| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Instituto de Neurociencias (E.A., E.C., P.B.), Universidad Miguel Hernández and Consejo Superior de Investigaciones Científicas (CSIC), San Juan, 03550 Alicante; and Instituto de Investigaciones Biomédicas "Alberto Sols" (R.L.-A., F.E.d.R., G.M.d.E.), CSIC and Universidad Autónoma de Madrid, 28029 Madrid, Spain
Address all correspondence and requests for reprints to: Pere Berbel, Instituto de Neurociencias, Universidad Miguel Hernández-Consejo Superior de Investigaciones Científicas, Apartado de correos 18, San Juan, 03550 Alicante, Spain. E-mail: pere.berbel{at}umh.es.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-aminobutyric acidergic neurons was, however, normal. The infusion of dams with T4 between E13 and E15 avoided these alterations, which were not prevented when the T4 infusion was delayed to E15–E18. In total, 52% of the pups born to the goitrogen-treated dams responded to an acoustic stimulus with wild runs, followed in some by seizures. When extrapolated to man, these results stress the need for prevention of hypothyroxinemia before midpregnancy, however moderate, and whichever the underlying cause. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-aminobutyric acidergic (GABAergic; inhibitory) neurons, respectively. The maternal thyroid is then the only source of T4 and T3 for the brain of the fetus because its thyroid gland does not start contributing to fetal requirements until midgestation in man, and E17.5–E18 in rats. Therefore, the amount of maternal T4 that the fetus receives early in pregnancy will determine thyroid hormone action in its brain because it depends on maternal T4 for its intracellular supply of the active form of the hormone, T3. Maternal hypothyroxinemia during this period of brain development may result in poor fetal neurodevelopment, a possibility that is strongly supported by an increasing number of epidemiological studies and case reports, which have been recently reviewed (7, 8, 9, 10). The term hypothyroxinemia indicates that, whether or not TSH is increased above normal values, as in primary hypothyroidism, T4 or free T4 (FT4) is low compared with values usually found at the same stage of pregnancy in normal women with an adequate iodine intake (7, 8). The most severe CNS brain damage, that of the neurological cretin, is the result of early and profound maternal hypothyroxinemia, caused by severe iodine deficiency (ID). This condition impairs the adequate production and secretion of T4 by the maternal thyroid. As discussed more extensively elsewhere (7, 11), the degree of the neurological damage is related to the severity of the maternal hypothyroxinemia and not to her circulating T3 or TSH levels, which are normal. Even more important, as a public health problem, is that in ID areas the mental development of the noncretin population is also affected. Both neurological cretinism and the mental impairment of the general population can only be prevented before midgestation.
We have recently described what we believe is the first direct experimental evidence of a potentially irreversible damage to the fetal brain caused by early maternal hypothyroxinemia, without hypothyroidism (11). Pups born to severely iodine-deficient dams showed alterations in the cytoarchitecture of the somatosensory cortex and hippocampus, and in the radial neuronal migration into the neocortex, which had been caused by events taking place before onset of active fetal thyroid secretion. Such alterations may well be causally related to the neurodevelopmental handicaps described in neurological cretins and the general population of areas with severe ID. The study, however, did not directly address some of the questions raised by an increasing number of reports, which strongly suggest that milder degrees of maternal hypothyroxinemia, caused by moderate ID, or by mild thyroid failure from other causes, increase the risk of bearing children with an intelligence quotient (IQ) of 85 or less, with a frequency that may be 150–200 times that of children born with congenital hypothyroidism (7).
Some of these questions could not be answered previously because of the experimental design that had been used (11). The dams were severely hypothyroxinemic before and throughout pregnancy and lactation, and therefore we could not define: 1) a window of special sensitivity of the developing brain to the decrease in maternal T4; 2) the degree of maternal hypothyroxinemia required to observe alterations of neuronal migration in the progeny; 3) the timing after which normalization of the maternal hypothyroxinemia might no longer prevent fetal CNS damage; and 4) whether the alteration of the migratory phase of cells labeled with 5-bromo-2'-deoxyuridine (BrdU) at E17–E19, was a consequence of derangement of the previous phase. It also appeared interesting to determine whether, or not, the tangential neuronal migration was also affected, for which different subtypes of GABAergic neurons were studied.
To answer this type of question, we have here restricted both the degree and the duration of the maternal thyroid dysfunction, by treating normal dams with a goitrogen for only 3 d at the beginning of neocorticogenesis.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Young adult females, weighing approximately 200 g, were mated with normal males on E0. The main experimental series are described in Fig. 1
, with the number of animals studied in each group for the different endpoints1 and consisted of: one group of normal rats and three groups of dams that drank 0.02% 2-mercapto-1-methylimidazole (methimazole, MMI) from E12–E15. One group received MMI treatment alone (3dMMI12), and the other two groups were infused with T4, again for 3 d, from E13–E16 (3dMMI12 + T4–13) and from E15–E18 (3dMMI12 + T4–15); the subscripts indicate the day of onset of treatments (Fig. 1). Onset of T4 infusion was delayed for 1 d after onset of MMI, considering that its effects on circulating hormone levels were likely to be more rapid than those of the oral administration of MMI. A pilot experiment (12) was carried out infusing for 3 d a dose of 2.4 µg T4/100 g body weight/d that is adequate to maintain euthyroidism of thyroidectomized adult rats (13). This dose increased circulating T4 and T3 about 4-fold. Thus, for the present study we infused 0.6 µg T4/100 g body weight·d. The solution was prepared as previously described (14) and infused using osmotic minipumps (ALZET, model 2001, Alza Corp., Palo Alto, CA), placed under the dorsal skin (13).
|
; see footnote1).
