NIAAA) Council Guidelines on Alcohol Administration (https://niaaa.nih.gov/Resources/ResearchResources/job22.htm).

Fig. 1.

Schematic outline of the intravenous ghrelin study: (A) intravenous alcohol self-administration (IV-ASA) experiment (two visits), (B) intravenous alcohol clamp (IV-AC) experiment (two visits). Each participant underwent up to four experimental sessions. During each session, a 10-min loading dose of intravenous ghrelin/placebo was administered prior to time 0 (the start of continuous infusion). Measurement of inflammatory markers in the blood samples of this study was a secondary investigation.

Fig. 2.

Schematic outline of the PF-5190457 phase 1b study: (A) dosing phase, (B) alcohol challenge phase. Three identical visits were run, during which five doses of placebo (visit 1), PF-5190457 50 mg (visit 2), or PF-5190457 100 mg b.i.d (visit 3) were administered before the alcohol challenge on day 3. Time 0 is the first drug dose on each day. Measurement of inflammatory markers in the blood samples of this study was a secondary investigation.

2.2. Study design and procedures

2.2.1. Study 1 (intravenous ghrelin study)

A schematic outline of study 1 is depicted in Fig. 1. Each participant underwent up to four alcohol administration sessions: two IV alcohol self-administration (IV-ASA) and two IV alcohol clamp (IV-AC) sessions. The IV-ASA and IV-AC experiments each had a within-subject, crossover, randomized, double-blind, placebo-controlled design. During each session, participants received a 10-min loading dose of IV ghrelin (3 mcg/kg) or placebo, followed by a continuous ghrelin (16.9 ng/kg/min) or placebo infusion until the end of the session. Human acyl-ghrelin (PolyPeptide Laboratories Inc., Torrance, CA) was used for IV ghrelin infusions. Participants were admitted to the NIH Clinical Center the day prior to each study session and discharged the day after. A washout period of ≥3 days was implemented between the study visits.

Intravenous alcohol self-administration:

Under the IV-ASA paradigm, participants self-administered alcohol via the Computer-Alcohol Infusion System (CAIS) by pressing a button in a progressive ratio manner. This was the primary outcome of the parent study. Each alcohol infusion was designed to increase the breath alcohol concentration (BrAC) by 7.5 mg% over 2.5 min with a subsequent fall of 0.5 mg %/min until the following infusion. A safety limit of BrAC = 120 mg% was set. As reported in the parent study, exogenous ghrelin administration, compared to placebo, significantly increased the amount of alcohol self-administered during this experiment (Farokhnia et al., 2018).

Intravenous alcohol clamp:

Under the IV-AC paradigm, participants received a fixed dose of IV alcohol. This infusion was designed to increase the BrAC linearly to 80 mg% within 20 min and maintain (clamp) the BrAC at this level for 15 min (therefore, 35 min in total). The IV-AC was performed as part of the brain functional magnetic resonance imaging (fMRI) experiments of the parent study, as previously reported (Farokhnia et al., 2018).

2.2.2. Study 2 (PF-5190457 phase 1b study)

A schematic outline of study 2 is depicted in Fig. 2. This was a within-subject, dose-escalating, single-blind, placebo-controlled, phase 1b clinical trial with PF-5190457, an orally bioavailable GHSR1a blocker. Each participant underwent three identical visits, the only difference being drug condition across visits: placebo (visit 1), PF-5190457 50 mg (visit 2), and PF-5190457 100 mg (visit 3). Each visit started with a dosing phase during which the study drug (placebo or PF-5190457 50 mg or PF-5190457 100 mg) was administered twice on day 1, twice on day 2, and once on the morning of day 3 (5 doses in total). On day 3, approximately 30 min after the fifth drug dose, an oral alcohol challenge was performed. Participants were admitted to the NIH Clinical Center the day prior to the first drug dose and discharged the day after the alcohol challenge. A washout period of ≥3 days was implemented between the study visits.

Oral alcohol challenge:

A drink containing the same type of alcohol (Smirnoff vodka, 40% alcohol by volume) and each participant’s preferred mixer (see Appendix 5) was administered. The quantity of alcohol was calculated individually for each participant based on total body water to achieve a BrAC of 60 mg%. As reported in the parent study, PF05190457 was safe and well-tolerated, and no significant or clinically meaningful drug-alcohol interactions were found (Lee et al., 2020).

