- Top of page
- Materials and methods
Attention deficit hyperactivity disorder (ADHD) is the most commonly diagnosed childhood psychiatric disorder. We have found that a transgenic mouse bearing a human mutant thyroid receptor (TRβ1) expresses all of the defining symptoms of ADHD—inattention, hyperactivity, and impulsivity—as well as a ‘paradoxical’ response to methylphenidate (MPH). As with ADHD, the behavioral phenotypes expressed by the TRβ transgenic mice are dynamic and sensitive to changes in environmental conditions, stress, and reinforcement. TRβ transgenic mice are euthyroid except for a brief period during postnatal development, but the behavioral phenotypes, elevated dopamine turnover, and paradoxical response to MPH persist into adulthood. Thus, like the vast majority of children with ADHD, the TRβ transgenic mice exhibit the symptoms of ADHD in the complete absence of thyroid abnormalities. This suggests that even transient perturbations in developmental thyroid homeostasis can have long-lasting behavioral and cognitive consequences, including producing the full spectrum of symptoms of ADHD.
Abnormal thyroid hormone levels during gestation can have profound effects on brain development and cognition (Bernal 2002; Thompson & Potter 2000) Neurodevelopmental processes such as cellular differentiation, neurite outgrowth, synaptogenesis, and myelination all depend on proper thyroid homeostasis (Konig & Moura Neto 2002). Normal development of the monoaminergic and cholinergic neurotransmitter systems are also thyroid dependent, and dysfunctions in these transmitter systems have been linked to attention deficits and hyperactivity (Evans et al. 1999; Jin et al. 2001; Oades & Muller 1997; Rastogi & Singhal 1976). It has long been known that cretinism, mental retardation, and other debilitating conditions can result from abnormal thyroid synthesis, action, or metabolism during development. However, the consequences of transient, subclinical thyroid abnormalities are just beginning to be appreciated (Klein et al. 2001; Pop et al. 1995; 2003; Smit et al. 2000 Surks et al. 2004).
In the pituitary gland, the β thyroid receptors (TRβ) are responsible for downregulating thyroid stimulating hormone (TSH), an action that results in decreased production of thyroid hormones thyroxine (T4) and triiodothyronine (T3). A number of mutations in the thyroid receptor β (Thrb) gene have been identified, most of which are in the ligand-binding domain of the TRβ receptor and act to impair the ability of the receptors to bind T3 (Jameson 1994; Yen 2003; Yen et al. 2003). With functional receptors, the mutant receptors form homo- and heterodimers that lack the ability to act at genomic response elements. The result of this dominant negative action is a failure to downregulate TSH. This condition, known as resistance to thyroid hormone (RTH), is a heritable syndrome that typically results in elevated thyroid hormones, normal or elevated TSH, short stature, hearing loss, and tachycardia (Beck-Peccoz et al. 1992; Weiss & Refetoff 2000). In addition, 70% of children with RTH syndrome have attention deficit hyperactivity disorder (ADHD) compared with an incidence of 3–5% in the population as a whole (Burd et al. 2003; National Institutes of Health 1998). Although RTH syndrome is rare, the strikingly high number of RTH patients that exhibit the full spectrum of ADHD symptoms suggests that common mechanisms downstream of the TRβ receptor may be responsible for manifestation of the behavioral phenotypes in both disorders.
Several mouse lines with mutations in the Thrb gene have been created. Knockout mice lacking the Thrb gene exhibit elevated TSH and free and total T3 and T4 but normal behavior (Forrest et al. 1996). Transgenic mice bearing the human PV mutant Thrb gene have been developed (Wong et al. 1997). The PV mutant TRβ1 was derived from a patient (PV) with a mutation in the ligand-binding domain of the Thrb gene, and RTH syndrome characterized by short stature, low body weight, and ADHD (Mixson et al. 1992; Parrilla et al. 1991; Wong et al. 1997). The mice were created using a β-actin promoter to induce ubiquitous expression of the mutant receptor. Consistent with a thyroid resistance phenotype, thyroid hormone levels in the TRβ(β-actin) transgenic mice are increased by approximately 50%, and TSH levels are inappropriately normal. Behaviorally, male TRβ (β-actin) transgenic mice are hyperactive relative to wild-type controls but have normal attention and are not impulsive (McDonald et al. 1998).