Blood (
0.5 ml) was obtained under slight ether anesthesia from the jugular vein from dams of the four main experimental groups on the mornings of E15 and E19. The plasma was spun off and kept at –20 C, for the determination of T4, T3, and TSH concentrations.
In all cases, both onset and withdrawal of treatments were carried out between 0800 and 1000 h.
At 2–3 d of postnatal age (P2–P3) each litter was culled to eight pups. At P39, pups from the four groups were tested for their audiogenic seizure susceptibility, and at P40 all pups were weighed, anesthetized, bled, and perfused with 50 ml saline followed by 200 ml 4% paraformaldehyde and 0.1 M sucrose in 10 mM phosphate buffer (PB, pH 7.3–7.4), containing 0.002% CaCl2. The brains were postfixed by immersion in PB at room temperature for 4 h, and stored at 4 C in PB containing 0.1 M sucrose.
Immunohistochemistry
The brain was coronally sectioned using a Vibratome PELCO 101 (Ted Pella Inc., Redding, CA) in eight parallel series of 100 µm-thick sections. Sections containing the primary somatosensory cortex (taken at –2.0 to –4.0 mm from bregma) and the mediocaudal portion of the hippocampal formation and the primary auditory cortex (both, at –3.8 to –4.3 mm from bregma) were collected in PB saline. One series was immunostained with anti-BrdU monoclonal antibody (mAb; 1:40), biotinylated horse antimouse antibody (Ab; 1:150), Vectastain ABC kit (Vector, Burlingame, CA; 1:200), and 0.05% 3,3'diaminobenzidine (DAB) (15). These dilutions were the same whenever these four reagents were used. An adjacent series was stained with cresyl violet. Six series were double immunostained. Two of them were double immunostained for fluorescence, starting with sheep anti-BrdU Ab (10 µg/ml), biotinylated rabbit antisheep Ab (5 µg/ml), and NeutrAvidin (Molecular Probes Europe, Leiden, The Netherlands), Rhodamine Red (Molecular Probes Europe) conjugate (1 µg/ml). One of these two series was then immunostained with antineuronal nuclei (NeuN) mAb (1:400), a marker for neurons; the other with anti-2',3'-cyclic nucleotide 3' phospodiesterase (CNP) mAb (5 µg/ml), a marker for oligodendrocytes. They were then incubated with biotinylated horse antimouse Ab and Avidin (Molecular Probes Europe), BODIPY FL (Molecular Probes Europe) conjugate (1 µg/ml). The remaining four series were immunostained for conventional light microscopy with markers of different subtypes of GABAergic neurons (16, 17, 18, 19). In two series; the sections were first processed with rabbit anticalretinin Ab (1:2000) or rabbit antineuronal nitric oxide synthase (nNOS) Ab (1:100), which was followed with biotinylated goat antirabbit Ab (1:150), ABC kit, and DAB. These sections were then double-immunostained with anti-BrdU mAb, followed with biotinylated horse antimouse Ab, ABC kit and DAB, intensified with 0.1% ammonium nickel sulfate (20). The remaining two series were first processed with sheep anti-BrdU Ab (10 µg/ml), followed with biotinylated rabbit antisheep Ab (5 µg/ml), ABC kit and DAB, intensified with 0.1% ammonium nickel sulfate (20). One series was immunostained with antiparvalbumin mAb (1:1000), the other, with anticalbindin D-28K mAb (1:2000). They were then immunostained with biotinylated horse antimouse Ab, ABC kit and DAB. Immunostained sections were mounted on gelatinized slides, air dried during 24 h, dehydrated in ethanol, cleared in xylol, and coverslipped.
For quantitative analysis, we have studied five to 11 pups of two to three litters of dams from the subgroups of the four main experimental groups,1 and also from 3dMMI10 and 3dMMI11 dams (see Results). In each instance, cells immunoreactive to BrdU, parvalbumin, calbindin D-28K, calretinin, and nNOS were counted in a total of nine to 12 probes in the primary somatosensory cortex (one to two probes per pup); the mean values corresponding to each pup were later used for statistical evaluation of the results. Probes 600-µm wide were placed over the primary somatosensory cortex and spanned from layer I to the subcortical white matter, as previously detailed (11). In all cases, the borders between layers and strata were placed at the same relative depth, measured from adjacent cresyl violet-stained sections. The relative frequencies of immunoreactive (ir) cells in each layer were averaged across probes and animals. We plotted only BrdU-ir cells that showed intensely and homogeneously labeled nuclei, or diffusely labeled nuclei of lesser intensity, but with clumped chromatin. These cells have entered the last S-phase of their cell cycle during uptake of the injected BrdU and correspond to types 1 and 2, respectively, of Takahashi et al. (21).
