Differential effects of early environmental enrichment on emotionality related behaviours in Huntington's disease transgenic mice


T. Renoir: Florey Neuroscience Institutes, Melbourne Brain Centre, University of Melbourne, VIC 3010, Melbourne, Australia. Email: thibault.renoir@florey.edu.au

Key points

  • • Clinical diagnosis of Huntington's disease (HD) is determined on the basis of motor symptoms; however, the pre-motor stages of the disease are commonly associated with psychiatric alterations including depression and anxiety.
  • • Using the R6/1 transgenic mouse model of HD, this study is the first report on the effects of environmental enrichment (EE) at a very early stage on a broad range of behavioural tests assessing stress-related measures.
  • • Environmental enrichment did not prevent despair- and anhedonia-like behaviours displayed by HD mice. However, EE reduced anxiety levels and corrected altered stress responses observed in HD mice.
  • • Despite the enhanced hypothermic response to the serotonin 5-HT1A receptor agonist 8-OH-DPAT exhibited by HD mice, we found a reduction in 5-HT1A receptor mediated stimulation of [35S]GTP-γ-S binding in the dorsal raphe nucleus and the hippocampus of HD animals.
  • • Our data suggest that early EE has beneficial effects on the anxiety-like, but not on the depression-like, behaviours in HD mice. We also provide evidence that 8-OH-DPAT induced hypothermia could be mediated by other targets besides the serotonin 5-HT1A receptors.

Abstract  Psychiatric disorders such as depression and anxiety are reported in patients with Huntington's disease (HD). Recent studies suggest beneficial effects of environmental enrichment (EE) on HD progression possibly through the serotonergic system. We investigated the potential effectiveness of EE in correcting the affective-like phenotype of female R6/1 HD mice. In addition to a behavioural battery of tests assessing depression and anxiety-related endophenotypes, we recorded physiological measures, including body temperature regulation and defecation rate as indices of stress reactivity. Finally, following identification of changes in serotonin (5-HT) receptor gene expression we measured the function of 5-HT1A auto- and hetero-receptors. We found that 8-week-old female HD mice exhibited higher immobility time in the forced swimming test and a decreased preference for saccharin solution. EE did not correct those depressive-like behaviours but reduced anxiety-related measures in unconditioned approach/avoidance conflict situations. Defecation rate in a large open field and change in temperature during exposure to the tail suspension test were both enhanced in HD compared to wild-type animals. Despite the enhanced hypothermic response to the 5-HT1A receptor agonist 8-OH-DPAT exhibited by HD mice, we found a reduction in 5-HT1A receptor-mediated stimulation of [35S]GTP-γ-S binding in the dorsal raphe nucleus and the hippocampus of HD animals. EE did not change 5-HT1A receptor function. Our data suggest that early EE has beneficial effects on the anxiety-like, but not on depression-like, behaviours in HD. This is the first evidence that these affective endophenotypes can be dissociated via this form of environmental stimulation. As 5-HT1A receptor dysfunction was not affected by EE, this receptor is unlikely to underlie the anxiety-related phenotype of HD. However, the specific regulatory role of the 5-HT1A autoreceptor in mediating depressive-like behaviour in HD remains to be elucidated. Interestingly, by comparing in vivo and in vitro results, our findings suggest that 8-OH-DPAT-induced hypothermia could be mediated by other targets besides the 5-HT1A autoreceptor, including hippocampal 5-HT7 receptors.


The level of complexity of the sensory stimuli present in an environment that an organism is exposed to can have profound effects on spontaneous behaviours, as well as on brain structure and function. Over the past five decades, the putative beneficial effects of sensory, cognitive and motor stimulation have been studied extensively in various animal models. Exposure to an enriched environment (EE) can indeed improve cognitive function in normal animals as well as in animal models of neurological disease (Nithianantharajah & Hannan, 2006). The majority of the pre-clinical studies in this field have assessed the effects of EE mainly on memory tasks focusing on neurodegenerative disorders and associated hippocampal neuroplasticity. However, recent studies have shown beneficial effects of enriched paradigms on behavioural abnormalities in rodent models of psychiatric disorders such as depression and anxiety (Laviola et al. 2008; Renoir et al. 2012a). For example using a prenatal stress rat model, studies reported an increased social interaction/play behaviour (Laviola et al. 2004), reduced fearfulness (Qian et al. 2008) and hastened normalization of stress hormone levels (Morley-Fletcher et al. 2003) following EE.

The beneficial effects of EE on any genetic animal model of a brain disorder was first discovered using the R6/1 HD mouse line to demonstrate an enrichment-driven delay in the onset of motor symptoms (van Dellen et al. 2000). A follow-up study then demonstrated an enrichment-mediated rescue of the cognitive deficits exhibited by this model (Nithianantharajah et al. 2008). HD is a fatal hereditary neurodegenerative disorder caused by an expanded CAG repeat in exon 1 of the huntingtin gene, which translates into an abnormally long polyglutamine tract in the huntingtin protein (The Huntington's Disease Collaborative Research Group, 1993). Clinical diagnosis of HD is determined on the basis of motor symptoms; however, the pre-motor stages of the disease are commonly associated with psychiatric alterations, including depression and anxiety (Paulsen et al. 2005; Duff et al. 2007; Julien et al. 2007; Marshall et al. 2007; van Duijn et al. 2008). Interestingly, apathy, irritability and depression are each associated with distinct longitudinal profiles (Thompson et al. 2012) and although antidepressant drugs have been shown to be effective in the treatment of depression in patients with HD (Holl et al. 2010) and depression-related behaviours in the R6/1 mouse model (Renoir et al. 2012c), non-pharmacological treatments devoid of any unpleasant side effects still need to be uncovered.

