Neuroscience Vol. 36, No. 2, pp. 473-482, 1990
0306-4522/90 $3.00 + 0.00 Pergamon Press plc © 1990 IBRO
Printed in Great Britain
IMMUNOCYTOCHEMICAL LOCALIZATION OF THYROID HORMONE NUCLEAR RECEPTORS IN CULTURED ACETYLCHOLINESTERASE-POSITIVE NEURONS: A CORRELATION BETWEEN THE PRESENCE OF THYROID HORMONE NUCLEAR RECEPTORS AND L-TRI-IODOTHYRONINE MORPHOLOGICAL EFFECTS R. GARZA,* J. PUYMIRATand J. H. DUSSAULT Unit6 de recherche en Ontogrnrse et Grnrtique molrculaire, Centre Hospitalier de l'Universit6 Laval, Qurbec, Canada GlV 4G2 Almtraet--A monoclonal antibody against the rat liver L-tri-iodothyronine nuclear receptor and acetylcholinesterase cytochemistry were used for the localization of thyroid hormone nuclear receptors in acetylcholinesterase-positive cell nuclei in fetal rat cerebral hemisphere neuronal cultures. After 3 days in vitro, the ratio of acetylcholinesterase-positive cells that were immunoreactive for the thyroid hormone nuclear receptor to those not stained for this receptor (74-26%, respectively) remains unchanged despite an increase in the number of acetylcholinesterase-positive cells with time (from day 3 to day 21) in culture. Furthermore, the addition of 3 x l0 -8 L-tri-iodothyronine in culture did not modify this ratio or have an effect on the number of acetylcholinesterase-positive cells, but significantly increased the neurite density in those acetylcholinesterase-positive cells that were immunoreactive for the thyroid hormone receptor. Conversely, no difference in the neurite densities of those acetylcholinesterase-positive cells not stained for this receptor was observed when cultured in the presence or absence of thyroid hormone. In other experiments with the same fetal brain cultures, treatment of cultures for 8 days with L-tri-iodothyronine, beginning on culture day 20, demonstrated the presence of a critical period which occurs in vitro around day 20, since the stimulatory effect of L-tri-iodothyronine on immunoreactive acetylcholinesterase-positive cell neurite density is lost after 20 days in vitro. These results demonstrate, for the first time, the presence of L-tri-iodothyronine nuclear receptors in fetal rat acetylcholinesterase-positive neurons and the existence of a cellular heterogeneity in the distribution of the thyroid hormone receptor. The presence of these receptors in fetal brain acetylcholinesterase-positive neurons suggests that some effects of L-tri-iodothyronine on the maturation of a subpopulation of acetylcholinesterase-positive neurons may result from a direct effect of this hormone through an interaction with its specific nuclear receptors.
The presence of thyroid hormones is essential for the normal development of the brain and these hormones have been shown to influence neuronal development. 7 Several findings indicate that thyroid hormones affect the development of acetylcholinesterasepositive (ACHE + ) central neurons. First, neonatal hypothyroidism is associated with a reduction of A C h E activity in various brain areas of experimental animals. 5'3° Second, L-tri-iodothyronine (L-T3) has been found to increase A C h E activity as well as enhance the morphological development of A C h E + neurons in aggregating and primary cultures of fetal *To whom correspondence should be addressed. ACHE, acetylcholinesterase; CDM, chemically defined medium; CHAT, choline acetyltransferase; FCS, fetal calf serum; FITC, fluorescein isothiocyanate; GAD, glutamate decarboxylase; GALC, galactocerebroside; GFAP, glial fibrillary acid protein; L-T 3, L-tri-iodothyronine; mab, monoclonal antibody; NF, neurofilaments; PBS, phosphate-buffered saline; TH, tyrosine hydroxylase; TNR, L-tri-iodothyronine nuclear receptor; TRITC, tetramethylrhodamine isothiocyanate.
Abbreviations:
brain cells. 4,~° These findings indicate that L-T 3 acts directly on brain cells; however, the molecular mechanisms of thyroid hormone action on neurons remain to be clarified. It is generally accepted that thyroid hormone action is initiated by the binding of L-T 3 to an L-tri-iodothyronine nuclear receptor ( T N R ) ) 8'19 Previous studies using Scatchard analysis revealed the presence of L-T 3 binding sites in neuron-enriched fractions in different areas of the adult brain as well as in primary cultures of neurons and astrocytes. 12A3'15'2°'25'27 These studies, however, d i d not show whether L-T 3 binding sites are located in A C h E + neurons. Our recent development of a monoclonal antibody (2B3 mab) against partially purified L-T3 rat liver nuclear receptor ~4 has made it possible to study the localization of T N R at the cellular level. In this study, 2B3 mab and A C h E cytochemistry were used to localize T N R in A C h E + neurons and to determine if the presence of T N R in this neuronal population was correlated with L-T 3 morphological effects.
473
R. GARZA et al.
474 EXPERIMENTAL PROCEDURES
Cerebral hemisphere neuronal cultures Primary cultures of dissociated cerebral hemisphere cells from 15- to 16-day-old Sprague-Dawley rat fetuses (Charles River, St Constant, Qurbec) were prepared as previously described. 4 After mechanical dissociation in serumsupplemented medium, the cells were centrifuged and resuspended in chemically defined medium (CDM). An aliquot of the cell suspension containing 2 x 105 cells (98,000 cells/cm2) was plated in 0.5ml of CDM on 16mm diameter tissue culture wells containing glass coverslips that had been treated previously with gelatin (250gg/ml), incubated overnight with poly-L-lysine (10/~g/ml in phosphatebuffered saline, PBS), rinsed with PBS and preincubated with 10% fetal calf serum (FCS) stripped of thyroid hormones according to the procedure of Samuels et al. 28 The composition of the CDM was that previously described for similar cultures: Cultures were maintained in a humidified atmosphere of 5% CO2, 95% air at 37°C. The medium was renewed 3 days after seeding and every third day thereafter. In experiments with thyroid hormone, L-T 3 was added at the initiation (except when studying the duration of L-T 3 treatment after 20 days in vitro) of the culture and renewed at every change of the medium. Based on previous dose-response studies,4 L-T 3 was used at a final concentration of 3 x 10-8 M. Preparation and specificity 2B3 monoclonal antibody against rat liver nuclear receptor TNR was purified from liver cytosol as previously described. 2 The preparation and characterization of the monoclonal anti-TNR have been described in detail, t4't6 In the present study, the specificity of the 2B3 mab was determined according to the following criteria: (1) cells incubated with control ascitic fluid or with 2B3 mab pre-adsorbed with purified receptor showed no specific staining; (2) in accordance with previous immunocytochemical localization of TNR in different tissues of the adult rat and in primary cultures of neurons and astrocytes, TNR immunoreactivity was strictly limited to the nucleus, t5:6"2~ Acetylcholinesterase cytochemistry AChE + neurons were identified in dissociated cell cultures by AChE cytochemistry according to the procedure described by Tago et al. 29 Briefly, cultures were allowed to grow for 3, 5, 8, 21 and 28 days, fixed with 3% paraformaldehyde (in PBS), rinsed thoroughly with PBS (pH 7.4) and rinsed twice in 0.I M maleate buffer (pH 6.0). The cells were then incubated for 30 min with a substrate solution containing 72 p M acetylthiocholine iodine, 10 p M K2Fe(CN)6, 6 0 # M CuSO 4 and 100#M sodium citrate. After removal of the substrate solution, the cells were rinsed and incubated with a mixture of 0.4% 3',Y-diaminobenzidine and 0.3% nickel ammonium sulfate for 10 min. Then 0.003% H202 was added and the cells were further incubated for 25 min to reveal optimal staining. Specificity of staining was tested using butyrylthiocholine as substrate or with the AChE inhibitor 1,5 bis(4-allyldimethylammoniumphenyl)pentan-3-one dibromide (BW284c51; Sigma, St Louis, MO), as previously described: lmmunocytochemistry For TNR immunocytochemistry following AChE staining, the stained cells were rinsed with 50 mM Tris HCI buffer (pH 7.6) and PBS. After incubation with 0.1% Triton X-100 (15 min, room temperature), the cells were incubated overnight at 4°C with a 1: 200 dilution (PBS) of 2B3 mab. After incubation, the cells were washed (PBS) and incubated at room temperature for 90 min with a 1: 50 dilution (PBS) of a fluorescein isothiocyanate (FITC)-labeled goat antimouse IgG (purchased from Boehringer, Mannheim, F.R.G.). After rinsing with PBS and water, the cells were mounted and viewed under a Leitz fluorescence microscope
