Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Most psychoactive drugs have multiple molecular targets in brain, and one of the leading challenges in neuropharmacology is to link actions on specific proteins with discrete behavioral effects of these drugs. A notable success in this area is demonstrated by recent work defining behavioral consequences of benzodiazepine action on specific subtypes of GABA_A receptors. Discovery of a mutation in the  subunit of GABA receptors that prevents actions of benzodiazepines without changing the GABA sensitivity of the receptor allowed construction of mice ("knock-in") with benzodiazepine-insensitive  subunits. This revealed that benzodiazepine modulation of the ₁, ₂, and ₃ subunits is linked to distinct behavioral actions of these drugs (Low et al., 2000; McKernan et al., 2000; Crestani et al., 2001). This paradigm provides a powerful approach to dissecting molecular determinants of behavioral effects of drugs with complex mechanisms of action. Ethanol is such a drug with effects on several neuronal ion channels, which may be important for its behavioral and physiological effects (Harris, 1999). The strategy used for benzodiazepines requires a mutation in a protein that abolishes drug action without changing the normal function of that protein. In this study we apply this approach to ethanol actions on brain glycine receptors.

The glycine receptor (GlyR) is a member of the ligand-gated ion channel superfamily that includes GABA_A, nicotinic acetylcholine, and serotonin₃ receptors. The strychnine-sensitive GlyR is an inhibitory pentameric receptor that conducts the chloride ion and inhibits neurotransmission in the spinal cord and brainstem. The GlyR is composed of three  subunits and two  subunits (Betz, 1992), and gephryn localizes GlyR at postsynaptic densities facing glycine-releasing nerve terminals (Triller et al., 1985; Todd, 1990) by simultaneously binding microtubules (Meyer et al., 1995) and the  subunit (Kirsch and Betz, 1995). GlyR function is potentiated in vitro by ethanol, longer chain alcohols, and volatile anesthetics (Celentano et al., 1988; Engblom and Akerman, 1991; Aguayo and Pancetti, 1994; Mascia et al., 1996a,b; Yamakura and Harris, 2000). Several electrophysiological experiments have shown ethanol's ability to potentiate GlyR function (Eggers et al., 2000; Ye et al., 2001). Additional behavioral tests also suggest involvement of the GlyR with the acute effects of ethanol. Glycine and the glycine precursor, serine, can enhance the depressant effects of ethanol in mice, as observed using the loss of righting reflex (LORR) test, and this effect can be blocked by strychnine (Williams et al., 1995). Ethanol also inhibits strychnine seizures in mice (McSwigan et al., 1984). However, it is still unclear whether specific behavioral effects of ethanol can be linked to potentiation of GlyR function. To address this question, we have created lines of transgenic mice that express a mutant (S267Q) ₁ subunit of the GlyR. This mutation was selected because it retains normal sensitivity to glycine but abolishes the enhancing effects of ethanol, and high concentrations of ethanol inhibit the function of this mutant receptor. Expression of this transgene in addition to endogenous expression of wild-type ₁ subunits should reduce alcohol action of GlyR; thus, any effects of ethanol that are mediated by GlyR should be reduced or eliminated in transgenic mice. We tested a variety of acute actions of ethanol, with an emphasis on those that might be related to spinal cord or motor control. To ensure that behavioral phenotypes were not due to the integration site, which will vary among transgenic lines (Whitelaw et al., 2001), we constructed and compared multiple independent lines of transgenic mice.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Oocyte Electrophysiology. Two-electrode voltage-clamp electrophysiological recordings were performed as previously described with Xenopus oocytes (Mascia et al., 2000; Findlay et al., 2001) by using wild-type ₁ and mutant ₁ S267Q GlyR. We tested the capacity of ethanol to enhance the effect of a glycine concentration that produced 5% of the maximal effect (EC₅) of glycine. Ethanol was perfused onto oocytes for 1 min and then coapplied with glycine for 30 s, or glycine was applied alone for 30 s. Maximal (peak) currents and EC₅ values were determined for each oocyte individually.

