- Email: [email protected]
Retinoic acid-induced decrease of DNA synthesis and peroxidase mRNA levels in human thyroid cells expressing retinoic acid receptor alpha mRNA
Life Sciences, Vol. 53, pp. 1039-1048 Printed in the USA
Pergamon Press
RETINOIC ACID-INDUCED DECREASE OF DNA SYNTHESIS AND PEROXIDASE mRNA LEVELS IN HUMAN THYROID CELLS EXPRESSING RET1NOIC ACID RECEPTOR ALPHA mRNA L. del Senno*, R. Rossi, D. Gandini*, R. Piva*, P. Franceschetti and E. C. degli Uberti Cattedra di Medicina Intema-Sezione di Endocfinologia, *Istituto di Chimica Biologica e Centro di Studi delle Patologie del Genoma Umano, Universifi degli Studi di Fen'ara, Italy. (Received in lrmal form July 12, 1993) Summarv In order to clarify the effect of retinoids on thyroid cell growth and function, the presence of retinoic acid receptors (RARs) and the action of retinoic acid (RA) on DNA synthesis and on thyroid peroxidase (TPO) and thyroglobulin (TGB) mRNA expression were investigated in primary cultures of human thyroid follicular cells. A time and dose-dependent reduction in 3H-thymidine (3H-thy) incorporation was found in cells exposed for 48 h to all-trans-RA up to 1 ~tM. A cytotoxic effect was found only with the higher dose of 50 ~tM. The RA-induced decrease of 3H-thy incorporation was reflected by parallel change in DNA content of cell monolayers. The inhibitory effect of 1 I.tM RA on 3H-thy incorporation ranged from 28.5 + 4.6 % in normal cells to 42 +3.2 % in adenomatous cells. In addition, 1 ~tM RA significantly reduced basal and TSH-induced TPO mRNA levels in normal, goiu'ous and adenomatous ceils, but did not alter TGB mRNA levels. Furthermore, in these cells the study of R A R c t and B mRNA showed the presence of two major R A R c t mRNA transcripts of approximately 3.5 and 2.8 Kb in size, whereas RAR 13mRNA was undetectable. Overall, our data indicate that RARct gene is expressed in human thyrocytes and that RA may be involved in the regulation of the human thyroid by reducing proliferation and function of follicular cells. The vitamin A derivative retinoic acid (RA) may play a fundamental role in regulating the growth and differentiation of a vmiety of cells (1). The cellular responses to RA are mediated by the binding to nuclear receptor proteins that are members of the steroid/thyroid hormone receptor superfamily of transcriptional regulators (2). Retinoids have been shown to affect thyroid function "in vivo". Vitamin A deficiency in rats has been found to be associated with an increase in serum concentrations of thyroid hormones (3). On the other hand, massive doses of vitamin A seem to inhibit thyroid function in man (4). In order to verify the possible role of RA in the regulation of growth and differentiation of the human thyroid we examined in primary cultures of human follicular cells the effect of RA on 3HCorresponding Author: Prof. E.C. degli Ubel~i, Endocrinology Section, Institute of Internal Medicine, University of Fen'ara, via Savonarola 9, 1-44100 Felxara, Italy. 0024-3205/93 $6.00 + .00 Copyright © 1993 Pergamon Press Ltd All rights reserved.
1040
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
thymidine (3H-thy) incorporation and mRNA expression of thyroglobulin (TGB) and peroxidase (TPO), the main markers of thyroid differentiation and function (5,6). Furthermore, to clarify the mechanism by which RA could act on thyroid ceils, we also investigated the mRNA expression of RA receptors (RARs) by measuring the mRNA levels of RARct and B, as they appear to have a relatively wider distribution than RAR ~ in mouse tissues (7). The expression of these genes was analysed in basal conditions and in response to RA and/or to TSH, the main regulator of thyroid growth and function (6,8). Methods Cell culture Follicles from 10 normal, 10 goitrous and 5 adenomatous human thyroid samples were prepared by collagenase (type IV, Boehringer Mannheim) and trypsin treatment as previously described (9). Normal tissues were taken from a normal area of thyroid glands from patients (8 women and 2 men) undergoing surgery for solitary non functioning nodules. Goitrous and adenomatous tissues were obtained from patients undergoing surgery for a non toxic multinodular goiter (7 women and 3 men) and for a follicular adenoma (4 women and 1 man) respectively. All patients were euthyroid and did not receive thyroid holinone therapy before surgery. Follicles were resuspended in RPMI 1640 medium (Gibco, Grand Island, NY) with 5% foetal calf serum (FCS, FLOW, Meckenheim), distributed in 24-well culture plates (16 mm diameter) or in plastic 250 ml flasks (Costar, Cambridge, Ma), whatever was more appropriate for the experimental protocol, and grown at 37°C in a humidified atmosphere of 5% CO2. Experiments were carried out within 5-10 days of plating. For 3H-thy incorporation analysis, cells, cultured in 24-well plates as indicated, were incubated with either serum free medium plus 0.4% bovine serum albumine (BSA) or 2% FCS and all-trans retinoic acid (RA, Sigma, Chemical Co.,St. Louis,MO) at various concentrations and for different times, according to the experimental protocol. RA from 10 mM stock solution in ethanol was further diluted with RPMI 1640 medium to final concentration as required by the protocol. The final concentration of ethanol was the same in both control and RA-treated cultures. Cell viability was determined by their adherence to plastic dishes and by the trypan blue dye exclusion technique. All experiments were performed in triplicate. For RNA analysis, cells, cultured in 250 ml flasks, as indicated, were mantained for two days in 2% FCS and then were incubated with 2% FCS alone or with RA (1, 10 ~tM) and/or with bovine TSH (10 mU/ml, Ambinon, Organon). DNA synthesis 3H-thy incorporation was measured in adherent cells (approximately 105 cells/well) in 24-multi-well plates obtained as detailed in the previous section. Cells were incubated with 0.4 % BSA, 3H-thy (1.5 I.tCi/ml, 87 Ci/mmol) as well as with and without RA for 24, 48 and 72 h, unless otherwise indicated. 3H-thy was added at time 0. After the incubation time, cells were washed three times with ice cold phosphate buffer saline (PBS) and twice with 10% ice-cold trichloro-acetic acid (TCA). TCA precipitable material was solubilized in 500 I.tl 0.2 M sodium hydroxide (NaOH) and 0.1% sodium dodecyl sulphate (SDS) as previously described (10). Cells associated radioactivity was then counted in a scintillation spectrometer. Results (count per min/well) were obtained in triplicate experiments.
