Retinoic Acid Receptor-P: lmmunodetection and Phosphorylation on Tyrosine Residues

Cecile Rochette-Egly, Marie-Pierre Gaub, Yves Lutz, Simak Ali, lsabelle Scheuer, and Pierre Chambon Laboratoire de Gkktique Molkculaire des Eucaryotes du CNRS Uniti! 184 de Bioloaie Molkulaire et de GEnie Gitnktique de I’INSERM lnstitut de Chimie &ologique Facultk de Mkdecine 67085 Strasbourg Cedex, France

INTRODUCTION

Polyclonal (RP) and monoclonal (Ab) antibodies were raised against synthetic peptides (or fusion proteins) corresponding to amino acid sequences unique to human and mouse retinoic acid receptorp @AR@) isoforms. Antibodies directed against the A2 region [Ab682(A2), Ab7P2(A2), and RPP2(A2)], the D2 region [RPP(DP)], or the F region [Ab88(F)2, RPP(F)l, and RPP(F)S] were selected. The monoclonal and polyclonal antibodies directed against the D2 and F regions specifically immunoprecipitated and recognized by Western blotting all human and mouse RARP isoforms (mRARP1, $32, -/33, and -p4), produced in COS-1 cells transfected with expression vectors containing the corresponding RARp cDNA. Furthermore, in gel retardation assays, the monoclonal antibodies supershifted RARB protein-RA response element oligonucleotide complexes. Antibodies directed against the A2 region were specific for the RARB2 isoform. The above antibodies allowed us to detect the presence of mRAR82 proteins in mouse embryos and to show that their presence in embryonal carcinoma cells (F9 and PI9 cell lines) is dependent upon RA treatment. The antibodies were also used to demonstrate that RARP proteins produced by transfection in COS-1 cells are phosphorylated. RARB2 phosphorylation was not affected by RA treatment, whereas the phosphorylation of RARBI and RARP3 isoforms was greatly enhanced by RA. We also show that, in contrast to RARal and RARyl, RARBP proteins contain phosphotyrosine residues and are only weakly phosphorylated in vitro by CAMP-dependent protein kinase. These results support our previous proposal that the various receptors have distinct functions in the RA-signaling pathway. (Molecular Endocrinology 6: 2197-2209, 1992) OmJ-8809/92/2197-2209$03.00/0 Molecular Endocrinology Copyright 0 1992 by The Endccnne

Retinoids are important signaling molecules for pattern formation and cellular differentiation during vertebrate development (for reviews, see Refs. l-6). It is well established that retinoids exert their biological activity via three types of proteins: the cellular retinoic acid (RA)-binding proteins CRABP-I and -II (for review, see Ref. 7), the cellular retinol-binding proteins CRBP-I and -II (for review, see Ref. 8), and the nuclear RA receptors (RARs), which are ligand-inducible transcriptional transregulators modulating the transcription of target genes via c&acting DNA response elements (RAREs) (for review, see Refs. 9 and 10 and references therein). In addition, a distinct family of nuclear receptors, referred to as retinoid X receptors (RXRs), has been recently identified (1 l-l 4). Three types of RARs have been characterized in the human and mouse, RARa, $3, and -y (see Ref. 9 for a review), which belong to the superfamily of steroid/ thyroid hormone nuclear receptors (15-I 7). Moreover, isoforms of RARa, -p, and -y have been identified which differ only in their N-terminal A regions (9, 18-20). In a given species, the amino acid sequences of the three RARs (-a, -p, -y, and their isoforms) are more divergent from one another than the sequences of a given RAR across species, suggesting that each RAR type or isoform may perform unique functions (9, 20). This idea is further supported by in situ hybridization studies, which revealed that transcripts encoding the three RARs exhibit unique patterns of expression during embryonic development (21-24), the distribution of RARP and RARy transcripts being specifically restricted compared with the ubiquitous pattern of expression of RARa. In the case of the RARP gene, four isoforms have been identified which are generated by alternative splicing of primary transcripts initiated from two promoters, Pl for RAR/31 and $3 isoforms, and P2 for RARP2 and $4 isoforms (19,25-28). Interestingly, the P2 promoter