Construction of Mice. The transgenic GlyR ₁ S267Q mouse was created using the rat synapsin I (Syn) promoter and inbred (FVB/NJ) mice by the following method. The Syn promoter was obtained (Hoesche et al., 1993). The rat 4.3-kb Syn promoter was cut from the pBL4.3 Syn-Cat construct with SalI and XhoI; both sites were then blunt-ended. The 4.3-kb Syn promoter was cloned into the EcoRV site of pBluescript SK- and subsequently cut out with SmaI and XhoI and both sites were blunt-ended. The blunted Syn promoter was cloned into the pCI vector (Promega, Madison, WI) from which the cytomegalovirus promoter was removed. Orientation of the Syn promoter was confirmed by restriction digestion and sequencing. GlyR ₁ S267Q was cloned (cut out of the pCis2 vector with SalI and blunt-ended) into the SmaI site in the multiple cloning site of the pCIS-Syn vector. This construct (pCIS-Syn GlyR ₁ S267Q) was successfully expressed in Xenopus oocytes before the transgenic mice were made. Using two-electrode voltage-clamp recordings, the ethanol-insensitive pharmacology of the GlyR ₁ S267Q cDNA was observed. In preparation for creation of transgenic mice, the Syn promoter and the transgene sequences were cut away from the vector backbone of the pCIS-Syn vector with ClaI digestion. The resulting 6.4-kb band was injected into oocytes of inbred (FVB/NJ) mice at the University of Colorado Health Sciences Center, Denver, CO.

ot

ot

Drugs. The following drugs were obtained from Sigma Chemical (St. Louis, MO): ethanol, strychnine, pentylenetetrazol, ketamine, and Tween 80. Flurazepam was obtained from Hoffman-LaRoche (Nutley, NJ), and pentobarbital was obtained from Sigma/RBI (Natick, MA). Ethanol was injected i.p. at 20% v/v in saline. Drugs were dissolved in 0.9% saline, except flurazepam, which was first dissolved in 3 to 4 drops of Tween 80 before saline was added. Strychnine and pentylenetetrazol were injected s.c.; all other drugs were injected i.p.

[³H]Strychnine Binding. Brain regions (cortex, midbrain, cerebellum, and combined brainstem and spinal cord) were dissected and placed into 2 ml of ice-cold assay buffer (145 M NaCl, 5 mM KCl, 1 mM MgCl₂, 10 mM glucose, 1 mM CaCl₂, 10 mM HEPES, pH 7.5). After homogenization by using a PowerGen 700 (Fisher Scientific, Pittsburgh, PA), samples were centrifuged at 30,000g for 25 min at 4°C and resuspended in assay buffer. Samples were again centrifuged at 30,000g at 4°C for 25 min and then resuspended in assay buffer. To measure specific binding of [³H]strychnine (250 µCi; Sigma/RBI, Boston, MA) to tissue membranes, aliquots of crude synaptic membranes (approximately 200 µg of protein) were incubated in a final volume of 0.5 ml of assay buffer in Eppendorf tubes at 0-4°C in a final concentration of [³H]strychnine (1-20 nM) with or without glycine (3 mM final concentration). Glycine was used to measure nonspecific binding. Incubations were terminated by rapid filtration and washing over a vacuum by using borosilicate microfiber filters (MFS, Dublin, CA). Filters were incubated overnight in Bio-Safe II scintillation liquid (4 ml; Research Products International, Mount Prospect, IL) before being analyzed with a scintillation counter (Beckman LS 6500; Beckman Coulter, Inc., Fullerton, CA). Specific [³H]strychnine binding was obtained by subtracting the nonspecific (in the presence of cold 3 mM glycine) binding from the total bound radioactivity. Values for K_D and B_max were determined by nonlinear regression analyses using the PRISM program (GraphPad Software, San Diego, CA).

Loss of Righting Reflex. Mice were given an injection of ethanol (3.6-4.3 g/kg), pentobarbital (45 mg/kg), flurazepam (225 mg/kg), or ketamine (150 mg/kg) and the length of drug-induced loss of righting reflex (sleep time) was measured. Upon loss of the righting reflex, mice were placed in a sleep trough (~90° angle) and the time to regain the righting reflex was measured. LORR is defined as the inability of a mouse to right itself within 30 s. Return of the righting reflex was defined as the ability of a mouse to right itself twice within 1 min. Duration of LORR is defined as the difference between loss and regaining of the righting reflex.

Ethanol-Induced Hypothermia. Basal body temperature was recorded then mice received an injection of ethanol (3.0-3.8 g/kg). Body temperature was measured rectally every 10 min for 180 min by using a RET-3 probe and thermolyte TH-5 monitor (Physitemp Instruments, Clifton, NJ). Data were statistically analyzed by comparing the area under the curves (change from baseline temperature × time) and maximal temperature alteration.