2.3. Blood collection, processing, and assays

For both studies, repeated blood samples were taken during each visit. The IV ghrelin study included two experiments, IV-ASA and IV-AC, with five blood sampling time-points each (Fig. 1). The PF-5190457 phase 1b study included two phases, a dosing phase and an alcohol challenge phase, with six and seven blood sampling time-points, respectively (Fig. 2). At each time-point, peripheral venous blood was collected into 3 mL BD Vacutainer tubes with 5.4 mg K2EDTA (Becton, Dickinson and Company, Franklin Lakes, NJ). Tubes were centrifuged within 30 min post-collection (relative centrifugal force: 1700×g, temperature: 4°C, centrifugation time: 15 min). The plasma supernatant was pipetted into 500 μL microtubes and the aliquots were stored at −80 °C until analysis. ProteinSimple Simple Plex Ella Immunoassay (ProteinSimple, Wallingford, CT) (Aldo et al., 2016) was used to quantify CRP (in study 1 only), IL-6, IL-10, IL-18, and TNF-α concentrations. Briefly, plasma samples were thawed on wet ice, agitated using a vortex mixer, and then centrifuged (in study 1 only; relative centrifugal force: 3000×g, temperature: 4 °C, centrifugation time: 10 min) to separate out lipid content from the plasma. Samples were diluted by a factor of 2 for IL-6, IL-10, IL-18, and TNF-α and a factor of 2000 for CRP using diluent supplied by the manufacturer. Fifty microliters of the diluted sample were loaded onto the Ella Simple Plex Immunoassay cartridge. IL-6, IL-10, IL-18, and TNF-α were assayed together on a multi-plex cartridge, while CRP was assayed individually using a single plex cartridge. Cartridges were processed using the provided ProteinSimple Ella software and equipment per manufacturer instructions.

2.4. Statistical analysis

Data were analyzed using Statistical Package for the Social Sciences (SPSS) (IBM Corp., version 25; Armonk, NY) and Statistical Analysis System (SAS) (SAS Institute Inc., version 9.4; Cary, NC). Prior to analysis, statistical outliers (±1.5 interquartile range per condition per time-point) were removed and normal distribution was assured. Repeated measurements of each marker were analyzed using linear mixed-effects models. Specifically, for study 1, the random effect of participant, the main effects of drug (ghrelin or placebo) and time (four post-baseline time-points), and the drug × time interaction effect on plasma concentration of each marker (as the outcome) were examined. Age, race, body mass index (BMI), session order (ghrelin first or placebo first), total number of alcohol infusions self-administered (IV-ASA paradigm only), and baseline concentration of each marker (first time-point) were tested as covariates in the initial run of each model, and significant covariates were retained in the final model. Pairwise comparisons between estimated marginal means of fitted models were adjusted by the Bonferroni procedure. A similar approach was employed to analyze data from study 2. Significance level was set at p < 0.05 (two-tailed) for all analyses.

3. Results

3.1. Study sample

Demographic characteristics of the study sample are presented in Table 1. The final sample included those participants who had blood samples available for this secondary analysis and finished both sessions (placebo and ghrelin) of each experiment for study 1 (n = 10 for the IV-ASA experiment, n = 6 for the IV-AC experiment) or all three visits (placebo, PF-5190457 50 mg, and PF-5190457 100 mg) for study 2 (n = 11).

Table 1.

Demographic characteristics of the study sample.