Vol. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
1041
In addition, in a parallel set of experiments, the DNA content of cell monolayers was evaluated as an estimation for cell number under the same concentrations. Cells were harvested by scraping, washed in saline and pooled by centrifugation at 2000 rpm for 10 min. Cells were lysed in 300 B1 of 0.5% NP40 solution and spun at 12000 rpm for 5 min. DNA from nuclei was obtained by RNAse and proteinase K treatment, then phenol-chloroform extracted (11). DNA content was estimated by optical density at 260 nm and expressed as ~g/well. 3H-leucine incorooration into thvroid oroteins In order to determine the effect of RA on protein synthesis, 3H-leucine incorporation into thyroid cells was evaluated. Adherent cells in 24-well culture plates (as for 3H-thy incorporation) were incubated without and with 0.01, 0.1, 1, 10 and 50 ~tM RA respectively, in presence of 0.4% BSA and 3H-leucine (20 BCi/ml, 141 Ci/mmol). After 24 h of incubation at 37 ° C, the labelling was stopped by removing the medium and washing the cells with ice-cold PBS, pH 7, and twice with ice-cold TCA. Radioactivity into the TCA-precipitable material was determined as for 3H-thy incorporation. Results (count per min/well) were obtained in triplicate experiments. RNA det¢rmin~ion, After incubation for 48 h in 250 ml flasks, adherent cells were harvested and lysed in 0.5% NP40 solution (containing the RNA inhibitor vanidylribonucleoside, Sigma), as described (12). Total RNA was isolated from cytoplasm by phenol-chloroform extraction, subdivided in three aliquots and ethanol precipitated. One aliquot per flask was used for Northern blot analysis. RNA (8-30 Bg) was denatured in formaldehyde, electrophoresed in 1% agarose-0.66 M formaldehyde gel with 0.5 Bg/ml ethidium bromide in 20 mM 3-morpholinopropansulphonic acid, pH 7 running buffer and transferred onto Gene Screen Plus membrane (NEN, Boston, MA). Filters were hybridized for 18-36 h at 42 ° C in 50 % formamide and dextrane sulphate with 32p_ labeled probes. The DNA probes were: pH23, a 1300 base pair fragment of human TGB c-DNA; pH2.8, a TPO c-DNA; pAct2, an actin c-DNA probe; pTM61A and pSKR13, the o~ and 13 RAR cDNA probes. The probes were labeled with 32p dCTP (3000 Ci/mmol) by multiprime DNAlabeling system (Boehringer, Mannheim, Germany). After hybridization and washing, filters were exposed to Kodak X OMAT films at -70°C with Dupont intensifying screens. The levels of TGB, TPO and RAR t~ RNA in each sample were measured by densitometric analysis of the autoradiographic signals (12). Scanning values were related to those of actin RNA and expressed as relative scanning units which reflect the relative abundance of each mRNA in a given RNA sample. Analvtical orocedure. Data are expressed as the mean + SEM. Statistical analysis of the data was performed using the unpaired and paired Student's t test, as applicable. Results Effects of RA on DNA synthesis To find out the effect of RA on DNA synthesis of human thyroid cells, 3H-thy incorporation was measured in cells treated for 48 h with increasing concentration of RA (0.01, 0.1, 1, 10, and 50 BM). As shown in Fig. 1A, RA induced a significant decrease of 3H-thy incorporation at all concentrations. 1 ~tM RA was more inhibiting than 10~tM (45.9+5.1 and 39.9 + 7.25 % inhibitions respectively after 1 and 10 I.tM RA) but maximal decrease occurred at 50 ~tM (64.65 + 8.7 % inhibition). However, at this maximal concentration, cells exhibited an abnormal shape,
1042
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
round in form, suggestive of a RA cytotoxic effect for these cells. No changes in shape, adherence and trypan blue staining were observed by phase-contrast microscope between cells treated with 0.01, 0.1, 1 and 10 ~tM RA. As shown in Fig. 1B, the inhibitory effect of RA on 3H-thy incorporation of human thyroid cells was also reflected by parallel changes in the DNA content of cells monolayers (22 _ 2.6%, 20 + 2.4% and 57 + 4.9% inhibitions respectively after 1, 10 and 50 I.tM RA). In cells treated for 24 h with RA, a reduction in protein synthesis measured by 3H-leucine incorporation was observed only at 50 IxM concentration (Fig. 1C). 120 i00
~
~o
~'~ # ~_~
20 o
-r
48
T
A
-r
6O
B
5O
~
40
~
2o
120
11InNl'p
i00
--
lO o
24
I"
-r
"1-
I-
l
i0
C
0
-'~ o
80
•
40 ~
"~--~
20
o m
0
0.010.i
50
R A ktM
FIG. 1. Effect of RA on DNA and protein synthesis in human thyroid follicular cells. A: Cells from cultures of goitrous thyrocytes were incubated for 48 h in serum-free medium, plus 0.4% BSA, with RA at various concentrations and with 3H-thy. The mean values (cpm/well) + SEM in RA-treated cells were compared to those of untreated cells and expressed as percent of control values. B: DNA content in cells cultured for 48 h was measured as described in Materials and Methods. The mean values in I.tg/well (+SEM) were shown. C: 3H-leucine incorporation was measured in cells cultured for 24 h (see Methods for technical details). The results were obtained from three independent experiments in triplicate. *, **, ***, p< 0.01, 0.05 and 0.001 respectively vs. untreated cells.
Vol. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
1043
B
A =
5-~
O
O
"~
4
4
= , -•"7.
3
~
-V. 3
1 g M RA ;~'~
o~,
normal
adenoma
q-
T
O
1
;~'~ ,a:z
1
!
0
I
!
I
24
48 Hours
72
0
|
0
1 RA
i
0 l.tM
1
FIG. 2. Effect of 1 p-M RA on 3H-thy incorporation in human normal and adenomatous thyrocytes. A: Time course study of 3H-thy uptake into normal cells incubated in serum free medium plus 0.4% BSA,with (A) and without (A) RA (1 p.M) for 24, 48 and 72 h. Results are the mean (+SEM) of a triplicate experiment. B: Effect of RA (lp-M) on 3H-thy incorporation in thyrocytes isolated from ten normal thyroids and six adenomas. Empty (0) and filled (1) columns represent untreated and RA-treated cell cultures. 3H-thy incorporation was measured as radioactivity in TCA-precipitable material, as described in Materials and Methods, and expressed as the mean (_+SEM) value of cpm/well. *, **, p< 0.05 and 0.01 respectively vs. untreated cells.
To study the effects of RA after different periods of incubation, 1 p-M concentration of RA was chosen, because the percent reduction in RA treated cells versus untreated control cells was at 1 p-M slightly higher than that observed at 10 p-M. Similar RA-induced reduction was also observed in presence of 2% FCS. As shown in Fig.2A, the concentration of 1 gM RA reduced 3H-thy incorporation in normal thyroid after both 48 h (from 35760 + 650 in control untreated cells to 23150 + 1400 cpm/well in RA-treated cells, p< 0.01) and 72 h (from 45120 + 3440 in untreated cells to 25270 + 592 cpm/well in RA-treated cells, p< 0.01). As indicated in Fig.2B, the analysis of 3H-thy incorporation in thyrocytes obtained from 10 normal thyroids and 6 adenomas showed that the 48 h treatment with 1 p-M RA caused a significant decrease (p< 0.01) in 3H-thy incorporation also in adenomatous cells. In adenomas the mean + SEM percent RA-induced reduction in 3H-thy incorporation was 42.7 + 3.2 % compared with that of 28.5 + 4.6 % obtained in normal thyroid.