fitted with epi-illuminescence and the appropriate filters, and under phase contrast. For neurofilament (NF), glial fibrillary protein (GFAP) and galactocerebroside (GALC) immunochemistry, antisera were purchased or obtained from the following sources: anti-NF (mol. wt 200,000) and anti-GFAP (monoclonal) antibodies were purchased from Boehringer, Mannheim, anti-NF (mol. wt 70,000) and anti-GALC (polyclonal) antibodies were kindly provided by Dr M. Vincent, C.H.U.L., Ste-Foy, Qurbec and Dr Joyce Benjamins, Wayne State University School of Medicine, Detroit, Michigan. Double immunostaining Cells were fixed with 3% paraformaldehyde (in PBS), washed thoroughly with PBS, incubated with Triton (see above) and then incubated (12 h, 4°C) simultaneously with either anti-NF (mol. wt 200,000) and anti-GALC (1:100, PBS) or anti-NF (mol. wt 70,000) and anti-GFAP (1:100, PBS). After incubation, the cells were washed (PBS) and then simultaneously incubated (90 min, room temperature) with 1:50 dilutions (PBS) of tetramethylrhodamine isothiocyanate (TRITC)-labeled sheep anti-rabbit IgG (Boehringer) and FITC-labeled goat anti-mouse IgG. The cells were washed, mounted and viewed as described for TNR immunocytochemistry. Specificity controls were carried out by replacing antisera with non-immune mouse or rabbit sera, non-immune mouse ascites fluid and with 2B3 mab adsorbed with partially purified TNR. In these experiments, 2B3 mab (1:200) was incubated overnight with 100/~g of partially purified TNR. The mixture was then used for immunocytochemistry. Morphometric analysis Morphometric analysis was performed on acetylcholinesterase-positive (ACHE +) cells. All AChE + cells present in 12 mm diameter glass coverslips representing controls and L-T:treated cells were counted after 3, 5, 8, 15 and 21 days in vitro. The percentage of AChE + cells that were immunoreactive for TNR (TNR + ) was then determined. To obtain an index of neurite density and branching in AChE + cells, neurite density measurements were performed on 48-50 microphotographs of AChE + cells representing three different experiments per group. An AChE + cell was placed at the center of the field. Photographs were then taken with a Leitz photomicroscope (25 objective) and enlarged to a final magnification of × 640. A square grid was then superimposed on a photograph and the neurite density was determined by counting all the neurites starting from the soma and intersecting with the 1-cm squares of the grid. For statistical analysis, the data are presented as the mean + S.E.M. The significance of differences between data groups was assessed by Student's t-test. RESULTS Morphological evolution o f the cultures The morphological d e v e l o p m e n t of fetal rat cerebral hemisphere cultures grown in C D M is similar to those described for a n e u r o b l a s t o m a cell line a n d rat h y p o t h a l a m i c cell cultures. 1'25The cells rapidly attach to the pretreated glass coverslips a n d begin to extend neurites after a few h o u r s in culture. On day 8, a r o u n d 9 0 % o f the cells were neurofilament-positive (Fig. la). A small percentage o f the cells (approximately 10%) were G F A P - p o s i t i v e (Fig. lb), while n o n e were stained by a n anti-galactocerebroside antibody (data not shown).
Thyroid hormone receptors in acetylcholinesterase-positive neurons
Fig. 1. Double immunolabeling (same field) of NF + cells ([a]: sheep anti-rabbit TRITC) and GFAP + cells ([b]: goat anti-mouse FITC) after 8 days in t, itro (x 640). NSC 36,2--G
475
476
R. GARZA et al.
ill!
Fig. 2. Morphological effect of L-T 3 on T N R + A C h E + cells visualized by A C h E cytochemistry and T N R immunocytochemistry after 8 days in vitro. (a) L-T3-treated culture: immunofluorescence. (b) Same as (a): phase contrast. (c) Control culture: immunofluorescence. (d) Same as (c): phase contrast. Arrow indicates A C h E + cells with T N R ( x 320).
Thyroid hormone receptors in acetylcholinesterase-positive neurons
477
Table 1. Distribution of thyroid hormone receptors in cerebral hemisphere acetylcholinesterase-positive neurons Days in culture Control L-T 3 Control L-T 3 Control L-T 3 Control L-T 3 Control L-T 3
3 5 8 15 21
Number of AChE + cells (total)
Number of AChE + cells (TNR +)
Number of AChE + cells (TNR)
% AChE + cells (TNR +)
37 + 4 44 + 4 103 + 15 98 __+10 146_ 18 149 _ 14 135 + 20 I17+15 129 __+19 110+ 13
13 + 3 12 + 2 37 + 7 36 + 9 55 + 11 56 + 8 35 _+ 13 46+15 36 + 12 47+ 17
74 78 74 73 73 73 79 72 78 70
50 + 7 56 + 6t 140 + 22 134 + 19 201 + 29 205 + 22 170 _+ 11 163+8 165 + 13"* 157+10"*
All the AChE + cells in control and L-T3-treated (3 x I0 s M) 12 mm diameter coverslips were counted. Each value represents the mean + S.E.M. of three independent experiments, each performed in duplicate. Statistical analyses were performed using Student's t-test. **The cell number at day 21 is significantly different from day 3, P < 0.001. tAt all ages none of the L-T~-treated values were significantly different from controls.
Visualization of aceo,lcholinesterase-positive neurons in culture The morphological development of A C h E + neurons identified by A C h E cytochemistry have been previously described in detail. 4 Briefly, after 8 days in vitro, A C h E + neurons exhibited great heterogeneity in their form (i.e. round, ovid, pyramidal) and size. Furthermore, a small percentage ( < 10%) of fusiform cells were also observed. The A C h E + neurons represented a small percentage of neuronal cells ( < 0.1%). With time in culture, the number of A C h E + neurons increased from 50 _+ 7 on day 3 to 165 _+ 13 on day 21 (Table 1). No positive cells were observed using butyrylthiocholine as substrate or in the presence of BW284c51.
lmmunocytochemical localization of L-tri-iodothyronine nuclear receptors After 8 days in vitro, we estimated that approximately 70-80% of the neuronal cells appeared labeled with 2B3 mab (Fig. 2a,c). However, we were unable to accurately determine the number of N F + cells stained by 2B3 mab nor perform morphometric analyses on these cells, since the cell density required to have enough A C h E + cells was too high. The staining was strictly limited to the nucleus. After 3 days in vitro, 74% of A C h E + cells were immunoreactive with 2B3 mab while 26% were not stained with the antibody (Table 1). This ratio remained unchanged despite an increase in the number of A C h E + cells with time (from day 3 to day 21) in culture. Furthermore, the addition of 3 x 10-SM L-T 3 in culture did not affect the proportion of A C h E + cells also expressing T N R to those not stained by 2B3 mab or have an effect on the number of A C h E + cells, at least until day 21 in vitro (Table 1). No staining was observed with 2B3 mab with normal mouse serum or control ascites fluid. Furthermore,
the staining disappeared when the primary antibody was diluted 1 : 800 and when the cells were incubated with a mixture of primary antibody and partially purified nuclear receptor (data not shown).