In order to better understand the role of thyroid resistance in mediating the behavioral features of ADHD, we tested a transgenic mouse that harbors the PV mutant Thrb gene (Zhu et al. 1999). In this mouse, mutant receptor expression is limited to the pituitary using the mouse glycoprotein hormone α-subunit (αGSU) promoter. The ratio of functional, endogenous receptors to non-functional, mutant receptors in the TRβ(αGSU) transgenic mice is 1:1 (Zhu et al. 1999). Thus, transgenic and wild-type control mice are identical except for 50% of the TRβ1 receptors in the pituitary. However, despite the robust expression of mutant transgene, TRβ(αGSU) transgenic mice exhibit normal thyroid levels as adults (Zhu et al. 1999). The present experiments were designed to test the hypothesis that transient thyroid resistance during development can result in persistent behavioral and neurochemical abnormalities in adulthood, in the TRβ(αGSU) transgenic mice.
- Top of page
- Materials and methods
Although the thyroid resistance phenotype in the TRβ transgenics was detected only at 33 days of age, behavioral abnormalities were evident in euthyroid, 56-day-old mice, and persisted into adulthood. Thus all of the demonstrated behavioral phenotypes in the TRβ transgenic mice occurred in the presence of normal levels of T3, T4, and TSH. This suggests that the transient thyroid resistance observed in 33-day-old transgenics resulted in permanent changes in the brain. Thyroid hormone directly regulates the development of several brain systems associated with attention, locomotor activity, motivation, and impulsive behavior. For example, genes that code for proteins involved in myelination and the development of cholinergic, dopaminergic, and noradrenergic neurotransmitter systems are all regulated by thyroid hormone. The attenuated response to methylphenidate and increase in DA turnover in the TRβ transgenic mice suggests that these permanent brain changes involve the catecholaminergic system. This is consistent with data showing that rats made transiently hyperthyroid as pups, but not as adults, are hyperactive and exhibit elevated DA turnover (Rastogi & Singhal 1976). A number of catecholaminergic receptors and transporters have been implicated in hyperactivity and attentional deficits, including the DA transporter, DA D4 receptor, and the α2 adrenergic receptor (Giros et al. 1996; Clifford et al. 1998; Arnsten 2000). Methylphenidate inhibits the DA and NE transporters, resulting in increased neurotransmitter concentrations in catecholaminergic synapses. Considerable evidence suggests that DA turnover rates are associated with DA transporter function. The DA transporter knockout and knockdown mice both exhibit hyperactivity, reduced DA uptake, and elevated DA turnover (Jones et al. 1998; Zhuang et al. 2001). Other perturbations that reduce DA uptake also increase turnover, such as aging, fatigue, or exposure to environmental toxins such as the ubiquitous polychlorinated biphenyls (PCBs) or the organophosphate insecticide chlorpyrifos (Karen et al. 2001; Haycock et al. 2003; Mizokawa et al. 2003; Richardson & Miller 2004).
Interestingly, locomotor hyperactivity in male transgenic mice was only evident after repeated sessions or after a saline injection in an initial activity session. The fact that multiple sessions were required before genotype differences emerged suggests that male transgenic mice were only hyperactive in a familiar environment. The asymptotic locomotor activity of the male transgenics increased with repeated dark-cycle sessions, suggesting that the hyperactivity was more than a simple failure to habituate. Importantly, the hyperactivity exhibited by the male TRβ transgenics was an increase in normal murine activity in a familiar environment, with none of the repeated circling, head-bobbing, or stereotypic behaviors characteristic of many animal models of hyperactive syndromes (Giros et al. 1996; Wilson 2000). Naïve male transgenics were only hyperactive after a saline injection (Fig. 3a). It is difficult to know to what extent this putative stress-induced hyperactivity interacted with the attenuated response to methylphenidate. If the male transgenics had not been hyperactive relative to wild-type controls under saline, the methylphenidate response would have been in the same direction as that of the wild-type mice, albeit a smaller response similar to what was observed in the female mice. Alternately, if injection stress was the source of behavioral change it is plausible to assume that both saline and methylphenidate groups were affected. Whether injection stress is manifest similarly under saline and methylphenidate is not known. Current investigations examining home-cage running-wheel activity with methylphenidate administered through the drinking water should address some of these concerns. This situational hyperactivity in the TRβ transgenic mice is consistent with expression of attention deficits and hyperactivity in humans and in animal models, which are exhibited primarily in familiar environments and are exacerbated under stressful conditions (American Psychiatric Association 1994; National Institutes of Health 1998; Drolet et al. 2002; Carboni et al. 2003).