1044
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
B
A
Kb
TPO
"--3.2
TGB
4 " - 8.7
---- 1.8
ACTIN TSH RA 1 RAIO
+ +
-+
+
+
+ --
-+
mRNA a r b i t r a r y units TSH RA RA +TSH TPO Actin
5.06 ±0.41
50.65 +11.94
3.33 ~ +_0.53
15.60 + *_5.70
TGB
2.19 +_0.20
14.13 ±5.19
2.09 ±0.32
11.82 ±4.24
Actin
FIG.3 Northern blot analysis of TPO and TGB mRNA transcripts in human thyrocytes treated with RA. A: Cells isolated from a goiter were cultured as indicated in Materials and methods and were incubated for 2 days in medium supplemented with 2% FCS, without (-) and with (+) TSH (10 mU/ml), RA (1 or 10 pM). Northern blotting was carried out using one aliquot of total RNA isolated from each 250 ml flask. RNA was electrophoresed on agarose-formaldehyde gel, then transfen'ed to a nylon membrane filter (Gene Screen Plus, NEN) and hybridized with 32p labelled TPO (upper panel) and TGB (middle panel) probes. Since different amounts of total RNA have been loaded on the gel, after appropriate exposure, the blots were rehybridized with the actin probe (lower panel) to check the amount of RNA loaded in the different lanes. The ratio betweeen the intensity of TPO and TGB mRNA and that of actin mRNA, as judged from the densitometric scanning of the con'esponding lanes in the autoradiography, was calculated in RNA blots of 5 thyrocyte cultures (2 normal thyroids and 3 goiters) treated with 1 laM RA (details of method are described in Materials and Methods). The mean (±SEM) values of the relative scanning units are given in B and expressed as mRNA arbitrary units. *, p< 0.05 vs. (-); +, p< 0.01 vs. (TSH) using the paired Student's t test.
Effect of RA on TPO and TGB mRNA levels. Northern blot analysis of TPO and TGB mRNA in Fig.3 showed that in thyrocytes, as expected, the levels of TPO and TGB mRNA markedly increased after treatment with TSH (10 mU/ml) for 48 h. When cells were incubated in the presence of RA (1 and 10 p.M), the TSH effect on TPO mRNA levels sharply decreased at both RA concentrations (Fig. 3A, upper panel). A slight reduction in both basal and TSH-induced TGB mRNA levels by 1 and 10 pM RA was also observed (Fig. 3A, middle panel). The densitometric analysis (TPO/actin and TGB/actin mRNA ratio, Fig. 3B) of Northern blots of RNA isolated from 5 cultures (2 normal thyroids and 3 goiters) confirmed these results and indicated that both basal and TSH-induced levels of TPO mRNA were significantly reduced by 1 ~tM RA, whereas basal and TSH-induced TGB mRNA levels were not significantly modified by RA.
Voi. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
NORMAL
A
GOITER I
1045
ADENOMA
I
I
I
!
Kb 4 3.5 2.8 2
RARr~
ACTIN
1.8
TSH
- + - +
RA 1
- - - - -I'- "t-
--+
++ +--+-+--+
RA10
B
+
m
-I-
Kb --,---4 •",--" 3.5 - - - 2.8
RARa
-.--.-2
--28S RARfl --18
-,,- 1.8
ACTIN
TSH RA1 RAIo
S
--
4"
~ 4-
~
4-I--
-I-
FIG. 4 Northern blot analysis of RAR ~ mRNA transcripts in human thyrocytes treated with RA. A: Cells isolated from normal, goitrous and adenomatous cells were cultured and processed as indicated in Fig.3 and in Materials and Methods. RNA was hybridized with 32p labelled RAR ct (upper panel) and actin probes. B: RNA isolated from goitrous cultured cells was hybridized with 32p labelled RAR tx (upper panel), RAR 13 (middle panel) and actin probes (lower panel).
1046
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
RAR g mRNA expression in thyrocytes The expression of RA receptors was examined by hybridizing the RNA samples of the cultured cells with two probes which can specifically detect RAR c~ and 13 transcripts (7). The Northern blot analysis of RNA isolated from normal, goitrous and adenomatous cells showed that after hybridization with the RAR ct probe two minor RNA species of approximately 4 and 2 Kb in size, and two major RNA species of approximately 3.5 and 2.8 Kb (Fig. 4 A and B, upper panels) were detected. The two major RARct mRNA species corresponded well to those shown in other human tissues in previous studies (7). After hybridization with the RAR 13probe (Fig. 4B, middle panel), no transcripts were detected in these samples, as well as in the other RNA samples examined (data not shown). Densitometric data (RAR c~/actin mRNA ratio) of Northern blots of RNA isolated from 5 cultures (2 normal thyroids and 3 goiters), indicated no significant variation in basal R A R c t mRNA levels after u'eatment with 1 p.M RA, TSH and with the combined treatments, in human thyroid cell cultures (data not shown). Discussion We assessed the effects of RA on DNA synthesis and TGB and TPO mRNA expression in human thyroid cells cultured in vitro. The reduction of 3H-thy incorporation by all-trans-RA suggests that this retinoid mainly at 1 I.tM concentration exerts an inhibitory action on DNA synthesis and retains a growth inhibitory activity on human thyroid, as observed in other cell systems (1,13). Although the inhibitory action of RA on 3H-thy incorporation was detected also at the dose of 10 ktM, which has been found to be cytotoxic in other cell lineages, such as breast and melanoma cell lines (14,15), no changes in 3H-leucine incorporation nor in shape and adherence, suggestive of RA cytotoxic effect on thyroid cells, were observed. In addition, RA did not appear to be cytotoxic at these doses in other human cell systems including thyroid carcinoma (16) or rhabdomyosarcoma (13). Our findings are consistent with the inhibitory effect of either 13-cis-RA or retinyl acetate at 10 I~M concentration on 3H-thy uptake observed in a human follicular carcinoma cell line (16) both in basal and in EGF-stimulated conditions. However, our data do not agree with that observed in cultures of porcine thyroid cells (17) in which retinol (1 ktM) did not modify basal 3H-thy incorporation whereas it increased EGF- or IGF-Iinduced 3H- thy incorporatiol~. Although this difference in RA activity on thyroid growth is difficult to explain, we suggest that RA could possess divergent effects, growth-stimulating as well as growth-inhibiting properties, depending on the different responsiveness to RA of cell cultures from different species (humans vs. pigs). The present study shows that RA inhibits basal and TSH-stimulated TPO mRNA levels both in normal and in goitrous cells. This is in agreement with the reduction of serum thyroid hormones observed "in vivo" using massive doses of vitamin A in the treatment of hyperthyroidism (4) and also with the inhibition of both TSH-induced TPO mRNA levels and iodide uptake by retinol in cultures of porcine thyroid cells (17). The biochemical mechanism leading to the reduction of TPO mRNA levels by RA at present is still unknown. Such modulation of gene expression may be achieved either by alteration at transcriptional or posttranscriptional level, including RNA processing and mRNA stability. In addition, this may results by interfering with TSH-responsive signal transduction pathways, which
