Association between the A107V substitution in the δ-opioid receptors and ethanol drinking in mice selected for high and low analgesia
Abstract
Experimental evidence suggests that endogenous opioids play an important role in the development of ethanol addiction. In this study, we employed two mouse lines divergently bred for opioid-mediated stress-induced analgesia. In comparison with HA (high analgesia line) mice, LA (low analgesia line) mice, having lower opioid receptor system activity, manifest enhanced basal as well as stress-induced ethanol drinking. Here, we found that recently discovered C320T transition in exon 2 of the δ-opioid receptor gene (EU446125.1), which results in an A107V substitution (ACA23171.1), leads to higher ethanol preference in CT mice compared with CC homozygotes. This genetic association is particularly evident under chronic mild stress (CMS) conditions. The interaction between stress and ethanol intake was significantly stronger in HA than in LA mice. Ethanol almost completely attenuated the pro-depressive effect of CMS (assessed with the tail suspension test) in both the CC and CT genotypes in the HA line. In the LA mice, a lack of response to ethanol was observed in the CC genotype, whereas ethanol consumption strengthened depressive-like behaviours in CT individuals. Our results suggest that constitutively active A107V substitution in δ-opioid receptors may be involved in stress-enhanced vulnerability to ethanol abuse and in the risk of ethanol dependence.
Introduction
Despite accumulating evidence supporting the contribution of genetic background to alcoholism susceptibility, specific genes and gene variants that explain ethanol-related traits remain largely unknown. However, several lines of evidence point to polymorphisms in genes encoding opioid receptors and endogenous opioid peptides that facilitate a propensity to ethanol dependence (Schinka et al. 2002). This notion is supported by the observation that the activity of the opioid system appears to be diminished in ethanol abusers and in animal models of this disorder (Barr et al. 2004; van der Kam et al. 2005; Sacharczuk et al. 2008; Bodnar 2010). Therefore, animals with an inherited alteration in opioid system activity are useful models for the study of the genetics of ethanol abuse.
In accordance with this premise, we used two mouse lines that have been selectively bred for high (HA) and low (LA) swim stress-induced analgesia (SSIA). HA mice exhibit increased opioid system activity compared with their LA counterparts (Panocka, Marek & Sadowski 1986a, 1991; Kest et al. 1993; Sadowski & Panocka 1993; Mogil et al. 1996a,b). These lines also differ in basal nociception and in the magnitude of ethanol-induced analgesia (EIA) (Panocka, Marek & Sadowski 1986b; Mogil et al. 1993). High SSIA and EIA in the HA line are partially reversed by the non-specific opioid antagonist naloxone and the NMDA receptor antagonist dizocilpine, which implies that these phenomena in the HA line have a mixed opioid/non-opioid character. Conversely, SSIA close to the baseline in the LA line is insensitive to naloxone (Marek et al. 1992, 1993; Mogil et al. 1996b). HA mice are also less sensitive to morphine analgesia than is the LA line (Lutfy et al. 1994), and manifest analgesic cross-tolerance between morphine and swim stress (Sadowski & Panocka 1993). As only mild differences in opioid receptor density were found in certain brain regions, the divergent opioid systems activity in the phenotypes of the HA and LA lines appears a complex phenomenon, thought to rely upon multiple molecular mechanisms (Mogil et al. 1994; Kest et al. 1999).
One factor contributing to the HA/LA divergence in opioid analgesia may be a C320T transition in exon 2 of the δ-opioid receptor gene mentioned in our previous publication (Sacharczuk et al. 2010a). The carriers of the CT genotype display lower baseline and swim analgesia in comparison with their CC counterparts. Because the CT mutation is more frequent in the LA than in the HA line, this genetic difference might account, albeit only partially, for the opioid activity difference observed between the cohorts of HA and LA mice sampled at random from the colony.
Apart from pain-related traits, LA mice displayed greater ethanol consumption and ethanol preference upon chronic mild stress (CMS) than without stress, when submitted to a two-bottle free-choice procedure (Sacharczuk et al. 2008). On the contrary, we observed only small effects of CMS on ethanol intake in HA mice, which display enhanced opioid system activity. Moreover, we showed line-specific effects of ethanol drinking on food intake and body weight (Sacharczuk et al. 2010b) as well as on pro-depressive and pro-nociceptive effects of chronic stress (Sacharczuk et al. 2009) under normal and CMS conditions.
