Thyroid hormones (THs) classically regulate the gene expression by transcriptional mechanisms. In pituitary, the encoding genes for growth hormone (GH) and thyroid-stimulating hormone (TSH) are examples of genes regulated by triiodothyronine (T3) in a positive and negative way, respectively. Recent studies have shown a rapid adjustment of GH and TSH synthesis/secretion induced by T3 posttranscriptional actions. In somatotrophs, T3 promotes an increase in Gh mRNA content, poly(A) tail length and binding to the ribosome, associated with a rearrangement of actin cytoskeleton. In thyrotrophs, T3 reduces Tshb mRNA content, poly(A) tail length and its association with the ribosome. In parallel, it promotes a redistribution of TSH secretory granules to more distal regions of the cell periphery, indicating a rapid effect of T3 inhibition of TSH secretion. T3 was shown to affect the content of tubulin and the polymerization of actin and tubulin cytoskeletons in the whole anterior pituitary gland, and to increase intracellular alpha (CGA) content. This review summarizes genomic and non-genomic/posttranscriptional actions of TH on the regulation of several steps of GH and TSH synthesis and secretion. These distinct mechanisms induced by T3 can occur simultaneously, even though non-genomic effects are promptly elicited and precede the genomic actions, coexisting in a functional network within the cells.
Nuclear thyroid receptors and the geneexpression regulation
Most of the well-characterized actions of thyroid hormones (THs) are mediated by thyroid hormone receptors (THRs), the alpha (THRA) and beta (THRB) receptors, which belong to the superfamily of nuclear receptors. THRs act as transcriptional factors interacting with specific sequences of the DNA called thyroid hormone response elements (TREs) and promoting the activation or inhibition of target gene transcription. THRs can also indirectly bind to DNA through proteins such as AP1 (jun/fos) complexes, influencing the gene expression (Flamant et al. 2017). The presence of THRA and THRB isoforms in humans, mice, rats, chickens and Xenopus laevis was described and indicates that this mechanism is conserved across a variety of species (Tata & Widnell 1966, Samuels & Tsai 1973, Sap et al. 1986, Weinberger et al. 1986, Yaoita et al. 1990).
The THRs are encoded by Thra and Thrb genes. The heterogeneous mRNAs are subjected to alternative splicing or alternative use of promoters generating nine different transcripts: Thra1, Thra2, Thra3, ThrΔa1, ThrΔa2, Thrb1, Thrb2, Thrb3 (found only in the kidney, liver and lung of rats) and ThrΔb (O’Shea & Williams 2002, Ortiga-Carvalho et al. 2014). The expression of these transcripts changes during the development and cellular differentiation processes, being also transcriptionally and posttranscriptionally regulated. The THRs are differentially expressed in most tissues. Thrb2 is highly expressed in the pituitary, paraventricular, ventromedial and arcuate nuclei of hypothalamus, inner ear and in some regions of the brain during development, while Thrb1 mRNA is mainly detected in the brain, skeletal muscle, liver and kidney of rats (Lazar 1993, Ortiga-Carvalho et al. 2016). Mutations in THR isoforms are associated with different phenotypes according to tissue distribution of the mutated receptor (Mendoza & Hollenberg 2017).
THRs bind to TREs under the form of homodimers or heterodimers, mainly with retinoid acid receptors and retinoid X receptors (RXRs) that also belong to the superfamily of nuclear receptors. The THRs interact with TH, coactivators or corepressors modulating the gene transcription (Shupnik 2000, Yen 2001, Cheng et al. 2010).
In the anterior pituitary gland, THs downregulate the expression of genes that encode CGA (glycoprotein hormones alpha chain) and TSHB subunits of thyroid-stimulating hormone (TSH) (Franklyn et al. 1988) and upregulate Gh transcription (Crew & Spindler 1986, Koenig et al. 1987, Lavin et al. 1988). The THRB2 isoform participates in the positive and negative regulations of Gh and Tshb expressions, respectively (Abel et al. 1999, Barra et al. 2004). In humans, mutations in THRB isoforms led to the classic form of resistance to thyroid hormone (RTHB) characterized by high TH levels and normal or slightly elevated serum TSH concentration, due to the resistance of pituitary and hypothalamus to respond to the negative feedback triggered by TH (Forrest et al. 1996, Refetoff et al. 2014). Mild-to-moderate growth retardation and delayed bone maturation have also been reported in RTHB (Refetoff et al. 1993, Kaneshige et al. 2000).
Hypothyroidism induced by thyroidectomy in male rats increases the expression of both TSH subunits (Bargi-Souza et al. 2013, 2015), while the expression of GH, which is involved in linear growth and intermediary metabolism modulation, is markedly reduced (Hervas et al. 1975, Volpato & Nunes 1994). On the other hand, studies performed in hyperthyroid rodents and in T3-treated TtT97 cells, a thyrotrophic tumor-transplantable cell line, have shown a marked increase in Gh mRNA content and a significant reduction in Cga and Tshb contents (Shupnik et al. 1985, Chin et al. 1993, Volpato & Nunes 1994, Shupnik 2000, Bargi-Souza et al. 2013, 2015).