Vol. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
1047
control both proliferation and gene expression in thyroid cells (18), or by a cis direct action of RA on TPO gene promoter. Since we demonstrated that RA inhibits not only TSH-induced TPO mRNA levels, but also those found in basal conditions, we suggest that RA may act independently from TSH or TSH-receptor. According to this are data reported on porcine cells (17) where cAMP response to TSH is not modified by retinol. Since TPO gene transcription is stimulated by TSH via a cAMP pathway through a cAMP responsive element present in its promoter (6), it is attractive to speculate that also a RA responsive element could be present upstream the TPO gene. Such an element, binding a RA-receptor complex, could negatively affect gene u'anscription. The inhibition of TPO gene expression by RA here reported is somewhat at variance with findings obtained in a human follicular carcinoma cell line where RA reduces cell proliferation, but increases 1251 uptake, thus driving tumor cells toward a more differentiated state (16). However, this may be explained by previous observations indicating that retinoid effects can vary according to the stage of differentiation of the responding cells. In relatively non differentiated cells, such as human myeloid leukemia cells (19), RA may reduce cell proliferation inducing a more differentiated state. On the other hands, in keratinocytes retinoids inhibit the expression of differentiated functions (20). In particular, in non neoplastic hepatocytes RA inhibits both proliferation and function (21) In contrast to the apparent inhibitory action of RA on TPO mRNA levels, no significant effect was observed on TGB expression. This suggests that TGB gene expression is not affected by RA. This different expression pattern is not an unusual finding in the thyroid. In fact, it has been shown that different thyroid specific functions, including iodide uptake, TPO and TGB mRNA levels may exibit different sensitivity to hormonal regulation and oncogene transformation (22): the most sensitive is iodide uptake and the less sensitive is TGB gene expression. This assumption agrees with the observations that maintenance of differentiated functions in cultured thyroid cells results from complex interactions among various factors and that different cellular processes are independently implicated in the modulation of TGB and TPO gene expression (5,6). Our "in vitro" results suggest that human thyrocytes possess specific receptors that may mediate RA action. At present there are no data on RAR gene expression in thyroid cells. Our investigation shows that RAR ct transcripts are present in normal, goitrous and adenomatous human thyroid cells, whereas levels of expression of RAR 13gene cannot be detected. In conclusion, our data demonstrate that the specific retinoic acid receptor tx gene is expressed in human thyrocytes and that RA may be involved in the regulation of the human thyroid by reducing proliferation and function, but not differentiation of follicular cells. However, further investigation of differential regulation of RAR gene expression by RA may provide a deeper insight into the action of RA on cell differentiation and growth.
Acknowledgments We would like to thank the Drs. F. Mavilio, E Chambon, G. Vassart, S. Kimura and J.L. Mandel for giving us the RARs, TGB, TPO and actin probes, respectively. We thank Dr. B. Anderson for correcting the English, Work supported in part by MURST 40-60 %, CNR PFs Biotec., ACRO, by AIRC and Regione Emilia Romagna.
1048
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
13 14. 15 16. 17. 18. 19. 20. 21. 22.
AB. ROBERTS and M B . SPORN, The Retinoids. Academic, Orlando, FL. 2209-286 (1984). LM. DE LUCA, FASEB J. 52924-2933 (1991). H. GARCIN, P. HIGUERET, J. Endocrinol. 84135-140 (1980). S. SIMKINS, J. Clin. Endocrinol. Metab. 7574-585 (1947). Y. NAGAYAMA, S. YAMASHITA, H.HIRAYU, et al, J. Clin. Endocrinol. Metab. 6811551159 (1989). CM. GERARD, A. LEFORT, D. CHRISTOPHE, et al.,Mol. Endocrinol.3 2110-2118 (1989). A. KRUST, P. KASTNER, M. PETKOVICH, A. ZELENT, P. CHAMBON,.Proc. Natl. Acad. Sci. USA. 865310-5314 (1989). P. ROGER, M. TATON, J. VAN SANDE, JE. DUMONT, J. Clin. Endocrinol, Metab.6....66 1158-1165 (1988). L. DEL SENNO, EC. DEGLI UBERTI, S. HANAU, R. PIVA, R. ROSSI, G. TRASFORINI, Mol. Cell. Endocrinol..6..~367-74 (1989). GP. SCHINCAGLIA, F. FORNITI, R. CAVALLINI, R. PIVA, G. CALURA, L. DEL SENNO, J.Oral Pathol. Med.. 2...! 181-185 (1992). J. SAMBROOK, E.T. FRITSCH and E MANIATIS, . Molecular Cloning. CSH Laboratory press, 29.16-9.19 (1989). EC. D E G L I UBERTI, S. HANAU, R. ROSSI, R. PIVA, A. M A R G U T T I , G. TRASFORINNI, G. PANSINI, L. DEL SENNO, J. Clin. Endocrinol. Metab. 7..221364-1371 (1991). GD. CROUCH,LJ. HELMAN, Cancer Res. 514882-4887 (1991). H. UEDA, M. ONO, Y. HAGINO, M. KUWANO, Cancer Res. 4..~_53332-3338 (1985). R. LOTAN, T. STOLARSKY, D. LOTAN, Cancer Res.4_.3_32868-2875 (1983). AJ. VAN HERLE, ML. AGATEP, DN. PADUA III, TL. TOTANES, DV. CANLAPAN, HML. VAN HERLE, and GJE JUILLARD, J. Clin. Endocrinol. Metab 71755-763 (1990). M. ARAI, T. TSUSHIMA, O. ISOZAKI, K. SHIZUME, I. EMOTO, H. DEMURA, M. MIYAKAWA, and N. ONODA, Endoclinology 1292827-2833 (1991). JE. DUMONT, E LAMY, P. ROGER, and C. MAENHAUT, Physiol. Rev. 72667-697 (1992). TR. BREITMAN, SE. SELONICK and SJ. COLLINS, Proc. Natl. Acad. Sci. USA. 7729362940 (1980). E. FUCHS and H. GREEN, Cell 2..5_5617-624 (1981). H. IKEDA and K. FUJIWARA, Biochem. Biophys. Res. Comm. 191675-680 (1993). RZ. ZARRILLI, S. FORMISANO and B. DI JESO, Molec. Endocrinol., _439-45 (1990).
