Developmental Thyroid Hormone Insufficiency and Brain Development: a Role for Brain-Derived Neurotrophic Factor (Bdnf)?

Neuroscience xxx (2013) xxx–xxx

REVIEW DEVELOPMENTAL THYROID HORMONE INSUFFICIENCY AND BRAIN DEVELOPMENT: A ROLE FOR BRAIN-DERIVED NEUROTROPHIC FACTOR (BDNF)?

M. E. GILBERT a* AND S. M. LASLEY b the available literature, the case for a direct role for BDNF a Toxicity Assessment Division, Neurotoxicology Branch, U.S. in thyroid-mediated effects in the brain is not compelling. Environmental Protection Agency, Research Triangle Park, NC, USA We conclude that delineation of the potential role of neurotrophins in TH-mediated neuronal development may be b Department of Cancer Biology and Pharmacology, University more fruitful by examining additional neurotrophins (e.g., of Illinois College of Medicine, Peoria, IL, USA nerve growth factor), moderate degrees of TH insufficiency, and younger ages. We further suggest that investigation of Abstract—Thyroid hormones (TH) are essential for normal BDNF invoked by synaptic activation (i.e., plasticity, enrich- brain development. Even modest degrees of TH disruption ment, trauma) may serve to elucidate a role of thyroid experienced in utero can result in neuropsychological defi- hormone in BDNF-regulated synaptic function. cits in children despite normal thyroid status at birth. Neuro- This article is part of a Special Issue entitled: Steroid hor- trophins have been implicated in a host of brain cellular mones and BDNF. Published by Elsevier Ltd. on behalf of functions, and in particular, brain-derived neurotrophic IBRO. factor (BDNF) has a well documented role in development and function of the nervous system. A number of laborato- Key words: BDNF, hypothyroidism, hippocampus, cortex, ries have reported the effects of TH administration or severe developmental, propylthiouracil. deprivation on neurotrophin expression in brain. This review provides an overview and update of recent develop- ments in the thyroid field as they relate to the nervous Contents system. Secondly, we describe an animal model of low level Introduction 00 TH insufficiency that is more relevant for studying the Some basics of thyroid biology 00 neurological consequences associated with the modest TH Thyroid hormone transporters in the brain 00 perturbations of subclinical hypothyroidism, or that would Thyroid hormone metabolism in the brain 00 be anticipated from exposure to environmental contami- Non-genomic actions and novel thyroid signaling molecules 00 nants with a mode-of-action that involves the thyroid. Effects of TH insufficiency in humans 00 Finally, we review the available in vivo literature on TH-med- Environmental factors and subclinical hypothyroidism 00 iated alterations in neurotrophins, particularly BDNF, and Pharmacological models of hypothyroidism using propylthio- discuss their possible contribution to brain impairments uracil and methimazole 00 associated with TH insufficiency. The observations of Low dose PTU model characterization 00 altered BDNF protein and gene expression have varied as Hippocampal synaptic transmission and plasticity 00 a function of hypothyroid model, age, and brain region Hippocampal-dependent learning 00 assessed. Only a handful of studies have investigated the Cell fate specificity and white matter abnormalities 00 relationship of neurotrophins and TH using models of TH Errors in neuronal migration 00 deprivation that are not severe, and dose–response informa- Is there a role for neurotrophins in the developmental effects of tion is sparse. Differences in the models used, species, thyroid hormone insufficiency? 00 doses, regions assessed, age at assessment, and method Increasing thyroid hormones in adults in vivo 00 employed make it difficult to reach a consensus. Based on Augmenting thyroid hormones in the neonate in vivo 00 Reducing thyroid hormones in the adult 00 Reducing thyroid hormones to neonate 00 *Corresponding author. Address: Toxicity Assessment Division, Summary and conclusions 00 Neurotoxicology Branch (MD-B105-05), National Health and Envi- Uncited reference 00 ronmental Effects Research Laboratory, U.S. Environmental Protec- Acknowledgments 00 tion Agency, Research Triangle Park, NC 27711, USA. Tel: +1-919- References 00 541-4394; fax: +1-919-541-4849. E-mail address: [email protected] (M. E. Gilbert). Abbreviations: BDNF, brain-derived neurotrophic factor; GD, gestational day; GFAP, glial fibrillary acidic protein; LTP, long-term potentiation; MCT, monocarboxylate transporter; MMI, methimazole; NGF, nerve growth factor; NT, neurotrophin; OATP, organic anion- INTRODUCTION transporting polypeptide; PTU, propylthiouracil; T1AM, 3- iodothyronamine; T3, 3,5,30-triiodothyronine; T4, thyroxine; TH, thyroid hormones; TRE, thyroid response element; TSH, thyroid Thyroid hormones (TH) are critical for normal on brain stimulating hormone. development. Severe TH deficiencies in the neonatal

0306-4522/12 $36.00 Published by Elsevier Ltd. on behalf of IBRO. http://dx.doi.org/10.1016/j.neuroscience.2012.11.022 1 2 M. E. Gilbert, S. M. Lasley / Neuroscience xxx (2013) xxx–xxx period are of clinical interest because they can result in is distinct from, and impairments more subtle than, those diminished mental capacity which can be mitigated with resulting from chemical or surgical thyroidectomy. early diagnosis and hormone supplementation. However, It is the TRE, not the liganded receptor, which directly a greater recognition of the impact of maternal regulates specific TH-responsive genes. Few genes have subclinical hypothyroidism on fetal CNS function is been identified that are directly activated by TH, and of emerging, and the contribution of TH-disrupting that group, most are transcription factors (e.g., hairless, environmental contaminants in this etiology remains to hr) that modulate the expression of other genes be determined. Because of the established roles of (Thompson and Potter, 2000). TH regulation of these neurotrophins in neuronal migration, differentiation, and downstream genes is therefore indirect as these genes plasticity, there has been a focus of efforts to understand do not themselves possess a TRE and do not bind the their role in TH-mediated neurological effects. T3-receptor complex. In this manner TH can modulate In this review we examined the available literature to the expression of a multitude of downstream genes, an evaluate the association between TH disruption and action that dramatically increases the sphere of their neurotrophin gene/protein expression. TH synthesis influence on brain development, as well as the difficulty inhibitors have been used to pharmacologically induce identifying developmentally important TH-responsive hypothyroidism in animals, and we review data genes (Anderson et al., 2003; Flamant and Samarut, generated by these models. The roles of neurotrophins, 2003; Quignodon et al., 2004). Furthermore, TRs do not with special attention to brain-derived neurotrophic regulate the same genes in all neuronal cell types, nor factor (BDNF), are then evaluated by examining their do they regulate the same gene in the same cell over responses to TH supplementation or deprivation in time during development, necessitating the existence of young and adult animals in vivo. This assessment of the mechanisms that control the cell specificity and timing of literature will allow us to define clearer roles for TH on TR action (Iniguez et al., 1996; Williams, 2008; neurotrophin function, to evaluate the potential for Hernandez et al., 2010). Selective expression of TH- neurotrophins to serve as biomarkers of developmental specific transporter proteins in the brain and local TH effects, and to elucidate the course of future studies. metabolizing enzymes represent two mechanisms whose relative contribution to the local fidelity of TH signaling is just beginning to be unravelled. SOME BASICS OF THYROID BIOLOGY Adding further to this complexity, the degree of change in the expression of TH-responsive genes in the brain is TH is essential for a number of physiological processes in subtle in nature, standing in marked contrast to TH- mammalian species. In the hypothalamic–pituitary– mediated change in gene expression in many other thyroid axis (HPTA), thyroid releasing hormone (TRH) tissues. Most known neural genes exhibit transient from the hypothalamus stimulates the pituitary to release responsiveness to TH and undergo changes in thyroid stimulating hormone (TSH). TSH acts on thyroid expression of only two to threefold in response to the gland receptors to activate synthesis and release of TH. hormone (Poguet et al., 2003; Quignodon et al., 2004; Thyroxine (T4) is the predominant form of TH released Royland et al., 2008). Consequently, it has proven into the bloodstream and is enzymatically deiodinated to challenging to associate the changes in expression of a 0 3,5,3 -triiodothyronine (T3), a high-affinity ligand for the particular gene or family of genes to the well known nuclear TH receptors TRa and TRb, whose activation effects of TH on brain development (Oppenheimer and regulates vertebrate development and physiology. In the Schwartz, 1997; Thompson and Potter, 2000; Bernal, circulation, these hormones are bound to serum proteins 2002). for transport and delivery to the many TH-dependent Difficulty linking alterations in gene expression and organs. The developing brain is one organ most brain function may also derive from the paradigm vulnerable to TH insufficiency (Oppenheimer and traditionally used in these studies. Specifically, models Schwartz, 1997; Santisteban and Bernal, 2005; Williams, of severe hypothyroidism may lead to a plethora of 2008). A key feature of TH action in the brain is the effects on the somatic development of many organ temporal sequence of events it supports, a feature that systems that may obscure the more direct actions of TH increases the complexity of research studies, while at on brain development. Observations at the molecular, the same time providing avenues to identify mechanisms anatomical, physiological, and behavioral level in such that must exist to control TH action. models may not only reflect the actions of TH TH acts through nuclear receptors encoded by two insufficiency but also may include those secondary to genes, TRa or TRb, to affect gene transcription. Thyroid the somatic insult resulting from severe TH deprivation. receptors bind to DNA sequences (thyroid response elements, TREs) in the unliganded state and for this Thyroid hormone transporters in the brain reason are called ‘‘aporeceptors’’. In the absence of T3, these aporeceptors typically behave as transcription Specific transporter proteins from the monocarboxylate repressors toward target genes (Bernal and Morte, transporter (MCT) and organic anion-transporting 2012). Upon binding of hormone, the aporeceptors then polypeptide (OATP) families actively take up T3 and T4 function as transcription activators. As such, the absence and are primarily expressed on blood vessels, astrocytes, of T3 exerts a more detrimental effect than the absence and neurons (Williams, 2008; Bernal, 2011). Individuals of receptors, as shown in various TR knockout models in suffering from mutations of the MCT8 transporter exhibit which the resulting somatic and neurological phenotype severe mental retardation, i.e., Allen–Herndon–Dudley