Socety

2197

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

MOL 2198

END0.1992

Vol6No.12

contains a RA response element (26-28). Both adult and embryonic brains contain high levels of the mRARP1 and mRARP3 transcripts relative to those of mRARP2 and $4, suggesting a possible role for RARPI and $3 isoforms in the development and function of the central nervous system (19,26). In situ hybridization studies have shown that RARP transcripts are restricted to the developing central nervous system and the mesenchyme of the face, gut, limb, urogenital system, teeth, and sensory organs (21, 22, 24) and, thus, do not correlate with a particular tissue type or cell fate. To further study RARP proteins, we raised polyclonal and monoclonal antibodies against synthetic peptides corresponding to different RAR/3 sequences. These antibodies provide tools to detect endogenous RARP proteins in mouse embryos and embryonal carcinoma (EC) cells. In the latter case, the presence of RARP proteins was found to be dependent on RA treatment. Furthermore, our antibodies allowed us to demonstrate that RARP proteins are phosphorylated. Whereas RARP2 was constitutively phosphorylated after transient expression in COS-1 cells, the phosphorylation of RARPl and RARp3 increased upon RA treatment. Finally, we demonstrate that RARP2 proteins contain phosphotyrosine residues, thus suggesting a role for tyrosine kinases in the regulation of RARP2 functions.

RESULTS Generation of Polyclonal and Monoclonal Antibodies against Synthetic Peptides Specific RAW3 lsoforms

to

The RARPP isoform is nearly 100% conserved between mouse and human and differs from the other RAR types in its N-terminal A region, central D2 region, and Cterminal F region (Ref. 9 and references therein). The four major RARfl isoforms, -pi, -02, -p3, and -04, differ from each other only in their N-terminal A regions (Al, A2, A3, and A4) (9, 19, 26) (see Table 1). Potential immunogenic amino acid sequences were selected to generate antibodies specific to all mouse (m) RARP isoforms (regions D2 and F) or unique to mouse and/or human (h) RARP2 isoforms (region A2; see Table 1). Two of these peptides (SP248 and SPA39, corresponding to regions A2 and D2, respectively) are fully conserved between human and mouse. Peptide SP172 (corresponding to human region F) as well as the whole human F region fused to the DHFR protein [FP(DHFR)] diverge from the mouse sequences by two and three amino acids, respectively. All of these peptides as well as the fusion protein Fp(DHFR) were antigenic in rabbits and resulted in the production of polyclonal antibodies. Two of them [RP/32(A2) and RPP(F)l], directed against peptides SP248 and SP172, have previously been described (29). The antisera raised against SPA39 [RPP(D2)] and the fusion protein F/I(DHFR) [RPP(F)2] recognized their cognate, but not other peptides or

fusion proteins, as determined by enzyme-linked immunosorbent assay (ELISA; data not shown). Peptide SP248 and the fusion protein FP(DHFR) were antigenic in mice and resulted in the production of specific hybridoma. Based on positive signals obtained by immunoblotting and immunoprecipitation, we selected two clones corresponding to SP248 [Ab6@2(A2) and Ab7P2(A2)] and one clone corresponding to F/I(DHFR) [Ab8P(F)2]. Each clone recognized specifically its cognate, but not other peptides or fusion protein, as checked by ELISA (data not shown). All these antibodies were identified as IgGl K (data not shown). All monoclonal and polyclonal antibodies were checked for their ability to immunoreact with RARP proteins produced in transfected COS-1 cells. In all cases, an intense nuclear staining was seen in mRARP2-transfected COS-1 cells (data not shown). Specific Detection of Cloned Human and Mouse RARB Proteins by Immunoblotting, Immunoprecipitation, and Gel Shift Assay lmmunoblotting The monoclonal antibodies and the rabbit polyclonal antisera were tested for their ability to specifically reveal by immunoblotting the cloned human or mouse RARP proteins produced by transfected COS-1 cells (see Materials and Methods). After fractionation of whole cell extracts (WCE) by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE), Western blotting analysis was performed by using the specific monoclonal or polyclonal antibodies together with [‘251]antimouse immunoglobulins or [‘*?I protein-A (Fig. 1 A). In extracts of mRAR/32-transfected COS-1 cells (COS-p2), the monoclonal and polyclonal antibodies gave strong signals (Fig. 1A, lanes 2, 5, 8, 11, 14, and 17, and Fig. 1 B, lanes 1, 4, 7, and 8), with an apparent mol wt of -51 kilodaltons (kDa). Similar signals were observed whether transfections were performed in the absence or presence of RA (Fig. 1 B, lanes 7 and 8). An additional signal with a lower apparent molecular mass (--47 kDa) was often detected with the polyclonal and monoclonal antibodies directed against the F domain of RARP: Ab8P(F)2 (Fig. lA, lane 5, and Fig. lB, lane l), RP@(F)l (Fig. lA, lane 14) and RPP(F)2 (Fig. IA, lane 17). This additional species was not seen with antibodies directed against the A2 domain (Fig. 1 A, lanes 2 and 8; Fig. lB, lane 4, and data not shown) and was not constantly seen in extracts corresponding to different transfections (e.g. Fig. lB, lanes 7 and 8). In extracts of hRARp2-transfected COS-1 cells, the same signals were also revealed by Western blotting with each of the monoclonal and polyclonal antibodies (data not shown). These signals were specific, since they were not seen in nontransfected COS-1 cells (Fig. 1 A, lanes 1, 4, 7, 10, 13, and 16) or when the ascites fluids or the antisera were immunoadsorbed with the corresponding peptide (see Materials and Methods; Fig. 1 A, lanes 3, 6, 9, 12, 15, and 18). Similar competition experiments with ovalbumin (used as a carrier protein

