The α2δ subunit of voltage-sensitive calcium channels (VSCCs) is the molecular target of pregabalin and gabapentin, two drugs marked for the treatment of focal epilepsy, neuropathic pain, and anxiety disorders. Expression of the α2δ subunit is up-regulated in the dorsal horns of the spinal cord in models of neuropathic pain, suggesting that plastic changes in the α2δ subunit are associated with pathological states. Here, we examined the expression of the α2δ-1 subunit in the amygdala, hippocampus, and frontal cortex in the trimethyltiazoline (TMT) mouse model of innate anxiety. TMT is a volatile molecule present in the feces of the rodent predator, red fox. Mice that show a high defensive behavior during TMT exposure developed anxiety-like behavior in the following 72 h, as shown by the light–dark test. Anxiety was associated with an increased expression of the α2δ-1 subunit of VSCCs in the amygdaloid complex at all times following TMT exposure (4, 24, and 72 h). No changes in the α2δ-1 protein levels were seen in the hippocampus and frontal cortex of mice exposed to TMT. Pregabalin (30 mg/kg, i.p.) reduced anxiety-like behavior in TMT-exposed mice, but not in control mice. These data offer the first demonstration that the α2δ-1 subunit of VSCCs undergoes plastic changes in a model of innate anxiety, and supports the use of pregabalin as a disease-dependent drug in the treatment of anxiety disorders.
Voltage-sensitive calcium channels (VSCCs) are formed by the assembly of different subunits, including the pore-forming α1 subunit (which defines the specific type of VSCC), acytosolic β subunit, a 4-TM-domain γ subunit, and an α2δ subunit, in which an extracellular α2 subunit is disulfide bonded to a transmembrane δ subunit (Catterall 2000; Arikkath and Campbell 2003; Davies et al. 2007, 2010). There are four isoforms of the α2δ subunit encoded by four separate genes with several known splice variants. The α2δ-1 subunit is expressed in the forebrain, spinal cord, skeletal muscle, and cardiac muscle; the α2δ-2 subunit is found in the forebrain and cerebellum; the α2δ-3 subunit is exclusively expressed in the brain; and the α2δ-4 subunit is found in the brain, pituitary, adrenal glands, and intestine (Dooley et al. 2007; Taylor and Garrido 2008).
If coexpressed with the α1 subunit, the α2δ subunit increases the amplitude of calcium channel currents and influences the rate of channel activation and inactivation (De Waard and Campbell 1995; Birnbaumer et al. 1998; Jones et al. 1998; Klugbauer et al. 1999; Wakamori et al. 1999; Cantí et al. 2005). The amplifying effect of calcium current has been ascribed to an increase trafficking of the VSCC to the plasma membrane (Cantí et al. 2005; Bernstein and Jones 2007). Interestingly, neuronal α2δ-1 subunit binds to the extracellular matrix proteins trombospondins, thereby promoting the formation of excitatory synapses (Eroglu et al. 2009). The interest for the α2δ subunit of VSCCs has grown considerably since the discovery that the α2δ-1 and α2δ-2 subunits are the pharmacological targets of gabapentin and pregabalin (Gee et al. 1996; Marais et al. 2001; Quin et al. 2002; Field et al. 2006; Dooley et al. 2007), two anti-epileptic drugs that are now gold standard in the treatment of neuropathic pain (Attal et al. 2010).Gabapentin and pregabalin are also effective in the treatment of generalized anxiety disorder, bipolar disorders, and fibromyalgia (reviewed by Yatham 2004; Keck et al. 2006; Bandelow et al. 2007; Owen 2007; Siler et al. 2011; Tzellos et al. 2010; Samuel et al. 2011). Studies carried out in spinal cord slices suggest that gabapentin and pregabalin behave like ‘disease-dependent’ drugs, i.e. they reduce pain transmission only under pathological conditions (Patel et al. 2000; Fehrenbacher et al. 2003). During neuropathic pain, the levels of the α2δ-1 subunit are up-regulated in the dorsal horns of the spinal cord, and pregabalin prevents the increased transport of the α2δ-1 subunit from dorsal root ganglia neurons to the pre-synaptic terminals of primary afferent fibers in the spinal cord (Bauer et al. 2009). So far, expression of the α2δ subunit has only been studied in the dorsal root ganglia/spinal cord under conditions of chronic pain, and how the protein behaves in models of anxiety is unknown.