Effect of L-tri-iodothyronine on L-tri-iodothyronine nuclear receptor-imrnunoreactive acetylcholinesterasepositive cells Eight days of L-T 3 treatment of the cultures resulted in an increase in size and neurite arborization of T N R + A C h E + cells over those T N R + A C h E + cells grown in the absence of the hormone (Fig. 2: compare a,b with c,d). Conversely, no change in neurite arborization was observed between L-T3-treated and control T N R A C h E + cells (Fig. 3: compare a,b with c,d). Neurite density measurements performed after 8 days in vitro in these two subpopulations of A C h E + neurons showed a >4-fold increase in the neurite index of the T N R + fraction of A C h E + cells when grown in the presence of L-T 3 over T N R + A C h E + cells grown in the absence of L-T3 (Table 2). In contrast, no significant difference in the neurite index of L-T3-treated T N R A C h E + cells was observed over T N R - A C h E + cells not treated with the hormone (Table 2).
Effect of L-tri-iodothyronine on L-tri-iodothyronine nuclear receptor immunoreactive acetylcholinesterasepositive cells after 20 days in vitro To determine a possible critical period of L-T3 action in vitro, the same fetal brain cultures were grown in the absence of L-T 3 (control) for 28 days, in the presence of L-T 3 for 8 days starting on culture day 20 and in the presence of L-T3 for the entire 28-day culture period. A C h E cytochemistry and T N R immunocytochemistry were performed after 28 days in vitro. When viewed under phase contrast, chronically L-T3-treated (those cells treated with L-T 3
478
R. GARZA et al.
k
(o) Fig. 3. Morphological effect of L-T 3 on T N R - A C h E + cells visualized by AChE cytochemistry and T N R immunocytochemistry after 8 days in r'itro. (a) L-T3-treated culture: immunofluorescence. (b) Same as (a): phase contrast. (c) Control culture: immunofluorescence. (d) Same as (c): phase contrast. Arrow indicates AChE + cells with TNR ( x 320).
Thyroid hormone receptors in acetylcholinesterase-positive neurons
479
Table 2. Effect of L-tri-iodothyronine on {~'~neiiri(edehsity of L-tri-iodothyronine nuclear receptor-immunoreactive and L-tri-iodothyronine nuclear receptor-negative acetylcholinesterase-positive cells after 8 days in vitro AChE + cells (TNR + ) Control L-T 3 (3 x 10 -s M) Neurite index
21.4 + 1.9
91.1 + 6.2*
AChE + cells (TNR-) Control L-T3 (3 × 10-8 M) 16.9 ___1.6
15.3 _+ 1.5
The values represent the number of neurite intersections falling in a 1 cm 2 area. Each value represents the mean of three different experiments performed in duplicate ___S.E.M. Statistical analyses were performed using Student's t-test. TNR +, TNR-immunoreactive; TNR-, TNR-negative. *Significantly different from control, P < 0.05.
for the entire culture period) 28-day-old A C h E + cells appeared larger and exhibited a greater degree of neurite arborization in contrast to 28-day-old A C h E + cells where the L-T3 treatment started on culture day 20 for 8 days or in 28-day-old control cultures (compare Fig. 4a with Fig. 4b). Indeed, neurite density measurements after 28 days in vitro showed a > 2 - f o l d increase in the neurite index of 28-day-old chronically L-T3-treated A C h E + cells over those 28-day-old A C h E + cells where the addition of the hormone was started on culture day 20 or in 28-day-old control cultures (Table 3). Cell counts to determine the relative percentages of T N R + A C h E + cells to T N R - A C h E + cells after 28 days in vitro were not possible due to the degree of morphological development at this stage of the cultures. However, all A C h E + cells taken for neurite density measurements were T N R +.
DISCUSSION In a previous in vitro study, 4 we demonstrated the sensitivity of central A C h E + neurons to L-T3 with an effect on the size of the perikarya and the neurite length. Table 3. Effect of the duration of L-tri-iodothyronine treatment at different periods of culture on the neurite density of acetylcholinesterase-positive cells L-T 3 (3 × 10 -8 M) treatment started (culture day) 0 20 1
Duration of L-T 3 treatment (days)
Neurite index after 28 days
0 8 28
76+9 75 + 8 182 + 20**
in vitro
Cells were grown in the absence of L-T3 for 28 days (no L-T 3 added at any time), in the presence of L-T 3 for 8 days starting on culture day 20 and in the presence of L-T 3 for 28 days. AChE staining was performed after 28 days in vitro and the neurite density determined. The values represent the number of neurite intersections falling in a 1 cm 2 area. Each value represents the mean of three different experiments performed in duplicate + S.E.M. Statistical analyses were performed using Student's ttest. **Significantly different from control and 8-day L-T 3treated cultures, P < 0.001.
The present study was undertaken to determine if L-T 3 nuclear receptors were present in A C h E + neurons and whether a relationship existed between the presence of T N R and the effects of L-T3 on the morphological development of this neuronal population. Here we show that 74% of fetal A C h E + neurons were immunoreactive for T N R , while 26% did not stain with 2B3 mab. This may suggest that some A C h E + neurons are not target cells for thyroid hormones or that, in our culture conditions, some A C h E + neurons do not express T N R or have lost T N R expression, depending on their degree of maturation. However, this latter hypothesis seems improbable since the percentage of A C h E + cells not stained by 2B3 mab remains the same with time in culture, at least until day 21. A cellular heterogeneity in the distribution of T N R in fetal neuronal cells is supported by several in vivo and in vitro data. The influence of neonatal dysthyroidism on the development of CHAT, tyrosine hydroxylase (TH) and glutamate decarboxylase ( G A D ) activities vary greatly depending on the brain area studied. 3"1°'21"23"26Furthermore, it has been reported that the developing cholinergic basal forebrain neurons but not the developing cholinergic pontomesencephalotegmental neurons are sensitive to thyroid hormone. 6 Similar results have been obtained in brain cell cultures where it has been shown that L-T 3 affects the development of mesencephalic and hypothalamic dopaminergic neurons differently. 22'24 These studies suggest that some suhpopulations of neurons are not responsive to thyroid hormones during brain development. This hypothesis is also supported by data showing the regional heterogeneity of T N R immunoreactivity in the adult rat brain. Comparing the regional distribution of A C h E + neurons H with the distribution of T N R immunoreactivity m the adult rat brain (Puymirat et al., unpublished observations), it appears that some brain areas containing A C h E + cells are weakly T N R immunoreactive, such as the caudate-putamen and the pallidum. It is therefore possible that some A C h E + cells located in these brain areas are not target cells for thyroid hormone. A double staining study should be performed to determine the target A C h E + cells for thyroid hormone in
(a) ii
tl
B
.
(b)
Q
IP
4
q
Fig. 4. Morphological effect of L-T 3 on AChE + cells after 28 days in t, itro. (a) AChE + cell treated with L-T 3 for 8 days starting on culture day 20. (b) AChE + cell treated with L-T 3 throughout the 28-day culture period ( × 640).
Thyroid hormone receptors in acetylcholinesterase-positiveneurons the rat brain. Additionally, our study shows a correlation between the presence of TNR immunoreactivity and the effect of L - T 3 o n the neurite density of AChE + neurons, which suggests that this effect may be directly mediated by an interaction of L-T3 with specific nuclear receptors. The direct effect of L-T3 on neurite elongation and branching is also supported by our recent finding showing that conditioned media prepared from astrocytes grown in the presence L - T 3 are unable to produce an increase in neurite density when compared with conditioned media obtained from astrocytes grown in the absence of L-T 3 (Garza et al., unpublished observations). To support this hypothesis, it remains to be determined whether the expression of TNR is associated with the morphological effect of L - T 3. However, this cannot be performed in our culture conditions since 80% of the cell nuclei are TNR + after 1 day in culture (data not shown). The failure of LoT3 to elicit a morphological response if added on culture day 20 suggests the existence of a critical period of L - T 3 action which coincides with in vitro studies where ChAT activity is stimulated only when the hormonal treatment is started before culture day 20. 4'1°This is in agreement with in vivo data showing that L - T 3 treatment of neonatal hypothyroid rats, beginning after the 14th postnatal day (which corresponds to our culture day 20), does not reverse the deficiency produced by hypothyroidism. 3'8The apparent discrepancy between
481
the absence of the morphological response to thyroid hormone and the presence of TNR immunoreactivity in 28-day-old AChE + cells may be explained by the following: our mab may recognize different forms of the TNR. Indeed, recent data ~7 have demonstrated not only the TNR, which binds L-T3, but also two variant forms which do not bind L-T 3. It is therefore possible that our mab recognizes both forms of the TNR, which may explain the absence of a morphological effect of L-T 3 despite the presence of TNR immunoreactivity. An alternative explanation is that our mab recognizes the active form of the TNR, but the gene(s) regulated by L - T 3 have lost their sensitivity to thyroid hormones after 20 days in vitro. CONCLUSION
This in vitro study showed that (1) fetal brain AChE + neurons are target cells for thyroid hormones. Since subpopulations of AChE + cells present in our cultures are also ChAT-positive,9 our results may suggest that some subpopulations of cholinergic neurons are target cells for thyroid hormone. (2) The presence of TNR in a subpopulation of AChE + neurons associated with L-T3 morphological effects suggests that some effects of L-T3 on the maturation of AChE + neurons may result from a direct effect of L - T 3 through an interaction with its specific nuclear receptors.