Transcriptional regulation of Gh by T3
During the 1980s, several studies carried out in rats demonstrated that the Gh expression is positively regulated by THs (Crew & Spindler 1986, Larsen et al. 1986, Koenig et al. 1987, Lavin et al. 1988). The increase in the Gh transcription rate triggered by T3 requires the interaction of the THR/T3 complex with sequences located close to the transcription start site (TSS) of Gh and the presence of specific transcriptional factors expressed in somatotrophs, since the insertion of Gh promoter in fibroblast or kidney cells did not evoke Gh transcription in response to T3 (Larsen et al. 1986, Koenig et al. 1987).
Studies on the mutational analysis of the TRE of the rat Gh gene (rGh) and other positively T3-regulated genes suggest that (G/A)GGT(C/G)A is a putative consensus hexamer half-site sequence (Flamant et al. 2017). Changes in the orientation, spacing and number of half-sites of TRE were also observed (Yen 2001, Ortiga-Carvalho et al. 2016). In the rGh promoter, the sequences could be oriented as palindromes, or direct repeats separated by 4 nucleotides (Yen 2001, Cheng et al. 2010).
In the case of THs positively regulated genes, such as Gh, the THRB2 heterodimerizes mainly with the RXRs and in the absence of the ligand, the THR/RXR complex is linked to the DNA, repressing the transcriptional process through the interaction with corepressors such as NCoR1 (nuclear receptor corepressor-1) and the SMRT (silencing mediator for retinoid and thyroid receptor), a process that involves the alteration of histone deacetylases, such as HDAC1 or HDAC3 (Harvey & Williams 2002). Numerous corepressors interact with THRs suggesting a high level of complexity in the basal suppression process in the absence of THs.
The presence of ligand promotes a conformational change in THRB2 that results in the dissociation of corepressors (CoR) and recruitment of coactivators (CoA) such as the nuclear receptor coactivator 1 (NCoA-1 or SRC-1) establishing a new complex that leads to histone acetylation or methylation, which, in turn, changes the interaction between RNA polymerase and other transcriptional factors inducing the gene transcription (Ortiga-Carvalho et al. 2014, Mendoza & Hollenberg 2017). It is worth mentioning that T3-mediated activation of rat Gh gene expression presents a synergism between transcriptional factor Pit-1 (POU domain, class 1, transcription factor 1) and T3, increasing the Gh mRNA content (Schaufele et al. 1992, Sinha & Yen 2000, Yen 2001).
The regulation of GH synthesis and secretion mediated by T3 differs in non-mammalian species. T3 increases GH synthesis and secretion in fish, but decreases Gh mRNA expression in birds and GH secretion in chickens (Melamed et al. 1998). These findings could be related to a differential expression of coactivators and repressors in different species, which affects the response to TH, explaining the diversity of T3 effects on GH regulation. Currently, the literature and knowledge are relatively poor for reptile and amphibian species.
Negative transcriptional regulation of Cga and Tshb genes by T3
The involvement of THRA and THRB isoforms in the control of hypothalamus–pituitary–thyroid (HPT) axis has been evaluated in patients with natural mutations. Clinical observations have shown that in patients with a mutation in the THRA, the HPT axis is minimally or not impaired, while the opposite is observed in patients with mutation of THRB isoform (Ortiga-Carvalho et al. 2014).
These clinical observations are supported by studies performed in different models of Thrb1 and Thrb2 knockout mice. The lack of all THRB isoforms (Thrb−/−) led to resistance to thyroid hormone (RTH-b) characterized by the enlargement of the thyroid gland (goiter), increased size and number of follicles, elevated T4 and T3 concentrations, normal or elevated serum TSH concentration and increased Tshb and Cga mRNA expressions due to defects in thyrotrophic response to THs (Forrest et al. 1996, O’Shea & Williams 2002, Mendoza & Hollenberg 2017). The generation of THRB2 knockout (Thrb2−/−) mice provided evidence that this isoform plays a major role in the negative feedback exerted by T3 in thyrotrophs and in the stimulation of Gh mRNA synthesis in somatotrophs (Abel et al. 1999, 2003).
The absence of both Thra1 and Thra2 in mice led to the development of an impaired thyroid gland and to the reduction of thyroxine (T4) synthesis (O’Shea & Williams 2002), while the overexpression of THRA1 results in low levels of T3 and T4 and normal levels of TSH (Mendoza & Hollenberg 2017). Mice with double knockout for Thra and Thrb (Thra−/−b−/−) present higher T4, T3 and TSH serum concentrations than Thrb−/− mice (Weiss et al. 1997, Abel et al. 1999, Chiamolera et al. 2012). In TαT1 cells, the THRB is essential for T3 inhibition of Tshb mRNA expression, while the THRA plays an important role only in the absence of THRB (Chiamolera et al. 2012). In humans, some RTH patients with mutations in THRA presented normal or low TSH levels and reduced T4/T3 ratio. All these data together suggest that the THRA isoform plays an important role in the negative feedback mechanism and that THRA and THRB are required for the adequate regulation of the HPT axis (Mendoza & Hollenberg 2017).