In the present study, we describe our recent intriguing results showing that the C320T polymorphism in the δ-opioid receptor gene can explain previously reported line-specific differences in ethanol preference. In a multifactorial experimental model, we attempted to assess the role of stress × genotype (opioid system) interaction in mediating susceptibility to ethanol intake and ethanol preference. We showed that the C320T transition in exon 2 in the opioid δ-receptor gene resulted in a genotype- and line-dependent increase in ethanol intake and ethanol preference during CMS.
Materials and Methods
Animals
Subjects were 6-week-old male Swiss-Webster mice from a colony maintained at the Institute of Genetics and Animal Breeding of the Polish Academy of Sciences in Jastrzebiec. The mice were selectively bred for 76 generations towards high (HA) or low (LA) analgesia produced by 3 minutes of swimming in 20°C water (e.g. Panocka et al. 1986b). After weaning, the mice were housed in groups of 4–5 siblings per cage at ambient temperature of 22 ± 2°C and 55 ± 5% relative humidity on a 12-hour light/dark cycle (lights on at 07:00). The mice had free access to tap water and food [murine chow pellets provided by LABOFEED H, MORAWSKI Co., Kcynia, Poland: 22% protein (1.5% lysine), 5% crude fibre, 4% crude fat, 6.5% crude ash, and 13,4 kcal/g of energy]. Ten days before the ethanol consumption test, the animals were transferred to individual polycarbonate shoebox cages, where they remained throughout the entire experiment. At the beginning of the experiment, HA mice weighed 35–36 g, and LA mice weighed 34–36 g.
A functionally relevant polymorphism in the δ-receptor gene was genotyped in 486 HA and 218 LA mice and subsequently studied in 80 mice. In total, 40 mice from each line were randomly assigned to four equal groups to evaluate the effects of line (HA versus LA), conditions [CMS (chronic mild stressing) versus NC (normal conditions)] and genotype (CC versus CT genotype) on ethanol intake and ethanol preference. Any interaction between CMS with genetic background (HA versus LA mice) and CMS with C320T transition was detected in replicate groups without access to ethanol.
Genotyping
DNA was isolated from frozen mouse tail samples using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany) following standard protocols and procedures. Amplification of an 892 bp fragment of the δ-receptor gene was conducted using a PT-200 thermocycler (MJ Research, Waltham, MA, USA). The total reaction volume of 10 μl included 10 ng of genomic DNA, 25 mM dNTPs, 25 μM of each specific primer (F-5′GGTCCTTTGTCGATCTTTGAGC3′ and R-5′ACTGGCTGACTTTCCACCCTAC3′) and 2 U/μl of Taq DNA polymerase (Polgen, Lodz, Poland) in a buffer containing 100 mM Tris-HCl, 20 mM MgCl2 and 500 mM KCl in injection quality water (Polpharma, Starogard Gdanski, Poland). The following reaction conditions were applied: 3:30 minutes at 94°C followed by 32 amplification cycles (30 seconds at 94°C, 50 seconds at 62°C, 1:30 minutes at 72°C) and a final elongation for 10 minutes at 72°C.
Automated sequencing was performed in both directions on an ABI PRISM™ 377 DNA Sequencer (Applied Biosystems, Foster City, CA, USA). Sequences were aligned and analysed with Sequencher Demo™ software (Gene Codes Corporation, Ann Arbor, MI, USA). The polymorphism was identified in sequence traces displayed by the software.
Behavioural procedures
Behavioural testing was performed to analyse the effect of the δ-opioid receptor C320T polymorphism on ethanol intake/preference and to determine whether this polymorphism alters the effect of ethanol on CMS-induced depressive-like behaviour.
Ethanol intake and preference testing
Mice could choose freely between 8% ethanol and tap water from two 25-ml glass bottles. To eliminate possible placement preferences, the positions of the ethanol- and water-containing bottles were switched each day. The bottles were controlled for spillage. The 8% ethanol solutions were prepared by diluting 96% ethanol (Chempur, Piekary Slaskie, Poland) with distilled water. The intake of ethanol and water was assessed by weighing the ethanol and water bottles every 24 hours. The mice were also weighed at that time. Daily ethanol intake was calculated in g per kg body weight after correction of the consumed ethanol solution for the specific gravity of ethanol.