Please cite this article in press as: Gilbert ME, Lasley SM. Developmental thyroid hormone insufficiency and brain development: A role for brain-derived neurotrophic factor (BDNF)? Neuroscience (2013), http://dx.doi.org/10.1016/j.neuroscience.2012.11.022 M. E. Gilbert, S. M. Lasley / Neuroscience xxx (2013) xxx–xxx 3 syndrome (Schweizer and Ko¨ hrle, 2012). The OATP family of TH have also been documented (Davis et al., 2008). of transporters, also capable of iodothyronine transport, is T4 itself, and its presumed iodothyronamine metabolites, expressed in hippocampal, cortical, and cerebellar have rapid physiological actions both in vitro and in vivo. neurons (Schweizer and Ko¨ hrle, 2012). Initial findings These data indicate that actions of TH can also occur at with mouse models with targeted deletions of these non-nuclear sites to alter developmental processes and transporters were disappointing as these mutants did not function of fetal, neonatal and adult nervous systems. exhibit characteristics of the neurodevelopmental One presumed TH metabolite, 3-iodothyronamine phenotype observed in humans or rodent models of (T1AM), produces a rapid and profound effect on body severe hypothyroidism (Heuer and Visser, 2009; Mayer temperature, heart rate, and metabolism when et al., 2012a). However, recent reports of double administered in vivo (Scanlan et al., 2004). Although knockouts of both transporters in mouse models reveal initially described as an extrathyroidal metabolic significant neuroanatomical abnormalities and behavioral derivative of T4, a recent report indicates T1AM impairments (Mayer et al., 2012b). Exciting new represents a distinct hormone entity of the same discoveries using transgenic models hold great promise biosynthetic machinery in the thyroid gland that is for the future elucidation of the role played by TH- necessary for T4 synthesis (Hackenmueller et al., 2012). transporters in the region-specific and time-dependent As such, this newly discovered TH may constitute a part presentation of T3 to the neuron. From both a clinical and of TH action in target tissues including the brain whose environmental contaminant perspective, it is important to effects are yet to be revealed. recognize that these mechanisms alter TH availability to the developing brain in a manner that is not accurately reflected in changes in serum T3 or T4. EFFECTS OF TH INSUFFICIENCY IN HUMANS In humans, severe deficiencies in TH during development Thyroid hormone metabolism in the brain are associated with irreversible damage to virtually all Enzymes that convert T4 to T3 (i.e., deiodinase 2, D2) or organ systems, a condition termed cretinism (Glinoer, deactivate T3 to physiologically inactive reverse-T3 or 2007). The primary causes of severe developmental T2 (i.e., deiodinase 3, D3) provide additional means of hypothyroidism in humans are mainly congenital control over TH-dependent gene regulation in the neuron hypothyroidism and iodine deficiency (Glinoer and (Bernal, 2011; Williams and Bassett, 2011). Neuronal T3 Delange, 2000; Glinoer, 2007). Each condition can is largely derived from D2-dependent metabolism of T4 produce a different spectrum of symptoms, the specific in glial cells. D2-deficient mice exhibit elevated nature of which is dependent on the timing, duration, circulating T4 and TSH levels but normal circulating T3 and severity of the deficiency. and as neonates display reduced T3 concentrations in Congenital hypothyroidism refers to a condition in the brain. In contrast to the glial expression of D2, the which children are born with very low levels of serum inactivating D3 enzyme is highly expressed in neurons. hormone, typically resulting from a malformation of the As neonates, D3-deficient mice have elevated levels of thyroid gland. Children appear normal at birth, their serum hormones, whereas serum hormone profiles mothers have normal thyroid function, and in the past, demonstrate hypothyroidism after PN15. Selective the lack of specific symptoms often delayed the deletions of D2 or D3 have modest effects on the diagnosis and treatment, with the prognosis deteriorating expression level of T3-target genes (e.g., RC2, Hr) and with the passage of time. The full clinical picture prior to offspring exhibit a mild general neurological phenotype the era of systematic neonatal screening included compared with the abnormalities of hypothyroidism. stunted growth and severe mental retardation (see Collectively, these findings with transgenic mouse reviews by Dussault and Ruel, 1987; Chan et al., 2005; models suggest that compensatory mechanisms exist to Glinoer, 2007). Neonatal screening programs for ameliorate the neurological consequences of transporter identification and early treatment, now widely instituted in and deiodinase deficiencies during the organism’s North America and abroad, have largely eliminated development, and that significant species differences congenital hypothyroidism as a primary cause of mental may exist between humans and rodents (i.e., as seen in retardation. However, lower global IQ scores, language MCT8 mutations). However, several laboratories have delays and weak verbal skills, motor weakness, recently described significant neurological impairments attentional deficits and learning impairments were and overlap in brain gene expression with chemical evident in children with delayed or inadequate treatment hypothyroid models with different combinations of (Derksen-Lubsen and Verkerk, 1996). Even in cases receptor, transporter, and deiodinase deletions (e.g., where the condition is diagnosed early and treated Hernandez et al., 2010, 2012; Bernal, 2011; Mayer et al., effectively, subtle impairments in mental function remain 2012b). (Zoeller and Rovet, 2004). Iodine is an essential element for the biosynthesis of TH and some of the most serious neurologic Non-genomic actions and novel thyroid signaling impairments from environmental causes are in children molecules born in regions of the world where dietary iodine T3-modulated transcription of target genes via activation deficiency is prevalent (Zimmermann, 2007). These of TRs is a slow process, the effects of which manifest children are characterized by stunted growth and a high over hours and days. However, rapidly occurring effects incidence of mental retardation. Neurological

Please cite this article in press as: Gilbert ME, Lasley SM. Developmental thyroid hormone insufficiency and brain development: A role for brain-derived neurotrophic factor (BDNF)? Neuroscience (2013), http://dx.doi.org/10.1016/j.neuroscience.2012.11.022 4 M. E. Gilbert, S. M. Lasley / Neuroscience xxx (2013) xxx–xxx deficiencies seen in endemic iodine deficiency beginning also for the ongoing debate of the cost, risk and benefits in utero and that continue throughout infancy are more of treatment of subclinical hypothyroidism. severe than those resulting from congenital hypothyroidism, indicating that early fetal thyroid Pharmacological models of hypothyroidism using function is also critical for normal brain development propylthiouracil and methimazole (Glinoer, 2007). A number of studies have now demonstrated that There are two prevalent models of developmental modest levels of TH insufficiency characteristic of hypothyroidism using the synthesis inhibitors subclinical hypothyroidism in pregnant women can result methimazole (MMI) or propylthiouracil (PTU). These in neuropsychological deficits in their offspring, despite agents block the action of the thyroperoxidase enzyme normal thyroid status of the child at birth (Haddow et al., in the thyroid gland that is necessary to couple iodine to 1999; Zoeller and Rovet, 2004; Berbel et al., 2009; tyrosine in order to form T3 and T4 (Cooper et al., 1983, Henrichs et al., 2010). These observations of 1984). These agents are typically delivered to pregnant neurological impairments following subclinical reductions dams through the drinking water beginning in early in maternal T4 have raised the level of concern over the gestation to compromise hormone supplies to the fetus influence of environmental contaminants which and neonate, or late in gestation to affect the near term decrease thyroid function (Porterfield, 1994; Blount fetus and the thyroid status of the nursing pup. et al., 2006; Ginsberg et al., 2007; Mendez et al., 2012). Typically, a single high dose of either PTU or MMI, The animal literature has lagged behind the human data ranging from 0.02% to 0.05% (equivalent to 200–500 because models of moderate TH disruption have not ppm) is administered to pregnant dams through the been available that parallel the human phenotype. As drinking water beginning early (gestational day 6, GD6) such, animal research is lacking on mechanisms, or late (GD17) in gestation. These treatments biomarkers of effect, and efficacy of intervention to dramatically reduce circulating levels of T3 and T4, and reverse subtle perturbations of the thyroid axis. produce exponential increases in pituitary release of TSH that stimulates the thyroid gland to increase hormone synthesis and release. These models generate a serum hormone profile consistent with hypothyroidism ENVIRONMENTAL FACTORS AND in humans. Phenotypically, these animals are SUBCLINICAL HYPOTHYROIDISM characterized by stunted growth, severe developmental delays, hair loss, and motor and sensory deficits The U.S. Environmental Protection Agency must regulate (Dussault and Ruel, 1987; Bernal, 2002; Gilbert and chemicals that have the potential to pose a hazard to Zoeller, 2010). Other models have incorporated human health. A large number of environmental thyroidectomy of the dam and the newborn pups to contaminants with diverse structures have been shown even more dramatically alter thyroid status (e.g., Iniguez to decrease circulating levels of TH (Brucker-Davis, et al., 1996; Kim et al., 2007). 1998). Animal studies have indicated that this action of These models have provided valuable information on xenobiotics on the thyroid system may contribute to the critical importance of TH to brain development in the alterations in nervous system development and function fetal and early neonatal period, but their utility in (see review by Gilbert and Zoeller, 2010). Some of assessing the impact of moderate degrees of TH these environmental contaminants have also been insufficiency is limited. evaluated in humans and been found to disrupt the thyroid axis (Koopman-Esseboom et al., 1994; Blount Low dose PTU model characterization et al., 2006; Chevrier et al., 2007; Turyk et al., 2007; Baccarelli et al., 2008). Moreover, human exposures to To begin to address this knowledge gap, we have some of these same contaminants are associated with developed a model of low level TH disruption using neurodevelopmental impairments (Jacobson and graded levels of PTU delivered through the drinking Jacobson, 1996; Koopman-Esseboom et al., 1996; water to the pregnant rat from early gestation and Lonky et al., 1996; Patandin et al., 1999), suggesting continuing throughout lactation (e.g., Sui and Gilbert, that environmental contaminants may produce 2003; Gilbert and Sui, 2006; Gilbert, 2011). In contrast neurotoxic effects on the developing human by to the standard models of 200–500 ppm, we have interfering with thyroid function or hormone action. examined concentrations of PTU in the drinking water Relative to severe iodine deficiency and congenital that range from 1 to 10 ppm. Water consumption is not hypothyroidism, environmental contaminants reduce altered at these concentrations, a significant concern circulating TH to a modest degree (Gilbert and Zoeller, when high concentrations of PTU are used, due to its 2010). What has not been established is the degree to bitter taste. This model leads to dose-dependent which thyroid status must be perturbed to produce reductions in serum TH, with reductions in T4 being the adverse neurological outcomes. Until recently most sensitive, and at the lower doses, the only experimental studies have uniformly examined severe indicator of hormone disruption. Body weight gain in TH deficiency with a limited information on the dams and pups is not altered at 1 and 2 ppm PTU, and neurological consequences of low-level thyroid only marginally and transiently in pups at 3 ppm in the dysfunction. Addressing this question has important later neonatal period. At the 10 ppm dose, body weight implications not only for environmental regulation but over the first 10 days of life is not different from controls,

Fig. 1. Serum thyroid hormone proﬁle in the low-dose PTU model. Dams were exposed to varying concentrations of PTU in the drinking water beginning on gestational day (GD6) and continuing through lactation. (A) Serum (mean ± SEM) T4 was reduced in a graded fashion as a function of PTU dose in pups and in dams (n = 5–15 litters/group). (B) More modest reductions in circulating T3 were seen at 3 and 10 ppm PTU concentrations in pups and dams. (C) Serum T4 and (D) serum T3 exhibited full recovery by PN90. ⁄p < 0.05 compared to controls (from Lasley and Gilbert, 2011).

