GENERAL

AND

COMPARATIVE

ENDOCRINOLOGY

85,

193-207 (1992)

Characterization of 3,5,3’-Triiodothyronine Receptors in Primary Cultures of Hepatocytes and Neurons from Chick Embryo’ J. GAGNON,N.GALLO-PAYET, *T2J.-G. Lmroux,‘! AND D. BELLABARBA*,~ Endocrine

Laboratory, Departments University of Sherbrooke,

of *Medicine, Medical School,

fBiochemistry, Sherbrooke,

S. BELISLE$~

and #Obstetrics Quebec, Canada

and Gynecology, JlH 5N4

Accepted May 19, 1991 We have detected the presence of nuclear 3,5,3’-triiodothyronine (T,) receptors in primary cultures of chick embryo hepatocytes and neurons. Hepatocytes were isolated from livers of embryos of 12, 16 and 19 days by treatment with 0.2% collagenase and hyaluronidase. They were plated at a density of 34 x 10’/35-mm petri dish in Ham’s F-10 medium containing fetal calf serum, tryptose phosphate, and antibiotics. Cells were used for the binding assay at Day 3 of culture. Neurons from &day-old embryo brains were cultured in a serum-free medium at a density of 1.2 x lo6 cells/35mm petri dish and used for the binding assay after 7 days of culture. Biological activity of hepatocytes was determined by measuring insulin binding, inositol phosphate formation, and 5’-monodeiodinase activity. Neurons or glial cells in culture were identified by immunostaining with anti-neurofilaments and anti-glial ftbrillary acidic protein antisera. Binding assay was performed with isolated nuclei and 0.4 M NaCl nuclear extracts. With the latter preparation, the Scatchard analysis showed, in both ceils, a single, high-affinity, low-capacity T, receptor. In the hepatocytes of 12-, 16-, and 19-day-old embryos association constants (K,) were, respectively, 0.93 2 0.02, 0.74 * 0.03, and 0.56 f 0.04 nM-‘, whereas the maximal binding capacities (MBC) were 2.26 +- 0.2,2.72 2 0.33, and 1.83 ? 0.19 fmol/pg DNA (mean f SE, n = 3). In neurons K, was 1.25 2 0.53 nM-’ and MBC 0.59 2 0.14 fmol/pg DNA (n = 3). The receptor had a sedimentation coefficient of 3.4 S, an estimated M, of 59 kDa, and the following relative affinity for thyroid hormone analogues: TRIAC > L-T~ > L-T~. These data indicate that cultured hepatocytes and neurons of chick embryo contained T3 receptors with properties similar to those described in intact tissues from this and other species. Only the MBC of neurons was 50% lower than that observed in whole brain of embryo, but was comparable to values observed in cultured neurons from other species. 8 Iwz Academic PRSS, II-C.

In many species, thyroid hormones have a critical effect on the growth and development of various tissues, including liver and brain (Romanoff and Laufer, 1956; Balazs, ’ This work was supported by a grant (MT-3943) of the Medical Research Council of Canada (MRC). * Recipient of a development award of the MRC (DG-301). 3 Present address: Department of Obstetrics and Gynecology, Facultt de Medecine, Universite de Montreal, C.P. 6128, Succ. A, Montreal, Quebec, Canada H3C 3J7. 4 To whom requests for reprints should be addressed.

1977; Legrand, 1982). It is well known that these hormones initiate their action by binding to nuclear receptors which are nonhistone proteins with M, of 50 kDa, soluble in high salt solution, and present in all target tissues. The formation of the hormonereceptor complex is followed by increases of various mRNAs and, therefore, by enhanced protein synthesis which determines the maturation of tissues during the embryo development (for review see Oppenheimer et al., 1987). Moreover, recent studies have shown that both the thyroid hormone and the glucocorticoid receptors belong to the 193 0016~6480/92 $1.50 Copyright 0 1992 by Academic Press. Inc. All rights of reproduction in any form reserved.

194

GAGNON

c-erbA family of ligand-dependent transcriptional regulatory proteins (Evans, 1988). These proteins, which share similar domain structures and possess significant sequence homology, determine the function of steroid and thyroid hormones in regulating cell proliferation and maturation (Franklyn and Sheppard, 1989). Most of the studies which have characterized the binding properties of 3,5,3’-triiodothyronine (T,) receptors have been made in nuclei or nuclear extracts from whole tissue. In order to overcome the problem of cell heterogeneity and to develop systems where it will be possible to study the direct effect of the hormones, various cell culture systems have been developed. A number of papers have demonstrated the presence of T3 binding sites in both neurons and glial cells from different regions of the brain of adult rat (Kolodny et al., 1985; Yokota et al., 1986; Puymirat and Faivre-Bauman, 1986; Luo et al., 1986; Yusta et al., 1988). These studies have revealed some differences in the binding properties of such sites according to the cell type considered. On the other hand, hepatocyte cell cultures, including those from chick embryo, have also been extensively developed in order to study various cellular functions including the effects of T, on lipogenic enzymes and of insulin and vasopressin on glucose metabolism (Goodridge, 1973; Goodridge et al., 1974; Goodridge and Adelman, 1976; Tarlow et al., 1977; Thomas et al., 1984; Winberry et al., 1983; Agius et al., 1986; Menuelle et al., 1988; Cake et al., 1989; Duerden et al., 1989). However, no study on T, binding in cultured hepatocytes from embryo have been performed. In our laboratory, we have previously characterized the 3,5,3’-triiodothyronine (T,) receptors in liver and brain homogenates from chick embryo and we have described their ontogenic patterns which show significant changes during embryo development, particularly in the brain (Bella-

ET AL.

barba and Lehoux, 1981; Bellabarba et al., 1983). Since such changes could be influenced by extracellular factors we sought to characterize these receptors in isolated cells to find out if their properties and ontogeny were similar to those of the whole tissues. Therefore in the present study, we describe the features of T, receptors in hepatocytes and neurons grown in monolayer culture. The biological activities of our cell cultures have been assessed by 5’deiodinase activity, insulin binding, vasopressin-induced phosphoinositide breakdown, and inositol phosphate production. Our results show that the T, receptors have properties essentially similar to those observed in intact tissues. Moreover, preliminary results on insulin binding and vasopressin-induced inositol phosphate production confirm that these cell cultures represent a useful system to investigate the receptor ontogeny and function. MATERIALS

