Norepinephrine strengthens the working memory, behavioral inhibition, and attentional functions of the prefrontal cortex through actions at postsynaptic α2-adrenoceptors (α2-AR). The α2-AR agonist guanfacine enhances prefrontal cortical functions in rats, monkeys, and human beings and ameliorates prefrontal cortical deficits in patients with attention deficit hyperactivity disorder. The present study examined the subtype of α2-AR underlying these beneficial effects. Because there are no selective α2A-AR, α2B-AR, or α2C-AR agonists or antagonists, genetically altered mice were used to identify the molecular target of the action of guanfacine. Mice with a point mutation of the α2A-AR, which serves as a functional knock-out, were compared with wild-type animals and with previously published studies of α2C-AR knock-out mice (Tanila et al., 1999). Mice were adapted to handling on a T maze and trained on either a spatial delayed alternation task that is sensitive to prefrontal cortical damage or a spatial discrimination control task with similar motor and motivational demands but no dependence on prefrontal cortex. The effects of guanfacine on performance of the delayed alternation task were assessed in additional groups of wild-type versus α2A-AR mutant mice. We observed that functional loss of the α2A-AR subtype, unlike knock-out of the α2C-AR subtype, weakened performance of the prefrontal cortical task without affecting learning and resulted in loss of the beneficial response to guanfacine. These data demonstrate the importance of α2A-AR subtype stimulation for the cognitive functions of the prefrontal cortex and identify the molecular substrate for guanfacine and novel therapeutic interventions.
- prefrontal cortex
- attention deficit hyperactivity disorder
- adrenoceptor subtype
- delayed alternation
- neuropsychiatric disorder
The prefrontal cortex (PFC) regulates behavior using working memory, guiding both overt behaviors and the contents of attentional focus, inhibiting inappropriate responses and organizing plans for the future (Goldman-Rakic, 1996;Robbins, 1996). Deficits in PFC function are common in many neuropsychiatric disorders (Weinberger et al., 1986; Barkley et al., 1992; Blumberg et al., 1999) and are a fundamental feature of ailments such as attention deficit hyperactivity disorder (ADHD), in which symptoms of impulsivity correlate with reduced size of the right PFC (Casey et al., 1997).
Recent studies have shown that norepinephrine (NE) has a critical beneficial influence on PFC functions through actions at postsynaptic α2-adrenergic receptors (α2-ARs) (for review, see Avery et al., 2000). For example, systemic administration of α2-AR agonists such as guanfacine (GFC) to rats, monkeys, or human beings enhances performance of PFC tasks (Arnsten et al., 1988; Carlson et al., 1992; Jakala et al., 1999). Similarly, direct infusion of α2-AR agonists into the PFC improves cognitive function, whereas infusions outside the relevant PFC region are ineffective (Tanila et al., 1996; Mao et al., 1999). On the basis of research in animals, guanfacine is currently in use for the treatment of ADHD in children and adults (Scahill et al., 2001; Taylor and Russo, 2001), particularly in patients for whom stimulant medication is contraindicated. However, the α2-AR subtype underlying these important beneficial effects is not known.
