Phosphorylation cell-specifically modulates
the activity of AF-1 of retinoic acid receptor -2
Gert E. Folkers, and Paul T. van der Saag
Hubrecht Laboratory, Netherlands Institute
for Developmental Biology, Utrecht, The Netherlands
to be submitted in an altered form
We have previously reported the presence of an autonomous transcriptional activation domain within the N-terminal part of RAR2, that can activate transcription in a cell-type specific fashion. A detailed mutational analysis revealed that this AF resembled the acidic activator VP16. Furthermore putative phosphorylation sites contribute to its activity. These residues are located within the putative amphipatic -helix, which is required for transcriptional activation. We now present evidence that the N-terminal part of the receptor is indeed phosphorylated in vivo. The presence of these phosphorylatable residues clearly contribute to the cell-specific activation by RAR AF-1 since mutation of these residues resulted in a cell type-dependent modulation of AF-1 activity. Furthermore, cotransfection of CKII and PKA expression vectors resulted in a cell type-specific response to overexpression of these kinases. Together these experiments suggest that phosphorylation of phosphorylatable residues modulates the activity of RAR2 AF-1 in a cell-type specific manner.
Activation of transcription is thought to be dependent on an interaction between the DNA-bound activator and the general transcription factors. Through stabilization of the pre-initiation complex or by increasing the formation of the pre-initiation complex activators modulate the activity of RNA polymerase II promoters (37). To perform this task, activators interact with one or more proteins which could be a GTF and/or a cofactor/intermediary protein that can communicate with GTFs (37,44). Alternatively activators could be involved in the disturbing or removal of the nucleosomes (46). In the cell DNA is packed in chromatin, arrays of highly ordered nucleosomes, that in term consist of DNA wrapped around histone particles. Packing of DNA in chromatin structure is thought to prevent transcriptional activation. Acetylation of histones results in a change in chromatin structure, thereby allowing binding of activators and/or GTFs, which in turn can promotes transcriptional activation. Therefore activators interact with proteins that are involved in histone acetylation through the histone acetyl transferases or alternatively activators are, in a SWI/SNF/Brahma-complex-dependent manner, remodelling or removing the nucleosomes (46).
Transcription is a highly regulated process, mRNA expression of most genes is regulated tissue-specifically both during embryonal development and in adult live. The mechanism underlying this cell-specificity is largely unknown. The most important aspect for transcriptional activation is the presence of activators (and/or absence of repressors); however also the activator activity is an important aspect in the cell-specific regulation of transcription. Activity can be regulated by the (limiting) amounts of cofactors/intermediary proteins (32). Also the strength of the interaction between the activator and the interacting protein may modulate transcriptional activation cell-specifically. One mechanism for cell-specific regulation of activator activity is protein phosphorylation. The activity of several transcription factors, e.g. CREB (16) and c-jun (39, 33) have been shown to be up-regulated by phosphorylation (for review see 19). Furthermore steroid hormone receptors are phosphorylated in vivo (45), and phosphorylation sites have been mapped (6,10,15). Although the precise function is mostly unknown, a role in modulating transactivation was suggested by phos-phorylation, as has been reported for the vitamin D3 receptor (20) and the estrogen receptor (ER) (1). Also the retinoic acid receptors (RARs) were shown to be phosphor-proteins, as both ligand-dependent and -independent phosphorylation was observed (14,34,35).
