Portal de Periódicos da CAPES

The significance of plasma membrane transport in the bioavailability of thyroid hormone

1998; Wiley; Volume: 48; Issue: 1 Linguagem: Inglês

10.1046/j.1365-2265.1998.00377.x

1365-2265

Hennemann, Everts, Igle J. de Jong, Lim, Krenning, Docter,

Ion channel regulation and function

3,3′,5-Triiodothyronine (T3) is the most important, if not the only, active thyroid hormone. Eighty percentage of total production of plasma T3 is effected by extra-thyroidal conversion of thyroxine (T4) into T3, in which process the liver plays a predominant role, while only 20% is normally secreted directly by the thyroid ( Hennemann, 1986). Most tissues like cardiac and skeletal muscle depend for the greater part, if not fully, on plasma T3 for nuclear occupancy. However, pituitary cells and even more so the brain derive at least 50% of their intracellular T3 from local production from T4 ( Larsen et al., 1981 ). It is thus evident that if transport of thyroid hormones across the plasma membrane was subject to regulation, this could have an impact on the intracellular bioavailability of thyroid hormone, provided that this transport process was rate limiting on the intracellular availability of thyroid hormones. Until two decades ago, it was presumed that the plasma membrane transport process of thyroid hormones consisted of passive diffusion as thyroid hormones are lipophilic, enabling diffusion through the lipid-rich plasma membrane. However, it has now been established by many laboratories using different cell types from various species that thyroid hormones are transported across the plasma membrane of target cells by carrier-mediated processes that are often found to be temperature, sodium, and energy dependent ( Docter & Krenning, 1990). Although the transport protein (s) has not yet been cloned (studies are in progress in our laboratory), it is now generally accepted that these transport processes are operative with regard to entry of thyroid hormones into many tissues. The main question that has to be answered now, is what is their (patho)physiological significance in the bioavailability of thyroid hormone, notably that of T3. It is the intention of this review to discuss step by step the present evidence that enables us to have some idea about the answer to this question. There are certain requirements that have to be fulfilled before it can be concluded that the transport process of thyroid hormones across the plasma membrane may be of (patho)physiological significance ( Table 1). We will discuss these different aspects mainly for the liver because this organ plays a predominant role in plasma T3 production. If T4 transport into the liver is a regulated process and of rate limiting significance, then this process may have significant influence on the total production of bioactive thyroid hormone. When a transport process is specific, it means that only structurally related compounds are transported or compete with the process (in the case of diffusion no such specificity is present). Specificity of transport in rat hepatocytes has been found for T4, T3 and reverse T3 (rT3). The published Km values of transport for these hormones are mostly in the nanomolar range (for review see Kragie, 1994). However, different findings have been reported with regard to energy dependency and specificity of thyroid hormone transport, in that our laboratory reported not only energy dependency of T3 transport but also of T4 transport, whereas we also found evidence for different carriers for T4 and T3 in rat hepatocytes ( Docter & Krenning, 1990). Others, however, although reporting energy dependency for T3 transport, found no such energy dependency for T4 transport, whereas the same group found evidence for a single carrier for both T3 and T4 ( Blondeau et al., 1988 ). The reason for this discrepancy is probably the fact that the latter authors used conditions that were suboptimal for full ATP restoration after isolation of hepatocytes. As T4 transport is more sensitive to decreases in cellular ATP than transport of T3, even a slight decrease in intracellular ATP concentration will impair T4 transport without appreciably affecting that of T3 ( Krenning et al., 1982 ). Under optimal concentrations of cellular ATP in cultured rat hepatocytes we found that, although T4 and T3 transport exhibit mutual inhibition, they were transported via different carriers. In addition, evidence is presented that rT3 is transported through the same system as T4 ( Docter & Krenning, 1990) (see addendum). Thyroid hormone transport is also inhibited by structurally related compounds like benzodiazepines, X-ray contrast agents, amiodarone, and aromatic amino acids ( Kragie, 1994). As published by Chalmers et al. (1993 ), hydrophobicity may also account for the inhibition of transport by these compounds. In a diffusion process of a compound through a membrane there is no saturability of uptake and unidirectional transport is solely dependent on the unbound concentration of the compound in solution. Even on theoretical grounds it is improbable that thyroid hormone, for instance T3, could traverse the plasma membrane of target cells by diffusion since the pore radius of the plasma membrane is calculated to vary between 