Fig. 2. Developmental exposure in the low-dose PTU model produces dose-dependent impairments in excitatory and inhibitory synaptic transmission and synaptic plasticity as measured in field potentials recorded from the dentate gyrus in vivo. (A) Excitatory synaptic transmission as measured via an input/output function of the amplitude of the EPSP slope of adult offspring of PTU-exposed dams is impaired at moderate degrees of TH insufficiency. Input/output curves from 2 and 3 ppm groups were significantly suppressed relative to controls (data taken from Gilbert, 2011). Previous work demonstrated greater reductions at a higher dose of 10 ppm PTU (Gilbert and Sui, 2006, not shown). (B) Paired pulse depression is reduced and paired pulse facilitation is enhanced in PTU-exposed animals (from Gilbert et al., 2007). Effects are most prominent at the lowest stimulus intensity (20% maximum) and highest PTU dose. (C) Theta-burst stimulation of the perforant path induced long-term potentiation of the EPSP slope, the magnitude of which was significantly suppressed in PTU-exposed offspring at all doses levels when collapsed across intensity (ANOVA with mean contrasts); indicates all dose groups significantly different from 0 ppm, p < 0.05 (modified from Gilbert, 2011).

et al., 2005) but the magnitude of the change appeared on water maze acquisition (unpublished observations), not as robust as that observed in vivo. This may but context learning deficits in a modified version of the possibly reflect differential sensitivity of these trace fear conditioning paradigm were evident in subregions to developmental hypothyroidism, but are offspring from the lowest 1 ppm dose group tested more likely attributable to inherent differences between (Fig. 4). Impairments in context learning were also seen in vivo and in vitro assessments. only in male offspring at the lower doses, but the more severe hypothyroidism induced by 10 ppm produced impairments in males and females, and in both context Hippocampal-dependent learning and cue learning (Gilbert and Taylor, 2012). Deficits in Learning and memory are clearly impaired in animals context fear learning were subtle in nature and could be following developmental hypothyroidism (Davenport and ameliorated in all but the 10 ppm dose group by Dorcey, 1972; Akaike et al., 1991; Darbra et al., 2004; increasing the duration of the shock, a manipulation that Gilbert and Sui, 2006; Axelstad et al., 2008). In the low- serves to enhance the robustness of the learning dose PTU model we have evaluated learning and (Fig. 4). These data are important as they reveal memory using the Morris water maze and trace fear significant and persistent declines in cognitive function conditioning in adult offspring. Both paradigms revealed in response to modest reductions in serum T4 in the learning deficits at doses of 3 and 10 ppm PTU (Fig. 4; dam (<20%) and the neonate (50%) and in the Gilbert and Sui, 2006; Gilbert, 2011). In recent studies, absence of change in serum T3, TSH, body weight, or lower doses of PTU (1 and 2 ppm) were without effect latency to eye opening. Generally, it has been quite

Fig. 3. Altered inhibitory circuitry as assessed by paired pulse functions in adult offspring of PTU-treated dams (see Fig. 2B) was associated with reductions in immunostaining for parvalbumin, a calcium-binding protein found only in inhibitory neurons. Staining was reduced in the cortex and hippocampus of PN21 offspring (A) exposed to 3 or 10 ppm of PTU. Cell counts of parvalbumin-positive cells (B) and protein content based on Western blot analysis (C) were similarly reduced, some of which persisted to adulthood (from Gilbert et al., 2007). ⁄p < 0.05 compared to controls. difficult to demonstrate impairments in these simple tasks composition of the corpus callosum and anterior of learning and memory when modest reductions in TH commissure, the primary fiber pathways that support are induced by PTU, the environmental contaminant communication between the two cerebral hemispheres. perchlorate (Gilbert and Sui, 2008), or by marginal At low doses, reductions in mRNA of myelin-associated dietary iodine deficiency (Gilbert et al., 2012). The lack glycoprotein (MAG), a marker of oligodendrocytes, were of effect of these modest alterations in thyroid status on coupled with increases in mRNA for the astrocyte neurobehavior may reflect the lack of sensitivity of the marker, glial fibrillary acidic protein (GFAP). The test to subtle deficits as much as the absence of subtle reduction in oligodendrocytes was dose-dependent and cognitive impairments accompanying low-level TH accompanied by an equivalent parallel increase in the insufficiency. number of GFAP-positive astrocytes – i.e., oligodendrocyte density declined and astrocyte density Cell fate specificity and white matter abnormalities rose linearly with serum T4 concentration (Fig. 5). Because both oligodendrocytes and astrocytes derive White matter formation and myelination are important from the same glial precursor cell and the total number developmental events that are regulated by TH and of cells of both phenotypes did not change, these data growth factors including BDNF (Ibarrola and Rodriguez- suggest that the lack of TH altered cell fate specificity, Pena, 1997; Ferreira et al., 2007; vonDran et al., 2010). i.e., in the absence of sufficient TH, glial precursors Children exposed to low levels of TH during critical were directed more to an astrocyte than to an periods of brain development have altered myelination oligodendrocyte lineage. patterns (Jagannathan et al., 1998) and reductions in myelin formation is a hallmark of congenital hypothyroid Errors in neuronal migration models. In these models, although the total number of axons is not altered, the number of myelinating Smaller brain size and disorganized cytoarchitecture in the oligodendrocytes and myelinated axons in major white cortex and cerebellum have been described in models of matter tracts is significantly reduced (Gravel and severe hypothyroidism (Rami et al., 1986; Madeira et al., Hawkes, 1990; Schoonover et al., 2004; Powell et al., 1991, 1992). Bernal and colleagues reported delayed 2012). In the low-dose PTU model, Sharlin et al. (2008) and disordered migration of cortical and hippocampal identified structural alterations in the white matter of neurons in response to iodine deficiency, a model of PTU-exposed pups using in situ hybridization transient maternal hypothyroidism, and the late-gestation techniques. They reported alterations in the cellular thyroidectomy model of preterm birth (Lavado-Autric

Fig. 4. Spatial learning acquisition deficits were evident in the Morris water maze at 3 and 10 ppm as represented in latencies to find the hidden platform (A), and for time spent searching in the correct quadrant on a series of probe trials (B) administered at intervals throughout acquisition (from Gilbert and Sui, 2006). (C) Deficits in context learning were revealed at low levels of TH disruption in a trace fear conditioning paradigm 24 h after training with two presentations of the conditioned stimulus (tone/light) and unconditioned stimulus (footshock) separated by a 30-s trace interval. All dose groups were impaired with short duration shocks (1 mA for 0.5 s). (D) Deficits were ameliorated in all but the highest PTU dose of 10 ppm when shock durations were increased to 2 s (from Gilbert and Taylor, 2012). ⁄p < 0.05 compared to controls. et al., 2003; Auso et al., 2004; Berbel et al., 2010). A been fully elucidated, the reliability of its positioning and pattern of aberrant neuronal migration has also been its detectability at low doses make it attractive as a described in the low-dose PTU model by our laboratory. simple, sensitive, and reliable biomarker of modest We have identified the presence of a misplaced cluster prenatal TH disruption. of neurons, a heterotopia, positioned in the corpus callosum of developmentally hypothyroid animals IS THERE A ROLE FOR NEUROTROPHINS IN (Fig. 6). TH insufficiency in the prenatal period is both THE DEVELOPMENTAL EFFECTS OF THYROID necessary and sufficient for its formation. Clearly HORMONE INSUFFICIENCY? observable with simple Nissl staining in offspring of 3 and 10 ppm animals (Goodman and Gilbert, 2007; Neurotrophins are a family of soluble proteins that regulate Saegusa et al., 2010; Powell et al., 2012), recent the initial steps in neuron–target interactions and promote findings indicate that it is discernible using in vivo MRI the survival and differentiation of various nerve cells in the imaging techniques (Powell et al., 2012). With neuron- developing and adult nervous system. Nerve growth factor specific antibodies and immunohisto-chemistry, we have (NGF) was the first neurotrophin discovered and is a key recently identified the presence of heterotopias in player in directing developing fibers toward their target offspring of dams exposed to the lowest 1 ppm dose of field to regulate the extent of innervation (Lewin and PTU, a dose that reduces maternal T4 by less than 20% Barde, 1996). Other members of the neurotrophin family (unpublished observations). Although the functional include BDNF and neurotrophins-3, -4, and -5 (NT-3, NT- significance of the presence of a heterotopia has not yet 4/5). Neurotrophins are largely produced and released by

Fig. 5. Astrocytes expressing glial fibrillary acidic protein (GFAP) are increased (A) and oligodendrocytes expressing myelin-associated glycoprotein (MAG) are decreased (B) in a dose-dependent manner in the corpus callosum of rat pups exposed through gestation and lactation to PTU. Decreases in MAG-positive cells were balanced by increases in GFAP-positive cells and these changes were linearly related to serum T4 (C). The latter relationships suggest that TH insufficiency may disrupt cell fate specificity as both MAG- and GFAP-expressing cells are derived from a common glial precursor (from Sharlin et al., 2008). ⁄⁄p < 0.01; ⁄⁄⁄p < 0.001 compared to controls.

Fig. 6. An abnormal cluster of heterotopic neurons is reliably detected in the corpus callosum following prenatal TH reductions. Large bilaterally positioned heterotopias are evident in Nissl-stained material at high doses (10 ppm, arrows, from Goodman and Gilbert, 2007), and even more prominently using immunohistochemistry for the neuron-speciﬁc antibody NeuN. Recently heterotopic neurons have been detected at the lowest doses of PTU assessed, the incidence and size of which increase dose-dependently. Occasionally NeuN-positive cells in the same position were seen sporadically in the corpus callosum of some control animals (bottom left). However, a prominent heterotopia is clearly visible in response to 1 ppm PTU, a dose that reduces maternal TH by <20% (unpublished observations).

