Regional Expression of the Type 3 Iodothyronine Deiodinase ...

0013-7227/99/$03.00/0 Endocrinology Copyright © 1999 by The Endocrine Society

Vol. 140, No. 2 Printed in U.S.A.

Regional Expression of the Type 3 Iodothyronine Deiodinase Messenger Ribonucleic Acid in the Rat Central Nervous System and Its Regulation by Thyroid Hormone* HELEN M. TU, GABOR LEGRADI, TIBOR BARTHA, DOMENICO SALVATORE, RONALD M. LECHAN, AND P. REED LARSEN Thyroid Division, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School (H.M.T., T.B., D.S., P.R.L.), Boston, Massachusetts 02115; Tupper Research Institute and Department of Medicine, Division of Endocrinology, Diabetes, Metabolism, and Molecular Medicine (G.L., R.M.L.), and Department of Neuroscience, Tufts University School of Medicine (R.M.L.), Boston, Massachusetts 02111 ABSTRACT Type 3 iodothyronine deiodinase (D3) is a selenoenzyme that inactivates thyroid hormone. It is necessary for T3 homeostasis in the central nervous system. D3 activity has been identified in many regions of the brain and parallels thyroid status, but the level at which it is regulated and its specific cellular locations are not known. We evaluated the effect of thyroid status on the expression of the D3 gene within the central nervous system using in situ hybridization histochemistry. D3 messenger RNA (mRNA) was identified throughout, but with high focal expression in the hippocampal pyramidal neurons,

granule cells of the dentate nucleus, and layers II–VI of the cerebral cortex. In every region, D3 mRNA abundance was correlated with thyroid status. Four different D3 transcripts were identified by Northern analyses, with evidence for region-specific processing, and D3 mRNA increased 4- to 50-fold from the euthyroid to the hyperthyroid state. D3 mRNA was not detectable in hypothyroid brain. In the central nervous system, the D3 gene is highly T3 responsive, and its focal localization within the hippocampus and cerebral cortex suggests an important role for T3 homeostasis in memory and cognitive functions. (Endocrinology 140: 784 –790, 1999)

T

HE INACTIVATION of thyroid hormones can be catalyzed by inner ring deiodination by either the type 3 (D3) or the type 1 (D1) iodothyronine deiodinase. As the maximum velocity/Km ratio of D3 for 5-deiodination of T4 and T3 is about 1000-fold higher than that of D1, it is likely that D3 is the more important enzyme for this reaction (1, 2). In mammals, D3 activity has been identified in placenta, neonatal skin, skeletal muscle, and central nervous system (reviewed in Refs. 3 and 4). D3 activity is heterogeneous in this latter tissue, being much higher in cortex, corpus striatum, midbrain, and hypothalamus than in cerebellum, brain stem, or spinal cord (5). Type 3, but not type 1, activity has been identified in the human central nervous system, suggesting that it is D3 that inactivates T3 in the human brain (6, 7). D3 is a selenoenzyme initially identified as a complementary DNA (cDNA) with significant nucleotide similarity to D1, which was enriched in a subtraction library prepared from T3-treated Xenopus laevis tadpoles (8). Other strategies have resulted in the cloning of D3 cDNAs from Rana catesbeiana, rat, human, and chicken (2, 9 –11). In their

original studies, Kaplan and Yaskowski showed that D3 activity in the central nervous system was positively correlated with thyroid status, in contrast to that of type 2 deiodinase (D2) which changes in the opposite direction (12–14). This led to the concept that T3 homeostasis in the central nervous system is a function of reciprocal compensatory changes in D3 and D2 activities. In the cortex and cerebellum from hypothyroid rats, compensatory decreases in D3 activity appear to be quantitatively as important as the increases in D2 activity in maintaining T3 concentrations constant in these tissues (15, 16). In the rat, D3 is expressed primarily in astroglial cells, and its activity in culture systems increases in response to thyroid hormone, retinoids, cAMP, phorbol esters, and fibroblast growth factor (8, 17–19). Despite the critical role played by D3 in T3 homeostasis in the central nervous system, it is not known where in this tissue the enzyme is expressed, nor is it clear whether the response of D3 to T3 in the central nervous system is local or general. Recent studies have shown, for example, that the expression of D2 is highly localized to the tanycytes lining the lower portions of the third ventricle (20, 21). The present studies were initiated to evaluate the localization of D3 messenger RNA (mRNA) expression in the rat central nervous system using in situ hybridization techniques and to determine the qualitative and quantitative aspects of the response of D3 mRNA in this tissue to alterations in thyroid status using Northern analysis.

Received May 29, 1998. Address all correspondence and requests for reprints to: P. Reed Larsen, M.D., Thyroid Division, Harvard Institute of Medicine, #560, 77 Louis Pasteur Avenue, Boston, Massachusetts 02115. E-mail: larsen@ rascal.med.harvard.edu. * This work was supported by NIH Grants DK-44128 (to P.R.L.), T-32-DK-07529 (to H.M.T.), and DK-37021 (to R.M.L.) and General Clinical Research Center Grant RR-02635.