AND METHODS

Materials Golden comet chick embryos were obtained from a local hatchery. Radiolabeled [‘*‘I]T, or [‘*‘I]rT, (sp act, 2800 mCi/ug) were purchased from Amersham (Mississauga, Ontario, Canada). myo-[3H]Inositol (sp act, 15-20 mCi/mmol) was from New England Nuclear (Boston, MA). L-T~, L-T~, and D-T, as free acids and TRIAC, deoxyribonuclease I, hyaluronidase, bovine insulin, human transferrin, putrescine, porcine skin gelatin, sodium selenite, L-polylysine, progesterone, 17B-estradiol, N-2-hydroxyethyl piperazine-W-2ethanesulfonic acid (Hepes), and compound Hoechst 33258 (bisbenzamyde) were obtained from Sigma Chemicals (St. Louis, MO). Ham’s F-10, Ham’s F-12, Dulbecco’s modified Eagle’s medium (DMEM), antibiotics (penicillin, streptomycin), L-glutamine solution, fetal calf serum (FCS), and tryptose phosphate broth (TPB) were from GIBCO (Grand Island, NY). Collagenase type I was obtained from Cooper Biochemicals (Freehold, NJ) and Linbro six-well culture plates (35-mm diameter) were obtained from Flow Laboratories (MacLean, VA). Antiglial fibrillary acidic protein (anti-GFAP) and antineurofilament (anti-NF) antisera were from Boehringer (Dorval, Quebec, Canada). Fluorescin isothiocyanate (FITC)labeled goat anti-rabbit immunoglobulin G (IgG) was

T3 RECEPTORS

IN

HEPATOCYTES

purchased from U.S. Biochemical OH).

Co. (Cleveland,

Preparations of Cell Cultures Hepatocytes. Hepatocytes were isolated and cultured using a modification of the methods of Gebhardt et al. (1978). The various steps of tissue isolation were performed at room temperature in DMEM culture medium supplemented with antibiotics (100 U/ml of penicilline and 100 pg/ml of streptomycin) and 2 pU/ml of heparine and 0.6 mM EGTA. The livers from 12-, 16-, and 19-day-old embryos were removed aseptically, rinsed with DMEM, finely minced, incubated for 10 min in the same medium, and then centrifuged for 10 min at 1OOg. This washing facilitated the removal of blood cells. The pellet was resuspended in DMEM containing 0.2% collagenase, 0.2% hyaluronidase, and 0.07% deoxyribonuclease and incubated for 30 min at 37”. The cells were dissociated by aspiration first through a lo-ml pipette and then through a firenarrowed Pasteur pipette. The cell suspension was tiltered through a 250-urn nylon filter and centrifuged once at 600g for 2 min and twice at 1OOgfor 5 min in order to eliminate cell fragments. Pure hepatocytes were obtained after centrifugation for 10 min over a Percoll gradient (density of 1.0853 g/ml). The entire hepatocyte isolation procedure requires approximately 3 hr. The hepatocytes were plated at an initial density of 3-4 x lo5 cells/petri dish (35 mm) in Ham’s F-10 medium supplemented with 10% FCS, 10% TPB, 200 U/ml of penicillin, 200 &ml of streptomycin, and 14.4 mM NaHCO,. Cell viability as determined by trypan blue exclusion was 97%. The cultures were incubated under an atmosphere of 95% air and 5% CO,. Twenty four hours after plating the medium was replaced by Ham’s F-10 with 10% (v/v) FCS depleted of thyroid hormones by charcoal treatment. Forty eight hours later (3 days of culture) the cells were used for the binding assay. At that time, each petri dish contained 2-3 x lo5 cells. Neurons. Primary neuron cultures were generated as described by Faivre-Bauman et al. (1984). Briefly, brains of I-day-old chick embryos were stripped of the meninges, minced with scissors in Ham’s F-10 supplemented with 10% FCS, and then mechanically dissociated by 10 passages through a tire-narrowed Pasteur pipette in 1 ml of the same medium. The suspension was left standing at room temperature for 5 min and the procedure was repeated three times on the decanted pellet. The pooled suspension was filtered through a 250~pm nylon filter and centrifuged at 450g. The cell pellet was then resuspended in serum-free medium (SFM) and plated at a density of 1.2 X lo6 cells/petri dish coated with swine skin gelatine (250 mg/ml) and L-polylysine (10 mg/ml). The SFM was as described by Bottenstein and Sato (1979) and consists

AND

NEURONS

OF

CHICK

EMBRYO

195

of 50% Ham’s F-12 and 50% DMEM supplemented with 14 mM NaHCO,, 15 mM Hepes, 0.5 mM ghrtamine, 5 mg/rnl of bovine insulin, 0.1 mM putrescine, 100 mg/ml human transferrin, 30 mM selenium, 20 nM progesterone, and 1 pM 17B-estradiol. The cell viability was 97% as determined by trypan blue exclusion. The medium was changed after 3 days of culture, and the cells were used for the binding assay after 7 days of culture. At that time, the cell density was 2.5-3 x IO5 cells/petri dish.

Insulin Binding Insulin binding was assayed as described by GalloPayet et al. (1984) and Guillon and Gallo-Payet (1986). Briefly, pure monoiodinated insulin (sp act, 130-150 mCi/pg) was obtained by the chloramine-T method after purification by high-pressure liquid chromatography. At the beginning of each experiment, the culture medium was aspirated and the cells were washed twice with 2 ml phosphate buffer saline (PBS: 1 mM KH,PO,, 1 mM Na2HP04, 137 mM NaCl, 3.5 mM KCI, 1 mM CaCI,, 1 mM MgCl,, 5 m&f Hepes, pH 7.4) supplemented with 1 giliter glucose. The hormone binding reaction was initiated by rapidly aspirating the PBS solution and adding to each petri dish 1 ml PBSglucose-BSA (1 mg/ml) containing 50 pM of [1251]insulin in the absence (total binding) or presence of an excess amount (1 JLM) of unlabeled hormone (nonspecific binding). After a 2-hr incubation at IS”, the incubation medium was removed, and 2 ml PBS was added to the cells. Cells were rapidly detached by scraping with a rubber policeman. The cell suspension was layered on the surface of a Millipore filter (EAWP, 0.45 urn) under continuous aspiration. Petri dishes and tilters were rinsed twice with PBS. The radioactivity retained on the filter, which represents the [‘251]insulin bound to the hepatocytes or neurons, was counted in a gamma counter.

