0013-7227/90/1262-1276102.00/0 Endocrinology Copyright © 1990 by The Endocrine Society

Vol. 126, No. 2 Printed in U.S.A.

J. ENRIQUE SILVA AND PETER RUDASf Thyroid Unit, Brigham and Women's Hospital, and Thyroid Unit, Beth Israel Hospital, Harvard Medical School, Boston, Massachusetts 02215

ABSTRACT. In view of the defective neurotubule assembly observed in congenital hypothyroidism and the striking morphological abnormalities of the cerebellum in this condition, we have investigated the expression of microtubule-associated protein-2 (MAP2) in the cerebellum of rats with congenital hypothyroidism. Analysis included the measurement of immunoreactive MAP2 and its mRNA. In addition, the intracellular distribution of MAP2 was studied by immunostaining of the appropriate histological preparations. The results showed that the developmental increase in MAP2 is delayed in congenital hypothyroidism, but eventually the concentration of this protein reached normal levels in animals with this condition, even if untreated. These abnormalities in the immunoreactive protein are not paralleled by abnormalities in the abundance of MAP2 mRNA, which was not affected by the thyroid status of the animals. In

C

ONGENITAL hypothyroidism has devastating effects on the development of the central nervous system if left untreated. Furthermore, there is a short period of time when these abnormalities can be reversed by the proper administration of thyroid hormone (1). The brain histology of individuals affected by untreated congenital hypothyroidism shows striking abnormalities, one of the most notable being the meager development of nerve processes and poor connectivity among neurons. One of the organs most strikingly affected is the cerebellum (2). Pioneer studies by Legrand and colleagues (2, 3) show markedly abnormal Purkinje cells with disorganized and sparse dendrites, lack of migration of cells of the external granular layer to deeper layers of the cerebral cortex, and abnormal or absent synapsis among several types of cells in the cerebellar cortex. Received August 3, 1989. Address all correspondence and requests for reprints to: J. E. Silva, M.D., Thyroid Unit, Department of Medicine, Beth Israel Hospital, 330 Brookline Avenue, Boston, Massachusetts 02215. * This work was supported by the Howard Hughes Medical Institute (while Dr. Silva was an Associate Investigator in this Institute at Brigham and Women's Hospital) and NIH Grant DK-18416. t Present address: Department of Physiology and Biochemistry, University of Veterinary Science, Budapest, H-1400 P0B 2, Hungary.

spite of the normalization of the content of the protein, the distribution of MAP2 in the Purkinje cells of hypothyroid rats remained abnormal. Whereas in euthyroid rats the protein rapidly migrated into the dendrites, in the Purkinje cells of hypothyroid pups, MAP2 remained largely confined to the body and the most proximal part of the dendrites. These results suggest that thyroid hormone affects the expression of MAP2 at translation or posttranslational levels. The abnormality in distribution may result from some posttranslational abnormality of the protein itself or some underlying defect in the function of the neurons. These observations are probably relevant to the abnormalities in cerebellar function seen in animals and humans with untreated congenital hypothyroidism. (Endocrinology 126: 1276-1282,1990)

In part because of the reduced nerve process development, Nunez and co-workers (4, 5) examined the polymerization of microtubules in brain of congenitally hypothyroid rats. They found that the in vitro assembly of microtubules is reduced in cytosol from hypothyroid rat brain (4) and that this is corrected by adding cytosol from normal rats. They ultimately came to the conclusion that the defective factor was r-protein (5). Interestingly enough, the abnormal microtubule assembly corrected spontaneously in rats with congenital hypothyroidism by 4-5 weeks of age (4). MAP2 is present in dendrites and is considered to be specific for neurons (6-8), although related peptides can be found in other cell types (9). As r-protein, MAP2 is thermostable and promotes the assembly of microtubules in vitro. However, MAP2 is a bigger and more complex molecule, probably involved in several other functions (10). Given the striking abnormalities of the dendrites of Purkinje cells in the cerebellum of rats with congenital hypothyroidism, we examined the expression of MAP2 in animals with congenital hypothyroidism. Our findings show that the concentration of this protein in the cerebellum of rats with congenital hypothyroidism is indeed reduced but, as happens with T-protein, the concentra-

1276

Downloaded from https://academic.oup.com/endo/article-abstract/126/2/1276/2533290 by University of Massachusetts/Amherst user on 03 February 2019

Effects of Congenital Hypothyroidism on MicrotubuleAssociated Protein-2 Expression in the Cerebellum of the Rat*

CEREBELLAR ABNORMALITIES IN CONGENITAL HYPOTHYROIDISM

tion normalizes with age. In addition to the abnormality in the concentration of this protein, we found that its distribution in the Purkinje cells was markedly and persistently altered by congenital hypothyroidism.