Please cite this article in press as: Gilbert ME, Lasley SM. Developmental thyroid hormone insufficiency and brain development: A role for brain-derived neurotrophic factor (BDNF)? Neuroscience (2013), http://dx.doi.org/10.1016/j.neuroscience.2012.11.022 10 M. E. Gilbert, S. M. Lasley / Neuroscience xxx (2013) xxx–xxx target tissues and are taken up by responsive neurons. 2005). Although delivery of T3 (1 lg/kg, sc) to the adult They direct various cellular mechanisms including cell rat in vivo also increases spine density in hippocampal fate determination, determination and process outgrowth neurons (Gould et al., 1990a,b), the involvement of in the developing brain and in response to CNS activity or neurotrophins in this action has not been directly injury in the adult brain. Neurotrophins have been demonstrated. implicated in a host of brain functions including neuronal Repeated administration of T4 (2 lg/kg/day, i.p. for cell survival, neurite outgrowth, cell migration, regulation 12 days) to the adult mouse does increase the of excitatory-inhibitory balance, and glutamate- expression of NGF mRNA and protein in cortex, dependent dendritic growth, synapse formation, cerebellum, and brainstem (Walker et al., 1979, 1981). stabilization and plasticity (Yoshii and Constantine-Paton, In contrast, in the rat, higher concentrations of T3 (50– 2010). BDNF and its associated receptor TrkB have a 500 lg/kg/day, s.c. for 10 days) did not alter the basal wide spectrum of biological actions: BDNF plays a role in levels of NGF, NT-3 or BDNF mRNA in the hippocampus promoting survival and differentiation of selected or pyriform cortex (Kim et al., 1998; Vaidya et al., 2001). neuronal populations, its expression and release are Giordano et al. (1992) reported selective increases in activity-dependent, and its presence can alter the NGF and NT-3 mRNA in the hippocampus but not cortex structure of dendrites and spines of neurons in the with high doses of T3 (500 lg/kg/day, i.p. for 8 days), developing and mature CNS. Consequently, BDNF has a but there was no change in BDNF mRNA expression in well documented role in brain development and in adult either site. In contrast to these older reports based on brain function and synaptic plasticity (e.g., Lewin and in situ hybridization, Sui et al. (2010) employed Barde, 1996; Lu and Figurov, 1997) as well as in the quantitative RT-PCR to evaluate BDNF expression. emergence and therapeutic reversal of nervous system They report that a single high dose of T3 (300 lg/kg) disorders including clinical depression and Alzheimer’s markedly increased BDNF mRNA expression and the disease (Martinowich et al., 2007; Allen et al., 2011). transcription of BDNF exons I, II, IV and V in the In view of the importance of BDNF in these various hippocampus for at least 24 h, the latest time-point roles, and because TH is clearly required for normal assessed. Increased BDNF gene expression was brain development, a number of laboratories have accompanied by an augmentation of protein expression examined the potential relationship between TH and in the hippocampus with a slightly delayed onset (4 h) neurotrophin expression in developing brains. It seems that also persisted for 24 h (Sui et al., 2010). reasonable to propose that alterations in BDNF may Collectively, these findings indicate that increasing underlie some of the persistent neurological impairments serum T3 concentrations to supra-physiological levels associated with developmental hypothyroidism. On by systemic administration of T3 has varied effects on review of the literature however, the results are generally neurotrophin expression in the adult rodent brain. When mixed with effects that are more consistent for other present, effects of TH administration are regionally members of the neurotrophin family (e.g., NGF) than specific, duration-dependent, and augmentation of BDNF. To investigate the relationship of TH and specific neurotrophins have been inconsistent – some neurotrophins, animals have been dosed with hormones reports favoring NGF and NT-3 with no change in BDNF or had circulating levels of TH dramatically reduced. (Giordano et al., 1992), others reporting increases in Exogenous delivery of T3 or T4 to the adult or neonate BDNF mRNA and protein (Sui et al., 2010). generally increases neurotrophin expression, but effects are brain region-dependent, and may vary with hormone Augmenting thyroid hormones in the neonate in vivo type, timing of delivery, animal species, and neurotrophin (see below). The impact of TH deprivation on BDNF also As in the adult, early work revealed an increase in NGF appears to be age- and region-specific, and with a few protein expression in the cortex, cerebellum, and exceptions inquiry has been limited to severe brainstem of mice treated as neonates with a constant hypothyroid models. Limited data are available for BDNF volume of T4 for the first 11 days of life (Walker et al., in animal models of moderate TH insufficiency that are 1982). This dosing regimen approximated dose levels of unencumbered by confounds of general developmental over 1 mg/kg, i.p. to the newborn and less than half that delays, concerns for undernutrition, severe growth much in the PN10 pup. Following a similar T4 dosing impairments, and excessive pup mortality. The following regimen in rats, Camboni et al. (2003) reported discussion is a brief summary of the available literature increased BDNF mRNA and protein expression in the concerning effects on neurotrophin expression in the septum of the PN10 pups which subsequently dissipated brain that accompany either exogenous supplementation by adulthood. Neonatal T4 treatment also increased or deprivation of TH in adult and developing rodents. BDNF, NGF, NT-3 and NT-4 mRNA in the hippocampus of the PN10 rat (Lu¨ esse et al., 1998). These increases in neurotrophin gene expression were not accompanied by Increasing thyroid hormones in adults in vivo alterations in the expression of their respective To examine the relationship of TH and neurotrophins, T3 neurotrophin receptors (p75, TrkA, TrkB, TrkC), leading or T4 have been added to brain cells in culture or the authors to conclude that neurotrophin content rather administered to euthyroid animals as adults or as than receptor function was the basis for their neonates. TH increases neurite extension and dendritic morphological effects produced in brain development expansion in neuronal cell cultures in a BDNF- (Roskoden et al., 1999). Assessments of neurotrophin dependent manner (Koibuchi et al., 2003; Farwell et al., protein expression, however, were not performed.

High doses of T3 (1 mg/kg/day, i.p., from PN7 to PN28) Animals have been typically exposed in utero and facilitated the development of choline acetyltransferase- throughout lactation by treating the drinking water of positive fibers in forebrain cholinergic projection areas pregnant dams with MMI or PTU. Koibuchi et al. (1999, and a precocious expression of cholinergic markers in 2001) reported reductions in BDNF mRNA and/or cortex, hippocampus, and amygdala (Gould and Butcher, protein in the cerebellum of neonatal rats and mice 1989; Oh et al., 1991). Supporting the observations of (PN15–30) exposed to high doses of PTU (500 ppm) or Roskoden et al. (1999), these findings in forebrain MMI (250 ppm) coupled with a very high dose of the structures were not accompanied by any change in NGF iodine uptake inhibitor perchlorate (20000 ppm) from receptor immunohistochemistry in the cortex. Similar mid-gestation to weaning. Reductions in transcription of doses of T3 in neonatal rats beginning at various ages BDNF exons II, III, IV, and V were reported for (PN1–PN28) increased mRNA expression of NT-3 in the hypothyroid animals on PN15 and PN30 relative to granule cells of the cerebellum but not in the hypothyroid animals co-treated with T4 (no untreated hippocampus, and the effects were limited to animals control animals were included in the design). No younger than PN21 (Lindholm et al., 1993). In summary, changes in any BDNF transcript were evident in these findings support those observed in the adult rat in younger (PN1 and PN7) or older (>PN30) offspring or which excess TH was found to augment neurotrophin in cortical tissue from the same animals. mRNA expression. Like Koibuchi et al. (2001), Godbole and colleagues (Sinha et al., 2009) also reported reductions in BDNF Reducing thyroid hormones in the adult mRNA and protein in the cerebellum following exposure of rat dams to MMI (250 ppm without perchlorate) Many fewer studies have examined the effects of adult beginning earlier in gestation (GD8 vs GD15 for onset hypothyroidism on BDNF expression. Basal Koibuchi et al., 2001). These authors observed levels of BDNF protein assessed by Western blot, in reductions in BDNF but at an earlier age (PN8) than addition to a number of the signaling molecules reported by Koibuchi et al. (2001). The study design implicated in synaptic plasticity (e.g., calcineurin, was somewhat limited as observations were based on CREB, MAP kinase) are reduced in rats following offspring from only two litters. Increases in mRNA for surgical removal of the thyroid gland in adulthood NGF and the neurotrophin receptor p75NTR were (Alzoubi et al., 2007). These changes were associated observed with no change in the BDNF receptor TrkB or with impairments in long-term potentiation in vivo in the NGF receptor TrkA. Paradoxically, increases in area CA1, but, distinct from developmental hypothyroid expression of NGF transcript in the cerebellum were models, LTP of the dentate gyrus remained intact coincident with decreases in NGF protein on PN0 and (Gerges et al., 2004; Alzoubi et al., 2007). In a PTU PN8 suggesting a post-translational modification in model of adult onset hypothyroidism, the opposite protein expression and demonstrating a lack of effect of enhanced expression of BDNF has recently consistency in message and protein levels of been reported. Corte´s et al. (2012) reported that a high neurotrophins (Caldwell et al., 2008). Opposite effects dose of PTU in the drinking water (500 ppm for were seen in a microarray study in mice in which 20 days) of adult rats increased BDNF protein in the p75NTR was downregulated in MMI-treated pups on telencephalon and mRNA expression in the cortex, and PN4, and this observation was confirmed by RT-PCR the hippocampal subfields, including the dentate gyrus. (Takahashi et al., 2008). BDNF has been implicated in postnatal neurogenesis in The Godbole laboratory also examined neurotophins the dentate gyrus of the hippocampus and the olfactory in the cortex, using a similar design and dosing bulb of the adult rat, a process whereby neurons regimen. Kumar et al. (2006) administered 250 ppm continue to be generated throughout life (Babu et al., MMI to dams from GD8 throughout lactation, and the 2009; Choi et al., 2009; Lazarini and Lledo, 2011). offspring were continued on this dose until sacrifice. Several reports have demonstrated reduced capacity Expression of the 53 kDa precursor form of NGF and for proliferation (Lemkine et al., 2005; Montero- the p75NTR receptor were markedly elevated from birth Pedrazuela et al., 2006) or survival of newly born to PN16, suggesting enhanced apoptosis, a function of neurons in both of these neurogenic niches of the adult p75NTR in development (Chen et al., 2009). At the same brain under conditions of thyroid compromise in time, however, TrkA mRNA was diminished. These adulthood (Ambrogini et al., 2005; DeSouza et al., effects had completely reversed by PN90 – i.e., pro- 2005; Kapoor et al., 2012). A reduction of adult NGF and p75NTR expression were significantly hippocampal neurogenesis under hypothyroid decreased and TrkA mRNA elevated compared to conditions may involve reduced expression of BDNF in euthyroid animals. These increases in NGF and the these brain regions. However, we found no reports p75NTR receptor, were consistent with mRNA results in directly assessing the possible relationship between the cerebellum (Sinha et al., 2009). Unlike the impaired neurogenesis and BDNF in animals with adult cerebellum, however, TrkA gene expression was onset hypothyroidism. decreased in the cortex of the pup, but elevated in the cortex of the adult. Reducing thyroid hormones in the neonate In offspring of dams exposed to 50 mg/kg/day PTU by gavage (400 ppm drinking water equivalent) beginning in More work has focused on reducing TH in the developing late gestation (GD17), Neveu and Arenas (1996) reported animal to evaluate the effect on neurotrophins in the brain.