Acute Functional Tolerance. Mice were tested for acute functional tolerance as previously described (Erwin and Deitrich, 1996). Mice were trained to remain on a rotarod (Economex; Columbus Instruments, Columbus, OH; speed of rod, 5 rpm) for 1 min then given a dose of ethanol (1.75 g/kg) and tested for recovery of balance every 5 min. When animals could remain on the rotarod again for 1 min, a 25-µl blood sample (BEC1) was taken from the retro-orbital sinus and the animals were immediately injected with a dose of 2 g/kg ethanol. When the animals again regained rotarod ability for 1 min, a final 25-µl blood sample (BEC2) was taken. BEC values, expressed as milligrams of ethanol per milliliter of blood was determined spectrophotometrically using an enzymatic assay (Lundquist, 1959).

Ethanol Pharmacokinetics. Mice received a 4.0-g/kg injection of ethanol. Blood samples (25 µl) were taken from the retro-orbital sinus by using capillary tubes (Drummond, Broomall, PA) at 15-, 60-, 120-, 180-, and 240-min time points. Ethanol content of blood samples was determined spectrophotometrically using an enzymatic assay (Lundquist, 1959).

Initial Sensitivity to Motor Incoordination. Mice were trained on a fixed speed rotarod (Economex; Columbus Instruments; speed of rod, 2.5 rpm), and training was complete when mice were able to remain on the rotarod for 120 s. Ethanol or saline was injected, and the mouse was placed back on the rotarod 2 min later and the ability of the mouse to remain on the rotarod for 60 s was determined. The 95% confidence limits were determined using the "up and down" method (see below) with an ethanol log dose interval of 0.0138, which corresponds to approximately a 0.1-g/kg ethanol dose difference at doses tested.

Initial Sensitivity to Loss of Righting Reflex. Mice were given saline or ethanol and 3 min later they were tested for an LORR (see above) greater than 1 min. The 95% confidence limits were determined using the "up and down" method (see below) with an ethanol log dose interval of 0.0138, which corresponds to approximately a 0.1-g/kg ethanol dose difference at doses tested.

Analysis of Data by Up and Down Method. The up and down method was used as described by Dixon and Massey (1969). Mice were injected with an initial dose and tested, and the results from each animal determined the dose that the next animal would receive. If the mouse was unable to successfully perform the task then the dose of drug administered would be decreased (by a log interval of the dose). If the mouse was able to successfully perform the task then the dose of drug administered would be increased (by a log interval of the dose). The ED₅₀ values were determined by the following equation: 95% CI = dosing increment ×  × 1.96, in which n = the last n trials and 1.96 reflects the 0.05  level (Dixon and Massey, 1969). For each trial with a given drug, six subsequent mice (n = 6) were used.

Spontaneous Motor Activity. Locomotor activity was measured in standard mouse cages with Opto-microvarimex (Columbus Instruments). Each cage had bedding, food, and water and was covered by a heavy plastic lid with holes for ventilation. Activity levels were monitored every hour for 24 h.

Elevated Plus-Maze. The elevated plus-maze was a modification of that validated by Lister (1987) and comprised two open (30 × 5 × 0.25 cm) and two enclosed (30 × 5 × 5 cm) arms, which extended from a common central platform (5 × 5 cm). The apparatus was constructed from black Plexiglas and was elevated to a height of 60 cm above floor level. All testing was conducted under room light. In accordance with established procedure (Rodgers and Johnson, 1995), mice were individually placed on the central platform of the maze facing an open arm. A normal 5-min test duration was used, with the maze thoroughly cleaned between subjects. All test sessions were recorded by a vertically mounted camera linked to a monitor and VCR in an adjacent laboratory. Parameters scored from videotape were the conventional spatiotemporal measures and a range of specific behaviors related to the defensive repertoire of the mouse (Rodgers and Johnson, 1995).