Transgenic mice performed as well as wild-type mice under the high reinforcement conditions (50% Ensure) in the reaction-time task. When the concentration of the reinforcer was reduced to 10%, performance of the male transgenic mice deteriorated significantly more than that of wild-type controls. This is consistent with the idea that ADHD-related phenotypes in humans and non-humans are particularly sensitive to changes in reinforcement (American Psychiatric Association 1994; Douglas & Parry 1983,1994; Sagvolden et al. 1993a, 1998). Although this may be conceived of as a motivational deficit (Haenlein & Caul 1987; Slusarek et al. 2001), our progressive ratio data suggest that there was no genotype difference in motivation at the lower reinforcer magnitude. In fact, the transgenic mice had higher progressive ratio break points when working for 50% Ensure, suggesting that attentional deficits at the higher reinforcer concentration may have been masked by increased motivation to work for the food reinforcer. When Ensure was replaced with water as a reinforcer, there were no differences in hit rate or reaction time between the two genotypes. Although this may seem inconsistent with the change from 50 to 10% Ensure, in fact there is a fundamental difference in that it is a change from reinforcement to no reinforcement, rather than from strong-to-weak reinforcement. The mice were not water deprived, and the rapid drop in hit rate to 20% in both groups after three sessions is evidence that water held little or no value as a reinforcer.
The persistence of TRβ transgenic mice to approach the food well with only a water outcome is characteristic of resistance to extinction (Sagvolden et al. 1998), which may be mediated by the presence of conditioned reinforcers in the operant-testing environment. Johansen and Sagvolden (2004) showed the importance of conditioned reinforcers using the spontaneously hypertensive rat (SHR) model of ADHD. Using a multiple-component reinforcement schedule, performance of water-restricted SHR rats for water reinforcement was similar to that of controls. However, under the extinction component of the schedule, SHR rats were significantly more likely to enter the water receptacle in the presence of conditioned reinforcer (a light that previously signaled reinforcement). In the absence of conditioned reinforcers, SHR mice were no more likely than controls to enter the water receptacle. In the present experiment, there were a number of conditioned reinforcers operating on each hit trial, the most salient of which was extinguishing of the cue light at the moment of each correct response. Each correct response was also signaled by a click, which was used during training to facilitate learning and maintained throughout the experiment. Immediately following these events, the dipper was raised, which was accompanied by the quiet but distinct sound of the motor driving the dipper arm. All of these events occurred on every hit trial, regardless of reinforcer concentration, and probably served as conditioned reinforcers. Resistance to extinction can be induced by lesions of the dorsal noradrenergic bundle (DNB) and reversed by inhibitors of the NE transporter (Mason & Fibiger 1979; Mason & Iversen 1979; Pisa et al. 1988). An extensive set of experiments conducted by Mason & Iversen (1977, 1978, 1979; Mason 1979) determined that resistance to extinction induced by DNB lesions could not be explained by conditioned reinforcement, locomotor activity, perseveration, internal inhibition, frustrative non-reward, or motivation. Instead, they found that this DNB extinction effect was the result of impaired selective attention in the failure to ignore relevant stimuli in the experimental situation. In contrast, Steketee et al. (1989) found that that the DNB extinction effect resulted from differential response to novelty and not attentional impairments. Although further investigation is required to determine the behavioral and neurochemical processes involved in the persistence of TRβ transgenics to enter the food well in the absence of reinforcement, it is consistent with deficits observed in the SHR model and with a disruption in the noradrenergic system.