RARs belong to the nuclear hormone receptor superfamily which share a common domain structure (11,17,25). The DNA-binding domain which is most conserved among the different members of this family consists of two zinc fingers. The hormone-binding domain is a highly structured region, located at the C-terminal part of the receptor, and is involved in the binding of ligand, dimerization, and hormone-dependent transactivation through an activation function (AF-2), present in the C-terminal part of the LBD. The N-terminal part of the receptor is most variable and has been shown to contain an autonomous activation function (AF-1) (11). We and others have previously reported the presence of two autonomous transcriptional activation functions in RAR (13,29,30) which both activate transcription by distinct, cell-type dependent mechanisms (13). The activation function present in the N-terminal part of the protein (AF-1, formerly called TAF-1), is located in the first 32 amino acids of the receptor, and functions both in the presence and absence of RA (13) Further characterization revealed that AF-1 is unstructured in solution and is dependent on the presence of acidic and hydrophobic amino acids, showing similarities with the viral acidic activator VP16 (12). Furthermore we found that several putative phosphorylation sites, present within this region, are involved in transcriptional activation. Mutation of these phosphorylation sites led to a small (50%) but significant decrease in activity of RAR2 AF-1 in COS cells. (12). The observed cell-specific activity of this activator (13) and the contribution of putative phosphorylation sites in the activity of this activator (12) led us to suggest that phosphorylation could be involved in the cell-specific activation of RAR.
We show here that RAR is phosphorylated at multiple sites, including the AB-region. Transfection of various AF-1 constructs with mutated putative phosphorylation sites revealed that the contribution of these residues to the activity of AF-1 is cell-specific. The involvement of phosphorylation in the cell-specific activity of AF-1 is further confirmed by the cell-specific effect of cotransfection of expression constructs for the kinases CKII and PKA.
Plasmids, cell culture and transient transfection.
All plasmids used have been described
before (12). Cell culture, transfections transfection, -galactosidase-
, CAT- and LUC-assays and Western blotting were performed as described
before (12,13).
In vitro phosphorylation
Bacterial GST-RAR fusion proteins
were purified as described (12). Kinase reactions were performed in 50
mM Tris pH 7.5, 4 mM MgCl2, 0.2 mM EGTA, 0.2 µl [32P]-ATP,
50 µM ATP, 0.05-0.5 µg gluthation-sepharose-bound GST-fusion
protein and 0-10 µg nuclear extract (23), or WCE (see above) in 20
µl for 30 min at 30oC. Subsequently GST-fusion proteins
were purified by four washes with PBST (PBSø containing 1% (v/v)
Triton X100) and separated by SDS-PAGE (12.5%).
In vivo labeling and immunoprecipitation
Cells were transfected as described and 12 hrs or 24 hrs post transfection medium was changed for methionine- or phosphate-free DF medium containing 7.5% (v/v) dialyzed serum and cells were labeled with 50 µCi [35S]-methionine/ml (Amersham) or 1 mCi orthophosphate/ml (ICN), for 14 and 4 hrs respectively. 1.0 µM RA was added together with orthophosphate (4 hrs). Cells were washed with cold PBSø (150 mM NaCl, 16 mM Na2HPO4 and 4 mM NaH2PO4), scraped in PBSø and lysed in 40 µl lysis buffer (see western blotting). After centrifugation the cell lysate was precleared in 750 µl IP-buffer (50 mM Tris pH 7.5, 100 mM NaCl, 5 mM EDTA, 1% (v/v) Triton X100) containing 0.2 mM PMSF and protease inhibitors using 1 µl normal mouse serum and 50 µl 50% (vol/vol) slurry protein A-sepharose in IP-buffer. For phosphate labelling 50 mM NaF, 40 mM -glycerophosphate and 0.2 mM NaVO4 was added to both lysis and IP buffer. The lysates were then incubated for 2 hrs with 75 µl 12CA5 hybrodoma supernatant, thereafter 50 µl 50% (vol/vol) slurry protein A-Sepharose in IP buffer was added and incubated for another hour. Beads were washed 3 times with 1 ml IP buffer, once with 100 mM Tris-HCl pH 7.5, 0.5 M LiCl (1 ml) and once with 10 mM Tris-HCl pH 7.5 (1 ml). Immune complexes were eluted by incubation at 100oC for 5 min in sample buffer. Immunoblotting was performed as described above. Phosphorylation and methionine incorporation was visualized by autoradiography (2-8 days).