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

RARP

on Tyrosine

Table

1. Amino

2199

Residues

Acid Sequences

of the Synthetic

1

Peptides

53 81 I 1B [ 60 88 I I B I 87 116

A* 1 Al 1

c c

I

A3

1B 1 1 5 34 I A41 B 1

Peptide SP248(A2

human mouse

Used to Generate

147 193 I I I I 1~1l~21~31 154 200 I I IDI\D~\D~~ 182 228 I

the RARP Antibodies

413 448 E

RAR-82 1 F 420 455

E

RAR-81 1 F 448 483 1 F RARj33 366 401

I

C

IDI D&31 100 146

E

C

IDI D~jDsj

E

1 F

RARj34

region) : Ab682(A2) Ab7P2(A2) RP82(A2)

SPGQILDFYTASPSSC

-----~~-~~~~~~--

11

26

Peptide SPA39(D2 region) :RPP(D2) human mouse

CTESYEMTAELD

___---------

167

178

Peptide SP172(human F region) : RPP(F)l human mouse

PSISPSSVENSGVSQSPLVQ

425v-------------L448

Fb(DHFR)(human F region fused to DHFR) : RPp(F)2 Ab8P(F)2 human mouse

HEPLTPSSSGNTAEHSPSISPSWENSGVSQSPLVQ

-----------I------V---------------L413

448

RAR@l , RARfl2, RARPB, and RARP4 proteins (455, 448, 483, and 401 amino acids long, respectively; same length in mouse and human) are schematically represented, with their six distinct regions designated A through F. The four RARP isoforms differ from each other only in their NH?-terminal A regions (Al for RARpl , A2 for RARPP, A3 for RAR/33, and A4 for RARB4). The amino acid sequences (single letter code) of the synthetic peptides used to generate RAR/3 antibodies are represented. The numbers flanking the peptide sequences refer to the positions of amino acid residues in the sequence of the RARP2 isoform. Amino acids differing between mouse and human RARP2 are indicated.

coupled to the synthetic peptides) did not affect the intensity of the specific signals (data not shown). Furthermore, no cross-reactions were seen with the same antibodies using extracts from COS-1 cells transfected with either mouse or human RAR71 or RARal expression vectors, indicating that these antibodies are specific for RARP proteins (data not shown). As expected, the antibodies directed against the F region of RARP [Ab8/3(F)2, RPP(F)2, and RPP(F)l] also reacted specifically with extracts from COS-1 cells transfected with mRARP1 (Fig. 1 B, lanes 2, 9, and lo), $3 (Fig. 1 B, lanes 3, 11, and 12), and $4 (Fig. 1 B, lanes 13 and 14, and data not shown) expression

vectors and revealed major proteins with the expected apparent mol wt of 54, 56, and 48 kDa, respectively. In contrast, antibodies directed against the A2 region

[Ab6@2(A2)and RPP2(A2)]did not recognizemRARP1, -@3, and $4 proteins not shown).

(Fig. 1 B, lanes 5 and 6, and data

lmmunoprecipitation The monoclonal and polyclonal antibodies were further tested for their ability to discriminate between the different isoforms by immunoprecipitation of WCE from COS-1 cells transfected with mRARP1, $32, $3, or -p4 expression vectors. The immunoprecipitated proteins were fractionated by SDSPAGE and analyzed with RPP(F)2 and [‘251]protein-A.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

MOL 2200

ENDO.

1992

Vol6No.12

kDa 1036751+ 412812

3

45

6

RA’-

7 8 9

101112

131415

I -+-f-+-+

)_

1617 18

I

kDa 103-

54

4J56 +i51 48

I’ ’/ ;’ 51 2812

3

4 56

7 8, ? 10, 1112,1,314 2.