Immunoblotting. Dissected cortex, midbrain, cerebellum, brainstem, and spinal cord from transgenic and wild-type mice were homogenized using a PowerGen 700 tissue tearor (Fisher Scientific) in a buffer containing 320 mM sucrose, 5 mM HEPES, 1 mg/ml leupeptin, 1 mg/ml aprotinin, and 1 mg/ml pepstatin. Homogenates were centrifuged at 5,000g for 5 min (4°C), and the supernatant fraction was subsequently centrifuged at 30,000g for 30 min (4°C). The resulting pellet was resuspended in homogenization buffer and analyzed for protein concentrations by using the Bradford method (1976). Loading dye (200 mM Tris, 400 mM dithiothreitol, 8% glycerol, 0.4% bromophenol blue) was added to samples, which were subjected to SDS-polyacrylamide gel electrophoresis (10%). Proteins were electrophoretically transferred to a polyvinylidene difluoride membrane, and nonspecific sites were blocked in 7.5% nonfat dry milk in Tris-buffered saline (135 mM NaCl, 2.5 mM KCl, 50 mM Tris, and 0.1% Tween 20, pH 7.4). Membranes were then incubated in the presence of a polyclonal antibody to the N terminus of GlyR (rabbit anti-glycine receptor antibody; Chemicon International, Temecula, CA; Wick et al., 1999) in Tris-buffered saline containing 2.5% nonfat milk. We used a horseradish peroxidase-tagged goat antibody to rabbit IgG (ICN Pharmaceuticals, Costa Mesa, CA) and visualized using chemiluminescence (Pierce Chemical, Rockford, IL). Densitometric analysis was performed on spinal cord and brainstem samples and calibrated to coblotted dilutional standards of control spinal cord and brainstem. Blots were then stripped for 20 min at 80°C (8 M urea, 100 mM 2-mercaptoethanol, and 62.5 mM Tris, pH 6.8) and reprobed with an antibody to -tubulin (Sigma Chemical).

Chemically Induced Seizures. Mice were given an injection of saline or ethanol (0-2 g/kg) and then a subcutaneous injection of either strychnine (0.9 mg/kg) or pentylenetetrazol (50 mg/kg). The mice were observed continuously for 20 min for the appearance of tonic-clonic seizures. Data were compared using logistic regression, and ED₅₀ values were determined using standard maximum likelihood methods.

Statistics. Statistical differences between groups were determined using a two-tailed Student's t test, one-way analysis of variance, or two-way analysis of variance where indicated using the PRISM program (GraphPad Software). Logistic regression was performed using SPSS for Windows (version 10; SPSS, Chicago, IL).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Our strategy required a transgene whose glycine sensitivity would be indistinguishable from wild-type GlyR but whose function would be unaffected or inhibited (rather than potentiated) by ethanol. Site-directed mutagenesis and oocyte expression were used to investigate point mutations in the GlyR that could change ethanol action without altering the glycine dose-response relationship. Xenopus oocytes were injected with the cDNA for wild-type or mutant GlyR ₁ subunits, and glycine-activated currents were measured in the presence and absence of ethanol. After examination of a number of mutants, S267Q was chosen for the transgene. This was based on statistically indistinguishable concentration-response relationships to glycine for GlyR ₁ and ₁(S267Q) receptors (Fig. 1A). Hill values were 2.1 ± 0.2 and 1.7 ± 0.1 and EC₅₀ values were 235 ± 24 and 214 ± 10 µM for GlyR ₁ and ₁(S267Q) receptors, respectively. Unlike the potentiation of receptor function that ethanol produced in the presence of glycine in the wild-type ₁ GlyR (Mascia et al., 1996a,b; Yamakura and Harris, 2000), ethanol inhibited EC₅ glycine responses in the mutant GlyR ₁(S267Q) (Fig. 1B). Syn transgene expression of this mutant GlyR was used, and four lines of transgenic mice (tg1-tg4) were produced using a Syn GlyR ₁(S267Q) construct (Fig. 2). Unless otherwise stated, all experiments were performed on tg3 mice. All families of transgenic mice were normal in appearance (e.g., weight, coat quality, whiskers).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Pharmacology of wild-type and mutant 1 GlyR expressed in Xenopus oocytes. A, dose-response relationships of wild-type 1 and mutant 1(S267Q) GlyR expressed in Xenopus oocytes. ED50 and Hill coefficient values were statistically indistinguishable between the wild-type and mutant receptors. B, ethanol (EtOH) alteration of EC5 glycine responses in 1 and 1(S267Q) GlyR. EtOH potentiated wild-type 1 glycine responses but inhibited 1(S267Q) receptor glycine responses. Values are expressed as mean ± S.E.M. (n = 5-6).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   GlyR 1 construct used for transgenic injection. A, transgene construct (pCIS-Syn GlyR 1 S267Q) used for production of mice. B, response to ethanol of mutant receptor expressed by injecting transgene construct (pCIS-Syn GlyR 1 S267Q) DNA into Xenopus oocytes and subsequently measured currents by using a two-electrode voltage clamp. The GLY response was obtained using an EC5 concentration of glycine, and Gly + EtOH responses were obtained by coapplying 100 mM ethanol with a EC5 concentration of glycine.