Congenital hypothyroidism was induced by adding 0.02% methimazole (MMI) to the drinking water of pregnant SpragueDawley rats at 16 days gestation. This regimen has been previously shown to produce marked hypothyroidism in both the mothers and pups. Hypothyroidism persisted in the pups if the nursing mothers were maintained on MMI. The efficacy of the treatment has been documented by the undetectable levels of T4 and T 3 in the sera of mothers and pups (11). MAP2 purification MAP2 was purified from the forebrain of adult normal rats. Cytosol was obtained and boiled, which precipitated all thermolabile proteins. The boiled cytosol was centrifuged, and the supernatant was submitted to gel filtration, initially in a 1.8 X 50-cm column with Ultrogel AcA44 (LKB, Rockville, MD). Subsequent preparations were made in a Superose 12 column in a fast protein liquid chromatography (FPLC) system (Pharmacia, Piscataway, NJ). In both cases, the excluded volume Containing proteins of greater than 250,000 mol wt was collected and analyzed by sodium dodecyl sulfate (SDS)-gel electrophoresis. Initially, the homogenizing buffer was 20 mM MES, 80 mM NaCl, 2 mM EGTA, 0.1 mM EDTA, 1 mM MgCl2, and 0.2 mM phenylmethylsulfonylfluoride, pH 6.4, but subsequently, when the FPLC system was used, this buffer was 0.05 M Tris, 0.4 M KC1, and 1 mM EDTA, pH 9.0. The column was eluted with the same buffer. The addition of salt to cytosol resulted in higher levels of purification of MAP2 and better flow rates. The excluded volume was boiled again and repurified by these same methods. Except for some traces of T-protein, the preparations were highly purified. MAP2 antibodies Polyclonal antibodies were obtained by injecting one such preparation of MAP2 into White New Zealand rabbits following standard procedures. Four of four rabbits gave antibodies, the best of which was rabbit 4 (R4). This antibody actually was better than a monoclonal antibody (kindly provided by Dr. Kenneth Kosik) in that it detected only intact or large fragments of MAP2, whereas the monoclonal antibody also reacted with smaller degradation products of this protein. MAP2 RIA Highly purified MAP2 was iodinated as follows. A 5-fold molar excess of succinimidyl-hydroxy-phenyl-propionate (SHPP) was allowed to react with MAP2 in 0.05 M borate buffer, pH 8.8, in the presence of dimethylformamide following the instructions of the manufacturer of SHPP (Pierce, Rockford, IL). After 2-3 h of incubation in an ice bath, unreacted SHPP was quenched with 1 vol 0.2 M Tris buffer, pH 8.8.

MAP2 was separated from unconjugated, inactivated, or hydrolyzed SHPP by chromatography in a 2-ml Sephadex G-25 (fine) column equilibrated in the same Tris buffer. The appropriate fractions, usually containing 100-200 ng MAP2, were labeled with 2-3 mCi 125I in the presence of 100 ng chloramine-T. The reaction was stopped 1 min later with the addition of 800 tig sodium metabisulfite. After adding 3 ^1 10% SDS (final concentration, 0.1-0.15%), the sample was left sitting 30-40 min. Then, the 125I-labeled MAP was separated from the unreacted materials by chromatography in Sephadex G-25 (fine), as described above, in the same Tris buffer containing 0.01% BSA. The specific activity of the iodinated material was 10 nCi/fig. Its purity was checked by SDS-gel electrophoresis and gel filtration in a Superose 6 column (FPLC, Pharmacia). This iodination procedure was designed in view of the difficulty of iodinating MAP2directly with chloramine-T, Iodogen (Pierce), or Bolton-Hunter reagent. The RIA buffer was 0.05 M Tris, 0.1 M NaCl, 1 mM EDTA, and 0.25% BSA, pH 8.5. Samples and standards were incubated for 20-24 h with a 1:200 dilution of the R4 antiserum. Bound and free MAP2 were separated with goat antirabbit immunoglobulin G, with which the assay tubes were incubated for 4 h at 4 C. Precipitates were cleared by centrifugation at 2000 X g for 10 min. Before centrifugation, 2 ml ice-cold assay buffer were added to dilute the unbound counts. Tubes were left upside down to allow full drainage of the liquid and subsequently were counted. Binding of tracer in the absence of added unlabeled MAP2 typically ranged between 12-20%. Standards ranged from 0.1-2 ng MAP2, the latter inducing a displacement of 70%. Binding curves subtended by increasing concentrations of the iodinated protein or by displacement of tracer amounts of iodinated protein with cold material were superimposed. The displacement of antibody-bound [125I]MAP2by crude or boiled cytosol was identical to that obtained with purified MAP2, indicating the lack of interference of other proteins present in the cytosol. Unless otherwise indicated, all of the assays were performed with boiled cytosol. Figure 1 shows Coomassie blue staining of SDS-polyacrylamide gel electrophoresis (SDSPAGE) of purified MAP2 used as standard. Western blots employing R4 antiserum are also shown. Note that in crude cytosol this antiserum detects only MAP2. Figure 2 shows a typical standard curve wherein the data has been linearized by log-logit transformation (see Results). Immunostaining of cerebella Cerebella from animals of several ages were fixed in 4% paraformaldehyde-0.2% glutaraldehyde in 0.1 M Na phosphate buffer, pH 7.4, at 4 C for 16-20 h. Fixed tissues were left overnight at 4 C in 30% sucrose-50 mM phosphate buffer, pH 7.4. With pups less than 1 week old, the cerebella were fixed by immersion, whereas older animals were infused through the heart successively with saline and then with the fixing solution in which the cerebella were subsequently left overnight as described above. One hundred-micron slices of tissue were affixed to glass slides, which were washed in 50 mM Tris-100 mM NaCl, pH 7.5. These preparations were subsequently exposed to 10% non-

