Inhibition, Na+/I- symporter (NIS) leads to Impairment, Learning and memory

During pre- and perinatal development, disruption of TH signaling leads to a multitude of neurological deficits. Multiple studies have shown that TH deprivation leads to defects in learning processes (for a comprehensive review, see Raymaekers and Darras, 2017). Congenital hypothyroidism has been shown to cause selective visuocognitive malfunctions, a lower IQ even in young adults (Oerbeck et al., 2003; Simic et al., 2013; Wheeler et al., 2012; Willoughby et al., 2014). On the other hand, adult-onset hyperthyroidism has been associated with a decrease in signal activity between the hippocampus and other cortical regions (Zhang et al., 2014), hyperactivity, attention deficits and changes in anxiety state (Raymaekers and Darras, 2017), which could impact learning potential.

Some epidemiological and in vivo studies have indicated associations between NIS inhibition (e.g., as a consequence of exposure to perchlorate or other pollutants, such as BPA and BDE-47, NIS inhibitors) (Wu Y et al., 2016) and decreased cognition.

BPA exposure has been also associated with hypothyroidism (i.e., decreased of free T3 and free T4, increase of TSH plasma levels, perturbation of thyroid gland morphological structure and thyroid cell function) in humans (i.e., inverse relationships between urinary BPA and total T4 and TSH) (Meeker and Ferguson, 2011), in young rats breast-fed from mothers treated with BPA (Mahmoudi et al. 2018), and in pregnant ewes and their newborn lambs (i.e., decrease of total T4 in BPA-treated pregnant ewes and in the cord and the jugular blood of their newborns (30% decrease), and of plasma free T4 levels in the jugular blood of the newborns) (Viguié et al. 2013).

Polybrominated diphenyl ether (PBDEs) and their hydroxylated metabolites (OH-PBDEs) can bind to the serum-binding proteins transthyretin and thyroxine-binding globulin, can affect deiodinases (DI 1, 2 and 3)  activity, and alter TH metabolism and excretion, leading to hypothyroidism in experimental animals (Butt et al. 2011; Marchesini et al. 2008; Meerts et al. 2000; Szabo et al. 2009; Zhou et al. 2002). Human studies also observed PBDE-associated TH disruption during pregnancy (Chevrier et al. 2010; Herbstman et al. 2008; Lin et al. 2011; Stapleton et al. 2011; Zota et al. 2011). Therefore, thyroid disruption may be a critical underlying mechanism related to the developmental neurotoxicity of PBDEs and their metabolites (Dingemans et al. 2011; Costa et al. 2008; Chen et al. 2014). 

- Wang et al., 2016: in this in vivo study, pregnant Sprague-Dawley female rats were orally treated with either vehicle or BPA (0.05, 0.5, 5 or 50 mg/kg BW/day) during days 9-20 of gestation. Male offspring were tested on PND 21 with the object recognition task. BPA-exposed male offspring underwent memory and cognitive impairments: they not only spent more time (~ 43% more, at 1.5 hr after training) in exploring the familiar object at the highest dose than the control, but also displayed a significant decrease in the object recognition index (at 50 mg/kg BW/day, ~ 54% lower short term memory measured 1.5 hr after training).

- Jang et al., 2012: In this in vivo study pregnant female C57BL/6 mice (F0) were exposed to BPA (0.1-10 mg/kg) from gestation day 6 to 17, and female offspring (F2) from F1 generation mice were analysed. High-dose BPA (10 mg/kg) caused neurocognitive deficit (i.e., reduced memory retention) as shown by passive avoidance testing (~ 33% decrease vs control) in F2 mice. These results suggest that BPA exposure (NIS inhibition) in pregnant mothers could decrease hippocampal neurogenesis and cognitive function in future generations.

- Taylor et al., 2014: In this historical cohort study of 21,846 women in Cardiff, United Kingdom, and Turin, Italy, who were pregnant from 2002 to 2006, levels of urinary perchlorate (a NIS inhibitor) in the highest 10% were associated with a higher risk for having children with IQ scores in the lowest decile at age three, as described in 487 mother–child pairs in mothers who were hypothyroid/hypothyroxinemic during pregnancy.