Variable	IV Ghrelin Study		PF-5190457 Phase 1b Study (n = 11)
	IV-ASA Experiment (n = 10)	IV-AC Experiment (n = 6)
Age, years, M (SEM)	39.13 (3.83)	39.45 (3.07)	39.22 (3.90)
Gender, males, n (%)	8 (80%)	5 (83%)	10 (91%)
Race, African Americans, n (%)	8 (80%)	5 (83%)	10 (91%)
BMI, kg/m2, M (SEM)	25.01 (0.85)	25.10 (1.10)	27.02 (1.48)
Education, years, M (SEM)	13.10 (0.46)	13.17 (0.83)	11.4 (0.91)
Age at first drink, M (SEM)	16.63 (0.73)	16.67 (0.76)	13.82 (1.20)
Alcohol drinking history1, M (SEM)
- Average drinks per drinking days	9.50 (2.33)	10.10 (2.76)	11.04 (1.29)
- Number of heavy drinking days	50.70 (7.65)	52.67 (11.52)	65.18 (5.36)
AUDIT score, M (SEM)
- Total score	21.2 (2.46)	27.00 (3.14)	21.18 (2.22)
- Hazardous alcohol consumption	9.78 (0.55)	11.00 (0.55)	10.09 (1.19)
- Alcohol dependence symptoms	6.11 (1.18)	9.00 (1.58)	5.73 (1.32)
- Alcohol-related problems	5.33 (1.51)	7.00 (1.30)	5.36 (1.19)
Current cigarette smoker, n (%)	8 (80%)	5 (83%)	8 (73%)
Cigarette pack years, M (SEM)	5.83 (2.27)	9.27 (3.21)	6.84 (2.67)
FTND score, M (SEM)	3.25 (1.00)	4.00 (1.23)	2.63 (0.56)

Abbreviations: AUDIT: Alcohol Use Disorders Identification Test; BMI: Body Mass Index; FTND: Fagerström Test for Nicotine Dependence; IV-AC: Intravenous Alcohol Clamp; IV-ASA: Intravenous Alcohol Self-Administration; M: Mean; n: Number; SEM: Standard Error of the Mean; TLFB: Timeline Followback.

¹Based on alcohol TLFB 90 days prior to the in-person screening visit.

3.2. Effects of intravenous ghrelin and PF-5190457, in combination with alcohol, on inflammatory markers

3.2.1. Study 1 (intravenous ghrelin study)

Table 2 and Fig. 3 outline the results during the IV-ASA experiment. Significant drug × time interaction effects were observed for IL-6 and IL-10. Post-hoc analyses showed that, under ghrelin compared to placebo, IL-6 concentrations were significantly lower at + 90 (p = 0.005) and + 120 (p = 0.016) time-points, and IL-10 concentrations were higher at + 60 (p = 0.022) and + 90 (p < 0.001) time-points. While the interaction term did not reach statistical significance for other markers, a significant drug main effect for CRP and IL-18, as well as a trend-level drug main effect for TNF-α was also observed, suggesting that plasma concentrations of pro-inflammatory cytokines were lower during the ghrelin than the placebo session.

Table 2.

Drug, time, and drug × time interaction effects on inflammatory markers during intravenous alcohol self-administration (IV-ASA) experiment of the intravenous ghrelin study.

Outcome	Drug Main Effect	Time Main Effect	Drug × Time Interaction Effect
CRP	F1,58.4 = 4.61, p = 0.036	F3,48.8 = 0.57, p = 0.635	F3,48.8 = 0.80, p = 0.499
IL-6	F1,36 = 10.56, p = 0.016	F3,36 = 3.93, p = 0.016	F3,36 = 3.34, p = 0.030
IL-10	F1,56.1 = 12.28, p = 0.001	F3,53.8 = 2.96, p = 0.040	F3,53.2 = 4.63, p = 0.006
IL-18	F1,57.3 = 6.18, p = 0.016	F3,49.7 = 0.51, p = 0.673	F3,49.3 = 0.64, p = 0.591
TNF-α	F1,61.6 = 2.89, p = 0.094	F3,52.9 = 3.11, p = 0.034	F3,52.5 = 0.58, p = 0.629

Abbreviations: CRP:C-reactive protein; IL-6: interleukin 6; IL-10: interleukin 10;IL-18: interleukin 18; TNF-α: Tumor necrosis factor alpha.

Fig. 3.