To date, only one study has assessed the effects of EE on depressive-like behaviours in HD animals (Pang et al. 2009). In addition, this was the first evidence that female (but not male) R6/1 HD mice displayed a depression-like phenotype. However in this later study, the authors used 12-week-old HD animals, an age with potential locomotor confounds (Hansson et al. 2001) and cognitive deficits (Nithianantharajah et al. 2008). Moreover, the potential effects of EE were assessed only on the forced swim test (FST), a widely used depression-related behavioural test sensitive to antidepressant drugs but with debatable reliability, particularly in terms of face validity (Nestler & Hyman, 2010). Taking these recent recommendations on board, we designed this present study to assess the effects of EE in 8-week-old female HD mice, broadening the scope of behavioural assays to include anhedonia (measuring animals’ interest in pleasurable activities such as preference for a saccharin solution) and measuring parameters of the physiological response to stress exposure. A dysregulation of the physiological stress response is commonly associated with the diagnosis of major depressive disorder, and along with early hypothalamic changes revealed by magnetic resonance imaging analysis (Soneson et al. 2010), the hypothalamic–pituitary–adrenal (HPA) axis, which controls the secretion of stress hormones have been reportedly altered in both patients with HD (Bjorkqvist et al. 2006; Aziz et al. 2009; van Duijn et al. 2010) and animal models (Petersen & Bjorkqvist, 2006; Du et al. 2012). Localized in the paraventricular nucleus of the hypothalamus, glucocorticoid receptors, which mediate negative feedback of the HPA axis through a tight interaction with the serotonergic system (Lanfumey et al. 2008), are also highly expressed in the hippocampus and dorsal raphe nucleus. Interestingly along with those brain regions, the hypothalamus is also known to be crucial for controlling temperature in a manner largely governed by serotonergic signalling (Morrison & Nakamura, 2011). The identification of the different serotonin (5-HT) receptor subtypes has facilitated our understanding of the contribution of 5-HT to regulate body temperature, and 5-HT1A receptor agonists such as 8-OH-DPAT are widely used to assess 5-HT function through its robust induction of hypothermia. Although the exact pathways mediating this effect remain to be elucidated, the relevance of the 8-OH-DPAT-induced hypothermia paradigm is furthermore highlighted by the role of the 5-HT1A receptor in the pathophysiology of affective-like disorders (Akimova et al. 2009; Renoir et al. 2012b) as well as the beneficial antidepressant/anxiolytic-like effects of several 5-HT1A receptor ligands (Caliendo et al. 2005; Ohno, 2010). Interestingly, there is a well-established role of the 5-HT1A receptor in mediating the beneficial effects of EE on hippocampal plasticity (Rasmuson et al. 1998).



R6/1 transgenic hemizygote males (Mangiarini et al. 1996) were originally obtained from the Jackson Laboratory (Bar Harbor, ME, USA) and bred with CBB6 (CBA × C57/B6) F1 females to establish the R6/1 colony at the Florey Neuroscience Institutes. After weaning, animals were group housed (four mice per cage with two of each genotype) and maintained on a 12 h light–dark cycle with access to food and water ad libitum. To avoid any possible ‘litter effects’, we ensured that animals from each litter were represented in each experimental condition using appropriate randomization. Unless otherwise specified, experiments were conducted on animals housed under standard conditions. EE mice were housed in larger cages from 4 to 8 weeks with various objects and shredded paper (Pang et al. 2009). They also were placed into large 120 litre boxes three times a week containing additional objects. All experiments were performed on female wild-type (WT) and R6/1 (HD) mice at 8 weeks of age and in accordance with the guidelines of the Florey Neuroscience Institutes Animal Ethics Committee (approval ID, AEC #08–100) and the National Health and Medical Research Council (NHMRC). Unless otherwise specified, different mice were tested in each behavioural paradigm.


[35S]GTP-γ-S was purchased from GE Healthcare Europe (Orsay, France). (±)-8-OH-DPAT was supplied by Sigma (Aldrich, NSW, Australia). Other compounds were WAY 100635 (Wyeth-Ayerst, Princeton, NJ, USA) and 5-carboxamido-tryptamine (5-CT; Res Biochem Int, Natick, MA, USA).

Body temperature

Core temperature was measured at ambient temperature of 23 ± 1°C in gently restrained mice using a thermocouple rectal probe (ID-Tech-Bioseb, France; 0.71 mm diameter).

Forced swimming test

A first cohort of animals (cohort 1, 63 mice) was used to assess depression-like behaviour using the FST. In this test, mice were individually placed into a glass beaker (13 cm diameter) filled with 12 cm deep water (25–26°C) and videorecorded for 300 s. Total immobility time was manually scored by an experienced experimenter blind to treatment and mouse genotype.

Saccharin preference test

A second cohort of animals (cohort 2, 48 mice) was used to assess anhedonia-like measures using the saccharin preference test. Based on a validated protocol (Harkin et al. 2002), single housed mice were exposed to both 0.1% saccharin and tap water solutions across four 15 h overnight periods (18.00 h to 09.00 h). To avoid place preference, the position of the two bottles was varied every day across the left or right side of the feeding compartment. To take into account time for the animals to become accustomed to the testing environment, the saccharin preference score (saccharin intake/total fluid intake) was calculated as the mean of the last two measures.

Open field test

A third cohort of animals (cohort 3, 29 mice) was used to assess exploratory behaviour in the open field. Following a minimum 1 h acclimatization period, the animal was removed from its home cage and then placed inside the centre of a square-shaped arena (1 m × 1 m). Its movements were then videotaped and later analysed by a tracking program (Ethovision software; Noldus Information Technology Inc., The Netherlands). The total duration of the test was 5 min.