z

22i

Fig. 1. Characterization of Monoclonal and Polyclonal Antibodies by lmmunoblotting A, COS-1 cell extracts were either from control untransfected CO&l cells (lanes 1, 4, 7, 10, 13, and 16) or from COS-1 cells transfected with mRARP2 (lanes 2,3, 5, 6, 8, 9, 11, 12, 14, 15, 17, and 18) expression vectors. Whole cell extracts (20 pg proteins) were fractionated by SDS-PAGE, electrotransferred onto NC filters, and then immunoprobed with the mouse monoclonal antibodies Ab602(A2) (lanes l-3) and Ab8/3(F)2 (lanes 4-6) or the rabbit polyclonal antibodies RP@2(A2) (lanes 7-9) RPP(D2) (lanes 1 O-l 2). RPfi(F)l (lanes 13-15), and RPP(F)2 (lanes 16-18). without (lanes 1, 2, 4, 5, 7, 8, 10, 11, 13, 14, 16, and 17) or with previous antibody depletion (lanes 3, 6, 9, 12, 15, and 18). The specific complexes were revealed with [‘251]protein-A or [‘251]antimouse immunoglobulins. The positions of the prestained mol wt standards (Bethesda Research Laboratories, Gaithersburg, MD) are indicated in kilodatons. B, Extracts were from COS-1 cells transfected with mRARP2 (lanes 1, 4, 7, and 8) mRARP1 (lanes 2, 5, 9, and lo), mRARP3 (lanes 3,6, 11, and 12). or mRARP4 (lanes 13 and 14) expression vectors, without (lanes l-6,7,9, 11, and 13) or with (lanes 8, 10, 12, and 14) a 4-h treatment with lo-’ M RA. lmmunoblotting was performed with Ab8P(F)2 (lanes l-3) Ab6@2(A2) (lanes 4-6) or RPP(F)P (lanes 7-l 4).

Monoclonal antibodies [Ab6P2(A2), Ab7P2(A2), and Ab8/?(F)2] as well as polyclonal antibodies [RP/32(A2), RPP(DP), RPP(F)l , and RP@(F)2] specifically immunoprecipitated mRARP2 from WCE of mRARP2-transfected COS-1 cells (Fig. 2A, lanes 13-l 9). These signals were specific, since they were not observed when immunoprecipitation

was

performed

with

preimmune

nonreactive serum, nonreactive ascite fluid (Fig. 2A, lanes 20 and 21) or WCE from cells transfected with the parental expression vector pSG5 (Fig. 2A, lanes 2-l 0).

As already mentioned, an additional signal with a lower apparent mol wt (--47 kDa) was often immunoprecipitated with the monoclonal and polyclonal antibodies directed against the F domain of RARP (Fig. 2A, lanes 15, 16, and 18; Fig. 2B, lane 3) and not with the antibodies directed against the A2 domain, particularly when the cell extracts were frozen and thawed before immunoprecipitation. This additional 47-kDa signal was also observed when immunoprecipitations were performed in RIPA buffer (150 mM NaCI, 1% Nonidet P-40, 0.5% sodium desoxycholate, 0.1% SDS, and 50

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

RAR/3 on Tyrosine Residues

@ Ab

281 2 3 4 5 6 7 8 9 10 11 I 2’ 2 pSG5 $ transfected COS-lcells g

Ab8P(F)2 Ab6P2(A2)

1213141516171819 202122 ‘22 2 I mFiAR-P2 a 8 transfected COS-lcells g

- - + - - + - - + - - + - + - - + - - + - - + -

kDa 10367-

:

5’--, 47 4128,1 2 3,f1 5 6,,7 8 9,jOll 12, 2 z 2 2%

Characterization of Monoclonal and Polyclonal RAR@ Antibodies by lmmunoprecipitation A, Extracts (50 pg) from either pSG5 (lanes 2-1 O)- or mRARp2 (lanes 13-21)-transfected COS-1 cells were immunoprecipitated with the following antibodies: Ab6P2(A2) (lanes 2 and 13), Ab7@2(A2) (lanes 3 and 14), Ab8B(F)2 (lanes 4 and 15), RP@(F)2 (lanes 5 and 16), RPP2(A2) (lanes 6 and 17), RPP(F)l (lanes 7 and 18), RPP(D2) (lanes 8 and 19), nonreactive serum (NRS; lanes 9 and 20), or nonreactive ascite fluid (NRA; lanes 10 and 21). Antigen-antibody complexes bound to protein-A-Sepharose beads were eluted, fractionated by SDS-PAGE, and electrotransferred to NC filters. The immunoprecipitated material was immunoprobed by incubation of the filters with RPp(F)2 and [‘251]protein-A. As positive controls, extracts (20 pg protein) of mRARflStransfected COS1 cells (lanes 1, 11, 12, and 22) were directly loaded on the gel without prior immunoprecipitation and then immunoprobed. B, Extracts (50 pg) from mRAR@2 (lanes 2 and 3)-, mRARP1 (lanes 5 and 6)-, mRARP3 (lanes 8 and 9)-, or mRARfl4 (lanes 11 and 1 P)-transfected COS-1 cells were immunoprecipitated with the monoclonal antibody Ab6P2(A2) (lanes 2, 5, 8, and 11) or Ab8fl(F)2 (lanes 3, 6, 9, and 12). The immunoprecipitated material was immunoblotted with RP@(F)2 as described in A. As positive controls, extracts (20 fig protein) of mRARP2 (lane l)-, mRARP1 (lane 4)-, mRARP3 (lane 7)-, and mRARP4 (lane lO)-transfected COS-1 cells were directly loaded on the gel without prior immunoprecipitation and then immunoprobed. Fig. 2.