Pergamon Press
RETINOIC ACID-INDUCED DECREASE OF DNA SYNTHESIS AND PEROXIDASE mRNA LEVELS IN HUMAN THYROID CELLS EXPRESSING RET1NOIC ACID RECEPTOR ALPHA mRNA L. del Senno*, R. Rossi, D. Gandini*, R. Piva*, P. Franceschetti and E. C. degli Uberti Cattedra di Medicina Intema-Sezione di Endocfinologia, *Istituto di Chimica Biologica e Centro di Studi delle Patologie del Genoma Umano, Universifi degli Studi di Fen'ara, Italy. (Received in lrmal form July 12, 1993) Summarv In order to clarify the effect of retinoids on thyroid cell growth and function, the presence of retinoic acid receptors (RARs) and the action of retinoic acid (RA) on DNA synthesis and on thyroid peroxidase (TPO) and thyroglobulin (TGB) mRNA expression were investigated in primary cultures of human thyroid follicular cells. A time and dose-dependent reduction in 3H-thymidine (3H-thy) incorporation was found in cells exposed for 48 h to all-trans-RA up to 1 ~tM. A cytotoxic effect was found only with the higher dose of 50 ~tM. The RA-induced decrease of 3H-thy incorporation was reflected by parallel change in DNA content of cell monolayers. The inhibitory effect of 1 I.tM RA on 3H-thy incorporation ranged from 28.5 + 4.6 % in normal cells to 42 +3.2 % in adenomatous cells. In addition, 1 ~tM RA significantly reduced basal and TSH-induced TPO mRNA levels in normal, goiu'ous and adenomatous ceils, but did not alter TGB mRNA levels. Furthermore, in these cells the study of R A R c t and B mRNA showed the presence of two major R A R c t mRNA transcripts of approximately 3.5 and 2.8 Kb in size, whereas RAR 13mRNA was undetectable. Overall, our data indicate that RARct gene is expressed in human thyrocytes and that RA may be involved in the regulation of the human thyroid by reducing proliferation and function of follicular cells. The vitamin A derivative retinoic acid (RA) may play a fundamental role in regulating the growth and differentiation of a vmiety of cells (1). The cellular responses to RA are mediated by the binding to nuclear receptor proteins that are members of the steroid/thyroid hormone receptor superfamily of transcriptional regulators (2). Retinoids have been shown to affect thyroid function "in vivo". Vitamin A deficiency in rats has been found to be associated with an increase in serum concentrations of thyroid hormones (3). On the other hand, massive doses of vitamin A seem to inhibit thyroid function in man (4). In order to verify the possible role of RA in the regulation of growth and differentiation of the human thyroid we examined in primary cultures of human follicular cells the effect of RA on 3HCorresponding Author: Prof. E.C. degli Ubel~i, Endocrinology Section, Institute of Internal Medicine, University of Fen'ara, via Savonarola 9, 1-44100 Felxara, Italy. 0024-3205/93 $6.00 + .00 Copyright © 1993 Pergamon Press Ltd All rights reserved.
1040
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
thymidine (3H-thy) incorporation and mRNA expression of thyroglobulin (TGB) and peroxidase (TPO), the main markers of thyroid differentiation and function (5,6). Furthermore, to clarify the mechanism by which RA could act on thyroid ceils, we also investigated the mRNA expression of RA receptors (RARs) by measuring the mRNA levels of RARct and B, as they appear to have a relatively wider distribution than RAR ~ in mouse tissues (7). The expression of these genes was analysed in basal conditions and in response to RA and/or to TSH, the main regulator of thyroid growth and function (6,8). Methods Cell culture Follicles from 10 normal, 10 goitrous and 5 adenomatous human thyroid samples were prepared by collagenase (type IV, Boehringer Mannheim) and trypsin treatment as previously described (9). Normal tissues were taken from a normal area of thyroid glands from patients (8 women and 2 men) undergoing surgery for solitary non functioning nodules. Goitrous and adenomatous tissues were obtained from patients undergoing surgery for a non toxic multinodular goiter (7 women and 3 men) and for a follicular adenoma (4 women and 1 man) respectively. All patients were euthyroid and did not receive thyroid holinone therapy before surgery. Follicles were resuspended in RPMI 1640 medium (Gibco, Grand Island, NY) with 5% foetal calf serum (FCS, FLOW, Meckenheim), distributed in 24-well culture plates (16 mm diameter) or in plastic 250 ml flasks (Costar, Cambridge, Ma), whatever was more appropriate for the experimental protocol, and grown at 37°C in a humidified atmosphere of 5% CO2. Experiments were carried out within 5-10 days of plating. For 3H-thy incorporation analysis, cells, cultured in 24-well plates as indicated, were incubated with either serum free medium plus 0.4% bovine serum albumine (BSA) or 2% FCS and all-trans retinoic acid (RA, Sigma, Chemical Co.,St. Louis,MO) at various concentrations and for different times, according to the experimental protocol. RA from 10 mM stock solution in ethanol was further diluted with RPMI 1640 medium to final concentration as required by the protocol. The final concentration of ethanol was the same in both control and RA-treated cultures. Cell viability was determined by their adherence to plastic dishes and by the trypan blue dye exclusion technique. All experiments were performed in triplicate. For RNA analysis, cells, cultured in 250 ml flasks, as indicated, were mantained for two days in 2% FCS and then were incubated with 2% FCS alone or with RA (1, 10 ~tM) and/or with bovine TSH (10 mU/ml, Ambinon, Organon). DNA synthesis 3H-thy incorporation was measured in adherent cells (approximately 105 cells/well) in 24-multi-well plates obtained as detailed in the previous section. Cells were incubated with 0.4 % BSA, 3H-thy (1.5 I.tCi/ml, 87 Ci/mmol) as well as with and without RA for 24, 48 and 72 h, unless otherwise indicated. 3H-thy was added at time 0. After the incubation time, cells were washed three times with ice cold phosphate buffer saline (PBS) and twice with 10% ice-cold trichloro-acetic acid (TCA). TCA precipitable material was solubilized in 500 I.tl 0.2 M sodium hydroxide (NaOH) and 0.1% sodium dodecyl sulphate (SDS) as previously described (10). Cells associated radioactivity was then counted in a scintillation spectrometer. Results (count per min/well) were obtained in triplicate experiments.