CMS procedure
After 1 week of habituation, CMS groups of the HA and LA lines were exposed to a set of various stressors (food deprivation, tiled cage, soiled cage, stroboscopic illumination, overnight illumination, removed bedding and noise emitted by a radio) that were varied in 12-hour cycles over the course of 6 weeks. Non-stressed groups were maintained under normal conditions (NC groups). Mice exposed to CMS were housed in separate animal rooms from those not exposed to stress. Detailed procedures for CMS were adapted from Willner et al. (1987) and those described previously (Sacharczuk et al. 2008, 2009, 2010b). To achieve an unpredictable CMS procedure, stressors were applied in a semi-random order. All mice received the same treatment schedule occurring in various orders in subsequent weeks.
Tail suspension test (TST)
The behavioural changes caused by CMS and alcohol were assessed with the TST, as suggested by Steru et al. (1985). The TST was conducted between 9:00 am and 4:00 pm on the day following either the last NC or CMS session in a separate experimental room. In order to avoid acute alcohol withdrawal during behavioural testing (which typically causes hyperexcitability), as a result of abstinence from voluntary alcohol drinking, alcohol bottles were removed shortly before behavioral testing. The person performing the tests was unaware of the line, conditions and treatment. Animals were observed in a 680 × 365 × 280 mm (H × W × L) wooden box with the front wall removed. A fabric ribbon (200 × 17 × 1 mm) was attached to a cover. The mouse was suspended by attaching its tail with an adhesive tape to the ribbon. Adhesive tape was placed 30 mm from the base of the tail. The suspended animal was 120 mm away from the box walls. The total duration of immobility was scored for 6 minutes using the EthoVision system (Noldus, Wageningen, the Netherlands). The only unquestionable attempts to escape (e.g. strong upward jerks, grabbing the base of the tail or climbing) qualified the animal as active.
Blood ethanol concentrations (BECs)
BEC (mg/ml) was measured in the same group of non-stressing mice without access to ethanol, used in the previous TST experiment. The measurements were made 2 weeks after the TST. In total, 8–10 mice per group received intraperitoneal injections of 4.0 g/kg of ethanol, and 20 μl of blood was obtained via tail vein puncture at 30 and 120 minutes post-injection. Blood samples were stored at −80°C until BECs were determined using an enzymatic ethanol analyser.
Ethical note
The State Ethics Commission, in conformity with Polish law, approved the experimental protocol. All the procedures are commonly used and considered ethically acceptable in all European Union countries and North America. They also conform to the NIH Guide for the Care and Use of Laboratory Animals (Zimmermann 1983).
Statistical analysis
Data were analysed with an analysis of variance (ANOVA) using Statistica® 7.1 (StatSoft Inc., Tulsa, OK, USA). Mouse line, genotype, ethanol and CMS were considered independent factors. Because all measurements were taken in the same animals at different timepoints, time was included as a repeated measure in the analysis. A Bonferroni test was conducted for post hoc comparisons. All results presented in the graphs are the means ± SE. Significance was set at a probability level of P = 0.05.
Results
Frequency of C320T transition
The C320T transition in exon 2 of the δ-opioid receptor gene resulted in an A107V substitution in the first extracellular loop (EL1) of the peptide chain and was only observed in heterozygous form (CT genotype). The frequency of heterozygotes with the CT genotype was significantly higher in LA (35.0%) mice than in HA mice (16.0%) (χ2 = 16.64, d.f. = 1, P < 0.001)—Table 1.
| Line | Number of genotypes | Frequency of genotypes | Frequency of alleles | |||
|---|---|---|---|---|---|---|
| CC | CT | CC | CT | C | T | |
| HA | 210 | 40 | 84 | 16 | 0.92 | 0.08 |
| LA | 74 | 40 | 65 | 35 | 0.82 | 0.18 |
Ethanol intake and ethanol preference
A four-way ANOVA showed that CT heterozygotes displayed higher ethanol intake [F(1,72) = 28.16; P < 0.001] and ethanol preference [F(1,72) = 22.38; P < 0.001] than CC homozygotes. A significant interaction between line and genotype indicated that the phenotypic consequences of this transition on drinking behaviour are greater in LA mice than in HA mice [F(1,72) = 7.16 (ethanol intake) and 9.18 (ethanol preference); both P < 0.01]. CMS significantly increased ethanol intake [F(1,72) = 19.21; P < 0.001] and ethanol preference [F(1,72) = 18.14; both P < 0.001] more effectively in LA mice than in HA mice [F(1,72) = 5.08 (ethanol intake) and 5.93 (ethanol preference); P < 0.05—line × CMS interaction]. Moreover, the effect of CMS was greater in CT genotype animals within each line [F(1,72) = 4.36 (ethanol intake) and 4.02 (ethanol preference); both P < 0.05—line × genotype × CMS interaction]. Significant effects of line × genotype × time interaction indicate a time-dependent difference between the lines manifested by a rapidly declining ethanol intake [F(5,360) = 2.84; P < 0.001] and ethanol preference [F(5,360) = 2.43; P < 0.001] in HA mice with the CT genotype (Figs 1 & 2).