[³H]Strychnine binding and immunoblot analysis were used to verify expression of the transgene. Transgene expression was observed in cortex, midbrain, and cerebellum of tg3 mice by using immunoblot analysis with a polyclonal anti-glycine receptor ₁ antibody (Fig. 3, inset). Other immunoblot experiments did not show a significant change in GlyR levels in the spinal cord and brainstem (data not shown); however, transgene expression was observed in (spd^ot/spd^ot) mice that possessed the transgene (Fig. 4). It is possible to observe the transgene protein on this genetic background because these mice lack GlyR ₁. Due to this genetic defect the (spd^ot/spd^ot) mice die at about 3 weeks of age. Unfortunately, the transgene was not able to "rescue" the mice and homozygous null spd^ot mice with the transgene died about 3 weeks after birth.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Increased levels of GlyR in transgenic mice. [3H]Strychnine binding in mouse brain regions. Wild-type () and transgenic () cortex, and combined spinal cord and brainstem tissue from wild-type () and transgenic () mice were examined. Increases in Bmax were observed in transgenic mouse cortex. KD values were indistinguishable in all tissues expressing significant amounts of GlyR. Inset, immunoblot analysis of membrane fractions by using a polyclonal anti-GlyR antibody on wild-type cortex (wt-c), midbrain (wt-m), cerebellum (wt-b), and spinal cord with brainstem (wt-s) and transgenic cortex (tg-c), midbrain (tg-m), cerebellum (tg-b), and spinal cord with brainstem (tg-s) tissues. Clear increases were observed in transgenic cortex, midbrain, and cerebellum.

View larger version (33K): [in this window] [in a new window]   Fig. 4.   Transgene expression in homozygous (spdot/spdot) mice. Transgenic mice were bred with (spdot) mice to produce (spdot/spdot) mice that possess the transgene. Tissue was harvested at postnatal day 20. Transgene (tg) expression was observed in both spinal cord with brainstem (sp) and cortex (ctx) of (spdot/spdot) mice, who lack the endogenous GlyR 1 subunit. The molecular masses of the GlyR and -tubulin bands were 48 and 55 kDa, respectively.

Transgene expression was also demonstrated in tg3 mice by using [³H]strychnine binding. For mouse cortical, and combined spinal cord and brainstem tissue from transgenic and wild-type mice, the maximal binding (B_max) values were increased in the cortex of transgenic mice, confirming transgene expression (Fig. 3). B_max values were 37 ± 10, 249 ± 26, 599 ± 28, and 536 ± 22 fmol/mg protein for wild-type cortex, transgenic cortex, transgenic spinal cord + brainstem, and wild-type spinal cord + brainstem, respectively (n = 6 animals/group). Using [³H]strychnine binding, approximately equal levels of transgene expression were observed in the cortex of all transgenic families (tg1-tg4) (data not shown). The dissociation constant (K_D) values were indistinguishable between transgenic and wild-type GlyR subunits (Fig. 3). K_D values were 3.3 ± 1, 3.5 ± 0.6, and 3.5 ± 0.5 nM [³H]strychnine for transgenic cortex, transgenic spinal cord + brainstem, and wild-type spinal cord + brainstem, respectively (n = 6 animals/group).

Because strychnine antagonizes and ethanol potentiates GlyR function, ethanol inhibition of strychnine seizures was determined for wild-type and transgenic mice (Fig. 5A). Ethanol (0.5-2.0 g/kg) inhibited strychnine seizures in wild-type mice and this protection was decreased by presence of the transgene. No gender or dose × gender interactions were observed. The odds for transgenic mice to experience seizures at any given dose of ethanol were 3.15 times greater compared with the wild-type mice (P = 0.016, logistic regression, n = 16-19/point). Using standard maximum likelihood methods, ED₅₀ values for ethanol inhibition of strychnine-induced tonic-clonic seizures were estimated to be 0.53 and 0.64 g/kg for wild-type and transgenic mice. Strychnine (0.9 mg/kg) produced tonic-clonic seizures in 100% and caused death in 38% of all mice (n = 21). Ethanol (0.75 g/kg) eliminated both morbidity (n = 32) and (1.0 g/kg) seizures (n = 32).