Elevated plus maze

A fourth cohort of animals (cohort 4, 58 mice) was used to assess anxiety-like behaviour using the elevated plus maze (EPM). The EPM comprised of two open arms (36 × 5 cm) and two closed arms (36 × 5 × 18 cm) that extended from a central platform (5 × 5 cm) and was elevated 50 cm from the floor. The time spent in the closed and open arms was tracked using Ethovision software (Noldus Information Technology Inc.).

Novelty suppressed feeding test

Two days after EPM, cohort 4 was used in the novelty suppressed feeding test (NSFT). As previously described (Pang et al. 2009), mice were then food deprived for 48 h before testing but allowed 2 h of feeding after an initial 24 h period. Water was available ad libitum. During the test, a single food pellet was placed on a piece of filter paper in the centre of the test arena (80 × 80 × 80 cm). Individual mice were then placed into a corner of the arena and the latency to grasp and feed on the food pellet within 5 min was recorded. Upon initiation of feeding or reaching the time limit, mice were removed and allowed to feed on a single food pellet of predetermined weight for 5 min (to assess the level of hunger).

Tail suspension test-induced hyperthermia

A fifth cohort of animals (cohort 5, 24 mice) was used to assess physiological stress response. Mice were suspended by the tail over a 6 min session. Tail suspension test (TST)-induced hyperthermia has been shown to be sensitive to antidepressants and may be an index of stress reactivity (Liu et al. 2003). The change of temperature induced by TST was expressed as the change between pre-TST (baseline, t = 0 min) and after-TST (t = 6 min) temperatures. Core temperature was measured at ambient temperature of 23 ± 1°C in gently restrained mice using a thermocouple probe (ID-Tech-Bioseb, Chaville, France; 0.71 mm diameter).

8-OH-DPAT-induced hypothermia

A sixth cohort of animals (cohort 6, 47 mice) was used to assess 5-HT1A receptor function. As previously described (Renoir et al. 2011), basal temperature was determined just before subcutaneous injection of the 5-HT1A receptor agonist 8-OH-DPAT (0.3 mg kg−1, 1 ml 100 g−1 body weight) and body temperature was measured every 10 min thereafter. The response to 8-OH-DPAT was calculated as the decrease (from baseline) in body temperature during 60 min post-injection and expressed as area under curve over this period.

Quantitative autoradiography of 5-HT1A-mediated [35S]GTP-γ-S binding

Mice naive to behavioural testing were killed by cervical dislocation and their brains were frozen by immersion in isopentane chilled at −30°C with dry ice, then stored at −80°C. Coronal sections (20 μm thick) were cut at −20°C, and thaw mounted on to gelatin-coated slides. Autoradiographic measurement of 5-HT1A receptor-stimulated [35S]GTP-γ-S binding in the dorsal raphe nucleus and the CA1 area of the hippocampus was performed according to Froger et al. (2004). Briefly, brain sections were incubated with 0.05 nm[35S]GTP-γ-S (1000 Ci mmol−1) in the presence (stimulated conditions, Stim) of 5-CT, a non-selective 5-HT1 receptor agonist with nanomolar affinity for 5-HT1A receptors at four different concentrations (10−8, 10−7, 10−6, 10−5 m). Non-specific (NonSpe) binding was determined by adding the specific 5-HT1A receptor antagonist WAY 100635 (10 μm). Optical density was measured on Biomax MR autoradiographic films (Kodak, Rochester, NY, USA) using computerized image software (Samba). 5-CT-stimulated [35S]GTP-γ-S binding was expressed as percentage over the non-specific signal ([(Stim-NonSpe)/NonSpe] × 100).

Real-time PCR for quantification of mRNA expression

Mice naïve to behavioural testing were killed by cervical dislocation for dissection of the hippocampus and hypothalamus, which were snap frozen in liquid nitrogen and stored at −80°C until further use. Total RNA was isolated using Qiagen RNeasy RNA Mini kits (Qiagen, NSW, Australia) and stored at −80°C. Quality of the RNA and concentration of the final elutes were determined using an Agilant 2100 Bioanalyser (Australian Genome Research Facility, Parkville, VIC, Australia). cDNA was reverse transcribed from 1 μg of total RNA per sample using Reverse Transcription kits (PE Applied Biosystems, Foster City, CA, USA) with random hexamers. The reverse transcription PCR conditions were: 25°C, 10 min; 48°C, 30 min; and 95°C, 5 min. cDNA products were stored at −20°C. Gene expression was determined by real-time quantitative PCR using the 7500 Fast Real-time PCR system sequence detection software version 1.4 (Applied Biosystems, Foster City, CA, USA). All primer pairs were optimized for working concentrations before use and primer sequences are as follows. 5-HT1A receptor (5-HT1AR) F: CCC CAA CGA GTG CAC CAT, R: GCG CCG AAA GTG GAG TAG AT; 5-HT7 receptor (5-HT7R) F: GCG CCC CCG GAC AAT GTC TC, R: CTG GTG ATC CCA AGG TAC CTG TCG A; cyclophilin F: CCC ACC GTG TTC TTC GAC A, R: CCA GTG CTC AGA GCT CGA AA. Cyclophilin was used as the endogenous housekeeping gene as it is not altered in the R6/1 mouse line at this age (Zajac & Pang, unpublished data). Five to six samples per group were run in triplicate together with appropriate negative controls. Real-time PCR conditions: 50°C for 2 min, 96°C for 10 min and 40 cycles of 95°C for 15 s and 60°C for 1 min. A final dissociation stage consisted of 95°C for 15 s, 60°C for 20 s, 95°C for 15 s and 60°C for 15 s.