Inositol Phosphate Production The hepatocytes were loaded 18 hr before the experiment with myo-[3H]inositol (2 mCi/ml). Thereafter the culture medium was removed and the cells were incubated for 1 hr in culture medium alone in order to remove as much myo-[3H]inositol as possible. Cells were then washed twice and incubated for 15 min at 37” in the presence of 10 mM LiCI, and then for an additional 15 min in the presence of vasopressin and LiCI. The incubation was terminated by rapidly removing the medium and adding 1 ml 5% HC104 and 200 ml BSA (20 mg/ml). After centrifugation, the inosit01 phosphates (IP, and IP, + IP,) in the supematant were separated by chromatography on Dowex AGlX8 columns as previously described by Guillon and Gallo-Payet (1986). Radioactivity found in the IPi and

196

GAGNON

the IP, plus the IP, fractions was determined by scintillation counting in gel phase in a beta counter.

5’-Deiodinase

Activity

The 5’-deiodinase activity has been determined by a modification of the method of Galton and Hiebert (1987). After washing with PBS-glucose, the cells were scraped from petri dishes, homogenized in PBS containing 0.1 M DTT and 1 mM EDTA, and sonnicated. Aliquots of 175 ml of the suspension (50 mg proteins) were preincubated for 15 min at 37” in the presence of 20 mM DTT. Reaction was initiated by addition of 0.05 pM of [1Z51]T, and increasing amounts of unlabeled rT,. The incubation was carried out at the same temperature and for the same length of time. The reaction was stopped by addition of 4% BSA and 20% TCA. After centrifugation, the 12’1 present in the supernatant was separated by chromatography on Ag50W-X8 (H*) resin columns, whereas the 3,3’[‘251]T2 was identified by paper chromatography on a tertiary amylo-ammonia-hexane system.

Immunochemical

Studies

Indirect immunofluorescent studies using antiGFAP (glial cells) and anti-NF (neurons) antisera counterstained with FITC-labeled goat anti-rabbit antiserum were performed on cells grown on L-polylysine coverslips and fixed in 95% ethanol-5% acetic acid. We used the method described by Leonard and Larsen (1985). Emitted fluorescence was observed on a Zeiss microscope using the BG-12 filter.

T3 Binding Assay Hepatocytes and neurons were washed twice before the assay with PBS containing 1% glucose. We then added to each petri dish the PBS-glucose plus 0.5% BSA, 0.05 nM [i2’I]T,, and increasing amounts of unlabeled hormone ranging from 0.1 nM to 1 pM. The optimal incubation time, established in preliminary experiments, was 24 hr at 4” for the hepatocytes and 30 min at 37” for the neurons. After the incubation, the cells were harvested with a rubber policeman and the petri dishes were washed with PBS to collect all the cells. They were transferred into tubes and centrifuged at 600g for 10 min. The pellets were washed twice, by gentle agitation on a vortex, with a 20-n-&f Tris-HCI buffer at pH 7.8, containing 0.25 M sucrose, 1.1 mM MgCI,, and 0.2% Triton X-100, and centrifuged at 300,OOOgfor 50 min to precipitate the nuclei. The radioactivity in the pellet, representing [‘251]T3 bound to nuclei, was then measured. The pellet was resuspended with the same Tris buffer containing 0.4 M NaCl, 2 mM EDTA, and 5 rnM mercaptoethanol. The suspension was incubated at 4” for 45 min, agitated every 5 min by vortex, and then centrifuged at 1OOOg

ET AL. for 10 min. The radioactivity measured in the supernatant represented the extracted T,-receptor complex. The relative affinity of the receptors for L-T~, D-T,, L-T~, and TRIAC was determined by incubating the cells with a trace amount (0.05 nM) of [‘251]T, and increasing concentrations (0.1 nM-I @4) of L-T, and each analogue.

Density Gradient Sedimentation Gel Chromatography

and

For these studies, the cells were cultured in 75-cm2 tissue culture flasks at a density of 5 X lo6 cells. The sedimentation coefficient of the complex was measured on 4.8 ml of a 5-30% linear glycerol density gradient (Michel et al., 1974) after a 3-hr ultracentrifugation at 340,OOOg using a vertical rotor. After the centrifugation, we collected 0.5-ml fractions using a Beckman sampler and we measured the amount of [‘251]T, in each fraction. In a parallel gradient, we ran a 5% solution of BSA and determined its sedimentation coefftcient (4.4 S) by absorbance at 280 nm. To further characterize the T,-receptor complex, the labeled nuclear extract was chromatographed on Sephadex C-100. A 100 x 1.5~cm column (total volume of 190 ml) was packed with gel and equilibrated with a IO-n&f Tris-HCI buffer containing 0.4 M NaCl, 1 mM EDTA, and 0.1% sodium azide, pH 7.5. We then applied 2 ml of labeled nuclear extracts and carried out the elution with the same Tris buffer. Fractions of 1 ml were collected using a Super-Rat Fraction collector (LKB, Turku, Finland). The radioactivity in each fraction was measured and the elution profile was compared with those of blue dextran and proteins of known molecular weight (BSA, ovalbumin, carbonic anhydrase, and cytochrome c) to estimate the approximate molecular weight of the T,-receptor complex.

Measurement of Radioactivity Data Treatment

and

The B or y radioactivities were determined in Beckman L-8000 beta or L-9000 gamma spectrometers with an efficiency of 60%. Data of the binding assay were analyzed by the method of Scatchard (1949). The association constant (K,) was calculated from the slope of the curve, whereas the maximal binding capacity (MBC) was assessed by the abscissa intercept and expressed as femtomoles per microgram of DNA present in the nuclear pellet. Results were expressed as means * SE of three or more experiments. Unpaired Student’s t test was used for statistical analysis.

Protein

and DNA Determination

Proteins were determined by the technique of Lowry et al. (1951) using bovine serum albumin as a standard. DNA was measured according to the modi-

T3

RECEPTORS

IN

HEPATOCYTES

tied method of Labarca and Paigen (1980) with the compound Hoechst 33258. Hepatocytes contained about 11 pg of DNA/lo6 cells and neurons contained 12.1 kg DNA/lo6 cells.

