Benzodiazepines and heightened aggressive behavior in rats: reduction by GABAA/α1 receptor antagonists
- 208 Downloads
Positive modulators of the benzodiazepine/GABAA receptor complex can heighten aggressive behavior; the GABAA/α1 subunit may play a critical role in benzodiazepine-modulated aggressive behavior.
The carboline derivatives, β-CCt and 3-PBC, antagonists with preferential action at the GABAA receptors with α1 subunits, may antagonize benzodiazepine-heightened aggression, thus implicating the α1 subunit in heightened aggression.
The GABAA receptor agonist 4,5,6,7-tetrahydroisoxazolo[5,4c]-pyridin-3-ol (THIP) (0.01–3.0 mg/kg), and the benzodiazepine receptor agonists midazolam (0.3–3.0 mg/kg) and triazolam (0.003–3.0 mg/kg) were administered to adult male resident rats to assess the drugs’ effects on their aggressive behavior toward an intruder. Then β-CCt (0.3–10.0 mg/kg) and 3-PBC (0.3–17.0 mg/kg) were each administered in conjunction with midazolam. The salient elements of aggressive and non-aggressive behavior were measured by analyzing video recordings and encoding each behavioral act and posture in terms of its frequency and duration of occurrence.
Midazolam significantly increased the duration of aggressive behaviors at 1.0 and 1.7 mg/kg, and triazolam increased attack bite frequency at 0.03 mg/kg, both implicating GABAA receptors with benzodiazepine binding sites in aggressive behavior. In the present dose range, THIP did not affect any behaviors. The broad-spectrum benzodiazepine antagonist, flumazenil (1.0 mg/kg), antagonized the aggression-heightening effects of midazolam. β-CCt (0.3–10.0 mg/kg) and 3-PBC (0.3–17.0 mg/kg) also antagonized the aggression-heightening effects of midazolam (1.0 mg/kg).
These results implicate both the GABAA γ and α1 subunits in benzodiazepine-heightened aggression.
KeywordsAggression Benzodiazepines GABAA receptor GABA Motor activity Anxiolytics
Rapid administration of benzodiazepines (BZs) is useful in the control of highly agitated and assaultive patients (Lion et al. 1975; Itil et al. 1978; Bond and Lader 1979; Leventhal and Brodie 1981; Sheard 1984; Pabis and Stanislav 1996). However, some clinical reports document increased hostility or assaultive behavior in humans administered chlordiazepoxide, oxazepam, triazolam, and diazepam (Rickels and Downing 1974; Cherek et al. 1991; Berman and Taylor 1995; Ben-Porath and Taylor 2002). In one study, patients who had taken alprazolam for 8 weeks to treat anxiety reported feeling less aggressive, yet behaved significantly more aggressively in response to provocation (Bond et al. 1995).
Preclinical research has strongly linked the BZ effects on aggression to GABAergic mechanisms, chiefly based on antagonism studies and comparable drug profiles of agonists (Miczek et al. 2002, 2003). In general, drugs that inhibit depolarization at the GABA receptor decrease aggression (Davanzo and Sydow 1979; Haug et al. 1980; Rodgers and DePaulis 1982; Simler et al. 1983; Potegal 1986; Ferreira et al. 2000; Podhorna and Krsiak 2000). The BZ full, partial, and inverse agonists, clobazam, diazepam, zolpidem, zopiclone, 3-carboethoxy-β-carboline, and bentazepam decrease offensive aggression in encounters between male mice (Krsiak 1975; Skolnick et al. 1985; Martin-Lopez and Navarro 1998). At higher doses, the BZ agonists alprazolam and triazolam decrease aggressive displays in male rats encountering an intruder (Plummer III and Holt 1987).