Statistical analysis

Statistical analyses were performed using SPSS statistics 17.0 and GraphPad Prism 5.0. Two-way analysis of variance (anova) was used to examine possible effects of genotype and/or housing condition. Repeated measures anovas were used to analyse 5-HT1A-mediated [35S]GTP-γ-S binding when considering the different concentrations of 5-CT used ([5-CT]) and the brain regions studied (raphe vs. hippocampus). To determine specific group differences in case of significant main effects (or interaction), anova was followed by Fisher's least significant difference or Bonferroni post-hoc tests. In all cases, the significance level was set at P < 0.05. For real-time PCR, statistical analyses were performed on relative fold changes determined from raw Ct values by the 2-ΔΔCt method (ABI User Bulletin #2). The mean fold changes of the various groups were normalized to the WT standard housed group.


Environmental enrichment does not prevent despair- and anhedonia-like behaviours displayed by Huntington's disease mice

FST is a widely used behavioural paradigm believed to reflect helplessness behaviour in rodents. It has been extensively used for the characterization of depressive-like behaviour in animal models as well as the assessment of potential antidepressant compounds. Two-way anovas of the immobility time in the FST (Fig. 1A) revealed a significant genotype effect (F1,59 = 20.4; P < 0.0001). There was no significant effect of housing condition (F1,59 = 1.3; > 0.05) or genotype × housing interaction (F1,59 = 0.8; > 0.05). Indeed, regardless of housing conditions, HD mice spent more time immobile than WT animals (P < 0.05). Similarly, there was a significant genotype effect (F1,44 = 8.3; P < 0.01) on saccharin preference suggestive of depressive-like behaviour in HD animals (Fig. 1B). There was no significant effect of housing condition (F1,44 = 2.6; > 0.05) or a significant genotype × housing interaction (F1,44 = 1.3; > 0.05).

Figure 1.

Effect of Huntington's disease (HD) mutation and environmental enrichment (EE) on despair- and anhedonia-like behaviours 
A, nalysing the immobility time (s) in the forced swimming tests, we found that HD mice spent more time immobile than wild-type (WT) animals. This depressive-like behaviour was observed in both standard housing (SH) and EE condition. B, we also revealed a genotype effect on saccharin preference (expressed as percentage of total fluid intake). Indeed regardless of housing conditions, HD animals exhibited reduced saccharin preference compared to WT, which suggested an anhedonia-like behaviour. WT vs. HD: *P < 0.05.

Exploratory behaviours in a novel environment are altered in Huntington's disease animals and modulated by environmental enrichment

There was an overall genotype difference (F1,25 = 6.1; P < 0.05) in exploration behaviour in large open field indicative of a reduction in HD animals (Fig. 2A). However, there was no effect of housing condition (F1,25 = 2.4; > 0.05) in total distance travelled. No significant effects of genotype or environment were revealed when analysing time spent in corners versus the centre arena (data not shown), although a significant effect of EE (F1,25 = 4.49; P < 0.05) on latency to leave the centre arena was observed (Fig. 2B). Finally, a two-way anova of the number of droppings during the 5 min test revealed significant overall effects of genotype (F1,25 = 7.2; P < 0.05) and housing condition (F1,25 = 13.5; P < 0.001) (Fig. 2C); however, there was no significant genotype × housing interaction (F1,25 = 0.03; > 0.05).

Figure 2.

Effect of Huntington's disease (HD) mutation and environmental enrichment (EE) on spontaneous behaviours in a large open field 
A, we found an overall significant effect of genotype on the total distance travelled (cm) during the 5 min session in the open-field arena. Housing conditions did not affect this measure. B, however, we observed a significant effect of EE on the latency to leave the centre arena. C, finally, a two-way anova analysis of the number of droppings during the open field test revealed a significant effect of environment and genotype but no interaction.

Environmental enrichment reduced anxiety levels based on response to unconditioned approach/avoidance conflicts

As a replication of a previous report, the performance in the NSFT was affected by the genotype (F1,54 = 6.9; P < 0.05) (Fig. 3A). Interestingly we also found a significant effect of environment (F1,54 = 4.7; P < 0.05). Indeed, the higher delay to feed in the NSFT observed in HD under standard housing condition was no longer observed following EE (Fig. 3A).

Figure 3.

Effect of Huntington's disease (HD) mutation and environmental enrichment (EE) on the response to unconditioned approach/avoidance conflicts 
A, we found a significant effect of genotype and housing conditions when analysing the time (expressed in seconds) to complete the task in the novelty suppressed feeding test (NSFT). Indeed, the higher delay to feed in the NSFT observed in HD under standard housing (SH) condition was prevented by EE. B, we found an overall effect of EE on the time spent in the open arms (s) in the elevated plus mazeEPM. WT vs. HD: *P < 0.05; ***P < 0.001. SH vs. EE: #P < 0.05.

There was no significant effect of genotype (F1,54 = 0.3; > 0.05) on time spent in the open arms of the EPM (Fig. 3B). However, a significant anxiolytic-like effect of housing was observed (F1,54 = 7.04; P < 0.05). The genotype × housing interaction was not significant (F1,54 = 0.4; > 0.05).

Environmental enrichment corrected altered stress response observed in Huntington's disease mice

We recently reported that, compared to WT animals, 8-week-old female R6/1 mice do not differ in terms of immobility times on TST (Renoir et al. 2011). Another output parameter of the TST is the stress-induced hyperthermia response which has been shown to be sensitive to antidepressant/anxiolytic drugs and may be regarded as an index of stress reactivity (Liu et al. 2003). Interestingly, the change in temperature induced by TST exposure (Fig. 4) was affected by the genotype (F1,19 = 6.7; P < 0.05). Furthermore, we found an antidepressant/anxiolytic-like effect of EE, regardless of genotype (F1,19 = 10.9; P < 0.01).

Figure 4.