Downloaded from https://academic.oup.com/endo/article-abstract/126/2/1276/2533290 by University of Massachusetts/Amherst user on 03 February 2019

Materials and Methods

1277

1278

CEREBELLAR ABNORMALITIES IN CONGENITAL HYPOTHYROIDISM

Endo • 1990 Voll26«No2

2r

MAP.

- • Stondord curve ® Crude cytosol ©Boiled cytosol

116 0.05

92.5

66

O.I O.2 O.5 ug MAP2 /tube

.0

2.0

FIG. 2. Standard curve of a MAP2 RIA. Binding data have been loglogit transformed to make the curve linear. The indicated amounts of crude or boiled cytosol were used to displace [125I]MAP2 from its antibody. The regression lines satisfying the experimental points generated with cytosol had a slope identical to the standard curve (r < —0.98; not shown). B/Bo, Bound to free ratio.

Results

FIG. 1. Left panel, Coomassie blue staining of partially purified MAP2 (left lane) and mol wt markers (right lane). MAP2has a calculated mol wt of 270,000. Right panel, Western blot of a partially purified MAP2 preparation (left lane) and crude cytosol (right lane). Note that while the antibody reacts with fragments of MAP2 resulting from spontaneous degradation, in a fresh cytosolic preparation it does not cross-react with other cytosolic proteins (right lane).

immune goat serum containing 1 mM EGTA and 0.2% Triton X-100 to block nonspecific binding of the antibody. Subsequently, they were incubated overnight at 4 C with a 1:1000 dilution of R4 antiserum in 1% goat serum containing 1 mM EGTA. Immunostaining was performed with Vectastin (Vector, Burlingame, CA). Briefly, the slides were washed in same Trissaline buffer mentioned above and incubated with biotinilated goat antirabbit immunoglobulin G. This second antibody was then visualized using avidin and biotinilated horseradish peroxidase and stained with H202 and diaminobenzyline for 5 min. Lastly, the samples were weakly stained with hemotoxylin for visualizing the nuclei. MAP2 mRNA The relative abundance of MAP2 mRNA was assessed with a human cDNA kindly provided by Dr. Rachel Neve from Children's Hospital Medical Center (Boston, MA) (12). This 2.4-kilobase cDNA was labeled by the random priming method (13) to a specific activity of 2-3 x 108 cpm/mg DNA. RNA was analyzed by both Northern blot and dot blot methods (14). RNA was extracted and purified by standard methods adapted to the brain by others (15).

The left panel of Fig. 1 shows Coomassie blue staining of SDS-PAGE of purified MAP2 from adult rat brain. The right panel represents the immunostaining of a purified MAP2 preparation (left lane) and crude cytosol after electrophoresis and electrotransfer onto nitrocellulose paper. Please observe that in the whole cytosol R4 recognizes only the band corresponding to MAP2. The smearing of the purified preparation probably represents some degradation of the protein at the time of the assay. Figure 2 shows a typical standard curve in the RIA and curves generated by increasing volumes of either crude or boiled cytosol from adult rat brain. Please observe that regardless of treatment, the cytosol generated superimposable curves which, in turn, superimposed nicely onto the standard curve. This indicates that other proteins present in crude cytosol do not interfere with MAP2 binding to the antibody. Because of the high specificity of this antibody, it is possible to use crude cytosol, as opposed to boiled cytosol or partially purified MAP2, to measure this protein in other areas of the brain. Nonetheless, in the present studies we performed the assay in boiled cytosol. Figure 3 shows the MAP2 content of cerebella of neonatal rats at various ages. The left panel represents the total content of MAP2 per cerebellum, which increases sharply in the two weeks after birth and reaches a plateau at 2 weeks of age. In rats with congenital hypothyroidism, the slope of this curve is less steep than in normal rats, giving rise to a wide difference between both groups at 1

Downloaded from https://academic.oup.com/endo/article-abstract/126/2/1276/2533290 by University of Massachusetts/Amherst user on 03 February 2019

200