Here, we examined the expression of the α2δ-1 subunit in mice exposed to trimethyltiazoline (TMT), a highly volatile sulfur-containing molecule that is found in the feces of the red fox, a natural predator of rodents (Vernet-Maury et al. 1984). Exposure to predator odor in rodents represents a model for innate anxiety and animal phobias with putative clinical relevance (reviewed by Kavaliers and Choleris 2001; Rosen et al. 2008).
Materials and methods
2,5-Dihydro-2,4,5-TMT was purchased from Phero-Tech (Delta, BC, Canada). Pregabalin [(S)-3-aminomethyl)-5-methylhexanoic acid] was purchased from Tocris Bioscience (Bristol, UK).
Experimental design and TMT exposure
Experiments were carried out according to the European (86/609/EEC) and Italia (D. Lgs 116/92) guidelines of animal care. All efforts were made to minimize animal suffering and the number of animals used. The experimental protocol was approved by the Italian Ministry of Health according to the procedure indicated in the D. Lgs. 116/92. Male CD1 adult mice (25–30 g, b.w.; Harlan Laboratories, Udine, Italy) were housed with free access to food and water and maintained on a 12-h light\dark cycle (lights off at 20:00). One week before behavioral testing, mice were individually handled daily. Handled mice were lifted by the base of the tail and permitted free exploration of a gloved hand for approximately 1 min. Three days before behavioral evaluation, all mice were exposed to the experimental protocol: on each day, they were leave free to explore a plexiglass cage (50 × 50 × 50 cm) containing an odor-free Petri dish for 10 min. After that, they were placed back in their home cage. After habituation, mice were randomly assigned to TMT or saline. Experiments were performed from 10:00 a.m. to 1 p.m. TMT (40 μL), the major component of the anal gland secretions of the red fox, was impregnated on a gauze in a Petri dish that was positioned in the open field. Mice were exposed individually for 10 min to TMT or saline and videotaped for behavioral analysis. Contact frequency with the odorant pad, defensive burying (movement toward the odorant cloth while pushing or spraying bedding material towards the cloth), freezing (immobility except for movement of breathing muscles), and grooming frequencies were scored. As expected, mice were not homogenous in displaying defensive behavior during TMT exposure. Only mice showing a high defensive response to TMT (i.e. mice displaying > 1-min of freezing during the 10-min exposure; Hebb et al. 2004) were recruited for biochemical analysis and assessment of anxiety-like behavior in the light–dark box. Of 72 mice exposed to TMT, 22 were ‘high responders’. ‘High responders’ to TMT were divided into three groups (n = 6–8); control mice exposed to odor-free gauge (n = 20) were also divided into three groups (n = 6–7). Individual groups of ‘high responders’ and their controls were tested for anxiety-like behaviors in the light–dark box after 4, 24, or 72 h, and killed 20 min after the end of the light–dark test (at 4, 24, or 72 h after exposure to TMT or odor-free Petri dish) for biochemical analysis of the α2δ-1 subunit.
Other independent groups of mice were used to examine the anxiolytic-like effect of pregabalin after exposure to TMT or an odor-free Petri dish. Again, only mice displaying > 1-min freezing during TMT exposure were included in the TMT groups. In a first experiment, mice exposed to an odor-free Petri dish or to TMT were used for the assessment of anxiety-like behavior in the light–dark box at 4 h after exposure. These mice were treated i.p. with pregabalin (30 mg/kg, dissolved into saline) or with an equal volume of saline 1 h prior to the light–dark test, i.e. 3 h after exposure to the odor-free gauge or to TMT (n = 7, in the group of odor-free + saline, i.p.; n = 6, in the group of odor-free + pregabalin; n = 8, in the group of high responder to TMT + saline; n = 8 in the group of high responders to TMT + pregabalin). In a second experiment, we tested the anxiolytic-like effect of pregabalin at 24 h after exposure to odor-free gauge or TMT. Pregabalin or saline were injected i.p. 1 h prior to the light–dark test, i.p. 23 h after the exposure (n = 6, in the group of odor-free + saline, i.p.; n = 6, in the group of odor-free + pregabalin; n = 10, in the group of high responder to TMT + saline; n = 11 in the group of high responders to TMT + pregabalin).
The apparatus consisted of a rectangular plexiglass box (20 × 50 × 20 cm) with a black chamber comprising 1\3 of the total volume. The two sections were separated by a plexiglass septum with an open door (12 × 5 cm) that permitted the passage from the illuminated chamber to the enclosed dark chamber. Mice were videotaped, and the time spent by each mouse in the two chambers was measured. Mice were considered to have entered a chamber when all four paws were positioned within the section.