Effect of Huntington's disease (HD) mutation and environmental enrichment on physiological response to acute stress 
Measuring the body temperature before and after a 6 min tail suspension test (TST), we found an overall effect of genotype and environmental enrichment (EE) on TST-induced hyperthermia. Indeed, the increased TST-induced hyperthermia (suggesting an altered stress response) exhibited by HD mice under standard housing (SH) conditions was prevented by EE.

Increased 8-OH-DPAT-induced hypothermia exhibited by Huntington's disease animals is not rescued by environmental enrichment

The acute administration of the 5-HT1A receptor agonist 8-OH-DPAT is known to decrease body temperature. In mice, this effect is believed to be mediated by somato-dendritic 5-HT1A autoreceptors (Bill et al. 1991). Using this paradigm we found a significant effect of genotype (F1,43 = 9.3; P < 0.01) when analysing the hypothermic response induced by DPAT over 1 h (Fig. 5). However, there was no effect of EE (F1,43 = 2.6; > 0.05) or any interaction (F1,43 = 0.13; > 0.05) on this physiological measure.

Figure 5.

Effect of Huntington's disease (HD) mutation and environmental enrichment on the 8-OH-DPAT-induced hypothermia 
Considering the area under curve (AUC) within the 1 h period following 8-OH-DPAT administration, we found an overall effect of genotype but no effect of environmental enrichment or any interaction.

Serotonin receptors in pre- and post-synaptic brain regions are not affected by environmental enrichment

Repeated measures anova revealed a significant effect of [5-CT] in the GTP-γ-S binding within (i) the raphe (F1,11 = 45.8; P < 0.001), and (ii) the hippocampus (F1,11 = 106; P < 0.001). Interestingly we also found a significant interaction between [5-CT] and genotype (raphe: F1,11 = 7.3; P < 0.05; hippocampus: F1,11 = 15.9; P < 0.01). However, in both brain regions there was no effect of EE regardless of genotype (Fig. 6A and B). Similar results were observed when comparing GTP-γ-S binding in raphe versus hippocampus at the higher concentration of 5-CT (Fig. 6C). Indeed, we found a significant effect of brain region (F1,11 = 169; P < 0.001) and genotype (F1,11 = 67.1; P < 0.001) but no effect of housing condition (F1,11 = 0.1; > 0.05).

Figure 6.

Effect of Huntington's disease (HD) mutation and environmental enrichment (EE) on quantitative autoradiography of 5-HT1A receptor-mediated [35S]GTP-γ-S binding 
For both brain regions studied (A) the dorsal raphe nucleus (DRN) and (B) the hippocampus, we found no effect of EE but a significant interaction between the concentration of the 5-HT1A receptor agonist [5-CT] and genotype when analysing 5-HT1A receptor-mediated [35S]GTP-γ-S binding [expressed as %Stim/NonSpe for labelling in the presence of 5-CT (Stim) over non-specific (NonSpe) signal obtained when adding the 5-HT1A receptor antagonist WAY100635]. C, similar results were observed when comparing GTP-γ-S binding in raphe versus hippocampus at the higher concentration of 5-CT (10 μm) suggesting that the reduced 5-HT1A auto- and heteroreceptor-related binding displayed by HD animals was not affected by EE. Wild-type (WT) vs. HD: ***P < 0.001.

We found a significant effect of genotype (F1,14 = 37.2; P < 0.001) but no effect of EE (F1,14 = 0.01; > 0.05) when analysing 5-HT1A receptor mRNA levels in the hippocampus (Fig. 7A). Indeed, compared to WT animals, hippocampal 5-HT1A receptor gene expression was reduced in HD mice regardless of housing condition. In contrast, the two-way ANOVA revealed an overall up-regulation of 5-HT7 receptor mRNA levels within the hippocampus of HD mice (F1,14 = 5.36; P < 0.05) with no significant effect of EE (F1,14 = 0.01; > 0.05) or any interaction (F1,14 = 1.44; > 0.05) (Fig. 7B). Finally, there was no effect of genotype or housing condition on the hypothalamic gene expression of 5-HT1A (Fig. 7C) and 5-HT7 (Fig. 7D) receptors.

Figure 7.

Effect of Huntington's disease (HD) mutation and environmental enrichment on 5-HT1A and 5-HT7 receptor gene expression in the hippocampus and hypothalamus 
Compared to wild-type (WT) animals, hippocampal 5-HT1A receptor gene expression was reduced in HD mice regardless of housing condition (A). We also found an overall significant effect of genotype when analysing 5-HT7 receptor mRNA levels in the hippocampus (B). However, there was no effect of genotype or housing condition on the hypothalamic gene expression of the 5-HT1A (C) and 5-HT7 (D) receptors. WT vs. HD: *P < 0.05,***P < 0.001. SH, standard housing.


Several studies have provided evidence for beneficial effects of EE by delaying the onset and progression of disease symptoms in rodent models of HD. However, this present report is the first to specifically assess the potential for EE as a therapeutic intervention for preventing the depression and anxiety-related behavioural changes. Indeed, so far, most of the work in this field (EE in R6 transgenic mouse models of HD) has focused on the mid-late stage symptoms such as locomotor impairments and cognitive deficits. Interestingly, EE has been shown to improve cognition (in the Morris water maze) in female, but not male R6/2 HD mice (Wood et al. 2010). Although no clinical data are available yet in terms of potential sex dimorphism in the development and treatment of mood disorders in patients with HD, our group recently reported that female (but not male) R6/1 HD mice displayed affective-like behaviours from 8 weeks of age (Renoir et al. 2011); this is an early time-point we consider as crucial within the scope of disease development. Indeed, huntingtin protein aggregates have been shown to first appear in the R6/1 striatum at 8 weeks of age in ≈15% of the cells and increase dramatically to ≈80% by 13 weeks of age, correlating with the development of motor deficits (Hansson et al. 2001). For all those reasons, we assessed the effects of early EE exposure (from 4 to 8 weeks of age) in R6/1 female HD mice.