RESULTS Characterization of Embryonic Liver Cells in Monolayer Cultures Morphology

The method used yielded a large number of liver cells with high viability and plating efftciency. When freshly prepared, 97% of the hepatocytes in the cell suspension excluded the trypan blue dye, and 24 hr after seeding, more than 95% of the cells were firmly attached to the surface of the petri dish. As shown in Fig. I, the cells have the same polygonal morphology seen in histological slices of embryo liver. With an initial density of 3-4 x lo5 cells/cm’, we obtained confluent monolayer cultures after 3 days. Biological Activity

Three activities have been measured in our hepatocyte cultures: (1) Insulin bind-

AND

NEURONS

OF

CHICK

EMBRYO

197

ing, (2) vasopressin-induced inositol phosphate production, and (3) deiodinase activity. The first one has been previously studied in rat fetal hepatocytes (Blazquez et al., 1976) and the two other parameters are considered good markers of hepatocyte activity in the adults liver. The results of all these biological activities are presented in Table 1. Insulin binding. Studies previously performed in our laboratory (Gallo-Payet and Hugon, 1984) have shown that the optimal insulin binding was achieved either after a 30-min incubation at 37” or after a 2-hr incubation at 15”. Under the latter conditions, and in the presence of 50 ph4 of [‘251]insulin, the specific binding was 1.00 + 0.09%/l x lo6 cells (n = 4), which represents 19.7 fmol insulin bound/mg cell protein or 118.2 pg/mg protein. As cell membranes represent 9% of the total cell protein content, binding capacity is 177 fmol/mg membrane protein. Znositol-phosphate production. In order to compare our results with those previously published (Thomas et al., 1984; Poggioli et al., 1986; Pittner and Fain, 1990), we have used vasopressin at the concentration

PIG. 1. Phase-contrast appearance of hepatocytes from 12-day-old chick embryos at Day 3 of culture. Primary culture conditions were described under Materials and Methods. The biological activity of the cells was determined by insulin binding, inositol phosphate production, and 5’monodeiodinase activity. Magnification, X320.

198

GAGNON

ETAL.

TABLE 1 BIOLOGICAL ACTIVITIES OF CULTURED HEPATOCYTES FROM 16-DAY-OLD EMBRYO Insulin binding Percentage specific binding Specific binding Inositol phosphate formation Stimulation by vasopressin (1 x IO-’ M) IP, IP, + IP, 5’-Deiodinase activity v max

1.00 2 0.09%/l x IO6 cells 118 2 11 pg/mg protein 44” 6%” 408 k 3S%b 9 I .O 2 8 pmol/minlmg protein

Note. Values are means 2 SE of three experiments. LI IP, accumulation rose from 4872 k 339 dpm/106 cells under control conditions to 7000 2 584 in vasopressinstimulated hepatocytes. b IP, + IP, accumulation increased from 140 f 10 dpm/106 cells in controls to 712 ? 99 in vasopressinstimulated cells.

of 100 nM, which maximally stimulates inositol phosphate production in a number of tissues. With this concentration we observed an increase of IP, and IP, + IP, of 43.6 + 5.7 and 408 * 35%, respectively over the control values. The absolute amount of IPs produced is higher in IP, than the amount of IP, and IP,, which confirms that also in fetal cells, LiCl predominantly inhibits IP phosphatase. .5’-Deiodinase activity. The V,,,,, of 5’deiodinase activity of isolated hepatocytes was 91 + 8 pmol/min/mg protein, a value close to those reported in the whole liver homogenate of chick embryo (Galton and Hiebert, 1987). Characterization of Neuronal Cell Cultures Using brains from 8-day-old embryos and culture conditions as described under Materials and Methods, the cells we obtained after 7 and 10 days in culture were mainly neurons. As shown in Fig. 2A, we observed numerous clusters of phase-bright cells with abundant processes which connect both individual cells and cell clusters. These are neurons, since they are positive for neurofilaments (Fig. 2B). Also, the arrows in Fig. 2A indicate the presence of a

few phase-dark polyedral cells which have been identified as glial cells (Fig. 2C). Under the conditions of culture described above, we obtained the growth of rather pure neuronal cells, since we had only a small contamination by glial cells, estimated to be less than 10%. When incubated under the same experimental conditions used for hepatocytes, the neuronal cultures have a specific insulin binding of 0.34 + 0.009%/l X IO6 cells (n = 4), equivalent to 6.1 fmol/mg protein. Time Course of [1251]T3 Binding in Hepatocytes and Neurons in Culture Hepatocytes Time course studies have been performed in nuclear extracts of cells from 19day-old embryo (67 x lo5 cells/petri dish) incubated with 0.05 ti of [‘251]T3 with or without 1 p,iV unlabeled T,. At 37” the maximal specific binding occurred after a 15min incubation, but decreased rapidly thereafter. At 22” the equilibrium was reached after 1 hr of incubation and lasted for 1 hr (data not shown). The same stable binding was obtained after 24 hr at 4”. Displacement curves made at the three temperatures demonstrated that the best reproduc-

FIG .2. Photomicrographs of brain cells of I-day-old chick embryos at Day 7 of culture. Neurons are I by phase-contrast microscopy (A) and by fluorescence microscopy after staining with anti-NF antisei rum and counterstaining with FITC-labeled goat anti-rabbit IgG (B). Few glial cells (>lO%) (C) were iidentified with anti-GFAP antiserum stain and counterstained with the same FITC goat antirabbit IgG. Magnification, x320. 199

200

GAGNON

0

4

6

12

16

20

24

26

Time ( hour )

ET AL.

with the same hormone concentration as hepatocytes (50 pM [1251]T,), the specific binding was at equilibrium between 15 and 60 min and decreases thereafter (Fig. 3B). Subsequent studies have been performed for 30 min at 37” in neurons cultured for 7 days. Under these conditions, the specific binding was 1.33 k 0.34% (n = 3) for 5 x IO5 neuronal cells from &day-old embryo. As seen with the hepatocytes, the nonspecific binding did not exceed 20% of the total counts and did not change during the incubation period. Binding

Time

( min )

FIG. 3. Time course of [‘251]T, binding by nuclear extracts of hepatocytes (Day 3 of culture) at 4” (A) and of neurons (Day 7 of culture) at 37” (B). Cells were incubated with 0.05 n&f [‘251]T, with or without 1 pM unlabeled T, for 24 hr (A) or 30 min (B). (A) Total binding; (0) nonspecific binding; (0) specific binding.

ible results were obtained after the incubation of 24 hr at 4”. Figure 3A illustrates the time course of [lz51]T3 binding to hepatocytes under the latter conditions. During all the incubations we did not detect any change of the nonspecific binding, which never exceeded 20% of total counts, whatever temperature of incubation we used. With 50 pM [‘251]T,, the mean specific binding was 3.75 k 0.87% (n = 6) for 6 x lo5 hepatocytes from 19-day-old embryo. Neurons At room or lower temperatures, neurons detached easily from the substrate. For this reason, we have chosen to perform the binding studies at 37” in order to avoid a thermic stress to the cells. When incubated

