165
Mechanisms of Development, 36 (1992) 165-172 © 1992 Elsevier Scientific Publishers Ireland, Ltd. 0925-4773/92/$05.00 MOD00074
Aggregation and cell cycle dependent retinoic acid receptor mRNA expression in P19 embryonal carcinoma cells Luigi J.C. Jonk, Marjolijn E.J. de Jonge, Frank A.E. Kruyt, Christine L. Mummery, Paul T. van der Saag and Wiebe Kruijer Hubrecht Laboratorium, Netherlands Institute for Developmental Biology, Utrecht, The Netherlands (Received 11 July 1991; revision received 15 November 1991; accepted 18 November 1991)
Differentiation of P19 EC cells along different pathways into derivatives resembling cells of the three embryonic germ layers is accompanied by characteristic differences in modulation of expression of each of the three retinoic acid receptor genes, RARa, -/3 and "3'. Differentiation induced by addition of RA to P19 EC cells cultured in monolayer is accompanied by a rapid increase in expression of both RARa and -/3. Induction of RAR/3 occurs in a characteristic biphasic manner, suggesting that multiple factors a n d / o r different mechanisms are involved in controlling its expression. RAR/3 mRNA is induced to a far higher level during early aggregation in the presence of RA than during early differentiation in monolayer, suggesting that the direction of differentiation depends on the number a n d / o r ratio of a and/3 type of RA receptors. Aggregation of PI9 EC cells in the presence of RA, but not DMSO, is accompanied by repression of RAR3", suggesting that the expression of RAR~ and RAR3" during neuroectodermal differentiation is mutually exclusive. The effects of RA on RAR expression are significantly greater in G1 than in S-phase of the cell cycle. These results extend previous observations that commitment to differentiation is cell cycle dependent and indicates that critical target gene regulation in response to RA has to take place in G1 for differentiation to occur. Retinoic acid receptor; Differentiation; Embryonal carcinoma cell; Cell cycle; Development
Introduction The vitamin A derivative retinoic acid (RA) has been implicated to play a central role in several processes occurring during vertebrate development. Gradients of this candidate morphogen are believed to provide positional information necessary for the proper formation of the nervous system (Durston et al., 1989) and extremities (Eichele, 1989). In vitro, RA can have a profound effect on cell proliferation and is a potent inducer of differentiation of several cell types (Amos and Lotan, 1990), notably embryonal carcinoma (EC) and embryonic stem (ES) cells (Strickland and Mahdavi, 1979; Mummery et al., 1990). The isolation of three highly homologous subtypes of the retinoic acid receptors (RARs) a, -/3, -3' and their isoforms (Leroy
Correspondence to: W. Kruijer, Hubrecht Laboratorium, Netherlands Institute for Developmental Biology, Uppsalalaan 8, 3584 CT Utrecht, The Netherlands.
et al., 1991; Zelent et al., 1991; Kastner et al., 1990) has greatly enlarged our insight into the mechanism of action of RA. Analogous to the mode of action of other members of the nuclear steroid/thyroid hormone receptor group, RARs act as ligand-inducible transcription factors capable of transactivating target genes containing cognate DNA recognition sites, like the RAR/3 gene (De Th6 et al., 1990) and genes previously known to be regulated by RA, e.g., laminin B1 (Vasios et al., 1991). Embryonal carcinoma cells (EC cells) have been widely used as a model system for studying early murine development. These cells resemble those comprising the inner cell mass of preimplantation blastocysts and a number of differentiated cell types, with characteristics of cells of the three germ layers can be obtained reproducibly by different treatments (Martin, 1980). Addition of RA at 10 - 6 M to monolayer cultures of the pluripotent murine EC cell line P19 induces differentiation to a heterogeneous population consisting of endoderm- and mesoderm-like cells~(Mummery et al.,
166 1986; Roguska and Gudas, 1985). The endogenous serum RA level ( ~ 10 -1° M) is sufficient to induce differentiation into a similar cell type, when the cells are aggregated in suspension to form embryoid bodies. Addition of RA at 10 - 6 M or 1% DMSO to aggregating P19 cells induces neuronal or mesodermal differentiation, respectively (McBurney et al., 1982). The importance of RARs as mediators of the differentiation inducing effect of RA has been demonstrated by the analysis of a non-differentiating mutant of P19, RAC65, carrying a mutated R A R a allele which encodes a truncated receptor acting as a trans-dominant negative variant (Pratt et al., 1990; Kruyt et al., in press). Previous investigations have indicated that the commitment to differentiation by RA in P19 EC cells is cell cycle dependent (Mummery et al., 1987). A 4 h pulse treatment with RA of cells in the G1 phase of the cell cycle, but not in S or asynchronously growing, is sufficient to induce their differentiation as determined by a reduction of their ability to grow in soft agar and expression of SSEA-1 antigen. In this report, we have investigated modulation of expression of the R A R a , -/3 and -y genes in P19 EC cells induced to differentiate along different pathways in endodermal, mesodermal and neuroectodermal derivatives. In addition, we examined the cell cycle dependence of R A R expression in relation to commitment to differentiation. Our results show that aggregation of P19 EC cells potentiates RA-induced RAR/3 expression, while it has a distinct negative effect on steady-state R A R y mRNA levels. Furthermore, a correlation between R A R a and -/3 inducibility and cell cycle dependent RA induced differentiation is demonstrated. The observed modulation of expression of R A R genes will be discussed in relation to the regulation of these genes as well as their role in cell differentiation and development.
