Obsessive-compulsive disorder (OCD) is the fourth most common mental disorder (Pittenger et al., 2005). It has a lifetime prevalence of 2.3% and a 12-month prevalence of 1.2% (Ruscio et al., 2010). Patients suffering from OCD suffer from persistent obsessive thoughts causing distress, and perform compulsive repetitive behaviors to alleviate uncomfortable feelings resulting from obsessions (Diniz et al., 2012). OCD can have disabling effects throughout the patient’s lifespan in both males and females (Attiullah et al., 2000).

Obsessions can be thematic, such as fear of contamination, pathological doubt, or need for symmetry/order, or somatic obsessions, like aggression. Repetitive compulsive behaviors involve, washing, seeking, counting, sorting, hoarding and searching (Doron and Moulding, 2009; Goit and Ghimire, 2014; Pauls et al., 2014). Although recently declassified as an anxiety disorder (American Psychiatric Association, 2015), many clinicians conceptualize OCD as a spectrum of related disorders (OCRD) sharing common clinical features of anxiety/fear and worry (Stein and Lochner, 2006; Storch et al., 2008; Fornaro et al., 2009). OCRD encompasses a wide range of diseases which includes somatoform (e.g., Hypochondriasis), impulse control (e.g., Trichotillomania, pathological gambling) and tic disorders (e.g., Tourette’s syndrome; Fornaro et al., 2009). Selective serotonin reuptake inhibitors (SSRIs) and cognitive behavioral therapy or their combination are often used as first line treatments. However, a large group of patients remain resistant to treatment either partially or completely (Jenike, 2004; Pittenger et al., 2005).

The cholinergic system in the brain is comprised primarily of nicotinic acetylcholine receptors (nAChRs; Paterson and Nordberg, 2000; Kalamida et al., 2007) and muscarinic acetylcholine receptors (mAChRs; Scarr, 2012; Thiele, 2013), members of the cys-loop superfamily of ligand gated ion channels and G protein-coupled receptors, respectively. Dysregulation of both mAChRs and nAChRs have been strongly associated with several neurological disorders (Janowsky et al., 1972; Freedman et al., 1995; Warpman and Nordberg, 1995; Breese et al., 2000; Salamone and Zhou, 2000; Perry et al., 2001; Woodruff-Pak and Gould, 2002; Ray et al., 2005; Quik et al., 2007; Scarr, 2012).

The α4β2 nAChR is one of the most prevalent nicotinic subtypes expressed in the brain (McGranahan et al., 2011). The α4β2 subtype is expressed in abundance in the dopamine pathways in the midbrain that influence the drug-induced reward system, mood disorders, stress, movement generation and learning (Wise, 2009; Maskos, 2010). α4β2 nAChRs have also been identified in the striatum, thalamus and cortex (Quik et al., 2013), brain areas implicated in OCD (Pena-Garijo et al., 2010; Fitzgerald et al., 2011). In the striatum, α4β2 receptors have also been shown to modulate GABA and dopamine release (McClure-Begley et al., 2009; Perez et al., 2012). In particular, a subtype of the α4β2 receptor with high sensitivity to acetylcholine (HS α4β2) appears to be involved in striatal dopamine release (Anderson et al., 2009). These studies support a modulatory role of α4β2 receptors in neurotransmitter release in circuits affected in OCD.

Positive allosteric modulators (PAMs) enhance agonist responses via increased agonist potency and/or efficacy. Desformylflustrabromine (dFBr) is a novel PAM capable of potentiating acetylcholine-induced whole cell responses by 370% for the HS and 260% for the low sensitivity (LS) α4β2 receptors with an EC₅₀ of 40 μM and 2.5 μM respectively (Weltzin and Schulte, 2010, 2015). It is currently the only selective PAM for α4β2 receptors capable of potentiating the HS form of the receptor involved in striatal dopamine release. As dFBr increases the efficacy of acetylcholine and does not directly activate receptors, it is postulated that its effect in the synapse would be to enhance acetylcholine mediated transmission. Application of dFBr, unlike application of exogenous agonists, would thus retain the control of synaptic activation via presynaptic release of acetylcholine, albeit with increased stimulation (Weltzin and Schulte, 2015). Only one in vivo study has been conducted to examine the effect of dFBr potentiation of α4β2 nAChR in an in vivo behavioral model. In this study dFBr was shown to attenuate nicotine self-administration in rats (Liu, 2013). The use of HS α4β2 receptors PAMs for the treatment of OCD has not been previously proposed or tested in any animal model. The aim of the current study was to evaluate our hypothesis that acute and chronic administration of dFBr, a novel PAM specific for α4β2 nAChRs and active at the HS α4β2 subtype, will attenuate compulsive-like and anxiety-like behaviors in our non-induced compulsive-like mouse model.

