Oxytocin differentially modulates specific dorsal and ventral striatal functional connections with frontal and cerebellar regions

Highlights

• Randomized placebo-controlled resting state pharmaco-fMRI study (n = 144, males).

• Effects of oxytocin (OXT) on basal ganglia sub-region connectivity were examined.

• OXT increased connectivity of ventral striatal & pallidal nodes with frontal regions.

Abstract

Interactions between oxytocin and the basal ganglia are central in current overarching conceptualizations of its broad modulatory effects on behavior. Whereas evidence from animal models emphasizes the critical role of the ventral striatum in the behavioral effects of oxytocin, region-specific contributions of the basal ganglia have not been systematically explored in humans. The present study combined the randomized placebo-controlled administration of oxytocin versus placebo in healthy men (n = 144) with fMRI-based resting-state functional connectivity to determine the modulatory role of oxytocin on the major basal ganglia pathways. Oxytocin specifically increased connectivity between ventral striatal and pallidal nodes with upstream frontal regions, whereas it decreased the strengths of downstream pathways between the dorsal striatum and posterior cerebellum. These pathways have previously been implicated in salience, reward and behavioral flexibility, thus shaping goal-directed behavior. Given the importance of aberrant striatal intrinsic organization in autism, addiction and schizophrenia the present findings may suggest new mechanistic perspectives for the therapeutic potential of oxytocin in these disorders.

Introduction

The evolutionary conserved neuropeptide oxytocin (OXT) is a key regulator of social behavior, particularly social bonding, reward and salience processing (Gao et al., 2016; Gordon et al., 2011; Love, 2014; Ma et al., 2016; Numan and Young, 2016; Shamay-Tsoory and Abu-Akel, 2016). Converging evidence from animal models and human oxytocin administration studies suggests that the oxytocinergic regulation of social behavior involves modulation of regional activity as well as interregional coupling in basal ganglia-limbic-cortical networks (Bethlehem et al., 2013; Johnson and Young, 2017; Liu et al., 2015; Wang et al., 2017; Wigton et al., 2015). Whereas previous OXT administration studies in humans predominantly focused on the modulation of the amygdala-centered functional network (Ebner et al., 2016; Eckstein et al., 2017; Koch et al., 2016; Sripada et al., 2013; Wang et al., 2014), animal models have established converging evidences that the effects of OXT on social bonding (Amadei et al., 2017), reward (e.g. King et al., 2016), and also its intrinsic functional effects (Moaddab et al., 2015) are critically mediated by basal ganglia core nodes, particularly the ventral striatum and the pallidal formation (overview in Young et al., 2011).

Traditionally conceived as motor control region it has become increasingly clear that the basal ganglia critically contributes to a broad range of behavioral domains in humans ranging from basic salience/reward and novelty processing to complex decision making (Arsalidou et al., 2013; Haber, 2016). The broad behavioral contribution is mirrored in the complex functional organization of the basal ganglia, with sub-division- and function-specific bidirectional connections to nearly all cortical regions (Alexander et al., 1986; Haber et al., 2000). Within this functional organization the striatum as major input site for afferent cortical projections and the pallidal formation projecting back to the cortex (Draganski et al., 2008; Haber, 2016) appears of major importance for the integrative function of the basal ganglia in orchestrating motor, cognitive and motivational/emotional processes. The striatum comprises a set of sub-regions which according to their functions and projections are assigned to the ventral striatum (nucleus accumbens, ventral putamen/caudate), interacting with limbic and prefrontal regions to promote motivational/emotional processes, and the dorsal striatum (dorsal putamen, caudate), interacting with dorsolateral-prefrontal and motor regions sub-serving cognitive and motor control functions (Haber, 2016; Kelly et al., 2009). Despite the central role of the basal ganglia in overarching conceptualizations on the role of OXT in human behavior (Love, 2014; Ma et al., 2016; Shamay-Tsoory and Abu-Akel, 2016) previous empirical evidence is mainly based on animal models indicating a central role of ventral striatal nodes in mediating behavioral effects of OXT (Young et al., 2011). Accumulating evidence from human research combining the intranasal application of OXT with task-based fMRI suggests modulatory effects in functional domains which are strongly related to the basal ganglia, including learning-based social adaptations (Baumgartner et al., 2008; Rilling et al., 2012), novelty (Wittfoth-Schardt et al., 2012) and reward processing (Hu et al., 2015; Mickey et al., 2016; Scheele et al., 2013). However, depending on the task paradigms employed different basal ganglia sub-regions have been found to be modulated by OXT, including both, dorsal and ventral striatal nodes (Hu et al., 2015; Scheele et al., 2013; see also Bethlehem et al., 2014 or Dölen et al., 2013). Overall these previous findings suggest that the specific effects of OXT observed are the result of an interplay between the task-specific network and the social-emotional stimuli employed.

With respect to demonstrating state-dependent basal ganglia functioning and mapping basal ganglia sub-region-specific networks, resting-state fMRI functional connectivity (rsFC) approaches have been extensively used (Di Martino et al., 2008). Moreover, neuropsychiatric disorders characterized by deficient reward, salience and social-emotional processing including autism, schizophrenia and depression have consistently been associated with aberrations in intrinsic basal ganglia networks directly related to core underlying deficits such as anhedonia (Gabbay et al., 2013) and lack of social interest (Delmonte et al., 2013). Resting-state pharmaco-fMRI approaches allow investigation of the effects of OXT on the intrinsic functional organization of the brain in the absence of context- or stimulus-specific influences. Indeed, previous amygdala-focused resting-state pharmaco-fMRI studies have revealed consistent findings with respect to OXT's modulatory effects on amygdala-prefrontal connectivity (Eckstein et al., 2017; Sripada et al., 2013; for a general discussion on the modulatory potential of OXT on the level of functional networks see also Bethlehem et al., 2013). Moreover, recent data-driven approaches have consistently revealed modulatory effects of OXT on basal ganglia-frontal circuits, particularly major afferent targets of the striatum such as anterior cingulate and frontal motor cortices (Bethlehem et al., 2017; Paloyelis et al., 2016). However, despite the central role of the basal ganglia in recent conceptualizations of OXT and the clinical relevance of basal ganglia pathology, specific intrinsic effects of OXT on the human basal ganglia sub-regions have not been examined. Therefore, the present study employed a randomized, double-blind, placebo-controlled, pharmaco-rsFC experiment in healthy men (n = 144) to determine effects of intranasal-OXT on sub-regional intrinsic functional connectivity of basal ganglia. Based on previous findings in humans (Bethlehem et al., 2017; Paloyelis et al., 2016), and the key role of the ventral striatum in OXT's functional effects in animal models (King et al., 2016; Moaddab et al., 2015), we expected to observe enhanced functional connectivity between the ventral striatum and its afferent projection sites in limbic and frontal regions.