RAR is phosphorylated in vivo
AF-1 contains relatively many serine/threonine
residues which could be putative phosphorylation sites to test this more
directly we first constructed a number of GST-fusion proteins containing
the AB region or mutants of this region and performed an in vitro
phosphorylation assay with nuclear extracts from various cell lines as
a source for kinases, as described in Material and Methods. We observed
a weak phosphorylation which was both extract- and fusion protein-dependent.
The levels of phosphorylation however were not significantly different
between extracts of the various cell lines and were not influenced by RA
treatment of cell extracts. Furthermore, no correlation was observed between
phosphorylation levels in vitro of the various mutants tested and
their capacity to activate transcription in transient transfection (data
not shown). We therefore decided to investigate the role of phosphorylation
in vivo. In order to be able to immunoprecipitate the receptor we
cloned a hemagglutinin tag (HA-tag) in front of the receptor and performed
immunoprecipitation with extracts of COS cells transfected with epitope-tagged
receptor grown in the presence of [35S]-methionine using a monoclonal
against this epitope (12CA5). Figure 1A shows that RAR and RAR AB could
be specifically immunoprecipitated with the 12CA5 antibody and, migrated
according to their molecular mass. Furthermore this labeled protein immuno-reacted
with the antibodies 12CA5 or RAR (Figure 1A, ECL; data not shown). Figure
1B shows the result of an immunoprecipitation of COS cells transfected
with epitope-tagged RAR, grown for four hours with 32P-orthophosphate,
in the presence or absence of RA. A 32P-labeled protein was
immunoprecipitated that comigrated with the simultaneously transfected
and immunoprecipitated 35S-labelled epitope-tagged RAR. Immunoprobing
with RAR-specific antibody confirmed that the labeled proteins were indeed
RAR (data not shown). The levels of retinoic acid receptor phosphorylation
were not influenced by RA addition to the culture medium for 4 hours, which
is in agreement with earlier findings (34). The levels of phosphorylation
of other nuclear hormone receptors is sometimes affected by hormonal treatment
as is the case for PR (38), GR (31) and ER (1,15), while the RARs (except
RAR1 and 3) were all reported to be phosphorylated irrespectively of the
presence or absence of ligand (14,34,35).
Multiple regions of RAR are phosphorylated


Involvement
of putative phosphorylation sites in cell specific transcriptional activation.
We have previously shown that the AF-1 activity
functions cell-specifically, and mutation of putative phosphorylation sites
within the AB-region contribute to the activity of AF-1 (12) together with
the above presented evidence for in vivo phosphorylation of AF-1,
this led us to investigate whether phosphorylation modulates the activity
of AF-1 in a cell-specific fashion. We therefore transfected a GAL-responsive
reporter construct together with fusion constructs consisting of the DNA-binding
domain of the yeast transcription factor GAL4 and the AB-region of RAR
or mutants of this region, in which putative phosphorylation sites were
altered. If phosphorylation indeed modulates the activity cell-specifically,
differences in the contribution of these residues between the various cell
lines are expected. As shown in Table 1, mutation of the three critical
acidic residues (D3,6,17A) resulted in a complete loss of AF-1 activity
in all cell lines similar as reported before (12). Mutation of Ser 9, 11
and Thr 20 to Ala did not significantly decrease activity in P19-EC and
COS cells while a 40-60% reduction in activity was found in 293 and T47D
cells. The activity of the S22,24,25A mutant was lower than wild-type in
all cell lines tested, suggesting that phosphorylation of these residues
contributed to transcriptional activation. Mutation of these residues to
aspartic acid increased the activity of this activator, suggesting that
negative charge, caused by phosphorylation, can be artificially mimicked
by introduction of negatively charged residues (Table 1). Similar findings