Although the precise sequence of negative TREs (nTRE) has not been described yet (Ortiga-Carvalho et al. 2016), previous studies demonstrated that they appear to be near the TSS of Cga/CGA and Tshb/TSHB genes of rats, mice and humans (Carr et al. 1987, Shupnik 2000). However, recent studies have shown that the negative regulation of human TSHB gene exerted by T3 is independent of nTRE (Matsushita et al. 2007, Matsunaga et al. 2015).
In the pituitary, the glycoprotein alpha subunit (CGA) is expressed in thyrotrophs and gonadotrophs and composes the mature molecule of TSH, luteinizing hormone (LH) and follicle-stimulating hormone (FSH) together with the beta-subunit, which, in turn, guarantees hormonal specificity (Pierce & Parsons 1981).
Regarding the transcriptional effects of triiodothyronine on the Cga expression, it is known that in the rat and mouse genomes, there are several hexameric TRE half-site motif regions in the Cga gene, supporting the hypothesis that there is a suppression of Cga transcription by T3 (Burnside et al. 1989). Studies in the human choriocarcinoma cell line JEG-3, transfected with THR, have shown that T3 treatment markedly suppresses the CGA expression. The presence of a binding site for THR located between −22 and −7 base pairs of TSS of CGA gene immediately downstream from the TATA box suggests a possible interference of THs/TATA box with transcriptional factors reducing the synthesis of CGA mRNA (Chatterjee et al. 1989).
Considering that the Cga subunit is negatively regulated by THs (Shupnik & Ridgway 1987, Bargi-Souza et al. 2015), it might be expected that during hypothyroidism and hyperthyroidism, significant alterations in gonadotrophin synthesis and secretion would occur since CGA is essential for the composition of FSH and LH and that binding between alpha and beta subunits is important for hormonal maturation, biological activity and the recognition of these glycoproteins by specific receptors on gonads (Pierce & Parsons 1981, Bargi-Souza et al. 2015). Our recent studies indicate that alterations in the HPT axis strongly affect the hypothalamus–pituitary–gonad axis in rats, which might explain some reproductive disorders observed under hypothyroid and hyperthyroid conditions (Romano et al. 2013).
Despite the efforts, the precise mechanism implicated on the molecular basis of T3 negative regulation of Tshb and Cga remains unknown. Four models were proposed for the negative regulation triggered by T3 involving the nTREs as follows: (a) the position of nTRE, which prevents the formation of RNA polymerase II complex and consequently the initiation of transcription; (b) the interaction of the THR–TH complex with co-coactivators altering the conformation of the DNA molecule and thus preventing the transcription process; (c) the interaction of THR–TH with transcription inhibitors and (d) the presence of an nTRE inside an enhancer sequence, thus preventing transcription (Chin et al. 1993). Moreover, the mutation in the activation domain of THRB abolishing CoA recruitment but preserving normal T3 binding and CoR interactions has shown that this portion is required for positive and, paradoxically, for negative regulation by TH, increasing the complexity of TH gene expression regulation (Ortiga-Carvalho et al. 2005). It is noteworthy to mention that transcriptional alterations occur concomitantly with alterations in GH and TSH secretion in rats, which implies that different mechanisms are triggered for the proper synthesis/secretion regulation of these hormones (Zoeller et al. 2007).
The regulation of HPT in non-mammals appears to be very similar to that in mammals. In teleost and amphibians, the activity of HPT is important for the normal development and metamorphosis of the animals and their larvae are sensitive to slight disruption of HPT function (Carr & Patino 2011). In the few fish and amphibians species studied, both T4 and T3 decrease TSH secretion by reducing Tshb gene transcription (Blanton & Specker 2007). In birds, THs present a shorter half-life than that in mammals but the negative feedback triggered by THs at the pituitary level is similar to that observed in mammals (McNabb 2007).
Non-genomic actions of THs
The non-genomic or posttranscriptional actions of THs are not dependent on the nuclear receptor interaction with TRE, even though the signaling resulting from these mechanisms could lead to the regulation of gene expression (Davis et al. 2013, 2016). These actions are rapidly triggered even in the presence of transcription inhibitors (Yusta et al. 1998, Lorenzo et al. 2002) by TH interactions with THRs coupled to specific enzymes in the cytosol or to other binding sites located on the plasma membrane and organelles such as mitochondria and endoplasmic reticulum (Moeller et al. 2006, Davis et al. 2008, 2009, Goulart-Silva et al. 2011).
There are a substantial number of studies connecting non-genomic actions of THs with complex physiological processes like the traffic of intracellular protein, the basal activity of transporter proteins, the regulation of the half-life of specific mRNAs (Davis et al. 2005, Hammes & Davis 2015), the cellular metabolism (Moeller et al. 2006), the brain development (Anton et al. 1999, Leonard 2008, Cao et al. 2009), the cardiac growth (Kenessey & Ojamaa 2005), angiogenesis (Bergh et al. 2005) and the osteoblast proliferation (Kalyanaraman et al. 2014).