784

D3 EXPRESSION IN THE CENTRAL NERVOUS SYSTEM Materials and Methods Animals For in situ histochemistry, adult male Sprague-Dawley rats (Taconic Farms, Inc., Germantown, NY; Charles River Laboratories, Inc., Worcester, MA; or Zivic Miller Laboratory, Zelienople, PA), weighing 125–200 mg, were acclimatized to a 12-h light, 12-h dark cycle (lights on between 0600 –1800 h) and a controlled temperature (22 6 1 C). Rat chow and tap water were provided ad libitum. All experimental protocols were reviewed and approved by the animal welfare committee at Brigham and Women’s Hospital and New England Medical Center (Boston, MA). Several groups of animals were made hypothyroid by the addition of 0.02% methimazole to their drinking water for 3 weeks, and another group was thyroidectomized by the supplier with parathyroid reimplantation and killed 10 days after surgery. Other groups were made hyperthyroid by ip injections of T3 with 50 g/100 g BW every other day for 8 days, and some groups of euthyroid animals of the same age and supplier were used as controls. The mean weights at the time of killing were: hypothyroid rats, 198 6 3 g (6se); euthyroid rats, 266 6 3 g; and T3-treated rats, 247 6 10 g. Thyroid status was confirmed for hypothyroid and T3-treated rats by measurements of serum T4 concentrations. For Northern analysis with polyadenylated [poly(A)1] RNA, male Harlan Sprague-Dawley, Inc. rats (Harlan Sprague-Dawley, Inc., Indianapolis, IN), weighing 125–200 g, were used. Animals were maintained at the Center for Animal Resources and Comparative Medicine at Harvard Medical School and were untreated (euthyroid) or T3 treated (hyperthyroid) as described. Animals were killed by pentobarbital or CO2 anesthesia, and various brain regions were removed quickly and frozen in liquid nitrogen. At the time of death, animals weighed 160 –250 g (euthyroid) or 193–226 g (T3 treated).

In situ hybridization histochemistry For in situ hybridization histochemistry, four animals in each of the three groups were weighed, overdosed with pentobarbital, and perfused through the ascending aorta with 0.01 m PBS, pH 7.2, containing 15,000 U/liter heparin sulfate for 30 – 60 sec, followed by 4% paraformaldehyde in PBS for 15 min. Venous blood was collected from the inferior vena cava for measurement of T4 and T3. Brains were removed, postfixed by immersion in the same fixative for 6 h, and placed in 20% sucrose in PBS at 4 C overnight. Eighteen-micron coronal tissue sections through the forebrain were obtained on a cryostat (Reichert-Jung 2800 Frigocut-E, Leica, Deerfield, IL) and adhered to SuperFrost Plus glass slides (Fisher Scientific International, Inc. Co., Pittsburgh, PA). The sections were desiccated overnight by heating at 42 C and were stored at 280 C until prepared for in situ hybridization histochemistry. The hybridization protocol was based on methods previously reported from our laboratories (20). Tissue sections were hybridized with a single stranded [35S]UTP-labeled complementary RNA probe generated from a linearized 0.84-kb rat D3 (rD3) cDNA in pCR II (Invitrogen, San Diego, CA) vector containing a mutation in its coding region resulting in a cysteine codon in place of the TGA selenocysteine codon. The original rD3 cDNA was a gift from Dr. D. St. Germain (2). The plasmid was linearized using the XbaI site in the vector’s polylinker and transcribed with SP6 polymerase. The rD3 antisense probe contained the entire 833-nucleotide coding region and 5 nucleotides of the 39-untranslated region. The same sequences were included in the sense RNA. Hybridization was performed under plastic coverslips in buffer containing 50% formamide, a 2-fold concentration of standard sodium citrate (2 3 SSC), 10% dextran sulfate, 0.25% BSA, 0.25% Ficoll 400, 0.25% polyvinylpyrrolidone 360, 250 mm Tris (pH 7.5), 0.5% sodium pyrophosphate, 0.5% SDS, 250 mg/ml denatured salmon sperm DNA, and 6 3 105 cpm radiolabeled probe for 16 h at 55 C. Slides were dipped into Kodak NTB2 autoradiography emulsion (Eastman Kodak Co., Rochester, NY), and the autoradiograms were developed after 6 days of exposure at 4 C. After being cleared in graded solutions of ethanol and Histosol (National Diagnostics, Atlanta, GA), the sections were coverslipped in Histomount (National Diagnostics), and the autoradiograms were visualized and photographed under darkfield illumination with a Zeiss (Carl Zeiss Inc., Thornwood, NJ) or Olympus (Olympus America Inc., Lake Success, NY) binocular microscope. Some sections were counterstained with cresyl violet to determine whether silver grains were

785

located in neurons as identified by their characteristic nuclear morphology.