Plasma concentrations of CRP, IL-6, IL-10, IL-18, and TNF-α during intravenous alcohol self-administration (IV-ASA) experiment of the intravenous ghrelin study. Time 0 marks the start of continuous ghrelin/placebo infusion, following a 10-min loading dose of ghrelin/placebo. Alcohol self-administration started at time 0. Mean (M) and standard error of the mean (SEM) are depicted at each time-point. For statistical results, see Table 2. Abbreviations: CRP: C-reactive protein; IL-6: interleukin 6; IL-10: interleukin 10; IL-18: interleukin 18; TNF-α: Tumor necrosis factor alpha.

Table 3 and Fig. 4 outline the results during the IV-AC experiment. No significant drug × time interaction effects were observed, while the drug main effect reached statistical significance for IL-10 and IL-18. Of note, the IV-AC experiment had a shorter duration and a smaller sample compared to the IV-ASA experiment.

Table 3.

Drug, time, and drug × time interaction effects on inflammatory markers during intravenous alcohol clamp (IV-AC) experiment of the intravenous ghrelin study.

Outcome	Drug Main Effect	Time Main Effect	Drug × Time Interaction Effect
CRP	F1,51.8 = 2.75, p = 0.103	F3,47.0 = 0.34, p = 0.790	F3,47.0 = 0.81, p = 0.494
IL-6	F1,29.3 = 1.22, p = 0.278	F3,27.9 = 1.81, p = 0.740	F3,27.9 = 0.33, p = 0.800
IL-10	F1,33.4 = 8.34, p = 0.007	F3,32.1 = 0.41, p = 0.740	F3,32.1 = 0.32, p = 0.808
IL-18	F3,29.1 = 6.83, p = 0.014	F3,29.1 = 3.19, p = 0.038	F3,29.1 = 0.58, p = 0.632
TNF-α	F1,36 = 0.14, p = 0.709	F3,36 = 1.19, p = 0.324	F3,36 = 1.80, p = 0.163

Abbreviations: CRP:C-reactive protein; IL-6: interleukin 6; IL-10: interleukin 10; IL-18: interleukin 18; TNF-α: Tumor necrosis factor alpha.

Fig. 4.

Plasma concentrations of CRP, IL-6, IL-10, IL-18, and TNF-α during intravenous alcohol clamp (IV-AC) experiment of the intravenous ghrelin study. Time 0 marks the start of continuous ghrelin/placebo infusion, following a 10-min loading dose of ghrelin/placebo. Alcohol clamp started at time 30. Mean (M) and standard error of the mean (SEM) are depicted at each time-point. For statistical results, see Table 3. Abbreviations: CRP: C-reactive protein; IL-6: interleukin 6; IL-10: interleukin 10; IL-18: interleukin 18; TNF-α: Tumor necrosis factor alpha.

3.2.2. Study 2 (PF-5190457 phase 1b study)

Tables S1 and S2 and Figs. S1 and S2 outline the results during the dosing phase and the alcohol challenge phase. No significant drug main effect or drug × time interaction effects were observed, except for a significant drug main effect for TNF-α during the alcohol challenge phase.

4. Discussion

To our knowledge, this is the first study investigating the effects of exogenous ghrelin administration and ghrelin receptor blockade, in combination with alcohol, on peripheral concentrations of inflammatory markers in a clinically relevant sample of heavy-drinking individuals. Results showed that IV ghrelin, compared to placebo, reduced blood concentrations of proinflammatory markers and increased the anti-inflammatory cytokine IL-10 levels – a finding in line with ghrelin’s known function in reducing inflammation that has been observed in previous studies with different samples and methodologies (Pereira et al., 2017; de Candia and Matarese, 2018; Dixit and Taub, 2005; Baatar et al., 2011; Prodam and Filigheddu, 2014). During the IV-ASA experiment, we observed drug × time interaction effects for the proinflammatory cytokine IL-6 and the anti-inflammatory cytokine IL-10, indicating that exogenous ghrelin administration, compared to placebo, decreased IL-6 levels and increased IL-10 levels in the context of alcohol administration. A drug main effect was also found for the other markers (CRP, IL-18, and TNF-α), suggesting lower blood concentrations of proinflammatory cytokines under ghrelin, compared to placebo. Less prominent changes were observed during the IV-AC experiment. We did not find any significant drug × time interaction effects, and only a drug main effect was observed for IL-18 and IL-10, the latter being the most robust (Fig. 4) and in the same direction as in the IV-ASA experiment.