In 2005, a binding site for THs was identified in the extracellular domain of alphaVbeta3 integrin present on the plasma membrane, which has a crucial role in pro-angiogenic actions triggered by THs (Bergh et al. 2005, Cody et al. 2007, Lin et al. 2011). This integrin has two binding sites, S1 and S2, for THs with different affinities for T4 and T3. The S1 site only recognizes T3 and, in the presence of the ligand, activates the Src and PI3K pathways, whereas the S2 site recognizes both T4 and T3 with higher affinity for T4 and activates the ERK1/2 pathway (Davis et al. 2011). Parts of TH actions elicited via alphaVbeta3 integrin are blocked by RGD (arginine-glycine-aspartate) peptide which strongly suggests that this TH binding site is located at or near the RGD recognition site on the alphaVbeta3 integrin (Davis et al. 2005, Cody et al. 2007). It was shown that the presence of RGD peptide prevented the trafficking of THRs from the cytosol to the nucleus induced by T3 treatment (Cheng et al. 2010), pointing out a crucial participation of the alphaVbeta3 integrin in gene transcription regulation.
More recently, a THRA of 30 kDa (p30 THRA) was evinced in osteoblasts associated with plasma membrane domains rich in caveolin. T3 binding to this receptor leads to an increase in intracellular Ca++ concentration and subsequent activation of NO-cGMP-PKGII followed by MEK-ERK and PI3K-Akt signaling cascades resulting in increased proliferation and survival of osteoblasts (Kalyanaraman et al. 2014). Another non-genomic mechanism of TH actions involves their interaction with nuclear receptors coupled to the regulatory p85 subunit of phosphoinositide 3-kinase (PI3K) in the cytoplasm regulating important proteins associated with glucose uptake, protein synthesis and proteolysis inhibition (Saji & Ringel 2010, Kimura et al. 2014). Truncated isoforms of THRs present in the cytoplasm also mediate non-genomic actions of T4 and reverse T3 (rT3) (Cheng et al. 2010). The interaction of THRΔA1 with T4 or rT3 is associated with the rearrangement of actin cytoskeleton and with the release of laminin, a matrix protein, by astrocytes (Farwell et al. 1995). The laminin and the rearrangement of the actin cytoskeleton are related to the induction of neuronal migration and growth of neurites, respectively (Farwell et al. 1995).
Thus, the identification of binding sites of THs at the plasma membrane and cytosol indicates a complex mechanism triggered by THs to establish their biological effects and the elucidation of TH non-genomic actions raises the possibility of therapeutic intervention through pharmacological agent actions in these binding sites.
Non-genomic regulation of GH and TSH synthesis and secretion
Posttranscriptional effects of T3 on Gh synthesis
In the beginning of 1990s, few studies addressed the non-genomic actions of THs and they were restricted to T4 and rT3. These studies pointed out the rapid modulation of type 2 deiodinase (DIO2) activity triggered by T4 and rT3 in astrocytes, but not by T3, suggesting that genomic actions were triggered by T3, while T4 and other metabolites were responsible for the non-genomic actions (Leonard 2008).
However, our group showed that a single T3 injection in thyroidectomized (Tx) rats promoted an increase in Gh mRNA content in 15 min compared to hypothyroid animals (Volpato & Nunes 1994), and in the molecular weight of Gh mRNA characterized by its slow migration on the agarose gel 30 min thereafter. After removing the poly(A) tail, Gh mRNA from Tx rats treated with T3 migrated on the gel to the same height of Gh mRNA from untreated Tx rats. In fact, sequencing analyses of Gh mRNA from T3-treated Tx rats revealed an increased number of adenine residues in its poly(A) tail, evincing a posttranscriptional action of T3 in this step of GH processing (unpublished data).
It is worth to highlight that Gh mRNA from Tx rats presents longer poly(A) tail making this transcript more stable than that of euthyroid animals (Jones et al. 1990, Murphy et al. 1992). It appears that in the absence of TH, mechanisms are triggered to ensure minimal GH synthesis since the transcriptional rate of Gh gene is reduced under the hypothyroidism condition. Furthermore, it has also been shown that suppression of GH synthesis and secretion induced by IGF-I were abolished in hypothyroidism (Melmed & Yamashita 1986), which reinforces that other mechanisms operate to regulate the GH expression in Tx animals.
In our study, T3 further induced a rapid increase in Gh mRNA poly(A) tail length, ensuring more stability for Gh transcripts (unpublished data), and GH labeling at the perinuclear region of somatotrophs of Tx rats suggesting that GH synthesis was stimulated by T3 (Silva et al. 2006). Our group has proceeded with the investigations on the molecular mechanisms involved in the rapid actions of T3 on Gh mRNA stability and we have also studied the involvement of the cytoskeleton in T3 non-genomic actions, as discussed later.
Rapid and negative posttranscriptional regulation of Cga and Tshb
Considering the positive effects of T3 on the posttranscriptional steps of Gh synthesis, our group has investigated whether T3 could trigger rapid effects on Tshb mRNA content and poly(A) tail length (Goulart-Silva et al. 2011, Bargi-Souza et al. 2013).
Using Tx rats, we observed that in hypothyroidism, the poly(A) tail length of Tshb mRNA is increased, followed by a significant rise in the translational rate of Tshb, besides the expected increase in the transcriptional process (Bargi-Souza et al. 2013). Thirty minutes after treatment with T3, the poly(A) tail length of Tshb mRNA of Tx rats was reduced to values similar to those observed in the poly(A) tail length of euthyroid rats and was positively correlated with a marked reduction in the content of Tshb in polysome fractions indicating that T3 rapidly reduced mRNA translation and TSH protein synthesis (Goulart-Silva et al. 2011, Bargi-Souza et al. 2013).