BZs may also increase aggression in laboratory rodents and non-human primates at low doses, similar to the “paradoxical” increases in aggressive outbursts observed in humans (DiMascio 1973). For example, the prototypic BZ, chlordiazepoxide, increases aggression in male rats toward opponents, an effect that has been replicated in male mice and in lactating mice (Miczek 1974; Rodgers and Waters 1985; Weerts et al. 1992; Palanza et al. 1996; Ferrari et al. 1997). Chlordiazepoxide prolongs aggressive encounters between male rats and increases threat displays and aggressive vocalizations in squirrel monkeys (Miczek 1974; Beck and Cooper 1986; Weerts and Miczek 1996). Chlordiazepoxide, alprazolam, triazolam, and diazepam dose-dependently increase attack bites in male rats (Miczek 1974; Plummer III and Holt 1987; Olivier et al. 1991).
GABAA receptors are an important site for mediating these aggression-heightening effects. For example, selectively bred β-carboline-3-carboxylate (β-CCM)-insensitive mice display significantly more aggressive behavior, implying that aggression decreases in the absence of binding at the GABAA BZ site (Guillot et al. 1999). The pro-aggressive effects of low doses of alcohol interact in an additive manner with chlordiazepoxide and allopregnanolone to further increase aggression (Miczek and O’Donnell 1980; Fish et al. 2001). BZ antagonists prevent alcohol-heightened and benzodiazepine-heightened aggression in both rats and squirrel monkeys (Olivier et al. 1991; Weerts et al. 1993). One strain of mice bred for high aggression shows decreased muscimol-activated chloride flux in the hippocampus, hypothalamus, and cerebral cortex (Weerts et al. 1992).
The ionotropic GABAA receptor consists of one or more of seven families of polypeptide subunits, each with multiple isoforms (Rudolph et al. 2001). Classic BZs, such as diazepam, interact with all BZ-sensitive GABAA receptor subunits (α1, α2, α3, and α5) with similar affinities (Mohler and Okada 1977). Pharmacological studies suggest that the α1 subunit of the GABAA receptor complex is involved in the anxiolytic and sedative effects of BZs, although point mutation research is not consistent with these results and implicates further subunits (McKernan and whiting 1996; Rudolph et al. 1999; Vekovischeva et al. 1999). β-CCt and 3-PBC are antagonists with 10- to 20-fold selectivity for the GABAA/α1 receptor versus other subunits (Cox et al. 1995; Huang et al. 2000). In behavioral studies, β-CCt effectively blocks several anti-anxiety effects of chlordiazepoxide, diazepam, and triazolam, implicating the α1 subunit in the anxiolytic effects of BZs (Belzung et al. 2000; Rowlett et al. 2001). β-CCt also antagonizes alcohol-heightened and species-typical aggression in mice (de Almeida et al. 2004). Though structurally similar to β-CCt, the behavioral profile of 3-PBC remains unexplored in rodents; only limited research has been conducted in squirrel monkeys (Harvey et al. 2002; Lelas et al. 2002; Rowlett et al. 2003). Based on the anti-aggressive effects of β-CCt in mice, it may be extrapolated that 3-PBC will be an effective agent in reducing aggression.
The present experiment explored the potentially aggression-heightening effects of the GABAA agonist THIP, as well as the BZ agonists, midazolam and triazolam. The extent to which the GABAA/α1-preferring antagonists reduced aggressive behavior points to a possible role of the α1 subunit in aggression at species-typical levels and when heightened by benzodiazepine treatment.
Materials and methods
Subjects were Long–Evans hooded rats (Charles River Breeding Laboratories, Wilmington, Mass., USA). Each of nine male rats, initially weighing 500–650 g, were housed as residents with a female rat in a large stainless steel cage (75×45×45 cm3) containing pine shaving bedding and a clear acrylic front panel. A wooden panel and a wooden structure provided cover and gnawing material. Eight-week-old male intruder rats weighed 300–350 g. All animals were allowed unrestricted access to food and water except during experimental trials. Animals were housed in a vivarium with a controlled photocycle (12:12-h, lights off at 0800 hours), and the temperature was controlled at 21±1° with 40–50% humidity. Experimental and housing procedures were carried out in accordance with the NIH Guide (NIH publication No. 85-23, revised 1996) for the care and use of laboratory animals. All procedures were approved by the Institutional Animal Care Committee of Tufts University.