mM Tris, pH 7.5) and was not affected by prolonged incubation at 37 or 56 C (data not shown). However, proteolytic degradation of the A2 domain, leading to the appearence of the 47-kDa signal, cannot be exeluded. Alternatively, the 47-kDa species may be generated by dephosphorylation (see below), which by

changing the conformation of RARP2 would mask the epitope recognized by the antibodies raised against the A2 domain. As expected, the antibodies directed against region F [Ab8@(F)2] immunoprecipitated the four isoforms mRARP1, $32, $33, and $34 (Fig. 28, lanes 3, 6, 9, and

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

MOL

ENDO.

Vol6No.12

1992

2202

12). In contrast, the antibodiesdirected againstthe A2 domain of mRARP2, [Ab6@2(A2)and Ab7P2(A2)] did not immunoprecipitatemRARP1,$3, and $4 (Fig. 2B, lanes5, 8 and 11, and data not shown). Gel Shift Assay To confirm the specificity of the antibodiesfor FiARP isoforms,gel retardation assays were performed usinga 32P-labeledoligodeoxynucleotide (RAREP; see Materials and Methods) containing the RA responseelement (RARE) of the RARP2 promoter (19, 27, 28). A specific complex was obtained when extracts of transfected COS-1 cells containing mRARP1(Fig. 3, lanesl-4), $2 (Fig. 3, lanes5-8) or $33 (Fig. 3, lanes 9-12) proteins were used. These complexes were not seen when a mutated RARE@ probe was used (Fig. 3, lanes 1, 5, and 9) or when extracts of untransfectedCOS-1 cellswere usedin the sameconditions(data not shown). The RARE@probe-receptor complex obtained with extracts of mRARp2-transfected cells was further shifted after incubationwith the monoclonalantibodies Ab8P(F)2 and Ab602(A2) (Fig. 3, lanes 7 and 8). As expected, the complexes formed with mRARpl- and mRARP3-transfectedcellswere clearly shifted with the monoclonalAb8P(F)2,directed againstthe F domainof

2--I-

1 2 3 4 ,5 6 7 8, ,9 10 11 12, mRAR$l mRAR-P2 mRAR-p3 Fig. 3. Gel Retardation Shift Assay Gel retardation reactions were carried out with 5 fig of extracts from COS-1 cells transfected with mRARP1 (lanes l4) mRARp2 (lanes 5-8) or mRARp3 (lanes 9-l 2) expression vectors. Arrow 7 indicates the specific complexes formed with the RARE-P probe. Arrow 2 indicates the shifted complex formed in the presence of the monoclonal antibodies Ab8P(F)2 (lanes 4, 7, and 12) and Ab6@2(A2) (lanes 3, 8, and 11). As a negative control, the mutated oligonucleotide (RARE-pm) was used (lanes 1, 5, and 9).

RARP(Fig. 3, lanes4 and 12) whereasthey were not affected by Ab6p2(A2) (Fig. 3, lanes 3 and 11) or Ab7@2(A2)(data not shown), directed against the A2 domainof RARP2. Mouse RARPQtransfected cell extracts gave the same results (23) (data not shown). Furthermore, these monoclonalantibodiesdid not induce any supershift of RARotl or RARyl complexes (30, 31) (data not shown), confirming their specificity for RARP. lmmunodetection of RARB lsoforms Embryos and EC Cells