3.5 and 5.5 Å ( Stitzer & Jacquez, 1975), whereas the mean radius of T3 has been reported to be between 7.2 and 7.5 Å ( Rao et al., 1976 ). Furthermore, as thyroid hormones are in a charged state, it is even more difficult to cross the plasma membrane passively. In addition to these theoretical considerations, experimental evidence has been provided that diffusion hardly takes place if at all. Thus, using electron spin resonance stop-flow technique, it was shown that T3 does not flip-flop at any appreciable rate in prepared phospholipid bilayers and that after partitioning into the membrane it remains in the outer half of the bilayer ( Lai et al., 1985 ). In other words, if no specific transport sites were present in the membrane of biological cells, T3 would not be able to cross the plasma membrane. There is also experimental evidence for absence of significant diffusion of thyroid hormones across the plasma membrane of rat hepatocytes. Using a monoclonal antibody raised against the putative carrier of thyroid hormones in rat liver cells, a relationship was found between the concentration of the monoclonal antibody and its inhibitory action on transport of T3 and T4. At a low dilution factor (1:100) 100% inhibition of T3 and T4 uptake was effected ( Mol et al., 1986 ). The same monoclonal antibody completely inhibited uptake of T4 in cultured human hepatocytes after 20 h incubation ( De Jong et al., 1993 ). In rat anterior pituitary cells and also in oocytes of Xenopus laevis, minimal, if any, uptake of T3 sulphate was detected while specific uptake of T4 and T3 is present in these cell types ( Everts et al., 1994 ; Docter et al., 1995 ). However, injection of fractionated rat liver RNA induced uptake of T3 sulphate in the oocytes ( Docter et al., 1997 ). The fact that no basal uptake of T3 sulphate is present in these cell types, but that membranal transport can be induced by transfection of RNA, strongly underlines the notion that diffusion plays little, if any, role in transmembrane transport of iodothyronines. As the concentration of free T4 and T3 in serum is in the picomolar range, whereas the Km values of uptake are in the nanomolar range, no regulation can evidently be exerted in vivo by the process of saturation. However, there is abundant evidence, from studies of hepatocytes and other types of cells in various species, for energy dependent uptake of T3 and T4. Thus, in mouse neuroblastoma cells, human glioma cells and rat synaptosomes and astrocytes, in which mostly T3 uptake was measured, energy dependency of specific uptake was found. Similar results were found for T4 regarding mouse neuroblastoma cells, but nonsaturability was found for T4 in rat synoptosomes. In rat anterior pituitary cells specificity and energy dependency were found both for T3 and T4. With regard to rat GH pituitary tumour cell lines, controversy exists concerning saturability of T3 and T4 uptake kinetics. In human and rat erythrocytes, T3 uptake is specific but energy dependency is not always found, while no saturable uptake is detected for T4. T3 saturability and energy dependency have also been reported in other cell types like choriocarcinoma, myoblasts, white blood cells, thymocytes, and fibroblasts of different species. T4 saturability, when studied, was not always present (for review see Kragie, 1994). Energy dependency of thyroid hormone transport in hepatocytes is not only found in vitro, but also in the perfused intact rat liver and in humans. When livers of starved rats are perfused using T4 or T3 as a substrate, a decrease in uptake is found ( Jennings et al., 1979 ; De Jong et al., 1992 ). Preperfusion of these livers with a combination of insulin plus cortisol and/or glucose restores uptake to normal within 1/2 h suggesting that no protein synthesis is involved. This led to the suggestion that decreased ATP apparently induced by fasting has been reversed. More direct evidence for the influence of intracellular ATP on uptake of T4 in the perfused rat liver was obtained by the addition of fructose to the perfusate which led to decreased intracellular ATP levels that were paralleled by a decrease in T4 uptake ( De Jong et al., 1994 ). In the same report it was also shown that intravenous administration of fructose to human volunteers resulted in a decrease of T4 transport into the liver. Although evidently liver ATP levels could not be measured in humans, an increase in serum lactic acid and uric acid indicated indeed that ATP content in the liver was decreased. In addition to a decrease in intracellular energy stores ( Krenning et al., 1983 ), it has been suggested that other factors also lower T4 uptake into the liver in vivo in nonthyroidal illness (NTI) and in starvation. Thus, it was found that in these states, compounds that exert inhibitory effects on T4 transport into cultured rat hepatocytes, circulate in increased concentrations in the serum. We reported that in patients with uraemia the increased concentrations of a furan fatty acid 3-carboxy-4-methyl-5-propyl-2-furan propanoic acid (CMPF), and indoxyl sulphate do inhibit T4 uptake in vitro. Furthermore, bilirubin and nonesterified fatty acids (NEFA), which are often found to be increased in critical illness, also inhibited T4 uptake