Background

Most smokers attempting to quit will fail often due to nicotine withdrawal symptoms including negative affect, irritability (enhanced responsivity to negative outcomes), and anhedonia (reduced responsivity to positive outcomes) [1, 2]. While such withdrawal symptoms are a product of dysregulated reward processing and the ventral striatum (VS), anterior cingulate cortex (ACC), and insula are regarded as constituents in the neurocircuitry of addiction [3, 4], emerging preclinical evidence implicates the habenula (Hb) as a contributor to negative reinforcement mechanisms perpetuating smoking [5]. The Hb is a small and understudied epithalamic nucleus that integrates information from limbic forebrain regions to modulate midbrain structures involved in monoamine neurotransmission. Specifically, the Hb inhibits dopamine releasing neurons following the absence of expected rewards [6], possesses a high density of nicotinic acetylcholine receptors [7], and is linked with the aversive effects of nicotine withdrawal and high nicotine doses [7, 8]. While the Hb’s small size limits its assessment in human fMRI studies [9], we utilized a performance feedback task previously shown to differentially activate the Hb, VS, ACC, and insula [10]. Our aims were three-fold, to: (a) characterize brain activity associated with positive and negative performance feedback (task effect), (b) elucidate brain activity differences as a function of a chronic smoking history (group effect: smokers vs. nonsmokers), and (c) delineate activity modulations as a function of acute pharmacological administration (drug effect: nicotine and varenicline).

Methods

Overnight-abstinent smokers (n = 17) and nonsmokers (n = 17) participated in 6 fMRI sessions during a two-drug, placebo-controlled, double-blind, crossover study. All participants were assessed twice (PATCH factor: once each wearing a transdermal nicotine patch and a placebo patch) at three points during a varenicline administration regimen (PILL factor: pre-pill, varenicline pill, placebo pill). Participants completed questionnaires to quantify trait-levels of addiction severity (FTND) and state-levels of negative affect and social anhedonia. To probe Hb functioning, we employed a ‘motion prediction task’ [10] in which participants predicted which of two movingballs, starting from different locations and traveling at different speeds, would reach a finish line first after viewing a short sequence of the balls’ motion. Task difficulty wasdynamically adapted thereby maintaining error rates at ~35% so that participants remained uncertain about their performance until feedback presentation. Participant responses (correct vs. error) were followed by feedback that did or did not provide information about trial outcomes (informative vs. non-informative). Task effects were assessed in a dependent samples t-test (informative-correct vs. informative-error trials, pcorrected < 0.001), group effects were assessed in an independent samples t-test (smokers vs. nonsmokers, pcorrected<0.05), and drug effects in an ANOVA (PILL by PATCH among smokers, pcorrected < 0.05).

Results

Regarding task effects, across all participants, we largely replicated a previous implementation of the task [10] and observed increased activity following negative feedback notably in the Hb, ACC, and bilateral anterior insula, and increased activity following positive feedback in the bilateral VS. Regarding group effects, smokers (vs. nonsmokers) showed reduced responsivity to positive feedback in the bilateral VS, yet increased responsivity to negative feedback in the left insula. Notably, greater reductions in VS responsivity following positive feedback correlated with higher FTND scores among smokers (r[16] = 0.60, p = 0.02) and more self-reported negative affect across all participants (PANAS, r[33] = 0.48, p = 0.004). Regarding drug effects, we observed a PATCH main effect such that placebo (vs. nicotine) administration was linked with elevated Hb activity following positive feedback among smokers, but not nonsmokers. A repeated measures correlation assessment among smokers indicated that greater Hb activity following positive feedback was linked with greater self-reported social anhedonia (r[74] = 0.28, p = 0.02), this relation was not observed among nonsmokers. No significant clusters were observed when considering the PILL or PILL by PATCH effects.

Conclusion

These results support contemporary views that an extended smoking history is associated with reduced responsivity to positive outcomes (anhedonia) and enhanced responsivity to negative outcomes (irritability). Critically, our results contribute to a growing literature demonstrating Hb involvement in positive and negative outcome processing and provide novel evidence that Hb activity in smokers can be modulated by nicotine.

Acknowledgements

The authors have no conflicts to declare and are supported by grants from NIDA (K01DA037819, R01DA041353), NICHD (U54MD012393, sub-project 5378), and by the NIDA Intramural Research Program.

References

[1] Piper et al., 2011, Psychopharm

[2] Cook et al., 2015, J. Abnorm Psych

[3] Koob & Volkow, 2010, Neuropsychopharm

[4] Sutherland et al., 2012, Neuroimage

[5] Baldwin et al., 2011, J Addict Res Ther

[6] Hikosaka, 2010, Nat Rev Neuro

[7] De Biasi & Salas, 2008, Exp Biol Med

[8] Fowler et al., 2011, Nature

[9] Salas et al., 2010, Frontiers in Human Neurosci

[10] Ullsperger & von Cramon, 2003, J Neurosci