have been reported for ER (1), c-jun (18) and p53 (28). The activity of
the Tyr19 mutant was most interestingly. Mutation of this residue had no
effect in COS cells, while a decrease in activity was found in 293, T47D,
and 3T3 cells and an increase in activity in P19-EC cells. This could mean
that phosphorylation of this tyrosine residue leads to a decrease in activity,
since upon mutation of this residue to a non-phosphorylateble residue (Ala)
an increase in activity is found in those cells that are able to phosphorylate
this residue (eg. P19-EC) but not in cells that apparently lack the particular
kinase activity. Interestingly, RAR but not RAR1 and RAR1 were reported
to be phosphorylated on tyrosine (34), and the tyrosine kinase signal transduction
pathway is functional in P19-EC cells (8). To further test the hypothesis
that phosphorylation of Tyrosine residue 19 decreases the activity of AF-1,
we tested the activity of WT and Y19A in P19-EC and in P19-RPTP, a P19
clonal cell line that overexpresses a receptor phosphor-tyrosine phosphatase
(RPTP), which can differentiate into the neuronal direction in the absence
of aggregation upon RA-treatment (9). As shown in Figure 3, the activity
of WT and mutant GAL-AF1 is similar in the RPTP clone but in the P19-EC
cells a two-fold higher activity is found for the Y19A mutant. These results
suggest that in the RPTP clone this tyrosine phosphorylation is absent,
similar to the situation in COS cells. It is tempting to speculate that
Y19 is dephosphorylated directly or indirectly by RPTP in the RPTP overexpressing
P19-EC cell line, leading to the loss of negative contribution of phosphorylated
Y19 to AF-1 activity. Cell-specific activation has been reported for ER
(5,7,42), and interestingly also mutation of identified phosphorylation
sites of the mouse ER S118 resulted in a cell type- and promoter-specific
decrease in activity (1) as is the case for RAR AF-1 (Table 1, 12), and
GR AF-1 (4). This site (S118) was shown to be phosphorylated by MAPK (22),
resulting in a two- to three-fold increase in activity of ER AF-1, both
in the presence or absence of ligand. This cross-talk between the growth
factor signaling pathways and ER emphasizes the role for phosphorylation
in cell-specific activation. Various phosphorylation sites for ER were
found in vivo by different groups in different cell lines, including
the presence of tyrosine phosphorylation (27), while others reported the
phosphorylation of Serine residues only (15,1). Furthermore others reported
besides the above described Ser118, an additional CKII phosphorylation
site within AF-1 in MCF7 cells (3) which was not found by other groups
in COS cells (1,15). Together these data are in agreement with the proposed
phosphorylation-dependent cell-specific activation of AF-1.
Overexpression of PKA and CKII influence AF-1 activity cell specifically.
The above described differences in the contribution
of putative phosphorylation sites in the activity of AF-1 suggests that
differences in phosphorylation levels are underlying the cell-
specific
activation by AF-1. Modulation of the activity of certain kinases present
in a cell might cell-specifically modulate AF-1 activity. The RA-dependent
activity of RARs has been reported to be activated by PKA (21,36) and PKC
(40,43). Dephosphorylation of RARs and RXRs by the addition of okadaic
acid led to a decrease in activity through a decrease in DNA binding activity
of the RAR/RXR heterodimer (24). To investigate whether kinase activity
contributes to cell specific RAR2 AF-1 activity we cotransfected expression
constructs for protein kinase A (PKA) and casein kinase II (CKII), together
with GAL-AF1 and a GAL-responsive reporter in P19-EC cells and in COS cells.
Interestingly cotransfection of both kinases caused a repression in P19-EC
cells while in COS cell a two- to three-fold activation was observed. These
cotransfection experiments suggest that phosphorylation modulate the activity
of AF-1, and that differences in phosphorylation between the cell lines
is possibly underlying the observed cell specificity.