RNA preparation and Northern analysis Two series of studies of D3 mRNA expression were performed. The first used total RNA prepared from euthyroid, hypothyroid, and hyperthyroid rats, and the second poly(A)1 RNA from euthyroid and hyperthyroid animals. Animals were anesthetized with pentobarbital or CO2 and decapitated, and core blood was collected to confirm thyroid status by measurement of serum T4 by RIA. The brains were rapidly removed from the calvarium, and discrete regions were dissected under direct vision including the olfactory bulb, cerebellum, cerebral cortex, hippocampus, hypothalamus, midbrain, and pons. The cortex was removed by incising the brain along the rhinencephalic fissure and reflecting it off the hippocampus with a microspatula. The two lobes of the hippocampus were then divided from a dorsal approach with a scapula, and each lobe removed separately with a microforceps. The hypothalamus was removed as a cone from the base of the brain using the mamillary bodies, optic chiasm, and temporal lobes as boundaries. The cerebellum was removed after transecting the cerebellar peduncles. Surface landmarks were used to separate the midbrain from the pons, including the medial mamillary bodies and the superior colliculi. All tissues were snap-frozen in liquid nitrogen and stored at 280 C. RNA was isolated according to the single step RNA isolation method of Chomczynski (22) by combining tissues from two or three identically treated animals. Total RNA was resuspended in 200 –500 ml diethylpyrocarbonate-treated water and quantitated by spectrophotometer at 260and 280-nm wavelengths. Poly(A)1 RNA was isolated from 300 –900 mg total RNA from four to six animals for the various tissues with the PolyAT Tract mRNA Isolation System (Promega Corp., Madison, WI). Three to 10 mg poly(A)1 RNA were electrophoresed through 1% agarose-1.85% formaldehyde-0.5 mg/ml ethidium bromide-1 3 MOPS (3(N-morpholino)-propanesulfonic acid) at room temperature according to the procedure of Lehrach et al. (23). RNA was transferred by capillary action with 20 3 SSC according to the manufacturer’s recommendation for the nylon membranes (GeneScreen Plus, New England Nuclear Life Sciences, Boston, MA), except that the gel was washed twice in 10 3 SSC for 30 min each time at room temperature. The completion of transfer was verified by the presence of the 0.24- to 9.5-kb RNA mol wt marker bands (Life Technologies, Gaithersburg, MD) as well as the 28S and 18S ribosomal RNAs. RNA was fixed to the nylon membrane by UV crosslinking with the Stratalinker (Stratagene, La Jolla, CA). Northern blots were hybridized with a rD3 cDNA under the stringency conditions of 5 3 SSPE (180 mm sodium chloride, 10 mm sodium phosphate, 1 mm EDTA, pH 7.7), 5 3 Denhardt’s solution, 0.5% (wt/vol) SDS, and 100 mg/ml denatured sonicated calf thymus DNA at 65 C. As a control for differences in loading and transfer, blots were stripped and reprobed with rat cyclophilin (rCYC; gift from W. W. Chin). For hybridization to the rD3 and rCYC probes, 1 3 106 and 0.5 3 106 cpm/ml, respectively, were applied to the blots and incubated for at least 16 h. Membranes were washed sequentially twice with 2 3 SSPE-0.1% SDS for 10 min each time at room temperature; once with 1 3 SSPE-0.1% SDS for 10 min at 42 C; once with 0.5% SSPE-0.1% SDS for 30 min at 42 C and for 30 min at 50 C; and once with 0.2 3 SSPE-0.1% SDS for 30 min at 50, 55, and 60 C. For the probing with rCYC, the final wash was in 0.2 3 SSPE-0.1% SDS for 1 h at 65 C. Between probings, the blots were stripped according to the manufacturer’s recommendations in 10 mm Tris-HCl (pH 7.5), 1 mm EDTA (pH 8.0), and 1% SDS and then autoradiographed to show that no residual signal was present. Blots were autoradiographed for at least 7 min at room temperature (rCYC) for 20 days at 280 C (rD3) with an intensifying screen. cDNA probes were made by random priming of template DNA with [a-32P]deoxy-CTP. An 843-bp BamHI/HindIII fragment of rD3 containing a selenocysteine to cysteine mutation subcloned into a pCRII vector (Invitrogen) was used. This fragment contains no 59-flanking region, 833 bp of coding region, and 5 bp of 39-flanking region including the stop codon. A 0.75-kb KpnI/SacI fragment of rCYC cDNA in pBSKS vector (Stratagene, La Jolla, CA) was used. This fragment contains the entire coding region of the rCYC (24). Quantitation of hybridization signals was performed with a scanning densitometer and Molecular Dynamics, Inc. software (version 3.1,

786

D3 EXPRESSION IN THE CENTRAL NERVOUS SYSTEM

Endo • 1999 Vol 140 • No 2

Sunnyvale, CA). The D3 mRNA expression was normalized to the rCYC hybridization signal.