Three α2-AR subtypes have been cloned in humans: the α2A-AR, α2B-AR, and α2C-AR, with the α2A-AR and α2C-AR, but not the α2B-AR, subtypes localized in the PFC (for review, see Aoki et al., 1994; MacDonald et al., 1997). Previous studies in monkeys used agonists with varying affinities for the α2A-AR, α2B-AR, and α2C-AR to attempt to dissect the subtype or subtypes underlying the cognitive-enhancing, hypotensive, and sedating actions of these agents (Arnsten et al., 1988; Arnsten and Leslie, 1991). These pharmacological analyses indicated that the cognitive-enhancing effects were mediated by the α2A-AR, whereas the hypotensive and sedating actions were mediated by the α2B-AR and α2C-AR. It is now known that these conclusions were erroneous. Studies of mice with genetically altered α2-AR subtypes demonstrated conclusively that the α2A-AR subtype mediates the hypotensive actions of α2-AR agonists (Link et al., 1996;MacMillan et al., 1996), and the sedative effects of these agents involve α2A-AR actions as well (for review, seeMacDonald et al., 1997). The faulty conclusions from the previous pharmacological studies most likely arose from several factors: (1) there are no highly selective α2A-AR, α2B-AR, or α2C-AR agonists or antagonists to dissociate actions between the subtypes; (2) many α2-AR compounds have nonadrenergic actions as well (e.g., at imidazoline I1 receptors) that can alter blood pressure, sedation, and cognition function (e.g., the potent hypotensive effects of clonidine most likely arise from a combination of α2-AR and I1 receptor actions); and (3) differences in bioavailability can confuse measures of potency (e.g., clonidine enters the brain more quickly than guanfacine). Thus, studies in mice with genetically altered α2-AR subtypes are needed to rigorously determine the α2-AR subtype underlying the cognitive-enhancing effects of NE and α2-AR agonists for the first time. A recent study showed that α2C-AR knock-out mice exhibit normal spatial working memory performance and normal enhancement of performance by an α2-AR agonist (Tanila et al., 1999). The present study performed analyses in mice with a mutation of the α2A-AR subtype (MacMillan et al., 1996).
MATERIALS AND METHODS
α2A-AR mutant mice were compared with wild-type controls in their ability to habituate to handling in a T maze and subsequently to learn and perform a task dependent on the PFC, spatial delayed alternation in a T maze (Larsen and Divac, 1978). This test of spatial working memory requires behavioral inhibition and attentional regulation, particularly as delays are increased. Wild-type and α2-AR mutant mice were also compared for their ability to learn a control task, spatial discrimination in the T maze, which has the same motor and motivational demands but does not depend on the PFC (Boyd and Thomas, 1977). Finally, wild-type and α2-AR mutant mice were assessed on the delayed alternation task for their response to guanfacine, an α2-AR agonist that in vitro prefers the α2A-AR by 10- to 60-fold (Uhlen and Wikberg, 1991; Devedjian et al., 1994), and in vivo improves PFC cognitive function (Arnsten et al., 1988; Jakala et al., 1999).
The present study performed analyses in mice with a point mutation (D79N) of the α2A-AR subtype (MacMillan et al., 1996), which has been shown to effect a functional knock-out of the receptor (Lakhlani et al., 1997). The α2A-AR mutant mice (C57BL/6D79NTG strain) were created by the laboratory of Dr. Lee Limbird (Vanderbilt University, Nashville, TN) and supplied by Dr. Brian Kobilka (Stanford University, Stanford, CA). These animals were shown previously to lack hypotensive α2A-AR mechanisms, although they demonstrated normal blood pressure under basal conditions (MacMillan et al., 1996). The expression of mutated α2A-AR in these mice results in reduction of receptor density by 80% compared with levels in wild-type mice. Furthermore, the residual α2A-AR cannot activate potassium currents or suppress calcium currents. Thus, the D79N α2A-AR mouse serves as a functional knock-out (Lakhlani et al., 1997). At the time that these behavioral studies were initiated, the D79N α2A-AR mice were the only mice altered at the α2A-AR locus to be backcrossed against C57BL/6 mice >12 generations, permitting direct comparison with C57BL/6 wild-type mice; α2A-AR and α2B-AR knock-out mice were not available for examination.
The D79N α2A-AR mice were bred at Yale and were 3–6 months of age at the start of the research. Wild-type C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Animals were purchased at 2 months of age and lived in the Yale University vivarium for 1–5 months before inclusion in the study to be age-matched to mutant animals. Wild-type and mutant strains (males, 19–27 gm) were housed singly, and their food was regulated to ensure motivation on the tasks. Animals were fed a 4 gm biscuit immediately after cognitive testing; water was available ad libitum. Animals showed a normal growth curve. Care of the animals followed the guidelines in the Guide For the Care and Use of Laboratory Animals and was approved by the Yale Animal Care and Use Committee.