Ethanol intake (±SE) in high analgesia (HA) and low analgesia (LA) mice with CC (homozygous) or CT (heterozygous) genotypes. The mice were allowed a choice between a bottle of 8% ethanol and a bottle of water under normal conditions (NCs) or during the chronic mild stress (CMS) conditions. Each group consisted of 10 mice. Post hoc comparisons within each line: +: NC/CT versus NC/CC; *: CMS/CT versus CMS/ /CC; #: CMS/CT versus NC/CT. One, two or three characters indicate P < 0.05, P < 0.01 or P < 0.001, respectively
Ethanol preference (±SE) in high analgesia (HA) and low analgesia (LA) mice with CC (homozygous) or CT (heterozygous) genotypes. The mice were allowed a choice between a bottle of 8% ethanol and a bottle of water under normal conditions (NCs) or during the chronic mild stress (CMS) conditions. Each group consisted of 10 mice. Post hoc comparisons within each line: +: NC/CT versus NC/CC; *: CMS/CT versus CMS/ /CC; +: CMS/CC versus NC/CC; #: CMS/CT versus NC/CT. One, two or three characters indicate P < 0.05, P < 0.01 or P < 0.001, respectively
A three-way ANOVA test of the data from the HA line showed that individuals with the CT genotype display higher ethanol intake [F(1,36) = 8.11; P < 0.01] and ethanol preference [F(1,36) = 9.27; P < 0.01] than homozygous (CC) HA individuals. CMS increased the ethanol intake and ethanol preference more effectively in CT animals as indicated by a significant genotype × CMS interaction [F(1,36) = 4.21 (ethanol intake) and 4.49 (ethanol preference); both P < 0.05] (Figs 1 & 2).
In the LA line, heterozygotes (CT) displayed higher ethanol intake [F(1,36) = 19.93; P < 0.001] and ethanol preference [F(1,36) = 18.45; P < 0.001] compared with homozygous (CC) LA individuals. The interaction between genotype and CMS reflects a higher increase of ethanol intake [F(1,36) = 5.41; P < 0.05] and ethanol preference [F(1,36) = 5.02; P < 0.05] by individuals with CT genotype in response to CMS conditions (Figs 1 & 2).
The effect of genotype and CMS on depression-like behaviour
The effect of CMS on the duration of immobility in the TST was studied in HA and LA mice given no access to ethanol (Fig. 3). CMS significantly prolonged the time of immobility as confirmed by a three-way ANOVA [F(1,72) = 9.19; P < 0.01]. CMS was more effective in HA mice than in LA mice, as shown by the line × CMS interaction [F(1,72) = 4.93; P < 0.05], and no significant difference was detected between CC and CT genotype animals within each line, as indicated by a non-significant line × genotype × CMS interaction. This finding was confirmed by the significant main effect of CMS in the HA line [F(1,36) = 16.02; P < 0.001] and LA line [F(1,36) = 7.56; P < 0.05], as well as the non-significant effects of genotype and the genotype × CMS interaction in both lines. A one-way ANOVA performed within genotypes in the HA line showed a significant effect of CMS in animals with both genotypes [F(1,18) = 17.14; P < 0.001 (CC genotype) and F(1.18) = 12.08; P < 0.01 (CT genotype)]. Significant effects of both genotypes were also confirmed in the LA line [F(1,18) = 6.48; P < 0.05 (CC genotype) and F(1,18) = 5.18; P < 0.05 (CT genotype)].