Specificity

The unlabeled T, reduced the tracer binding in a dose-dependent pattern with a 50% inhibition obtained with 1.2 nM unlabeled T, in both hepatocytes and neurons. The binding specificity of [‘*‘IIT, to receptors solubilized from hepatocytes and neurons was determined by the ratio of the molar concentrations of L-T~ and other analogs required to produce 50% inhibition of [‘*‘IIT binding. Figure 4 shows that the receptors of hepatocytes have a high affinity for TRXAC, followed by L-T+ D-T,, and L-T~. On an equimolar basis the affinity of TRIAC was six-fold higher than that of T,, whereas that of T, was 4.5-fold lower. Similar results were obtained with the soluble receptors of neurons. Ontogenic Characteristics Binding to Hepatocytes

of the [1251]T, and Neurons

Scatchard analyses of the binding data obtained with either nuclei or nuclear extracts of hepatocytes reveal the presence of a single class of high-affinity, low-capacity receptors. Figure 5 shows representative Scatchard analyses of binding data obtained with nuclei (Figs. 5A and 5C) and nuclear extracts (Figs. 5B and SD) of hepatocytes from embryos of 12 and 16 days of age. Table 2 summarizes the results obtained with both preparations from hepatocytes of 12- to IPday-old embryo. The as-

T3 RECEPTORS

55

80

4 =

60

ri Y

40

Z

IN HEPATOCYTES

AND

NEURONS

OF

CHICK

been estimated to be 15,110 + 1590, 19,796 + 2485, and 14,788 + 2481, respectively, for hepatocytes from 12-, 16-, and lPdayold embryos and 4609 2 1354 for neurons from g-day-old embryo. Physicochemical T,-Receptor

OJ C

201

EMBRYO

-10

-9

ANALOGUE

-8

(LOG

-7

I

-6

M)

FIG. 4. Relative binding affinity of the nuclear T, receptor from hepatocytes of 3-day culture for various thyroid hormone analogues. Cells were incubated with 0.05 n&f [‘251]T, and increasing amounts (0.1 nM-1 p,M) of unlabeled L-T~ (0) and other competitors (TRIAC, Cl; D-T,, 0; L-T~, n ). Incubation, extraction of receptors, and determination of percentage bound T, were performed as described under Materials and Methods. The relative afftnity of TRIAC and L-T, to L-T~ were calculated from the concentration of analogue which reduced by 50% the bindings of [‘251]T,. The TRIAUL-T, ratio was 6, whereas the T,/T, ratio was 0.2. D-T, affinity did not differ from that of L-T~. We obtained similar results with cultured neurons.

sociation constants and the MBC did not differ significantly, although the latter appeared lower in nuclear extracts. The mean dissociation constant for hepatocytes (1.3 1 ? 0.02) was identical to the ED,, of the displacement curve of [‘251]IT, by unlabeled Ts. The MBC showed a slight, nonsignificant, increase from Day 16 to Day 19 in nuclei and from Day 12 to Day 16 in the extracts. The results of the neurons are presented in Fig. 6 and Table 2. Figure 6 shows a representative Scatchard analysis with nuclei (Fig. 6A) and nuclear extracts (Fig. 6B). The association constant for the neurons from g-day-old embryo was higher than that of the hepatocytes, i.e., 1.38 + 0.85 and 1.25 f 0.53 nM-‘, respectively, for nuclei (Table 2) and nuclear extracts, whereas the MBC values were lower (1.18 + 0.42 and 0.59 + 0.14 fmol/pg DNA). As the binding occurred in whole cells, the number of binding sites per cell, calculated from the data of the nuclear extracts, has

Properties Complex

of the

In both types of cells, the T,-receptor complex sedimented at 3.4 S on a glycerol density gradient. Figure 7A illustrates the results obtained with hepatocytes. Figure 7B shows the elution profile of the solubilized T,-receptor complex from hepatocytes on a Sephadex G-100 column. We observed three peaks. The first was eluted with the void volume, the second was situated between BSA and ovalbumin and, by comparison with standard protein, had an estimated M, of 59 kDa. A third peak (not shown) was eluted between fractions 250 and 300 and represented free T,. DISCUSSION Our data indicate that the hepatocytes and neurons of chick embryo, grown in primary culture, possess nuclear T, receptors similar to those of the whole liver and brain. This report is the first to demonstrate the presence of such receptors in the hepatocytes and to characterize the neuronal receptor. Saturable binding sites have already been described in cultured neurons of chick embryo (Pascual et al., 1986), but no immunocytochemical analysis or characterization of the receptor was performed. During the past 8 years many studies have been performed using hepatocyte cell cultures from adult rats in order to investigate the mechanisms involved in liver function (Morin et al., 1982; Duerden et al., 1989; Winberry et al., 1983; Thomas et al., 1984; Agius et al., 1986) and some have been published from fetal sources (Goodridge et al., 1974; Menuelle et al., 1988; Cake et al., 1989). We have observed the same cell morphology as previously de-

202

GAGNON

20

40

ET AL.

60

10

15

B (PM >

B (PM >

B (PM >

B (PM >

20

25

30

FIG. 5. Rep*.j entative Scatchard plots of binding data obtained with nuclei (A, C) and nuclear extracts (B, D) of hepatocytes from 12-day-old (left panel) and 16-day-old (right panel) embryo. The assay was performed by incubating 0.05 nM r1251]Tj and increasing amounts (0.1 n&f-l t&f) of unlabeled T, with 6 x 10’ cells for 24 hr at 4”. Details of the assay are under Materials and Methods. For these particular experiments, K, values for nuclei and nuclear extracts were respectively 0.6 and 0.9 niW’ with Day 12 hepatocytes and 0.9 and 0.7 nM-’ with cells from Day 16. MBC values were 6.0 and 2.0 fmol/pg DNA (12 days) and 2.53 and 2.34 fmoUpg DNA (16 days) for nuclei and nuclear extracts, respectively.

scribed by Tarlow et al. (1977) who reported that the fetal cells have the properties to proliferate, although cells from the adult did not. TABLE ONTOGENIC

Age

PATTERN

The biologic4 activities we have tested provided interesting information. The insulin-binding capacity we observed (177 fmol/ mg membrane protein) was similar to that 2

OF K, AND MBC DETERMINED IN ISOLATED NUCLEI AND NUCLEAR HEPATOCYTES AND NEURONS OF CHICK EMBRYO

Nuclei

12 16 19

0.80 -+ 0.16 0.91 5 0.25 0.58 f 0.12

8

1.38 2 0.85

OF

MBC

K,

(days)

EXTRACTS

Extracts Hepatocytes 0.93 + 0.02 0.74 t 0.03 0.56 2 0.04 Neurons 1.25 f 0.53

Nuclei

Extracts

4.98 k 0.87 3.09 2 0.36 4.28 +- 0.83

2.26 k 0.20 2.72 2 0.33 1.83 -t 0.19

1.18 2 0.42

0.59 + 0.14

NOW. Data are means 2 SE of three experiments. K, is expressed as nk-r; MBC is expressed as fmol/pg DNA. Results obtained with nuclear extracts were used to calculate the number of receptors per cell.