Mice were killed by decapitation 20 min after the end of the light–dark session, and the hippocampus, prefrontal cortex and amygdaloid complex were dissected and stored at −80°C. Anatomical dissection was standardized with a mouse stainless steel coronal brain matrix with a slice width of 0.5 mm using chilled razor blades. Dissection was performed at 2°C. To dissect the prefrontal cortex, we cut the portions of cerebral hemispheres anterior to a point roughly corresponding to +2.00 mm (AP) from bregma (according to the Franklin and Paxinos mouse brain atlas, 1997) . Both the medial and lateral portions of the prefrontal cortex were dissected. To dissect the amygdaloid complex, we made a coronal slice 1.5 mm width with the anterior portion roughly corresponding to −0.5 mm, and the posterior portion to −2.00 mm from bregma. To obtain the amygdaloid complex, we dissected the ventral portion of the slice between the lateral border of the hypothalamus and the medial border of the piriform cortex. The hippocampus was dissected out from the remaining portion of the brain. For western blot analysis, tissue was homogenized at 4°C in ice-cold 0.1% sodium dodecyl sulfate (SDS)-lysis buffer containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride , 1 μg/mL aprotinin and 1 μg/mL of leupeptin) and phosphatase inhibitors (1 mMNaF, 1 mM Na3VO4 and 1 mM glycerol-2-phosphate) with a motor-driven Teflon-glass homogenizer (1700 rpm). Homogenates were centrifuged at 17 000 g, 4°C, for 20 min and the supernatant was used for protein determinations. Samples containing 30 μg protein were resuspended in SDS-bromophenol blue reducing buffer containing 40 mM dithiothreitol. Samples were electrophoresed on 8% SDS polyacrylamide gels, which were then electroblotted on nitrocellulose membranes (Bio-Rad, Hercules, CA, USA). Filters were blocked overnight at 4°C in Tween-tris-buffered saline (TTBS) buffer (100 mM Tris–HCl; 0.9% NaCl; 0.1% Tween 20; pH 7.4) containing 5% non-fat dry milk and then, respectively, incubated for 2 h at 20°C with 1 μg/mL of a polyclonal antibody that recognizes the N-terminal region of the α2δ-1 subunit of VSCCs (Millipore, Temecula, CA, USA), or with 0.2 μg/mL of a monoclonal anti-β-actin antibody (Sigma-Aldrich, St. Louis, MO, USA). Blots were washed three times with TTBS buffer and then incubated for 1 h with secondary antibodies (peroxidase-coupled anti-rabbit or anti-mouse; Millipore) diluted 1 : 5000 with TTBS. Immunostaining was revealed by enhanced chemiluminescence (Amersham Biosciences, Milan, Italy).
Real-time PCR analysis
Total RNA was extracted from the amygdala, hippocampus, or frontal cortex with Trizol reagent (Invitrogen, Milan, Italy) and subjected to DNaseI treatment (Promega, Milan, Italy) according to the manufacturer's instructions. One microgram of total RNA was then employed for cDNA synthesis, using ImProm-II Reverse Trascriptase (Promega) and random hexamer primers according to the manufacturer's instructions. One microliter of cDNA was employed for amplification. Amplification of Cacna2d1 cDNA was carried out using the following primers: forward: CAGCAACGCTCAGGATGTAA; reverse: ATCTGTGATCCCCTTTGCTG; β-actin: forward: GTTGACATCCGTAAAGACC; reverse: TGGAAGGTGGACAGTGAG. Real-time quantitative PCR was performed using a 2X SensiMix SYBR & Fluorescein Kit (Bioline, Rome, Italy) containing the double-stranded DNA-binding fluorescent probe Syber Green and all necessary components except primers. Quantitative PCR conditions included an initial denaturation step of 94°C/10 min followed by 35 cycles of 94°C/30 s and 58°C/30 s. Standards, samples, and negative controls (no template) were analyzed in triplicate. Concentrations of mRNA were calculated from serially diluted standard curves simultaneously amplified with the unknown samples and corrected for β-actin mRNA levels.
Statistical analysis was carried out by Student's t-test (Fig. 1a) or by two- or three-way anova followed by Fisher's LSD Multiple Comparison test (Figs 1b, 2, 3, and 4).