Drugs and reagents
BrdU was from Roche Molecular Biochemicals España (Barcelona, Spain). T4, MMI, DAB, and cresyl violet were obtained from Sigma-Aldrich (St. Louis, MO). Anti-BrdU mAb was from Amersham Pharmacia Biotech Ltd. (Little Chalfont, Buckinghamshire, UK). Sheep anti-BrdU and biotinylated rabbit antisheep Abs were from Abcam Ltd. (Cambridge, UK). Horse antimouse and goat antirabbit biotinylated Abs, Anti-NeuN, and anti-CNP mAb were from Chemicon International Inc. (Temecula, CA). Antiparvalbumin and anticalbindin D-28K mAbs and rabbit anticalretinin Ab were from Swant (Bellinzona, Switzerland). Rabbit anti-nNOS Ab was obtained from Zymed Laboratories Inc. (San Francisco, CA).
Determination of serum T3, T4, and TSH
Their concentrations were determined by very sensitive and highly specific RIAs for T4 and T3 (22) and TSH (23). TSH values are expressed as weight equivalents of the National Institute of Diabetes and Digestive and Kidney Diseases rat TSH-RP-3 reference preparation.
Audiogenic seizure susceptibility
At P39, pups were exposed to the electric doorbell used by van Middlesworth (24, 25) to study audiogenic seizure susceptibility of rat pups born from iodine-deficient or goitrogen-treated dams. Rats were individually placed in a plastic cage. The electric bell was placed 20 cm from the bottom of the cage, so that the intensity of the sound reaching the rat was 95–100 dB. Once the bell was switched on, stimulation was maintained for 90 sec, unless the animal responded with a tonic-clonic seizure, when rapidly interrupted. The animals were observed for the onset of wild runs, or wild runs followed by seizures, and the onset of each type of reaction was timed. Rats were returned to their initial cages until they were killed for the immunohistochemical studies. The dams used for this part of the study were from the same strain previously used to assess audiogenic seizure susceptibility in pups from iodine-deficient dams (14).
Statistical analysis
Plots and counts of ir cells were obtained using the Neurograph system (Microptic, Barcelona, Spain). The Systat statistical software (Systat Inc., Evanston, IL) was used for two-way ANOVAs of cell density, factors being layers and experimental groups. Significant layers-groups interactions (P < 0.001) were found in the primary somatosensory cortex the primary somatosensory cortex both for the 3dBrdU14 and 3dBrdU17 subgroups. One-way ANOVAs of ir-cell densities were then used to identify layers that were affected by the experimental condition, followed by Tukey’s test to identify significant differences (P 
0.05) between means.
The SPSS statistical package (SPSS Inc., Chicago, IL) was used for detection of differences between concentrations of hormones, number of pups/litter, and body weight, with one-way ANOVA and the protected least significant difference test for multiple comparisons, after validation of the homogeneity of variances by the Bartlett-Box F test; for the audiogenic seizure susceptibility variables using the Mann-Whitney nonparametric test for frequencies with which the pups responded with wild runs, or wild runs followed by seizures. The Kaplan-Meier survival analysis was applied to the timing of onset of wild runs or seizures. The test calculates the percentage of pups that survive the 90-sec acoustic challenge without a response.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), maternal plasma T4 concentration of the 3dMMI12 and 3dMMI12 + T4–15 dams decreased significantly (P < 0.05) to approximately 70% of values of normal dams. On the contrary, the plasma T4 of the 3dMMI12 + T4–13 dams was higher than that of 3dMMI12 and normal dams. Plasma T3 changed similarly, decreasing in both 3dMMI12 and 3dMMI12 + T4–15 dams; T3 in 3dMMI12 + T4–13 dams was higher than that of 3dMMI12 dams and not different from normal. Although mean plasma TSH values were higher in 3dMMI12 and 3dMMI12 + T4–15 dams compared with those of the control group differences among these three groups were not statistically significant. This lack of a rapid and commensurate increase of circulating TSH in the 3dMMI12 and 3dMMI12 + T4–13 dams, despite the decrease in circulating T4 and T3, supports the concept of hypothyroxinemia (7), namely, that low T4 and T3 concentrations are not necessarily accompanied by an increased TSH. TSH was lower in the 3dMMI12 + T4–13 dams than in dams from the other three groups.
|
), plasma T4 values of the 3dMMI12 and 3dMMI12 + T4–13 dams were no longer different from that of normal ones, whereas the mean value was higher in the 3dMMI12 + T4–15 dams, although the difference was not statistically significant. There were no differences among groups in plasma T3. Plasma TSH was the same as that of normal dams for all groups except for the lower value of the 3dMMI12 + T4–15 dams. No differences were found among experimental groups at P0 in the number of pups born to the dams (9.4 ± 1.03), or at P40 in their body weight (131.9 ± 0.90 g), plasma T4 (31.2 ± 1.0 ng/ml) and T3 (0.97 ± 0.02 ng/ml), or TSH (0.48 ± 0.02 ng/ml).