In conclusion, the autonomous activation function
(AF-1) present at the N-terminus of RAR2 activates transcription cell-specifically
(13). Here we have presented evidence that putative phosphorylation sites
within this AF are involved in this cell specificity (Table 1, Fig. 3).
Furthermore the N-terminal part of the receptor containing AF-1, was found
to be phosphorylated (Fig 2) and activity was cell-specifically influenced
by kinase expression constructs (Fig. 4). This suggests that cell-specific
differences in phosphorylation of AF-1 residues are modulating the activity
of AF-1 in a cell-specific manner. To directly confirm this hypothesis
we are currently mapping the phosphorylation sites of RAR in different
cell lines.
Acknowledgements
We are grateful to P. Chambon for providing receptor constructs and antisera against RAR. Patricia Swanink for technical assistance, and F.J.M. Vervoordeldonk and J. Heinen for photography. G.E.F is supported by the Dutch Cancer Society.
1. Ali, S., D. Metzger, J.-M. Bornert, and P. Chambon. 1993. EMBO J. 12:1153-1160.
2. Almlöf, T., P.H. Wright, and J.-A. Gustafsson. 1995. J. Biol. Chem. 270:7535-17540.
3. Arnold, S.F., J.D. Obourn, H. Jaffe, and A.C. Notides. 1994. Mol. Endocrin. 8:1208-1214.
4. Bai, W., and N.L. Weigel. 1996. J. Biol. Chem. 271:12801-12806.
5. Berry, M., D. Metzger, and P. Chambon. 1990. EMBO J. 9:2811-2818.
6. Bodwell, J.E., E. Orti, J.M. Coull, D.J.C. Pappin, L.I. Smith, and F. Swift. 1991. J. Biol. Chem. 266:7549-7555.
7. Bocquel, M.T., V. Kumar, C. Stricker, and P. Chambon. 1989. Nucleic Acid Res. 17:2581-2595.
8. Den Hertog, J., R. Eman, L.G.J. Tertoolen, S.W. de Laat, and W. Kruijer. 1991. Exp. Cell. Res. 196:226-232.
9. Den Hertog, J., C.E.G.M. Pals, M.P. Peppelenbosch, L.G.J. Tertoolen,, S.W. de Laat, and W. Kruijer. 1993. EMBO J. 12:3789-3798.
10. Denner, L.A., W.T. Schrader, B.W. O'Malley, and N.L. Weigel. 1990. J. Biol. Chem. 265:10025-10029.
11. Evans, R.M. 1988. Science 240:889-895.
12. Folkers, G.E., E.C. van Heerde, and P.T. van der Saag. 1995. J. Biol. Chem. 270:23552-23559.
13. Folkers, G.E., B.J.M. van der Leede, and P.T. van der Saag. 1993. Mol. Endocrinol. 7:616-627.
14. Gaub, M.P., C. Rochette-Egly, Y. Lutz, S. Ali, H. Matthes,
I. Scheuer, and P. Chambon. 1992. Exp. Cell Res. 201:335-346.
15. Le Goff, P., M.R. Montano, D.J. Schodin, and B.S. Katzenellenbogen. 1994. J. Biol. Chem. 269:4458-4466.
16. Gonzalez, G.A., and M.R. Montminy. 1989. Cell 59:675-680.
17. Green,S. and P. Chambon. 1988. Trends Genet. 4:309-314.
18. Hoeffler,W.K., A.D. Levinson, and E.A.Bauer. 1994. Nucleic Acid Res. 22:1305-1312.
19. Hill, C.S., and R. Treisman. 1995. Cell 80:199-211.
20. Hsieh,J.-C., P.W. Jurutka, M.A. Calligan, C.M. Terpening, C.A. Haussler, D.S. Samuels, Y. Shimizu, N. Shimizu, and M.R. Haussler. 1991. Proc. Natl. Acad. Sci. USA 88:9315-9319.