Aggression by each resident rat towards an intruder was evaluated before initiating drug testing in confrontations between residents and a smaller, socially naive intruder male 4 weeks after the residents were housed with females as breeding pairs. The female rat, wooden structure, water bottle, and pups were removed from the cage. A male intruder rat was introduced into the resident’s home cage, and the confrontation was terminated 5 min after the resident’s first attack bite (Miczek 1979). If no bite occurred after 5 min, the intruder was removed. Confrontations occurred at least once a week for 3 weeks to establish a stable baseline of aggression. Animals that attacked intruders during three consecutive confrontations were included in the experiments. Nine of the ten animals screened for aggressive behavior reliably attacked intruders before the experiments began and were included in the studies. After the session, females and water bottles were returned to the resident home cages. Intruder animals were used at most twice and then killed by CO2. Experiments that included drug treatments followed the same procedure and were conducted no more than 4 times per week between 1100 and 1600 hours.
Apparatus and measurements
The behavior of rats during agonistic interactions was recorded with an infrared-sensitive video camera (Canon model IR-20W) for later analysis. Trained observers used a commercial data acquisition program (“The Observer”; Noldus, Waggening, The Netherlands). The frequency and duration of each behavior were coded by depressing a specific key on a specialized keyboard, similar to that previously described (Miczek 1982). Aggressive behaviors were operationally defined as attack bites, aggressive posture, sideways threat, and pursuit. Time spent walking, rearing, autogrooming, and digging represented non-aggressive motor behaviors.
THIP (4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridine-3-ol) (0.03, 0.3, 3.0 mg/kg) (Sigma-Aldrich, St Louis, Mo., USA) and midazolam maleate (0.3, 0.56, 1.0, 1.7, 3.0 mg/kg) (Hofmann La Roche Pharmaceuticals) were prepared daily in a saline vehicle and administered 30 min before the resident-intruder confrontation. Triazolam (0.003, 0.01, 0.3 mg/kg) was prepared daily in a 17% propylene glycol, 2% Tween 80 solution and administered 10 min before the introduction of an intruder. Each drug administration was counterbalanced with a vehicle administration.
Flumazenil (1.0 mg/kg) was prepared daily in a 15% propylene glycol, 2% Tween 80 solution. β-CCt (β-carboline-3-carboxylate-t-butyl ester) (1.0, 3.0, 10.0 mg/kg) and 3-PBC (3-propoxy-β-carboline) (1.0, 3.0, 10.0, and 17.0 mg/kg) were prepared within 48 h of administration in a 14% propylene glycol, 1% Tween 80 solution. All antagonists were injected in combination with midazolam 30 min prior to the introduction of an intruder, with the exception of one dose each of 1.0 mg/kg flumazenil, 10.0 mg/kg β-CCt, and 17.0 mg/kg 3-PBC, administered to explore the effects of the antagonists on aggressive behavior in the absence of midazolam. Antagonists were synthesized in Dr. Cook’s laboratory. All drugs were injected intraperitoneally in a volume of 1.0 ml/kg. With one exception, all rats received each drug in the following order: THIP, triazolam, midazolam, flumazenil, β-CCt, 3-PBC. Drug doses were counterbalanced between subjects, and a 1-week wash-out period between drugs was allowed. One animal was excluded from the antagonist studies due to a complete and sudden cessation of aggressive behavior, even in the absence of injection. The wash-out period and the within-subjects experimental design appeared appropriate, as aggressive behavior after intermittent saline injection remained stable throughout the series of experiments and reflected values observed during pre-drug screening assessments.
Behavior durations and frequencies were analyzed using one-way repeated measures analyses of variance (ANOVA). When a statistically significant effect was present with α<0.05, Bonferroni t-tests compared individual drug doses to vehicle baseline for post hoc comparisons. The multiple vehicle values were averaged after one-way repeated measure ANOVAs revealed no differences across administrations.