Results In situ hybridization histochemistry

By in situ hybridization histochemistry, D3 mRNA was widely distributed throughout the forebrain (Fig. 1), but was particularly striking in layers II–VI of the cerebral cortex, hippocampal pyramidal cells, granule cells of the dentate, and layer II of the pyriform cortex (Figs. 1–3). Localization of silver grains overlying neurons in the CA3 region of the hippocampus was confirmed in sections counterstained with cresyl violet (Fig. 4). Thyroidectomy resulted in a general

FIG. 1. Low (320) power photomicrograph montage (darkfield illumination) of D3 mRNA in the forebrain of euthyroid (A) and hyperthyroid (B) rats. Note the generalized distribution of silver grains with particularly high concentrations in the cerebral cortex (CC) and hippocampus (H). Hyperthyroidism results in an increase in D3 mRNA in all regions.

FIG. 2. High (3100) power photomicrograph (darkfield illumination) of D3 mRNA in the cerebral cortex of euthyroid (A), hypothyroid (B), and hyperthyroid (C) rats. Silver grains accumulate over the majority of cells in layers II–VI. Note the increase in density during hyperthyroidism and the decrease in density during hypothyroidism.


787

FIG. 4. Brightfield photomicrograph of the CA3 region of the hippocampus counterstained with cresyl violet. Note the accumulation of silver grains over the neurons. Magnification, 3360.

hypothalamus, moderate hybridization was observed throughout, with the exception of the supraoptic nucleus, where in some hyperthyroid animals a higher signal was observed. Northern analysis

FIG. 3. High (3100) power photomicrograph (darkfield illumination) of D3 mRNA in the hippocampus of euthyroid (A), hypothyroid (B), and hyperthyroid (C) rats. Hybridization is present primarily in pyramidal cells and in granule cells of the dentate. As in other regions of the brain, hyperthyroidism increases and hypothyroidism decreases silver grain accumulation in all hybridized regions.

reduction in the accumulation of silver grains over all hybridized structures in the forebrain (Figs. 2 and 3). In contrast, T3 administration resulted in a marked increase in D3 mRNA, particularly in the cerebral cortex and hippocampus (Figs. 1–3). In the cerebellum, silver grains accumulated primarily over the granule cell layer. Similar to other regions of the forebrain, hypothyroidism tended to reduce D3 hybridization, whereas excess T3 increased hybridization. In the

In the first series of experiments, D3 mRNA expression was analyzed using 20 –30 mg total RNA from euthyroid, hypothyroid, and hyperthyroid animals. By Northern analysis of total RNA, D3 mRNA expression was detected in the cortex, cerebellum, hippocampus, hypothalamus, olfactory bulb, amygdala, and midbrain but not in pons, and only in tissues from hyperthyroid rats (data not shown). No discrete bands were seen in the hypothalamic RNA of euthyroid or hyperthyroid rats. Poly(A)1 selection of RNA was used to enhance the sensitivity of the Northern analysis to allow comparison of changes in D3 mRNA expression in response to T3 administration. Four different sized transcripts (3.3, 3.0, 2.6, and 2.0 kb) were found in the cortex, hippocampus, cerebellum, and olfactory bulb of hyperthyroid animals (Fig. 5). There was insufficient mRNA from the small pieces of hypothalamic tissue to obtain significant quantities of mRNA. The 3.0-/3.3-kb doublet was easily seen on shorter exposures. The same transcripts were present in much reduced amounts in the cortex, hippocampus, and olfactory bulb of euthyroid animals. In the cerebellum from normal rats, levels of the 2.6-kb transcript and the doublet were barely detectable, and the 2.0-kb band was undetectable. To compare differences in expression levels for the different transcripts, the densities of the different D3 mRNAs were normalized to the CYC transcript, and their relative abundances were compared between tissues and conditions (Tables 1 and 2). In the euthyroid animals, the 2.6-kb transcript was the major transcript, but was barely detectable in the cerebellum and olfactory bulb. The quantities of the different sized transcripts in hyperthyroid rats varied between tissues, with the 2.0-kb mRNA predominant in the cerebellum and olfactory bulb and the 2.6-kb band predominant in the cortex and hippocampus of the T3-treated animals (Table 1). Hyperthyroidism caused a 4.4- to 53-fold increase in total

788

Endo • 1999 Vol 140 • No 2


TABLE 1. Quantification of the different type 3 iodothyronine deiodinase transcripts in different regions of the central nervous system of euthyroid (EU) and hyperthyroid (T3) rats normalized to cyclophilin mRNA expression Transcript size (kb)

3.3 1 3.0 2.6 2.0 3.3 1 3.0 1 2.6 1 2.0

Cortex

Cerebellum

Hippocampus

Olfactory bulb

EU

T3

EU

T3

EU

T3

EU

T3

0.18 0.44 0.11 0.73

1.9 2.6 1.8 6.3

0.07 0.03 0.00 0.1

1.0 0.58 3.7 5.3

0.80 2.2 0.19 3.2

4.5 7.7 1.9 14

0.56 1.1 1.2 2.9

2.4 3.2 11 17

Each entry is the density of the D3-specific transcript normalized to the density of the cyclophilin mRNA-specific band. TABLE 2. Ratio of the expression of D3 mRNA between hyperthyroid (T3) and euthyroid (EU) rats in different areas of the central nervous system Transcript size (kb)