A number of factors may have contributed to statistically different results between the two alcohol paradigms of the IV ghrelin study. The IV-AC experiment included a much smaller sample size (n = 6) than the IV-ASA experiment (n = 10). Both exogenous ghrelin and alcohol were administrated for a longer period of time during the IV-ASA than the IV-AC experiment. We speculate that the lack of significant drug × time interaction effects during the IV-AC session could be due, at least in part, to the smaller sample and/or to not extending the sampling time enough to observe an effect. Specifically, during the IV-ASA experiment, blood samples were collected every 30 min, until 120 min, following intravenous alcohol initiation, whereas the last blood sample was collected 40 min after intravenous alcohol initiation for the IV-AC experiment (Fig. 1). Moreover, participants had the opportunity to administer more alcohol during the IV-ASA experiment than the IV-AC experiment (BrAC limit of 120 mg% and 80 mg% for IV-ASA and IV-AC, respectively). These differences are intrinsic limitations due to the secondary nature of this investigation, given that the two experiments were designed to address the main endpoints reported in the parent study (Farokhnia et al., 2018). Notwithstanding the methodological differences between the two experiments, an advantage of this study was that the same individuals participated in both IV-ASA and IV-AC (except for the dropouts), thus reducing possible inter-individual variabilities between the two samples. It is also important to note that the IV-AC experiment, compared to the IV-ASA experiment, had a longer period of pretreatment with exogenous ghrelin prior to alcohol infusion (40 versus 10 min; see: Fig. 1). On the other hand, the duration of exposure to alcohol was longer during the IV-ASA than the IV-AC experiment (120 versus 35 min; see: Fig. 1). Given that ghrelin’s anti-inflammatory properties were more prominent during the IV-ASA experiment, one may speculate that, at least within the context of this study, a robust alcohol-induced shift in the inflammatory system is needed in order to observe and detect ghrelin’s anti-inflammatory effects. Mechanistic studies are, however, required to prospectively test this hypothesis.

The observation that IV ghrelin administration produced anti-inflammatory effects in the context of alcohol administration, at least interms of peripheral concentrations of the markers investigated here, begs the question whether ghrelin receptor blockade may lead to proinflammatory effects. Notably, the results of the second study indicate that, oral administration of the ghrelin receptor blocker PF-5190457 did not significantly change blood concentrations of IL-6, IL-10, IL-18, or TNF-α, neither during the dosing phase (drug alone), nor during the alcohol challenge phase (drug plus alcohol). These negative findings are important from a safety perspective and further expand on our previous reports on the safety of PF-5190457 when co-administered with alcohol (Lee et al., 2020; Lee et al., 2019). It is important to note that our findings do not provide a direct side-by-side comparison of ghrelin administration versus GHSR1a blockade and their potential impact on inflammatory pathways. Methodological differences between studies 1 and 2, inherent pharmacological differences between the two compounds (exogenous ghrelin versus PF-5190457), and different routes of administration (IV versus oral) are just a few factors that must be taken into account. Furthermore, for the IV ghrelin study, an acute and robust supraphysiologic hyperghrelinemia was created by administering a high dose of exogenous ghrelin for a relatively short period of time before and during IV alcohol infusions. For the PF-5190457 phase 1b study, on the other hand, steady state was achieved by administering five doses of the PF-5190457 compound over the course of three days, before oral alcohol administration (see Figs. 1 and 2). These differences must be considered when interpreting and comparing the results of the two aforementioned studies.