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Ethanol inhibition of seizures. A, ethanol inhibition of strychnine seizures. Strychnine (0.9 mg/kg) was injected s.c. in male and female mice and mice were observed continuously for 20 min for the appearance of tonic-clonic seizures. Ethanol (0.5-2.0 g/kg) inhibited strychnine seizures in the mice and this protection was decreased by presence of the transgene (P = 0.016). By using logistic regression, the odds for transgenic mice to experience seizures were 3.15 times greater compared with the wild-type mice at any given dose of ethanol (n = 16-19/point). B, ethanol inhibition of pentylenetetrazol seizures. Ethanol (1.00, 1.05, 1.25 g/kg i.p.) inhibited pentylenetetrazol seizures equally well in wild-type and transgenic mice (n = 12-13/point). Data for male and female mice are combined.

To determine whether the transgenic attenuation of the antistrychnine effect of ethanol was nonspecific and shared by other convulsants, ethanol inhibition of tonic-clonic seizures produced by pentylenetetrazol was also evaluated. Pentylenetetrazol (50 mg/kg) caused seizures in 100% of mice (n = 18). Although ethanol inhibited pentylenetetrazol-induced seizures (ED₅₀ = 1.05 g/kg i.p. ethanol for all mice), there were no differences between transgenic and wild-type mice. A dose of 1.25 g/kg ethanol eliminated seizures in 96% of all mice (n = 26) (Fig. 5B).

The ethanol-induced LORR was examined for several doses of ethanol in male and female mice. Ethanol metabolism did not differ between transgenic and wild-type mice (Fig. 6), and male mice showed decreased LORR duration at higher doses of ethanol, which was confirmed using the different families of transgenic mice (Fig. 7). Female wild-type and transgenic mice did not differ in duration of ethanol-induced LORR (data not shown). To determine whether the reduction in ethanol-induced LORR by the transgene was nonspecific, we also measured LORR responses to pentobarbital, ketamine, and flurazepam. The duration of LORR produced by these drugs was not altered in the transgenic mice (Table 1).

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Ethanol metabolism in transgenic and wild-type mice. Ethanol (4.0 g/kg i.p.) was injected and blood samples taken at 15, 60, 120, and 180 min were analyzed for ethanol content. No (transgene × time) interaction was observed between transgenic and wild-type mice by using a two-way analysis of variance (F1,69 = 0.52; P > 0.05), n = 8/group.

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Ethanol-induced LORR. A, duration of LORR was examined at several doses of ethanol (3.4-4.2 g/kg i.p.). A significant transgene (F1,85 = 7.9; P = 0.006) and dose (F2,85 = 40.9; P < 0.0001) effect were observed using a two-way analysis of variance (n = 10-32/point). B, ethanol induced LORR in male mice from the different transgenic lines produced (tg1-tg4). Reductions in the duration of ethanol LORR were observed in three of four lines of transgenic mice. All studies were performed on male mice. No differences were observed between transgenic and wild-type female mice. Values are mean ± S.E.M. (n = 10-36/group). , P < 0.05, , P < 0.01, Student's t tests.

The ethanol ED₅₀ for a LORR greater than 1 min was used as a measure of acute sensitivity to ethanol. The ethanol ED₅₀ for this behavior was higher in male transgenic mice than in wild-type male mice (Fig. 8).

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Ethanol ED50 values for initial LORR. Values are displayed for male transgenic (m-Tg) and wild-type (m-WT) and female transgenic (f-Tg) and wild-type (f-WT) mice. A, ethanol ED50 required to produce an LORR duration of greater than 1 min was greater in male transgenic mice than in wild-type control. B, no differences were observed in ethanol ED50 between transgenic and wild-type female mice. Values are mean ± 95% confidence interval. , significantly different, P < 0.05.

We also determined the ethanol ED₅₀ for inability to remain on a rotarod. For both sexes, transgenic mice required approximately 0.2 g/kg more alcohol than wild-type mice to produce incoordination on the rotarod (Fig. 9). The ED₅₀ of pentobarbital required to produce ataxia on the rotarod did not differ between transgenic and wild-type mice (Fig. 9).

View larger version (28K): [in this window] [in a new window]   Fig. 9.   Ethanol ED50 for rotarod incoordination. Values are displayed for male transgenic (m-Tg) and wild-type (m-WT) and female transgenic (f-Tg) and wild-type (f-WT) mice. A and B, ethanol ED50 values to produce rotarod incoordination were increased in transgenic males and females compared with wild-type littermates. C and D, no differences between transgenic and wild-type mice were observed in pentobarbital ED50 for rotarod incoordination. Values are mean ± 95% confidence interval. , significantly different, P < 0.05.