Vol. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
1041
In addition, in a parallel set of experiments, the DNA content of cell monolayers was evaluated as an estimation for cell number under the same concentrations. Cells were harvested by scraping, washed in saline and pooled by centrifugation at 2000 rpm for 10 min. Cells were lysed in 300 B1 of 0.5% NP40 solution and spun at 12000 rpm for 5 min. DNA from nuclei was obtained by RNAse and proteinase K treatment, then phenol-chloroform extracted (11). DNA content was estimated by optical density at 260 nm and expressed as ~g/well. 3H-leucine incorooration into thvroid oroteins In order to determine the effect of RA on protein synthesis, 3H-leucine incorporation into thyroid cells was evaluated. Adherent cells in 24-well culture plates (as for 3H-thy incorporation) were incubated without and with 0.01, 0.1, 1, 10 and 50 ~tM RA respectively, in presence of 0.4% BSA and 3H-leucine (20 BCi/ml, 141 Ci/mmol). After 24 h of incubation at 37 ° C, the labelling was stopped by removing the medium and washing the cells with ice-cold PBS, pH 7, and twice with ice-cold TCA. Radioactivity into the TCA-precipitable material was determined as for 3H-thy incorporation. Results (count per min/well) were obtained in triplicate experiments. RNA det¢rmin~ion, After incubation for 48 h in 250 ml flasks, adherent cells were harvested and lysed in 0.5% NP40 solution (containing the RNA inhibitor vanidylribonucleoside, Sigma), as described (12). Total RNA was isolated from cytoplasm by phenol-chloroform extraction, subdivided in three aliquots and ethanol precipitated. One aliquot per flask was used for Northern blot analysis. RNA (8-30 Bg) was denatured in formaldehyde, electrophoresed in 1% agarose-0.66 M formaldehyde gel with 0.5 Bg/ml ethidium bromide in 20 mM 3-morpholinopropansulphonic acid, pH 7 running buffer and transferred onto Gene Screen Plus membrane (NEN, Boston, MA). Filters were hybridized for 18-36 h at 42 ° C in 50 % formamide and dextrane sulphate with 32p_ labeled probes. The DNA probes were: pH23, a 1300 base pair fragment of human TGB c-DNA; pH2.8, a TPO c-DNA; pAct2, an actin c-DNA probe; pTM61A and pSKR13, the o~ and 13 RAR cDNA probes. The probes were labeled with 32p dCTP (3000 Ci/mmol) by multiprime DNAlabeling system (Boehringer, Mannheim, Germany). After hybridization and washing, filters were exposed to Kodak X OMAT films at -70°C with Dupont intensifying screens. The levels of TGB, TPO and RAR t~ RNA in each sample were measured by densitometric analysis of the autoradiographic signals (12). Scanning values were related to those of actin RNA and expressed as relative scanning units which reflect the relative abundance of each mRNA in a given RNA sample. Analvtical orocedure. Data are expressed as the mean + SEM. Statistical analysis of the data was performed using the unpaired and paired Student's t test, as applicable. Results Effects of RA on DNA synthesis To find out the effect of RA on DNA synthesis of human thyroid cells, 3H-thy incorporation was measured in cells treated for 48 h with increasing concentration of RA (0.01, 0.1, 1, 10, and 50 BM). As shown in Fig. 1A, RA induced a significant decrease of 3H-thy incorporation at all concentrations. 1 ~tM RA was more inhibiting than 10~tM (45.9+5.1 and 39.9 + 7.25 % inhibitions respectively after 1 and 10 I.tM RA) but maximal decrease occurred at 50 ~tM (64.65 + 8.7 % inhibition). However, at this maximal concentration, cells exhibited an abnormal shape,
1042
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
round in form, suggestive of a RA cytotoxic effect for these cells. No changes in shape, adherence and trypan blue staining were observed by phase-contrast microscope between cells treated with 0.01, 0.1, 1 and 10 ~tM RA. As shown in Fig. 1B, the inhibitory effect of RA on 3H-thy incorporation of human thyroid cells was also reflected by parallel changes in the DNA content of cells monolayers (22 _ 2.6%, 20 + 2.4% and 57 + 4.9% inhibitions respectively after 1, 10 and 50 I.tM RA). In cells treated for 24 h with RA, a reduction in protein synthesis measured by 3H-leucine incorporation was observed only at 50 IxM concentration (Fig. 1C). 120 i00
~
~o
~'~ # ~_~
20 o
-r
48
T
A
-r
6O
B
5O
~
40
~
2o
120
11InNl'p
i00
--
lO o
24
I"
-r
"1-
I-
l
i0
C
0
-'~ o
80
•
40 ~
"~--~
20
o m
0
0.010.i
50
R A ktM
FIG. 1. Effect of RA on DNA and protein synthesis in human thyroid follicular cells. A: Cells from cultures of goitrous thyrocytes were incubated for 48 h in serum-free medium, plus 0.4% BSA, with RA at various concentrations and with 3H-thy. The mean values (cpm/well) + SEM in RA-treated cells were compared to those of untreated cells and expressed as percent of control values. B: DNA content in cells cultured for 48 h was measured as described in Materials and Methods. The mean values in I.tg/well (+SEM) were shown. C: 3H-leucine incorporation was measured in cells cultured for 24 h (see Methods for technical details). The results were obtained from three independent experiments in triplicate. *, **, ***, p< 0.01, 0.05 and 0.001 respectively vs. untreated cells.
Vol. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
1043
B
A =
5-~
O
O
"~
4
4
= , -•"7.
3
~
-V. 3
1 g M RA ;~'~
o~,
normal
adenoma
q-
T
O
1
;~'~ ,a:z
1
!
0
I
!
I
24
48 Hours
72
0
|
0
1 RA
i
0 l.tM
1
FIG. 2. Effect of 1 p-M RA on 3H-thy incorporation in human normal and adenomatous thyrocytes. A: Time course study of 3H-thy uptake into normal cells incubated in serum free medium plus 0.4% BSA,with (A) and without (A) RA (1 p.M) for 24, 48 and 72 h. Results are the mean (+SEM) of a triplicate experiment. B: Effect of RA (lp-M) on 3H-thy incorporation in thyrocytes isolated from ten normal thyroids and six adenomas. Empty (0) and filled (1) columns represent untreated and RA-treated cell cultures. 3H-thy incorporation was measured as radioactivity in TCA-precipitable material, as described in Materials and Methods, and expressed as the mean (_+SEM) value of cpm/well. *, **, p< 0.05 and 0.01 respectively vs. untreated cells.
To study the effects of RA after different periods of incubation, 1 p-M concentration of RA was chosen, because the percent reduction in RA treated cells versus untreated control cells was at 1 p-M slightly higher than that observed at 10 p-M. Similar RA-induced reduction was also observed in presence of 2% FCS. As shown in Fig.2A, the concentration of 1 gM RA reduced 3H-thy incorporation in normal thyroid after both 48 h (from 35760 + 650 in control untreated cells to 23150 + 1400 cpm/well in RA-treated cells, p< 0.01) and 72 h (from 45120 + 3440 in untreated cells to 25270 + 592 cpm/well in RA-treated cells, p< 0.01). As indicated in Fig.2B, the analysis of 3H-thy incorporation in thyrocytes obtained from 10 normal thyroids and 6 adenomas showed that the 48 h treatment with 1 p-M RA caused a significant decrease (p< 0.01) in 3H-thy incorporation also in adenomatous cells. In adenomas the mean + SEM percent RA-induced reduction in 3H-thy incorporation was 42.7 + 3.2 % compared with that of 28.5 + 4.6 % obtained in normal thyroid.