Please cite this article in press as: Gilbert ME, Lasley SM. Developmental thyroid hormone insufficiency and brain development: A role for brain-derived neurotrophic factor (BDNF)? Neuroscience (2013), http://dx.doi.org/10.1016/j.neuroscience.2012.11.022 12 M. E. Gilbert, S. M. Lasley / Neuroscience xxx (2013) xxx–xxx reductions in cerebellar BDNF, TrkC, and NT-3 mRNA on cerebellum in the PN14 neonate (Fig. 7, Lasley and PN15 and PN30, whereas NGF and NT-4 expression Gilbert, 2011). We have since evaluated expression for levels were increased. High-dose PTU (500 ppm) from total BDNF and BDNF exons I and IV in the hippocampus GD15-PN21 decreased hippocampal BDNF exon II using the low-dose PTU model and found modest transcripts and BDNF protein concentrations on PN1, 5, reductions in mRNA in similarly aged pups (unpublished 15, and 30, with recovery evident by PN60 after observations). In adulthood however, BDNF protein termination of treatment at weaning (Sui and Li, 2010). expression was diminished in a U-shaped relationship as The profile of BDNF expression in this cohort of animals a function of PTU dose in both cortex and hippocampus however showed an increase in expression with age, a (Fig. 7, Lasley and Gilbert, 2011), changes that were not pattern in direct contrast with numerous reports on the reflected in mRNA (unpublished observations). The normal ontogeny of BDNF in early brain development in bases for this unusual dose–effect pattern, observed long rodents (e.g., Das et al., 2001; Kim et al., 2007). after PTU administration had ceased and hormone levels Possible epigenetic modifications were suggested to had returned to normal, are unknown. The dose– underlie these changes with up-regulation of DNA response characteristics of the expression pattern of methyltransferase activity and down-regulation of histone BDNF do not align well with any functional measures we acetylation prior to weaning associated with the have observed in the hippocampus of similarly treated decreases in BDNF gene and protein expression. animals (see above). What these findings do suggest, Further support for these observations was obtained however, is that the dosimetry of alterations in BDNF by Wang et al. (2012). A complex model of maternal protein expression is complex, the changes are post- hypothyroidism was developed via thyroidectomies translational in nature and are not dependent on an performed in rats prior to mating, and the dams existing state of hypothyroidism, and that these changes received partial or full replacement doses of T4 during in the adult may be mediated by epigenetic mechanisms gestation. On PN3, offspring of dams given no or partial (Lasley and Gilbert, 2011). replacement doses of T4 had 30–35% decreases in Opazo et al. (2008) did not see significant reductions in hippocampal BDNF protein compared to controls that BDNF protein in adult offspring of dams exposed were not surgically altered. Full replacement doses of transiently to a high dose of MMI (200 ppm) in late T4 initiated later during gestation also further gestation, but the data did trend toward a decline in the ameliorated the reductions in BDNF. hippocampus. Although the concentration of MMI may In contrast, Alvarez-Dolado et al. (1994) observed no seem relatively high in this study, the duration of change in BDNF mRNA in the hippocampus, cortex, or exposure was very brief so the level of hypothyroidism in cerebellum at PN15 or PN90 in response to severe the dam was relatively mild and transient. The hypothyroidism induced by a high dose of MMI persistence of BDNF declines into adulthood following (200 ppm) given to the dams beginning on GD9, and transient exposure limited to early development and coupled with thyroidectomy of the pups on PN5. following moderate levels of TH insufficiency are Decreases in NGF mRNA were seen and are consistent consistent with the results of Lasley and Gilbert (2011) with those reported by Neveu and Arenas (1996), but described above. unlike that report, increases rather than decreases in In a different and more complex model of subclinical NT-3 mRNA were observed in the adults hypothyroidism, Liu et al. (2010) evaluated pups born to thyroidectomized as pups. Increases in NT-3 mRNA also thyroidectomized dams. In one group, severe TH stand in contrast to the observations of Lindholm et al. deficiency was evidenced by dam serum hormone (1993) who also reported a 50% decrease in NT-3 concentrations below the level of detection throughout mRNA expression in severely hypothyroid rats that could pregnancy as well as in pups at birth, while dam serum be reversed with T3 administration. TSH was elevated 30- to 60-fold. Under these In more recent studies, Bastian et al. (2010) found no conditions, hippocampal BDNF protein and mRNA were difference in the expression of whole brain transcripts of reduced in offspring on PN3, PN7, and PN21. T4 BDNF exon IV in PN12 rat pups exposed in utero from supplementation throughout pregnancy created a GD6 to a much lower dose of PTU (10 ppm), but one transient state of hyperthyroidism, resulting in elevated which still resulted in a rather severe state of postnatal serum T4 30–70% above control levels in mid-pregnancy hypothyroidism (see Gilbert and Sui, 2006). In a followup (GD10 and GD13), followed by a modest degree of study, a moderate degree of hypothyroidism induced by hypothyroidism in dams during late gestation and in pups a 3 ppm concentration of PTU in the drinking water of at birth (30% reduction in T4). TSH was elevated >10- pregnant dams that reduced T3 and T4 > 50% and 95%, fold in newborn pups and in dams in early gestation, and respectively, did not alter BDNF mRNA or transcripts of more modestly in dams on GD17 (4-fold). In contrast BDNF exons IV or VI in the hippocampus or cortex of to non-supplemented dams with severe hypothyroidism, PN10 hypothyroid brains (Bastian et al., 2012). Neither significant reductions in hippocampal BDNF mRNA and did Royland et al. (2008) observe changes in expression protein expression under these more modest of BDNF transcript in PN14 cortex or hippocampus in a hypothyroid conditions were limited to pups sacrificed on microarray study with varying doses of PTU (0, 1, 2 and PN3 and PN7, with no difference from control on PN21 3 ppm) delivered to dams beginning on GD6. Consistent (Liu et al., 2010). Limitation of disrupted neurotrophin with these observations, no changes in BDNF protein expression to very early times in the postnatal period concentrations were seen in the hippocampus, cortex, or following more modest perinatal hypothyroidism is

Fig. 7. BDNF protein expression was not altered in the (A) cortex, hippocampus or cerebellum of the neonate on PN14 or PN21 following exposure to PTU that began in utero. (B) Reductions in protein expression, however, emerged in adulthood and exhibited a biphasic dose–response profile. Effects were somewhat more prominent in the hippocampus and cortex and in males rather than females. No changes were detected in the cerebellum at neonatal or adult time points. Values are expressed as percent control. ⁄p < 0.05 compared to the control group of the same gender. Marginal effects seen in female hippocampus at 10 ppm and cortex at 2 ppm are depicted as +p = 0.06 when contrasted to control females (taken from Lasley and Gilbert, 2011). consistent with a recent report by Chakraborty et al. (2012) SUMMARY AND CONCLUSIONS using developmental PTU exposure (4 ppm). Whether declines in BDNF protein concentrations paralleling the For the most part investigations of alterations in effects of Lasley and Gilbert (2011) and Opazo et al. neurotrophins occurring as a result of developmental (2008) would re-emerge in adulthood was not established. hypothyroidism have been conducted under conditions