There are few animal models that exhibit consistent and spontaneous differences in compulsive-like behaviors. We have previously shown that our mice exhibit face and predictive validity as a spontaneous non-induced model for OCD-like behaviors (Greene-Schloesser et al., 2011). The current model was achieved by bidirectionally selecting house mice, Mus musculus, for nest-building behavior for 56 generations (Lynch, 1980; Bult and Lynch, 2000). The stock population for the original selection experiment (Lynch, 1980) was a cross among eight inbred strains, i.e., A, AKR, BLB/c, C3H/2, C57BL, DBA/2, Is/Bi and RIII, to yield the HS/Ibg outbred strain (McClearn Ge and Meredith, 1970; Lynch, 1980). Bidirectional selection resulted in three levels of nesting behavior (NB). All BIG mice exhibit consistent excessive NB engaging in rapid and repetitive pulling of cotton through the cage top metal bars amounting to 6–7 g of cotton on an average in 24 h when compared to normal NB (no significant hyperactivity and repetitiveness when introduced to cotton averaging around 0.50–0.70 g in 24 h) by the Control strain (non-compulsive) and very little NB (most of them do not indulge in nesting) by the SMALL strain (non-compulsive). The Control mice therefore serve as a selection control with intermediate levels between compulsive-like BIG and non-compulsive SMALL strains (Bult and Lynch, 2000). NB is homologous to hoarding in humans with OCD (Warneke, 1993), which is considered to be a measure of compulsive-like phenotype in mice (Greene-Schloesser et al., 2011; Wolmarans De et al., 2016). The BIG mice also uniformly display repetitive marble burying (MB) behavior burying on an average 19–20 marbles. Both these behaviors are significantly attenuated by SSRIs (e.g., fluoxetine) used to treat OCD but not with normal antidepressants (e.g., desipramine; Greene-Schloesser et al., 2011) substantiating the face and predictive validity of the NB and MB phenotype of the BIG mice for investigating compulsive disorders. Hence in the current context of investigation compulsive-like BIG mice have been considered.

Evidence exists for cholinergic involvement in OCD (Lucey et al., 1993; Yankelevitch-Yahav and Joel, 2013). Some studies have indicated exacerbation of OCD symptoms induced by nicotine (Abramovitch et al., 2015). In contrast to the higher rates of smoking in patients with psychiatric disorders, such as schizophrenia, bipolar disorder and ADHD, OCD patients report less smoking behavior (Bejerot and Humble, 1999; Bejerot et al., 2000; McCabe et al., 2004; Abramovitch et al., 2014). It has been suggested that nicotinic activation of an already hyperactivated fronto-striatal circuit worsens OCD symptoms (Abramovitch et al., 2015). However, other studies have shown that nicotine augmentation improves clinical symptoms in patients with OCD (Carlsson, 2001; Pasquini et al., 2005). Glutamatergic hyperactivity associated with OCD may also be due to mediation of glutamate release by nicotinic receptor activation. (Araki et al., 2002; Mansvelder et al., 2002; Pasquini et al., 2005). Studies investigating cholinergic involvement in glutamatergic hyperactivation suggest that nicotine promotes glutamatergic transmission and stabilizes hyperactivity of the neural circuit that originates in the orbitofrontal cortex and projects to the cingulate gyrus, the striatum and the thalamus (Pasquini et al., 2005). PET and fMRI studies in OCD subjects have shown elevated cerebral blood flow, metabolism and activation (indicators of hyperactivity) in the orbitofrontal cortex and amygdala in OCD (Busatto et al., 2000; Carlsson, 2000; Menzies et al., 2008). These regions receive substantial cholinergic innervations (Mesulam et al., 1986; Carlsson, 2000). Based on these prior studies, we investigated the modulatory role of α4β2 nAChRs in compulsive-like and anxiety-like behaviors in the compulsive-like mice model.

Administration of the novel α4β2 PAM, dFBr produced a reduction in compulsive-like NB and MB, but did not alter anxiety-like and locomotor activity in the OF for the acute study. A very similar response to chronic dFBr was observed where the treatment groups showed rapid suppression of NB (1 h and 2 h) and MB (2 h) after dFBr administration. OF behaviors however remained unaffected by the chronic treatment. These results indicate an apparent selectivity of dFBr for compulsive-like behaviors corroborating the hypothesis that potentiation of α4β2 nAChRs could be an alternative approach for suppressing compulsive-like phenotype thereby posing significant translational potential.

In the acute administration, 4 mg/kg and 6 mg/kg dFBr doses had the largest attenuating effects on NB 2 h after injection, while for the chronic administration the suppression effects on NB was visible after the first hour and endured in the second hour with 6 mg/kg showing a more consistent effect. Interestingly, an earlier effect of dFBr (1 h after administration) on NB was observed for the chronic study indicating potential sensitization to dFBr due to repeated treatment. The attenuating effects gradually decreased during the next 4 h for both the treatment, showing that dFBr had a rapid effect. This result is consistent with the finding that peak levels of dFBr in the cerebrospinal fluid occur 90 min after administration in rats (Liu, 2013). The 2 mg/kg dFBr dose had no immediate attenuating effect on NB. A long term effect of this dose was however seen in both acute (after 24 h) and chronic (week 5) administration indicating that this dose was effective over a longer time period.