Adaptation to testing procedures. Animals were tested in a T maze made smaller (15 × 21 × 3 inches) to be appropriate for testing mice. The maze was constructed of wood and painted black, with a guillotine door separating the start box from the main stem of the maze. Testing occurred in a small room near the colony room under normal light conditions. A sink was located on the wall to the left of the maze, and a dustpan and broom on the wall to the right of the maze; the maze was maintained in the same position in the room throughout the duration of the study. All cognitive testing was performed between 8:00 A.M. and 5:00 P.M. during the animals' light cycle; each animal was tested by the same experimenter at the same time of day (e.g., 11:00 A.M.) every day, Monday through Friday. Before cognitive training, all animals were exposed to the food rewards (halved, peeled sunflower seeds) in their home cage. Importantly, all animals were fully habituated to the T maze and to eating food rewards on the maze (criterion of eating 10 rewards in 8 min for two consecutive days) and were subsequently habituated to handling procedures on the maze (criterion of eating 10 food rewards in 5 min while being picked up five times for two consecutive days). This was done to minimize any affective differences between the mutant and wild-type animals that might obscure changes in cognitive performance. It is common for genetically altered mice to exhibit changes in emotional responding; thus, it was particularly important that all mice achieved the same criterion of comfort with each procedure before continuing with the cognitive aspects of the study.
Cognitive assessment. α2A-AR mutant mice subsequently were compared with wild-type mice on the number of days needed to reach a criterion performance of two consecutive days of ≥90% correct under 0 sec delay conditions for the delayed alternation and spatial discrimination tasks. Half of the animals were trained on delayed alternation first, and the other half were trained on spatial discrimination first to control for any possible order effects. In the spatial delayed alternation task, the animal is rewarded for choosing either the left or right arm on the first trial (not counted), but from then on must always choose the arm not entered on the previous trial. In contrast, in the spatial discrimination task, half the animals are trained to always choose the left arm to receive reward, whereas the other half are trained to always choose the right arm. In both tasks, the animal is picked up immediately after eating the reward (or picked up with no reward if he was incorrect) and placed in the start box between trials. The choice point of the maze is wiped with alcohol to prevent olfactory cues from guiding behavior. This process takes ∼1–2 sec and is designated a 0 sec delay. Each daily test session consisted of 10 trials.
For the assessment of the effects of guanfacine on delayed alternation performance, delays were increased as needed to maintain performance of ∼70% correct, thus allowing room for either improvement or impairment with drug treatment. The delay was increased in 5 sec intervals if a mouse performed at ≥90% for two consecutive testing sessions (rarely, delay was also decreased if baseline deteriorated to <60–70%). Thus, the delay needed to maintain baseline performance at 70% correct can be used as a measure of general performance on the task.
Mice were considered ready for drug administration when they had been at their current delay for at least three test sessions and had performed at between 60 and 70% for two consecutive test sessions. Furthermore, a washout period of at least 7 d was required between drug treatments. Guanfacine was dissolved in saline and was injected (0.15 ml, i.p.) 1 hr before cognitive testing. Animals were extensively adapted to the injection procedure before the experiment. Guanfacine was generously provided by Wyeth-Ayerst (Princeton, NJ). An initial pilot study identified the proper dose range for examination; doses of 0, 0.0001, 0.001, 0.01, 0.1, 1.0, or 10.0 mg/kg were injected in random order in wild-type mice (n = 4–7) and in α2A-AR mutants (see below). Mean percentage correct ± SEM for each dose is shown in Table1. The 0.01 mg/kg dose tended to impair performance in wild-type mice (p > 0.1;n = 6), as has been seen occasionally in monkeys (Arnsten et al., 1988; Franowicz and Arnsten, 1998) and is probably a result of presynaptic drug actions reducing endogenous NE release. The 10 mg/kg dose induced sedation that interfered with testing in wild-type and mutant mice. Only the 1.0 mg/kg dose improved the working memory performance of wild-type mice, and this dose is similar to that which reliably improves working memory performance in young adult monkeys as well (0.5–1.0 mg/kg). Thus, the 1.0 mg/kg dose was compared with saline in a larger number of animals (n = 12) and was repeated to ensure reliability of response within subjects. All doses of guanfacine were tested in a large sample of α2A-AR mutants (n = 10) to ensure that no other dose of guanfacine was effective (Table 1).