Distribution of cells
The radial distribution of BrdU-ir nuclei in the primary somatosensory cortex was abnormal in many pups of the 3dBrdU14 subgroups of 3dMMI12 and 3dMMI12 + T4–15 dams, with heterotopic BrdU-ir cells at locations different from those corresponding to their birth date (Fig. 2). In these pups, the percentage of BrdU-ir cells decreased in layers IV and V, and increased in layer VI and subcortical white matter (see P values in Fig. 2), whereas no differences were found between 3dMMI12 + T4–13 and normal pups. These alterations in radial migration, revealed by the presence of heterotopic neurons in the subcortical white matter, were milder in two of 12 of the 3dMMI12 and one of nine of the 3dMMI12 + T4–15 pups. They were never observed in the pups from normal dams. Heterotopic BrdU-ir cells were also found in the subcortical white matter of the auditory cortex (not shown). On visual inspection, their number appeared similar to that found in the primary somatosensory cortex.
|
). Compared with pups from the normal dams, the proportion of BrdU-ir cells in pups from the 3dMMI12 dams decreased in layers II–III and IV, and increased in layers V, VI, and in the subcortical white matter. Such alterations were not observed in pups born to 3dMMI12 + T4–13 mothers.
|
and 2 for 3dMMI12 pups.
In contrast with the radial distribution of the BrdU-ir cells, the radial distribution of neurons immunoreactive to parvalbumin, calbindin D-28K, calretinin, and nNOS was similar in all groups to that of pups from normal dams (Fig. 2). Two types of calbindin D-28K-ir neurons (heavily and lightly labeled) were observed according to the intensity of the immunostaining. Heavily labeled calbindin D-28K-ir neurons showed a morphology corresponding to local circuit neurons, whereas most of the lightly stained calbindin D-28K-ir neurons resembled pyramidal cells, mostly located in layers II-III. In the four main experimental groups, the parvalbumin-ir and calbindin D-28K-ir neurons were mostly found in cortical layers II–III to VI. The radial distribution of parvalbumin-ir neurons was very uniform with a moderate high density in layer V (Fig. 2). The proportion of heavily labeled calbindin D-28K-ir neurons was similar in layers II–III, V, and VI with a decrease in layer IV (Fig. 2). Calretinin- and nNOS-ir neurons were also found in layers II–III to VI, with the proportion being lower in layer IV. In contrast to parvalbumin- and heavily labeled calbindin D-28K-ir neurons, a small but significant number of calretinin- and nNOS-ir neurons were observed in layer I and in the subcortical white matter (Fig. 2).
Cytoarchitectonic changes
In addition to an abnormal migration of BrdU-ir cells, we also observed changes in cresyl violet-stained coronal sections obtained from pups of the 3dMMI12 and 3dMMI12 + T4–15 dams, both as regards the cytoarchitecture of the barrel cortex of the primary somatosensory cortex and of the hippocampus, whereas these features were normal in the 3dMMI12 + T4–13 pups (Fig. 4, A and B). Although not shown, changes observed in pups from the 3dMMI10 and 3dMMI11 dams were again comparable to those from the 3dMMI12 dams. The major features in the normal pups were typical barrels in layer IV, and easily distinguishable borders between layers. In contrast, these features are less prominent in the progeny of 3dMMI12 and 3dMMI12 + T4–15 dams (Fig. 4, C and D). These changes were not seen in the 3dMMI12 + T4–13 pups (Fig. 4B). No statistically significant differences were found in the thickness of the neocortex among the four groups (on average, 1740 ± 67 µm). Again, changes in all the pups from the 3dMMI10 and 3dMMI11 dams (not shown) were comparable to those from the 3dMMI12 dams.
|
) was similar to that of normal pups (Fig. 4E). In the hippocampal CA1, the pyramidal cell layer showed clear-cut borders with the adjacent strata. In contrast, in the hippocampal CA1 of the 3dMMI12 and 3dMMI12 + T4–15 progeny, the border of the pyramidal layer with the strata oriens was more blurred (Fig. 4, K–L). The same was observed in pups from the 3dMMI10 and 3dMMI11 dams.
Double-immunolabeling
Double-labeling for BrdU and NeuN showed that all BrdU-ir cells that were located in heterotopic positions were neurons, whether found in the subcortical white matter (Fig. 5, C and F) or in the hippocampal stratum oriens and the hippocampal alveus (Fig. 5, G and H). This result was further confirmed by the absence of double-labeling for BrdU and cyclic nucleotide CNP. Double-labeling for CNP was only occasionally observed (Fig. 5I) for those BrdU-ir cells that were classified as type 3 after Takahasi et al. (21).
|
, J, K, L, and P). However, most of the heterotopic neurons in the subcortical white matter that stained with the above markers were BrdU-negative (Fig. 5, L and O); only in a few cases were cells double-labeled for BrdU and nNOS observed (Fig. 5, K and M). Other features of immunostaining with these markers such as size and shape of the cells were similar to those of normal pups (Fig. 5, N, P, and Q). In 3dMMI12 pups, perisomatic parvalbumin-ir terminals were prominent around the soma of pyramids in layers II–III and Va and similar to found in normal pups (Fig. 5R).