A 8
CID
F
E
I
Fig.
1. S c h e m a t i c
representation
of RAR
mRNA
ETA
I
sequence.
The
encoded RAR functional domains are designated by A-F. Alpha, beta and gamma denote the relative position and length of sequences from the corresponding RARa, -/3 and -y cDNAs used to generate complementary RNA for RNAase protection analysis.
Differentiation in monolayer In monolayer culture, P19 EC cells differentiate into endoderm- and mesoderm-like cells when treated with 10 - 6 M RA. The results of the RNAase protection analysis on RNA isolated from asynchronously growing cells treated with RA is shown in Figs. 2 and 3. At 24 h after plating, RA was added and cells were lysed after 0-12 h (Figs. 2 and 3A-C) or 0-5 days (Figs. 3 D - F ) of incubation. Lanes 1-3 show the portion of the R A R a , -/3 and -3' probes protected from RNAase T1 digestion following hybridization of each probe individually with 10 /zg of total RNA from P19 cells treated for 24 h with RA. As can be seen from the autoradiogram (Fig. 2) and the quantification data (Fig. 3) the amount of R A R a mRNA exceeds that of RAR/3 and -% The level of R A R y mRNA expression remains fairly constant during both short (0-12 h) as well as prolonged CONTROLS
)~ /~
a
HOURS IN RA
p
t
0
'~
',~ 1
V,~ 2 2',~ 3
4
5
6
8
10
12
m
Results
P19 EC cells can be induced to differentiate along different pathways into endodermal, mesodermal and neuroectodermal derivatives. We analyzed R A R a , -/3 and -3' mRNA expression in response to the different modes of differentiation by RNAase protection. The R A R probes used in the protection experiments, were derived from regions of their cDNAs exhibiting high sequence divergence, thus excluding cross-detection during simultaneous RNAase protection (Fig. 1). A probe detecting transcripts from the uniformly expressed glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was included as an internal reference to which all quantification data were normalized.
;11 1
2
3
4
-Ol t
-/31 5
6
7
8
9
10 11 12 13
14 15 16
17
18 19
Fig. 2. Modulation of RAR mRNA expression during RA induced differentiation of asynchronously growing P19 cells. Autoradiogram showing the results from RNAase protection analysis, a, /3 and y (lanes 1-3) denote the position of RARa, -/3 and -y specific protected fragments ( f ) following hybridization of each probe to 10/xg total RNA from 24 h 10 -6 M RA treated P19 EC cells. Lane 4: mixed undigested RARa, -/3 and -y hybridization probes (top: y, middle: /3, bottom: a probe). Lane 5: hybridization of mixed a, /3 and 3' probes with tRNA followed by digestion with RNAase T1. Lanes 6-19: protected fragments resulting from hybridization of the mixed a,/3 and 3' probes with 10 p.g total RNA from cells incubated with RA for the times indicated and RNAase T1 digestion.