Previous studies regarding T3 effects on the Tshb mRNA poly(A) tail length have already shown a decrease in its polyadenylation degree in hypothyroid pituitary cell cultured with T3 for 8 h and this reduction was more pronounced after 24 h (Krane et al. 1991, Staton & Leedman 1998). Euthyroid animals treated for 4 h with higher doses of T3 also presented a reduction of Tshb mRNA poly(A) tail length (Krane et al. 1991).
In hypothyroidism, we have also observed an increase in the Cga mRNA content and in its association with ribosomes, indicating a rise in the translational rate of CGA. Thirty minutes after T3 administration, the content of Cga mRNA seen in the polysome fraction was reduced suggesting that the translation of both alpha and beta subunits is posttranscriptionally modulated by T3 in hypothyroid animals (Bargi-Souza et al. 2015).
In contrast to the results observed for Tshb, the poly(A) tail length of Cga was not altered under different thyroidal conditions, which led us to hypothesize that T3 acts through different pathways modulating the elongation of Tshb poly(A) tail and the translation of both transcripts. Thus, the regulation of poly(A) tail length by T3 is specific for Tshb and it could be speculated that such specificity would reduce the impact on Cga synthesis since this subunit is also important for LH and FSH hormones.
Cytoskeleton as a target of non-genomic actions of T3: consequences to GH and TSH secretion
The cytoskeleton is directly involved with the mechanical support system of a cell, its shape and the release of molecules stored in secretory granules by exocytosis (Apodaca 2002), being a target of hormones such as T3 (Banovac & Koren 2000).
The first experiments showing cytoskeletal rearrangement induced by THs were performed in cultured astrocytes. The treatment with T4 or rT3 promoted a cytoskeleton reorganization characterized by an increase in polymerized actin content and led to reduction of the DIO2 enzymatic activity in astrocytes (Siegrist-Kaiser et al. 1990).
Considering that DIO2 is responsible for the intracellular generation of T3 from T4 and that T3 is an important Gh and Tshb gene expression modulator (Ortiga-Carvalho et al. 2016), our group attempted to investigate the repercussions of hypothyroidism and T3 effects in the cytoskeleton of pituitary cells and the consequences to GH and TSH secretion (Silva et al. 2006, Bargi-Souza et al. 2013).
We have observed a marked disarrangement of the cytoskeleton characterized by a deep reduction of actin and tubulin filament staining in anterior pituitary cells of hypothyroid rats and an evident increase in blood vessel diameter (Silva et al. 2006, Bargi-Souza et al. 2013). Actin and tubulin polymerization leads to the assembly of microfilaments and microtubules, respectively. The latter is involved in the transport of vesicles and granules within the cell and their release to the extracellular milieu (Fokin et al. 2014). Thus, changes in the arrangement of the cytoskeleton in pituitary could be associated with alterations in pituitary hormone secretion.
Studies carried out by different research groups have shown a link between the animal thyroid condition and the microtubule polymerization in the central nervous system (Chaudhury et al. 1985, Ravindra & Grosvenor 1990, Lorenzo et al. 2002, Gutierrez 2012), as well as a reduction in the amount of actin filaments in astrocyte cytoskeleton (Siegrist-Kaiser et al. 1990) in the absence of T4. The results obtained in the anterior pituitary of hypothyroid rats showed a disarrangement of actin and tubulin filaments, which reinforces the importance of THs in the cytoskeleton integrity of cells.
The repercussions on pituitary hormone secretion were then evaluated by immunohistochemistry. The results showed a decrease in actin filament content (F-actin) in GH immunoreactive cells of hypothyroid rats in which GH granules were predominantly observed at the cell periphery. Thus, cytoskeleton disarrangement in somatotrophs could contribute to the well-established reduction of GH secretion observed in hypothyroidism (Silva et al. 2006). However, 30 min after T3 administration of hypothyroid rats, an increase was observed in the F-actin content mainly near the plasma membrane, characterizing a rapid cytoskeleton rearrangement, even though the total actin protein remained unchanged, which led us to suppose that T3 only affected actin rearrangement (Silva et al. 2006). In parallel, there was a remarkable increase in GH labeling at the perinuclear region and a decrease in GH staining at the cell periphery, strongly indicating that GH was released, which was confirmed by a rapid increase in liver Igf1 mRNA expression in these animals (Silva et al. 2010).
Furthermore, a possible relationship between T3 effects on the cytoskeleton and the repercussions in GH synthesis cannot be discarded. In fact, the cytoskeleton enables the polysome to anchor to it. Polysomes correspond to the functional units of the translational machinery and they arrest the mRNAs to initiate the translation mechanism. Thus, anchoring polysomes to the cytoskeleton also means attaching and distributing mRNAs to specific cell regions, where they will be translated (Campbell & Hesketh 1996, Veyrune et al. 1996, Czaplinski & Singer 2006, Coulon et al. 2013). In this sense, the cytoskeleton may be involved in the increased GH synthesis observed after hypothyroid rats had been acutely treated with T3 and in the marked GH staining at the somatotroph perinuclear region. In the hypothyroid state, immunostaining for GH was almost absent near the nucleus due to the reduced transcriptional rate of the Gh gene (Crew & Spindler 1986, Larsen et al. 1986, Lavin et al. 1988, Samuels et al. 1988).