Simple comparisons between the wild-type and mutant mice were analyzed with an independent t test (Tind). Delay achieved over time was analyzed by ANOVA with a mixed design: a within-subjects factor of time (25th vs 50th test session) and a between-subjects factor of α2A-AR mutation (wild type vs α2A-AR mutation). Statistical analyses were performed with Systat (Evanston, IL) software.
Habituation to test procedures
The mice were initially adapted to the maze, food rewards, and handling procedures to minimize contamination of affective influences on cognitive assessment. Animals were required to reach criterion levels of responding before continuing to the next level of the study. These procedures are particularly important for mice, given their neophobic response to novel environments and procedures. The α2A-AR mutants were not different from wild-type mice in their time to habituate to eating food rewards in the maze (wild-type mice, n = 6, 4.8 ± 0.9 d to reach criterion; α2A-AR mutant mice,n = 6, 3.3 ± 0.7 d to reach criterion;T ind, p > 0.1); however, they were markedly slower to habituate to handling procedures in the maze (Fig. 1 A) (T ind; p < 0.002). These results are consistent with recent data suggesting that mice with genetically altered α2A-AR can show heightened signs of anxiety (Schramm et al., 2001), whereas in contrast, those with reduced α2C-AR exhibit decreased response to stress (Sallinen et al., 1999).
Acquisition of the cognitive tasks
Once habituated to the maze and handling procedures, the mice were trained in the T maze on either the spatial working memory task, delayed alternation, or the reference memory task, spatial discrimination. There were no significant differences between the α2A-AR mutants and the wild-type controls in the time needed to learn either the delayed alternation task (0 sec delays; T ind, p = 0.92) (Fig. 1 C) or the spatial discrimination task (T ind, p = 0.82) (Fig.1 B) to criterion levels of responding. Thus, learning was unaffected by the mutation of the α2A-AR.
Delay achieved on the delayed alternation task
A second group of animals was used to evaluate the effects of the α2-AR agonist guanfacine on the delayed alternation performance of the α2A-AR mutant versus wild-type mice (n = 12). After habituation to the maze as described above, the mice were trained on the delayed alternation task, and a stable baseline of performance was established for assessment of drug effects on working memory performance. Delays were increased as needed (see Materials and Methods) to maintain performance of ∼70% correct, thus allowing room for either improvement or impairment with guanfacine treatment. Thus, the delay needed to maintain baseline performance at 70% correct can be used as a measure of general performance on the task.
Wild-type mice were able to withstand significantly greater delays than the mutant mice, as shown in Figure 2, comparing mice at the 25th and 50th testing sessions. Although all mice were able to withstand increasing delays as the study progressed (ANOVA; within-subjects effect of time;F (1,22) = 15.21; p = 0.001), there was a highly significant effect of the α2A-AR mutation on the delay needed to maintain baseline performance at 70% correct (between-subjects effect of the α2A-AR mutation;F (1,22) = 27.23; p < 0.0001). The effects of the mutation became increasingly evident over time (significant interaction between the effects of the mutation and time: F (1,22) = 12.57;p = 0.002). Thus, the differences between the mice had already emerged by the 25th test session (mean ± SEM delay needed: wild-type, 5.0 ± 2.2 sec; α2A-AR mutants, 0 ± 0 sec; p < 0.03) but were more evident by the 50th test session (mean ± SEM delay needed: wild-type, 13.7 ± 2.1 sec; α2A-AR, 0.4 ± 0.4 sec; p < 0.00001). Thus, by the 50th test session, no α2A-AR mutant mouse had experienced a delay >5 sec, whereas the majority of wild-type mice were able to show solid baseline performance with delays of 15–20 sec. These data indicate that the α2A-AR mutant mice have measurably poorer performance when working memory abilities are challenged.