Audiogenic seizure susceptibility
Acoustic stimulation showed that there were significant differences between the pups born to normal and 3dMMI12 dams in the proportion of pups that responded with wild runs alone, or with wild runs followed by a seizure (Fig. 6). The time required by the 3dMMI12 pups to respond (Fig. 6) was also much shorter. The results obtained in the 3dMMI12 + T4–13 and 3dMMI12 + T4–15 pups differed from those of the 3dMMI12 group, either in the proportion of animals responding with wild runs or seizures, in the length of time required for the response, or in both. The frequency of wild runs and seizures was similar in pups from both groups of T4-infused 3dMMI12 dams, and was between those observed for the normal and the 3dMMI12 pups. There was a difference, however, in the time needed for the wild runs to start: the pups from the 3dMMI12 + T4–15 responded as quickly as the 3dMMI12 pups, whereas those from the 3dMMI12 + T4–13 group required a more prolonged stimulation, comparable with that needed by the few pups from normal dams that reacted to the acoustic stimulus (Fig. 6).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The present experimental design also included dams that were infused with T4 during, or after, the 3-d treatment with MMI. The decrease of the maternal T4 and T3 levels was indeed avoided by infusion of T4 during this 3 d of treatment with MMI, showing that the neurodevelopmental changes found in their pups are not caused by MMI per se. They also show that the changes persist if the T4 infusion is delayed.
The relatively mild degree of transient hypothyroxinemia in pregnant rats, induced with the present design at the beginning of fetal corticogenesis, was enough to alter the organization of the neocortex in most of their pups, as revealed by an abnormal migration and cytoarchitecture, and to result in a functional deficit, as documented by an increased frequency of abnormal responses to an acoustic stimulus. These are, to our knowledge, novel findings. Thus the present experimental design has contributed to answering points 1–4 in the introductory section, permitted further insight into the type of neuronal migration that is affected, and posed new challenges for further research, especially as regards possible functional consequences of a relatively mild degree of transient maternal hypothyroxinemia.
Effects on histogenesis and cytoarchitecture
The alterations of the distribution of BrdU-ir cells among different layers of the primary somatosensory cortex and the hippocampal CA1, and the changes in the cytoarchitecture of these cortical areas found in the progeny of the 3dBrdU14 and 3dBrdU17 subgroups of the 3dMMI12 dams are qualitatively quite similar to those observed in pups born to iodine-deficient dams that were severely hypothyroxinemic because of marked, and chronic, ID (11). Another similarity between the pups of the iodine-deficient and the 3dMMI12 dams is that the heterotopic BrdU-ir cells found in the subcortical white matter are neurons. There might, however, be a slight quantitative difference: all pups from iodine-deficient dams were affected, whereas some of the 3dMMI12 pups (<15%) were not. Although there were only two litters per group, all pups from the 3dMMI10 and 3dMMI11 dams were abnormal, and this suggests that E12, the only day on MMI common to 3dMMI10, 3dMMI11 and 3dMMI12 dams, might be especially sensitive to maternal hypothyroxinemia with respect to cytoarchitecture and neuronal migration into the neocortex and hippocampus.
The present evidence that even a mild and transient decrease of maternal thyroid hormone insufficiency during early pregnancy effectively alters early brain development is in conceptual agreement with the findings described by others (27, 28, 29, 30), who used different experimental designs and studied different end-points. They have shown that severe maternal hypothyroidism, induced by surgical thyroidectomy or MMI treatment starting 2 wk before mating, alters the expression of different cerebral genes at gestation d 16 (equivalent to our E15), and that their expression responds within hours to single injections of T4 into the dam. These observations increase the likelihood that the alterations in neuronal migration and cytoarchitecture described here are the consequence of inadequate and/or untimely expression of one or more cerebral genes that are thyroid hormone sensitive during this developmental period (31). The unique features of the present model might be quite useful for the identification of genes specifically involved in the radial migration of neurons into the cortex, which are involved in processes such as neuron-glia and neuron-substrate adhesion or signal molecules that might guide juvenile neurons to their final destination (32).
The present study presents additional information with respect to the study of iodine-deficient pups (11) regarding the possible effects of this early maternal hypothyroxinemia on radial migration of neurons generated in the cortical ventricular zone vs. the tangential and radial migration of cells originated in the medial ganglionic eminence and their final radial positioning (4, 5, 6). These two populations of migrating neurons will give rise to glutamatergic (excitatory) and to GABAergic (inhibitory) neurons, respectively. Present results confirm alteration in the radial migration of the former, and exclude that of the latter, because the radial distribution of cells ir to parvalbumin, calbindin D-28K, calretinin, and nNOS was not affected in 3dMMI12 pups. This, however, does not exclude the possibility that the tangential migration of GABAergic neurons might also be abnormal in 3dMMI12 pups, despite their normal radial distribution.