in Mouse

Mouse RARP2, $1, $3, and -64 RNAs have been found in EC cells (F9 and P19 cell lines)and in mouse embryos at different stagesof development(9, 19,26). All of these RNAs are expressedduring the course of mouseembryogenesis.However, mRARp2 transcripts and, to a lesser extent, mRARP1, $33, and $4 transcripts are induced by RA treatment in EC cells (19, 26). To correlate transcription and translation product abundance,we examinedwhether the RARP proteins could also be detected in these cells and tissueswith RARP-specificantibodies. Equalamountsof EC cell nuclearextracts (NE) were tested by immunoblottingwith RPP(F)P(Fig. 4A). In undifferentiatedF9 and P19 cells,no RARPsignalcould be detected (Fig. 4A, lanes4 and 6), whereasa 51-kDa mRARP2specieswas present in RA-treated cells (Fig. 4A, lanes5 and 7). A similarinduction of mRAR@upon RA treatment also occurred in END-2 and MES-1 cell lines (Fig. 4A, lanes 9 and 11) which are permanent clonal endodermaland mesodermalcell lines, respectively (32, 33). However, the RA-dependentincreasein RARP proteins differed from one cell line to the other; the strongest signal was observed with the P19 and MES-1 cell lines (Fig. 4A, compare lanes5, 7, 9, and 11). The 47-kDa specieswas also presentat a low level (Fig. 4A, lanes 7 and ll), as well as a higher mol wt signal (-65 kDa; Fig. 4A, lanes 4-11). All of these signals were specific, since they disappeared when immunoblottingwas performedwith a depletedRPP(F)2 antibody (data not shown). No 54- or 56-kDa signals correspondingto the RARPl and $3 isoformswere observed. Unfortunately, none of our monoclonalantibodies [Ab8P(F)2 and Ab6P2(A2)] was able to reveal endogenousRARp protein by direct Western blotting (data not shown). To determine whether the signalsobserved in RAtreated EC cellscorrespondto mRARP2or to the other RARPisoforms,we performedimmunoprecipitationexperimentsusingthe sameNE and monoclonalantibodies directed against either region A2 [Ab6P2(A2)] that specificallyrecognize mRARP2or those against region F [Ab8P(F)2] that recognize all mRAR@isoforms. As shown by subsequentimmunoblotting[using RP@(F)2 and [1251]protein-A] of the immunoprecipitatesthus obtained, a 51-kDa signalwas obtained with the antibodies raised against either the A2 or F region (Fig. 48, lanes 2-7) indicating that the revealed species do

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

RARp

on Tyrosine

Residues

2203

EMBWONAL @h

CARCINOMA

CELLS

polyclonal antibodies against region F [RPP(F)P]. A doublet (51-47 kDa; Fig. 5A) was present with a constant intensity as embryogenesis progressed, except for a slight increase at 10.5 days pc. This signal was also obtained after immunoprecipitation of mouse embryo NE with monoclonal antibodies raised against either the A2 (Fig. 48, lanes 10 and 12) or the F (Fig.

lmmunoblotting

kDa 103-

I

41 -

.’

I

@

/

281 2 3 4 5

0B

6

7

8

91011

12 1314

-

+ - + - + -

+

+ - +

-

-

+ -

- + + -

1 uuuDDuu mmmnu7m ojd,l --r--T--

tic++

kDa 67-

lmmunopreclpltatlon

Ab6P2(A2) Ab82(F)2

Mouse embryos

- + -

kDa 103e 51412812345 $%iF

67 hi&E;

8 9 110 Ii] 12 13,14 C-J 95d 10.5dr‘l 9-9 Mouse go embryo

Fig. 4. Characterization of RAR/3 in EC Cells A, Immunoblotting. NE (70 pg protein) of F9 (lanes 4 and 5) P19 (lanes 6 and 7). END-2 (lanes 8 and 9) and MES-1 (lanes 10 and 11) cells were fractionated by SDS-PAGE, electrotransferred to NC filters, and immunoprobed with RPP(F)2 together with [‘251]protein-A. Cells were grown as monolayers in the absence (lanes 4, 6, 8, and lo), or presence (lanes 5, 7, 9, and 11) of 1 O-’ M RA for 48 h. As positive controls, extracts (20 pg protein) of mRARfi1 (lanes 1 and 14)-, mRARP3 (lanes 2 and 13), and mRARP2 (lanes 3 and 1 P)-transfected COS-1 cells were run in parallel. B, Immunoprecipitation. NE (300 pg protein) were from RA-treated F9 cells (lanes 2 and 3). RAtreated P19 cells (lanes 4 and 5). RA-treated MES-1 cells (lanes 6 and 7) 9.5-day-old mouse embryos (lanes 10 and 1 l), or 10.5-day-old mouse embryos (lanes 12 and 13). The extracts were immunoprecipitated with the monoclonal antibody Ab6P2(A2) (lanes 2, 4, 6, 10, and 12) or Ab8/3(F)2 (lanes 3, 5, 7, 11, and 13) and the immunoprecipitated material was immunoprobed with RPP(F)P and [‘251]protein-A. As positive controls, extracts (20 vg protein) of mRARB2 (lanes 1 and 14), mRARfl1 (lane 8)-, and mRARP3 (lane 9)-transfected COS-1 cells were directly loaded on the gel without prior immunoprecipitation and then immunoprobed.

