Placental Transport of Thyroid Hormone and Iodide

The major role of the thyroid is to synthesise and secrete thyroid hormones (TH). It does this by a complex process that begins with extraction of iodide from circulating blood via the sodium iodide symporter (NIS) (Dai, et al. 1996). Intracellular iodide is oxidised, under the influence of a thyroperoxidase leading to iodination of the amino acid tyrosine on the abundant thyroglobulin that occupies thyroid follicles. Iodinated tyrosines are combined to form thyroxine (4 iodine atoms, T4) and triiodothyronine (3 iodine atoms, T3). Both T4 and T3 are secreted from the thyroid gland and circulate bound to a family of thyroid binding proteins so that only a tiny fraction of T4 and T3 remain unbound (Benvenga 2005). T4 is avidly taken up by liver and deiodinated by a type 1 deiodinase (D1) (Bianco, et al. 2002) to the biologically more active T3 and the biologically inactive reverse T3 (rT3). Most circulating T3 is of hepatic origin. T4, and to a lesser extent T3 feed back at the pituitary level. Intrapituitary T4 is deiodinated to T3 by a Type 2 deiodinase (D2) and this together with T3 from the circulation inhibits synthesis and secretion of thyroid stimulating hormone (TSH), also known as thyrotropin (Shupnik, et al. 1985). TSH is a highly glycosylated protein with alpha and beta chains and is under the tonic control of the inhibitory hypothalamic hormone somatostatin (Weeke, et al. 1975) and the stimulatory thyrotropin releasing hormone (TRH) (Shupnik, et al. 1986). TSH via thyroid cell membrane TSH receptors stimulates iodide uptake (Levy, et al. 1997) and TH synthesis and secretion. Serum TH levels are controlled by the pituitary feedback mechanism.


The thyroid gland
The major role of the thyroid is to synthesise and secrete thyroid hormones (TH). It does this by a complex process that begins with extraction of iodide from circulating blood via the sodium iodide symporter (NIS) (Dai, et al. 1996). Intracellular iodide is oxidised, under the influence of a thyroperoxidase leading to iodination of the amino acid tyrosine on the abundant thyroglobulin that occupies thyroid follicles. Iodinated tyrosines are combined to form thyroxine (4 iodine atoms, T 4 ) and triiodothyronine (3 iodine atoms, T 3 ). Both T 4 and T 3 are secreted from the thyroid gland and circulate bound to a family of thyroid binding proteins so that only a tiny fraction of T 4 and T 3 remain unbound (Benvenga 2005). T 4 is avidly taken up by liver and deiodinated by a type 1 deiodinase (D1) (Bianco, et al. 2002) to the biologically more active T 3 and the biologically inactive reverse T 3 (rT 3 ). Most circulating T 3 is of hepatic origin. T 4 , and to a lesser extent T 3 feed back at the pituitary level. Intrapituitary T 4 is deiodinated to T 3 by a Type 2 deiodinase (D2) and this together with T 3 from the circulation inhibits synthesis and secretion of thyroid stimulating hormone (TSH), also known as thyrotropin (Shupnik, et al. 1985). TSH is a highly glycosylated protein with alpha and beta chains and is under the tonic control of the inhibitory hypothalamic hormone somatostatin (Weeke, et al. 1975) and the stimulatory thyrotropin releasing hormone (TRH) (Shupnik, et al. 1986). TSH via thyroid cell membrane TSH receptors stimulates iodide uptake (Levy, et al. 1997) and TH synthesis and secretion. Serum TH levels are controlled by the pituitary feedback mechanism.

Thyroxine (T 4 )
The thyroid gland is the only known source of T 4 in the body (Chopra 1996). T 4 is the most abundant iodothyronine in the circulation, present at around twenty times the concentration of T 3 , up to one hundred times more than rT 3 and more than one thousand times the concentration of any other iodothyronine derivative. Iodine constitutes about 65% (by weight) of the T 4 molecule and T 4 accounts for up to 90% of protein bound iodine in serum. The extent of overall protein binding is great, such that the serum free T 4 concentration is usually less than 0.1% of total T 4 concentration. The major TH binding proteins are; thyroxine binding globulin (TBG), transthyretin (TTR) and albumin as well as several minor carriers. The less biologically active T 4 is largely deiodinated in peripheral tissues to the bioactive form of TH, T 3 ( Figure 1). Alternatively, T 4 may be converted to the inactive metabolite, rT 3 . Both T 3 and rT 3 can be further metabolised in peripheral tissues to 3,3'diiodothyronine (T 2 ).

3,5,3'-triiodothyronine (T 3 )
T 3 was first discovered in human serum in 1951 (Gross and Leblond 1951) and was found to be several times more biologically potent than T 4 in producing the classic effects of THs (Gershengorn, et al. 1979). T 3 is formed by the removal of an iodine atom, by deiodinase enzymes, from the phenolic ring of T 4 (Figure 1). Like T 4 , T 3 in serum is bound to TBG, TTR and albumin. As indicated above the main source of T 3 is peripheral conversion from T 4 in addition to some limited direct thyroid gland secretion.

3,3',5'-triiodothyronine (rT 3 )
Reverse T 3 (rT 3 ) differs from T 3 in that iodine is removed from the inner or tyrosyl ring of T 4 rather than the outer or phenolic ring (Chopra 1996) (Figure 1). rT 3 was first found in the blood of rats in 1956, it has little or no activity when administered to animals and its metabolism is extremely rapid. The main source of rT 3 is inner-ring deiodination of T 4 in peripheral tissue, predominantly liver (Chopra, et al. 1975).

