Introduction
Obsessive–compulsive disorder (OCD) is a mental disorder mainly characterized by obsessive and compulsive behaviors. The prevalence of OCD is ~ 2.3% in the population [
1], and most of the patients develop symptoms before 35 years old [
2]. Selective serotonin reuptake inhibitors comprise the first-line treatment of OCD patients, but approximately half of the OCD patients fail to fully respond to this treatment [
3]. Therefore, it is urgent to elucidate the mechanisms and causes of OCD.
OCD symptoms include obsessive (e.g., fear of contamination, the need to order things symmetrically, and aggressive, sexual, or religious thoughts) and compulsive (e.g., excessive washing, checking, ordering, counting, and repeating) traits [
4]. Besides, OCD is a highly heterogeneous disease, with which many patients experience anxiety and cognitive deficits additionally. The symptoms vary widely among patients [
5], and different dimensions of OCD symptoms may be caused by distinct neurobiological mechanisms [
6]. Previous studies proposed that the dysfunction of parallel, partly segregated cortico–striato-thalamo-cortical (CSTC) loops, including serotoninergic, dopaminergic, and glutamatergic systems, are related to different cognitive-affective processes in OCD [
4,
7‐
9]. For example, Mataix-Cols et al. used functional magnetic resonance imaging (fMRI) to measure neurological activity in patients with different symptom dimensions of OCD. They found that the activities of the bilateral ventromedial prefrontal regions and the right caudate nucleus were activated in patients with washing symptoms than in the control population.
Moreover, checking symptoms were accompanied by increased activity of the putamen/globus pallidus, thalamus, and dorsal cortical areas, whereas the activities of the left precentral gyrus and right orbitofrontal cortex in patients with hoarding symptoms were increased [
6]. Therefore, each symptom may be mediated by relatively distinct brain regions or circuits. Targeted treatment on relevant brain regions may enable us to develop precise treatment, thereby improving treatment effectiveness in OCD.
Animal models are widely used to explore the physiological and pathological characteristics of OCD. However, the behavioral phenotypes of the OCD animal model were quite different among different models. Additionally, CSTC circuit anomalies were only detected in some models [
10,
11]. The studies that systematically compare the behavioral traits and abnormal brain circuits between different animal models are entirely lacking.
In this study, we systematically examined the behavioral phenotypes of three OCD mouse models induced by pharmacological reagents (RU24969, 8-OH-DPAT, and MCPP), and compared the activated brain regions in each model, respectively. Our results showed that different OCD mouse models exhibited distinct behavioral traits, which may be mediated by the activation of different brain regions.
Discussion
Pharmacological animal models of OCD have been widely used in previous studies [
10,
11]. However, most of the published studies used rats instead of mice. In this report, three mouse models of OCD were established to provide more options for drug-induced OCD models and offered more possibilities for researchers to explore the mechanisms of OCD.
RU24969, 8-OH-DPAT, and MCPP all belong to serotonin receptor agonists. The choice of these reagents was made based on previous clinical experience with OCD medications and hypothesized involvement of the serotonin system in OCD. They all augment the sensitivity of 5-HT receptor subtypes, which caused OCD [
13].
The dysfunction of 5-HT receptor 1B (5-HT1B) in OCD has been suggested in previous studies. For example, in pharmacological studies, the use of 5-HT1B agonists aggravated symptoms in patients with OCD, and mutations of the serotonin transporter
SCL6A4 were associated with OCD [
14]. As serotonin 5-HT1A/1B receptor agonist, RU24969 treatment in rats induced locomotor stereotypy, prepulse inhibition (PPI) deficits, and impairments in delayed alternation, all related to common signs and symptoms in OCD patients [
12]. Consistently, in the present study, RU24969 administration induced repetitive circling behavior in mice (Additional file
2: Fig. S2). Moreover, the mice exhibited anxiety in the OFT. Impaired memory is one of the cognitive dysfunctions of OCD [
32]. In the NOR, RU24969-treated mice had a decreased time to explore novel objects indicative of memory impairment, further confirming the validity of this model.
8-OH-DPAT, a 5-HT receptor 1A (5-HT1A) agonist, could bind to its receptor with high selectivity [
33]. 8-OH-DPAT-treated animals were commonly used as an OCD model. In the spontaneous alternation behavior test, the animals repeatedly selected the same arm, similar to the OCD perseveration symptoms [
34]. Importantly, treatment with selective serotonin reuptake inhibitors could eliminate this repetitive behavior in the 8-OH-DPAT rat model [
35]. However, other OCD-like behavioral tests have been rarely tested in the rat model. In the present study, 8-OH-DPAT administration induced spray-induced grooming in mice. Interestingly, self-grooming behaviors were not increased in 8-OH-DPAT mice compared to control. This may be due to the necessity of a trigger to induce OCD-like behavior in this model, similar to patients with excessive washing or cleaning behavior whose symptoms were triggered by contamination stimuli in clinical [
4]. Thus, the OCD-like behaviors in 8-OH-DPAT-treated mice differed from those in RU24969-treated mice. On the other side, similar to RU24969-treated mice, 8-OH-DPAT-treated mice showed recognition memory impairments in the NOR, whereas anxiety was not detected in the OFT, which might be related to a lack of related stimuli in these experiments.