to mRAR/32. As already described for extracts of mRARPP-transfected cells, the additional 47kDa signal was also immunoprecipitated with the antibodies directed against region F (Fig. 4B, lanes 3, 5, and 7). Mouse embryo NE [from 9.5-14.5 days postcoitum (pc)] were also tested by immunoblotting with the rabbit

correspond

kDa 58c 51-w 41-

y’-

123456

gg 0

kDa 6751-w, 41-

Y-F

123456

Fig. 5. Characterization of RARP, RARq and RARr in Mouse Embryos at Different Stages of Development NE (70 pg) of mouse embryos were fractionated by SDSPAGE, electrotransferred to NC filters, and immunoprobed with RPP(F)P (A) or RPcx(F) (B), followed by [‘251]protein-A. For the detection of RARy (C), NE (1 mg protein) were first immunoprecipitated with the monoclonal antibody AbPr(mF), as previously described (30) and then immunoprobed with RPr(mF) and [‘251]protein-A. As positive controls, extracts (20 rg protein) of mRARPl-, -p2-, -p3-, -al-, -(~2-, -yl-, and -r2transfected COS-1 cells were run in parallel. Mouse embryos were tested at 9.5 (lane l), 10.5 (lane 2) 11.5 (lane 3) 12.5 (lanes 4). 13.5 (lane 5) and 14.5 (lane 6) days pc.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

MOL 2204

ENDO.

1992

46, lanes 11 and 13) region, indicating that it corresponds to mRARB2. As shown in Fig. 5B, the three signals (51, 54, and 58 kDa) corresponding to mRARa1, [as detected by immunoblotting with RPa(F) according to Gaub et al. (31)] were also uniformly present during the course of embryogenesis. A faint faster migrating band with a mol wt of 48 kDa, which might correspond to that of the mRARcu2 isoform appeared transiently at 10.5 days pc (Fig. 58, lane 2). In contrast to this ubiquitous expression of mRARP2 and mRARu1 proteins in embryos at various stages, mRARr1 proteins appeared only around 11.5 days pc (Fig. 5C) and were present at very low levels, since immunoprecipitation of high amounts of NE (1 mg protein/assay) was required for their detection (30). RARp Phosphorylation The existence of the 51- to 47-kDa doublet detected by immunoprobing RARP with the antibodies raised against the F region [RP@(F)2; see above] suggested to us the possible occurence of posttranslational modifications, such as phosphorylation, which is known to alter the electrophoretic mobility of some proteins (3436) and was previously reported for RARotl and RAR-yl (30, 31). To establish the possible nature of the modifications involved, we have analyzed the effects of treating mRARB proteins with phosphatases. Immunoprecipitates of mRARP1, $2, and $3 proteins synthesized in transfected COS-1 cells were treated with phosphatase in the presence or absence of sodium phosphate (a phosphatase inhibitor) and then immunoblotted with RPB(F)2. Neither calf intestinal nor bacterial alkaline phosphatase affected the mobility of mRARB proteins (data not shown), though they altered the electrophoretic mobility of RAR-yl (30) and RARotl (31) proteins. However, potato acidic phosphatase (PAP) treatment of RARPl, $2, and $3 immunoprecipitates induced the appearance of a faster migrating 47-kDa species (Fig. 6A, lanes 3, 6, and 9). Interestingly, and in contrast to RARal (31) and RARrl (30) the 51-kDa species did not disappear even after longer incubation times or with increased phosphatase concentrations (data not shown). The presence of sodium phosphate during the incubation prevented the appearance of the 47-kDa species, suggesting that this species corresponded to the removal of phosphate from RARP proteins (Fig. 6A, lanes 2, 5, and 8) rather than to a proteolytic product. Whether this 47-kDa species is identical to that seen in Figs. 1 and 2 is unknown. To confirm the phosphorylation of RARP proteins, transiently transfected COS-1 cells were labeled in vivo with 32P in the presence or absence of RA, and extracts were immunoprecipitated with Ab8P(F)2. After SDSPAGE and electrotransfer, subsequent autoradiography of the nitrocellulose (NC) filters showed that the three mRAR@ proteins were phosphorylated (Fig. 66, lanes 7-l 2). No phosphorylation was observed in COS1 cells transfected with the control expression vector