Thyroid hormone receptors
Thyroid hormones act by binding to specific nuclear receptors that interact with DNA causing activation or repression of transcription (Tata and Widnell 1966). In the 1960s it was noted that nuclear RNA transcription preceded many of the physiological effects of T 3 and this led to the discovery of high affinity nuclear receptors for T 3 (Oppenheimer, et al. 1972;Samuels and Tsai 1973). The cloning from many species of multiple cDNAs encoding proteins with the characteristics of TH receptors (TR) brought about the realisation that there is a family of TRs (Evans 1988). These have molecular weights of 50 to 55kDa and bind T 3 with high affinities (Sap, et al. 1986;Weinberger, et al. 1986). The TR isoforms have substantial amino acid sequence homology with the steroid hormone receptors (Evans 1988). Levels of nuclear TRs correlate well with the developmental and tissue specific effects of T 3 (Chan, et al. 2002).

Mild maternal thyroid dysfunction and impaired neuro-cognitive function in the offspring
There is now ample evidence that even mild maternal hypothyroidism is associated with impaired fetal neuro-cognitive outcome. World wide, the most common cause of maternal hypothyroidism is iodine deficiency (Andersson, et al. 2010;Pharoah and Connolly 1991). Severe iodine deficiency is well known to cause severe mental retardation, neuro-muscular impairment and short stature, a syndrome known as cretinism. While cretinism was described hundreds of years ago, the link between milder degrees of iodine deficiency and reduced intelligence of the offspring was first described described in the Himalayas in the early 20 th century. Subsequent work in the highlands of Papua New Guinea by Pharoah confirmed the link between milder degrees of iodine deficiency and reduced neurocognitive function and provided evidence of correlations between reduced maternal T 4 (but not T 3 ) levels and reduced intelligence and coordination (Pharoah and Connolly 1991). A small, more recent Italian study (Vermiglio, et al. 2004) confirmed reduced IQ levels and a high incidence of attention deficit disorder in offspring of women from an area of moderate iodine deficiency. There was a strong inverse correlation between maternal mid gestation www.intechopen.com free T 4 (FT 4 ) and offspring IQ. A Spanish study emphasised the importance of early iodine supplementation in preventing neurological damage (Berbel, et al. 2009).
A link between mild maternal hypothyroidism and impaired intellectual development of offspring was suggested in the 1960s. Over several years, from the late 1960s to the 1970s Man and co-workers published data from 1349 women whose serum T 4 was estimated by measuring butanol-extractable iodine during early and late pregnancy (Man, et al. 1991); three percent were hypothyroxinemic. Developmental and intellectual outcomes of progeny of 210 euthyroxinemic women, 15 hypothyroxinemic women adequately treated with thyroxine and 21 women inadequately treated with T 4 were compared at eight months, four and seven years of age. Mothers were well matched for intelligence, years of education and chronological age. At each age children of mothers with inadequately treated hypothyroidism had lower mean developmental and intellectual scores.
Subsequent studies using more precise measurements of TH status have yielded similar results. Haddow and co-workers (Haddow 1999) measured TSH levels in stored blood taken from over 25,000 pregnant women. Seventy-five women (0.3%) had levels at or above the 99.7 th percentile; 47 were contacted and agreed to allow neuropsychological testing of their children at seven to nine years of age. Children of an additional 15 women with serum TSH levels between the 98 th and 99.6 th percentiles were also tested; 77 % of these 62 women had positive thyroid antibodies (markers of potential autoimmune thyroid disease). Results were compared with those of 124 women with normal thyroid function (14% of whom had positive antithyroid antibodies) and demonstrated significantly (p=0.06) reduced full-scale IQ scores, reduced verbal IQ scores (p=0.06) and word discrimination (p=0.01).
Children of women with normal thyroid stimulating hormone (TSH) levels but FT 4 levels less than the 10 th percentile during early pregnancy (i.e. technically normal thyroid function) had significantly lower Bayley Psychomotor Development Index scores at ten months of age than children of women with higher FT 4 levels. Mothers with FT 4 levels below the 10 th percentile had a significantly higher incidence of positive antithyroid antibodies (Pop, et al. 1999).
These, and other (Ghassabian, et al. 2011;Klein, et al. 2001;Kooistra, et al. 2006;Li, et al. 2010), clinical studies provide strong evidence of a relationship between reduced or low normal early pregnancy FT 4 levels and neuro-cognitive functioning of offspring. This relationship holds whether maternal thyroid dysfunction results from iodine deficiency or autoimmune thyroid disease. These findings suggest that maternal thyroxine is required for early fetal brain development and that maternal TH crosses the placental barrier.