In addition to the changes observed for the actin cytoskeleton, T3 also promoted a microtubule rearrangement in pituitary as well as an increase in tubulin content and a rapid reduction of the pituitary blood vessel diameter, which, in turn, could alter the dynamics of pituitary blood flow (Bargi-Souza et al. 2013). Considering the role of microtubules in hormone secretion, our findings strengthen the important role of THs in the secretion of pituitary hormones consequently affecting metabolism, growth and organic system functions, considering the central position of pituitary in the control of other endocrine glands’ activity.
Different from the somatotrophs, no colocalization of TSH and F-actin or microtubule labeling was observed. In fact, despite F-actin and microtubule disarrangement in the pituitary in hypothyroidism, TSH secretion is highly increased. Indeed, the immunohistochemistry and electron microscopy analysis of hypothyroid rat pituitaries showed strong TSH staining near and inside the pituitary blood vessels which is in accordance with the high serum TSH concentration detected in the hypothyroid animals (Bargi-Souza et al. 2013).
In contrast, T3 promptly reduced TSH labeling inside blood vessels and promoted a rapid redistribution of TSH granules in thyrotrophs, increasing their abundance in the whole cytoplasm and reducing their presence in regions close to the plasma membrane. These findings indicate a rapid T3 effect inhibiting TSH secretion (Bargi-Souza et al. 2013).
Recently, studies from our group have shown that the CGA subunit content and secretion are also regulated by T3 (Bargi-Souza et al. 2015). In hypothyroidism, the CGA amount in the pituitary is decreased and, considering only thyrotrophs, this reduction was even more evident. Moreover, T3 acute administration in physiological and supraphysiological doses promoted a rapid increase in pituitary CGA intracellular content which is highly suggestive of a blockage in CGA secretion, since the Cga mRNA content in polysomal fractions was rapidly reduced, suggesting a reduction in CGA protein synthesis (Bargi-Souza et al. 2015).
Indeed, studies carried out by our group in anterior pituitary primary culture showed a reduction of CGA content in the cell culture medium after acute T3 treatment, reinforcing the inhibitory T3 action on CGA secretion (Bargi-Souza et al. 2015). These data together with the observation that acute T3 administration reduces the Cga mRNA translation rate strengthen the evidence that T3 acts non-genomically reducing the synthesis and secretion of CGA, which might also impact LH and FSH synthesis and secretion, as suggested (Krassas & Pontikides 2004, Romano et al. 2013). Hence, the increase in pituitary TSHB and CGA contents acutely triggered by T3 strongly indicates a rapid inhibition of TSH secretion. In this sense, studies carried out in GH4C1 pituitary cells shows that T3 increases the activity of the Potassium Voltage-Gated Channel (KCNH2) by non-genomic mechanisms that involve the PI3 kinase pathway activation, antagonizing the thyrotropin-releasing hormone effects on PRL secretion (Storey et al. 2006).
Interaction between cytoskeleton and mRNAs
Transcripts anchored to the cytoskeleton may be positioned in specific regions of the cell to be translated where the new protein may exert its function. In fact, the actin transcript has been shown to be attached to cytoskeleton and addressed to cell periphery to be translated to monomeric actin (G-actin), which could be polymerized to F-actin (Bassell et al. 1998, Eom et al. 2003).
The mRNA interaction with the cytoskeleton may alter transcript stability, which, in turn, affects protein synthesis efficiency (Kosmidou et al. 2007, Challa & Stefanovic 2011). T3 increases the Gh transcript stability by adding adenine residues to the 3′end of its poly(A) tail and the F-actin content; both alterations could contribute to increased GH protein synthesis. On the other hand, there are reports showing that some mRNAs become more stable after cytoskeleton disarrangement as it occurs with proinflammatory cytokine mRNAs, expressed on airway epithelial cells, whose translation rate is increased under this condition (van den Berg et al. 2006). Based on these findings, we can infer that actin and microtubule disorganizations, both observed in the hypothyroid rat pituitary, could differentially regulate GH and TSH synthesis and secretion.
As mentioned earlier, mRNAs are anchored to the cytoskeleton through interaction with specific proteins. EF1 alpha is one of these proteins, and it is worth mentioning that it has an important role in the protein synthesis elongation phase by transporting the aminoacyl-tRNA to the ribosome via a GTP-dependent mechanism (Greganova et al. 2011).
As pointed out, in hypothyroid states, the actin cytoskeleton of pituitary cells is disarranged. Besides that, we observed a reduction of both EF1 alpha and Gh mRNA in the F-actin fraction of hypothyroid rat pituitaries (Silva et al. 2010), which indicates a reduction of GH synthesis. In fact, the polysome profile analysis of hypothyroid rat pituitary showed a decrease in Gh mRNA amount in polysomes. Even though there is a low rate of Gh gene transcription in hypothyroid states, we cannot exclude the contribution of posttranscriptional mechanisms to this event since EF1 alpha and Gh mRNA binding to F-actin cytoskeleton was shown to be reduced (Silva et al. 2010). In fact, T3 acutely administered to hypothyroid rats rapidly increased the EF1 alpha and Gh mRNA amount in the F-actin cytoskeleton, the Gh mRNA content in the polysome fraction and promoted cytoskeleton rearrangement (Silva et al. 2006, 2010). These events may favor the increase in GH synthesis, and could explain the increase in GH labeling in the perinuclear region of somatotrophs, both observed in pituitary of hypothyroid rats acutely treated with T3.