- Chen et al., 2014: In this prospective birth cohort, maternal serum concentrations of BDE-47 and other PBDE congeners were measured in 309 women at 16 weeks of gestation, and associated with neurodevelopment in children. Importantly, BDE-47 and other chemicals, such as triclosan, triclocarban and BPA, have been reported to disturb TH homeostasis by inhibiting NIS-mediated iodide uptake and altering the expression of genes involved in TH synthesis in rat thyroid follicular FRTL-5 cells (Wu Y et al., 2016). A 10-fold increase in prenatal BDE-47 exposure was associated with a 4.5-point decrease in Full-Scale IQ and a 3.3-point increase in the hyperactivity score at age 5 years in children.

- Roze et al., 2009: Similarly, this epidemiological study assessed the level of several compounds, including BDE-47 (i.e., 2,2'-bis-(4 chlorophenyl)-1,1'-dichloroethene, pentachlorophenol (PCP), PCB-153, 4OH-CB-107, 4OH-CB-146, 4OH-CB-187, BDE-47, BDE-99, BDE-100, BDE-153, BDE-154, and hexabromocyclododecane), in 62 mothers during the 35^th week of pregnancy, and possible associations with the neuropsychological level in their children at 5-6 years of age. THs were determined in umbilical cord blood. Brominated flame retardants correlated with worse fine manipulative abilities, worse attention, better coordination, better visual perception, and better behavior. Chlorinated OHCs correlated with less choreiform dyskinesia. Hydroxylated polychlorinated biphenyls correlated with worse fine manipulative abilities, better attention, and better visual perception. The wood protective agent (PCP) correlated with worse coordination, less sensory integrity, worse attention, and worse visuomotor integration.

- van Wijk et al., 2008: This in vivo study assessed the behavioural effects of perinatal and chronic hypothyroidism during development in both male and female offspring of hypothyroid rats. To induce hypothyroidism, dams and offspring were fed an iodide-poor diet and drinking water with 0.75% sodium perchlorate (NIS inhibitor). Treatment was started in dams 2 weeks prior to mating, and in pups either until the day of killing (i.e., chronic hypothyroidism) or only until weaning (i.e., perinatal hypothyroidism) to test for reversibility of the effects observed. Early neuromotor competence, as assessed in the grip test and balance beam test, was impaired by both chronic and perinatal hypothyroidism. The open field test, assessing locomotor activity, revealed hyperactive locomotor behavioural patterns in chronic hypothyroid animals only. The Morris water maze test, used to assess cognitive performance, showed that chronic hypothyroidism affected spatial memory in a negative manner. Perinatal hypothyroidism was found to impair spatial memory in female rats only. In general, the effects of chronic hypothyroidism on development were more pronounced than the effects of perinatal hypothyroidism. This suggests that the early effects of hypothyroidism on functional alterations of the developing brain may be partly reversible.

- Kosugi et al. 1998; Ferrandino et al. 2017: Three Japanese children inherited two NIS mutations (V59E and T354P) from their healthy mother and father, respectively. V59E NIS was reported to exhibit as much as 30% of the activity of wild-type NIS (Fujiwara et al. 2000). The T354P and V59E NIS mutant proteins, when expressed in COS7 cells, were both trafficked to the cell surface, but totally inactive. The three siblings displayed different degrees of mental retardation, including heavy learning and memory deficits. The oldest one was nursed for longer than the second oldest, and evinced a less severe cognitive deficit. The youngest was not nursed, and displayed a more severe cognitive deficit than either of her siblings. It was discovered that the mother was addicted to laminaria, an alga extremely rich in I− (Ferrandino et al. 2017). These studies will be also cited in support of Essentiality for KE (MIE).