Please cite this article in press as: Gilbert ME, Lasley SM. Developmental thyroid hormone insufficiency and brain development: A role for brain-derived neurotrophic factor (BDNF)? Neuroscience (2013), http://dx.doi.org/10.1016/j.neuroscience.2012.11.022 14 M. E. Gilbert, S. M. Lasley / Neuroscience xxx (2013) xxx–xxx of severe maternal hypothyroidism with assessments Acknowledgments—This document has been subjected to re- taken during the preweaning period when offspring view by the National Health and Environmental Effects Research were still TH-deficient. With the exception of Liu et al. Laboratory and approved for publication. Approval does not sig- (2010) and Chakraborty et al. (2012), in the handful of nify that the contents reflect the views of the Agency, nor does studies in which less than severe hypothyroidism was mention of trade names or commercial products constitute endorsement or recommendation for use. We thank Drs. William invoked, little or no change in expression of BDNF or Mundy and David Sharlin for their critiques on an earlier version other neurotrophins was detected in exposed neonates. of this manuscript, and Keith Tarpley and Alan Tennant for assis- When present, effects of hypothyroidism were to tance with graphic images. reduce expression of BDNF during the fetal (Pathak et al., 2011) an early postnatal period (Liu et al., 2010; Chakraborty et al., 2012), and effects were regionally REFERENCES specific. Our work with moderate TH insufficiency has revealed modest reductions in BDNF mRNA in Akaike M, Kato N, Ohno H, Kobayashi T (1991) Hyperactivity and neonates at PN14 (unpublished observations) that spatial maze learning impairment of adult rats with temporary were not accompanied by alterations in protein neonatal hypothyroidism. Neurotoxicol Teratol 13:317–322. expression (Lasley and Gilbert, 2011). A delayed Allen SJ, Watson JJ, Dawbarn D (2011) The neurotrophins and their reduction in BDNF protein expression in the adult role in Alzheimer’s disease. Curr Neuropharmacol 9:559–573. offspring, in the absence of mRNA alteration, is Alvarez-Dolado M, Iglesias T, Rodrı´guez-Pen˜ a A, Bernal J, Mun˜ oz A (1994) Expression of neurotrophins and the trk family of suggestive of an indirect or epigenetic modification of neurotrophin receptors in normal and hypothyroid rat brain. BDNF following developmental TH insufficiency. Brain Res Mol Brain Res 27:249–257. However, given the importance of TH in fetal brain Alzoubi KH, Aleisa AM, Alkadhi KA (2007) Nicotine prevents development (Bernal, 2011), further exploration of the disruption of the late phase LTP-related molecular cascade in dose–response characteristics and ontogeny at earlier adult-onset hypothyroidism. Hippocampus 17:654–664. time points and lower degrees of thyroid disruption Ambrogini P, Cuppini R, Ferri P, Mancini C, Ciaroni S, Voci A, may be instructive. Gerdoni E, Gallo G (2005) Thyroid hormones affect neurogenesis in the dentate gyrus of adult rat. Neuroendocrinology 81:244–253. To conclude, on review of the available literature, no Anderson GW, Schoonover CM, Jones SA (2003) Control of thyroid clear consensus can be reached on the effects of TH on hormone action in the developing rat brain. Thyroid 13: BDNF expression in the brain. The results using 1039–1056. hypothyroid models have varied from declines, to no Auso E, Lavado-Autric R, Cuevas E, Del Rey FE, Morreale de significant alteration, to BDNF induction. Differences in Escobar G, Berbel P (2004) A moderate and transient deficiency the models used, doses, regions assessed, age of of maternal thyroid function at the beginning of fetal neocorticogenesis alters neuronal migration. Endocrinology 145: assessment, endpoint assessed (mRNA vs protein), and 4037–4047. method employed to determine differences (microarray, Axelstad M, Hansen PR, Boberg J, Bonnichsen M, Nellemann C, rt-PCR, in situ hybridization, Western blots, ELISA) may Lund SP, Hougaard KS, Hass U (2008) Developmental all contribute to a lack of consistency across studies and neurotoxicity of propylthiouracil (PTU) in rats: relationship obfuscate any current synthesis of finding. The between transient hypothyroxinemia during development and complexity of this endocrine system exemplified by the long-lasting behavioural and functional changes. Toxicol Appl Pharmacol 232:1–13. time- and region-specific requirements for TH over the Babu H, Ramirez-Rodriguez G, Fabel K, Bischofberger J, developmental period, the diversity of regulatory Kempermann G (2009) Synaptic network activity induces mechanisms that control presentation of the ligand to neuronal differentiation of adult hippocampal precursor cells the nuclear receptor, and the non-genomic effects through BDNF signaling. Front Neurosci 3:49. described for TH and their metabolites further obscures Baccarelli A, Giacomini SM, Corbetta C, Landi MT, Bonzini M, the emergence of a coherent picture. In the search to Consonni D, Grillo P, Patterson DG, Pesatori AC, Bertazzi PA define a unifying mechanism of neurodevelopmental (2008) Neonatal thyroid function in Seveso 25 years after maternal exposure to dioxin. PLoS Med 5:e161. deficits induced by TH insufficiency, and in the Bastian TW, Anderson JA, Fretham SJ, Prohaska JR, Georgieff MK, regulatory context to identify a sensitive biomarker of Anderson GW (2012) Fetal and neonatal iron deficiency reduces effect, the current case for BDNF is complicated and thyroid hormone-responsive gene mRNA levels in the neonatal rat less than compelling. hippocampus and cerebral cortex. Endocrinology 153: Future work using models of moderate TH 5668–5680. insufficiency, examining dose–response relationships, Bastian TW, Prohaska JR, Georgieff MK, Anderson GW (2010) Perinatal iron and copper deficiencies alter neonatal rat circulating with concurrent gene and protein determinations over a and brain thyroid hormone concentrations. Endocrinology broad developmental time window that encompasses 151:4055–4065. the fetal period may serve to elucidate the relationship Berbel P, Navarro D, AusoÉ, Varea E, Rodrı´guez AE, Ballesta JJ, between TH and neurotrophins. Furthermore, because Salinas M, Flores E, Faura CC, de Escobar GM (2010) Role of BDNF involvement in network circuit formation and late maternal thyroid hormones in cerebral cortex development: synaptic plasticity is rapidly regulated by neuronal an experimental model for human prematurity. Cereb Cortex 20:1462–1475. activation, the role of BDNF in TH-mediated neuronal Berbel P, Mestre JL, Santamarıá A, Palazoń I, Franco A, Graells M, development and function may be more evident when Gonza´lez-Torga A, de Escobar GM (2009) Delayed studied under conditions of BDNF induction in response neurobehavioral development in children born to pregnant to synaptic activation, plasticity, enrichment, or injury women with mild hypothyroxinemia during the first month of (e.g., Bramham et al., 1996; Shulga et al., 2009). gestation: the importance of early iodine supplementation. Thyroid 19:511–519.

Berghuis P, Dobszay MB, Sousa KM, Schulte G, Mager PP, Ha¨ rtig brain-derived neurotrophic factor and neurotrophin-3 mRNA and W, Go¨ rcs TJ, Zilberter Y, Ernfors P, Harkany T (2004) Brain- protein levels in developing regions of rat brain. Neuroscience derived neurotrophic factor controls functional differentiation and 103:739–761. microcircuit formation of selectively isolated fast-spiking Davenport JW, Dorcey TP (1972) Hypothyroidism: learning deficit GABAergic interneurons. Eur J Neurosci 20:1290–1306. induced in rats by early exposure to thiouracil. Horm Behav Bernal J (2002) Action of thyroid hormone in brain. J Endocrinol 3:97–112. Invest 25:268–288. Davis PJ, Leonard JL, Davis FB (2008) Mechanisms of nongenomic Bernal J (2011) Thyroid hormone transport in developing brain. Curr actions of thyroid hormone. Front Neuroendocrinol 29:211–218. Opin Endocrinol Diab Obes 18:295–299. Derksen-Lubsen G, Verkerk PH (1996) Neuropsychologic Bernal J, Morte B (2012) Thyroid hormone receptor activity in the development in early treated congenital hypothyroidism: absence of ligand: physiological and developmental implications. analysis of literature data. Pediatr Res 39:561–566. Biochim Biophys Acta, in press. Desouza LA, Ladiwala U, Daniel SM, Agashe S, Vaidya RA, Vaidya Blount BC, Pirkle JL, Osterloh JD, Valentin-Blasini L, Caldwell KL VA (2005) Thyroid hormone regulates hippocampal neurogenesis (2006) Urinary perchlorate and thyroid hormone levels in in the adult rat brain. Mol Cell Neurosci 29:414–426. adolescent and adult men and women living in the United Dussault JH, Ruel J (1987) Thyroid hormones and brain States. Environ Health Perspect 114:1865–1871. development. Annu Rev Physiol 49:321–334. Bramham CR, Southard T, Sarvey JM, Herkenham M, Brady LS Farwell AP, Dubord-Tomasetti SA, Pietrzykowski AZ, Stachelek SJ, (1996) Unilateral LTP triggers bilateral increases in hippocampal Leonard JL (2005) Regulation of cerebellar neuronal migration neurotrophin and trk receptor mRNA expression in behaving rats: and neurite outgrowth by thyroxine and 3,30,50-triiodothyronine. evidence for interhemispheric communication. J Comp Neurol Dev Brain Res 154(1):121–135. 368:371–382. Ferreira AA, Pereira MJ, Manha˜ es AC, Barradas PC (2007) Brucker-Davis F (1998) Effects of environmental synthetic chemicals Ultrastructural identification of oligodendrocyte/myelin proteins in on thyroid function. Thyroid 8:827–856. corpus callosum of hypothyroid animals. Int J Dev Neurosci Caldwell KK, Sheema S, Paz RD, Samudio-Ruiz SL, Laughlin MH, 25:87–94. Spence NE, Roehlk MJ, Alcon SN, Allan AM (2008) Fetal alcohol Flamant F, Samarut J (2003) Thyroid hormone receptors: lessons spectrum disorder-associated depression: evidence for from knockout and knock-in mutant mice. Trends Endocrinol reductions in the levels of brain-derived neurotrophic factor in a Metab 14:85–90. mouse model. Pharmacol Biochem Behav 90:614–624. Gerges NZ, Alkadhi KA (2004) Hypothyroidism impairs late LTP in Camboni D, Roskoden T, Schwegler H (2003) Effect of early CA1 region but not in dentate gyrus of the intact rat hippocampus: thyroxine treatment on brain-derived neurotrophic factor mRNA MAPK involvement. Hippocampus 14:40–45. expression and protein amount in the rat medial septum/diagonal Gilbert ME (2004) Alterations in synaptic transmission and plasticity band of Broca. Neurosci Lett 350:141–144. in area CA1 of adult hippocampus following developmental Chakraborty G, Magagna-Poveda A, Parratt C, Umans JG, hypothyroidism. Brain Res Dev Brain Res 148:11–18. MacLusky NJ, Scharfman HE (2012) Reduced hippocampal Gilbert ME (2011) Impact of low-level thyroid hormone disruption brain-derived neurotrophic factor (BDNF) in neonatal rats after induced by propylthiouracil on brain development and function. prenatal exposure to propylthiouracil (PTU). Endocrinology Toxicol Sci 124:432–445. 153:1311–1316. Gilbert ME, Hedge JM, Valentıń-Blasini L, Blount BC, Kannan K, Chan S, Franklyn JA, Kilby MD (2005) Maternal thyroid hormones Tietge J, Zoeller RT, Crofton KM, Jarrett JM, Fisher JW (2012) An and fetal brain development. Curr Opin Endocrinol Diab animal model of marginal iodine deficiency during development: 12:23–30. the thyroid axis and neurodevelopmental outcome. Toxicol Sci, in Chen Y, Zeng J, Cen L, Chen Y, Wang X, Yao G, Wang W, Qi W, press. Kong K (2009) Multiple roles of the p75 neurotrophin receptor in Gilbert ME, Paczkowski C (2003) Propylthiouracil (PTU)-induced the nervous system. J Int Med Res 37(2):281–288. hypothyroidism in the developing rat impairs synaptic Chevrier J, Eskenazi B, Bradman A, Fenster L, Barr DB (2007) transmission and plasticity in the dentate gyrus of the adult Associations between prenatal exposure to polychlorinated hippocampus. Brain Res Dev Brain Res 145(1):19–29. biphenyls and neonatal thyroid-stimulating hormone levels in a Gilbert ME, Sui L (2006) Dose-dependent reductions in spatial Mexican-American population, Salinas Valley, California. Environ learning and synaptic function in the dentate gyrus of adult rats Health Perspect 115:1490–1496. following developmental thyroid hormone insufficiency. Brain Res Choi SH, Li Y, Parada LF, Sisodia SS (2009) Regulation of 1069:10–22. hippocampal progenitor cell survival, proliferation and dendritic Gilbert ME, Sui L (2008) Developmental exposure to perchlorate development by BDNF. Mol Neurodegener 4:52. alters synaptic transmission in hippocampus of the adult rat. Cooper DS, Kieffer JD, Halpern R, Saxe V, Mover H, Maloof F, Environ Health Perspect 116:752–760. Ridgway EC (1983) Propylthiouracil (PTU) pharmacology in the Gilbert ME, Sui L, Walker MJ, Anderson W, Thomas S, Smoller SN, rat: II. Effects of PTU on thyroid function. Endocrinology Schon JP, Phani S, Goodman JH (2007) Thyroid hormone 113:921–928. insufficiency during brain development reduces parvalbumin Cooper DS, Kieffer JD, Saxe V, Mover H, Maloof F, Ridgway EC immunoreactivity and inhibitory function in the hippocampus. (1984) Methimazole pharmacology in the rat: studies using a Endocrinology 148:92–102. newly developed radioimmunoassay for methimazole. Gilbert ME, Taylor MM (2012) Mild developmental hypothyroidism Endocrinology 114:786–793. and trace fear conditioning: role of gender and shock duration. Corte´s C, Eugenin E, Aliaga E, Carren˜ o LJ, Bueno SM, Gonzalez PA, Toxicologist 126:421. Gayol S, Naranjo D, Noches V, Marassi MP, Rosenthal D, Jadue Gilbert ME, Zoeller RT (2010) Thyroid hormone – impact on the C, Ibarra P, Keitel C, Wohllk N, Court F, Kalergis AM, Riedel CA developing brain: possible mechanisms of neurotoxicity. In: Harry (2012) Hypothyroidism in the adult rat causes incremental GJ, Tilson HA, editors. Neurotoxicology. New York: Informa changes in brain-derived neurotrophic factor, neuronal and Healthcare USA, Inc. p. 79–111. astrocyte apoptosis, gliosis, and deterioration of postsynaptic Ginsberg GL, Hattis DB, Zoeller RT, Rice DC (2007) Evaluation of the density. Thyroid 22:951–963. U.S. EPA/OSWER preliminary remediation goal for perchlorate in Darbra S, Balada F, Martı´-Carbonell MA, Garau A (2004) Perinatal groundwater: focus on exposure to nursing infants. Environ hypothyroidism effects on step-through passive avoidance task in Health Perspect 115:361–369. rats. Physiol Behav 82:497–501. Giordano T, Pan JB, Casuto D, Watanabe S, Arneric SP (1992) Das KP, Chao SL, White LD, Haines WT, Harry GJ, Tilson HA, Thyroid hormone regulation of NGF, NT-3 and BDNF RNA in the Barone Jr S (2001) Differential patterns of nerve growth factor, adult rat brain. Mol Brain Res 16:239–245.