Mean time of immobility (±SE) in tail suspension test (TST) in high analgesia (HA) and low analgesia (LA) mice with CC (homozygous) or CT (heterozygous) genotypes. Mice were housed in normal conditions (NC) or in chronic mild stress conditions (CMS) with free access to 8% ethanol (Ethanol groups) or to water (No Ethanol groups). Each group consisted of 10 mice. Post hoc comparisons within genotypes in each line: x: CMS/No ethanol versus NC/No ethanol; +: CMS/Ethanol versus NC/Ethanol. Interactions within genotypes: *: CMS × Ethanol. Interactions between genotypes: #: Genotype × CMS × Ethanol. One, two or three characters indicate P < 0.05, P < 0.01 or P < 0.001, respectively
Alterations of depressive-like behaviour by ethanol
A three-way ANOVA of the data from the HA line showed that ethanol attenuated the depressive effect of CMS (Fig. 3). This finding is supported by the significant effect of the CMS × ethanol interaction [F(1,76) = 8.35; P < 0.01]. The significant effect of the CMS × genotype × ethanol interaction [F(1,76) = 4.51; P < 0.05] indicates that ethanol was much less effective in heterozygotes (CT). A significant CMS × genotype × ethanol interaction [F(1,76) = 5.04; P < 0.05] in the LA line reflects the lack of response to ethanol in CC genotype animals but a prolonged time of immobility in the CT genotype animals (Fig. 3).
Ethanol metabolism
To determine whether the difference in the depression-like behaviour under the action of ethanol might be due to alterations in ethanol metabolism and clearance in the CC and CT genotype mice of the HA and LA lines, we examined BECs (mg/ml) at 30 minutes and 120 minutes after intraperitoneal injection of 4.0 g/kg ethanol (Fig. 4). No differences in BEC were observed between genotypes and between lines. Therefore, we assume that the concentration of ethanol in the blood of mice of each line and CC/CT genotype was maintained on a level corresponding to the amount of daily ethanol consumption. These data indicate that the A107V substitution modified the anti-depressive properties of ethanol without affecting ethanol metabolism and clearance.
Blood ethanol concentration [BEC (mg/ml)] in high analgesia (HA) and low analgesia (LA) mice with CC (homozygous) and CT (heterozygous) genotypes after an injection of 4.0 g/kg ethanol indicating no difference between genotypes and lines. The BEC is shown in mg/ml over time. Each group consisted of 8–10 mice. Error bars denote the SEM
Discussion
Numerous studies have found that long-lasting stress increases individual vulnerability to alcohol abuse (Heath et al. 1997; Dai, Thavundayil & Gianoulakis 2002; Bertholomey et al. 2011; Lopez, Doremus-Fitzwater & Becker 2011). However, the biological factors that contribute to the stress-enhanced development of alcohol addiction are not completely understood. In our research, we identified one of the putative links between ethanol intake, chronic stress and genetic background. As we showed previously, HA animals displayed a higher sensitivity to the depressive effect of the CMS conditions than the LA mice. However, the LA mice but not the HA mice subjected to CMS displayed an increased ethanol intake (Sacharczuk et al. 2008). We found that the depressogenic effect of CMS in HA mice is effectively attenuated by low doses of ethanol. Due to the low depressogenic effect of CMS, high doses of ethanol are not effective or even increase the CMS-induced depressive behaviour in LA mice. In accordance with these observations, we hypothesize that depressive behaviour under stress conditions may have an adaptive value (Rybakowski & Rybakowski 2006). Such a hypothesis explains the low ethanol intake by HA mice, whereas the lack of an adaptive response to CMS in LA mice favours an increased ethanol intake. The negative effects of CMS in LA mice reflected by the development of obesity and hormonal changes were confirmed in one of our previous studies (Sacharczuk et al. 2010b).
In this study, we showed that the amount of ethanol consumption in the HA and LA mice is associated with a recently discovered C320T transition in exon 2 of the δ-opioid receptor gene. Within the cohort of mice examined for ethanol preference, we found that 16% of the HA and 35% of the LA mice were carriers of the C320T transition. The LA carriers of the CT genotype displayed a higher basal and stress-induced ethanol intake than the CC mice of the same line. Such genotype-dependent differences in ethanol intake were much less pronounced in the HA line. We found that the higher ethanol intake by LA mice might be related to a higher frequency and greater effect of the C320T transition in this line.
The association between genotype and ethanol intake is enhanced under CMS conditions, and the ethanol preference was often increased to 90% in some LA animals with the CT genotype. To validate our CMS procedure, we studied depression-like behaviour in TST. No distinction in the pro-depressive effects of CMS was observed between the CC and CT genotypes in the HA or LA line. Therefore, the higher ethanol intake in the CT genotype specimens was not directly caused by a higher sensitivity of the CT individuals to CMS. Our hypothesis that depression does not alter ethanol intake was corroborated by the data obtained from the CD mice (Pelloux et al. 2005) and is in agreement with the low efficiency of antidepressive drugs in alcohol abuse treatment (Kranzler et al. 2006).