1044
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
B
A
Kb
TPO
"--3.2
TGB
4 " - 8.7
---- 1.8
ACTIN TSH RA 1 RAIO
+ +
-+
+
+
+ --
-+
mRNA a r b i t r a r y units TSH RA RA +TSH TPO Actin
5.06 ±0.41
50.65 +11.94
3.33 ~ +_0.53
15.60 + *_5.70
TGB
2.19 +_0.20
14.13 ±5.19
2.09 ±0.32
11.82 ±4.24
Actin
FIG.3 Northern blot analysis of TPO and TGB mRNA transcripts in human thyrocytes treated with RA. A: Cells isolated from a goiter were cultured as indicated in Materials and methods and were incubated for 2 days in medium supplemented with 2% FCS, without (-) and with (+) TSH (10 mU/ml), RA (1 or 10 pM). Northern blotting was carried out using one aliquot of total RNA isolated from each 250 ml flask. RNA was electrophoresed on agarose-formaldehyde gel, then transfen'ed to a nylon membrane filter (Gene Screen Plus, NEN) and hybridized with 32p labelled TPO (upper panel) and TGB (middle panel) probes. Since different amounts of total RNA have been loaded on the gel, after appropriate exposure, the blots were rehybridized with the actin probe (lower panel) to check the amount of RNA loaded in the different lanes. The ratio betweeen the intensity of TPO and TGB mRNA and that of actin mRNA, as judged from the densitometric scanning of the con'esponding lanes in the autoradiography, was calculated in RNA blots of 5 thyrocyte cultures (2 normal thyroids and 3 goiters) treated with 1 laM RA (details of method are described in Materials and Methods). The mean (±SEM) values of the relative scanning units are given in B and expressed as mRNA arbitrary units. *, p< 0.05 vs. (-); +, p< 0.01 vs. (TSH) using the paired Student's t test.
Effect of RA on TPO and TGB mRNA levels. Northern blot analysis of TPO and TGB mRNA in Fig.3 showed that in thyrocytes, as expected, the levels of TPO and TGB mRNA markedly increased after treatment with TSH (10 mU/ml) for 48 h. When cells were incubated in the presence of RA (1 and 10 p.M), the TSH effect on TPO mRNA levels sharply decreased at both RA concentrations (Fig. 3A, upper panel). A slight reduction in both basal and TSH-induced TGB mRNA levels by 1 and 10 pM RA was also observed (Fig. 3A, middle panel). The densitometric analysis (TPO/actin and TGB/actin mRNA ratio, Fig. 3B) of Northern blots of RNA isolated from 5 cultures (2 normal thyroids and 3 goiters) confirmed these results and indicated that both basal and TSH-induced levels of TPO mRNA were significantly reduced by 1 ~tM RA, whereas basal and TSH-induced TGB mRNA levels were not significantly modified by RA.
Voi. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
NORMAL
A
GOITER I
1045
ADENOMA
I
I
I
!
Kb 4 3.5 2.8 2
RARr~
ACTIN
1.8
TSH
- + - +
RA 1
- - - - -I'- "t-
--+
++ +--+-+--+
RA10
B
+
m
-I-
Kb --,---4 •",--" 3.5 - - - 2.8
RARa
-.--.-2
--28S RARfl --18
-,,- 1.8
ACTIN
TSH RA1 RAIo
S
--
4"
~ 4-
~
4-I--
-I-
FIG. 4 Northern blot analysis of RAR ~ mRNA transcripts in human thyrocytes treated with RA. A: Cells isolated from normal, goitrous and adenomatous cells were cultured and processed as indicated in Fig.3 and in Materials and Methods. RNA was hybridized with 32p labelled RAR ct (upper panel) and actin probes. B: RNA isolated from goitrous cultured cells was hybridized with 32p labelled RAR tx (upper panel), RAR 13 (middle panel) and actin probes (lower panel).
1046
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
RAR g mRNA expression in thyrocytes The expression of RA receptors was examined by hybridizing the RNA samples of the cultured cells with two probes which can specifically detect RAR c~ and 13 transcripts (7). The Northern blot analysis of RNA isolated from normal, goitrous and adenomatous cells showed that after hybridization with the RAR ct probe two minor RNA species of approximately 4 and 2 Kb in size, and two major RNA species of approximately 3.5 and 2.8 Kb (Fig. 4 A and B, upper panels) were detected. The two major RARct mRNA species corresponded well to those shown in other human tissues in previous studies (7). After hybridization with the RAR 13probe (Fig. 4B, middle panel), no transcripts were detected in these samples, as well as in the other RNA samples examined (data not shown). Densitometric data (RAR c~/actin mRNA ratio) of Northern blots of RNA isolated from 5 cultures (2 normal thyroids and 3 goiters), indicated no significant variation in basal R A R c t mRNA levels after u'eatment with 1 p.M RA, TSH and with the combined treatments, in human thyroid cell cultures (data not shown). Discussion We assessed the effects of RA on DNA synthesis and TGB and TPO mRNA expression in human thyroid cells cultured in vitro. The reduction of 3H-thy incorporation by all-trans-RA suggests that this retinoid mainly at 1 I.tM concentration exerts an inhibitory action on DNA synthesis and retains a growth inhibitory activity on human thyroid, as observed in other cell systems (1,13). Although the inhibitory action of RA on 3H-thy incorporation was detected also at the dose of 10 ktM, which has been found to be cytotoxic in other cell lineages, such as breast and melanoma cell lines (14,15), no changes in 3H-leucine incorporation nor in shape and adherence, suggestive of RA cytotoxic effect on thyroid cells, were observed. In addition, RA did not appear to be cytotoxic at these doses in other human cell systems including thyroid carcinoma (16) or rhabdomyosarcoma (13). Our findings are consistent with the inhibitory effect of either 13-cis-RA or retinyl acetate at 10 I~M concentration on 3H-thy uptake observed in a human follicular carcinoma cell line (16) both in basal and in EGF-stimulated conditions. However, our data do not agree with that observed in cultures of porcine thyroid cells (17) in which retinol (1 ktM) did not modify basal 3H-thy incorporation whereas it increased EGF- or IGF-Iinduced 3H- thy incorporatiol~. Although this difference in RA activity on thyroid growth is difficult to explain, we suggest that RA could possess divergent effects, growth-stimulating as well as growth-inhibiting properties, depending on the different responsiveness to RA of cell cultures from different species (humans vs. pigs). The present study shows that RA inhibits basal and TSH-stimulated TPO mRNA levels both in normal and in goitrous cells. This is in agreement with the reduction of serum thyroid hormones observed "in vivo" using massive doses of vitamin A in the treatment of hyperthyroidism (4) and also with the inhibition of both TSH-induced TPO mRNA levels and iodide uptake by retinol in cultures of porcine thyroid cells (17). The biochemical mechanism leading to the reduction of TPO mRNA levels by RA at present is still unknown. Such modulation of gene expression may be achieved either by alteration at transcriptional or posttranscriptional level, including RNA processing and mRNA stability. In addition, this may results by interfering with TSH-responsive signal transduction pathways, which