Glinoer D (2007) Clinical and biological consequences of iodine Kim SY, Smith MA, Post RM, Rosen JB (1998) Attenuation of deficiency during pregnancy. Endocr Dev 10:62–85. kindling-induced decreases in NT-3 mRNA by thyroid hormone Glinoer D, Delange F (2000) The potential repercussions of maternal, depletion. Epilepsy Res 29:211–220. fetal, and neonatal hypothyroxinemia on the progeny. Thyroid Koibuchi N, Yamaoka S, Chin WW (2001) Effect of altered thyroid 10:871–887. status on neurotrophin gene expression during postnatal Goodman JH, Gilbert ME (2007) Modest thyroid hormone development of the mouse cerebellum. Thyroid 11:205–210. insufficiency during development induces a cellular malformation Koibuchi N, Fukuda H, Chin WW (1999) Promoter-specific regulation in the corpus callosum: a model of cortical dysplasia. of the brain-derived neurotrophic factor gene by thyroid hormone Endocrinology 148:2593–2597. in the developing rat cerebellum. Endocrinology 140: Gould E, Allan MD, McEwen BS (1990a) Dendritic spine density of 3955–3961. adult hippocampal pyramidal cells is sensitive to thyroid hormone. Koibuchi N, Jingu H, Iwasaki T, Chin WW (2003) Current Brain Res 525:327–329. perspectives on the role of thyroid hormone in growth and Gould E, Butcher LL (1989) Developing cholinergic basal forebrain development of cerebellum. Cerebellum 2(4):279–289. neurons are sensitive to thyroid hormone. J Neurosci Koopman-Esseboom C, Morse DC, Weisglas-Kuperus N, 9:3347–3358. Lutkeschipholt IJ, Van der Paauw CG, Tuinstra LG, Brouwer A, Gould E, Westlind-Danielsson A, Frankfurt M, McEwen BS (1990b) Sauer PJ (1994) Effects of dioxins and polychlorinated biphenyls Sex differences and thyroid hormone sensitivity of hippocampal on thyroid hormone status of pregnant women and their infants. pyramidal cells. J Neurosci 10:996–1003. Pediatr Res 36(4):468–473. Gravel C, Hawkes R (1990) Maturation of the corpus callosum of the Koopman-Esseboom C, Weisglas-Kuperus N, de Ridder MA, Van der rat: I. Influence of thyroid hormones on the topography of callosal Paauw CG, Tuinstra LG, Sauer PJ (1996) Effects of projections. J Comp Neurol 291:128–146. polychlorinated biphenyl/dioxin exposure and feeding type on Hackenmueller SA, Marchini M, Saba A, Zucchi R, Scanlan TS infants’ mental and psychomotor development. Pediatrics (2012) Biosynthesis of 3-iodothyronamine (T1AM) is dependent 97:700–706. on the sodium-iodide symporter and thyroperoxidase but does not Kumar A, Sinha RA, Tiwari M, Pal L, Shrivastava A, Singh R, Kumar involve extrathyroidal metabolism of T4. Endocrinology K, Kumar Gupta S, Godbole MM (2006) Increased pro-nerve 153:5659–5667. growth factor and p75 neurotrophin receptor levels in developing Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight GJ, hypothyroid rat cerebral cortex are associated with enhanced Gagnon J, O’Heir CE, Mitchell ML, Hermos RJ, Waisbren SE, apoptosis. Endocrinology 147:4893–4903. Faix JD, Klein RZ (1999) Maternal thyroid deficiency during Lasley SM, Gilbert ME (2011) Developmental thyroid hormone pregnancy and subsequent neuropsychological development of insufficiency reduces expression of brain-derived neurotrophic the child. N Engl J Med 341:549–555. factor (BDNF) in adults but not in neonates. Neurotoxicol Teratol Henrichs J, Bongers-Schokking JJ, Schenk JJ, Ghassabian A, 33:464–472. Schmidt HG, Visser TJ, Hooijkaas H, de Muinck Keizer- Lavado-Autric R, AusoÉ, Garcıá-Velasco JV, Arufe Mdel C, Escobar Schrama SM, Hofman A, Jaddoe VV, Visser W, Steegers EA, del Rey F, Berbel P, Morreale de Escobar G (2003) Early Verhulst FC, de Rijke YB, Tiemeier H (2010) Maternal thyroid maternal hypothyroxinemia alters histogenesis and cerebral function during early pregnancy and cognitive functioning in early cortex cytoarchitecture of the progeny. J Clin Invest childhood: the generation R study. J Clin Endocrinol Metab 111:1073–1082. 95:4227–4234. Lazarini F, Lledo PM (2011) Is adult neurogenesis essential for Hernandez A, Morte B, Belinchoń MM, Ceballos A, Bernal J (2012) olfaction? Trends Neurosci 34:20–30. Critical role of types 2 and 3 deiodinases in the negative Lemkine GF, Raj A, Alfama G, Turque N, Hassani Z, Alegria-Pre´vot

regulation of gene expression by T3 in the mouse cerebral O, Samarut J, Levi G, Demeneix BA (2005) Adult neural stem cell cortex. Endocrinology 153:2919–2928. cycling in vivo requires thyroid hormone and its alpha receptor. Hernandez A, Quignodon L, Martinez ME, Flamant F, St Germain DL FASEB J 19:863–865. (2010) Type 3 deiodinase deficiency causes spatial and temporal Lewin GR, Barde YA (1996) Physiology of the neurotrophins. Annu alterations in brain T3 signaling that are dissociated from serum Rev Neurosci 19:289–317. thyroid hormone levels. Endocrinology 151:5550–5558. Lindholm D, Castreń E, Tsoulfas P, Kolbeck R, Berzaghi MP, Heuer H, Visser TJ (2009) Minireview: pathophysiological importance Leinga¨ rtner A, Heisenberg CP, Tessarollo L, Parada LF, of thyroid hormone transporters. Endocrinology 150:1078–1083. Thoenen H (1993) Neurotrophin-3 induced by tri-iodothyronine Ibarrola N, Rodriguez-Pena A (1997) Hypothyroidism coordinately in cerebellar granule cells promotes Purkinje cell differentiation. J and transiently affects myelin protein gene expression in most rat Cell Biol 122:443–450. brain regions during postnatal development. Brain Res Liu D, Teng W, Shan Z, Yu X, Gao Y, Wang S, Fan C, Wang H, 752:285–293. Zhang H (2010) The effect of maternal subclinical hypothyroidism Iniguez MA, De Lecea L, Guadano-Ferraz A, Morte B, Gerendasy D, during pregnancy on brain development in rat offspring. Thyroid Sutcliffe JG, Bernal J (1996) Cell-specific effects of thyroid 20:909–915. hormone on RC3/neurogranin expression in rat brain. Lonky E, Reihman J, Darvill T, Mather Sr J, Daly H (1996) Neonatal Endocrinology 137:1032–1041. behavioral assessment scale performance in humans influenced Jacobson JL, Jacobson SW (1996) Intellectual impairment in children by maternal consumption of environmentally contaminated Lake exposed to polychlorinated biphenyls in utero. N Engl J Med Ontario fish. J Great Lakes Res 22:198–212. 335:783–789. Lu B, Figurov A (1997) Role of neurotrophins in synapse Jagannathan NR, Tandon N, Raghunathan P, Kochupillai N (1998) development and plasticity. Rev Neurosci 8:1–12. Reversal of abnormalities of myelination by thyroxine therapy in Lu¨ esse HG, Roskoden T, Linke R, Otten U, Heese K, Schwegler H congenital hypothyroidism: localized in vivo proton magnetic (1998) Modulation of mRNA expression of the neurotrophins of resonance spectroscopy (MRS) study. Brain Res Dev Brain Res the nerve growth factor family and their receptors in the septum 109:179–186. and hippocampus of rats after transient postnatal thyroxine Kapoor R, Desouza LA, Nanavaty IN, Kernie SG, Vaidya VA (2012) treatment: I. Expression of nerve growth factor, brain-derived Thyroid hormone accelerates the differentiation of adult neurotrophic factor, neurotrophin-3, and neurotrophin-4 mRNA. hippocampal progenitors. J Neuroendocrinol 24:1259–1271. Exp Brain Res 119:1–8. Kim JK, Jeon SM, Lee KM, Park ES, Cho HJ (2007) Expression of Madeira MD, Cadete-Leite A, Andrade JP, Paula-Barbosa MM (1991) brain-derived neurotrophic factor in the rat forebrain and upper Effects of hypothyroidism upon the granular layer of the dentate brain stem during postnatal development: an gyrus in male and female adult rats: a morphometric study. J immunohistochemical study. Neuroscience 146:1128–1136. Comp Neurol 314:171–186.