Genotype-dependent changes in ethanol intake were accompanied by a genotype-dependent effect of ethanol on the depression-like behaviour. Under normal conditions, ethanol did not affect the TST time of immobility in animals with the CC or CT genotypes from either the HA or LA line. However, ethanol almost completely attenuated the pro-depressive effects of CMS in the CC HA mice and was less effective in CT genotype specimens. In the LA mice, we observed no effect of ethanol on the CC genotype (due to the small effect of CMS) but a prolonged time of immobility in CT animals (despite the small effect of CMS). Therefore, we assume that ethanol exerts an antidepressive effect in the HA mice of either genotype but a pro-depressive effect in the CT LA mice. Ethanol consumption appears to influence the depressive-like behaviour, but this effect in our mice depends on the opioid profile of the animal. Simultaneously, we observed that even high doses of ethanol taken by the CT genotype group were ineffective in attenuating the anxiogenic effect of CMS (unpublished data). The anti-depressive effect of ethanol consumed in the free choice paradigm was also shown in other animal models of depressive behaviour, e.g. the Wistar-Kyoto rat strain (Paré, Paré & Kluczynski 1999). However, chronic alcohol exposure itself constitutes a potent stressor and reliably enhances voluntary alcohol consumption (Becker, Lopez & Doremus-Fitzwater 2011), which may reflect a depressive-like condition in the animal.
The differential effects of ethanol observed in the HA and LA mice support a hypothesis that selective breeding for high and low stress-induced analgesia modified the degree of opioid involvement in the neural mechanisms that control mood and motivation. This finding is in agreement with the prevailing opinion that the behavioural effects of ethanol are mediated through the opioid system because the antidepressant impact of ethanol is antagonized by the opioid receptor antagonist naltrexone (Froehlich et al. 1998; Mangold et al. 2000; Gianoulakis 2009). Chronic exposure to stress was also shown to enhance δ-opioid receptor function and induce a redistribution of δ-opioid receptors to the cell membrane (van Rijn, Brissett & Whistler 2012). Thus, a functional expression of the A107V substitution might be more pronounced under CMS conditions, particularly in the LA line.
In contrast to many examples of polymorphisms in the μ- or κ-opioid receptor genes that are associated with ethanol-related phenotypes (e.g. Saito et al. 2003; Kimura & Higuchi 2011; Kumar, Chakraborty & Das 2012), the importance of variability within the δ-opioid receptor gene has remained largely unknown. Most of the data related to the δ-opioid receptor functions were obtained from gene knockout mice (Roberts et al. 2001) or pharmacological analyses utilizing δ-opioid receptor antagonists (Nielsen et al. 2008; Gianoulakis 2009). In both cases, the function of the δ-opioid receptor is weakened or disabled and may mimic the C320T transition (Sacharczuk et al. 2010a).
Endogenous opioid peptides (EOPs) play a significant role in ethanol intake and the effect of ethanol on the tail suspension test in mice (Barfield et al. 2010). For example, enkephalin-deficient mice are resistant to the stress-induced elevation of ethanol consumption (Racz et al. 2008). Although HA and LA mice do not differ in the baseline and stress-induced levels of EOP (Sacharczuk et al. 2010a), the V107 δ-opioid receptor variant may display a lower affinity for enkephalins and may be linked with a higher binding affinity for the μ-opioid receptor (relative increase in the activity of the μ-opioid receptor system). However, further research is needed to confirm these hypotheses.
The A107V substitution is located in the second extracellular loop of the δ-opioid receptor. This residue occupies a remote position that is a critical determinant of high-affinity binding for δ receptor agonists. Such sites reside in the C-terminal domains TM6 and TM7 and in the third extracellular loop of the δ-opioid receptor (Befort et al. 1999; Décaillot et al. 2003). The A107V substitution caused δ-opioid receptor binding dysfunctions of stress-induced endogenous opioid peptides, exogenous specific δ-opioid receptor agonists, such as SNC80, and antagonists, such as naltrindole. This finding indicates a decreased functionality of δ-opioid receptors with this substitution and potent consequences in the ethanol-induced changes in opioid neurotransmission (Sacharczuk et al. 2010a).