In summary, in parallel to the transcriptional regulation of Gh, Tshb and Cga genes triggered by triiodothyronine which was for a long period considered the unique mechanism for regulation of GH and TSH synthesis and consequently secretion, T3 acts non-genomically on thyrotrophs and somatotrophs regulating the polyadenylation of Gh and Tshb mRNAs, and GH and TSH translation/synthesis and secretion. These findings significantly increase the comprehension on novel aspects of the negative feedback mechanism exerted by T3 on thyrotrophs, as well as add new elements to the understanding of T3 induction of Gh gene expression and secretion which are processes that are considered to be mainly regulated by growth hormone–releasing hormone. These non-genomic actions of T3 on GH and TSH synthesis/secretion may have an important physiological relevance for the optimal control of the rate of GH and TSH basal secretion, minute by minute. Figure 1 summarizes the alterations observed in GH and TSH expressions under hypothyroid conditions and the non-genomic actions triggered by T3 modulating GH and TSH synthesis/secretion.
It is worth mentioning that the non-genomic actions elicited by T3 precede and are in line with its transcriptional actions on both genes since they lead to a positive and negative regulation of GH and TSH, respectively. Other genes/proteins that are targeted by T3 also present an overlap of non-genomic and genomic mechanisms. The T4–αVβ3 interaction promotes αVβ3 internalization (Lin et al. 2013) and the αV monomer forms a complex with MAPK-p300 inside the cell that binds to the Thrb1 promoter increasing its transcriptional rate (Hammes & Davis 2015). The increased amount of THRB1, in turn, increases the response to genomic actions of T3 mediated by this isoform (Hammes & Davis 2015).
Thus, non-genomic actions are responsible for rapid adjustments of cellular basal activity, whereas in parallel, genomic actions promote long-term control of cellular activity, being responsible for the specific properties of a tissue/system (Hammes & Davis 2015). In this sense, the physiological response to TH is the result of both genomic and non-genomic actions, as illustrated in Figure 2.
The understanding of non-genomic molecular pathways activated by T3 would provide new insights into the physiological role of TH at a cellular level enabling the development of new analogs that at certain degree could compensate the lack of proper TH regulation observed in THR mutation disorders or specific antagonists to avoid tumor cell proliferation and angiogenesis, triggered by non-genomic actions of T4 through interaction with αVβ3 integrin at the plasma membrane (Hammes & Davis 2015).
It is noteworthy to mention that the literature on the non-genomic actions of TH is recent and there is still much to be investigated. Moreover, some actions are observed in a period of time of only a few seconds to minutes, which represents a limitation for clinical studies. The development of tools to explore the non-genomic actions in endocrine and metabolic disorders is also required. The use of microarray, RNA-Seq, proteomics and functional assays would also be useful to characterize/describe new non-genomic actions triggered by THs. Due to all these reasons, clinical applications of non-genomic actions are still unclear.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as hindering the impartiality of this review.
This work was supported by grants from the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP: 2006/61713-0 and 2013/05629-4) to M T N. M T N is the recipient of a Conselho Nacional de Pesquisa e Desenvolvimento (CNPq) fellowship and P B-S is the recipient of a FAPESP fellowship (2009/17813-9, 2012/06643-8 and 2014/25888-7).
The authors would like to thank Isidro Moisés Cervantes Galicia for the cell illustration model.
