Unraveling the molecular basis for successful thyroid hormone replacement therapy

The thyroid gland with its secreted hormones T4 (thyroxine) and T3 (triiodothyronine) plays a pivotal role in the pre – and postnatal development and is related to growth, neural differentiation and metabolic regulation in mammals as a part of the endocrinologic system. Because of this important role, the release of the thyroid hormones (THs) is tightly regulated. Liberation of hypothalamic thyrotropin releasing hormone (TRH) leads to a release of thyroid stimulating hormone (TSH) from the pituitary, which is recognized by thyrocytes. These follicular thyroid cells respond by secretion of T4 and to a minor degree by T3. In a negative feedback loop thyroid hormones act on hypothalamic TRH secretion and inhibit by an additional direct pituitary interaction the release of TSH from thyrotrophic cells. This regulatory circuit enables mammals to maintain a stable thyroid hormone homeostasis but as well to respond to exo – and endogenous stimuli with increased or decreased thyroid hormone levels. There are different regulatory elements with tissue specific temporal and spatial patterns to provide the regional sensitivity to thyroid hormones, namely transport proteins, deiodinases and thyroid hormone receptors (TRs). Thyroid hormones are specifically transported via the cellular membrane by various families of substrate specific transporters such as MCT8/10 or members of the OATP family and not by passive diffusion, as it was assumed due to their lipophilic chemical structure. Intracellularly, a family of deiodinases can convert T4 which has a low affinity to the nuclear T3 receptor to highly active T3 by cleavage of one iodide. These deiodinases act organ-specific and require cofactors for full activity. The cleavage site determines whether bioactive T3 or potentially less active but generally not well characterized metabolites will result. This is exemplified by the action of deiodinase type 3 (Dio3) which is able to convert T4 to biologically inactive reverse T3 (rT3) and T3 to the less active 3,3’-T2.

This cell and cofactor specific diversity is also reflected on the level of thyroid hormone receptors (TRs) which are divided into two major subclasses, TRα and TRβ, differing in spatial expression. Both act as dimers, which can be formed even with other nuclear receptors, e.g. retinoic acid receptors (RAR), liver X receptor (LXR) or peroxisome proliferator-activated receptor (PPAR) and bind to thyroid hormone response elements (TRE). Binding occurs as well when no thyroid hormones are present, but in that case they act in opposite regulatory direction on the target genes. TRα is primarily expressed in bones, heart, kidney and the CNS and only the isotype TRα1 has a T3 binding site to respond to different TH levels. All the other isoforms presumably act as dominant negative inhibitors. The most common TRβ isotypes are TRβ1 and TRβ2 which are expressed abundantly in liver and kidney (TRβ1) and hypothalamus, pituitary, cochlea and retina (TRβ2). Because of their expression pattern mutations in TRβ are clinically more obvious. Due to the involvement of the regulatory pathway which is dominantly controlled by TRβ patients have high levels of both TSH and thyroid hormones. Tissues bearing the mutation need higher thyroid hormone levels for normal function whereas organs predominantly expressing TRα such as brain, heart or bowel are not protected against the high circulating thyroid hormone levels and are thyrotoxic. This relates to a variety of clinical symptoms such as growth and developmental delay and tachycardia. In contrast, mutations in TRα1 do not affect the regulatory pathways of the thyroid and TSH as well as thyroid hormone levels are only mildly changed in patients with TRα mutations.

The correlation between the metabolic rate of an organism and the level of TH is known for a long time, since patients with severe hypothyroidism showed a 50 % reduced energy exposure. But there are some more metabolic factors influenced by TH, e.g. the uptake of free fatty acids (FFA), triglycerides and cholesterol by white and brown adipose tissue (WAT and BAT), liver and skeletal muscle. BAT is known for a long time from rodents and new born children, where it is responsible for the non-shivering adaptive thermogenesis, a process where the uncoupling protein 1 (UCP-1) bypass the F_oF₁ ATPase in the mitochondria and release the stored energy as heat. Incidentally BAT depots were found in adult humans, but just in a small portion and it became clear that TH is a main regulator of BAT development and maintenance, as well as for UCP-1 expression in BAT, thereby providing an important route of communication between brain and adipose tissue.

To keep the TH within a physiological range is very important for mammals, not only for its metabolic, but also for its cardiovascular and neurological functions. There are several drugs for the treatment of hyperthyroidism (thiamazol, carbimazol or methimazol) and hypothyroidism (levothyroxine), but special clinical situations like central hypothyroidism or mutations in TRβ result in TSH levels (main marker for thyroid gland function) that does not represent the status in special tissues. This ends up in inadequate replacement therapies with sides effects in the patients. Consequently, new biomarkers for TH action are urgently needed.

The aim of the current project is to identify novel biomarkers for thyroid dysfunction and thus for successful therapy. In our preliminary studies (see above) we identified a large number of new thyroid hormone targets including a dysregulation of many metabolically important pathways. By comparing human and mouse data we will be able to define under which circumstances the mouse may serve as an adequate model for hyper- or hypothyroidism. It will then be possible to define changes on the organ level and define the likely source of the alterations seen in the biomarker studies.