About 30% of mice exposed to TMT displayed a strong defensive behavior characterized by a total duration of freezing > 60 s in the 10-min observation period (t(40) = −38.124, p < 0.05, as compared to mice exposed to odor-free gauge). During exposure to TMT, ‘high responder’ mice showed reduced contact frequencies with the odorant pad (t(40) = 5.813; p < 0.05), an increased defensive burying (t(40) = −2.675; p < 0.05), and a reduced grooming behavior (t(40) = 3.383; p < 0.05) with respect to control mice exposed to odor-free gauge (Fig. 1a). All these mice were subdivided into different groups and analyzed for anxiety-like behavior in the light–dark box and for α2δ-1 subunit expression by immunoblot analysis and real-time PCR at 4, 24, or 72 h following exposure to TMT or saline. The same mice tested at the light–dark box were killed 20 min later for biochemical analysis. ‘High responder’ mice spent less time in the light chamber (i.e. they were more ‘anxious’) than mice exposed to odor-free gauge at 4, 24, and 72 h after odor exposure (group effect – TMT vs. odor-free -, F(1,41) = 75.56; p < 0.05) (Fig. 1b).
Immunoblot analysis showed an increased expression of the α2δ-1 subunit of VSCCs in the amygdaloid complex of ‘high responder’ mice exposed to TMT as compared to control mice exposed to saline. The increase was already substantial at 4 h and persisted at 72 h (the last time point examined) (group effect – TMT vs. odor-free, F(1,41) = 78.07; p < 0.05) (Fig. 2). No changes in the expression of the α2δ-1 subunit were seen in the hippocampus and frontal cortex of mice exposed to TMT (Fig. 2).
We also measured the transcript of the α2δ-1 subunit in mice exposed to TMT or saline. Real-time PCR analysis showed no changes in mRNA levels of α2δ-1 subunit in the amygdala, hippocampus or frontal cortex at 4, 24, or 72 h following TMT exposure (Fig. 3).
In a separate set of experiments, we examined the anxiolytic-like effects of pregabalin in control mice exposed to odor-free gauge and in mice exposed to TMT and classified as ‘high responders’ (> 1-min freezing during exposure). Pregabalin (30 mg/kg) or saline were injected i.p. 1 h prior the light–dark test either, i.e. 3 or 23 h following exposure to odor-free gauge or TMT. Different groups of mice were used for the assessment of anxiety-like behavior at 4 or 24 h after exposure to avoid that pregabalin injected after 3 h could interfere with measurements at 24 h. At both 4 and 24 h, injection of pregabalin did not affect behavior in the light–dark test in mice exposed to odor-free gauge, but abolished anxiety-like behavior in mice exposed to TMT (Fig. 4a and b). Interestingly, mice exposed to TMT and treated with pregabalin showed less anxiety-like behavior with respect to all other groups of mice at 4 h, but not at 24 h (Fig. 4a and b) (group effect – TMT vs. odor-free -, F(1,61) = 26.38; p < 0.05; treatment effect – pregabalin vs. saline -, F(1,61) = 56.66; p < 0.05).
Our data offer the first demonstration that expression of the α2δ-1 subunit of VSCCs displays plastic changes in response to a behavioral paradigm that results in anxiety-like behavior in mice. Here, mice were exposed to the predator odor, TMT, which is known to produce unconditioned fear in rodents, and causes behavioral reactions that recapitulate animal phobias in humans (reviewed by Kavaliers and Choleris 2001; Rosen et al. 2008). Lesion studies suggest that a ‘medial hypothalamic defensive circuit’ involving the anterior hypothalamic nucleus, the dorsomedial part of the ventromedial hypothalamic nucleus, and the dorsal premammillary nucleus, is critically involved in the unconditioned fear response to predator odor exposure in rodents (Canteras et al. 2001; Canteras 2002). This hypothalamic circuit receives major inputs from the medial nucleus of the amygdala and the bed nucleus of the striaterminalis, which carry the olfactory information to the hypothalamus (Shipley et al. 2004). Lesions of the medial amygdala or the bed nucleus of the striaterminalis reduce freezing in response to predator odor exposure (Fendt et al. 2003; Li et al. 2004; Müller and Fendt 2006; Rosen et al. 2008). We have found that mice characterized by a robust defensive response to TMT exposure developed anxiety-like behavior that was still visible after 72 h. Hebb et al. (2004) found an increased anxiety-like behavior after 30 min, but not after 24, 48, or 168 h following TMT exposure by comparing all mice exposed to TMT with mice exposed to saline, or, alternatively, comparing ‘high responders’ with ‘low responders’. We compared ‘high responders’ versus mice exposed to saline, because our aim was to examine the expression of the α2δ-1 subunit in groups of mice that were highly divergent in terms of anxiety-like behavior. We found that anxiety was associated with increases in the α2δ-1 protein levels in the amygdaloid complex of mice exposed to TMT. This increase, with the expected amplification of calcium channel current and neurotransmitter release, might contribute to the induction and maintenance of anxiety in response to unconditioned fear. The dissection procedure we have used did not allow us to determine where precisely the α2δ-1 subunit was up-regulated inside the amygdaloid complex. Immunohistochemical analysis (see Taylor and Garrido 2008) may help to specifically address this important issue. The association between an increased expression of the α2δ-1 subunit in the amygdala and unconditioned fear-induced anxiety was strengthened by the lack of changes in the hippocampus and prefrontal cortex, two brain regions that are only minimally involved in the behavioral response to predator odor exposure (Rosen et al. 2008).However, we should highlight that no effort was made in dissecting the subregions of the prefrontal cortex, and, therefore, we cannot exclude that anatomically restricted changes in the expression of the α2δ-1 subunit occur in the prefrontal cortex in response to TMT.