Another important issue has been clarified with the present experimental design, which could not be adequately answered in iodine-deficient pups (11). The abnormal distribution of cells incorporating BrdU at E17–E19, as well as the cytoarchitectural alterations described here, are the result of the previous transient exposure to MMI. It would appear that, once the initial radial neuronal migration is altered, other phases of neurodevelopment might also be affected.
Functional and behavioral alterations
The abnormalities in cell migration and the subtle changes in cytoarchitectonic organization found in the progeny of 3dMMI-treated dams indicate that the normal process of brain maturation and, consequently, the establishment of normal brain functions, are likely to be impaired. Indeed, progeny of severely hypothyroid mothers, caused by chronic MMI treatment, with altered neuronal migration and deranged barrel formation, present alterations of GABAergic (33), callosal (15, 34) and thalamo-cortical connections (20). In addition, alterations of behavior and biogenic amine metabolism have been also observed in euthyroid pups of hypothyroid pregnant rats (35). The increased audiogenic seizure susceptibility reported here for the 3dMMI12 pups supports the possibility that inhibitory circuits are altered. In the 3dMMI12 pups, about 20% of glutamatergic neurons are in abnormal locations, as shown by their concomitant BrdU labeling. Therefore, although the radial distribution of GABAergic neurons is similar between normal and 3dMMI12 pups, many of these heterotopic neurons, mostly located in layers Vb and VI are most probably abnormally innervated by GABAergic axons because in these layers we have not observed perisomatic parvalbumin-ir basquet terminals, similar to those found in layers II–III and Va.
We tried the present crude behavioral test because we had applied it before to the progeny of iodine-deficient dams (14), following the procedure initially described for such animals by van Middlesworth (24) and later applied to the progeny of goitrogen-treated dams (25). Such studies disclosed a relationship between audiogenic seizure susceptibility, hearing loss and a derangement in the development of the organ of Corti. The latter was known to be a postnatal event in rodents and treatment of the dams with goitrogens was usually started from a few days before birth up to P14–P21 (25). Such observations, therefore, did not suggest that early maternal hypothyroxinemia might also result in similar behavioral defects. But it was also known that expression of the ß1 and ß2 thyroid hormone receptors (TR) transcripts appear very early in the auditory system, at E12.5, and are restricted to that portion of the inner ear that gives rise to the cochlea (36). The later finding of audiogenic seizure susceptibility and hearing loss in ß1 and ß2 TR-deficient mice (37) prompted us to test the possible influence of maternal hypothyroxinemia, at about E12, on the audiogenic responses of their pups. They were indeed found to be abnormal in many of the pups. We wish to remark, however, that this finding might not be reproduced in rat strains that are genetically resistant to audiogenic seizure susceptibility or less susceptible to audiogenic seizure (38) than ours (14).
Epilepsy is a multifactorial disease in which not only central and peripheral circuitry seems to be involved but also the alteration of glutamatergic (excitatory) and GABAergic (inhibitory) cortical neuronal systems (39). In fact, selective alterations in GABAA receptor subtypes in human temporal lobe epilepsy have been recently reported (40). However, these comments do not necessarily imply that the behavioral alterations observed in 3dMMI12 and 3dMMI12 + T4–15 are causally related to the abnormal radial migration of neurons: differences in the effects of T4 infusion on behavior compared with those on corticogenesis suggest different underlying mechanisms.
Possible clinical implications for adequate human development
In humans, neurogenesis in the neocortex begins by E35–E42 and ends by E140, whereas neural radial migration begins by E56 and ends by E168 (1, 2). Thus, neurogenesis and migration takes place in a total of about 125 d, from wk 5–6 to wk 24 after conception. In the rat, the same events occur in about 10 d, between E11 and E21 (3). Therefore, a period of time of 3 d in rats might correspond roughly to a period of 37–38 d in humans. Extending these speculations somewhat further, the present critical period of sensitivity to maternal hypothyroxinemia would correspond roughly from late first trimester of pregnancy up to midgestation. Thus, our results strongly suggest that abnormal neuronal migrations could be found in the progeny of pregnant mothers that have been transiently and moderately hypothyroxinemic/hypothyroid at the beginning of fetal neocorticogenesis, even if only during as short a period of time as 5–6 wk. This hypothesis cannot be assessed for ethical reasons, but a deficit of T4 for the fetus during such a period of time may alter normal migration of cortical neurons and the cytoarchitectonic organization of the neocortex and hippocampus. These alterations would become irreversible if T4 treatment were delayed. The subtle changes in migration and cytoarchitecture described here may be an underlying cause of the decreased mental development and neurological alterations described in children born in areas of mild and moderate ID (9, 41) or in children born from mothers whose hypothyroxinemia early in pregnancy might be due to other causes (7, 42, 43, 44). Vermiglio et al. (45) have recently reported the results of a more than 10 yr follow-up of 16 children, euthyroid since birth, born to European mildly iodine-deficient mothers, compared with 11 age- and sex-paired children born to iodine-sufficient women from a nearby control village. Thyroid function of the women had been assessed clinically and biochemically at three time points during gestation, with the mean FT4 of the iodine-deficient group being significantly lower than that of the controls at 8 wk (13.9 ± 2.0 vs. 15.7 ± 1.7 pmol/liter) and 13 wk (13.0 ± 1.7 vs. 14.7 ± 1.7 pmol/liter), whereas T3 and TSH were not different. There was a difference of 18.1 points in the mean IQ, in favor of the controls, with a correlation between maternal FT4 before midgestation and the IQ of the offspring. Moreover, almost 70% (11/16) of the children born to the moderately iodine-deficient mothers, and none of the children of the control group, had attention deficit hyperactivity disorders. This abnormally high frequency of attention deficit hyperactivity disorders is surprisingly similar to that reported for the syndrome of generalized resistance to thyroid hormone, which is frequently the consequence of mutations in ß1 and ß2 TR isoforms (46, 47). In both conditions, the underlying cause of the disorder may well be an inadequate availability of receptor-bound T3 during early brain development.