Ethanol-induced hypothermia was tested with several ethanol doses (3.0-3.8 g/kg, n = 8/group/point). Transgenic and wild-type mice did not differ in the hypothermic response (Fig. 10). Acute functional tolerance to ethanol was also tested for the mice. No differences were observed in the development of acute functional tolerance between the mice (Table 1).

View larger version (23K): [in this window] [in a new window]   Fig. 10.   Ethanol-induced hypothermia. The area under the curve of ethanol-induced hypothermia was determined in transgenic (tg) and wild-type (wt) mice. A, no differences were observed between transgenic and wild-type male mice as tested by a genotype × ethanol concentration interaction (F2,18 = 0.72; P = 0.50). B, no differences were observed between transgenic and wild-type female mice as tested by a genotype × ethanol concentration interaction (F2,18 = 0.07; P = 0.94).

Other basal behaviors were unchanged by presence of the transgene. Wild-type and transgenic mice did not differ in spontaneous locomotion: both groups of mice adapted to the novel test chamber environment at equal rates, and subsequent locomotor activity did not differ (Fig. 11). Transgenic and wild-type mice also did not differ in responses to the elevated plus-maze anxiety test (Fig. 12).

View larger version (28K): [in this window] [in a new window]   Fig. 11.   Spontaneous locomotor activity. A, no (transgene × time) interaction (F23,336 = 0.60; P = 0.93) was observed for female mice (n = 8/group). B, no (transgene × time) interaction (F23,312 = 0.94; P = 0.54) was observed for male mice (n = 7-8/group).

View larger version (28K): [in this window] [in a new window]   Fig. 12.   Elevated plus-maze anxiety test. No differences were observed between wild-type and transgenic mice for the plus-maze anxiety test for the following parameters: percentage of time in the open arms (A), percentage of entries into the open arms (B), total number of entries into the arms (C), and number of entries into the closed arms (D) (n = 10/group, P > 0.05, Student's t tests). Data for male and female mice are combined.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Many in vitro studies have shown that GlyR function is positively modulated by anesthetics and alcohols (Celentano et al., 1988; Engblom and Akerman, 1991; Aguayo and Pancetti, 1994; Mascia et al., 1996a,b; Eggers et al., 2000; Yamakura and Harris, 2000; Ye et al., 2001). These observations have led to molecular studies of ethanol actions on GlyR. For example, mutation of a single amino acid, Ser-267, in transmembrane domain 2 of the GlyR ₁ subunit to isoleucine eliminated all enhancing effects of ethanol on the receptor (Mihic et al., 1997), and the size of the amino acid at position 267 altered the efficacy of n-alcohols (Wick et al., 1998; Ye et al., 1998). Furthermore, mutation of Ser-267 to cysteine allowed propanethiol, an unconventional anesthetic, to irreversibly increase function of the receptor by binding at this site and preventing subsequent potentiation of the receptor by octanol (Mascia et al., 2000). These data provide strong support for the hypothesis that alcohols potentiate GlyR function by binding in a cavity involving Ser-267.

Several behavioral tests have also implicated GlyR as an important target for ethanol action. Glycine and the glycine precursor serine are able to enhance the depressant effects of ethanol as measured by the duration of LORR, and this action was blocked by strychnine (Williams et al., 1995). Ethanol also inhibits strychnine seizures in mice (McSwigan et al., 1984). To combine the molecular and behavioral approaches and to provide a more direct test of the role of GlyR in alcohol action, we used transgenic expression of a mutant subunit.

The GlyR ₁(S267Q) mutant subunit was chosen as the transgene because this mutation eliminates the potentiation of receptor function at low concentrations of ethanol and inhibits receptor function at higher concentrations of ethanol without altering the glycine dose-response relationship. The FVB/NJ inbred mouse line was chosen because it allows a homogenous genetic background between wild-type and transgenic mice and thus avoids problems noted previously for mixed backgrounds (Gerlai, 1996; Crawley, 2000). The Syn promoter was chosen because it should provide expression in all neurons and reaches a maximal expression level on about postnatal day 20, which should prevent complications that can occur with transgene expression during embryogenesis and development (Hoesche et al., 1993). Thus, to address the hypothesis that GlyR potentiation is responsible for some of the acutely intoxicating effects of ethanol we produced a novel transgenic mouse by using the GlyR ₁(S267Q) subunit, which was expressed using the Syn promoter on an inbred (FVB/NJ) mouse background.