Maternal hypothyroxinemia
Under normal circumstances maternal FT 4 levels rise in the mid first trimester in response to a surge in maternal human chorionic gonadotropin (hCG) levels (Fisher 1983) (Figure 2). HCG is a double chain glycosylated protein secreted by placenta from early pregnancy that shares a common alpha subunit with TSH. The beta subunits of each hormone and their cell membrane receptors are also significantly homologous. HCG stimulates the normal maternal thyroid via the TSH receptor to synthesise and secrete TH. Very high hCG levels as are seen in women with excessive pregnancy induced vomiting (hyperemesis) and women with placental malignancy (choriocarcinoma) can cause maternal hyperthyroidism. As www.intechopen.com normal pregnancy progresses hCG levels fall and this is reflected in falling FT 4 levels ( Figure  2). Impairment of maternal TH secreting capacity from iodine deficiency or autoimmune thyroid disease can blunt the physiological first trimester surge in maternal TH secretion, or if more severe can result in maternal hypothyroidism. Fig. 2. Ontogenic changes in maternal and fetal thyroid gland and hormone function (compiled with data from (Fisher 1983;Fuse 1996)). hCG, human chorionic gonadotropin; TSH, thyroid stimulating hormone/thyrotropin; T 4 , thyroxine.
Mild iodine deficiency is prevalent in many parts of the world, including some European countries, the USA and Australia and as discussed above is a common cause of maternal hypothyroxinemia. Iodine deficiency can be exacerbated by maternal smoking, which increases blood thiocyanate levels. Thiocyanate competitively blocks the sodium iodide cell membrane symporter (NIS) responsible for transfer of iodide into maternal and fetal thyroid and for materno-fetal transfer of iodide by the placenta (Manley, et al. 2005) . Thyroid autoimmunity is common in pregnant women. Autoimmune thyroid disease (AITD) is associated with autoantibodies to thyroperoxidase (TPO). This enzyme oxidises iodine in the presence of hydrogen peroxide, facilitating iodination of tyrosine and synthesis of TH. Enzyme activity is blocked by anti-TPO autoantibodies. About 10% of pregnant women have positive anti-thyroperoxidase antibodies at 14 weeks gestation and about 2.5% have asymptomatic hypothyroidism (Lazarus 2005).

Ontogeny of the human fetal hypothalamic-pituitary-thyroid axis
The thyroid gland is the first endocrine organ to develop in man where it originates as a thickening in tissue destined to become the tongue (Fuse 1996). This bud descends to the level of the larynx, forming two lobes connected by an isthmus. Recognisable thyrocytes are present by the end of 7 weeks gestation but mature thyroid follicles containing colloid do not appear until after 13.5 weeks gestation. The fetal thyroid appears to be able to accumulate iodine by about 12 weeks gestation ( Figure 2). Iodine concentrating capacity increases from about 12 weeks, reaching a peak at about 24 weeks gestation. T 4 is synthesised by about 19 weeks gestation (Fuse 1996) (Figure 2).
The pituitary gland has two separate origins, the glandular component (appearing in the developing mouth at about 3 weeks gestation) and the neural primordium (extending from the hypothalamus about 5.5 weeks). The glandular primordium forms Rathke's Pouch, which loses its connection with the oral cavity by about 8.5 weeks. This adenohypophyseal primordium forms the anterior pituitary and forms a close association with the developing neural structure forming the neurohypophysis. Cells staining for the alpha subunit of TSH can be seen from 8-12 weeks gestation. The beta subunit can be identified by 13-15 weeks gestation. Immunoreactive TSH can be identified in fetal serum by 12 weeks gestation but levels are low until a rapid increase at 18-22 weeks.
These data suggest that while the fetal thyroid can concentrate iodine and synthesise TH in the late first trimester. TSH regulated TH secretion may not occur until as late as 18-20 weeks gestation ( Figure 2).

Iodide
Detectable iodine in the amniotic fluid of pregnant rabbits fed potassium iodide was reported in 1859 and in 1872 similar results were reported in man (quoted by (Gersten 1954). Transfer of radioiodine from the maternal to fetal circulations of the guinea pig was reported in 1955 (Logothetopoulos and Scott 1955) who noted that transport was blocked by sodium thiocyanate, suggesting an active transport process. In thyroid iodide is transferred from blood to the thyroid cell by the sodium iodide symporter (NIS) and iodide efflux from the thyrocyte is mediated by another transporter called Pendrin. NIS (Bidart, et al. 2000;) and Pendrin (Bidart et al. 2000) have been reported in trophoblasts and functional studies in a trophoblast cell line suggest that these are responsible for iodide influx and efflux in placenta (Manley et al. 2005). Sodium thiocyanate is a powerful inhibitor of NIS. Recently SLC5A6, a placental sodium/multivitamin transporter has been identified as an iodide transporter (de Carvalho and Quick 2011) but its role in placental iodide transfer is as yet unclear.