T3

RECEPTORS

IN

HEPATOCYTES

AND

B (PM > B 6.0

-

6.0

m 0

c o.oI

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

B (PM > FIG. 6. Representative Scatchard plots of binding data obtained with nuclei (A) and nuclear extracts (B) of neurons from I-day embryos at Day 7 of culture. The binding assay was performed as described under Materials and Methods and in the legend to Fig. 5. For this particular experiment, K, values were 3.1 and 2.3 mW ’ and MBC values 0.36 and 0.34 fmo@g DNA for nuclei and nuclear extracts, respectively.

obtained by Alvarez and Blazquez (1987) in 21-day-old rat fetus. Moreover, the specific binding and the binding capacity were more than two-fold lower than the values found by Morin et al. (1982) and Kalant et al. (1984) in adult rat hepatocytes. Our results clearly show that the hepatocyte maturation and differentiation are marked by an increase in insulin binding and corroborate the report of Blazquez et al. (1976) who find that insulin binding increased by two- to threefold from the end of the neonatal period to the adult age. On the other hand, our results indicate that the hepatocyte sensitivity to vasopressin in the embryo had already reached the level observed in cells from adult rat. In fact, the fourfold increase

NEURONS

OF

CHICK

EMBRYO

203

of the vasopressin-induced inositol phosphate production we observed was of the same order of magnitude as that reported in hepatocytes of adult rats. (Thomas et al., 1984; Poggioli et al., 1986; Pittner and Fain, 1990). Determination of the 5’-monodeiodinase activity showed that final products, enzyme kinetics and K,,, values were rather similar to those reported by Galton and Hiebert (1987) in embryo liver homogenates. These results indicate that the cultured hepatocytes are capable of metabolizing T, and that the hormone regulates the activity of the enzyme. Therefore the measurement of this enzyme represents a good marker of the biological effect of thyroid hormone on the hepatocytes. Other studies have also demonstrated that T, exerts its biological effect of hepatocytes in culture by regulating the synthesis of lipogenic enzymes and the gene expression of malic enzymes (Goodridge, 1973; Goodridge et al., 1974; Goodridge and Adelman, 1976; Winberry et al., 1983). Neurons were obtained by dissociation of cerebral hemisphere of 8-day-old embryos using a serum free medium and L-polylysine as substrate. With this technique first described by Faivre-Bauman et al. (1984), we obtained a rather pure culture as shown by the abundance of cells stained with anti-NF serum and the few cells that reacted with GFAP antiserum. The low number of glial cells was in part due to the fact that gliogenesis occurs later than 8 days of embryogenesis and to the inhibitory effect of polylysine on glial cell growth (Pettman et al., 1979). The culture conditions did not allow growth and maintenance of neurons from embryos older than 8 days. Specific insulin binding yielded values lower than those obtained by Bassas et al. (1985), also in brain of chick embryo. Differences could be explained on the basis of cell heterogeneity, since these authors employed the whole brain tissue, which, in particular, contained a large number of glial cells.

204

GAGNON

A

ET AL.

Several of our findings show the similarity of the nuclear T, binding site of hepatolOOOcytes and neurons to the receptor previously described in intact tissues from the 600. same or other species. Thus, the values of 5 600. K, and MBC observed in nuclear extracts 0 4007 of hepatocytes were similar to those we have reported for the T, receptor in whole 200. liver of embryo (Bellabarba and Lehoux, 0 Y........ 1981) and very close to those of rat liver 0 8 16 24 32 40 bottom top (Oppenheimer et al., 1987). The ontogenic Fractions (0.5 ml ) pattern showed a slight rise of MBC which was similar to our previous observations (Bellabarba and Lehoux, 1981). As for the B neurons, the K, was also identical to that 3000 which we have observed in whole brain, 2500 but the MBC was approximately 50% of the value we reported with the solubilized brain 2000 receptor (Bellabarba et al., 1983). This difference may be explained by the fact that brain homogenates contain other cells which have T, receptors (Luo et al., 1989). Actually the MBC of our cultured neurons was comparable to those of other cultured 20 40 60 60 100 120 neurons from rat, mouse, and a line of tuFractions (1 .O ml ) mor cells (Kolodny et al., 1985; Luo et al., FIG. 7. (A) Glycerol density gradient of [‘Z51]T,1986; Fuymirat and Faivre-Bauman, 1986; labeled nuclear extracts from hepatocytes of chick embryo. The 0.4 ml of extracts was layered on 4.8 ml of Ortiz-Caro et al., 1986; Yokota et al., a 5-30% linear glycerol density gradient. Sedimenta1986). Among these studies, only Kolodny tion was performed as decribed under Materials and et al. (1985) have characterized the T, reMethods. Five percent BSA, as standard protein, was ceptor from fetal neurons of rats. Furtherrun in a parallel gradient and its position was determore, our T, binding sites had a sedimenmined by absorbance at 280 nm. The solubilized T, receptor sedimented at 3.4 S. Identical results were tation coefftcient of 3.4 S, an estimated iI4, obtained with neuronal extracts. (B) Elution pattern of of 59 kDa, and a relative affinity for T, anthe [‘2sI]T,-labeled nuclear extracts from hepatocytes. alogues, which were also similar to those Cells were incubated with 0.05 nM “‘1 with or without 1 t.&f unlabeled T, and extracts were prepared as de- reported for the T, receptor in whole embryo liver (Bellabarba and Lehoux, 1981) scribed under Materials and Methods. Two milliliters of the extracts was filtered through a Sephadex G-100 and in all other species (Apriletti et al., column. Standard proteins of known M, were eluted 1983). Sephadex G-100 chromatography through the same column (A, BSA, 67 kDa; 0, Ovalalso identified a binding protein present in bumin, 4.5 kDa; CA, carbonic anhydrase, 3 1 kDa; Cytthe void volume corresponding to a second c, cytochrome c, 13 kDa). Void volume (V,) was debinding site, mostly for T,, that we have termined by dextran blue. The elution pattern shows two main peaks with proteins which bind T,, one recently described (Bellabarba et al., 1991). eluted in the void value and another between BSA and The recent discovery that the chick emovalbumin, which had a M, of 59 kDa and correbryo c-erbA proto-oncogene codes for a sponded to the T, receptor. A third peak (not shown) protein which is the thyroid hormone rewas eluted between fractions 25&300 and contained free T,. The continuous line at the bottom of each peak ceptor has allowed considerable advances represents the nonspecific binding. in the understanding of the nature, struc1200

1

T3 RECEPTORS

IN

HEPATOCYTES

ture, and action of the receptor on responsive genes (De Groot et al., 1989). In particular several forms (0~ and p) of receptor whose mRNA are prevalent in specific tissues have been identified. Our data cannot distinguish between such forms, but preliminary results (Gig&e et al., 1991) from our laboratory suggest that the distribution of the cx and p mRNA is closed to that described by Forrest ef al. (1990) in chick embryo and by Strait et al. (1990) in rat. With these in vitro systems, we can avoid the complexities of the in vivo studies and we may perform more direct manipulations of the environmental conditions and observe the cellular response. Moreover, preliminary results on insulin binding and vasopressin-induced inositol phosphate production confirm that these cell cultures represent a useful tool to study the effect of thyroid hormone and other regulators of the maturation and function of embryonic cells. ACKNOWLEDGMENT The authors thank Mrs. Suzanne Fortier for her skillful technical assistance.