in vitro in hepatocytes at pathophysiological concentrations. Finally, it was also found that in caloric deprivation, NEFA was increased to concentrations that exerted a T4 transport inhibiting effect ( Lim et al., 1993a ; Lim et al., 1993b ; Lim et al., 1994 ) ( Table 2). Transport is rate limiting when any change in this process (increase or decrease) positively or negatively influences subsequent metabolism. This implies that influx of thyroid hormones is independent of intracellular metabolic capacity. When rat hepatocytes in primary culture were incubated long term with T4, T3 and rT3 in the presence of a monoclonal antibody directed against the putative iodothyronine transport carrier, or ouabain to lower the sodium gradient over the plasma membrane, a decreased clearance from the medium of these iodothyronines was found that paralleled a decreased iodide production in the medium ( Table 3). As it was shown that the added compounds had no effect on the intracellular deiodination process, it was concluded that the decrease in clearance (i.e. uptake) was rate limiting on subsequent deiodination ( Hennemann et al., 1986 ). In addition it was reported from different laboratories that compounds that inhibit T3 uptake at the plasma membrane level and did not influence nuclear binding of T3 per se, effected a decrease in nuclear occupancy that paralleled the inhibition of uptake, indicating that cellular uptake was rate limiting on nuclear binding of T3 ( Halpern & Hinkle, 1982; Hennemann et al., 1984 ; Pontecorvi et al., 1987 ). In these publications pituitary tumour cells, rat hepatocytes and skeletal myoblasts were investigated, respectively. Furthermore, uptake of T3 sulphate induced in Xenopus laevis oocytes by injection of fractionated rat liver mRNA was similar in oocytes with or without cotransfection with cRNA of type I deiodinase. In other words, increasing the capacity of oocytes to metabolize (i.e. to deiodinate) T3 sulphate, did not affect the uptake characteristics of T3 sulphate, underlining the rate limiting potential of the uptake mechanism ( Fig. 1) ( Docter et al., 1997 ). Initial uptake of T3 sulphate (T3S) in Xenopus laevis oocytes injected with water (control), fractionated rat liver mRNA, cRNA of rat liver type I deiodinase (G21) or both (G21 + mRNA). (mean ± SEM; *= P < 0.001 vs water) (modified from Docter et al., 1997 ). Also in vivo in the human there is evidence that inhibition of T4 transport into the liver is rate limiting on total plasma T3 production. In a female in her sixties an increased serum free T4 concentration was present in combination with a low plasma T3 concentration in the absence of any abnormality of serum thyroid hormone binding proteins and of nonthyroidal illness. Iodothyronine kinetic studies revealed that T4 uptake in the rapid equilibrating compartment, comprising mainly the liver, was inhibited but uptake in the slowly equilibrating compartment representing the other body tissues was not ( Fig. 2). T3 uptake in both compartments was normal. Plasma T3 production was subnormal and calculations revealed that T3 production vs substrate T4 in the liver was normal, but total liver T4 was decreased. It was concluded that the lowered plasma T3 production was caused by inhibition of T4 uptake into the liver, leading to a decreased substrate for production of T3 ( Hennemann et al., 1993 ). We have identified this abnormal thyroid hormone profile in another subject more than a decade ago ( Jansen et al., 1982 ). In this latter subject serum TBG was elevated and normalized upon administration of physiological amounts of T3. As TBG may be elevated in hypothyroidism, this indicates that the lowered T3 production caused hypothyroidism at the level of the liver. From these human studies it is suggested that inhibition of T4 transport into the liver, leading to lowered T3 production, may have biological consequences. T4 uptake into (a) the rapidly equilibrating compartment (REC) and (b) the slowly equilibrating compartment (SEC) of the subject (●) and the controls (○) during the first 400 min of T4 tracer kinetics. Values are the mean ± SEM. (From Hennemann et al., 1993 .) In NTI and in starvation, plasma T3 production is reduced and this is considered to serve as an adaptation mechanism in these pathological conditions as energy and protein (i.e. organ function) are conserved ( Kaptein, 1986). Recently it was claimed that no evidence for such an adaptive effect existed, however, the results of this study are not interpretable because of lack of a proper control group of subjects without T3 supplementation during fasting ( Beyerley & Heber, 1996). This ‘low T3 syndrome’ has at least in part been suggested to be due to alterations in thyroid hormone deiodination, i.e. a decrease in 5′deiodination in the liver ( Kaptein, 1991; Chopra, 1991). However, the point to be discussed here is to what extent lowered T3 production in these situations can additionally be explained by inhibition of T4 transport into the liver which would then be an appropriate mechanism from the pathophysiological standpoint. Several reports underline this possibility. When human obese volunteers were subjected to a calorie restricted diet (240 kcal per day), thyroid hormone kinetic studies