21. Huggenvik, J.I., M.W. Collard, Y.-W. Kim, and R.P. Sharma. 1993. Mol. Endocrin. 7:543-550.
22. Kato, S., H. Endoh, Y. Masuhiro, T. Kitamoto, S. Uchiyama, H. Sasaki, S. Masushige, Y. Gotoh, E. Nishida, H. Kawashima, D. Metzger, and P. Chambon. 1995. Science 270:1491-1494.
23. Lee,K.A.W., A.S.Bindereif, and M.R. Green. 1988. Gen. Anal.Technol. 5:22-31.
24. Lefebvre, P., M.P. Gaub, A. Tahayoto, C. Rochette-Egly, and P. Formstecher. 1995. J. Biol. Chem. 270:10806-10816.
25. Leid,M., P. Kastner, and P. Chambon. 1992. Trends Biochem. Sci. 17:427-433.
26. Mason, S.A., and P.R. Housley. 1993. J. Biol. Chem. 268:21501-21504.
27. Migliaccio, A., A. Rotondi, and F. Aurricchio. 1986. EMBO J. 5:2867-2872.
28. Milne,D.M., R.H. Palmer, and D.W. Meek. 1992. Nucl. Acid Res. 20:5565-5570.
29. Nagpal,S., S. Friand, H. Nakshatri, and P. Chambon. 1993. EMBO J. 12:2349-2360.
30. Nagpal,S., M. Saunders, P. Kastner, B. Durand, H. Nakshatri, and P. Chambon. 1992. Cell 70:1007-1019.
31. Orti, E., D.B. Mendel, L.I. Smith, and A. Munck. 1989. J. Biol Chem. 264:9728-9731.
32. Ptashne, M. 1988. Nature 335:683-689.
33. Pulverer, B., J.M. Kyriakes, J. Avruch, E. Nikolokaki and
J.R. Woodgett. 1991. Nature 353:670-674.
34. Rochette-Egly,C., M.-P. Gaub, Y. Lutz, S. Ali, I. Scheuer, and P. Chambon. 1992. Mol. Endocrin., 6:2197-2209.
35. Rochette-Egly,C., Y. Lutz, M. Saunders, I. Scheuer, M.-P. Gaub, and P. Chambon. 1991. J. Cell Biol. 115:535-545.
36. Rochette-Egly, C., M. Oulad-Abdelghani, A. Staub, V. Pfister, I. Scheuer, P. Chambon, and M.-P. Gaub. 1995. Mol. Endocrinol. 9:860-871.
37. Roeder, R.G. 1996. TIBS 21:327-335.
38. Sheridan, P.L., R.M. Evans, and K.B. Horwitz. 1989. J. Biol. Chem. 264:6520-6528.
39. Smeal,T., B. Binetruy, D. Mercola, M. Birrer, and M. Karin. 1991. Nature 354:494-496.
40. Tahayoto, A., P. Levebvre, P. Formstecher, and M. Dautrevaux. 1993. Mol. Endocrinol. 7:1642-1653.
41. Takimoto, G.S., A. Rudie Hovland, D.M. Tasset, M.Y. Melville, L. Tung, and K.B. Horwitz. 1996. J. Biol. Chem. 271:13308-13316.
42. Tora,L., J. White, C. Brou, D. Tasset, N. Webster, E. Scheer, and P. Chambon. 1989. Cell 59:477-487.
43. Yang, Y., S. Minucci, D.J. Zand, K. Ozato, and J.D. Ashwell. 1994. Mol. Endocrinol. 8:1370-1376.
44. Verrijzer,C.P., and R. Tjian. 1996. TIBS 21:338-442.
45. Weigel, N.L. 1996. Biochem. J. 319:657-667.
46. Wolffe, A.P., and D. Pruss. 1996. Cell 84:817-819.
47. Zhou, Z.X., Kemppainen, J.A. and Wilson, E.M., 1995. Mol. Endocrin. 9:605-615.