We found that 8-week-old female HD mice recorded greater immobility time in the FST and were surprised that the presence of this depressive-like behavioural phenotype was not altered by EE, in contrast with the corrective nature of enrichment on 12-week-old female HD mice (Pang et al. 2009). It is important to note that we had adopted a similar length of enrichment exposure (4 weeks); therefore, the amount of enrichment stimulation provided did not differ and could not account for the difference in outcomes. One possible explanation for these contradictory results could be due to potential differences between 8- and 12-week-old HD animals when considering hippocampal neuroplasticity. While we previously found a reduction in cell proliferation in 12-week-old female HD mice (Renoir et al. 2012c), such measures have not yet been made at an earlier age and are worthy of further investigation. However, whether older mice with more mature neural connections could be more responsive to enrichment compared to maturing brains in adolescent mice has already been raised in a previous report of EE-induced changes in newborn granule neurons in both WT and HD mice (Lazic et al. 2006). Interestingly a recent study looking at the detailed effects of EE on adult neurogenesis over time (Llorens-Martin et al. 2010) suggested both an early selective, long-lasting effect of EE on the neurons born in the initial stages of enrichment, and a quick response when the environment again becomes impoverished (i.e. following a long-term 2 month period of EE). A potential direct relationship between the rate of adult hippocampal neurogenesis and immobility time in the FST has been suggested, as greater numbers of immature (but not mature) neurons in the dentate gyrus correlated positively with better FST performance (lower immobility times) (Llorens-Martin et al. 2007). Interestingly, age-dependent decline in cell proliferation and neurogenesis, usually associated with impaired cognitive function, have been characterized in animal models of other progressive neurodegenerative disorders (e.g. Alzheimer's disease) as well as in WT mice (Hamilton & Holscher, 2012; Kim et al. 2012). For instance, it was found that bromodeoxyuridine, a marker of proliferating cells, and doublecortin, a marker of immature neurons, steadily declined in female C57Bl/6 WT mice, with a 25% reduction in doublecortin-positive cells between 3- and 5-month-old animals (Hamilton & Holscher, 2012). Another study on male C57Bl/6 mice reported that the total number of neuroblasts generated from 3-month-old mice was reduced by more than half compared to that of 1-month-old animals (Kim et al. 2012).

Owing to recent concerns raised about the etiological relevance of FST for the study of clinical depression (Nestler & Hyman, 2010), we broadened the variety of behavioural assays employed. Anhedonia, defined as the loss of ability to enjoy pleasure (Gorwood, 2008), is a key feature of clinical depression and is routinely modelled in rodents by taking advantage of their natural preference for sweet solutions. We found that HD mice displayed anhedonic tendencies with reduced preference for saccharin-sweetened solution, and this was not altered by EE. It is possible that a mere 4 weeks of exposure to enriched environment before any overt cellular pathology occurring limits the ability to detect any preventative effects of enrichment on the behavioural changes. Therefore, future assessments of longer-term EE on anhedonic behaviour would be required to test this hypothesis. However, it would be too early to discount the possibility that despair behaviour and anhedonic tendencies are impacted on differently by enrichment. For instance, female mice reared in a communal nest were recently found to exhibit a higher sucrose preference although spending longer time immobile in the FST when compared to mice reared in standard laboratory conditions (D’Andrea et al. 2010).

Our observed anxiolytic effects of enrichment housing in EPM and open field tests were consistent with previous findings based on a longer-term 2 month EE protocol (Nithianantharajah et al. 2008). Notably, this anxiolytic-like effect of EE was observed regardless of the genotype suggesting that in contrast to what has been recently reported in aged mice with cognitive decline and sensorimotor deficits (Bouet et al. 2011), 8-week-old HD animals remain responsive to the beneficial effect of EE in terms of anxiety-related behaviours. Interestingly, previous studies have shown that mice value space and this might be modulated by environmental factors. In that regard, the latency to leave the starting spot of the open field can been interpreted as an indicator of stress and that was found to be reduced by EE for both HD and WT mice suggesting an enhanced desire to explore after EE. Defecation under stress is also considered as a valid index of emotionality in rodent models of depression (Kinn et al. 2008), sometimes in a female-specific manner (Renard et al. 2007). Interestingly, we found higher numbers of faecal boli excreted by female HD mice compared to WT animals during open field testing.

EE reduces the number of droppings in all groups, suggesting an anxiety-like effect of EE regardless of the genotype. Indeed, high incidence of defecation in the open field test usually translates to enhanced fear-related emotion. For example, normal control animals show significant decreases in defecation over time when exposed to the open field test throughout several days, while socially defeated rats continued to exhibit high defecation scores during each exposure to the test (Kinn et al. 2008). Other studies report increased defecation in open field after exposure to chronic stress as an index of high emotionality and anxiety-like behaviours (Renard et al. 2007; Zhu et al. 2011). Although hardly comparable to defecation behaviours of rodents during exposure to the open field test, data from patients with irritable bowel syndrome and major depression suggest an increased gut permeability-induced HPA axis alteration in these syndromes (Dinan & Cryan, 2012; Suarez-Hitz et al. 2012). Furthermore, there is evidence that shared peripheral and/or central inflammatory and oxidative stress pathways may underpin the pathophysiology of depression and other co-morbidities such as HD and leaky gut (Maes et al. 2011). Interestingly, gastrointestinal dysfunction has recently been found in HD mice (van der Burg et al. 2011) and transient gastric irritation was shown to induce long-lasting depression- and anxiety-like behaviours in rats (Liu et al. 2011). Determining, whether brain–gut axis dysfunction is involved in impaired stress–response in HD mice would require further investigations.