Thyroid hormone
Early human studies suggested that there was significant materno-fetal transfer of TH (Raiti, et al. 1967) and this was supported by detection of significant amounts of TH in cord blood of infants unable to synthesise TH (Vulsma, et al. 1989). Investigation of transfer of T 4 in the isolated perfused human placenta demonstrated that the abundant type 3 deiodinase (D3) www.intechopen.com significantly limited transfer so that the fetal circuit T 4 reached only 0.008% of maternal levels. Inhibition of D3 by iopanoic acid increased fetal T 4 levels to 30% of those in the maternal circuit (Mortimer, et al. 1996). Membrane TH transporters were demonstrated in human trophoblasts and choriocarcinoma cell lines (Mitchell, et al. 1992) mediating uptake and efflux of TH. The identities of these transporters were subsequently ascertained.
There is now clear evidence of THs in fetal serum, coelomic and amniotic fluid and brain in early pregnancy, a time when the fetal thyroid gland has not yet developed the capacity to secrete TH (Calvo, et al. 2002). These studies were done in human tissues obtained from fetuses as early as 5-6 weeks gestation. The early feto-placental unit (up to 12-13 weeks gestation) consists of the fetus floating in amniotic fluid (AF) within an amniotic sac. This in turn is contained within an exocoelomic cavity containing coelomic fluid (CF) in which floats a prominent secondary yolk sac. The yolk sac, an extension of the fetal gastrointestinal tract and circulation, secretes and resorbs a variety of proteins. The CF is contained within a uterus lined by the early placenta. The early placenta is poorly vascularised, relatively hypoxic and covers the surface of the chorionic sac. The CF is protein rich, containing albumin and TTR whereas AF is essentially a low protein ultrafiltrate of maternal serum containing placental and yolk sac secretory products. Its volume is increased by urine secreted by the developing fetal kidneys.
With increasing gestational age, the yolk sac and the majority of the placenta regress and the exocoelomic cavity is largely obliterated by an expanding amniotic sac. The placenta forms a circumscribed disc-like structure and trophoblasts, the epithelial cells that mediate maternofetal exchange, invade the uterine vasculature. This allows development of a mature maternal and fetal circulation within the placenta and a rise in placental oxygen levels. There is increasing evidence that the changing oxygen levels within the developing placenta have major effects on trophoblast function (Patel, et al. 2010b).
Human studies have provided considerable insight into TH levels in fetal serum (FS), CF and AF. It is clear that maternal TH crosses the placenta, entering the human embryonic cavities and fetal blood well before the fetal thyroid is secreting its own TH (Calvo et al. 2002;Contempre B 1993) (Figure 3). Total T 4 was detectable in several sets of CF fluid obtained between 5.8 and 11 weeks gestation, with set means ranging from 950 to 1280 pmol/litre. Total T 3 levels were very much lower (2.50 to 2.82 pmol/litre). rT 3 levels were high, ranging from 2.1 to 5.48 nmol/liter. T 4 , T 3 and rT 3 were also found in AF sampled from 8 weeks on. Total T 4 ranged from 63 to 2041 pmol/litre, whereas total T 3 was 6 to 12 pmol/litre. rT 3 levels were again relatively high at 210 to 3430 pmol/litre. The high ratios of rT 3 to T 4 suggest active Type 3 deiodination (see below). The yolk sac synthesises and secretes the T 4 binding protein TTR and low levels of TTR were detected in CF from 7 weeks gestation, increasing with increasing gestational age. Mean free T 4 levels of 2.5 to 2.82 pmol/litre could be estimated in CF. Mean FT 4 in AF ranged from 6.45 to 20.43 pmol/litre.
These TH levels, expressed as a percentage of corresponding maternal levels in early pregnancy, are shown in Figure 3. Although fetal total T 4 is only about 5% of maternal levels and CF and AF levels are less than 1% of maternal values, free T 4 levels are much higher (approaching maternal levels) in these compartments due to very low TBG levels (Calvo et al. 2002).  (Calvo et al. 2002)).
Maternal T 4 and FT 4 levels increase during the first and second trimester. T 4 levels in CF correlate with those in maternal serum and in the second trimester fetal blood levels also correlate with maternal values. Interestingly T 3 and free T 3 levels in CF do not correlate with maternal levels (Calvo et al. 2002). These data strongly argue for significant transfer of maternal THs to the coelomic and amniotic fluids and into the fetal circulation. The significant gradient of total T 4 from maternal to fetal circulations and high levels of rT 3 in CF and AH suggest that active conversion of T 4 to rT 3 by D3 in placenta, placental membranes and the fetus strongly modulates fetal T 4 supply. Despite this fetal T 4 levels are strongly determined by maternal levels.
There is also strong clinical evidence that, at least in the presence of a hypothyroid fetus, transfer of maternal T 4 to the fetus continues throughout pregnancy. In 1989 Vulsma and colleagues reported that cord blood T 4 levels from infants with complete thyroid agenesis or a complete defect in organification of iodine and capacity to synthesise TH had T 4 levels of 30 -75 nmol/liter, which must have been of maternal origin (Vulsma et al. 1989). Placental D3 levels appear normal in these infants (Koopdonk-Kool, et al. 1996).

Thyroid hormone and development of the fetal brain
Although fetal serum T 3 levels are very low at that time, T 3 and TH receptors have been identified in human fetal brain as early as 9 weeks gestation (Bernal and Pekonen 1984;Chan et al. 2002). Brain T 3 is locally produced by Type 2 deiodination of T 4 and rat studies indicate that brain T 3 in the hypothyroid rat fetus cannot be replenished by maternal T 3 administration but requires maternal T 4 . T 4 enters brain via the cerebral circulation but a significant proportion also appears to be transferred through the choroid plexus, a potent source of cerebro-spinal fluid (CSF) transthyretin (TTR). TTR represents about 20% of CSF www.intechopen.com protein and is the major TH transporter in CSF. Membrane TH transporters, notably monocarboxylate transporter (MCT) 8, are important mediators of neuronal T 3 uptake. Intracellular T 3 is translocated to the cell nucleus where it binds to TRs and, following homodimerisation or heterodimerisation with retinoic acid receptors, it binds to TH response elements present in many genes, positively or negatively regulating them. T 3 has major effects on neurogenesis, cell migration and myelination in the developing brain (Patel, et al. 2011).