REFERENCES
1. Bottenstein J. E. and Sato G. H. (1979) Growth o f a neuroblastoma cell line in serum-free supplemented medium. Proc. natn. Acad. Sci. U.S.A. 76, 514-517.
2. Faure R., Ruel J. and Dussault J. H. (1986) Production of an antibody against rat liver nuclear T 3 receptor. Biochem. Cell Biol. 64, 377-380. 3. Garcia Argiz C. A., Pasquini J. M., Kaplan B. and Gomez C. J. (1967) Hormonal regulation of brain development. II. Effect of neonatal thyroidectomy on succinate dehydrogenase and other enzymes in developing cerebral cortex and cerebellum of the rat. Brain Res. 6, 635 646. 4. Garza R., Dussault J. H. and Puymirat J. (1988) Influence of tr/iodothyronine (L-T3) on the morphological and biochemical development of fetal brain acetylcholinesterase-positiveneurons cultured in a chemicallydefined medium. Devl Brain Res. 43, 287-297. 5. Geel S. E. and Timiras P. S. (1967) Influence of neonatal hypothyroidism and of thyroxine on the acetylcholinesterase and cholinesterase activities in the developing central nervous system of the rat. Endocrinology 80, 1069-1074. 6. Gould E. and Butcher L. L. (1989) Developing cholinergic basal forebrain neurons are sensitive to thyroid hormone. J. Neurosci. 9, 3347-3358. 7. Grave G. D. (1977) Thyroid Hormones and Brain Development. Raven Press, New York. 8. Hamburg M. and Flexner L. B. (1957) Biochemicaland physiologicaldifferentiation during morphogenesis. XXI. Effect of hypothyroidism and hormone therapy on enzymeactivities of the developingcerebral cortex of the rat. J. Neurochem. 1, 273-288. 9. Hefti F., Hartikka J., Eckenstein H., Gnaln H., Heumann R. and Schwab M. (1985) Nerve growth factor increases choline acetyltransferase but not survival or fiber outgrowth of cultured fetal septal cholinergic neurons. Neuroscience 14, 55-68. 10. Honegger P. and Lenoir D. (1980) Triiodothyronine enhancement of neuronal differentiation in aggregating fetal rat brain cells cultured in a chemically defined medium. Brain Res. 199, 425-434. 1l. Jacobowitz D. M. and Palkovits M. (1974) Topographic atlas of catecholamine and acetylcholinesterase-containing neurons in the rat brain. I. Forebrain (telencephalon, diencephalon). J. comp. Neurol. 57, 13. 12. Kolodny J. M., Leonard J. L., Larsen P. R. and Silva J. (1985) Studies of nuclear 3,5,Y-triiodothyronine binding in primary cultures of rat brain. Endocrinology 117, 1848-1853. 13. Luo M., Faure R. and Dussault J. H. (1986) Ontogenesis of nuclear T~ receptors in primary cultured astrocytes and neurons. Brain Res. 381, 275-280. 14. Luo M., Faure R., Ruel J. and Dussault J. H. (1988) A monoclonal antibody to the rat nuclear triiodothyronine receptor: production and characterization. Endocrinology 123, 180-186. 15. Luo M., Puymirat J. and Dussault J. H. (1989) Immunocytochemical localization of nuclear 3,5,3'-triiodothyronine (L-T3) receptors in astrocyte cultures. Devl Brain Res. 46, 131 136.
482
R. GARZAet al.
16. Luo M., Faure R., Tong Y. A. and Dussault J. H. (1989) Immunocytochemicat localization of the nuclear 3,5,Y-triiodothyronine receptor in the adult rat: liver, kidney, heart, lung and spleen. Acta endocr. Copenh. 120, 451-458. 17. Mitsuhashi T., Tennyson G. E. and Nikodem V. M. (1988) Alternative splicing generates messages encoding rat c-erb A proteins that do not bind thyroid hormone. Proc. natn. Acad. Sci. U.S.A. 85, 5804-5808. 18. Oppenheimer J. H. (1979) Thyroid hormone action at the cellular level. Science 203, 971-979. 19. Oppenheimer J. H., Schwatz H. L., Mariash C N., Kinlan W. E., Wong N. C. W. and Freake H. C. (1987) Advances in our understanding of thyroid hormone action at the cellular level. J. Endocr. Invest. 8, 288-308. 20. Pascual A., Aranda A., Ferret-Senat V., Gabellec M. M., Rebel G. and Sarlieve L. L. (1986) Triiodothyronine receptors in developing mouse neuronal and glial cell cultures and in chick-cultured neurons and astrocytes. Devl Brain Res. 8, 89-101. 21. Patel A. J., Hayashi M. and Hunt A. (1987) Selective persistent reduction in choline acetyltransferase activity in basal forebrain of the rat after thyroid deficiency during early life. Brain Res. 422, 182-185. 22. Puymirat J., Barret A., Picart R., Vigny A., Loudes C., Faivre-Bauman A. and Tixier-Vidal A. (1983) Triiodothyronine enhances the morphological maturation of dopaminergic neurons from fetal mouse hypothalamus cultured in serum-free medium. Neuroscience 10, 801-810. 23. Puymirat J. (1985) Effects of dysthyroidism on central catecholaminergic neurons. Neurochem. Int. 7, 969-977. 24. Puymirat J., Faivre-Bauman A., Barret A., Loudes C. and Tixier-Vidal A. (1985) Does triioodothyronin influence the morphogenesis of fetal mouse mesencephalic dopaminergic neurons cultured in chemically defined medium? Devl Brain Res. 23, 315-317. 25. Puymirat J., Luo M. and Dussault J. H. (1988) Immunocytochemical localization of thyroid nuclear receptors in cultured hypothalamic dopaminergic neurons. Neuroscience 30, 443-449. 26. Rastogi R. B. and Singhal R. L. (1976) Influence of neonatal and adult hyperthyroidism on behavior and biosynthetic capacity for norepinephrine, dopamine and 5-hydroxytryptamine in rat brain. J. Pharmac. exp. Ther. 198, 609-618. 27. Ruel J., Faure R. and Dussault J. H. (1985) Regional distribution of nuclear T 3 receptors in rat brain and evidence for preferential localization in neurons. J. Endocr. Invest. 8, 343-348. 281 Samuels H. H., Stanley F. and Casanova J. (1979) Depletion of L-3,5,Y-triiodothyronine and L-thyroxine in euthyroid calf serum for use in cell culture studies of the action of thyroid hormone. Endocrinology 105, 80-85. 29. Tago H., Kimura H. and Maeda T. (1986) Visualization of detailed acetylcholinesterase fiber and neuron staining in rat brain by a sensitive histochemical procedure. J. Histochem. Cytochem. 34, 1431-1438. 30. Valcana T. (1971) Effect of neonatal hypothyroidism on the development of acetylcholinesterase and choline acetyltransferase activity in rat brain. In Influence of Hormones on the Nervous System (ed. Ford D. H.), pp. 174-184. S. Karger, Basel. (Accepted 17 January 1990)
0306-4522/90 $3.00 + 0.00 Pergamon Press plc © 1990 IBRO
Printed in Great Britain