We still have a poor understanding of mechanisms underlying EE-induced changes on affective-related behaviours and our present study failed to show any potential involvement of 5-HT1A or 5-HT7 receptors in that regard (see discussion about those specific 5-HT receptors below). Neuroplasticity-related mechanisms have been the focus of most previous studies in that field and remain first-line options to explain behavioural effects of EE.

Among others, brain-derived neurotrophic factor (BDNF) is hypothesized to be critically involved in the regulation of age-related processes in the hippocampus (von Bohlen und Halbach, 2010) as well as in the development of depression and mechanisms of action of antidepressant therapeutics (Martinowich et al. 2007; Chourbaji et al. 2011; Wolkowitz et al. 2011). Therefore the temporal dynamics of BDNF dysregulation could help explain the age-dependent effect of EE that our present findings suggest when compared to previous studies performed at 12 weeks of age (Pang et al. 2009). Indeed, a progressive reduction of BDNF mRNA levels has been described in HD models. For example, while no difference of BDNF exon III mRNA level was observed in R6/2 HD animals before 8 weeks of age, a 50% reduction was found at 12 weeks in HD mice compared to controls (Zuccato et al. 2005). The authors also showed that the progressive loss of mRNAs transcribed from BDNF exon II, III and IV followed a different pattern. Interestingly, we recently reported differential changes in exon-specific expression in the hippocampus of R6/1 HD mice exposed to environmental paradigms such as wheel running or EE (Zajac et al. 2010). For example, BDNF exon III mRNA levels were up-regulated in both female WT and HD animals by running while EE induced a down-regulation in WT mice only. Epigenetic regulation of the BDNF gene seems also to be differentially affected by both the HD mutation and environmental condition. Interestingly, epigenetic modifications play an important role in long-lasting gene expression changes induced by environmental factors and are believed to be involved in the pathophysiology of psychiatric disorders (Boulle et al. 2012; Sun et al. 2012). Antidepressant treatments have been shown to modulate epigenetic mechanisms (such as histone modifications and DNA methylation) in animal models (Sales et al. 2011; Melas et al. 2012). Interestingly, distinct epigenetic mechanisms seem to be involved in stress coping capabilities (Collins et al. 2009). Along with regulation of BDNF gene expression, recruitment of the N-methyl-d-aspartate receptor and extracellular signal-regulated kinases ERK signalling pathway has been found to be activated during the FST (Chandramohan et al. 2008). Potential associations between epigenetic regulation of the glucocorticoid receptors and mood disorders have been reported (McGowan et al. 2009; Alt et al. 2010). Since we recently found that EE rescued HPA axis dysregulation of 12-week-old female HD mice (Du et al. 2012), further exploration of stress-related pathways at an earlier stage of the HD process would be worthwhile. In that regard, we confirm here that the acute stress response was altered in HD mice from 8 weeks of age. Indeed, we found an exaggerated TST-induced hyperthermic response in HD compared to WT animals. This physiological response has previously been shown to be sensitive to anxiolytic drugs (Liu et al. 2003) and we provide further evidence of the anxiolytic properties of enrichment by reporting that the hypothermic responses of HD and WT mice were minimized. It would be very interesting to explore the corrective properties of EE on depression and anxiety-related behaviours that have been described in alternative mouse models of HD (Orvoen et al. 2012).

Numerous studies have demonstrated a role of the 5-HT1A receptor in the pathophysiology of affective-like disorders (Akimova et al. 2009; Renoir et al. 2012b). This is further evidenced by the beneficial antidepressant/anxiolytic-like effects of several 5-HT1A receptor ligands (Caliendo et al. 2005; Ohno, 2010). An initial report suggested that the EE-induced hippocampal plasticity may be mediated, in part, through 5-HT1A receptor subtypes (Rasmuson et al. 1998). Interestingly, an increased corticosterone response to buspirone has been observed in rats kept under enriched environments (Moncek et al. 2004). Because 5-HT1A receptor mRNA levels are altered in the brains of HD mice with partial rescue by EE (Pang et al. 2009), we initially expected to observe some modification of the hypothermic response to 8-OH-DPAT administration after EE. However, the unanticipated non-effect of enrichment could be due to a more complex nature of this physiological response, which was thought to exclusively involve 5-HT1A autoreceptors (Bill et al. 1991) but is likely to involve other post-synaptic 5-HT receptors (Bert et al. 2006). Further research built upon this initial finding will be required to better understand the modification of specific 5-HT receptors by EE in relation to thermoregulation. Continuing our investigation of the 5-HT1A receptor, we attempted to distinguish between pre- and post-synaptic receptors with the [35S]GTP-γ-S binding assay (Hensler & Durgam, 2001). For practical reasons, such measures were feasible in both raphe and hippocampus but not in hypothalamus (mainly because of the multiplicity of heterogeneous small nuclei within this brain region). Using this method we found a reduction in the 5-HT1A receptor-stimulated binding in the hippocampus of 8-week-old HD mice, which is the earliest age at which this change has been observed as previous reports of this were based on studies of post-synaptic areas of older HD animals (Pang et al. 2009; Renoir et al. 2012c). A similar decrease of 5-HT1A autoreceptor function was also found in the HD raphe region. It was interesting to observe that [35S]GTP-γ-S binding induced by the high concentrations of 5-HT1 receptor agonist 5-CT in the raphe was approximately three times less than the signal obtained in the hippocampus. Such regional differences in coupling efficiency are in contrast with other reports and may be due to several factors, including differences between in vivo versus in vitro receptor activation (Meller et al. 2000). More importantly, the reduced 5-HT1A receptor function found in the raphe of HD mice appears to contradict the enhanced 8-OH-DPAT-induced hypothermia observed in this same mouse model (see Fig. 5 of the present study and Renoir et al. 2011). In addition, we previously reported no change of 5 HT1A receptor gene expression within the raphe of 8-week-old female HD mice (Renoir et al. 2011). However, as discussed above the 8-OH-DPAT-induced hypothermia is likely to involve other signalling mechanism besides the 5-HT1A autoreceptor of the raphe.