167 after 3 h and a second rise commences 8 h after RA addition (Fig. 3B). Prolonged exposure to RA further induces RAR/3 mRNA to a maximum at four days of
(up to 5 days) exposure of the cells to RA. In contrast, RAR/3 mRNA is rapidly (within 1 h) induced, apparently in a two step fashion: a first maximum is reached
120
40
D
A
T
100
_
Mechanisms of Development, 36 (1992) 165-172 © 1992 Elsevier Scientific Publishers Ireland, Ltd. 0925-4773/92/$05.00 MOD00074
Aggregation and cell cycle dependent retinoic acid receptor mRNA expression in P19 embryonal carcinoma cells Luigi J.C. Jonk, Marjolijn E.J. de Jonge, Frank A.E. Kruyt, Christine L. Mummery, Paul T. van der Saag and Wiebe Kruijer Hubrecht Laboratorium, Netherlands Institute for Developmental Biology, Utrecht, The Netherlands (Received 11 July 1991; revision received 15 November 1991; accepted 18 November 1991)
Differentiation of P19 EC cells along different pathways into derivatives resembling cells of the three embryonic germ layers is accompanied by characteristic differences in modulation of expression of each of the three retinoic acid receptor genes, RARa, -/3 and "3'. Differentiation induced by addition of RA to P19 EC cells cultured in monolayer is accompanied by a rapid increase in expression of both RARa and -/3. Induction of RAR/3 occurs in a characteristic biphasic manner, suggesting that multiple factors a n d / o r different mechanisms are involved in controlling its expression. RAR/3 mRNA is induced to a far higher level during early aggregation in the presence of RA than during early differentiation in monolayer, suggesting that the direction of differentiation depends on the number a n d / o r ratio of a and/3 type of RA receptors. Aggregation of PI9 EC cells in the presence of RA, but not DMSO, is accompanied by repression of RAR3", suggesting that the expression of RAR~ and RAR3" during neuroectodermal differentiation is mutually exclusive. The effects of RA on RAR expression are significantly greater in G1 than in S-phase of the cell cycle. These results extend previous observations that commitment to differentiation is cell cycle dependent and indicates that critical target gene regulation in response to RA has to take place in G1 for differentiation to occur. Retinoic acid receptor; Differentiation; Embryonal carcinoma cell; Cell cycle; Development
Introduction The vitamin A derivative retinoic acid (RA) has been implicated to play a central role in several processes occurring during vertebrate development. Gradients of this candidate morphogen are believed to provide positional information necessary for the proper formation of the nervous system (Durston et al., 1989) and extremities (Eichele, 1989). In vitro, RA can have a profound effect on cell proliferation and is a potent inducer of differentiation of several cell types (Amos and Lotan, 1990), notably embryonal carcinoma (EC) and embryonic stem (ES) cells (Strickland and Mahdavi, 1979; Mummery et al., 1990). The isolation of three highly homologous subtypes of the retinoic acid receptors (RARs) a, -/3, -3' and their isoforms (Leroy
Correspondence to: W. Kruijer, Hubrecht Laboratorium, Netherlands Institute for Developmental Biology, Uppsalalaan 8, 3584 CT Utrecht, The Netherlands.
et al., 1991; Zelent et al., 1991; Kastner et al., 1990) has greatly enlarged our insight into the mechanism of action of RA. Analogous to the mode of action of other members of the nuclear steroid/thyroid hormone receptor group, RARs act as ligand-inducible transcription factors capable of transactivating target genes containing cognate DNA recognition sites, like the RAR/3 gene (De Th6 et al., 1990) and genes previously known to be regulated by RA, e.g., laminin B1 (Vasios et al., 1991). Embryonal carcinoma cells (EC cells) have been widely used as a model system for studying early murine development. These cells resemble those comprising the inner cell mass of preimplantation blastocysts and a number of differentiated cell types, with characteristics of cells of the three germ layers can be obtained reproducibly by different treatments (Martin, 1980). Addition of RA at 10 - 6 M to monolayer cultures of the pluripotent murine EC cell line P19 induces differentiation to a heterogeneous population consisting of endoderm- and mesoderm-like cells~(Mummery et al.,
166 1986; Roguska and Gudas, 1985). The endogenous serum RA level ( ~ 10 -1° M) is sufficient to induce differentiation into a similar cell type, when the cells are aggregated in suspension to form embryoid bodies. Addition of RA at 10 - 6 M or 1% DMSO to aggregating P19 cells induces neuronal or mesodermal differentiation, respectively (McBurney et al., 1982). The importance of RARs as mediators of the differentiation inducing effect of RA has been demonstrated by the analysis of a non-differentiating mutant of P19, RAC65, carrying a mutated R A R a allele which encodes a truncated receptor acting as a trans-dominant negative variant (Pratt et al., 1990; Kruyt et al., in press). Previous investigations have indicated that the commitment to differentiation by RA in P19 EC cells is cell cycle dependent (Mummery et al., 1987). A 4 h pulse treatment with RA of cells in the G1 phase of the cell cycle, but not in S or asynchronously growing, is sufficient to induce their differentiation as determined by a reduction of their ability to grow in soft agar and expression of SSEA-1 antigen. In this report, we have investigated modulation of expression of the R A R a , -/3 and -y genes in P19 EC cells induced to differentiate along different pathways in endodermal, mesodermal and neuroectodermal derivatives. In addition, we examined the cell cycle dependence of R A R expression in relation to commitment to differentiation. Our results show that aggregation of P19 EC cells potentiates RA-induced RAR/3 expression, while it has a distinct negative effect on steady-state R A R y mRNA levels. Furthermore, a correlation between R A R a and -/3 inducibility and cell cycle dependent RA induced differentiation is demonstrated. The observed modulation of expression of R A R genes will be discussed in relation to the regulation of these genes as well as their role in cell differentiation and development.