Effects of GABAA agonists on aggressive behavior
Effects of antagonists on midazolam-heightened aggressive behavior
The current findings confirm the critical role of the GABAA BZ receptor complex in both heightening and reducing aggressive behavior. THIP, a broad-spectrum GABAA agonist, given systemically, affected aggressive behavior across a wide range of doses in a modest, mostly insignificant manner as previously reported (Rudissaar et al. 2000 report decreases at higher doses). One study suggests that THIP may increase aggression in non-aggressive rats only (DePaulis and Vergnes 1983). In this study, THIP was administered directly into the ventricles of rats that showed no aggression towards mice during prolonged test periods. THIP heightened aggression in these “non-killers.” It is noteworthy, however, that THIP attenuated killing in rats that reliably attacked mice in this study. It is difficult to make comparisons between this study and the one presented here because we administered THIP systemically to rats that displayed stable baseline levels of aggression toward other adult male rats. Overall, THIP produces only modest behavioral effects, reinforcing the value of highly specific pharmacological compounds, as opposed to broad-spectrum compounds like THIP, in deciphering the modulatory effects of GABA compounds on aggressive behavior.
In this study, the BZ agonist, midazolam, a positive modulator at the GABAA receptor, significantly increased aggressive behavior towards an intruder, confirming this drug’s aggression-heightening effects as described in aggressive resident mice (Waters and Rodgers 1984). Triazolam increased the frequency of attack bites, as previously reported (Plummer III and Holt 1987), and the BZ receptor antagonist flumazenil antagonized agonist-heightened aggressive behavior, further implicating the BZ receptor in aggressive behavior and confirming similar findings in rats, monkeys, and humans (Weerts et al. 1993; Bond et al. 1995; Weisman et al. 1998).
Prototypic BZs produce bidirectional dose-dependent effects on rodent aggressive behavior; specifically, most BZs heighten aggressive behavior at low doses and sedate at higher doses (Miczek 1974; Miczek and O’Donnell 1980; Miczek et al. 1995, 2002). However, while the moderately long-acting BZ, midazolam, increased the duration of various aggressive behaviors, it did not affect the frequency of attack bites, suggesting midazolam lengthened the sequence of aggressive behaviors that precede and follow an attack, but did not increase the frequency of attacks. In support of this hypothetical scenario, the duration of walking decreased, but the duration of lying remained unchanged at the aggression-heightening dose of 1.0 mg/kg, suggesting that, instead of engaging in exploratory behaviors, the rats spent an increased amount of time threatening intruders. Conversely, triazolam increased the frequency of attack bites but did not affect the duration of aggressive postures. Triazolam thus induced an unusual pattern of behaviors, as flurries of attack bites and pursuits generally follow bouts of sideways threats and aggressive postures. An increase in attack bites often predicts an increase in aggressive postures, but such effects were not observed here. Triazolam is a short-acting BZ with a high preference for GABAA/BZ receptors containing α1, 2, 3, and 5 subunits (Huang et al. 2000). Triazolam also has a higher intrinsic efficacy, as measured by chloride ion flux, than the long-lasting BZ, diazepam, which suggests that triazolam’s historically inconsistent and unusual effects on aggressive behavior may be related to the drug’s pharmacokinetic characteristics (Ducic et al. 1993). Alternatively, further analysis of the midazolam and triazolam time courses might reveal that the BZs actually induce a similar profile of aggressive behaviors characterized by a distinct period of prolonged aggressive postures followed by a robust increase in attack bites.
As evidenced here, positive modulators at the GABAA receptor do not consistently increase the same elements of aggressive behavior, and individual differences likely contribute to statistically non-significant behavioral trends. Individual differences in the aggression-heightening effects of alcohol, another positive modulator of the GABAA receptor, are well documented (e.g., Miczek and Barry 1977; Blanchard et al. 1987; Lister and Hilakivi 1988; Miczek et al. 1992; van Erp and Miczek 1997; Miczek et al. 1998; Fish et al. 2001). Several BZs do not exert aggression-heightening effects at all: even at low doses, oxazepam, clorazepam, and zolpidem consistently produce anti-aggressive effects (Bond and Lader 1988; Martin-Lopez and Navarro 2002; de Almeida et al. 2004), and triazolam fails to increase aggressive behavior in various experimental protocols (Cherek et al. 1991; Kruk 1991; de Almeida et al. 2004).