3.3 1 3.0 2.6 2.0 3.3 1 3.0 1 2.6 1 2.0

Cortex T3/EU

Cerebellum T3/EU

Hippocampus T3/EU

Olfactory bulb T3/EU

10 5.9 16 8.6

14 19 ND 53

5.6 3.8 9.9 4.4

4.3 2.9 9.3 5.9

Each entry is the increase in the normalized density of T3-treated vs. euthyroid animals for the transcripts indicated. ND, Not determined due to the absence of a signal for the euthyroid tissues.

D3 mRNA expression (Table 2), with the cerebellum showing the greatest increase in response. The increases in the quantities of the different sized transcripts varied markedly from region to region, however. For instance, T3 treatment resulted in a 2.9-fold increase in the 2.6-kb mRNA in the olfactory bulb, but a 21-fold increase in the cerebellum. Because of differences in loading (see CYC expression), the high expression of D3 mRNA transcripts in the olfactory bulb was not obvious on the Northern blot. Discussion

The present studies show for the first time the wide distribution of D3 mRNA throughout the central nervous system. In general, this is in agreement with the results of earlier D3 activity analyses (5, 12). However, these studies did not quantitate the levels of D3 expression in the hippocampus, which shows a high focal expression of D3 mRNA in regions CA1-CA3. Throughout the central nervous system, D3 mRNA expression paralleled thyroid status, being markedly increased in T3-treated rats and decreased relative to euthyroid levels in hypothyroid animals. These results also agree with those of D3 activity assays (12–14). The expression of D3 mRNA was quite low in the hypothyroid rat brain, was undetectable in Northern blots using 20 –30 mg total RNA, and was barely visualized with 3–10 mg poly(A)1 RNA from euthyroid rats. The areas in which D3 mRNA is highly expressed, including the pyramidal cells of the hippocampus and layers II–VI of the cerebral cortex, presumably are those in which T3 homeostasis can be inferred to be most important. The presence of D3 mRNA in neurons in these areas was confirmed by cresyl violet counterstains. Although it is well recognized from studies of primary cultures that astrocytes and oligodendroglia can express D3 activity, this is the first documentation of D3 mRNA expression in neurons (19, 25–27). In this regard, it is of interest that these regions also contain the highest concentration of thyroid hormone receptors in the central nervous system (28, 29) and have critical roles in learning, memory, and higher cognitive functions (30). Al-

though thyroid insufficiency clearly impairs the proliferation, arborization, and synapse formation of hippocampal and cerebral cortical neurons and can result in cretinism and a variety of neurological manifestations (see Refs. 28 and 31 for discussion), excess thyroid hormone also can have adverse effects on both the developing and the mature brain. For example, thyrotoxicosis can increase neuronal excitability and induce epileptic seizures (32, 33), whereas the administration of excess T4 to neonatal rats disrupts learning and memory in adults (34, 35). Along these lines, hyperthyroidism can reduce the density of apical dendritic spines in region CA1 of the hippocampus (36), stimulate aberrant mossy fiber projections to region CA3 and the dentate (37), attenuate long term potentiation induction (38), and downregulate benzodiazepine and g-aminobutyric acid receptors in the cerebral cortex (39, 40), possibly explaining some of the detrimental effects of excess thyroid hormone on the hippocampus and cerebral cortex. Similarly, the cerebellum is adversely affected by either insufficient or excess thyroid hormone; the administration of high doses of T4 to neonatal rats leads to abnormal development of the cerebellum, manifested by a reduction in the total number of granule cells (41) and a reduction in total synaptic profiles (42). It is of interest, therefore, that the highest concentration of D3 mRNA was identified in the granule cells. A precise homeostatic mechanism to regulate the concentration of thyroid hormone in the brain, therefore, may be essential to prevent potential destructive influences of excess thyroid hormone on neurons and to reduce T3 and T4 degradation in the hypothyroid state. There are few data available with respect to expression of D3 mRNA in mammalian tissues. Northern analyses illustrate that there are at least four differently sized mRNAs in the rat central nervous system (Fig. 5). Croteau et al. found that the 1.6-, 3.3-, and 3.6-kb mRNAs were barely visible in 5–10 mg poly(A)1 RNA from the cerebral cortex of euthyroid animals, with the 3.3- and 3.6-kb transcripts increased in intensity in hyperthyroid rats (2). Likewise, in our studies of human tissues, we were unable to detect any bands specific to D3 in 2 mg poly(A)1 RNA extracted from total cerebral