The effects of dFBr, 2 h after injection on MB behavior were generally similar to the effects on NB. However, at the 2 h time point in the acute treatment 2 mg/kg moderately and significantly reduced MB behavior. This effect was not significant in the chronic regimen. No significant effect was observed on NB at the same dose and time point in the acute study but had an effect in the chronic study. These different effects of dFBr treatment may indicate subtle differences in the brain mechanisms that control NB and MB behavior. Clinical studies have shown that some OCD patients with specific types of symptoms do not respond to first line therapies in a similar way (McKay et al., 2004). The doses that act to attenuate obsessions and compulsions in general OCD patients typically fail to produce results in treatment resistant ones (Albert et al., 2013). Moreover, recommended doses for first line treatments might vary depending on the severity of the disorder, co-morbid symptoms like anxiety and potential side effects (Hanna et al., 2011; Albert et al., 2013). Though, a common agreement on OCD subtypes is lacking, therapeutic response and results for each OCD subtype are different (Alonso et al., 2001). For example, fluoxetine, a common OCD drug has greater efficacy in washers and obsessive thoughts when compared to checkers (Farnam et al., 2008). Therefore, the variation in dose response to dFBr of compulsive-like MB and NB behavior adds additional heterogeneity to the BIG mouse for assessing drug effects on various compulsive-like phenotypes.

Acute and chronic dFBr regimen failed to modulate anxiety-like (time spent in center) and locomotor (total distance traveled) behaviors in the OF test. Previous studies using the BIG mice have shown a similar effect of SSRIs like fluoxetine, which failed to reduce overall wheel-running locomotion in the compulsive-like BIG mice but significantly attenuated NB and MB behavior (Greene-Schloesser et al., 2011). Separate brain regions and signaling pathways influencing compulsive-like and anxiety-like symptoms are most likely the explanation for the observed lack of a dFBr effect in the OF test. Anxiety is attributed primarily to the amygdala and ventral hippocampus (McHugh et al., 2004), whereas compulsions and obsessions have been linked to dorsolateral prefrontal cortex (Hirosawa et al., 2013), anterior cingulate cortex (Fitzgerald et al., 2005), orbitofrontal cortex (Beucke et al., 2013) and dysregulation of the corticostriatal-thalamo-cortical circuitry (CSTC; Ting and Feng, 2011). These regions receive projections from the amygdala and hippocampus (McDonald, 1991; Eblen and Graybiel, 1995; Welch et al., 2007; Toyoda et al., 2011; Chen and Etkin, 2013) explaining the co-existence of anxiety along with OCD, which appears to be specific to anxiety related to compulsive-like behaviors rather than more generalized anxiety.

Removal or inhibition by antagonists of α4β2 nAChRs abolishes the anxiolytic effects of nicotine, while stimulating these nAChRs receptors with an agonist decreases anxiety-like behavior. In contrast, anxiogenic effects of nicotine withdrawal are enhanced by stimulation of α7 nAChRs and decreased by inhibition of these nAChRs receptors (Kutlu and Gould, 2015). Allosteric modulation of α4β2 nAChRs by dFBr did not affect anxiety-like behavior in the OF test in the BIG mice, suggesting that these nAChRs receptors may not be involved in the control of anxiety in nicotine-naïve mice. A partial agonist of α4β2 nAChRs (ABT-089) caused anxiogenic effects in nicotine-naïve mice (Yohn et al., 2014). Whether this result contradicts our findings or could be due to low affinity of ABT-089 for α7 nAChRs remains to be determined.

In summary, both acute and chronic dFBr was effective in reversing compulsive-like NB and MB, without exerting any influence on anxiety-like and locomotory behaviors. This indicates the therapeutic potential of modulation of α4β2 nAChRs by dFBr for compulsive phenotypes. Due to the rapid rate of onset (a few hours) of the attenuating effects of dFBr on compulsive-like behaviors, this class of specific nicotinic subtype modulators might also provide more immediate suppression effects thereby provide a bridging option to other first line therapies (e.g., SSRIs) that display longer time courses for onset of effectiveness. dFBr maintained its attenuating effects on NB and MB during chronic treatment, and may therefore also represent a novel first line treatment. However, the cellular mechanisms leading to such acute and chronic suppression of compulsive-like behavior and the role of upstream and downstream targets that ultimately modulate phenotypic expression of the behaviors remains to be elucidated. It also remains to be determined if this effect of dFBr is consistent across all rodent models of compulsive-like phenotype. The current study thereby provides a strong impetus for further exploration of these factors in otherwise sparsely explored area of the role of nAChRs in OCD.

SM, MM and SNK conducted all experiments and performed data analysis. SM and SNK lead manuscript writing efforts. AB-I, RG and MKS made significant contributions to research design, data interpretations and manuscript preparation.

Research reported in this publication was supported by an Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health under grant number P20GM103395 to SM and AB-I. This work was also supported by an Undergraduate Research and Scholarly Activity grant (#S15-24) to MM. The College of Natural Sciences and Mathematics also supported this work. These funding sources did not have a role in the study design, collection, analysis and interpretation of data and submission of this article for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.