Despite their dissimilar effects, BZs likely affect aggressive behavior via action on the GABAA receptors that inhibit 5-HT impulse flow in a circuit that originates in the raphe nucleus and projects to the ventral tegmental area, nucleus accumbens, prefrontal cortex, amygdala, and hippocampus. Within the dorsal raphe nucleus (DRN) and the surrounding periaqueductal grey, GABAergic neurons make monosynaptic connections with serotonergic neurons, and pharmacological studies demonstrate that, in general, GABAA agonists decrease, while antagonists increase, extracellular 5-HT (Nishikawa and Scatton 1986). For example, the GABAA receptor blocker bicuculline and the GABAA antagonist picrotoxin produce increases in extracellular 5-HT when directly microinjected into the rat DRN (Tao et al. 1996; Tao and Auerbach 2000). Conversely, GABAA receptor agonists inhibit 5-HT activity in the DRN (Pan and Williams 1989). 5-HT release appears to directly influence 5-HT’s anti-aggressive effects: in vivo monitoring of extracellular 5-HT during aggressive attacks in rats showed a decrease in cortical 5-HT following the initiation and termination of the aggressive attacks (van Erp and Miczek 2000; Ferrari et al. 2003). In the present study, midazolam and triazolam may be enhancing aggressive behavior by increasing GABA inhibition of 5-HT projection neurons in the DRN, and our antagonist data suggest that activation of the GABAA receptors containing γ and α1 subunits are particularly critical.
The GABAA α1 subunit may mediate aggressive behavior
When administered in combination with midazolam, the GABAA/α1-preferring β-carboline, β-CCt, systematically and robustly decreased midazolam-heightened aggressive behavior, extending similar effects found in the mouse, in which β-CCt was 5 times more potent than flumazenil in antagonizing alcohol-heightened aggressive behavior (de Almeida et al. 2004). Another β-carboline, 3-PBC, antagonized aggressive behavior at even lower doses than β-CCt. Although the behavioral effects of 3-PBC have been explored in non-human primates, particularly in antagonizing the discriminative and rewarding effects of alcohol, these data represent the first effort to behaviorally characterize the compound in rats (Harvey et al. 2002; Lelas et al. 2002; Rowlett et al. 2003).
β-CCt and 3-PBC also effectively antagonized decreases in the duration of walking characteristically induced by midazolam, and β-CCt had no intrinsic effects on locomotion, confirming previous observations (Löw et al. 2000; June et al. 2003). These data also reinforce evidence from GABAA/α1 knockout mice that suggests the α1 subunit mediates the sedative effects of GABAA agonists (Rudolf et al. 1999; McKernan et al. 2000). Why 3-PBC alone decreased durations of both walking and aggressive behavior is unclear, although like β-CCt, this compound displays moderate affinity for the GABAA α2 and α3 subunits where it acts as an inverse agonist (Huang et al. 2000). The decreases in walking and aggressive behaviors observed at the highest 3-PBC dose tested may thus reflect GABA-mediated sedation; it seems unlikely that lower doses would induce the same decreases. Once more fully characterized, 3-PBC may prove to be an important tool with which to explore the α1 subunit, despite its affinity for α2 and α3 subunits.
β-CCt has been extensively studied at both the molecular and behavioral level because of its selectively for the GABAA α1 subunit, a more than 10-fold selectivity over the α2 and α3 subunits and a more than 110-fold selectivity for the α1 over the α5 subtype (Cox et al. 1995). A 10-fold selectivity may appear rather modest, but considering that the α1 subunit appears in more than 50% of GABAA receptors, even a small degree of selectivity translates to a significant degree of receptor occupancy. β-CCt also binds competitively with BZ agonists over a range of doses in various in vivo and in vitro assays (Shannon et al. 1984; Carroll et al. 1991; Cox et al. 1998; Griebel et al. 1999; Paronis et al. 2001; Rowlett et al. 2001). It is not surprising that this mixed inverse agonist/antagonist produces varied results according to experimental preparation. For example, it acts as an anxiolytic in the plus maze but antagonizes diazepam’s anxiolytic effects in a light/dark test without affecting diazepam’s ataxic, relaxant, or nonspecific motor suppression effects (Carroll et al. 1991; Griebel et al. 1999).