- Babu et al. 2011: in this in vivo study 50-day-old female rats weighing 120–150 g were switched to a low iodine diet (LID) and given 1% KClO4 (NIS inhibitor) in drinking water for 10 days. Animals were then separated into an iodine sufficient groups (or euthyroid) and a low iodine diet (or hypothyroxinemic) group (0.005% KClO4) and kept on above diet regimen for 3 months. Based on the hormonal estimations and urinary iodine, female rats were further divided into euthyroid and hypothyroxinemic and were mated with normal males. In a separate group of age-matched female rats, hypothyroidism was induced in rats by giving MMI (0.025% wt/vol) in drinking water to the pregnant rats from gestational day 8 and continued thereafter until sacrifice of pups born to these dams (hypothyroid group).

Data showed a significant reduction in total serum T4 and T3 levels of rat pups administered with MMI compared to euthyroid controls (3-fold decrease of T3 vs ctr and 7-fold decrease of T4 at P16). Hypothyroxinemic pups (on low iodine diet and KClO4) showed a reduction in serum T4 (~ 70% decrease of T4 vs ctr) but not in T3, which was increased compared to euthyroid levels at P16 (~ 40% increase of T3 vs ctr). Even in the presence of elevated circulating T3 levels, hypothyroxinemic pups showed significantly impairment of TH responsiveness in developing rat neocortex.

Both hypothyroid (MMI) and hypothyroxinemic (KClO4) pups demonstrated a significant increase in D2 levels compared to controls (~ 11 fold in hypothyroidism, and ~ 4 fold in hypothyroxinemia). The expression of D3 mRNA was also decreased significantly (by ~ 3.3 fold in hypothyroidism and ~ 3 fold in hypothyroxinemic group compared to controls), whilst MCT8 and TH nuclear receptors α1 and β1 expression did not change. Additionally, myelin basic protein (MBP) protein levels and gene were decreased in both groups (for MBP gene: by ~ 60% and ~ 70% respectively in hypothyroidism and hypothyroxinemic groups vs Ctr). Moreover, increased number of apoptotic neurons was found evenly distributed in all the layers of the neocortex under both hypothyroxinemic and hypothyroid conditions. As stated in this study, altogether these data suggest that hypothyroxinemia induced by low iodine diet and KClO4 may lead to learning and memory impairment in this model. However, memory or cognitive tests were not assessed in this study.

- Buras et al. 2014: in this in vivo study 9-10 week old mice were administered with drinking water containing 0.05% MMI and 1% KClO4 for 4 weeks to render them hypothyroid. After 4 weeks, the hypothyroid group was further divided into 3 groups: the hypothyroid (0.05% MMI + 1% KCL04), T3 (0.05% MMI + 1% KCL04 + T3 (0.5 μg/ml) in drinking water) and T4 (0.05% MMI + 1% KCL04 + T4 (5 μg/ml) in drinking water) groups for weeks 5 and 6. T3 serum levels were decreased by ~ 40% in hypothyroid group vs Ctr, and T4 was totally depleted in hypothyroid group vs Ctr. Several tests were performed to evaluate fear-anxiety behaviour. In the elevated plus maze, the hypothyroid mice showed significantly lower distance and time in the open arms than the T3-treated group (~ 50% for both parameters) than the euthyroid controls. The hypothyroid group also showed greater distance and time in the closed arms (~ 10% and 20% more than Ctr respectively for distance and time scores) than the T3-treated group. Administration of T3 and T4 rescued these effects. Moreover, hypothyroid mice froze more than Ctr (~ 35% more) and T3 and T4 treatments reversed this effect.

- Navarro et al. 2015: in this in vivo study 0.02% MMI and 1% KClO4 were added to the drinking water in rats starting at embryonic day 10 (E10, developmental hypothyroidism) and E21 (early postnatal hypothyroidism) until day of sacrifice at PND 50. Behavior was studied using the acoustic prepulse inhibition (somatosensory attention) and the elevated plus-maze (anxiety-like assessment) tests. Total plasmatic T4 levels of both E10 (1.86 ng/ml) and E21 (1.08 ng/ml) pups were significantly lower than those of Ctr (36.29 ng/ml) pups. Total plasmatic T3 levels of E10 (0.10 ng/ml) and E21 (0.10 ng/ml) were significantly lower than in Ctr (0.45 ng/ml) pups. E10 and E21 treated pups showed abnormal laminar organization of the hippocampus, critical brain structure for learning and memory processes. The distribution, density and size of VGluT1 and VGAT boutons in the hippocampus and somatosensory cortex was abnormal in hypothyroid pups (in both groups) and these changes correlated with behavioral changes: prepulse inhibition of the startle response amplitude was reduced (23.3% in E10, 43.0% in E21 and 79.0% in Ctr pups), indicating severe pre-attention deficit in treated pups, while the percentage of time spent in open arms increased (57.0% time spent in open arms in E21 and 81.1% in E10 pups, vs 17.1% Ctr pups, indicative of increased anxiety).