FIG. 5. Poly(A)1 RNA was isolated from cortex (CO), cerebellum (CE), hippocampus (HI), and olfactory bulbs (OLF) of euthyroid (EU) and T3-treated (T3) rats. Variable quantities of poly(A)1 RNA were electrophoresed through a 1% agarose-1 3 MOPS-5% formaldehyde gel. For the CO, CE, HI, and OLF, these were 10, 9.3, 5, and 3.4 mg, respectively. Northern blots were probed with 32P-labeled cDNAs for the coding regions of D3 (A) or CYC (B) as described in Materials and Methods. Mol wt (MW) sizes are shown. Exposure times were 20 days at 280 C for D3 and 7 min at room temperature for CYC.

cortex despite the fact that D3 activity is present in the human central nervous system (6, 10). Of the tissues examined by in situ hybridization studies, the hippocampus showed the highest ratio of D3/CYC on Northern analysis. The increase in D3 mRNA after a short term T3 treatment was dramatic. At this time it is not possible to determine whether this reflects T3-induced increases in gene transcription, mRNA stabilization, or a combination of these. Qualitatively, the pattern of mRNA transcript expression in the hyperthyroid state is similar in the cortex and hippocampus, on the one hand, and the cerebellum and olfactory bulb, on the other. In the cortex and hippocampus, the 2.6-kb mRNA is predominant, whereas in the cerebellum and olfactory bulb, the major band is 2.0 kb in size. In the cortex and hippocampus, T3 treatment increased the 3.0-/3.3-kb and 2.0-kb mRNA by the largest amount, whereas in the cerebellum and olfactory bulb, the greatest increase was in the 2.0-kb band. As traces of the larger bands were present in euthyroid cortex and hippocampus, their prominence after T3 treatment could simply reflect an increase in the D3 mRNA induced by hyperthyroidism. On the other hand, in the euthyroid state, the 2-kb transcript was barely detectable. In the hyperthyroid animals, transcripts of the same size as those in the hippocampus appear in the cerebellum and olfactory bulb, suggesting again that the different transcripts represent a quantitative, rather than a qualitative, change. We could not identify D3 mRNA in hypothalamic extracts, suggesting that it is expressed at low levels in this tissue. By in situ hybridization histochemistry, we detected only modest quantities of D3 mRNA in the hypothalamus with accumulations in the supraoptic nucleus only in some hyperthyroid animals. Low levels of D3 activity have been previously reported in hypothalamic tissue, consistent with these re-

789

sults. As there is no information available on the structure of the rat D3 gene, it is not known whether the differences in transcript sizes are due to alternative splicing, differences in polyadenylation, and/or the use of different poly(A) addition signals. Although it is difficult to provide precise quantitative information due to the low levels of D3 mRNA in the euthyroid brain, hyperthyroidism causes a 4- to 50-fold increase in D3 expression in these four subfractions of the central nervous system. In addition to the above-mentioned studies of Kaplan et al. (12–14), T3 responsiveness of D3 activity has been demonstrated in several tissues of R. catesbeiana and X. laevis tadpoles (8, 9). However, D3 activity does not increase in the placenta of the hyperthyroid rat, indicating that the same gene is differentially responsive to thyroid hormone in different tissues (43, 44). That such marked increases in D3 mRNA should occur is all the more remarkable given the fact that T3 receptors are nearly saturated in the cerebral cortex of the euthyroid rat due to the substantial contributions to nuclear receptor T3 of locally produced T3 derived from the action of D2 (45– 48). Although it seems physiologically appropriate that D3 activity should increase in the hyperthyroid state and decrease during hypothyroidism to maintain T3 homeostasis in the central nervous system, D3 may also have an important role during iodine deficiency. Iodine deficiency, a significant stress to the thyroid economy of all vertebrates, has also been shown to be associated with a decrease in D3 activity as well as an increase in that of D2 (49). This indicates that in the central nervous system, increases in D2 combined with compensatory decreases in D3 activity, such as those found in the hypothyroid state, are an integral component of the response of the brain to the challenge of iodine deficiency (15, 16). The present in situ results show that D3 mRNA is diffusely expressed throughout the central nervous system, although it is particularly concentrated in the pyramidal cells of the hippocampus and cerebral cortex. These results suggest that the entire central nervous system, and these regions in particular, are highly sensitive to intracellular T3 concentrations, and that potent strategies have evolved to maintain T3 constant when T4 production is impaired. References 1. Visser TJ, Kaptein E, Terpstra OT, Krenning EP 1988 Deiodination of thyroid hormone by human liver. J Clin Endocrinol Metab 67:17–24 2. Croteau W, Whittemore SL, Schneider MJ, St.Germain DL 1995 Cloning and expression of a cDNA for a mammalian Type III iodothyronine deiodinase. J Biol Chem 270:16569 –16575 3. St Germain DL, Galton VA 1997 The deiodinase family of selenoproteins. Thyroid 7:655– 668 4. Huang T, Chopra IJ, Boado R, Solomon DH, Chua Teco GN 1988 Thyroxine inner ring monodeiodinating activity in fetal tissues of the rat. Pediatr Res 23:196 –199 5. Kaplan MM, McCann UD, Yaskoski KA, Larsen PR, Leonard JL 1981 Anatomical distribution of phenolic and tyrosyl ring deiodinases in the nervous system of normal and hypothroid rats. Endocrinology 109:397– 402 6. Campos-Barros A, Hoell T, Musa A, Sampaolo S, Stoltenburg G, Pinna G, Eravci M, Meinhold H, Baumgartner A 1996 Phenolic and tyrosyl ring iodothyronine deiodination and thyroid hormone concentrations in the human central nervous system. J Clin Endocrinol Metab 81:2179 –2185 7. Mori K, Yoshida K, Kayama T, Kaise N, Fukazawa H, Kiso Y, Kikuchi K, Aizawa Y, Abe K 1993 Thyroxine 5-deiodinase in human brain tumors. J Clin Endocrinol Metab 77:1198 –1202 8. St. Germain DL, Schwartzman RA, Croteau W, Kanamori A, Wang Z, Brown DD, Galton VA 1994 A thyroid hormone-regulated gene in Xenopus laevis