The molecular nature of the increased expression of the α2δ-1 subunit we have found in the amygdaloid complex of mice exposed to TMT is unknown. The increase was already substantial at 4 h after TMT exposure, and was not associated with changes in mRNA levels. Thus, it is unlikely that TMT exposure changed the de novo synthesis of the α2δ-1 subunit. A series of elegant studies have shown that gabapentin inhibits the Rab11-dependent insertion of the α2δ-2 subunit from post-Golgi compartments to the plasma membrane (Hendrich et al. 2008; Tran-Van-Minh and Dolphin 2010). It has been hypothesized that gabapentin displaces an endogenous ligand that promotes cell-surface expression of the α2δ-1 subunit (Hendrich et al. 2008). We speculate that unconditioned fear caused by predator odor exposure affects the intracellular trafficking of the α2δ-1 subunit in the amygdala, thereby enhancing protein recycling to the plasma membrane at the expenses of the endosomal pool that is destined to proteasomal degradation. This hypothesis warrants further investigation.
It was important to demonstrate that treatment with pregabalin, which is one of the two marketed ligands of the α2δ-1 subunit of VSCCs, was ‘therapeutically’ effective in the TMT model of innate anxiety. In spite of the wide use of pregabalin in the treatment of anxiety disorders (see Introduction and references therein), there are only a few studies showing the anxiolytic effect of the drug in animal models. Pregabalin has shown anxiolytic effects in conflict tests in rat and mice (Field et al. 2001; Lotarski et al. 2011), and in the elevated X maze in rats (Field et al. 2001). In addition, pregabalin was found to reduce anxiety-like behavior induced by cannabinoid withdrawal in mice (Aracil-Fernández et al. 2011). Of particular relevance to our finding is the evidence that pregabalin was effective in a rat model of post-traumatic stress disorder induced by a 10-min exposure to predator urine scent (well-soiled cat litter for 10 min). Using this model, pregabalin, at a dose of 30 mg/kg, reduced anxiety in the elevated plus maze and acoustic startle response paradigms at 1 h following predator urine scent exposure (Zohar et al. 2008). Here, pregabalin (30 mg/kg, i.p.) showed a clear-cut anxiolytic effect at both 4 and 24 h following TMT exposure, two time points that coincide with the increased levels of the α2δ-1subunit in the amygdala Remarkably, pregabalin treatment did not affect behavior in the light–dark test in control mice (i.e. in mice exposed to an odor-free gauge), suggesting that pregabalin behaved as an anxiolytic drug in a context characterized by an up-regulation of the α2δ-1 subunit. We have no explanation for the substantial anxiolytic-like effect of pregabalin seen at 4 h following TMT exposure, when TMT-exposed mice treated with pregabalin were even less ‘anxious’ than control mice exposed to the odor-free gauge.
In conclusion, our data support the hypothesis that α2δ ligands act as ‘disease-dependent’ drugs and raise a number of interesting questions that need to be addressed, i.e. (i) where precisely the α2δ-1 subunit is up-regulated within the amygdaloid complex and perhaps in other brain structures involved in anxiety; (ii)whether and how the α2δ-1 subunit responds to other behavioral paradigms of anxiety, including fear conditioning; and (iii) how anxiety affects specific functions related to α2δ, including trafficking and activity of voltage-sensitive calcium channels or changes in neuroplasticity resulting from the interaction between the α2δ subunit and thrombospondins (Eroglu et al. 2009).
The authors have no conflict of interest to declare.