The finding that a pharmacological challenge with exogenous ghrelin produced anti-inflammatory effects in the present study, especially on IL-6 and IL-10, and to a lesser extent on IL-18 and TNF-α, is consistent with numerous previous studies that support an immunoregulatory role for ghrelin (Gonzalez-Rey et al., 2006; Pereira et al., 2017; de Candia and Matarese, 2018; Dixit and Taub, 2005; Dixit and Taub, 2005; Baatar et al., 2011; Prodam and Filigheddu, 2014; Dixit et al., 2004; Taub, 2008; Wei et al., 2015). For example, both in vitro and in vivo preclinical studies indicate that ghrelin administration inhibits the production of the proinflammatory cytokines IL-6, IL-18, and TNF-α, while elevating blood concentrations of the anti-inflammatory cytokine IL-10 (Azizzadeh et al., 2017; Gonzalez-Rey et al., 2006; Dixit et al., 2004; Li et al., 2004). Unlike these cytokines, CRP, whose production is regulated by cytokines (particularly IL-6), has not been commonly assessed in relation to ghrelin administration. In the only study, to our knowledge, exogenous ghrelin administration did not change blood CRP levels in diet-induced obese mice or controls, but it did reduce IL-6 concentrations in the obese mice (Khazaei and Tahergorabi, 2015). In the present study, we found a significant drug main effect on CRP only during the IV-ASA experiment, where CRP levels were decreased under ghrelin, compared to placebo. No significant drug × time interaction effects on CRP was observed during either paradigm. Obviously, a side-by-side comparison of the two studies ((Khazaei and Tahergorabi, 2015) versus our IV ghrelin study) would not be relevant, given that the studies used different species (mice versus humans) and focused on different diseases (obesity versus alcohol dependence), just to name two main differences among many others.

The present report has some important limitations. First and foremost, this was a secondary investigation based on two previously published studies (Farokhnia et al., 2018; Lee et al., 2020); therefore, the experiments were not designed to a priori examine the outcomes presented in this report. The sample sizes were relatively small. There were a few missing samples, which were accounted for in the linear mixed-effects models. Having strict inclusionary and exclusionary criteria for enrollment (see Appendices S1 and S2) resulted in a homogenous sample of participants, thus reducing random variability in our analyses, but this factor also limits the generalizability of the results. The experimental design (drug dosage, meals, etc.) of both studies was standardized across different sessions and blood sampling was performed around the same time of the day, thus controlling for possible circadian changes of the inflammatory markers investigated (Morris et al., 2017; Keller et al., 2009). Having a predominantly African American male sample is also a noteworthy limitation of both studies. Racial and gender differences in inflammatory processes have been widely reported (Berkley et al., 2006; Casimir et al., 2011; Casimir et al.,2018; Ferguson et al., 2013; Feairheller et al., 2011; Gardner et al., 2015; Stewart, 2016; Schmeer and Tarrence, 2018), and future studies should be powered to examine ghrelin’s interaction with the immune system in different racial and gender groups. Finally, while both studies had a placebo to be compared with the study drug (ghrelin/PF-5190457), there was no control condition for alcohol in either study, thus making it difficult to disentangle the effects of ghrelin/PF-5190457 versus alcohol on these inflammatory markers – a limitation that could be addressed in the future via fully-factorial 2 (ghrelin/PF-5190457 versus placebo) × 2 (alcohol versus control) designs.

In conclusion, the present human laboratory studies indicate that A) a supraphysiologic ghrelin challenge in chronic heavy alcohol users produces anti-inflammatory effects which are not inhibited by co-administration of intravenous alcohol; B) ghrelin receptor blockade in chronic heavy alcohol users is safe and does not lead to changes in the inflammatory markers here investigated, either alone or co-administered with oral alcohol, at least within the doses and time exposure of this study. As shown in the parent studies, ghrelin administration leads to increased alcohol consumption (Farokhnia et al., 2018) and blocking the ghrelin signaling is currently being investigated as a potential treatment for AUD (Lee et al., 2020). While for obvious ethical reasons, the present study was limited to assessing peripheral inflammatory markers, a working hypothesis for future research is to see whether the downstream mechanisms of how the ghrelin system modulates addictive behaviors is mediated by neuroimmune pathways. Given that GHSR1a blockade is a pharmacological approach currently under investigation for AUD treatment, future work will need to further assess the potential impact this approach may have on inflammation and immunomodulation, especially in the context of alcohol use.

HIGHLIGHTS.

• Ghrelin system is under investigation as a potential therapeutic target for alcohol use disorder.

• Both ghrelin and alcohol have complex interactions with inflammatory pathways.

• Exogenous ghrelin, combined with alcohol, produced anti-inflammatory effects.

• Ghrelin receptor blockade, combined with alcohol, had no effect on inflammatory markers.