In conclusion
Present experimental results are the first to show that a moderate and transient maternal hypothyroxinemia during a period roughly corresponding in man from the end of the second month up to midgestation, causes irreversible alterations of brain development. These findings should contribute significantly to our awareness of the increasing number of reports on the poorer mental development of children born to mothers who have been hypothyroxinemic before midgestation, often only moderately so, especially if this occurs at the beginning of cortical neurogenesis and no palliative T4 therapy is applied to prevent the possible damage. As pointed out by Utiger (48), until 5 yr ago there was a general consensus that thyroid hormone was necessary for adequate brain development already during the third trimester of pregnancy, but its role until the middle trimester was not generally recognized. Attention was drawn, however, to the possible damage caused by maternal thyroid hormone insufficiency during early development, especially because environmental conditions, such as ID, were increasingly detected in women of child-bearing age, even in the United States. Present experimental results fully support that thyroid hormone of maternal origin is necessary in adequate amounts for normal brain development before and during the middle trimester.
A condition of transient maternal hypothyroxinemia is not likely to be detected early in pregnancy because it may not result in clinical or even subclinical hypothyroidism. T4 treatment of the latter condition is being strongly recommended (49) because the increased TSH indicates that thyroidal hormone production is already inadequate for that person, even if circulating concentration are within the very broad normal range. Present results suggest that this recommendation ought to be extended to maternal hypothyroxinemia because production of T4 may already be inadequate, although not sufficient to trigger an increase in circulating TSH, physiologically suppressed during the first trimester of pregnancy (8). This condition appears to be quite frequent, even in Western societies and, as shown here, is potentially damaging to the developing brain. For this reason, our present study strongly supports our previous conclusions (7, 11, 50) regarding the need to implement mass screening of pregnant women early in gestation, at the beginning of neocortical neurogenesis. For the sake of the unborn child, these programs should be based primarily on first trimester parameters related to circulating FT4, and not only on the detection of antithyroid Ab positivity and increased serum TSH, as previously proposed for the sake of the mother (51).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
E.A., R.L.-A., and E.C. contributed equally to this study.
Abbreviations: Ab, Polyclonal antibody; BrdU, 5-bromo-2'-deoxyuridine; CNS, central nervous system; CNP, 2',3'-cyclic nucleotide 3'-phophodiesterase; DAB, 3,3'-diaminobenzidine; E, embryonic day; FT4, free T4; GABAergic, -aminobutyric acidergic; ID, iodine deficiency; IQ, intelligence quotient; ir, immunoreactive; MMI, 2-mercapto-1-methyl-imidazole or methimazole; mAb, monoclonal antibody; nNOS, neuronal nitric oxide synthase; NeuN, neuronal nuclei; PB, phosphate buffer; P, postnatal age; TR, thyroid hormone receptor.
1 The number of animals used per group for the different endpoints in normal (a), 3dMMI12 (b), 3dMMI12 + T4–13 (c) and 3dMMI12 + T4–15 (d) groups was as follows: Number of dams for sampling for hormone determinations of pregnant dams at E15: a (n = 5); b (n = 7); c (n = 4); and d (n = 8). Number of dams for hormone determination at E19: a (n = 9); b, (n = 8); c (n = 4); and d (n = 8). Number of pups for hormone determinations at P40: a (n = 13); b (n = 15); c (n = 11); and d (n = 15). Number of pups per litter used for audiogenic seizure susceptibility studies: a, 46 pups of nine litters (46/9); b, 38/5; c, 32/4; and d, 55/10. The number of dams used for morphological studies was: a (n = 6); b (n = 12); c (n = 12); and d (n = 4). Number of pups per litter used for quantitative immunohistochemistry, except for 3dBrdU17 subgroups, were: a, eight pups of two litters (8/2); b, 11/3; c, 11/3; d, 10/3. Number of pups per litter used for quantitative analysis of 3dBrdU17 subgroups, were: a, seven pups of two litters (7/2); b, 9/3; c, 9/3. 