REFERENCES Agius, L., Chowdhury, M. H., Davis, S. N., and Alberti, G. M. M. (1986). Regulation of ketogenesis, gluconeogenesis and glucogen synthesis by insulin and proinsulin in rat hepatocyte monolayer cultures. Diabetes 35, 1286-1293. Alvarez, E., and Blazquez, E. (1987). Lack of insulin effect on its receptors in fetal rat hepatocytes. Horm. Metab. Res. 19, 458-463. Apriletti, J. W., David-Inouye, Y ., Baxter, J. D., and Eberhardt, N. L. (1983). Physicochemical characterization of intranuclear thyroid hormone receptor. In “Molecular Basis of Thyroid Hormone Action” (J. H. Oppenheimer and H. H. Samuels, Eds.), pp. 67-69. Academic Press, New York. Balazs, R. (1977). Effect of thyroid hormones and undernutrition on cell acquisition in the rat brain. In “Thyroid Hormone and Brain Development.” (G. D. Grave, Ed.), pp. 287-302. Raven Press, New York. Bassas, L., DePablo, F., Lesniak, M. A., and Roth, J. (1985). Ontogeny of receptors for insulin-like peptides in chick embryo tissues: Early dominance of

AND

NEURONS

OF

CHICK

EMBRYO

205

insulin-like growth factor over insulin receptors in brain. Endocrinology 117, 2321-2329. Bellabarba, D., and Lehoux, J. G. (1981). Triiodothyronine nuclear receptor in chick embryo: Nature and properties of hepatic receptor. Endocrinology 109, 1017-1025. Bellabarba, D., Bedard, S., Fortier, S., and Lehoux, J. G. (1983). 3,5,3’-Triiodothyronine nuclear receptors in chick embryo: Properties and ontogeny of lung and brain receptors. Endocrinology 112, 353-359. Bellabarba, D., Fortier, S., Gagnon, J., and Gig&e, A. (1990). Thyroid hormone binding during chick embryogenesis. Characterization of an additional hepatic binding site. In “Proceedings of the 10th International Thyroid Conference, The Hague, The Netherlands, February 3-8, 1991” p. 45. [Abstract 781 Blazquez, E., Rubalcava, B., Montesano, R., Orci, L., and Unger, R. H. (1976). Development of insulin and glucagon binding and the adenylate cyclase response in liver membranes of prenatal, post-natal and adult rat: Evidence of glucagon “resistance.” Endocrinology 98, 1014-1023. Bottenstein, J. E., and Sato, G. H. (1979). Growth of rat neuroblastoma cell line in serum free supplemented medium. Proc. Natl. Acad. Sci. USA 76, 51&517. Cake, M. H., Ho, K. K. W., Shelly, L., Milward, E., and Yeoh, G. C. T. (1989). Insulin antagonism of dexamethasone induction of tyrosine aminotransferase in cultured fetal hepatocytes. Eur. J. Biothem. 182, 429-435. De Groot, L. J., Nakai, A., and Macchia, E. (1989). The molecular basis of thyroid hormone action. J. Endocrinol. Invest. 12, 843-861. Duerden, J. M., Bartlett, S. M., and Gibbons, G. F. (1989). Long-term maintenance of high rates of very-low-density-lypoprotein secretion on hepatocytes cultures. Biochem. J. 263,937-943. Evans, R. M. (1988). The steroid and thyroid hormone receptor superfamily. Science 240, 889-895. Faivre-Bauman, A., Puymirat, J., Loudes, C., and Tixier-Vidal, A. (1984). Differentiated mouse fetal hypothalamic cells in serum free medium. In “Cell Culture Methods for Molecular and Cell Biology,” Vol. 4, Methods for Serum-Free Cultures of Neuronal and Lymphoid Cells (D. W. Barnes, S. A. Sirbasku, and G. H. Sato, Eds.), pp. 37-56. A.R. Liss, New York. Forrest, D., Sjoberg, M., and Vennestrom, B. (1990). Contrasting developmental and tissue-specific expression of (Y and 8 thyroid hormone receptor genes. EMBO J. 9, 1519-1528. Franklyn, J. A., and Sheppard, M. C. (1989). Thyroid and steroid hormone regulation of proto-oncogene expression. Trends Endocrinol. Metab. 1, 35-39.