790

9. 10.

11.

12. 13. 14. 15. 16. 17. 18. 19. 20.

21. 22.

23. 24. 25. 26. 27. 28. 29.


encodes a type III iodothyronine 5-deiodinase. Proc Natl Acad Sci USA 91:7767–7771 Becker KB, Schneider MJ, Davey JC, Galton VA 1995 The type III 5-deiodinase in Rana catesbeiana tadpoles is encoded by a thyroid hormone-responsive gene. Endocrinology 136:4424 – 4431 Salvatore D, Low SC, Berry MJ, Maia AL, Harney JW, Croteau W, St.Germain DL, Larsen PR 1995 Type 3 iodothyronine deiodinase: cloning, in vitro expression and functional analysis of the placental selenoenzyme. J Clin Invest 95:2421–2430 Van Der Geyten S, Sanders JP, Kaptein E, Darras VM, Ku¨hn ER, Leonard JL, Visser TJ 1997 Expression of chicken hepatic type I and type III iodothyronine deiodinases during embryonic development. Endocrinology 138:5144 – 5152 Kaplan MM, Yaskoski KA 1980 Phenolic and tyrosyl ring deiodination of iodothyronines in rat brain homogenates. J Clin Invest 66:551–552 Leonard JL, Kaplan MM, Visser TJ, Silva JE, Larsen PR 1981 Cerebral cortex responds rapidly to thyroid hormones. Science 214:571–573 Visser TJ, Leonard JL, Kaplan MM, Larsen PR 1982 Kinetic evidence suggesting two mechanisms for iodothyronine 59-deiodination in rat cerebral cortex. Proc Natl Acad Sci USA 79:5080 –5084 Silva JE, Matthews PS 1984 Production rates and turnover of triiodothyronine in rat-developing cerebral cortex and cerebellum: responses to hypothyroidism. J Clin Invest 74:1035–1049 Silva JE, Larsen PR 1986 Regulation of thyroid hormone expression at the prereceptor and receptor levels. In: Hennemann G (ed) Thyroid Hormone Metabolism. Marcel Dekker, New York and Basel, pp 441–500 Courtin F, Liva P, Gavaret JM, Toru-Delbauffe D, Pierre M 1991 Induction of 5-deiodinase activity in astroglial cells by 12-O-tetradecanoylphorbol 13acetate and fibroblast growth factors. J Neurochem 56:1107–1113 Esfandiari A, Gagelin C, Gavaret JM, Pavelka S, Lennon AM, Pierre M, Courtin F 1994 Induction of type III-deiodinase activity in astroglial cells by retinoids. Glia 11:255–261 Esfandiari A, Courtin F, Lennon AM, Gavaret JM, Pierre M 1992 Induction of type III deiodinase activity in astroglial cells by thyroid hormones. Endocrinology 131:1682–1688 Tu HM, Kim SW, Salvatore D, Bartha T, Legradi G, Larsen PR, Lechan RM 1997 Regional distribution of type 2 thyroxine deiodinase messenger ribonucleic acid in rat hypothalamus and pituitary and its regulation by thyroid hormone. Endocrinology 138:3359 –3368 Guadano-Ferraz A, Obregon MJ, St. Germain DL, Bernal J 1997 The type 2 iodothyronine deiodinase is expressed primarily in glial cells in the neonatal rat brain. Proc Natl Acad Sci USA 94:10391–10396 Chomczynski P 1995 Single-step isolation of RNA from cultured cells or tissues. In: Ausubel FM, Brent R, Kingston RE, Seidman JG, Smith JA, Struhl K (eds) Current Protocols in Molecular Biology. Wiley Interscience, Boston, pp 4.2.4 – 4.2.8 Lehrach H, Diamond D, Wozney JM, Boedtker H 1977 RNA molecular weight determinations by gel electrophoresis under denaturing conditions, a critical reexamination. Biochemistry 16:4743– 4751 Danielson PE, Forss-Petter S, Brow MA, Calavetta L, Douglass J, Milner RJ, Sutcliffe JG 1988 p1B15: a cDNA clone of the rat mRNA encoding cyclophilin. DNA 7:261–267 Leonard JL, Larsen PR 1985 Thyroid hormone metabolism in primary cultures of fetal rat brain cells. Brain Res 327:1–13 Courtin F, Chantoux F, Francon J 1986 Thyroid hormone metabolism by glial cells in primary culture. Mol Cell Endocrinol 48:167–178 Cavalieri RR, Gavin LA, Cole R, De Vellis J 1986 Thyroid hormone deiodinases in purified primary glial cell cultures. Brain Res 364:382–385 Puymirat J 1992 Thyroid receptors in the rat brain. Prog Neurobiol 39:281–294 Puymirat J, Miehe M, Marchand R, Sarlieve L, Dussault JH 1991 Immuno-