IMMUNOCYTOCHEMICAL LOCALIZATION OF THYROID HORMONE NUCLEAR RECEPTORS IN CULTURED ACETYLCHOLINESTERASE-POSITIVE NEURONS: A CORRELATION BETWEEN THE PRESENCE OF THYROID HORMONE NUCLEAR RECEPTORS AND L-TRI-IODOTHYRONINE MORPHOLOGICAL EFFECTS R. GARZA,* J. PUYMIRATand J. H. DUSSAULT Unit6 de recherche en Ontogrnrse et Grnrtique molrculaire, Centre Hospitalier de l'Universit6 Laval, Qurbec, Canada GlV 4G2 Almtraet--A monoclonal antibody against the rat liver L-tri-iodothyronine nuclear receptor and acetylcholinesterase cytochemistry were used for the localization of thyroid hormone nuclear receptors in acetylcholinesterase-positive cell nuclei in fetal rat cerebral hemisphere neuronal cultures. After 3 days in vitro, the ratio of acetylcholinesterase-positive cells that were immunoreactive for the thyroid hormone nuclear receptor to those not stained for this receptor (74-26%, respectively) remains unchanged despite an increase in the number of acetylcholinesterase-positive cells with time (from day 3 to day 21) in culture. Furthermore, the addition of 3 x l0 -8 L-tri-iodothyronine in culture did not modify this ratio or have an effect on the number of acetylcholinesterase-positive cells, but significantly increased the neurite density in those acetylcholinesterase-positive cells that were immunoreactive for the thyroid hormone receptor. Conversely, no difference in the neurite densities of those acetylcholinesterase-positive cells not stained for this receptor was observed when cultured in the presence or absence of thyroid hormone. In other experiments with the same fetal brain cultures, treatment of cultures for 8 days with L-tri-iodothyronine, beginning on culture day 20, demonstrated the presence of a critical period which occurs in vitro around day 20, since the stimulatory effect of L-tri-iodothyronine on immunoreactive acetylcholinesterase-positive cell neurite density is lost after 20 days in vitro. These results demonstrate, for the first time, the presence of L-tri-iodothyronine nuclear receptors in fetal rat acetylcholinesterase-positive neurons and the existence of a cellular heterogeneity in the distribution of the thyroid hormone receptor. The presence of these receptors in fetal brain acetylcholinesterase-positive neurons suggests that some effects of L-tri-iodothyronine on the maturation of a subpopulation of acetylcholinesterase-positive neurons may result from a direct effect of this hormone through an interaction with its specific nuclear receptors.
The presence of thyroid hormones is essential for the normal development of the brain and these hormones have been shown to influence neuronal development. 7 Several findings indicate that thyroid hormones affect the development of acetylcholinesterasepositive (ACHE + ) central neurons. First, neonatal hypothyroidism is associated with a reduction of A C h E activity in various brain areas of experimental animals. 5'3° Second, L-tri-iodothyronine (L-T3) has been found to increase A C h E activity as well as enhance the morphological development of A C h E + neurons in aggregating and primary cultures of fetal *To whom correspondence should be addressed. ACHE, acetylcholinesterase; CDM, chemically defined medium; CHAT, choline acetyltransferase; FCS, fetal calf serum; FITC, fluorescein isothiocyanate; GAD, glutamate decarboxylase; GALC, galactocerebroside; GFAP, glial fibrillary acid protein; L-T 3, L-tri-iodothyronine; mab, monoclonal antibody; NF, neurofilaments; PBS, phosphate-buffered saline; TH, tyrosine hydroxylase; TNR, L-tri-iodothyronine nuclear receptor; TRITC, tetramethylrhodamine isothiocyanate.
Abbreviations:
brain cells. 4,~° These findings indicate that L-T 3 acts directly on brain cells; however, the molecular mechanisms of thyroid hormone action on neurons remain to be clarified. It is generally accepted that thyroid hormone action is initiated by the binding of L-T 3 to an L-tri-iodothyronine nuclear receptor ( T N R ) ) 8'19 Previous studies using Scatchard analysis revealed the presence of L-T 3 binding sites in neuron-enriched fractions in different areas of the adult brain as well as in primary cultures of neurons and astrocytes. 12A3'15'2°'25'27 These studies, however, d i d not show whether L-T 3 binding sites are located in A C h E + neurons. Our recent development of a monoclonal antibody (2B3 mab) against partially purified L-T3 rat liver nuclear receptor ~4 has made it possible to study the localization of T N R at the cellular level. In this study, 2B3 mab and A C h E cytochemistry were used to localize T N R in A C h E + neurons and to determine if the presence of T N R in this neuronal population was correlated with L-T 3 morphological effects.
473
R. GARZA et al.
474 EXPERIMENTAL PROCEDURES
Cerebral hemisphere neuronal cultures Primary cultures of dissociated cerebral hemisphere cells from 15- to 16-day-old Sprague-Dawley rat fetuses (Charles River, St Constant, Qurbec) were prepared as previously described. 4 After mechanical dissociation in serumsupplemented medium, the cells were centrifuged and resuspended in chemically defined medium (CDM). An aliquot of the cell suspension containing 2 x 105 cells (98,000 cells/cm2) was plated in 0.5ml of CDM on 16mm diameter tissue culture wells containing glass coverslips that had been treated previously with gelatin (250gg/ml), incubated overnight with poly-L-lysine (10/~g/ml in phosphatebuffered saline, PBS), rinsed with PBS and preincubated with 10% fetal calf serum (FCS) stripped of thyroid hormones according to the procedure of Samuels et al. 28 The composition of the CDM was that previously described for similar cultures: Cultures were maintained in a humidified atmosphere of 5% CO2, 95% air at 37°C. The medium was renewed 3 days after seeding and every third day thereafter. In experiments with thyroid hormone, L-T 3 was added at the initiation (except when studying the duration of L-T 3 treatment after 20 days in vitro) of the culture and renewed at every change of the medium. Based on previous dose-response studies,4 L-T 3 was used at a final concentration of 3 x 10-8 M. Preparation and specificity 2B3 monoclonal antibody against rat liver nuclear receptor TNR was purified from liver cytosol as previously described. 2 The preparation and characterization of the monoclonal anti-TNR have been described in detail, t4't6 In the present study, the specificity of the 2B3 mab was determined according to the following criteria: (1) cells incubated with control ascitic fluid or with 2B3 mab pre-adsorbed with purified receptor showed no specific staining; (2) in accordance with previous immunocytochemical localization of TNR in different tissues of the adult rat and in primary cultures of neurons and astrocytes, TNR immunoreactivity was strictly limited to the nucleus, t5:6"2~ Acetylcholinesterase cytochemistry AChE + neurons were identified in dissociated cell cultures by AChE cytochemistry according to the procedure described by Tago et al. 29 Briefly, cultures were allowed to grow for 3, 5, 8, 21 and 28 days, fixed with 3% paraformaldehyde (in PBS), rinsed thoroughly with PBS (pH 7.4) and rinsed twice in 0.I M maleate buffer (pH 6.0). The cells were then incubated for 30 min with a substrate solution containing 72 p M acetylthiocholine iodine, 10 p M K2Fe(CN)6, 6 0 # M CuSO 4 and 100#M sodium citrate. After removal of the substrate solution, the cells were rinsed and incubated with a mixture of 0.4% 3',Y-diaminobenzidine and 0.3% nickel ammonium sulfate for 10 min. Then 0.003% H202 was added and the cells were further incubated for 25 min to reveal optimal staining. Specificity of staining was tested using butyrylthiocholine as substrate or with the AChE inhibitor 1,5 bis(4-allyldimethylammoniumphenyl)pentan-3-one dibromide (BW284c51; Sigma, St Louis, MO), as previously described: lmmunocytochemistry For TNR immunocytochemistry following AChE staining, the stained cells were rinsed with 50 mM Tris HCI buffer (pH 7.6) and PBS. After incubation with 0.1% Triton X-100 (15 min, room temperature), the cells were incubated overnight at 4°C with a 1: 200 dilution (PBS) of 2B3 mab. After incubation, the cells were washed (PBS) and incubated at room temperature for 90 min with a 1: 50 dilution (PBS) of a fluorescein isothiocyanate (FITC)-labeled goat antimouse IgG (purchased from Boehringer, Mannheim, F.R.G.). After rinsing with PBS and water, the cells were mounted and viewed under a Leitz fluorescence microscope