It may be tempting to attribute BZs’ anxiolytic and disinhibitory effects to the α2 subunit and the sedative effects to the α1 subunit, as was suggested earlier. However, recent research with subunit-specific pharmacological compounds demonstrates that such a compartmentalization over-simplifies the roles the α1 subunit plays in mediating the anxiolytic and disinhibitory effects of BZs. For example, β-CCt potentiates anticonflict responses produced by the GABAA α1-preferring compounds, zolpidem and zaleplon, in non-human primates (Paronis et al. 2001). However, β-CCt also antagonizes the anticonflict responses produced by midazolam, suggesting that this β-carboline acts differentially at GABA receptors that contain both the α1 and γ receptor subunits. Both the potentiation and antagonism effects on anticonflict behavior demonstrate that the α1 subunit modulates more than simply GABAA receptor-mediated sedation. It should also be noted that the 30 mg/kg dose used in these experiments preferentially occupies GABAA receptors with α1 subunits in the cerebellum.
Work with α1 subunit agonists has further confirmed the α1 subunit’s role in disinhibitory and anxiolytic effects of BZ compounds: zolpidem has disinhibitory effects equivalent to midazolam, chlordiazepoxide, clonazepam, and triazolam on punishment-suppressed behavior in rats and squirrel monkeys (Depoortere et al. 1986; Vanover et al. 1999; Lobarinas and Falk 2000; Paronis et al. 2001). Further, zaleplon increases punished drinking (but not shock-suppressed behaviors) in rats (Sanger 1995; Lobarinas and Falk 2000). However, it should be noted that zolpidem and zaleplon can produce sedative effects at the same doses that increase punished responding (Depoortere et al. 1986; Sanger and Zivkovic 1988; Sanger 1995).
Although β-CCt produces complex and sometimes divergent effects on behaviors related to anxiety, its antagonistic effects on other behaviors mediated by the GABAA receptor are provocative. For example, the GABAA agonist THIP increases alcohol consumption in rats, and although there has been no direct link between the reinforcing properties of alcohol and the α1 subunit, β-CCt has been proposed as a possible pharmacotherapeutic agent to reduce alcohol drinking in humans because of its suppression of alcohol-reinforced behaviors in alcohol-preferring rats when microinjected into the ventral pallidum (VP) or central nucleus of the amygdala (June et al. 2003; Foster et al. 2004). In the VP, flumazenil failed to alter EtOH-maintained responding, suggesting these behaviors are motivated by highly α1-specific mechanisms. In support of this hypothesis, it should be noted that GABAA/α1 subunit knockout mice will not work for alcohol reinforcements compared to wild-type animals (June et al. 2003).
That β-CCt antagonizes the reinforcing effects of alcohol and the aggression-heightening effects of both alcohol and BZs in diverse experimental preparations suggests that underlying neural mechanisms may be related and linked by GABAA receptors with both α1 and γ subunits. The antagonists in this experiment also restored walking frequency that was suppressed by midazolam. In sum, these data suggest a lack of differentiation between behaviorally excitatory and inhibitory effects at these receptors. Further studies with pharmacological or gene-targeting tools need to identify the distinct molecular mechanisms involved in heightening and reducing aggressive behavior within the BZ/GABAA complex and how these mechanisms translate into the modulation of relevant GABAergic synaptic connections.
We acknowledge support by USPHS research grants AA13983, DA02632, and grants from the Alcoholic Beverage Medical Research Foundation. We are grateful to Mr. J. Thomas Sopko and Ms. Sara Faccidomo who provided outstanding support.
