Relaxin-3 receptor (Rxfp3) gene knockout mice display reduced running wheel activity: Implications for role of relaxin-3/RXFP3 signalling in sustained arousal

Highlights

• Evidence suggests relaxin-3 can activate RXFP3, but also RXFP1.

• Study assessed basal behavioural phenotype of novel Rxfp3 KO mouse line.

• Rxfp3 KO mice demonstrate reduced voluntary running wheel activity and anxiety.

• Phenotype of Rxfp3 KO mice mirrors reported relaxin-3 KO phenotype.

• Findings support relaxin-3/RXFP3 pairing and their role in arousal.

Abstract

Anatomical and pharmacological evidence suggests the neuropeptide, relaxin-3, is the preferred endogenous ligand for the relaxin family peptide-3 receptor (RXFP3) and suggests a number of putative stress- and arousal-related roles for RXFP3 signalling. However, in vitro and in vivo evidence demonstrates exogenous relaxin-3 can activate other relaxin peptide family receptors, and the role of relaxin-3/RXFP3 signalling in specific brain circuits and associated behaviours in mice is not well described. In this study, we characterised the behaviour of cohorts of male and female Rxfp3 gene knockout (KO) mice (C57/B6J^{RXFP3TM1/DGen}), relative to wild-type (WT) littermates to determine if this receptor KO strain has a similar phenotype to its ligand KO equivalent. Rxfp3 KO mice displayed similar performance to WT littermates in several acute behavioural paradigms designed to gauge motor coordination (rotarod test), spatial memory (Y-maze), depressive-like behaviour (repeat forced-swim test) and sensorimotor gating (prepulse inhibition of acoustic startle). Notably however, male and female Rxfp3 KO mice displayed robust and consistent (dark phase) hypoactivity on voluntary home-cage running wheels (∼20-60% less activity/h), and a small but significant decrease in anxiety-like behavioural traits in the elevated plus maze and light/dark box paradigms. Importantly, this phenotype is near identical to that observed in two independent lines of relaxin-3 KO mice, suggesting these phenotypes are due to the elimination of ligand or receptor and RXFP3-linked signalling. Furthermore, this behavioural characterisation of Rxfp3 KO mice identifies them as a useful experimental model for studying RXFP3-linked signalling and assessing the selectivity and/or potential off-target actions of RXFP3 agonists and antagonists, which could lead to an improved understanding of dysfunctional arousal in mental health disorders, including depression, anxiety, insomnia and neurodegenerative diseases.

Introduction

Since its discovery in 2002 [1], [2] the neuropeptide, relaxin-3, has been identified as the ancestral member of the relaxin peptide family, and has been associated with several putative biological functions including stress response signalling, motivation for reward and orexinergic effects in rats (see e.g. [3], [4], [5], [6], [7], [8], [9] for review). Comprehensive data on the conserved anatomical distribution of GABAergic relaxin-3 neuron populations in the brainstem, in the nucleus incertus (see [10], [11], [12]), pontine raphe nucleus, medial/ventrolateral periaqueductal grey, and dorsal to substantia nigra [13], [14], [15]; and their projections in different species to a range of hypothalamic, limbic, cortical and septohippocampal circuits [13], [14], [15], [16], [17], has led to the hypothesis that these neurons constitute an ascending stress-related arousal system (see [5], [9]). In this regard, this system resembles other brainstem and hypothalamic arousal networks, including the noradrenaline/locus coeruleus [18], serotonin/dorsal and median raphe [19] and orexin/lateral hypothalamus [20] systems.

Relaxin-3 is thought to modify neuronal activity primarily via activation of the 7-transmembrane G_i/o-protein-coupled receptor - relaxin family peptide-3 receptor (RXFP3; also known as GPCR135) [6], [21], [22]. Several lines of evidence support this cognate peptide/receptor relationship: relaxin-3 binds to RXFP3 with high affinity (IC₅₀/EC₅₀ 0.5 nM) [21], [22], [23]; the genes encoding the peptide and receptor protein have phylogenetically co-evolved and are both highly conserved across species [21], [24]; there is a strong overlap between the distribution of RXFP3 mRNA/binding sites and relaxin-3 positive fibres within rodent brain [14], [15], [25]; and relaxin-3 is the only member of the relaxin peptide family that can activate RXFP3 [21] (see [6], [26] for review).

However, despite strong evidence that relaxin-3/RXFP3 forms a physiologically relevant cognate signalling pair, other data suggest some degree of ligand promiscuity under experimental conditions. For example, relaxin-3 has been shown in vitro to bind and activate receptors structurally related to RXFP3, namely RXFP1, albeit at a lower affinity and potency than at RXFP3. In addition, in vivo administration of exogenous relaxin-3 in the rat appears to activate RXFP1, as expression of the activity marker, Fos, is increased within RXFP1-rich circumventricular organs and related hypothalamic nuclei, and rats display a dipsogenic response characteristic of that following infusion of relaxin (designated relaxin-2 or H2 relaxin in humans), which is the cognate ligand for RXFP1 [6], [26], [27]. Several areas in rodent brain express RXFP3 and RXFP1 [3], [15], [25], [28], [29], and relaxin-3-positive axons and neuronal terminals are present within some RXFP1-rich regions [14], [15], [29]. Therefore, although the majority or all endogenous relaxin-3 signalling may be mediated via RXFP3 in rodents, currently the possibility that RXFP1 signalling contributes to brain relaxin-3 function cannot be dismissed.

The prior development of relaxin-3 null mutation (knockout, KO) mice, on a C57BL/6 background provided useful insights into the biological functions of relaxin-3 signalling [30], [31], [32], [33]. Examination of C57BL/6J relaxin-3 KO mice (backcrossed for >10 generations) revealed a chronic hypoactive voluntary running wheel phenotype compared to wild-type littermates [32], in line with the suggestion that relaxin-3 promotes behavioural activation. However, no genotype differences were observed under ‘basal conditions’ in acute behaviours relating to motor coordination, spatial memory, depressive-like behaviour and sensorimotor gating. In a second, independently generated C57BL/6N backcrossed colony, relaxin-3 KO mice displayed ‘slightly reduced’ anxiety-like behaviour in the elevated plus maze paradigm compared to WT littermates, which was reflected as an increased ratio of time spent in, and number of entries into, the aversive open arms [33].

The more recent development of an Rxfp3 KO mouse line (C57/B6J^{RXFP3TM1/DGen}) has provided the opportunity to assess the consequences of a loss of RXFP3-linked signalling on numerous behaviours, and to determine if the phenotypes previously observed in relaxin-3 KO mice are RXFP3 mediated or whether some of the traits of relaxin-3 KO mice may be due to other factors such as a flanking gene effect [34], [35]. In the current study, we observed that Rxfp3 KO mice display similar performance to WT littermates in acute behavioural paradigms designed to gauge motor coordination, spatial memory, depressive-like behaviour and sensorimotor gating. Notably however, Rxfp3 KO mice display hypoactivity on voluntary home-cage running wheels, and display a small, yet significant, decrease in some anxiety-like behaviours. This profile is near identical to that in two lines of relaxin-3 KO mice [32], [33], indicating the phenotype is due to the elimination of RXFP3 signalling. Furthermore, together with the ongoing development of pharmacologically active peptides [36], [37], [38] and viral peptide delivery technology directed towards RXFP3modulation [39], [40], this novel description of the behavioural phenotype of Rxfp3 KO mice provides an additional experimental approach to assess the selectivity and potential off-target actions of novel RXFP3-directed agonist and antagonist ligands.