It is now well established that the mammalian brain is a direct target organ of thyroid hormone, both during development and also in adult individuals. Most molecular studies have been directed towards the actions of thyroid hormone during development, and much less is known on its actions in the adult brain. During development, the role of thyroid hormone is the coordination of seemingly unrelated maturational processes. These processes are influenced by the hormone only temporarily during overlapping windows of development with regional specificity. In the brain, as in other systems, the active hormone is T3 which acts by regulating gene expression after binding to specific nuclear receptors. Besides this, there are convincing reports that thyroid hormone also has extra nuclear and extra genomic actions (1), but the extent to which these actions contribute to the general effects of the hormones in the brain is unknown. For details on the physiological and biochemical processes underlying thyroid hormone action in the brain the reader is referred to more extensive recent reviews by the author (2, 3). In this review, I will deal with three important topics, sometimes controversial and which still are not completely settled: what are the cellular targets of thyroid in brain, what is the role of the thyroid hormone receptors, and during which periods of development are thyroid hormones important.

Cellular targets of thyroid hormone in the brain
By situ hybridization analysis, T3 receptor mRNAs are located predominantly, if not exclusively, on neurons (4, 5). Little signal is present in the white matter. T3 receptors in specific neurons likely mediate the effects of the hormone in neuronal cell migration and differentiation, including for example, migration of neurons in the cerebral cortex and in the cerebellum, differentiation of Purkinje cells and cholinergic cells and the control of dendritic spine density in pyramidal cells of the cerebral cortex. T3 receptors are also present in high amounts in neurons in primary culture and in the neuronal line GT1-7 and a number of neuron-specific genes are regulated by T3 directly at the transcriptional level (3). Some of these genes contain thyroid hormone responsive elements (TRE).
It is also beyond doubt that the oligodendrocytes are also cellular targets of thyroid hormone. In vitro, thyroid hormone is required for normal timing of oligodendrocyte differentiation and in vivo thyroid hormone also controls the timing of myelination and the expression of oligodendrocyte-specific genes (6). The myelin basic protein gene promoter contains a TRE, suggesting direct transcriptional regulation. Despite the difficulties in detecting T3 receptor mRNA in brain slice preparations, mature oligodendrocytes in vitro express TRa1 and TRb1, whereas oligodendrocyte precursor cells (OPC) express only TRa1. This receptor isoform mediates therefore the effects of T3 on OPC differentiation (7).
It is controversial whether other cells are direct targets of thyroid hormone. Thyroid hormone has effects on astrocytes in vivo and in vitro; however it is not clear whether these effects are mediated by an action of T3 through the nuclear receptor. There are contradictory reports on the presence of T3 receptors in astrocytes and some evidence has been provided for extra genomic effects of both, T4 and T3 on astrocytes (8). On the other hand, T3 transcriptionally regulates the expression of some astrocyte-specific genes, so that the issue is not entirely settled. The effects of thyroid hormone on other types of cells in the central or peripheral nervous system have not been studied in much detail. Thyroid hormone is needed for proliferation and maturation of microglia (9). Interestingly, these cells in culture express the thyroid hormone receptor isoforms TRa1 and TRb1, but not TRa2. Finally, Schwann cells have been reported to express T3 receptors during development and after nerve regeneration, and T3 acutely increases the expression of early genes in cultured Schwann cells (10).

Role of thyroid hormone receptors
In mammals T3 receptors are the products of two genes known as TRa and TRb. The TRa gene encodes four protein products (TRa1, TRa2, and two truncated products) from which only TRa1 binds T3. The TRb gene encodes four T3 binding proteins, of which TRb1, TRb2 and TRb3 bind also to responsive elements in DNA. In addition, a truncated protein, delta-TRb3 binds T3 but not DNA.
One important question, not entirely settled, is why there are so many receptor isoforms and related proteins. Are the receptors equivalent, or do they regulate different genes and physiological functions? (11). The most prevalent view is that the receptor isoforms are mostly equivalent in their biological activity and that their different physiological role is due to their different patterns of expression and tissue concentrations (12). For example, whereas TRb1 is involved in cochlear development it can be replaced by TRa1 provided it is expressed at sufficiently high levels (13). In the cerebellum TRa1 is expressed in the granular cells, whereas the Purkinje cells express TRb1. Therefore, the effects of T3 on migration if granular cells are mediated by TRa1, whereas those on differentiation of Purkinje cells are mediated by TRb1(14).
A prominent role of TRa1 in brain development and function may be deduced from its relative expression in cerebrum and cerebellum, accounting for about 70-80% of total T3 receptor binding (15). In addition, TRa1 is expressed earlier in development than TRb1. Therefore it was puzzling that TRa1 null mutant mice did not display obvious signs of developmental abnormalities. One possible explanation for this paradox is that in the absence of ligand, transcriptional repression by the unliganded receptor is responsible for the effects of profound hypothyroidism. According to this, we found that when hypothyroidism is induced in TRa1 mutant mice there is no delay in granular cell migration as observed in the hypothyroid wild type mice (14). Also, congenitally hypothyroid, Pax 8-deficient mice, which die during the first weeks of life, can be rescued by TRa1 deletion (16). It appears therefore that many important effects of profound hypothyroidism may in fact be due to the presence of unliganded TRa1. It follows from this conclusion that many of the biochemical processes influenced by thyroid hormone could really take place in the absence of both, the hormone and the receptor. It is likely that the interplay between the repressor activity of the receptor and the derepressor role of the hormone serves to finely tune the coordinating features of thyroid control on developmental processes.

Timing of thyroid hormone action in brain
An important question is whether thyroid hormone is needed throughout all phases of development or there are limited windows of thyroid hormone action. These questions are important in the analysis of the effects of maternal hypothyroidism and hypothyroxinemia on the fetus, and also on prematurity (17). In the rat model, the peak of thyroid hormone sensitivity for the brain, judging from the highest occupancy of T3 receptors would be around postnatal day 15, and most developmental effects of thyroid hormone action in the brain appear to take place during the first three postnatal weeks. Most thyroid hormone-regulated genes identified to date are sensitive to the hormone at different phases within the period corresponding from about E18 to P25. However, there is strong evidence that the rat fetal brain is under thyroid hormone control before that age and, therefore before onset of thyroid gland function. Receptor mRNAs can be detected as early as E11.5 (5), and receptor protein in nuclear preparations of the whole brain is detectable around E14 (18). Of course, in the absence of the fetal thyroid, maternal hormones would play an important role at these early stages of development. In support of this, the progeny of pregnant dams on low iodine diets had permanent changes in the migratory patterns of cells migrating on E14-E16 in the neocortex and hippocampus (19). It is therefore important to dissect the pathways of thyroid hormone action at these early phases of development and identify the target genes mediating these actions (20).
There is still not much data to understand in molecular terms the role of thyroid hormone in the adult brain (21). Thyroidal status influences neurotransmitter systems, but the mechanisms of regulation are unknown. As during development, thyroid hormone influences gene expression (22) and we have shown recently that deletion of TRa1 leads to alterations of behaviour (23). Expression of a mutated version of TRa1 with dominant negative activity leads to dramatic anxiety-like features, which is normalized by T3 treatment (Collaborative work between Vennström’s and Bernal’s groups, to be published). The implication of thyroid hormone receptors in behaviour is important, because it may be possible to modulate behaviour by using highly specific agonists of receptor isoforms. Elucidating the mechanisms of thyroid hormone action in the adult brain, including the role of receptor isoforms in behaviour remain therefore an important open field of research for the near future.

Summarizing table:
Thyroid hormone and brain development