- Bond A, Lader M (1979) Benzodiazepines and aggression. In: Sandler M (ed) Psychopharmacology of aggression. Plenum, New York, pp 173–182Google Scholar
- Carroll M, Woods JE II, Seyoum RA, June HL (1991) The role of GABAA α1 subunit in mediating the sedative and anxiolytic properties of benzodiazepines. Alcohol Clin Exp Res 25:12AGoogle Scholar
- Cox ED, Hagen TJ, McKernan RM, Cook JM (1995) BZ1 receptor subtype specific ligands: synthesis and biological properties of β-CCt, a BZ1 receptor subtype specific antagonist. Med Chem Res 5:710–718Google Scholar
- Cox ED, Diaz-Arauzo H, Huang Q, Reddy MS, Ma CR, Harris B, McKernan R, Skolnick P, Cook JM (1998) Synthesis and evaluation of analogues of the partial agonist 6-(propyloxy)-4-(methoxymethyl)-β-carboline-3-carboxylic acid ethyl ester (6-PBC) and the full agonist 6-(benzyloxy)-4-(methoxymethyl)-β-carboline-3-carboxylic acid ethyl ester (Zk 93423) at wild type and recombinant GABAA receptors. J Med Chem 41:2537–2552CrossRefPubMedGoogle Scholar
- Itil TM, Seaman PA, Huque M, Mukhopadhyay S, Blasucci D, Nq KT, Ciccone PE (1978) Clinical and quantitative EEG effects and plasma-levels of fenobam (MCN-3377) in subjects with anxiety—open rising dose tolerance and efficacy study. Curr Ther Res Clin Exp 24:708–724Google Scholar
- June HL, Foster KL, McKay PF, Seyoum R, Woods JE, Harvey SC, Eiler WJA, Grey C, Carroll MR, McCane S, Jones CM, Yin WY, Mason D, Cummings R, Garcia M, Ma CR, Sarma PVVS, Cook JM, Skolnick P (2003) The reinforcing properties of alcohol are mediated by GABAA1 receptors in the ventral pallidum. Neuropsychopharmacology 28:2124–2137PubMedGoogle Scholar
- Leventhal BL, Brodie HKH (1981) The pharmacology of violence. In: Hamburg DA (ed) Biobehavioral aspects of aggression. Alan R. Liss, New York, pp 85–106Google Scholar
- Martin-Lopez M, Navarro JF (1998) Behavioural profile of bentazepam, an anxiolytic benzodiazepine, in social encounters between male mice. Med Sci Res 26:335–337Google Scholar
- McKernan RM, Rosh TW, Reynolds DS, Sure C, Wafford KA, Attack JR, Farrar S, Myers J, Cook G, Ferris P, Garrett L, Bristow L, Marshall G, Macaulay A, Brown N, Howell O, Moore KW, Carling RW, Street LJ, Castro JL, Ragan CI, Dawson GR, Whiting PJ (2000) Sedative but not anxiolytic properties of benzodiazepines are mediated by the GABAA receptor alpha1 subtype. Nat Neurosci 3:587–592CrossRefPubMedGoogle Scholar
- Miczek KA (1982) Ethological analysis of drug action on aggression defense and defeat. In: Spiegelstain MY, Levy A (eds) Behavioral models and the analysis of drug action. Elsevier, Amsterdam, pp 225–239Google Scholar
- Miczek KA, Fish EW, DeBold JF, de Almeida RMM (2002) Social and neural determinants of aggressive behavior: pharmacotherapeutic targets at serotonin, dopamine and gamma-aminobutyric acid systems. Psychopharmacology 163:434–458Google Scholar
- Plummer III HK, Holt IV (1987) Effects of alprazolam and triazolam on isolation-induced aggression in rats. Ohio J Sci 87:107–111Google Scholar
- Rudolph U, Crestani F, Möhler H (2001) GABAA receptor subtypes: dissecting their pharmacological functions. Trends Pharmacol Sci 22:188–194Google Scholar
- Shannon HE, Guzman F, Cook JM (1984) β-Carboline-3-carboxylate-t-butyl ester: a selective BZ1 benzodiazepine receptor antagonist. Life Sci 35:2227–2236Google Scholar
- Waters AJ, Rodgers RJ (1984) Differential effects of chlordiazepoxide and midazolam on agonistic behavior in male albino mice. Aggress Behav 10:177–178Google Scholar