Madeira MD, Sousa N, Lima-Andrade MT, Calheiros F, Cadete-Leite Saegusa Y, Woo GH, Fujimoto H, Kemmochi S, Shimamoto K, A, Paula-Barbosa MM (1992) Selective vulnerability of the Hirose M, Mitsumori K, Nishikawa A, Shibutani M (2010) hippocampal pyramidal neurons to hypothyroidism in male and Sustained production of Reelin-expressing interneurons in the female rats. J Comp Neurol 322:501–518. hippocampal dentate hilus after developmental exposure to anti- Martinowich K, Manji H, Lu B (2007) New insights into BDNF function thyroid agents in rats. Reprod Toxicol 29:407–414. in depression and anxiety. Nat Neurosci 10:1089–1093. Santisteban P, Bernal J (2005) Thyroid development and effect on Mayer S, Visser TJ, Darras VM, Horn S, Heuer H (2012a) Impact of the nervous system. Rev Endocr Metab Disord 6:217–228. Oatp1c1 deficiency on thyroid hormone metabolism and action in Scanlan TS, Suchland KL, Hart ME, Chiellini G, Huang Y, Kruzich PJ, the mouse brain. Endocrinology 153:1528–1537. Frascarelli S, Crossley DA, Bunzow JR, Ronca-Testoni S, Lin ET, Mayer S, Visser TJ, Bauer R, Darras VM, Heuer H (2012b) Hatton D, Zucchi R, Grandy DK (2004) 3-Iodothyronamine is an Consequences of brain-specific thyroid hormone deprivation in endogenous and rapid-acting derivative of thyroid hormone. Nat MCT8/OATP1c1 double knockout mice. Thyroid 22(S1). O-2. Med 10:638–642. Mendez Jr W, Eftim SE (2012) Biomarkers of perchlorate exposure Schoonover CM, Seibel MM, Jolson DM, Stack MJ, Rahman RJ, are correlated with circulating thyroid hormone levels in the 2007– Jones SA, Mariash CN, Anderson GW (2004) Thyroid hormone 2008 NHANES. Environ Res 118:137–144. regulates oligodendrocyte accumulation in developing rat brain Montero-Pedrazuela A, Venero C, Lavado-Autric R, Fernańdez-Lamo white matter tracts. Endocrinology 145:5013–5020. I, Garcıá-Verdugo JM, Bernal J, Guadan˜ o-Ferraz A (2006) Schweizer U, Ko¨ hrle J (2012) Function of thyroid hormone Modulation of adult hippocampal neurogenesis by thyroid transporters in the central nervous system. Biochim Biophys hormones: implications in depressive-like behavior. Mol Acta, in press. Psychiatry 11:361–371. Sharlin DS, Tighe D, Gilbert ME, Zoeller RT (2008) The balance Neveu I, Arenas E (1996) Neurotrophins promote the survival and between oligodendrocyte and astrocyte production in major white development of neurons in the cerebellum of hypothyroid rats matter tracts is linearly related to serum total thyroxine. in vivo. J Cell Biol 133:631–646. Endocrinology 149:2527–2536. Oh JD, Butcher LL, Woolf NJ (1991) Thyroid hormone modulates the Shulga A, Blaesse A, Kysenius K, Huttunen HJ, Tanhuanpa¨ a¨ K, development of cholinergic terminal fields in the rat forebrain: Saarma M, Rivera C (2009) Thyroxin regulates BDNF expression to relation to nerve growth factor receptor. Brain Res Dev Brain Res promote survival of injured neurons. Mol Cell Neurosci 42:408–418. 59:133–142. Sinha RA, Pathak A, Kumar A, Tiwari M, Shrivastava A, Godbole MM Opazo MC, Gianini A, Pancetti F, Azkcona G, Alarcoń L, Lizana R, (2009) Enhanced neuronal loss under perinatal hypothyroidism Noches V, Gonzalez PA, Marassi MP, Mora S, Rosenthal D, involves impaired neurotrophic signaling and increased Eugenin E, Naranjo D, Bueno SM, Kalergis AM, Riedel CA (2008) proteolysis of p75(NTR). Mol Cell Neurosci 40:354–364. Maternal hypothyroxinemia impairs spatial learning and synaptic Sui L, Anderson WL, Gilbert ME (2005) Impairment in short-term but nature and function in the offspring. Endocrinology 149: enhanced long-term synaptic potentiation and ERK activation in 5097–5106. adult hippocampal area CA1 following developmental thyroid Oppenheimer JH, Schwartz HL (1997) Molecular basis of thyroid hormone insufficiency. Toxicol Sci 85:647–656. hormone-dependent brain development. Endocr Rev 18:462–475. Sui L, Gilbert ME (2003) Pre- and postnatal propylthiouracil-induced Patandin S, Lanting CI, Mulder PG, Boersma ER, Sauer PJ, hypothyroidism impairs synaptic transmission and plasticity in Weisglas-Kuperus N (1999) Effects of environmental exposure area CA1 of the neonatal rat hippocampus. Endocrinology to polychlorinated biphenyls and dioxins on cognitive abilities in 144:4195–4203. Dutch children at 42 months of age. J Pediatr 134:33–41. Sui L, Li BM (2010) Effects of perinatal hypothyroidism on regulation Pathak A, Sinha RA, Mohan V, Mitra K, Godbole MM (2011) Maternal of reelin and brain-derived neurotrophic factor gene expression in thyroid hormone before the onset of fetal thyroid function rat hippocampus: role of DNA methylation and histone regulates reelin and downstream signaling cascade affecting acetylation. Steroids 75:988–997. neocortical neuronal migration. Cereb Cortex 21:11–21. Sui L, Ren WW, Li BM (2010) Administration of thyroid hormone Poguet AL, Legrand C, Feng X, Yen PM, Meltzer P, Samarut J, increases reelin and brain-derived neurotrophic factor expression Flamant F (2003) Microarray analysis of knockout mice identifies in rat hippocampus in vivo. Brain Res 1313:9–24. cyclin D2 as a possible mediator for the action of thyroid hormone Takahashi M, Negishi T, Tashiro T (2008) Identification of genes during the postnatal development of the cerebellum. Dev Biol mediating thyroid hormone action in the developing mouse 254:188–199. cerebellum. J Neurochem 104:640–652. Porterfield SP (1994) Vulnerability of the developing brain to thyroid Thompson CC, Potter GB (2000) Thyroid hormone action in neural abnormalities: environmental insults to the thyroid system. development. Cereb Cortex 10:939–945. Environ Health Perspect 102(Suppl. 2):125–130. Turyk ME, Anderson HA, Persky VW (2007) Relationships of thyroid Powell MH, Nguyen HV, Gilbert M, Parekh M, Colon-Perez LM, hormones with polychlorinated biphenyls, dioxins, furans, and Mareci TH, Montie E (2012) Magnetic resonance imaging and DDE in adults. Environ Health Perspect 115:1197–1203. volumetric analysis: novel tools to study the effects of thyroid Vaidya VA, Castro ME, Pei Q, Sprakes ME, Grahame-Smith DG hormone disruption on white matter development. (2001) Influence of thyroid hormone on 5-HT(1A) and 5-HT(2A) Neurotoxicology 33:1322–1329. receptor-mediated regulation of hippocampal BDNF mRNA Quignodon L, Legrand C, Allioli N, Guadan˜ o-Ferraz A, Bernal J, expression. Neuropharmacology 40:48–56. Samarut J, Flamant F (2004) Thyroid hormone signaling is highly vonDran MW, Clinton-Luke P, Honeywell JZ, Dreyfus CF (2010) heterogeneous during pre- and postnatal brain development. J BDNF+/ mice exhibit deficits in oligodendrocyte lineage cells of Mol Endocrinol 33:467–476. the basal forebrain. Glia 7:848–856. Rami A, Patel AJ, Rabie A (1986) Thyroid hormone and development Walker P, Weichsel Jr ME, Eveleth D, Fisher DA (1982) Ontogenesis of the rat hippocampus: morphological alterations in granule and of nerve growth factor and epidermal growth factor in submaxillary pyramidal cells. Neuroscience 19:1217–1226. glands and nerve growth factor in brains of immature male mice: Roskoden T, Heese K, Otten U, Schwegler H (1999) Modulation of correlation with ontogenesis of serum levels of thyroid hormones. mRNA expression of the neurotrophins of the nerve-growth-factor Pediatr Res 16:520–524. family and their receptors in the septum and hippocampus of rats Walker P, Weichsel Jr ME, Fisher DA, Guo SM, Fisher DA (1979) after transient postnatal thyroxine treatment: II. Effects on p75 Thyroxine increases nerve growth factor concentration in adult and trk receptor expression. Exp Brain Res 127:307–313. mouse brain. Science 204:427–429. Royland JE, Parker JS, Gilbert ME (2008) A genomic analysis of Walker P, Weil ML, Weichsel Jr ME, Fisher DA (1981) Effect of subclinical hypothyroidism in hippocampus and neocortex of the thyroxine on nerve growth factor concentration in neonatal mouse developing rat brain. J Neuroendocrinol 20:1319–1338. brain. Life Sci. 28:1777–1787.

Wang S, Teng W, Gao Y, Fan C, Zhang H, Shan Z (2012) Early Yoshii A, Constantine-Paton M (2010) Postsynaptic BDNF–TrkB levothyroxine treatment on maternal subclinical hypothyroidism signaling in synapse maturation, plasticity, and disease. Dev improves spatial learning of offspring in rats. J Neuroendocrinol Neurobiol 70:304–322. 24:841–848. Zimmermann MB (2007) The adverse effects of mild-to-moderate Williams GR (2008) Neurodevelopmental and neurophysiological iodine deficiency during pregnancy and childhood: a review. actions of thyroid hormone. J Neuroendocrinol 20:784–794. Thyroid 17:829–835. Williams GR, Bassett JH (2011) Deiodinases: the balance of thyroid Zoeller RT, Rovet J (2004) Timing of thyroid hormone action in the hormone: local control of thyroid hormone action – role of type 2 developing brain: clinical observations and experimental findings. deiodinase. J Endocrinol 209:261–272. J Neuroendocrinol 16:809–818.

(Accepted 6 November 2012) (Available online xxxx)