Response to the α2-AR agonist guanfacine
The effects of the α2-AR agonist guanfacine were compared in the wild-type and α2A-AR mutant mice after establishment of stable baseline performance. Pilot studies of a wide range of guanfacine doses indicated that the 1.0 mg/kg dose was appropriate for cognitive enhancement in mice. Thus, the 1.0 mg/kg dose was compared with saline in wild-type versus α2A-AR mutant mice. The results are shown in Figure 3. A two-way ANOVA with a between-subjects factor of α2A-AR mutation (wild-type versus α2A-AR mutation) and a within-subjects factor of drug (saline vs guanfacine) found a significant effect of α2A-AR mutation (F (1,19) = 5.73; p = 0.027), no significant overall effect of drug (F (1,19) = 1.39; p = 0.25), and a significant α2A-AR mutation by drug interaction (F (1,19) = 5.021; p = 0.037). Thus, guanfacine significantly improved delayed alternation performance in the wild-type mice (p = 0.04;n = 12) but had no effect on the α2A-AR mutant mice (p = 0.46; n = 12). Wild-type mice were improved by 1.0 mg/kg guanfacine irrespective of whether they received guanfacine early in the study when their delays were short (e.g., 0 sec) or later in the study when their delays were longer. Thus, guanfacine improved performance in WT mice whether their delays were 0 sec (saline 65%; GFC 77%; p = 0.02), 15 sec (saline 74%; GFC 85%;p = 0.02), or 20 sec (saline 68%; GFC 87%;p = 0.05) at the time when they received guanfacine. A similar pattern has been observed in monkeys, in which guanfacine improves spatial working memory performance on trials with either short or long delays (Franowicz and Arnsten, 1998). Other doses of guanfacine were without effect in the α2A-AR mutant mice (Table 1).
Studies of rats, monkeys, and human beings have found that α2-AR agonists, such as guanfacine, improve the cognitive functions of the PFC but have no effect on or impair the cognitive abilities of posterior cortical or subcortical regions. For example, systemic administration of α2-AR agonists to rats, monkeys, or human beings enhances performance of tasks that require working memory, planning, behavioral inhibition, and attention regulation and depend on the integrity of the PFC (Arnsten and Goldman-Rakic, 1985; Jackson and Buccafusco, 1991; Arnsten and Contant, 1992; Carlson et al., 1992; Coull et al., 1995; Rama et al., 1996; Steere and Arnsten, 1997; Jakala et al., 1999; O'Neill et al., 2000). These beneficial effects are dissociable from the hypotensive and sedating actions of these compounds (Arnsten et al., 1988) and are consistent with the known role of NE in attention regulation (Carli et al., 1983). In contrast, the learning and memory abilities of medial temporal lobe (Arnsten and Goldman-Rakic, 1990; Sirviö et al., 1991; Genkova-Papazova et al., 1997) and parietal cortex (Witte and Marrocco, 1997) are unaffected or even impaired by α2-AR stimulation. Thus, the beneficial effects on cognitive function appear to be selective to the PFC.