30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.

45. 46. 47. 48.

49.

Endo • 1999 Vol 140 • No 2

cytochemical localization of thyroid hormone receptors in the adult rat brain. Thyroid 1:173–184 Squire LR 1986 Mechanisms of memory. Science 342:1612–1619 Dussault JH, Ruel J 1987 Thyroid hormones and brain development. Annu Rev Physiol 49:321–334 Jabbari B, Huott AD 1980 Seizures in thyrotoxicosis. Epilepsia 21:91–96 Sandrini M, Marrama D, Vergoni AV, Bertolini A 1992 Repeated administration of triiodothyronine enhances the susceptibility of rats to isoniazid- and picrotoxin-induced seizures. Life Sci 51:765–770 Eayrs JT 1964 Effect of neonatal hyperthyroidism on maturation and learning in the rat. Anim Behav 12:195–199 Lipp HP, Schwegler H, Heimrich B, Driscoll P 1984 Postnatal modification of hippocampal circuitry alters avoidance learning in adult rats. Science 228:80 – 82 Gould E, Allan MD, McEwen BS 1990 Dendritic spine density of adult hippocampal pyramidal cells is sensitive to thyroid hormone. Brain Res 525:327–329 Represa A, Tremblay E, Ben-Ari Y 1987 Aberrant growth of mossy fibers and enhanced kainic acid binding sites induced in rats by early hyperthyroidism. Brain Res 423:325–328 Pavlides C, Westlind-Danielsson AI, Nyborg H, McEwen BS 1991 Neonatal hyperthyroidism disrupts hippocampal LTP and spatial learning. Exp Brain Res 85:559 –564 Go T, Ito M, Okuno T, Mikawa H 1988 Effect of thyroid hormones on benzodiazepine receptors in neuron-enriched primary cultures. J Neurochem 51:1497–1500 Sandrini M, Marrama D, Vergoni AV, Bertolini A 1991 Effects of thyroid status on the characteristics of a1-, a2-, b, imipramine and GABA receptors in the rat brain. Life Sci 48:659 – 666 Nicholson JL, Altman J 1972 The effects of early hypo- and hyperthyroidism on the development of rat cerebellar cortex. I. Cell proliferation and differentiation. Brain Res 44:13–23 Nicholson JL, Altman J 1972 The effects of early hypo- and hyperthyroidism on the development of the rat cerebellar cortex. II. Synaptogenesis in the molecular layer. Brain Res 44:25–36 Emerson CH, Bambini G, Alex S, Castro MI, Roti E, Braverman LE 1988 The effect of thyroid dysfunction and fasting on placenta inner ring deiodinase activity in the rat. Endocrinology 122:809 – 816 Mori K, Yoshida K, Fukazawa H, Kiso Y, Sayama N, Kikuchi K, Aizawa Y, Abe K 1995 Thyroid hormone regulates rat placental type III iodothyronine deiodinase activity by inducing kinetic changes different from those in the same isozyme in rat brain. Endocr J 42:753–760 Crantz FR, Silva JE, Larsen PR 1982 Analysis of the sources and quantity of 3,5,39-triiodothyronine specifically bound to nuclear receptors in rat cerebral cortex and cerebellum. Endocrinology 110:367–375 Kolodny JM, Leonard JL, Larsen PR, Silva JE 1985 Studies of nuclear 3,5,39triiodothyronine binding in primary cultures of rat brain. Endocrinology 117:1848 –1857 Oppenheimer JH, Schwartz HL 1985 Stereospecific transport to triiodothyronine from plasma to cytosol and from cytosol to nucleus in rat liver, kidney, brain and heart. J Clin Invest 75:147–154 Larsen PR 1996 Normal physiology of the hypothalamic-pituitary-thyroidal system and its relation to the neural system and other endocrine glands. In: DeGroot LJ, Larsen PR, Hennemann G (eds) The Thyroid and Its Diseases. Churchill Livingston, New York, pp 113–136 Meinhold H, Campos-Barros A, Walzog B, Kohler R, Muller F, Behne D 1993 Effects of selenium and iodine deficiency on type I, type II and type III iodothyronine deiodinases and circulating thyroid hormones in the rat. Exp Clin Endocrinol 101:87–93