Administration of MCPP, which binds to 5-HT1A and 5-HT2C receptors, could aggravate symptoms in patients with OCD [
13,
36]. In a preclinical study, the administration of MCPP blocked the beneficial effects of fluoxetine for OCD treatment. Moreover, MCPP could induce the occurrence of repetitive stereotypes such as increasing the number of buried marbles in the MBT [
13], promoting directional persistence in spontaneous alternation behavior [
11], and reinforcing self-grooming [
16]. In the current study, MCPP-induced OCD-like behaviors comprised excessive self-grooming behavior, whereas the induced-grooming test showed no significant differences to control mice. These results contrasted with those in 8-OH-DPAT mice. Self-grooming and spray-induced grooming are different grooming forms [
25]. Self-grooming is spontaneous; thus, excessive self-grooming in MCPP mice resembles trichotillomania (hair-pulling disorder) appearing spontaneously or under high pressure. Besides, the MCPP model presented no anxiety or memory impairment according to the OFT and NOR. Most patients with OCD chose to obey obsessive–compulsive thoughts and perform obsessive–compulsive behaviors to alleviate the anxiety caused by obsessive–compulsive thoughts and impulses. Thus, the reduction of anxiety in MCPP-treated mice may be due to the performance of obsessive–compulsive behavior.
Together, the three newly established OCD mouse models exhibited distinct OCD-like behavioral traits and presented different levels of anxiety and memory dysfunction.
The expression of c-fos is one of the commonly used indicators to measure the activity of neurons because repetitive action potentials were often accompanied by neuronal c-fos expression [
37]]. Based on the drugs’ pharmacological properties and the expression of the receptors, we speculate that the increase in c-fos expression was majorly raised from neurons. All three drugs are 5-HT receptor agonists and can act on almost all neuronal cells. The distribution of their corresponding receptors is wide. 5-HT 1A receptors were expressed in the pyramidal neurons of the cortex, hippocampus, and raphe nuclei and the cholinergic neurons in the septum [
38‐
40]. 5-HT 1B receptors were also expressed in the medium spiny neurons in the caudate putamen, probably GABAergic [
41,
42]. 5-HT 2C receptors were expressed in most GABAergic cells. Astrocyte was also reported to express 5-HT receptors; thus, the astrocyte could also be a source for the c-fos signal. Previous studies proposed that the dysfunction of CSTC loops is related to different cognitive-affective processes in OCD. The brain regions chosen in this study belong to the CSTC loop (except for BSTLD and IPACL). Here, neuronal activity was increased in the OFC, PrL, IL, ACC, CPu, NAc, BSTLD, IPACL, and hypothalamus of RU24969-treated mice. In 8-OH-DPAT-treated mice, activated brain regions included the ACC, PrL, IL, OFC, and AcbSh (Additional file
3: Table S1). This was similar to previous studies in mouse and rats [
33,
43‐
45] that 5-HT1B receptors are distributed in the striatum, cerebellum, and basal ganglia, 5-HT1A receptors were distributed in the neocortex, olfactory areas, hippocampal formation, cortical subplate, pallidum, hypothalamus, and mesencephalic raphe nuclei, although the brain regions with increased neuronal activity in these two model mice did not completely coincide with those expressing 5-HT1A or 5-HT1B receptors. One possible reason is that the OCD-like behaviors were not induced by directly activating all relevant brain regions containing 5-HT1A or 5-HT1B receptors, but acting within the same receptor-associated brain region instead, therefore indirectly increasing the neuronal activity of downstream brain regions and causing the occurrence of OCD-like behaviors. Notably, most of the activated brain regions were related to the CSTC loop elements corresponding to the postulated CSTC loop dysfunction in OCD.
In mice and rats, the 5-HT receptor 2C (5-HT2C) is mainly distributed in the choroid plexus and other areas, including the nucleus accumbens, patches of the caudate-putamen, the olfactory tubercle, claustrum, septum, cingulate cortex, amygdala, dentate gyrus, periaqueductal gray, and entorhinal cortex [
37,
44,
46]. Interestingly, after injection with the 5-HT2C agonist MCPP, we observed a weak c-fos expression in the mouse brains, with no difference to the control group. Even after increasing the dose, we obtained the same results (data not shown). In this MCPP model, mice only showed excessive self-grooming behavior, whereas all other behaviors appeared normal. To a certain extent, this was consistent between behavioral and c-fos expression. We speculate that MCPP injection might cause inhibition in some brain regions, which could not be reflected by c-fos elevation. Indeed, previous studies found that OCD patients have decreased activity in certain brain regions [
47,
48].
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.