Iodothyronine deiodinases
The deiodinases play an important role in coordinating TH action during vertebrate development and they regulate TH action within selected tissues during development and adulthood. Three deiodinases have been identified, deiodinase type 1 (D1), deiodinase type 2 (D2) and deiodinase type 3 (D3), and all are integral membrane proteins and are selenoenzymes that have regions of high homology surrounding the selenocysteine residue at the active site (Bianco and Larsen 2005). Interestingly, the deiodinases differ in tissue distribution, substrate specificities, catalytic profile, physiological functions and regulation. Of the three deiodinating enzymes, only D2 and D3 have been identified in the placenta (Koopdonk-Kool et al. 1996). Placental D3 activity is much greater (~200 times in first trimester and ~400 times at term) than D2 activity, however the activity and expression of both D2 and D3 falls as gestation progresses (Chan, et al. 2003;Koopdonk-Kool et al. 1996;Stulp, et al. 1998). D2 is an outer ring deiodinase (Nelson, et al.) found primarily in brain, pituitary, brown adipose tissue, thyroid and placenta with a preference for T 4 > rT 3 as a substrate (Figure 1). D2 converts the biologically inactive T 4 to the active T 3 . Conversely, D3 catalyses inner ring deiodination (IRD) of T 4 and T 3 and is mainly present in placenta, brain and skin with T 3 being the preferred substrate over T 4 (Figure 1) (Gereben, et al. 2008). D3 inactivates T 3 to T 2 or T 4 to rT 3 . T 2 and rT 3 were previously considered inactive metabolites because they do not bind TH receptors, however more recently rT 3 has been implicated in actin polymerisation (Farwell, et al. 2005) and T 2 in stimulation of mitochondrial respiration (O'Reilly and Murphy 1992). D2 and D3 have been found to be present in the placenta throughout gestation (Chan et al. 2003). D2 has been localised to the villous cytotrophoblasts cells in the first trimester with expression in villous syncytiotrophoblasts (ST) variable and weak. In contrast, D3 has been localised to the villous ST cells and syncytial sprouts with expression in villous CTs focal and weak. In the third trimester villous ST expressed D2 and D3, whilst villous CTs were stronger for D2 than D3 (Chan et al. 2003). The localisation of the deiodinases suggests that they may regulate the amount of maternal TH reaching fetal circulation. D3 localised to the villous ST layer, which is in direct contact with the maternal circulation can protect the fetus from excessive maternal TH.

Placental TH membrane transporters
TH membrane transporters mediate cellular uptake and efflux of TH (Hennemann, et al. 2001;. Trophoblast membrane transport of TH was first reported in 1992 by Mitchell et al, using the human placenta choriocarcinoma cell line, JAR (Mitchell et al. www.intechopen.com 1992). Since this time significant findings in placental TH transporters have been made. TH transporters identified in the human placenta to date include the monocarboxylate transporters MCT8 (Chan, et al. 2006) & MCT10 ), L-type amino acid transporters LAT1 (Okamoto, et al. 2002) and LAT2 (Park, et al. 2005) and organic anion transporting polypeptides OATP1A2 (Patel, et al. 2003) and OATP4A1 (Patel et al. 2003) (Sato, et al. 2003). However, their individual contribution to placental TH transport has yet to be elucidated.

MCT8 & MCT10
Freisema et al identified MCT8 as a TH transporter with a preference for T 3 over T 4 (Friesema, et al. 2003). Both MCT8 and MCT10 mRNAs have been identified in placenta however it is only recently that both mRNAs were identified in early human placenta (from 6 weeks gestation) and both increased in expression throughout pregnancy (Loubiere, et al. 2010). Immunohistochemical studies have localised MCT8 and MCT10 proteins to villous ST, CT and extra villous trophoblasts (EVTs) in first trimester placental tissue, with marked immunostaining of MCT10 in the CT layer (Chan et al. 2006;Loubiere et al. 2010). Both proteins localised to villous STs in term placental tissue.

OATP1A2 & OATP4A1
The expression of OATP1A2 and OATP4A1 in placenta and their ability to mediate transport of T 4 , T 3 and rT 3 have implicated both proteins as TH transport mechanisms in the placenta (Patel et al. 2003;Sato et al. 2003). RT-PCR analysis revealed OATP1A2 mRNA increases in human placental tissue throughout gestation, whilst OATP4A1 decreases to mid-gestation followed by an increase towards term. Western blotting results suggest no significant change in both proteins throughout gestation.
OATP1A2 and OATP4A1 proteins have been localised to villous STs in the first trimester with OATP1A2 also found moderately strong in villous CT and extra villous trophoblasts (EVTs). In term tissue both proteins revealed diffuse, weak expression, with OATP4A1 preferentially localised to the apical surface in STs (Loubiere et al. 2010;Sato et al. 2003).

LAT1 & LAT2
LAT1 protein has been localised to the ST layer in placenta and LAT2 to the apical and basal membranes of ST at term (Hoeltzli and Smith 1989;Lewis, et al. 2007;Okamoto et al. 2002;Ritchie and Taylor 2001). Localisation of the proteins throughout gestation has yet to be elucidated. A more recent study of LAT1 mRNA expression in human placenta revealed that it increases with gestation, whilst LAT2 did not alter.
Considering in the first trimester STs are in direct contact with maternal blood, MCT8, OATP4A1 and LAT1 may be the key transporters for TH uptake from the maternal circulation as they are preferentially localised to the apical membrane of STs (Chan et al. 2006;Ritchie and Taylor 2001;Sato et al. 2003), whilst OATP4A1 and LAT2 may play more prominent roles later in gestation. TH membrane transporters in the placenta most likely act in concert to regulate the passage of TH transported from the maternal to the fetal circulation throughout gestation. www.intechopen.com

Placental TH binding proteins
Thyroid hormone is extremely hydrophobic and carried in serum bound to three hepatically secreted binding proteins, thyroxine binding globulin (TBG), transthyretin (TTR) and albumin (Schussler 2000). Previously our group has described the synthesis of the TH binding proteins TTR and albumin by human placenta (McKinnon, et al. 2005).