It is likely that the improved performance induced by guanfacine in the present mouse study similarly reflects enhanced PFC cognitive function. Although guanfacine was administered systemically and thus influenced the entire neuroaxis, studies in rats and monkeys have demonstrated that these beneficial effects of α2-AR agonists arise from direct actions in the PFC. In monkeys, infusion of the α2-AR agonist guanfacine directly into the PFC produces a delay-related improvement in working memory performance (Mao et al., 1999), and similar beneficial effects have been seen with agonist infusion into the PFC of aged rats (Tanila et al., 1996). Conversely, infusion of the α2-AR antagonist yohimbine produces a delay-related impairment in performance (Li and Mei, 1994), emphasizing the importance of endogenous NE stimulation of these receptors. At the cellular level, either systemic or iontophoretic application of an α2-AR agonist increases the delay-related activity of PFC neurons (Li et al., 1999), the electrophysiological measure of working memory and behavioral inhibition (Funahashi et al., 1993). Conversely, the iontophoretic application of an α2-AR antagonist decreases the delay-related activity of PFC neurons (Sawaguchi, 1998; Li et al., 1999), indicating that endogenous NE release has a critical effect on PFC neuronal responding. The importance of α2-AR stimulation to PFC function has been corroborated in imaging studies in which systemic administration of the α2-AR agonist guanfacine increases regional cerebral blood flow in the PFC of both monkeys and human beings (Avery et al., 2000; Swartz et al., 2000). Given the consistency of these findings and the clinical use of guanfacine to treat PFC cognitive disorders, it was of great interest to determine the α2-AR subtype underlying these important beneficial effects of NE and α2-AR agonists on PFC function.
The present study found that mutation of the α2A-AR significantly weakened working memory performance on the delayed alternation task and prevented cognitive enhancement by guanfacine. Mice with mutations of this receptor were not able to achieve the longer delays acquired by wild-type mice, indicating weaker PFC regulation of behavior. Poorer achievement on the delayed alternation task by the mutant mice is unlikely to be accounted for by nonspecific changes in performance or affective variables, because the α2A-AR mutant mice learned the task at a rate similar to that of wild-type mice under 0 sec delay conditions and were able to perform the spatial discrimination task similarly to wild-type animals. These findings are consistent with previous studies in which α2A-AR stimulation did not improve learning (Sirviö et al., 1991). Special care was taken to minimize any affective influences on cognitive assessments, because emotional changes are common in mouse mutants and have been specifically identified in mice with altered α2A-AR (Schramm et al., 2001). Mice were carefully adapted to all potentially stressful procedures to set criteria to ensure that mutants and wild-type animals were equally comfortable with manipulations such as handling. After these adaptive measures, mice were indistinguishable in their behavior in the maze, with the exception that mutant mice were unable to achieve the longer delays acquired by wild-type animals. This profile is consistent with weaker PFC regulation of behavior. Deficits on working memory tasks have been observed in other genetically altered mice, e.g., those with overexpression of the mutant human amyloid precursor (Dodart et al., 1999). However, these working memory deficits occurred in the constellation of broader cognitive deficits, i.e., impaired reference memory and object recognition memory, consistent with widespread cortical and hippocampal plaque deposition (Dodart et al., 1999). In contrast, the present data provide the first evidence of a genetic mutation selectively impairing performance of a working memory task dependent on the PFC and emphasize the importance of α2A-AR stimulation for the strength of working memory function. These findings echo those observed in monkeys with blockade of α2-AR in PFC by local infusions of yohimbine, who exhibited markedly impaired working memory performance at delays as short as 4 or 6 sec (Li and Mei, 1994) and weakened delay-related neuronal activity in the PFC (Sawaguchi, 1998; Li et al., 1999). The present findings identify the molecule in which endogenous NE acts to increase delay-related firing and thus to strengthen PFC regulation of behavior. These findings represent the first, stringent dissociation of the α2-AR subtype contributions to higher cortical function. Because the functions of the PFC are fundamental to many of the highest-order cognitive abilities in humans, identification of a molecule critical to the functional integrity of the PFC is of wide-ranging significance.