Using [³H]strychnine binding, we documented transgene expression. Because GlyR  subunit levels are normally very low in the cortex, which results in little or no detectable binding, increases in binding in the cortex represent transgene expression. Immunoblots confirmed increases in GlyR in the transgenic mice in the cortex, midbrain, and cerebellum. Although no clear increase in spinal cord and brainstem GlyR levels was seen, it is not unreasonable to propose that down-regulation of the endogenous gene occurred to compensate for transgene expression. This is supported by our analysis of transgene expression on a genetic background lacking functional GlyR ₁ subunits. In these mice, we have clear evidence for expression of the transgene in the spinal cord.

No obvious physical differences were observed between the mice, and no differences were observed between wild-type and transgenic mice by using the spontaneous motor activity and the elevated plus-maze anxiety tests. Thus, the impact of overexpression of the transgene in cortex and other regions appears to be minimal. This is likely due to the lack of glycinergic innervation in these areas rendering the transgene receptors functionally silent (Betz et al., 1999).

Because strychnine antagonizes the GlyR, ethanol should inhibit strychnine seizures if it significantly potentiates GlyR function. In agreement with this hypothesis, ethanol eliminated seizures and all morbidity associated with an injection of a CD₁₀₀ dose of strychnine. Although ethanol could be inhibiting strychnine seizures through a target (or targets) separate from GlyR, the simplest explanation for the data is that ethanol potentiation of GlyR function is (largely) responsible for eliminating seizures caused by strychnine inhibition of GlyR function. We hypothesized that if our mutant GlyR subunit was being expressed and effectively incorporated into GlyR then ethanol's ability to eliminate strychnine seizures would be reduced. Consistent with this hypothesis, in transgenic mice a clear right-shift of the dose-response relationship for ethanol inhibition of strychnine seizures was observed. Specificity for the GlyR is suggested by the finding that pentylenetetrazol, a GABA receptor antagonist, induced seizures were inhibited by ethanol equally well in transgenic and wild-type mice.

Because the GlyR is important for motor and spinal function (Betz et al., 1999), we focused on testing behaviors related to the acute effects of ethanol on coordination and motor function. Male transgenic mice slept for a shorter period of time compared with wild-type mice and this effect was observed in three of four families of mice. It is unclear why one transgenic family (tg2) did not display a reduced LORR duration, but the lack of a transgene effect could have been caused by the transgene insertion site. It is also unclear why the transgene effect was not observed in female mice. However, sex differences in ethanol actions on transgenic (Meliska et al., 1995; Rikke et al., 2001) and knockout (Hall et al., 2001) mice were previously observed, and it is possible that the influence of GlyR on this behavior is greater in male mice. Transgenic alteration of LORR duration was limited to ethanol and was not observed for the GABA receptor modulators pentobarbital or flurazepam, or the N-methyl-D-aspartate receptor inhibitor ketamine; thus, the transgene was not merely responsible for a pharmacologically nonspecific change in LORR sensitivity. Because changes in hypothermia may change LORR (Alkana et al., 1988), we tested ethanol-induced hypothermia. No transgenic effect was observed for hypothermia. Male transgenic mice also required a higher dose of ethanol to produce a LORR of greater than 1 min, supporting the idea that GlyRs are responsible for some of the acute anesthetic effects of ethanol.

Additionally, because all transgenic mice required a higher dose of ethanol to produce incoordination on the rotarod (an effect that was not observed for pentobarbital), this supports the idea that ethanol potentiation of GlyR function is also responsible for some of the incoordinating effects of ethanol. No differences were observed in acute functional tolerance by using the rotarod test, suggesting that GlyR may be more important for initial sensitivity than for tolerance development.

It is generally accepted that ethanol sensitivity is a polygenic trait; concerning ethanol-induced LORR, for example, seven loci have been identified (Markel et al., 1997). Assuming equal participation by these loci for an LORR duration of 100 min, one would predict a 14-min change in LORR due to alteration of one loci. We observed a LORR decrease greater than this in the male mice of three of four transgenic families. However, it should be noted that our transgene model may underestimate the importance of GlyR in ethanol actions because of the residual presence of the endogenous alcohol-sensitive GlyR. To investigate this possibility will require new types of mutant mice. For example, construction of knock-in mice with the mutant GlyR subunit would result in complete replacement of the ₁ subunit.

Collectively, our results show a decreased sensitivity to specific acute effects of ethanol in the transgenic mice. This study supports the hypothesis that strychnine-sensitive glycine receptors are an important target for motor incoordinating and anesthetic properties of ethanol action, and provides the first animal model using a receptor point mutation to alter ethanol action.