Transthyretin (TTR)
Studies have reported high levels of TTR in fetal serum as early as 13 wks gestation (Fryer, et al. 1993) and considering little TTR mRNA is detectable in fetal liver at this time (Jacobsson 1989) we propose that the fetal TTR present may be of placental origin. Using placental explants and the choriocarcinoma cell line JEG3, we demonstrated that TTR was not only synthesised (McKinnon et al. 2005) but also secreted mainly through the apical cell membrane of these cells (Landers, et al. 2009).
We have also shown internalisation of TTR by placental explants and JEG3 cells which increased in the presence of T 4 (Landers et al. 2009). This increased internalization occurred under TTR: T4 ratios that favoured TTR tetramer formation (Landers et al. 2009) Similar increases have been described in astrocytoma cells (Divino and Schussler 1990b). Crosslinking studies of TTR bound to 125 I-T 4 suggest that TTR-T 4 is internalised by JEG3 cells as a TTR-T 4 complex. However, further research is required to confirm this finding and elucidate the mechanisms by which TTR or TTR-T 4 is internalised by the placenta. The protective role of TTR was postulated when binding of TTR to T 4 in placental cytosol was inhibited by addition of mefanamic acid, resulting in an increase in T 4 deiodination as determined by HPLC (McKinnon et al. 2005). The exact mechanisms of this are yet to be confirmed. TTR internalisation has previously been observed in ependymoma cells (Kuchler-Bopp, et al. 2000), chicken oocytes (Vieira, et al. 1995) and kidney proximal tubules (Sousa, et al. 2000) via megalin-mediated endocytosis. Receptor-mediated uptake of TTR was first described in HepG2 cells, primary rat hepatocytes, renal adenocarcinoma cells, neuroblastoma and transformed lung cells (Divino and Schussler 1990a). Similarly apical secretion of TTR has been described in the choroid plexus (Dickson 1986) and retinal pigment epithelium (Jaworowski, et al. 1995). Apical secretion of TTR by trophoblast cells into what would be the maternal circulation would increase local serum TTR concentrations at the surface of trophoblast cells. We propose that this would result in increased binding of maternal T 4 to placental TTR where TTR may serve to protect T 4 from deiodination and deliver T 4 or the TTR-T 4 complex to trophoblast cells of the placenta for eventual delivery to fetal circulation.
Many chemicals, including a variety of environmental pollutants, bind to TTR and displace T 4 . These agents can cross the placenta and interfere with fetal thyroid function (Koopman-Esseboom, et al. 1994). Their role in interfering with TTR TH transfer is however yet to be investigated.

Albumin
There is an abundance of albumin during human pregnancy that comes into direct contact with the trophoblast cell layer. Early studies have demonstrated that maternal albumin is internalised by placental explants and in the syncytiotrophoblast layer the protein is either apically recycled into the maternal circulation or degraded (Lambot, et al. 2006). The exact role albumin plays at the trophoblast surface remains unclear and requires further investigation www.intechopen.com particularly in first trimester tissue and in the presence of TH. Furthermore, the fate of albumin synthesized by placenta (McKinnon et al. 2005) is also of interest as, like TTR, it may also play a protective role for TH and aid in the transport of TH to fetal circulation.

Iodide transporters
It has been long recognised that maternal iodide crosses the placenta to the fetal circulation (Logothetopoulos and Scott 1956). Two transporters carry out placental iodide transport from the maternal to the fetal circulation: the sodium-iodide symporter (NIS) and Pendrin. Both transporters were first described in thyroid (Dai et al. 1996) followed by placenta ) and kidney (Spitzweg, et al. 2001). NIS is a membrane-bound glycoprotein and the fifth member of the solute carrier family (SLC5A5) . As its name suggests, NIS simultaneously takes up two Na+ and one Iion from extracellular fluid (i.e. blood) into cells (Figure 4). This process is an active transport powered by the sodium gradient across the cell membrane generated by sodium potassium pumps, (Na+/K+ ATPase). Pendrin is an anion exchanger, encoded by the Pendred syndrome gene (PDS) (Manley et al. 2005;Royaux, et al. 2001;Scott and Karniski 2000) activated by high concentration of intracellular iodide (Yoshida, et al. 2004). Pendrin activity is dependent on NIS transporting iodide into the cells (Figure 4). The physiological functions of these two transporters are similar in placenta and thyroid. In thyroid gland, NIS is localised at the basal membrane of thyrocytes and takes up iodide from the blood stream into the cells (Caillou, et al. 1998;Castro, et al. 1999;Dai et al. 1996;Royaux, et al. 2000;Yoshida, et al. 2002). Pendrin is expressed on the apical membrane of thyrocytes and releases iodide into thyroid follicles for TH synthesis Mian, et al. 2001;Spitzweg, et al. 2000). In placenta, NIS is localised to the apical membrane (maternal side) of syncytiotrophoblasts, which directly contacts with maternal blood and influxes iodide into the cells (Figure 4). Conversely, Pendrin is located in the basal membrane (fetal side) of syncytiotrophoblasts and effluxes iodide into the extracellular space (Bidart et al. 2000;Manley et al. 2005;Mitchell et al. 2001). Mutations of NIS have been found in patients previously found to have congenital hypothyroidism due to an iodide transport defect. Some of these NIS mutations have been confirmed to cause failure of membrane targeting (Kosugi, et al. 1999;Kosugi, et al. 1998;Matsuda and Kosugi 1997;Pohlenz, et al. 2000). Pendred syndrome, a recessively inherited disorder causing congenital deafness and thyroid goitre is caused by a genetic defect in the PDS gene (Everett, et al. 1997;Kopp, et al. 1999;Royaux et al. 2000;Taylor, et al. 2002).