Vol. 53, No. 13, 1993
Retinoic Acid and Human Thyroid Growth
1047
control both proliferation and gene expression in thyroid cells (18), or by a cis direct action of RA on TPO gene promoter. Since we demonstrated that RA inhibits not only TSH-induced TPO mRNA levels, but also those found in basal conditions, we suggest that RA may act independently from TSH or TSH-receptor. According to this are data reported on porcine cells (17) where cAMP response to TSH is not modified by retinol. Since TPO gene transcription is stimulated by TSH via a cAMP pathway through a cAMP responsive element present in its promoter (6), it is attractive to speculate that also a RA responsive element could be present upstream the TPO gene. Such an element, binding a RA-receptor complex, could negatively affect gene u'anscription. The inhibition of TPO gene expression by RA here reported is somewhat at variance with findings obtained in a human follicular carcinoma cell line where RA reduces cell proliferation, but increases 1251 uptake, thus driving tumor cells toward a more differentiated state (16). However, this may be explained by previous observations indicating that retinoid effects can vary according to the stage of differentiation of the responding cells. In relatively non differentiated cells, such as human myeloid leukemia cells (19), RA may reduce cell proliferation inducing a more differentiated state. On the other hands, in keratinocytes retinoids inhibit the expression of differentiated functions (20). In particular, in non neoplastic hepatocytes RA inhibits both proliferation and function (21) In contrast to the apparent inhibitory action of RA on TPO mRNA levels, no significant effect was observed on TGB expression. This suggests that TGB gene expression is not affected by RA. This different expression pattern is not an unusual finding in the thyroid. In fact, it has been shown that different thyroid specific functions, including iodide uptake, TPO and TGB mRNA levels may exibit different sensitivity to hormonal regulation and oncogene transformation (22): the most sensitive is iodide uptake and the less sensitive is TGB gene expression. This assumption agrees with the observations that maintenance of differentiated functions in cultured thyroid cells results from complex interactions among various factors and that different cellular processes are independently implicated in the modulation of TGB and TPO gene expression (5,6). Our "in vitro" results suggest that human thyrocytes possess specific receptors that may mediate RA action. At present there are no data on RAR gene expression in thyroid cells. Our investigation shows that RAR ct transcripts are present in normal, goitrous and adenomatous human thyroid cells, whereas levels of expression of RAR 13gene cannot be detected. In conclusion, our data demonstrate that the specific retinoic acid receptor tx gene is expressed in human thyrocytes and that RA may be involved in the regulation of the human thyroid by reducing proliferation and function, but not differentiation of follicular cells. However, further investigation of differential regulation of RAR gene expression by RA may provide a deeper insight into the action of RA on cell differentiation and growth.
Acknowledgments We would like to thank the Drs. F. Mavilio, E Chambon, G. Vassart, S. Kimura and J.L. Mandel for giving us the RARs, TGB, TPO and actin probes, respectively. We thank Dr. B. Anderson for correcting the English, Work supported in part by MURST 40-60 %, CNR PFs Biotec., ACRO, by AIRC and Regione Emilia Romagna.
1048
Retinoic Acid and Human Thyroid Growth
Vol. 53, No. 13, 1993
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
13 14. 15 16. 17. 18. 19. 20. 21. 22.
AB. ROBERTS and M B . SPORN, The Retinoids. Academic, Orlando, FL. 2209-286 (1984). LM. DE LUCA, FASEB J. 52924-2933 (1991). H. GARCIN, P. HIGUERET, J. Endocrinol. 84135-140 (1980). S. SIMKINS, J. Clin. Endocrinol. Metab. 7574-585 (1947). Y. NAGAYAMA, S. YAMASHITA, H.HIRAYU, et al, J. Clin. Endocrinol. Metab. 6811551159 (1989). CM. GERARD, A. LEFORT, D. CHRISTOPHE, et al.,Mol. Endocrinol.3 2110-2118 (1989). A. KRUST, P. KASTNER, M. PETKOVICH, A. ZELENT, P. CHAMBON,.Proc. Natl. Acad. Sci. USA. 865310-5314 (1989). P. ROGER, M. TATON, J. VAN SANDE, JE. DUMONT, J. Clin. Endocrinol, Metab.6....66 1158-1165 (1988). L. DEL SENNO, EC. DEGLI UBERTI, S. HANAU, R. PIVA, R. ROSSI, G. TRASFORINI, Mol. Cell. Endocrinol..6..~367-74 (1989). GP. SCHINCAGLIA, F. FORNITI, R. CAVALLINI, R. PIVA, G. CALURA, L. DEL SENNO, J.Oral Pathol. Med.. 2...! 181-185 (1992). J. SAMBROOK, E.T. FRITSCH and E MANIATIS, . Molecular Cloning. CSH Laboratory press, 29.16-9.19 (1989). EC. D E G L I UBERTI, S. HANAU, R. ROSSI, R. PIVA, A. M A R G U T T I , G. TRASFORINNI, G. PANSINI, L. DEL SENNO, J. Clin. Endocrinol. Metab. 7..221364-1371 (1991). GD. CROUCH,LJ. HELMAN, Cancer Res. 514882-4887 (1991). H. UEDA, M. ONO, Y. HAGINO, M. KUWANO, Cancer Res. 4..~_53332-3338 (1985). R. LOTAN, T. STOLARSKY, D. LOTAN, Cancer Res.4_.3_32868-2875 (1983). AJ. VAN HERLE, ML. AGATEP, DN. PADUA III, TL. TOTANES, DV. CANLAPAN, HML. VAN HERLE, and GJE JUILLARD, J. Clin. Endocrinol. Metab 71755-763 (1990). M. ARAI, T. TSUSHIMA, O. ISOZAKI, K. SHIZUME, I. EMOTO, H. DEMURA, M. MIYAKAWA, and N. ONODA, Endoclinology 1292827-2833 (1991). JE. DUMONT, E LAMY, P. ROGER, and C. MAENHAUT, Physiol. Rev. 72667-697 (1992). TR. BREITMAN, SE. SELONICK and SJ. COLLINS, Proc. Natl. Acad. Sci. USA. 7729362940 (1980). E. FUCHS and H. GREEN, Cell 2..5_5617-624 (1981). H. IKEDA and K. FUJIWARA, Biochem. Biophys. Res. Comm. 191675-680 (1993). RZ. ZARRILLI, S. FORMISANO and B. DI JESO, Molec. Endocrinol., _439-45 (1990).