Indeed, more recent studies using selective drugs and knockout mice have suggested some interaction between the 5-HT1A and 5-HT7 receptors in mediating thermoregulation (Hedlund et al. 2004) and 5-HT7 receptors have been implicated in neuropsychiatric disorders (Hedlund, 2009). Based on our previous findings and mRNA distribution profile of 5-HT1A/5-HT7 receptors, we then assessed the effect of the HD mutation and housing conditions on hypothalamic and hippocampal gene expression. We report no effect of genotype or EE on the hypothalamic gene expression of the 5-HT1A and 5-HT7 receptors. In contrast and reflecting regional differences, 5-HT1A receptor mRNA levels were found down-regulated in the hippocampus of HD animals, regardless of housing conditions. Interestingly, hippocampal 5-HT7 receptor gene expression was slightly increased in HD compared to WT mice. However, whether such small (although statistically significant) molecular changes could translate to the physiological differences exhibited by HD animal such as the altered temperature regulation, is not clear and requires further investigation.

Overall, we have identified novel depression and anxiety-related behavioural measures in the R6/1 mouse model of HD, evident as early as from 8 weeks of age. Interestingly, our findings suggest that some of these affective endophenotypes (i.e. anxiety-like but not depression-like behaviours) can be modulated by environmental stimulation. Indeed, our results indicate differential gene–environment interactions in HD when considering depression- versus anxiety-like behaviours. Surprisingly, environmental intervention at a very early stage (immediately after weaning in our current study) of the disease could correct anxiety-related disorders whereas depression-like behaviours seemed to respond to EE only when applied at a later stage (8-week-old mice are usually considered as ‘young adult’). It is difficult to transpose in the clinic a similar equivalent time-window and controlled environmental conditions used in our study. However, a retrospective study of the impact of lifestyle of 154 adults diagnosed with HD, suggested that a passive lifestyle (which could be translated as the ‘poor’ condition used for the standard housed animals compared to environmentally enriched mice) may contribute to the earlier onset of HD symptoms (Trembath et al. 2010). Further clinical investigations looking at the potentially time-dependent effects of ‘stimulating environments’ (of any kind and by any means) would be interesting in terms of potential therapeutic implications.

We have also found that both pharmacologically and stress-induced thermoregulatory responses were altered in HD animals, which is likely due to dysregulations within the raphe–hippocampus signalling in a complex mechanism involving the 5-HT1A autoreceptor and 5-HT7 post-synaptic receptor. However, investigating the potential role of other targets such as 5-HT2A receptors could also be informative as these are known to be involved in the modulation of hyperthermia, corticosterone and hippocampal 5-HT release following stress exposure (Doyle & Yamamoto, 2010). As 5-HT is functionally integrated in the neuroanatomical and neurotransmitter model of thermoregulation (Morrison & Nakamura, 2011), studies looking at the role of huntingtin protein on fundamental neural pathways such as thermogenic neurons in the dorsomedial hypothalamus and hippocampus, as well as the sympathetic and somatic pre-motor neurons in the rostral medullary raphe region, would further inform mechanisms of pathogenesis.

Finally, although only female R6/1 HD mice displayed depressive-like behaviours when assessed in the tests based on acute stress response, affective-like disorders are known to also occur in male patients. Unfortunately, no clinical data are available yet in terms of potential sex dimorphism in the development and treatment of mood disorders in patients with HD, including those who have not yet developed motor symptoms. Ultimately, it would be of interest to compare the incidence of mood disorders in both sexes, as well as the effects of specific environmental and pharmacological interventions.

Translational perspective

It has been decades since the gene mutation causing Huntington's disease (HD) was discovered, but a cure has so far eluded us. Therefore, recent research has focused on the management of symptoms occurring during the progression of HD. Part of the management plan includes having a clear understanding of the disease pathophysiology. To that end, the R6/1 transgenic mouse model of HD has proved to be useful and, among other features, exhibits a depression-related behavioural phenotype associated with serotonergic impairments. Notably, those behavioural alterations were observed in pre-motor symptomatic animals, consistent with the higher incidence of depression in patients with prodromal HD. This present study has extended our knowledge of the pre-symptomatic stage by uncovering altered physiological responses to acute stress, which may indicate a maladaptation of stress coping mechanisms attributable to the HD mutation. These abnormal stress responses might be a physiological contributor to the development of affective disorders in patients with HD. Therefore, our findings provide the rationale to explore this further in the clinical population through physiological parameters known to be altered by stress regulation such as heart rate, cardiac output and body temperature. In addition, our findings suggest that enhanced cognitive stimulation and physical activity through ‘environmental enrichment’ improve stress responses and reduce anxiety levels. Collectively, our data indicates that environmental factors, including levels of stress, cognitive stimulation and physical exercise, are important for disease onset and progression, and should be important considerations for future clinical studies.


Author contributions

All work was carried out at the Division of Behavioural Neuroscience, Florey Neuroscience Institutes, University of Melbourne. Conception and design of experiments: T.R., T.Y.P., L.L. Collection, analysis and interpretation of data: T.R., T.Y.P., C.M., G.C., C.C., L.L. Drafting the article or revising it critically for important intellectual content: T.R., T.Y.P., L.L., A.J.H.


This work was supported by NHMRC-INSERM Exchange Fellowship (T.R.), Project Grant funding (A.J.H.) from the NHMRC and an ARC Future Fellowship (A.J.H.). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. The authors do not have any conflicts of interests to disclose.