206

GAGNON

Gallo-Payet, N.. and Hugon, J. S. (1984). Insulin receptors in isolated adult mouse intestinal cells: Studies in vivo and in organ culture. Endocrinology 114, 1885-1892. Gallo-Payet, N., Guillon. G., Balestre, M. N.. and Jard, S. (1986). Vasopressin induces breakdown of membrane phosphoinositides in adrenal glomerulosa and fasciculata cells. Endocrinology 119, 1042-1047. Galton, V. A., and Hiebert, A. (1987). The ontogeny of the enzyme systems for 5’- and 5deiodination of thyroid hormone in chick embryo liver. Endocrinology 120, 2604-2610. Gebhardt, R., Belleman, P.. and Mecke, D. (1978). Metabolic and enzymatic characteristics of adult rat liver parenchimal cells in non-proliferating primary monolayer cultures. Exp. Cell Res. 112, 431441. Giguere, A., Fortier, S., Lehoux, J.-G., and Bellabarba, D. (1991). Ontogeny of c-erbA mRNA and thyroid hormone receptor in chick embryos. J. Cell. Biochem. (Suppl. 15E), 23. [Abstract LlOl] Goodridge, A. G. (1973). Regulation of fatty acid synthetase in isolated hepatocytes from neonatal chick. J. Biol. Chem. 248, 1924-1931. Goodridge, A. G., Garay, A., and Silipananta, P. (1974). Regulation of lipogenesis and the total activities of lipogenic enzymes in a primary culture of hepatocytes from prenatal and early postnatal chick. J. Biol. Chem. 249, 1469-1475. Goodridge, A. G., and Adelman, T. C. (1976). Regulation of malic enzyme synthesis by insulin, T3 and glucagon in liver cells in culture. J. Biol. Chem. 251, 3027-3032. Guillon, G., and Gallo-Payet, N. (1986). Specific vasopressin binding to rat adrenal glomerulosa cells. Biochem. J. 235, 209-214. Kalant, N., Osaki, S., Maekubo, H., Mitmaker, B., and Cohen-Khallas, M. (1984). Down-regulation of insulin binding by human and rat hepatocytes in primary culture: The possible role of insulin internalization and degradation. Endocrinology 114, 3743. Kolodny, J. M., Leonard, J. L., Larsen, P. R., and Silva, J. E. (1985). Studies of nuclear 3,5,3’triiodothyronine binding in primary cultures of rat brain. Endocrinology 117, 1848-1857. Labarca, C., and Paigen, K. (1980). A simple, rapid, and sensitive DNA assay procedure. Anal. Biothem. 102, 344-352. Legrand, J. (1982). Hormone thyroi’diennes et maturation du systeme nerveux. J. Physiol. (Paris) 78, 604-652. Leonard, J. L., and Larsen, P. R. (1985). Thyroid hormone metabolism in primary culture of fetal rat brain cells. Bruin Res. 327, l-13.

ET AL. Lowry, 0. H., Rosebrough, N. J., Fan-, A. R.. and Randall, R. J. (1951). Protein measurement with the Folin phenol reagent. J. Biol. Chem. 139, 265275. Luo. M.. Faure. R., and Dussault. J. H. (1986). Ontogenesis of nuclear T3 receptors in primary culture of astrocytes and neurons. Brain Res. 381, 275-280. Luo, M., Puymirat, J., and Dussault, J. H. (1989). Immunocytochemical localization of nuclear 3,5,3’triiodothyronine (L-T3) receptors in astrocytes cultures. Dev. Bruin Res. 461, 131-136. Menuelle, P., M’Zali, H., Forest, N., and Plas, C. (1988). Compared roles of glucose, galactose and fructose as glycogen precursors during the acute response to insulin in cultured rat foetal hepatocytes. Eur. J. Biochem. 20, 777-782. Michel. G., Jung, I., Baulieu, E. E., Aussel, C., and Uriel, J. (1974). Two high affinity estrogen binding proteins of different specificity in immature rat uterus cytosol. Steroids 24, 437-449. Morin, 0.. Felhman, M., and Freychet, P. (1982). Binding and action of insulin and glucagon in monolayer cultures and fresh suspensions of rat hepatocytes. Mol. Cell. Endocrinol. 25, 33%352. Oppenheimer, J. H., Schwartz, H. L., Mariash, C. N., Kinlaw. W. B., Wong, N. C. W., and Freake, H. C. (1987). Advances in our understanding of thyroid hormone action at the cellular level. Endocr. Rev. 8, 288-308. Ortiz-Caro, J., Yusta, B., Montiel, F., Villa, A., Arande, A., and Pascual. A. (1986). Identification and characterization of L-triiodothyronine receptor in cells of glial and neuronal origin. Endocrinology 119, 2163-2167. Pascual, A., Aranda, A., Ferret-Sena, V., Gabellec, M. M., Rebel. G., and Sarlitve, L. L. (1986). Triiodothyronine receptors in developing mouse neuronal and glial cell culture and in chick cultured neurons and astrocytes. Dev. Neurosci. 8, 89101. Pettman, B., Louis, J. C., and Sensenbrenner, M. (1979). Morphological and biochemical maturation of neurons cultured in the absence of glial cells. Nature 281, 378-380. Pittner, R. A., and Fain, J. N. (1990). Vasopressin and norepinephrine stimulation of inositol phosphate accumulation in rat hepatocytes are modified differently by protein kinase C and protein kinase A. Biochim. Biophys. Acta 1043, 211-217. Poggioli, J., Mauger, J. P., and Claret, M. (1986). Effect of cyclic AMP-dependant hormones and Cazf-mobilizing hormones on the Ca’+ influx and polyphosphoinositide metabolism in isolated rat hepatocytes. Biochem. J. 235, 663-669. Puymirat, J., and Faivre-Bauman, A. (1986). Evolu-

T3

RECEPTORS

IN

HEPATOCYTES

tion of triiodothyronine nuclear binding sites in hypothalamic serum-free cultures: Evidence for their presence in neurons and astrocytes. Neuroxi. Lett. 69, 299-308. Romanoff, A., and Laufer, H. (1956). The effect of injected thiourea on the development of some organs of chick embryo. Endocrinology 59,61 l-619. Scatchard, G. (1949). The attraction of proteins for small molecules and ions. Ann. N. Y. Acad. Sci. 51, 660672. Strait, K. A., Schwartz, H. L., Perez-Castillo, A., and Oppenheimer, J. H. (1990). Relationship of c-erbA mRNA content to tissue triiodothyronine nuclear binding capacity and function in developing and adult rats. J. Biol. Chem. 265, 10,51410,521. Tarlow, D. M., Watkins, P. A., Reed, R. E., Miller, R. S., Zwergel, E. E., and Lane, M. D. (1977). Lipogenesis and synthesis and secretion of very low density lipoprotein by avian liver cells in non-

AND

NEURONS

OF

CHICK

EMBRYO

207

proliferating monolayer cultures. J. Cell Biol. 73, 332-353. Thomas, A. P., Alexander, J., and Williamson, J. R. (1984). Relationship between inositol polyphosphate production and the increase of cytosolic free Ca*’ induced by vasopressin in isolated hepatocytes. .I. Biol. Chem. 259, 55745584. Winberry, L. K., Morris, S. M., Fisch, J. E., Glynias, M. J., Jenik, R. A., and Goodridge, A. G. (1983). Molecular cloning of cDNA sequences for anion malic enzymes. J. Biol. Chem. 258, 1337-1342. Yokota, T., Nakamura, H., Akamizu, T., Mori, T., and Imura, H. (1986). Thyroid hormone receptors in neuronal and glial nuclei from mature rat brain. Endocrinology 118, 1770-1776. Yusta, B., Besnard, F., Ortiz-Caro, J., Pascual, A., Aranda, A., and Sarlieve, L. (1988). Evidence for the presence of nuclear 3,5,3’-triiodothyronine receptors in secondary cultures of pure rat oligodendrocytes. Endocrinology 122, 2278-2284.