Biology Department University of Massachusetts Morrill Science Center Amherst, Massachusetts 01003
Address all correspondence and requests for reprints to: Dr. R. Thomas Zoeller, University of Massachusetts, Morrill Science Center, Biology Department, Amherst, Massachusetts 01003. E-mail: tzoeller{at}bio.umass.edu.
Frog metamorphosis has long been a fascinating example of thyroid hormone actions on development (1), and insights gained from studies of frog metamorphosis are helping us understand the role of thyroid hormone in the development of a completely different tissuethe human brain. In frogs such as Xenopus laevis, thyroid hormone controls the dramatic transformation from the larval to the adult form (2, 3), in which many larval tissues are lost (e.g. gills and tail), adult structures formed (e.g. limbs), and other organs are remodeled to support adult functioning. Importantly, frog metamorphosis is characterized by an orderly sequence of events; thus, different tissues undergo thyroid hormone-dependent metamorphic changes at different times and at different rates, all in the face of elevated circulating levels of thyroid hormone. A seminal observation is that local metabolism of thyroid hormone is a major factor controlling the timing of tissue responsiveness to thyroid hormone during frog metamorphosis and thus the sequence of metamorphic events (4). In this issue of the JCEM, Kester et al. (5) describe results of a new study indicating that local metabolism of thyroid hormone in different regions of the developing human brain likely contributes to the timing of thyroid hormone-driven development.
Like frog metamorphosis, development of the mammalian brain is characterized by an orderly sequence of developmental events (6). Moreover, the relative timing of maturational events within the brain is quite similar among mammalian species (7). Recent work in both humans and experimental animals demonstrates that thyroid hormone exerts effects on the developing brain throughout a broad period of fetal and neonatal development (8) and that the developmental events and brain structures affected by thyroid hormone differ as development proceeds. Therefore, it is possible that the human brain uses a strategy for "timing" thyroid hormone sensitivity of different brain regions that is similar to that used by Xenopus. The work by Kester et al. represents a key observation suggesting that this is indeed the case.
Kester et al. (5) report that in several brain regions, especially the cerebral cortex, levels of T3 increase during fetal development and this is correlated with an increase in the activity of type 2 deiodinase (D2), whereas the activity of the type 3 deiodinase (D3) is low to undetectable. D2 controls the conversion of T4 to the hormonally active T3, but D3 controls the conversion of T4 to the hormonally inactive reverse T3. Because T3 levels in the fetal cerebral cortex increased to an extent that could not be accounted for simply on the basis of the age-dependent increase in T4, it indicates that D2 is causing the age-dependent increase in T3 from 1420 wk (postmenstrual age). Importantly, during this same period, the fetal cerebellum has high levels of D3 and low levels of T3. Finally, at later gestational ages, D3 activity in the cerebellum declines and T3 levels increase.
These data further support the concepts that thyroid hormone plays a role in brain development during the fetal period, that different parts of the brain are differentially sensitive to thyroid hormone at any one time during development, and that the sensitivity to thyroid hormone is controlled, in part, by local control of hormone production. In turn, these observations imply that the consequences of thyroid hormone insufficiency during fetal development will differ from those of thyroid hormone insufficiency during postnatal development. In fact, this implication is supported by empirical evidence. For example, Smit et al. (9) studied a small group of infants of women with hypothyroidism diagnosed before pregnancy who were seemingly adequately treated. Although tests indicated that their children displayed normal neurophysiological and motor development, they had significantly lower mental development indices at 6 and 12 months. Importantly, Rovet et al. (10) followed a relatively large group of infants whose mothers had hypothyroidism diagnosed before or during pregnancy and found mild effects on specific cognitive abilities, particularly visual attention and visuospatial processing abilities. Compared with offspring of euthyroid women, these children showed poorer attention, slower and more variable reaction times to visual stimuli, and visual deficits, particularly reduced contrast sensitivity. Moreover, the specific types of visual deficits appeared to reflect the timing of thyroid hormone insufficiency during pregnancy (11).
The concept that the fetal brain is sensitive to thyroid hormone is of relatively recent origin. Early work indicated that thyroid hormone is not transferred from the mother to the fetus because the human placenta and fetal membranes contain high levels of D3 that degrade thyroid hormones and might prevent such transfer (12, 13). Thus, it was somewhat paradoxical that, in the 1960s and 1970s, Man et al. (14) published the results of a series of studies that found an association between "butenol-extractable iodine" in pregnant women and measures of cognitive function in the offspring, indicating that thyroid hormone may play a role in fetal brain development. This paradox was reconciled in part by Vulsma et al. (15) who reported that newborns with a genetic incapacity to synthesize thyroid hormones have T4 levels that are nearly the same as normal neonates, indicating that the fetus obtains a considerable proportion if its T4 from maternal circulation and this is likely to be true throughout gestation.
The observations of Vulsma et al. were pivotal because the fetus does not begin to produce its own thyroid hormone until around the end of the first trimester (16); therefore, if thyroid hormone acts on the fetal brain in the first trimester, the only source of hormone would be the mother. In fact, thyroid hormones are detected in human coelomic fluid as early as 8 wk gestation (17, 18), several weeks before the onset of thyroid function at 1012 wk (16), and these levels of thyroid hormones are biologically relevant (19). In addition, all the major thyroid receptor (TR) isoforms are present in human cerebral cortex as early as 8 wk gestation, with immunostaining being reported for TR expression in cerebellar pyramidal cells and Purkinje cells at this time (20, 21). TRs in fetal brain appear to be occupied by thyroid hormone as early as 9 wk gestation (17) and the proportion of TR occupancy by thyroid hormone is in the range known to produce physiological effects.
More recent studies have confirmed that thyroid hormone of maternal origin exerts functional effects on the fetus. Pop et al. (22, 23) showed that levels of free T4, and the presence of circulating antibodies for thyroid peroxidase, were strong predictors of infant mental development and childrens IQ. In addition, these authors found that children of women with hypothyroxinemia at 12 wk gestation had delayed mental and motor function compared with controls; the two groups were different by 810 index points on the mental and motor scales at both 1 and 2 yr of age (24). Finally, Haddow and colleagues (25, 26) showed that the children of women with low circulating levels of thyroid hormone exhibit a number of measurable neurological deficits depending on the severity of the hormonal insufficiency. Thus, the literature supports the conclusion that thyroid hormone insufficiency in pregnancy can lead to cognitive deficits in the offspring, clearly indicating that thyroid hormone plays an important role in fetal brain development.
Recent authors discuss the relative merits of developing a routine screening program for thyroid function in pregnant women (27, 28, 29, 30, 31). The relative lack of information about the potential adverse consequences, to the mother or to the fetus, of T4 replacement therapy in pregnant women is one of the critical arguments that screening programs should not presently be implemented, although T4 replacement is recommended for pregnant women who are clinically hypothyroid (29) and an increase in the dose of T4 replacement is recommended for pregnant women currently on T4 replacement (32). Recent animal studies indicate that thyroid hormone insufficiency in the mother can influence cortical neuronal migration in the absence of effects on maternal TSH. Specifically, Ausò et al. (33) found that 3 d of methimazole treatment to pregnant rats could alter neuronal migration in the cerebral cortex without affecting maternal TSH. In combination with the observations of Kester et al. (5), it is important to recognize that, although the deiodinases may account for tissue differences in thyroid hormone sensitivity, they may not always compensate for changes in circulating levels of thyroid hormone.
Footnotes
Abbreviations: D2, Type 2 deiodinase; D3, type 3 deiodinase; TR, thyroid receptor.
Received May 19, 2004.
Accepted May 20, 2004.
References