A 8
CID
F
E
I
Fig.
1. S c h e m a t i c
representation
of RAR
mRNA
ETA
I
sequence.
The
encoded RAR functional domains are designated by A-F. Alpha, beta and gamma denote the relative position and length of sequences from the corresponding RARa, -/3 and -y cDNAs used to generate complementary RNA for RNAase protection analysis.
Differentiation in monolayer In monolayer culture, P19 EC cells differentiate into endoderm- and mesoderm-like cells when treated with 10 - 6 M RA. The results of the RNAase protection analysis on RNA isolated from asynchronously growing cells treated with RA is shown in Figs. 2 and 3. At 24 h after plating, RA was added and cells were lysed after 0-12 h (Figs. 2 and 3A-C) or 0-5 days (Figs. 3 D - F ) of incubation. Lanes 1-3 show the portion of the R A R a , -/3 and -3' probes protected from RNAase T1 digestion following hybridization of each probe individually with 10 /zg of total RNA from P19 cells treated for 24 h with RA. As can be seen from the autoradiogram (Fig. 2) and the quantification data (Fig. 3) the amount of R A R a mRNA exceeds that of RAR/3 and -% The level of R A R y mRNA expression remains fairly constant during both short (0-12 h) as well as prolonged CONTROLS
)~ /~
a
HOURS IN RA
p
t
0
'~
',~ 1
V,~ 2 2',~ 3
4
5
6
8
10
12
m
Results
P19 EC cells can be induced to differentiate along different pathways into endodermal, mesodermal and neuroectodermal derivatives. We analyzed R A R a , -/3 and -3' mRNA expression in response to the different modes of differentiation by RNAase protection. The R A R probes used in the protection experiments, were derived from regions of their cDNAs exhibiting high sequence divergence, thus excluding cross-detection during simultaneous RNAase protection (Fig. 1). A probe detecting transcripts from the uniformly expressed glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was included as an internal reference to which all quantification data were normalized.
;11 1
2
3
4
-Ol t
-/31 5
6
7
8
9
10 11 12 13
14 15 16
17
18 19
Fig. 2. Modulation of RAR mRNA expression during RA induced differentiation of asynchronously growing P19 cells. Autoradiogram showing the results from RNAase protection analysis, a, /3 and y (lanes 1-3) denote the position of RARa, -/3 and -y specific protected fragments ( f ) following hybridization of each probe to 10/xg total RNA from 24 h 10 -6 M RA treated P19 EC cells. Lane 4: mixed undigested RARa, -/3 and -y hybridization probes (top: y, middle: /3, bottom: a probe). Lane 5: hybridization of mixed a, /3 and 3' probes with tRNA followed by digestion with RNAase T1. Lanes 6-19: protected fragments resulting from hybridization of the mixed a,/3 and 3' probes with 10 p.g total RNA from cells incubated with RA for the times indicated and RNAase T1 digestion.
167 after 3 h and a second rise commences 8 h after RA addition (Fig. 3B). Prolonged exposure to RA further induces RAR/3 mRNA to a maximum at four days of
(up to 5 days) exposure of the cells to RA. In contrast, RAR/3 mRNA is rapidly (within 1 h) induced, apparently in a two step fashion: a first maximum is reached
120
40
D
A
T
100
_