- Vasilopoulou et al. 2016:  this in vivo study investigated the effects of adult onset hypothyroidism (induced by administration of 1% w/v KClO4 in their drinking water for 8 weeks in adult male Balb/cJ mice) on acetylcholinesterase (AChE) activity and on related behavioral parameters. They found that adult onset hypothyroidism (TH levels were not measured in this study) caused decrease of memory and increased fear/anxiety (i.e., 51% decrease of time spent in open arms / [times spent in open + closed arms], 47% decrease of the number of entries into the open arms of the apparatus, and 42% decrease in the total number of arm entries), and activity of both isoforms of AChE was reduced in all examined brain regions.

Single NIS mutations, causing decreased thyroidal iodide uptake, may not necessarily lead to cognitive disorders. In this regard, Nicola and coworkers (Nicola et al., 2015) recently identified a new NIS mutation (V270E) in a patient (full-term girl born to healthy, non-consanguineous Jamaican parents), who resulted to be heterozygous for this NIS mutation (R124H/V270E). The presence of the mutation V270E markedly reduces iodide uptake (5.4% 24 hours after the oral administration of 100 μCi ¹²³I− (normal range, 10–40%)) via a pronounced (but not total) impairment of the protein's plasma membrane targeting. However, the retaining of a minimal iodide uptake was enough to enable sufficient TH biosynthesis and prevent cognitive impairment.

It should be noted that the van Wijk et al. 2008  study was performed with only one dose group exposed to perchlorate during development, and the behavioural assessments were performed using a limited group size of 5-8, possibly reducing the reliability of this study. In general, chronic hypothyroidism effects on development were more pronounced than the effects of perinatal hypothyroidism, suggesting that functional alterations occurring as a consequence of hypothyroidism may be partly reversible depending on developmental stage of the deficiency.

Opposite, other in vivo studies do not support associations between perinatal perchlorate exposure and neurobehavioural effects. For example, York et al. (2004) could not observe meaningful behavioral effects in rat offspring exposed as high as 10.0 mg/kg/day, as evaluated by passive avoidance, swimming water maze, motor activity, and auditory startle. In their re-evaluation of the data (York et al. 2005), authors concluded that rat pups exposed to perchlorate both during pregnancy and after 10 days of lactation, despite showing alterations of neurohistopathological features, did not show altered development of gross motor movements. Moreover, Gilbert and Sui (2008) found that adult male offspring born from rat dams exposed to 0, 30, 300, or 1,000 ppm perchlorate in drinking water from gestational day 6 until weaning, underwent reduction of T3 (10–14% reduction) and T4 (~ 9–20% reduction) reduction on postnatal day 21 (at the highest perchlorate dose), significant reductions in baseline synaptic transmission (~ 20% increase in excitatory postsynaptic potential slope amplitude), but without changes of motor activity, spatial learning, or fear conditioning.

Taylor et al. 2004 (CATS study) identified 1050 pregnant women with hypothyroidism or hypothyroxinemia; half were in the immediate T4 treatment group, and half were in the group tested and treated after pregnancy. 487 (46.4%) mother-child pairs completed psychological testing and urinary iodine and perchlorate measurements. Therefore, the 487 women-child pairs represent approximately two-thirds of those reported in the study of T4 treatment effects on cognitive outcome. Taking this into account, the absence of a direct effect of perchlorate on maternal thyroid function (Pearce et al. 2010), suggests that developmental effects of perchlorate may not necessarily be linked to maternal thyroid hormone levels, as commented in (Brent, 2014).