The pharmacological blockage of μ-opioid receptors prevents the ethanol-induced activation of the dopamine system and reduces ethanol consumption (Nielsen et al. 2008; Kasai & Ikeda 2011). The δ-opioid receptor antagonist naltrindole also decreases ethanol intake (Matsuzawa et al. 1999). However, transgenic mice with a δ-opioid receptor knockout showed a higher preference for ethanol and consumed more ethanol than their wild-type counterparts, suggesting that a decrease in δ-receptor activity is associated with increased ethanol-drinking behaviour (Roberts et al. 2001). These seemingly conflicting results might be explained by opposing phenotypes in the opioid receptor mutants, which is in contrast with the classical notion of similar activities of the μ-opioid and δ-opioid receptors. Mice lacking μ-opioid receptors display decreased depressive and anxious behaviours, whereas mice lacking δ-opioid receptors show an intensification of these behaviours. Consistent anxiogenic and depressive-like responses in Oprd1-/- mice indicate that δ-receptor activity contributes to the improvement of mood states (Filliol et al. 2000).
We suggest that the constitutive deletion of δ-opioid receptors resembles the lower affinity of the V107 isoform for endogenous opioids. This model partially explains the association of the C320T polymorphism with the differences in ethanol preference in the HA and LA lines. We also hypothesized that the A107 and V107 isoforms of the δ-opioid receptor may possess different susceptibilities to μ-δ heterodimerization. Because μ-δ opioid receptor heterodimers (δ1-opioid receptors) and δ-opioid receptor homodimers (δ2-opioid receptors) have opposite effects on ethanol consumption, leading to the inhibition and enhancement of ethanol intake, respectively, the increased ethanol intake in the LA line and CT genotype may be discussed with respect to less μ-δ heterodimerization. The results obtained confirm that opioid system support-based therapy, especially in the area of delta receptors, could be effective in the treatment of alcohol addiction (van Rijn & Whistler 2009).
The C320T polymorphism causing an Ala107Val substitution was first described in our HA-LA mouse model. In accordance with our knowledge, there is no analogous polymorphism in another mouse strain or in other species, e.g. rats. So far, data published on the C320T polymorphism have involved its influence on the magnitude of stress-induced analgesia and sensitivity to exogenous agonists and antagonists of delta opioid receptors (Sacharczuk et al. 2010a). The present data connecting this polymorphism to a higher ethanol preference and intake may suggest similar tests in other mouse models and species with an increased ethanol preference. The clinical implications of the examined polymorphism may originate from both the predisposition to increased ethanol intake and a decreased reaction to naltrexone, which is used in addiction therapy.
In summary, the results show that LA mice (which are less prone to depressive-like behaviour) display a higher susceptibility to a CMS-induced increase in ethanol intake and that this linkage is enhanced by the A107V substitution in δ-opioid receptors. The higher ethanol consumption of the LA line is determined by a higher frequency and larger effect of the C320T transition in the gene. The effect of the C320T transition was especially evident in CT carriers under CMS conditions, in which the ethanol preference increased, even to the level of 90% in some individuals. In the HA line, the frequency of the C320T transition was approximately twofold lower, and the overall effect of the genotype × condition interaction was significantly smaller than in LA mice. Moreover, ethanol almost completely attenuated the pro-depressive effect of CMS in both homozygous CC and heterozygous CT HA mice. In the LA line, ethanol enhanced the pro-depressive effects of CMS in heterozygotes and was ineffective in homozygotes. Importantly, the C320T transition represents a decrease-of-function mutation for the murine δ-opioid receptor, which may have an effect on opioid-regulated behaviours or ethanol addiction in vivo. The comparison of the HA and LA mouse lines provides an interesting model to study the consequences of genetic heterogeneity in opioid systems, which are a clinically important target for both pain and drug abuse research.
Acknowledgements
This work was supported by the Polish Committee for Scientific Research (KBN) Grant nos. 2P05A13029. The selective breeding programme has been financed by the Institute of Genetics and Animal Breeding at the Polish Academy of Science (Project No. S.I.1.3).
Authors Contribution
MS, AL and BS were responsible for the study concept and design. MS, AL, BS and AWL contributed to the acquisition of animal data. MS, AL, BS, MK, AWL and RP assisted with data analysis and interpretation of findings. MS, AL, MK, RP, AWL and BS drafted the manuscript. All authors critically reviewed content and approved the final version for publication.