prior and during calorierestriction revealed that there was substantial decrease of T4 transport into tissues, including the liver, of about 50%, whereas that of T3 was much less decreased, i.e. about 25%. Only about half of the lowered T3 production in these subjects could be accounted for by inhibition of 5′deiodination and it was concluded that inhibition of T4 transport into the liver contributed substantially to this process ( Van der Heyden et al., 1986 ). When T4 metabolism was studied in the isolated intact rat liver in the presence of normal serum or serum of patients with nonthyroidal illness, no decrease in liver T4 metabolism was detected, but a 50% decrease of T4 transport into the liver was found in the presence of NTI serum ( Vos et al., 1991 ). This inhibitory activity of NTI serum on T4 transport is in agreement with the above mentioned findings of inhibitory activity of compounds circulating in increased concentrations in this type of patient. A significant finding with regard to the pathophysiological appropriateness of T4 transport inhibition into the liver causing low plasma T3 was recently reported by us ( Vos et al., 1991 ; Vos et al., 1995 ). When rat hepatocytes in primary culture were incubated with T4 in the presence of serum of patients with NTI, it was found that there was a correlation (r = 0.69) between the residual transport of T4 into rat hepatocytes and the serum T3 concentration of the same serum sample tested for T4 transport inhibition ( Fig. 3). In other words, the more inhibition of T4 transport exerted by the serum, the lower the serum T3 concentration of that particular patient. Relationship between iodide production from T4 (corrected for differences in free hormone concentration) in the presence of 10% NTI serum, expressed as percentage of production in the presence of 10% serum of healthy controls and serum T3. r = 0.69. (From Vos et al., 1995 .) It is known that in NTI and starvation in the human serum TSH is either normal or depressed despite low serum concentrations of T3 and sometimes (in critical illness) also of T4 ( Kaptein, 1986). It would be inappropriate from the teleological point of view if, as a consequence of low serum T3 (and T4), serum TSH would increase (as would normally be expected), because in that case T4 secretion would be stimulated and this would lead to increased T3 production that would counteract the beneficial low T3 state. Having this in mind, it would then also be inappropriate if the compounds that inhibit T4 transport into the liver would also inhibit thyroid hormone entry into the pituitary, which would lead to an unwanted increased release of TSH. So far three of four of the mentioned compounds, i.e. CMPF, indoxyl sulphate and bilirubin were tested for their possible effects on the uptake of T4 and T3 into the rat pituitary. No inhibitory activity could be found ( Everts et al., 1995 ; Wassen et al., 1996 ). Although beyond the scope of this review, we would like to mention that we recently reported evidence that 3,3′,5-triiodothyroacetic acid (T3-AC), a thyroid hormone analogue that is produced in high amounts in NTI ( LoPresti & Dlott, 1992) and has superior TSH suppressing activity to T3 and possibly also 3,3′,5,5′-tetraiodothyroacetic acid (T4-AC) ( Carlin & Carlin, 1993), could play a role in TSH suppression during NTI ( Everts et al., 1994 , 1995). Fig. 4 illustrates the discussed mechanisms that may be operative in the low T3 syndrome. Factors operative in the low T3 syndrome in nonthyroidal illness and starvation (for explanation see text). T3-Ac: 3,3′,5-triiodothyroacetic acid (Triac), T4-Ac: 3,3′,5,5-tetraiodothyroacetic acid (Tetrac), CMPF: 3-carboxyl-4-methyl-5-propyl-2-furan propanoic acid, NEFA: nonesterified fatty acids, CHO: carbohydrates. It also seems possible that in physiological circumstances such as variations in dietary composition ( Otten et al., 1980 ) or in total amount of energy intake vs energy expenditure, intra and extracellular factors, like some of those mentioned above, are operative. These factors thus may also play a role in the determination of overall thyroid hormone bioactivity within physiological limits. Evidence is presented from in vitro experiments, intact liver perfusion and in vivo studies, that thyroid hormone transport into tissues is a regulated process and is rate limiting on thyroid hormone metabolism, notably in the liver. Furthermore, reported studies suggest that appropriate changes in thyroid hormone transport occur in pathophysiological situations, like nonthyroidal illness and starvation, and that these changes contribute to the low T3 state that is considered to be a defence mechanism in these pathological conditions. It should be stressed that many other factors than those mentioned here may be operative in the induction of the low T3 syndrome and that these factors may play a different role in the lowered T3 production depending on the type of caloric deprivation and disease entity. Discussion of these aspects is not pertinent to the purpose of this communication, but for a general review on the subject the reader is referred to Docter et al., (1993 ). Recently, Kaptein (1997) found evidence for separate transport mechanisms for T4, T3 and rT3 in human livers.