Although the incidence is greater among children with ADHD than in children without ADHD, detected thyroid abnormalities, including RTH, are rare in children with ADHD (Weiss et al. 1993). Like the TRβ transgenic mice, most children with ADHD are hyperactive, impulsive, and inattentive in the presence of normal thyroid hormone profiles. In the TRβ transgenics, these symptoms arise from a brief period of thyroid resistance during brain development, but symptoms persist much longer. In contrast, the initiating cause of the vast majority of ADHD cases is unknown. A recent study suggests that subclinical maternal thyroid abnormalities may be involved in a greater number of cases of ADHD than previously thought—even among children with normal thyroid levels at the time the symptoms are expressed (Haddow et al. 1999). Estimates of the prevalence of subclinical thyroid abnormalities range as high as 23% (Canaris et al. 2000; Glinoer & Delange 2000; Hollowell et al. 2002; Klein et al. 1991; Ladenson et al. 2000; Volzke et al. 2003). In addition, exposure to a number of environmental toxins has been implicated in ADHD (Berger et al. 2001; Hauser et al. 1998; Rice 2000). Many of the most common environmental toxins act directly on the thyroid system (Gauger et al. 2004), and in some cases may induce thyroid hormone resistance (Kuriyama et al. 2003). In many if not most cases, subclinical maternal or neonatal thyroid dysfunction or perinatal exposure to environmental toxins goes undetected or unmeasured. Thus, it is not known the extent to which these factors contribute to the prevalence of ADHD in the general population. Our data are consistent with the notion that transient, undetected, or subclinical thyroid disruption during gestation or postnatal brain development may contribute to some cases of ADHD in which juvenile thyroid levels are completely normal.
Interestingly, the thyroid resistance phenotype was not observed in female transgenics, despite their expressing largely similar behavioral phenotypes as male mice. The fact that female transgenics exhibited an attenuated response to methylphenidate and increased DA turnover suggests that the changes in the catecholamine system are also similar in male and female transgenics. At this point, we do not know the cause of the behavioral, pharmacological, or neurochemical phenotypes in the female transgenics, but as with male transgenics, it probably involves a transient thyroid resistance. There are a number of possible reasons for this gender difference. There is evidence in rats that the postnatal TSH surge begins later in females than in males (Simpkins et al. 1976). This temporal difference may be attributable to the fact that female rats have a higher density of pituitary TRβ than male rats (Donda et al. 1987) and thus may be better able to maintain physiological TSH levels. There is sexual dimorphism in both the number of substance P (SP)-containing cells in the anterior pituitary (AP) as well as the number of those cells that colocalize with TSH (Brown et al. 1991). The function of the colocalization is not known, but it is thought that SP modulates TSH release. SP levels in the AP are regulated both by thyroid hormones and estrogen (Brown et al. 1991). Estrogen receptors are also found in the AP and may compete with thyroid receptors for coactivators and DNA response elements (Pfaff et al. 1994). Estrogen increases thyroidal 131I uptake in mice, starting at 30 days of age (Cidlowski et al. 1975), and modulates thyroid growth in a gender-specific manner from 30 to 45 days of age (Banu et al. 2001). Other changes, such as gender differences in hippocampal granule cell number, are directly mediated by thyroid hormone during development (Madeira et al. 1988). Any or all of these mechanisms and more may contribute to the observed gender difference in the thyroid resistance phenotype in the TRβ transgenics at 33 days of age.
The TRβ transgenic mice exhibit all of the core symptoms and several adjunct features of ADHD, including a paradoxical response to methylphenidate. It is notable that the behavioral phenotypes expressed by the TRβ transgenic mice are dynamic and sensitive to changes in environmental conditions, stress, and reinforcement. Such differential reactivity has also been observed in children with ADHD, is altered by methylphenidate, and may be responsible for the observed gender differences (Iaboni et al. 1997; Johansen et al. 2002). This suggests that the mechanisms involved in ADHD converge at some point with mechanisms downstream of inactive TRβ1 receptors in the transgenic mice. The increased DA turnover in the transgenic mice supports this notion. Thyroid hormones regulate complex cascades of gene expression during development, including many of the genes implicated in ADHD. Exactly which genes are perturbed to produce the constellation of ADHD symptoms is not yet known. Thus, the TRβ transgenic mouse is a powerful tool that may be used to discover basic mechanisms and novel therapeutics involved in a large number of cases of ADHD of differing etiologies.