Cell model for study of placental iodide transport
During development of the human placenta, cytotrophoblasts fuse to form multinucleated syncytiotrophoblasts that form the surface of the placental villi and are directly bathed in maternal blood. Syncytiotrophoblast cells conduct maternal-fetal nutrient and gas exchange and have a distinct endocrine function to produce and secrete pregnancy hormones such as hCG and placental lactogen. BeWo cells are a human trophoblast-derived choriocarcinoma cell line that shares many features with primary trophoblasts in culture. They form a welldifferentiated monolayer, undergo syncytialization and, secrete hCG (Bode, et al. 2006;Liu, et al. 1997;Sullivan 2004). They have been used widely in placental transport studies, e.g. glucose (Antony, et al. 2007;Araujo, et al. 2008;Baumann, et al. 2007;Di Simone, et al. 2009;Mark and Waddell 2006), amino acid (Jones, et al. 2006a, b;Novak, et al. 2006), iron (Danzeisen andMcArdle 1998;Gambling, et al. 2001), fatty acid (Johnsen, et al. 2009;Tobin, et al. 2009), and drug and toxicity studies (Araujo, et al. 2009;Hirano, et al. 2008;Magnarin, et al. 2008;Prouillac, et al. 2009). BeWo cells express both of the iodide transporters, NIS and Pendrin (Manley et al. 2005). NIS proteins are located in the apical membrane of polarized BeWo cells while Pendrin is located to the basolateral membrane. BeWo cells demonstrate significant uptake and efflux of iodide, with kinetic and inhibitory characteristics consistent with these transporters (Li, et al. 2007;Manley et al. 2005). The human JAr placental choriocarcinoma cell line has also been used in iodide transport studies but radio-labelled iodide (I 125 ) uptake was dependent on the presence of exogenous hCG in the culture medium (Arturi, et al. 2002a) limiting their use in many studies. The JEG-3 cell line also expresses both NIS and Pendrin, however JEG-3 cells do not take up measureable amounts of I 125 , even after hCG treatment, since the NIS protein is not localized to the apical membrane (unpublished observation). Primary trophoblasts and placental explant cultures might appear to be the ideal model for iodide transport studies, but due to variations in sample quality and low NIS expression their use is limited. Clearly, BeWo cells possess the physiological properties required and are the best cell model for studies of iodide transport.

Regulation of placental iodide transport
In thyroid follicular cells, NIS is regulated by serum levels of the pituitary derived TSH (Ajjan, et al. 1998;. However, unlike thyroid, placental syncytiotrophoblasts not only express NIS but also produce hCG. NIS expression and iodide uptake is increased in Jar cells exposed to hCG, and withdrawing hCG from the culture medium, leads to decreased NIS expression and iodide uptake (Arturi et al. 2002a). In BeWo cells NIS mRNA and membrane protein is up-regulated by hCG, and is accompanied by increased levels of iodide uptake (Li et al. 2007). Clearly hCG is an important regulator of placental iodide transport. In thyroid, hCG can stimulate NIS expression, subsequently increasing iodide uptake and TH synthesis and secretion (Arturi, et al. 2002b;Kraiem, et al. 1994).
An inhibitory effect of excess iodide on iodide organification in the normal thyroid (Wolff-Chaikoff effect) was reported by Wolff andChaikoff in 1949 (Wolff, et al. 1949). Following the discovery of NIS, persuasive evidence suggested that the inhibitory effect of iodide is associated with a decrease in NIS mRNA and protein levels, subsequently reducing iodide transport to the thyroid (Eng, et al. 1999;Eng PH 2001;Glatt, et al. 2005). In iodine deficient rats, NIS mRNA is up regulated in fetal thyroid, as well as in the placenta (Schroder-van der Elst, et al. 2001). In BeWo cells iodide also caused a significant decrease in NIS mRNA and apical membrane protein, followed by a decrease in levels of iodide uptake (Li et al. 2007). These studies suggest that self-regulation of iodide uptake by intracellular iodide occurs in thyroid and placenta. In BeWo cells, iodide decreases hCG mRNA expression and protein secretion. Interestingly cord blood TH levels in neonates of mothers with moderate iodine deficiency and hypothyroxinemia are significantly higher than maternal levels (Glinoer 1997;Glinoer, et al. 1992). Although no measurements of serum inorganic iodine concentrations were made in these cases it is tempting to hypothesise that in the face of moderate maternal iodide deficiency up-regulated placental NIS expression and increased materno-fetal placental iodide transport may allow the fetus to maintain normal TH levels. Excessive maternal iodide intake may, on the other hand, down-regulate NIS expression in placenta and reduce iodide transport to the fetus.