AbelEDMouraEGAhimaRSCampos-BarrosAPazos-MouraCCBoersMEKaulbachHCForrestDWondisfordFE 2003 Dominant inhibition of thyroid hormone action selectively in the pituitary of thyroid hormone receptor-beta null mice abolishes the regulation of thyrotropin by thyroid hormone. Molecular Endocrinology 17 1767–1776. (doi:10.1210/me.2003-0109)
Bargi-SouzaPRomanoRMSalgado RdeMGoulart-SilvaFBrunettoELTenorio ZornTMNunesMT 2013 Triiodothyronine rapidly alters the TSH content and the secretory granules distribution in male rat thyrotrophs by a cytoskeleton rearrangement-independent mechanism. Endocrinology 154 4908–4918. (doi:10.1210/en.2013-1508)
Bargi-SouzaPRomanoRMGoulart-SilvaFBrunettoELNunesMT 2015 T3 rapidly regulates several steps of alpha subunit glycoprotein (CGA) synthesis and secretion in the pituitary of male rats: potential repercussions on TSH, FSH and LH secretion. Molecular and Cellular Endocrinology 409 73–81. (doi:10.1016/j.mce.2015.04.002)
BerghJJLinHYLansingLMohamedSNDavisFBMousaSDavisPJ 2005 Integrin alphaVbeta3 contains a cell surface receptor site for thyroid hormone that is linked to activation of mitogen-activated protein kinase and induction of angiogenesis. Endocrinology 146 2864–2871. (doi:10.1210/en.2005-0102)
van den BergAFreitasJKelesFSnoekMvan MarleJJansenHMLutterR 2006 Cytoskeletal architecture differentially controls post-transcriptional processing of IL-6 and IL-8 mRNA in airway epithelial-like cells. Experimental Cell Research 312 1496–1506. (doi:10.1016/j.yexcr.2006.01.010)
ChiamoleraMISidhayeARMatsumotoSHeQHashimotoKOrtiga-CarvalhoTMWondisfordFE 2012 Fundamentally distinct roles of thyroid hormone receptor isoforms in a thyrotroph cell line are due to differential DNA binding. Molecular Endocrinology 26 926–939. (doi:10.1210/me.2011-1290)
DavisPJDavisFBLinHYMousaSAZhouMLuidensMK 2009 Translational implications of nongenomic actions of thyroid hormone initiated at its integrin receptor. American Journal of Physiology: Endocrinology and Metabolism 297 E1238–E1246. (doi:10.1152/ajpendo.00480.2009)
KaneshigeMKaneshigeKZhuXDaceAGarrettLCarterTAKazlauskaiteRPankratzDGWynshaw-BorisARefetoffS 2000 Mice with a targeted mutation in the thyroid hormone beta receptor gene exhibit impaired growth and resistance to thyroid hormone. PNAS 97 13209–13214. (doi:10.1073/pnas.230285997)
KenesseyAOjamaaK 2005 Ligand-mediated decrease of thyroid hormone receptor-alpha1 in cardiomyocytes by proteosome-dependent degradation and altered mRNA stability. American Journal of Physiology: Heart and Circulatory Physiology 288 H813–H821. (doi:110.1152/ajpheart.00804.2004)
KimuraKChengXWInoueAHuLKoikeTKuzuyaM 2014 beta-Hydroxy-beta-methylbutyrate facilitates PI3K/Akt-dependent mammalian target of rapamycin and FoxO1/3a phosphorylations and alleviates tumor necrosis factor alpha/interferon gamma-induced MuRF-1 expression in C2C12 cells. Nutrition Research 34 368–374. (doi:10.1016/j.nutres.2014.02.003)
MatsunagaHSasakiSSuzukiSMatsushitaANakamuraHNakamuraHMHiraharaNKurodaGIwakiHOhbaK 2015 Essential role of GATA2 in the negative regulation of type 2 deiodinase gene by liganded thyroid hormone receptor beta2 in thyrotroph. PLoS ONE 10 e0142400. (doi:10.1371/journal.pone.0142400)
MatsushitaASasakiSKashiwabaraYNagayamaKOhbaKIwakiHMisawaHIshizukaKNakamuraH 2007 Essential role of GATA2 in the negative regulation of thyrotropin beta gene by thyroid hormone and its receptors. Molecular Endocrinology 21 865–884. (doi:10.1210/me.2006-0208)
MelmedSYamashitaS 1986 Insulin-like growth factor-I action on hypothyroid rat pituitary cells: suppression of triiodothyronine-induced growth hormone secretion and messenger ribonucleic acid levels. Endocrinology 118 1483–1490. (doi:10.1210/endo-118-4-1483)
MoellerLCCaoXDumitrescuAMSeoHRefetoffS 2006 Thyroid hormone mediated changes in gene expression can be initiated by cytosolic action of the thyroid hormone receptor beta through the phosphatidylinositol 3-kinase pathway. Nuclear Receptor Signaling 4 e020. (doi:10.1621/nrs.04020)
MurphyDPardyKSeahVCarterD 1992 Posttranscriptional regulation of rat growth hormone gene expression: increased message stability and nuclear polyadenylation accompany thyroid hormone depletion. Molecular and Cellular Biology 12 2624–2632. (doi:10.1128/MCB.12.6.2624)
Ortiga-CarvalhoTMShibusawaNNikrodhanondAOliveiraKJMachadoDSLiaoXHCohenRNRefetoffSWondisfordFE 2005 Negative regulation by thyroid hormone receptor requires an intact coactivator-binding surface. Journal of Clinical Investigation 115 2517–2523. (doi:10.1172/JCI24109)
RefetoffSBassettJHBeck-PeccozPBernalJBrentGChatterjeeKDe GrootLJDumitrescuAMJamesonJLKoppPA 2014 Classification and proposed nomenclature for inherited defects of thyroid hormone action, cell transport, and metabolism. Thyroid 24 407–409. (doi:10.1089/thy.2013.3393.nomen)
StatonJMLeedmanPJ 1998 Posttranscriptional regulation of thyrotropin beta-subunit messenger ribonucleic acid by thyroid hormone in murine thyrotrope tumor cells: a conserved mechanism across species. Endocrinology 139 1093–1100. (doi:10.1210/endo.139.3.5799)
VeyruneJLCampbellGPWisemanJBlanchardJMHeskethJE 1996 A localisation signal in the 3' untranslated region of c-myc mRNA targets c-myc mRNA and beta-globin reporter sequences to the perinuclear cytoplasm and cytoskeletal-bound polysomes. Journal of Cell Science 109 1185–1194.