fitted with epi-illuminescence and the appropriate filters, and under phase contrast. For neurofilament (NF), glial fibrillary protein (GFAP) and galactocerebroside (GALC) immunochemistry, antisera were purchased or obtained from the following sources: anti-NF (mol. wt 200,000) and anti-GFAP (monoclonal) antibodies were purchased from Boehringer, Mannheim, anti-NF (mol. wt 70,000) and anti-GALC (polyclonal) antibodies were kindly provided by Dr M. Vincent, C.H.U.L., Ste-Foy, Qurbec and Dr Joyce Benjamins, Wayne State University School of Medicine, Detroit, Michigan. Double immunostaining Cells were fixed with 3% paraformaldehyde (in PBS), washed thoroughly with PBS, incubated with Triton (see above) and then incubated (12 h, 4°C) simultaneously with either anti-NF (mol. wt 200,000) and anti-GALC (1:100, PBS) or anti-NF (mol. wt 70,000) and anti-GFAP (1:100, PBS). After incubation, the cells were washed (PBS) and then simultaneously incubated (90 min, room temperature) with 1:50 dilutions (PBS) of tetramethylrhodamine isothiocyanate (TRITC)-labeled sheep anti-rabbit IgG (Boehringer) and FITC-labeled goat anti-mouse IgG. The cells were washed, mounted and viewed as described for TNR immunocytochemistry. Specificity controls were carried out by replacing antisera with non-immune mouse or rabbit sera, non-immune mouse ascites fluid and with 2B3 mab adsorbed with partially purified TNR. In these experiments, 2B3 mab (1:200) was incubated overnight with 100/~g of partially purified TNR. The mixture was then used for immunocytochemistry. Morphometric analysis Morphometric analysis was performed on acetylcholinesterase-positive (ACHE +) cells. All AChE + cells present in 12 mm diameter glass coverslips representing controls and L-T:treated cells were counted after 3, 5, 8, 15 and 21 days in vitro. The percentage of AChE + cells that were immunoreactive for TNR (TNR + ) was then determined. To obtain an index of neurite density and branching in AChE + cells, neurite density measurements were performed on 48-50 microphotographs of AChE + cells representing three different experiments per group. An AChE + cell was placed at the center of the field. Photographs were then taken with a Leitz photomicroscope (25 objective) and enlarged to a final magnification of × 640. A square grid was then superimposed on a photograph and the neurite density was determined by counting all the neurites starting from the soma and intersecting with the 1-cm squares of the grid. For statistical analysis, the data are presented as the mean + S.E.M. The significance of differences between data groups was assessed by Student's t-test. RESULTS Morphological evolution o f the cultures The morphological d e v e l o p m e n t of fetal rat cerebral hemisphere cultures grown in C D M is similar to those described for a n e u r o b l a s t o m a cell line a n d rat h y p o t h a l a m i c cell cultures. 1'25The cells rapidly attach to the pretreated glass coverslips a n d begin to extend neurites after a few h o u r s in culture. On day 8, a r o u n d 9 0 % o f the cells were neurofilament-positive (Fig. la). A small percentage o f the cells (approximately 10%) were G F A P - p o s i t i v e (Fig. lb), while n o n e were stained by a n anti-galactocerebroside antibody (data not shown).
Thyroid hormone receptors in acetylcholinesterase-positive neurons
Fig. 1. Double immunolabeling (same field) of NF + cells ([a]: sheep anti-rabbit TRITC) and GFAP + cells ([b]: goat anti-mouse FITC) after 8 days in t, itro (x 640). NSC 36,2--G
475
476
R. GARZA et al.
ill!
Fig. 2. Morphological effect of L-T 3 on T N R + A C h E + cells visualized by A C h E cytochemistry and T N R immunocytochemistry after 8 days in vitro. (a) L-T3-treated culture: immunofluorescence. (b) Same as (a): phase contrast. (c) Control culture: immunofluorescence. (d) Same as (c): phase contrast. Arrow indicates A C h E + cells with T N R ( x 320).
Thyroid hormone receptors in acetylcholinesterase-positive neurons
477
Table 1. Distribution of thyroid hormone receptors in cerebral hemisphere acetylcholinesterase-positive neurons Days in culture Control L-T 3 Control L-T 3 Control L-T 3 Control L-T 3 Control L-T 3
3 5 8 15 21
Number of AChE + cells (total)
Number of AChE + cells (TNR +)
Number of AChE + cells (TNR)
% AChE + cells (TNR +)
37 + 4 44 + 4 103 + 15 98 __+10 146_ 18 149 _ 14 135 + 20 I17+15 129 __+19 110+ 13
13 + 3 12 + 2 37 + 7 36 + 9 55 + 11 56 + 8 35 _+ 13 46+15 36 + 12 47+ 17
74 78 74 73 73 73 79 72 78 70
50 + 7 56 + 6t 140 + 22 134 + 19 201 + 29 205 + 22 170 _+ 11 163+8 165 + 13"* 157+10"*
All the AChE + cells in control and L-T3-treated (3 x I0 s M) 12 mm diameter coverslips were counted. Each value represents the mean + S.E.M. of three independent experiments, each performed in duplicate. Statistical analyses were performed using Student's t-test. **The cell number at day 21 is significantly different from day 3, P < 0.001. tAt all ages none of the L-T~-treated values were significantly different from controls.
Visualization of aceo,lcholinesterase-positive neurons in culture The morphological development of A C h E + neurons identified by A C h E cytochemistry have been previously described in detail. 4 Briefly, after 8 days in vitro, A C h E + neurons exhibited great heterogeneity in their form (i.e. round, ovid, pyramidal) and size. Furthermore, a small percentage ( < 10%) of fusiform cells were also observed. The A C h E + neurons represented a small percentage of neuronal cells ( < 0.1%). With time in culture, the number of A C h E + neurons increased from 50 _+ 7 on day 3 to 165 _+ 13 on day 21 (Table 1). No positive cells were observed using butyrylthiocholine as substrate or in the presence of BW284c51.
lmmunocytochemical localization of L-tri-iodothyronine nuclear receptors After 8 days in vitro, we estimated that approximately 70-80% of the neuronal cells appeared labeled with 2B3 mab (Fig. 2a,c). However, we were unable to accurately determine the number of N F + cells stained by 2B3 mab nor perform morphometric analyses on these cells, since the cell density required to have enough A C h E + cells was too high. The staining was strictly limited to the nucleus. After 3 days in vitro, 74% of A C h E + cells were immunoreactive with 2B3 mab while 26% were not stained with the antibody (Table 1). This ratio remained unchanged despite an increase in the number of A C h E + cells with time (from day 3 to day 21) in culture. Furthermore, the addition of 3 x 10-SM L-T 3 in culture did not affect the proportion of A C h E + cells also expressing T N R to those not stained by 2B3 mab or have an effect on the number of A C h E + cells, at least until day 21 in vitro (Table 1). No staining was observed with 2B3 mab with normal mouse serum or control ascites fluid. Furthermore,
the staining disappeared when the primary antibody was diluted 1 : 800 and when the cells were incubated with a mixture of primary antibody and partially purified nuclear receptor (data not shown).