Changes in placental oxygen concentration through gestation
The adaptive processes of the developing placenta have long been studied demonstrating that under rapid physiological changes specific genes and associated proteins are affected, leading to altered nutrient, hormone and waste exchange between the mother and fetus. Many of these physiological changes relate to changing oxygen concentrations in the placenta that relate to placental vascularisation by the end of the first trimester of pregnancy. In the first trimester, EVT cells invade into the decidua, occluding uterine spiral arteries (Jauniaux, et al. 2003). This restricts blood flow into the intervillous space (IVS) resulting in a low oxygen environment that is essential for placental and embryonic development (Burton, et al. 1999;Genbacev, et al. 1997;Huppertz and Peeters 2005;Osol and Mandala 2009). Measurements with oxygen sensitive probes during ultrasonography at 8 weeks gestation have established that the oxygen concentration within the IVS is <20 mmHg or 3-5% O 2 (Rodesch, et al. 1992). Oxygen concentrations within the underlying www.intechopen.com maternal decidua are approximately 60 mmHg or 8-10% O 2 . Between weeks 11-12 of gestation, uterine spiral arterioles become patent, allowing significant maternal blood flow and increasing oxygen levels (Carter 2009;Jauniaux, et al. 2000;Rodesch et al. 1992). Recent in vitro studies have demonstrated the capacity of EVTs to initiate apoptosis of vascular smooth muscle cells (VSMC) and endothelial cells. This may represent the mechanism of the physiological modification of the uterine spiral arterioles that leads to the increased vascular compliance and circumference of early pregnancy (Ashton, et al. 2005;Harris, et al. 2006;Moffett-King 2002;Zygmunt, et al. 2003). The resulting increased blood flow and growth of the vascular and capillary network meets the demands of the growing fetus (Burton 2009).
Many placental transport processes are regulated by low oxygen levels, including hormonal, glucose, amino acid (system A -a sodium dependant transport process of amino acids) and iodide transporters. Here we describe potential placental adaptations to T 4 uptake through regulation of TTR expression and iodide uptake through regulation of the NIS cell membrane transporter.

Low oxygen in the placenta and NIS expression and function
As described earlier, placental iodide transport to the developing fetus is essential to allow the fetal thyroid to produce TH from about week 12 of gestation. We have demonstrated down regulation of mRNA and protein expression of the NIS transporter in human BeWo placental cells cultured at 1% oxygen in comparison to controls cultured at 8% oxygen (Li, et al. 2011). A significant reduction in iodide uptake in cells cultured at low oxygen was also observed (Li et al. 2011). This suggests that the increasing oxygenation of the placenta at about 12 weeks gestation may up regulate NIS expression leading to increased iodide transport at a time when the developing fetal thyroid requires maternal iodide. hCG expression measured in the same study mirrored the expression of NIS. hCG regulates Pax8, an essential protein that must bind to the NIS promoter and enhancer region to up regulate NIS transcription (Schmitt, et al. 2001). This highlights the complexity of placental NIS expression, with oxygen concentrations and hormonal expression both playing a role in NIS regulation.

Low oxygen, the placenta and transthyretin (TTR) expression and function
As described above, the low oxygen environment within the placenta clearly regulates a number of important genes including those related to specific transport processes. Recently, our group demonstrated that low oxygen levels up-regulate expression, secretion and reuptake of TTR (Patel, et al. 2010a). Human placental JEG-3 and primary trophoblast cells cultured under low oxygen conditions (1-3% O 2 ), showed an increase in TTR mRNA and protein expression. Using fluorescent and 125 I labelled TTR, increased up-take into trophoblast cells was observed using the same low oxygen culture conditions. The uptake studies were conducted in the presence of excess T 4 which causes TTR tetramerisation, a process that appears critical for significant TTR uptake by cells (Landers et al. 2009). This study was the first to demonstrate physiological regulation of trophoblast TTR uptake. Although speculative at this stage, this could suggest increased transplacental delivery of thyroxine (T 4 ) during the first trimester of pregnancy, when fetal requirement for maternal T 4 is higher (as detailed earlier in this chapter) and when a physiological low oxygen environment is present.

www.intechopen.com
Increased concentrations of placental TTR protein have been demonstrated in patients with preeclampsia (Gharesi-Fard, et al. 2010). This increased TTR expression probably relates to placental hypoxia, which is common in pre-eclampsia, but further investigation of TTR in the cause and/or diagnosis of pre-eclampsia is warranted.

Conclusion
Adequate supplies of maternal TH and iodide are essential for normal fetal brain development, with TH critical in the first trimester and iodide from the second trimester on. It is increasingly apparent that even very mildly reduced maternal T 4 levels may impair the offspring's neuro-cognitive function. Impaired maternal thyroid function from iodine deficiency or autoimmune thyroid disease is common and may represent a major public health issue. The mechanisms underlying materno-fetal transport of iodide and TH are slowly being unravelled. NIS and Pendrin mediate iodide transfer in placenta, as they do in the thyroid gland. HCG, iodide and placental oxygen levels regulate placental NIS. NIS transport is blocked by thiocyanate, a component of tobacco smoke, which may exacerbate marginal iodide deficiency in smoking mothers.
TH transfer appears to involve trophoblast membrane TH transporters but the important role of placental TTR requires further evaluation. There is increasing evidence that placental TTR participates in a shuttle in which TTR secreted by the apical trophoblast membrane is taken up by the trophoblast. This shuttle appears to be involved in TH transfer but whether this is by carriage of TH across the placenta, delivery of TH to the membrane transporters or both is as yet unclear. The low placental oxygen level of early pregnancy up regulates both TTR expression and reuptake. Many agents interfere with T 4 binding to TTR and impair fetal thyroid function but a role in interfering with placental transfer of TH is yet to be studied. Placental D3, which converts T 4 to the biologically inactive rT 3 , is an important modulator of TH transfer. The interaction of TTR, if any, with the deiodinase requires investigation.
Lastly, TTR is up regulated in placentas of women with pre-eclampsia. While this may be the effect of placental hypoxia, which is prevalent in this condition, its role as a marker of pre-eclampsia deserves further attention.

Acknowledgements
We would like to thank Mimi Kersting for help with preparing figures. www.intechopen.com