2 This lack of an increase in circulating TSH is in conceptual agreement with previous observations (26 ) showing that in pregnant rats an increase in circulating TSH is both delayed with respect to a decrease in T4 and T3 and less marked than expected.
Received March 3, 2004.
Accepted for publication April 8, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß-thyroid hormone receptor (TR) gene expression during auditory neurogenesis: evidence for TR isoform-specific transcriptional regulation in vivo. Proc Natl Acad Sci 91:439–443This article has been cited by other articles:
 |
|
 |
|
  M. C. Opazo, A. Gianini, F. Pancetti, G. Azkcona, L. Alarcon, R. Lizana, V. Noches, P. A. Gonzalez, M. Porto, S. Mora, et al. Maternal Hypothyroxinemia Impairs Spatial Learning and Synaptic Nature and Function in the Offspring Endocrinology, October 1, 2008; 149(10): 5097 - 5106. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Sinha, A. Pathak, V. Mohan, S. Bandyopadhyay, L. Rastogi, and M. M. Godbole Maternal Thyroid Hormone: A Strong Repressor of Neuronal Nitric Oxide Synthase in Rat Embryonic Neocortex Endocrinology, September 1, 2008; 149(9): 4396 - 4401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Moleti, V. P. Lo Presti, M. C. Campolo, F. Mattina, M. Galletti, M. Mandolfino, M. A. Violi, G. Giorgianni, D. De Domenico, F. Trimarchi, et al. Iodine Prophylaxis Using Iodized Salt and Risk of Maternal Thyroid Failure in Conditions of Mild Iodine Deficiency J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2616 - 2621. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Taylor, J. Swant, J. J. Wagner, J. W. Fisher, and D. C. Ferguson Lower Thyroid Compensatory Reserve of Rat Pups after Maternal Hypothyroidism: Correlation of Thyroid, Hepatic, and Cerebrocortical Biomarkers with Hippocampal Neurophysiology Endocrinology, July 1, 2008; 149(7): 3521 - 3530. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Chevrier, B. Eskenazi, N. Holland, A. Bradman, and D. B. Barr Effects of Exposure to Polychlorinated Biphenyls and Organochlorine Pesticides on Thyroid Function during Pregnancy Am. J. Epidemiol., June 25, 2008; (2008) kwn136v2. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Dosiou, G. D Sanders, S. S Araki, and L. M Crapo Screening pregnant women for autoimmune thyroid disease: a cost-effectiveness analysis. Eur. J. Endocrinol., June 1, 2008; 158(6): 841 - 851. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. E. Wasserman, K. Nelson, N. R. Rose, W. Eaton, J. P. Pillion, E. Seaberg, M. V. Talor, L. Burek, A. Duggan, and R. H. Yolken Maternal Thyroid Autoantibodies during the Third Trimester and Hearing Deficits in Children: An Epidemiologic Assessment Am. J. Epidemiol., March 15, 2008; 167(6): 701 - 710. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Saunier How Low Can Maternal Thyroxin Go? Endocrinology, June 1, 2007; 148(6): 2591 - 2592. [Full Text] [PDF]
|
|
|
|
|
|
J. H. Goodman and M. E. Gilbert Modest Thyroid Hormone Insufficiency during Development Induces a Cellular Malformation in the Corpus Callosum: A Model of Cortical Dysplasia Endocrinology, June 1, 2007; 148(6): 2593 - 2597. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Gilbert, L. Sui, M. J. Walker, W. Anderson, S. Thomas, S. N. Smoller, J. P. Schon, S. Phani, and J. H. Goodman Thyroid Hormone Insufficiency during Brain Development Reduces Parvalbumin Immunoreactivity and Inhibitory Function in the Hippocampus Endocrinology, January 1, 2007; 148(1): 92 - 102. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Roberts, M. Srinivas, D. Forrest, G. Morreale de Escobar, and T. A. Reh Making the gradient: Thyroid hormone regulates cone opsin expression in the developing mouse retina PNAS, April 18, 2006; 103(16): 6218 - 6223. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Kooistra, S. Crawford, A. L. van Baar, E. P. Brouwers, and V. J. Pop Neonatal Effects of Maternal Hypothyroxinemia During Early Pregnancy Pediatrics, January 1, 2006; 117(1): 161 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Dufault, G. Poles, and L. L. Driscoll Brief Postnatal PBDE Exposure Alters Learning and the Cholinergic Modulation of Attention in Rats Toxicol. Sci., November 1, 2005; 88(1): 172 - 180. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Sui, W. L. Anderson, and M. E. Gilbert Impairment in Short-Term but Enhanced Long-Term Synaptic Potentiation and ERK Activation in Adult Hippocampal Area CA1 Following Developmental Thyroid Hormone Insufficiency* Toxicol. Sci., May 1, 2005; 85(1): 647 - 656. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Forrest The Developing Brain and Maternal Thyroid Hormone: Finding the Links Endocrinology, September 1, 2004; 145(9): 4034 - 4036. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