Effect of L-tri-iodothyronine on L-tri-iodothyronine nuclear receptor-imrnunoreactive acetylcholinesterasepositive cells Eight days of L-T 3 treatment of the cultures resulted in an increase in size and neurite arborization of T N R + A C h E + cells over those T N R + A C h E + cells grown in the absence of the hormone (Fig. 2: compare a,b with c,d). Conversely, no change in neurite arborization was observed between L-T3-treated and control T N R A C h E + cells (Fig. 3: compare a,b with c,d). Neurite density measurements performed after 8 days in vitro in these two subpopulations of A C h E + neurons showed a >4-fold increase in the neurite index of the T N R + fraction of A C h E + cells when grown in the presence of L-T 3 over T N R + A C h E + cells grown in the absence of L-T3 (Table 2). In contrast, no significant difference in the neurite index of L-T3-treated T N R A C h E + cells was observed over T N R - A C h E + cells not treated with the hormone (Table 2).
Effect of L-tri-iodothyronine on L-tri-iodothyronine nuclear receptor immunoreactive acetylcholinesterasepositive cells after 20 days in vitro To determine a possible critical period of L-T3 action in vitro, the same fetal brain cultures were grown in the absence of L-T 3 (control) for 28 days, in the presence of L-T 3 for 8 days starting on culture day 20 and in the presence of L-T3 for the entire 28-day culture period. A C h E cytochemistry and T N R immunocytochemistry were performed after 28 days in vitro. When viewed under phase contrast, chronically L-T3-treated (those cells treated with L-T 3
478
R. GARZA et al.
k
(o) Fig. 3. Morphological effect of L-T 3 on T N R - A C h E + cells visualized by AChE cytochemistry and T N R immunocytochemistry after 8 days in r'itro. (a) L-T3-treated culture: immunofluorescence. (b) Same as (a): phase contrast. (c) Control culture: immunofluorescence. (d) Same as (c): phase contrast. Arrow indicates AChE + cells with TNR ( x 320).
Thyroid hormone receptors in acetylcholinesterase-positive neurons
479
Table 2. Effect of L-tri-iodothyronine on {~'~neiiri(edehsity of L-tri-iodothyronine nuclear receptor-immunoreactive and L-tri-iodothyronine nuclear receptor-negative acetylcholinesterase-positive cells after 8 days in vitro AChE + cells (TNR + ) Control L-T 3 (3 x 10 -s M) Neurite index
21.4 + 1.9
91.1 + 6.2*
AChE + cells (TNR-) Control L-T3 (3 × 10-8 M) 16.9 ___1.6
15.3 _+ 1.5
The values represent the number of neurite intersections falling in a 1 cm 2 area. Each value represents the mean of three different experiments performed in duplicate ___S.E.M. Statistical analyses were performed using Student's t-test. TNR +, TNR-immunoreactive; TNR-, TNR-negative. *Significantly different from control, P < 0.05.
for the entire culture period) 28-day-old A C h E + cells appeared larger and exhibited a greater degree of neurite arborization in contrast to 28-day-old A C h E + cells where the L-T3 treatment started on culture day 20 for 8 days or in 28-day-old control cultures (compare Fig. 4a with Fig. 4b). Indeed, neurite density measurements after 28 days in vitro showed a > 2 - f o l d increase in the neurite index of 28-day-old chronically L-T3-treated A C h E + cells over those 28-day-old A C h E + cells where the addition of the hormone was started on culture day 20 or in 28-day-old control cultures (Table 3). Cell counts to determine the relative percentages of T N R + A C h E + cells to T N R - A C h E + cells after 28 days in vitro were not possible due to the degree of morphological development at this stage of the cultures. However, all A C h E + cells taken for neurite density measurements were T N R +.
DISCUSSION In a previous in vitro study, 4 we demonstrated the sensitivity of central A C h E + neurons to L-T3 with an effect on the size of the perikarya and the neurite length. Table 3. Effect of the duration of L-tri-iodothyronine treatment at different periods of culture on the neurite density of acetylcholinesterase-positive cells L-T 3 (3 × 10 -8 M) treatment started (culture day) 0 20 1
Duration of L-T 3 treatment (days)
Neurite index after 28 days
0 8 28
76+9 75 + 8 182 + 20**
in vitro
Cells were grown in the absence of L-T3 for 28 days (no L-T 3 added at any time), in the presence of L-T 3 for 8 days starting on culture day 20 and in the presence of L-T 3 for 28 days. AChE staining was performed after 28 days in vitro and the neurite density determined. The values represent the number of neurite intersections falling in a 1 cm 2 area. Each value represents the mean of three different experiments performed in duplicate + S.E.M. Statistical analyses were performed using Student's ttest. **Significantly different from control and 8-day L-T 3treated cultures, P < 0.001.
The present study was undertaken to determine if L-T 3 nuclear receptors were present in A C h E + neurons and whether a relationship existed between the presence of T N R and the effects of L-T3 on the morphological development of this neuronal population. Here we show that 74% of fetal A C h E + neurons were immunoreactive for T N R , while 26% did not stain with 2B3 mab. This may suggest that some A C h E + neurons are not target cells for thyroid hormones or that, in our culture conditions, some A C h E + neurons do not express T N R or have lost T N R expression, depending on their degree of maturation. However, this latter hypothesis seems improbable since the percentage of A C h E + cells not stained by 2B3 mab remains the same with time in culture, at least until day 21. A cellular heterogeneity in the distribution of T N R in fetal neuronal cells is supported by several in vivo and in vitro data. The influence of neonatal dysthyroidism on the development of CHAT, tyrosine hydroxylase (TH) and glutamate decarboxylase ( G A D ) activities vary greatly depending on the brain area studied. 3"1°'21"23"26Furthermore, it has been reported that the developing cholinergic basal forebrain neurons but not the developing cholinergic pontomesencephalotegmental neurons are sensitive to thyroid hormone. 6 Similar results have been obtained in brain cell cultures where it has been shown that L-T 3 affects the development of mesencephalic and hypothalamic dopaminergic neurons differently. 22'24 These studies suggest that some suhpopulations of neurons are not responsive to thyroid hormones during brain development. This hypothesis is also supported by data showing the regional heterogeneity of T N R immunoreactivity in the adult rat brain. Comparing the regional distribution of A C h E + neurons H with the distribution of T N R immunoreactivity m the adult rat brain (Puymirat et al., unpublished observations), it appears that some brain areas containing A C h E + cells are weakly T N R immunoreactive, such as the caudate-putamen and the pallidum. It is therefore possible that some A C h E + cells located in these brain areas are not target cells for thyroid hormone. A double staining study should be performed to determine the target A C h E + cells for thyroid hormone in
(a) ii
tl
B
.
(b)
Q
IP
4
q
Fig. 4. Morphological effect of L-T 3 on AChE + cells after 28 days in t, itro. (a) AChE + cell treated with L-T 3 for 8 days starting on culture day 20. (b) AChE + cell treated with L-T 3 throughout the 28-day culture period ( × 640).
Thyroid hormone receptors in acetylcholinesterase-positiveneurons the rat brain. Additionally, our study shows a correlation between the presence of TNR immunoreactivity and the effect of L - T 3 o n the neurite density of AChE + neurons, which suggests that this effect may be directly mediated by an interaction of L-T3 with specific nuclear receptors. The direct effect of L-T3 on neurite elongation and branching is also supported by our recent finding showing that conditioned media prepared from astrocytes grown in the presence L - T 3 are unable to produce an increase in neurite density when compared with conditioned media obtained from astrocytes grown in the absence of L-T 3 (Garza et al., unpublished observations). To support this hypothesis, it remains to be determined whether the expression of TNR is associated with the morphological effect of L - T 3. However, this cannot be performed in our culture conditions since 80% of the cell nuclei are TNR + after 1 day in culture (data not shown). The failure of LoT3 to elicit a morphological response if added on culture day 20 suggests the existence of a critical period of L - T 3 action which coincides with in vitro studies where ChAT activity is stimulated only when the hormonal treatment is started before culture day 20. 4'1°This is in agreement with in vivo data showing that L - T 3 treatment of neonatal hypothyroid rats, beginning after the 14th postnatal day (which corresponds to our culture day 20), does not reverse the deficiency produced by hypothyroidism. 3'8The apparent discrepancy between
481
the absence of the morphological response to thyroid hormone and the presence of TNR immunoreactivity in 28-day-old AChE + cells may be explained by the following: our mab may recognize different forms of the TNR. Indeed, recent data ~7 have demonstrated not only the TNR, which binds L-T3, but also two variant forms which do not bind L-T 3. It is therefore possible that our mab recognizes both forms of the TNR, which may explain the absence of a morphological effect of L-T 3 despite the presence of TNR immunoreactivity. An alternative explanation is that our mab recognizes the active form of the TNR, but the gene(s) regulated by L - T 3 have lost their sensitivity to thyroid hormones after 20 days in vitro. CONCLUSION
This in vitro study showed that (1) fetal brain AChE + neurons are target cells for thyroid hormones. Since subpopulations of AChE + cells present in our cultures are also ChAT-positive,9 our results may suggest that some subpopulations of cholinergic neurons are target cells for thyroid hormone. (2) The presence of TNR in a subpopulation of AChE + neurons associated with L-T3 morphological effects suggests that some effects of L-T3 on the maturation of AChE + neurons may result from a direct effect of L - T 3 through an interaction with its specific nuclear receptors.