Weakened spatial working memory performance in the mutant mice may not be solely a result of loss of a beneficial, postsynaptic α2A-AR substrate for NE actions in the PFC but also of a disinhibition of detrimental factors. For example, the mutant mice also have lost much presynaptic regulation of NE release and most likely have excessive NE release (Trendelenburg et al., 2001). We have shown that such conditions during stress exposure impair working memory via stimulation of α1-AR (Birnbaum et al., 1999). This possibility could be tested in the future by determining whether an α1-AR antagonist improves the performance of mutant mice but not WT animals under nonstress conditions. However, α1-AR may have been downregulated as a consequence of sustained stimulation, in which case this factor may not contribute to working memory impairment. An additional possibility is that loss of the α2A-AR leads to an imbalance between α2A-AR and α2C-AR subtypes that is harmful. Cognitive deficits on non-PFC tasks have been observed in mice overexpressing the α2C-AR subtype (Bjorklund et al., 1998, 2000), and an imbalance between these subtypes may similarly alter PFC function. Finally, it is possible that differences in early environmental rearing conditions could contribute to differences in working memory abilities as adults, but changes of this kind probably would have been expressed in our measures of spatial learning as well (Liu et al., 2000).
The second major finding of this study was that administration of the α2-AR agonist guanfacine significantly improved the working memory performance of wild-type mice but had no effect on the α2A-AR mutant mice. These results contrast with those found in mice with a genetic alteration of the α2C-AR, which showed normal improvement with α2-AR agonist treatment on a spatial working memory task (Tanila et al., 1999). Together, these data demonstrate that stimulation of the α2A-AR subtype underlies the beneficial effects of α2-AR agonists on PFC function.
Guanfacine is currently being used to treat adults and children with ADHD (Chappell et al., 1995; Horrigan and Barnhill, 1995; Hunt et al., 1995; Scahill et al., 2001; Taylor and Russo, 2001), and double-blind, placebo-controlled trials have confirmed that guanfacine treatment enhances performance of PFC tasks in these patients (Scahill et al., 2001; Taylor and Russo, 2001). Preliminary research also suggests that guanfacine may benefit other neuropsychiatric disorders that involve PFC dysfunction, such as posttraumatic stress disorder (Horrigan, 1996) and schizophrenia (Friedman et al., 1999). The data from this study, in combination with previous research in rats, monkeys, and human beings, identifies the α2A-AR as the molecular substrate for the therapeutic actions of guanfacine. However, given that the α2A-AR also mediates many of the hypotensive effects of these agents, it will not be possible to develop α2A-AR agonists devoid of hypotensive side effects.
A rekindling of interest in NE mechanisms underlying neuropsychiatric disorders such as ADHD has begun to emerge in recent years (Biederman and Spencer, 2000). For example, whereas previous research has often emphasized the contribution of dopaminergic mechanisms in the actions of stimulants such as methylphenidate and d-amphetamine, recent data indicate that the lower doses of stimulants used to treat ADHD patients preferentially release NE in rat brain (Kuczenski and Segal, 2001). In concert with this basic finding, selective NE reuptake blockers are now being developed for the treatment of ADHD, and this avenue appears promising (Spencer et al., 1998; Biederman and Spencer, 2000). The data from the present study indicate that the therapeutic actions of these agents probably involve facilitation of endogenous NE stimulation of α2A-AR in the PFC. Thus, either endogenous NE or guanfacine-like compounds stimulate α2A-AR in the PFC, which leads to increased delay-related activity of PFC neurons, which results in stronger PFC regulation of behavior and reduced symptoms of impulsivity, distractibility, and disorganization. These data represent a rare example whereby one can establish a direct connection between highly specified drug actions in the brain and therapeutic efficacy in mental disorders.
This work was supported by United States Public Health Service (USPHS) Grant R37-AG-06036 to A.F.T.A.. The mice were created with the support of USPHS Grant HL-43671 to L.E.L. We thank Tracy White, Lisa Ciavarella, and Sam Johnson for their invaluable technical expertise.
Correspondence should be addressed to Dr. Amy F. T. Arnsten, Section of Neurobiology, Yale University School of Medicine, New Haven, CT 06510. E-mail:.