REFERENCES
1. Bottenstein J. E. and Sato G. H. (1979) Growth o f a neuroblastoma cell line in serum-free supplemented medium. Proc. natn. Acad. Sci. U.S.A. 76, 514-517.
2. Faure R., Ruel J. and Dussault J. H. (1986) Production of an antibody against rat liver nuclear T 3 receptor. Biochem. Cell Biol. 64, 377-380. 3. Garcia Argiz C. A., Pasquini J. M., Kaplan B. and Gomez C. J. (1967) Hormonal regulation of brain development. II. Effect of neonatal thyroidectomy on succinate dehydrogenase and other enzymes in developing cerebral cortex and cerebellum of the rat. Brain Res. 6, 635 646. 4. Garza R., Dussault J. H. and Puymirat J. (1988) Influence of tr/iodothyronine (L-T3) on the morphological and biochemical development of fetal brain acetylcholinesterase-positiveneurons cultured in a chemicallydefined medium. Devl Brain Res. 43, 287-297. 5. Geel S. E. and Timiras P. S. (1967) Influence of neonatal hypothyroidism and of thyroxine on the acetylcholinesterase and cholinesterase activities in the developing central nervous system of the rat. Endocrinology 80, 1069-1074. 6. Gould E. and Butcher L. L. (1989) Developing cholinergic basal forebrain neurons are sensitive to thyroid hormone. J. Neurosci. 9, 3347-3358. 7. Grave G. D. (1977) Thyroid Hormones and Brain Development. Raven Press, New York. 8. Hamburg M. and Flexner L. B. (1957) Biochemicaland physiologicaldifferentiation during morphogenesis. XXI. Effect of hypothyroidism and hormone therapy on enzymeactivities of the developingcerebral cortex of the rat. J. Neurochem. 1, 273-288. 9. Hefti F., Hartikka J., Eckenstein H., Gnaln H., Heumann R. and Schwab M. (1985) Nerve growth factor increases choline acetyltransferase but not survival or fiber outgrowth of cultured fetal septal cholinergic neurons. Neuroscience 14, 55-68. 10. Honegger P. and Lenoir D. (1980) Triiodothyronine enhancement of neuronal differentiation in aggregating fetal rat brain cells cultured in a chemically defined medium. Brain Res. 199, 425-434. 1l. Jacobowitz D. M. and Palkovits M. (1974) Topographic atlas of catecholamine and acetylcholinesterase-containing neurons in the rat brain. I. Forebrain (telencephalon, diencephalon). J. comp. Neurol. 57, 13. 12. Kolodny J. M., Leonard J. L., Larsen P. R. and Silva J. (1985) Studies of nuclear 3,5,Y-triiodothyronine binding in primary cultures of rat brain. Endocrinology 117, 1848-1853. 13. Luo M., Faure R. and Dussault J. H. (1986) Ontogenesis of nuclear T~ receptors in primary cultured astrocytes and neurons. Brain Res. 381, 275-280. 14. Luo M., Faure R., Ruel J. and Dussault J. H. (1988) A monoclonal antibody to the rat nuclear triiodothyronine receptor: production and characterization. Endocrinology 123, 180-186. 15. Luo M., Puymirat J. and Dussault J. H. (1989) Immunocytochemical localization of nuclear 3,5,3'-triiodothyronine (L-T3) receptors in astrocyte cultures. Devl Brain Res. 46, 131 136.
482
R. GARZAet al.
16. Luo M., Faure R., Tong Y. A. and Dussault J. H. (1989) Immunocytochemicat localization of the nuclear 3,5,Y-triiodothyronine receptor in the adult rat: liver, kidney, heart, lung and spleen. Acta endocr. Copenh. 120, 451-458. 17. Mitsuhashi T., Tennyson G. E. and Nikodem V. M. (1988) Alternative splicing generates messages encoding rat c-erb A proteins that do not bind thyroid hormone. Proc. natn. Acad. Sci. U.S.A. 85, 5804-5808. 18. Oppenheimer J. H. (1979) Thyroid hormone action at the cellular level. Science 203, 971-979. 19. Oppenheimer J. H., Schwatz H. L., Mariash C N., Kinlan W. E., Wong N. C. W. and Freake H. C. (1987) Advances in our understanding of thyroid hormone action at the cellular level. J. Endocr. Invest. 8, 288-308. 20. Pascual A., Aranda A., Ferret-Senat V., Gabellec M. M., Rebel G. and Sarlieve L. L. (1986) Triiodothyronine receptors in developing mouse neuronal and glial cell cultures and in chick-cultured neurons and astrocytes. Devl Brain Res. 8, 89-101. 21. Patel A. J., Hayashi M. and Hunt A. (1987) Selective persistent reduction in choline acetyltransferase activity in basal forebrain of the rat after thyroid deficiency during early life. Brain Res. 422, 182-185. 22. Puymirat J., Barret A., Picart R., Vigny A., Loudes C., Faivre-Bauman A. and Tixier-Vidal A. (1983) Triiodothyronine enhances the morphological maturation of dopaminergic neurons from fetal mouse hypothalamus cultured in serum-free medium. Neuroscience 10, 801-810. 23. Puymirat J. (1985) Effects of dysthyroidism on central catecholaminergic neurons. Neurochem. Int. 7, 969-977. 24. Puymirat J., Faivre-Bauman A., Barret A., Loudes C. and Tixier-Vidal A. (1985) Does triioodothyronin influence the morphogenesis of fetal mouse mesencephalic dopaminergic neurons cultured in chemically defined medium? Devl Brain Res. 23, 315-317. 25. Puymirat J., Luo M. and Dussault J. H. (1988) Immunocytochemical localization of thyroid nuclear receptors in cultured hypothalamic dopaminergic neurons. Neuroscience 30, 443-449. 26. Rastogi R. B. and Singhal R. L. (1976) Influence of neonatal and adult hyperthyroidism on behavior and biosynthetic capacity for norepinephrine, dopamine and 5-hydroxytryptamine in rat brain. J. Pharmac. exp. Ther. 198, 609-618. 27. Ruel J., Faure R. and Dussault J. H. (1985) Regional distribution of nuclear T 3 receptors in rat brain and evidence for preferential localization in neurons. J. Endocr. Invest. 8, 343-348. 281 Samuels H. H., Stanley F. and Casanova J. (1979) Depletion of L-3,5,Y-triiodothyronine and L-thyroxine in euthyroid calf serum for use in cell culture studies of the action of thyroid hormone. Endocrinology 105, 80-85. 29. Tago H., Kimura H. and Maeda T. (1986) Visualization of detailed acetylcholinesterase fiber and neuron staining in rat brain by a sensitive histochemical procedure. J. Histochem. Cytochem. 34, 1431-1438. 30. Valcana T. (1971) Effect of neonatal hypothyroidism on the development of acetylcholinesterase and choline acetyltransferase activity in rat brain. In Influence of Hormones on the Nervous System (ed. Ford D. H.), pp. 174-184. S. Karger, Basel. (Accepted 17 January 1990)