Vol6No.12

pSG5 (30) (data not shown). The phosphorylation of mRARP2 proteins occurred at a high level independently of RA and did not increase in the presence of RA (Fig. 66, lanes 9 and 10). In contrast, there was a low basal level of mRARP1 and mRARP3 phosphorylation, which was markedly increased in the presence of RA (Fig. 66, lanes 7, 8, 11, and 12). As indicated by immunoblotting of the same NC filters (Fig. 6B, lanes l-6), this increase cannot be ascribed to a RA-induced increase in the expression of the receptor proteins. It may reflect a higher rate of phosphate residue turnover in the presence of RA. The cDNA-deduced amino acid sequence of mRARP proteins shows the presence of a number of putative phosphorylation sites on serine, threonine, and tyrosine residues. Since CAMP has been implicated in concert with RA in F9 cell differentiation (37-39) we investigated whether RARP could be a substrate for CAMPdependent protein kinase (PKA). Bacterially expressed hRARP2 was incubated in the presence of [Y-~*P]ATP and PKA (see Materials and Methods). RARP2 proteins were immunoprecipitated, resolved by SDS-PAGE, electrotransferred onto NC filters, and autoradiographed. As shown in Fig. 7, PKA phosphorylated faintly the hRARP2 protein (Fig. 7, lane 2) revealed by subsequent immunoblotting of the same NC filters with RPp(F)2 (Fig. 7, lane 1). No signal at this position was observed when immunoprecipitation was performed with nonreactive ascite fluid (Fig. 7, lanes 3 and 4) or with extracts from untransformed bacteria (Fig. 7, lanes 5 and 6). In contrast, hRARy1, (Fig. 7, lanes 7 and 8) and hRARLu1 (data not shown) proteins expressed in E. co/i appeared to be phosphorylated by PKA to a higher level than RARPP. We then investigated whether RARP2 was phosphorylated on tyrosine residues with specific antibodies against phosphotyrosine (Sigma Chemical Co., St. Louis, MO). Using extracts of mRAR@2-transfected COS-1 cells and immunoblotting (Fig. 8A, lane 1) or immunoprecipitation (Fig. 8B, lane 2) the antiphosphotyrosine antibodies gave a signal with an apparent mol wt of 51 kDa, which comigrated with that obtained with an antibody raised against RARP [RPP(F)2; Fig. 88, lane I]. An identical signal was obtained with hRARP2 (Fig. 8A, lane 2). In contrast, no signal was seen with mRARa1 and mRARr1 expressed in COS-1 cells (Fig. 8A, lanes 3 and 4) or in control nontransfected cells (Fig. 8A, lane 5, and Fig. 88, lane 3). Mouse RAR/I2 was also synthesized in a rabbit reticulocyte lysate system by using in vitro generated mRARP2 mRNAs with the hope of obtaining a nonphosphorylated mRARP2 species (see Materials and Methods). A single mRARP2 species was revealed (Fig. 9, lanes 2-6) which comigrated with the 51-kDa species produced in COS-1 cells (Fig. 9, lane 1). No 47kDa species was generated by this method whether the 35S-labeled translation products were directly analyzed by autoradiography (Fig. 9, lane 5) or by immunoblotting with antibody raised against region F [RPP(F)2; Fig. 9, lanes 2-41. Moreover, no 47-kDa

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

RARB on Tyrosrne

Residues

2205

@ PAP NaPr kDa 103-

- + + - + -

-++ -+-

lmmunoblot

i3’p1

- ++ -+RA -+-+-+ 56

54- 67-

51-t

56-

56

54

4741--

54 a

51

51

28,1 cos-pi

,4 coq32

,7 cos-p3

Fig. 6. Phosphatase Treatment and in Vivo phosphorylation of mRARP Proteins A, Extracts from mRARP1 (lanes l-3)-, mRARP2 (lanes 4-6)-, mRARP3 (lanes 7-9)-transfected COB1 cells were immunoprecipitated using the Ab8P(F)2 (lanes l-3 and 7-9) or Ab6@2(A2) (lanes 4-6) monoclonal antibodies. Then, the antigen-antibody complexes immobilized on protein-A-Sepharose beads were incubated with (lanes 3, 6, and 9) or without (lanes 1, 4, and 7) 1 U PAP in the absence (lanes 1, 3, 4, 6, 7, and 9) or presence (lanes 2, 5, and 8) of 10 mM sodium phosphate. The incubated immunoprecipitates were then eluted, subjected to electrophoresls, and electrotransferred to NC filters. The mRAR@ proteins were identified by incubation of the filters with RPP(F)2 and [‘251]protern-A. B, COS-1 cells transfected with the mRARP1 (lanes 1, 2, 7, and 8) mRAFiP2 (lanes 3, 4, 9, and lo), and mRARP3 (lanes 5, 6, 11, and 12) expression vectors were labeled with 32P in the presence (lanes 2, 4, 6, 8, 10, and 12) or absence (lanes 1, 3, 5, 7, 9, and 11) of RA (lo-’ M) and subjected to immunoprecipitation with Ab8P(F)2. After electrophoresis and electrotransfer to NC filters, the phophorylated proteins were visualized by autoradiography (lanes 7-12). Proteins were identified as RARP by immunoblotting of the same NC filter with RPP(F)2, followed by alkaline phosphatase-labeled antirabbit antibodies (lanes l-6). Empty arowheads indicate contaminating immunoglobulins revealed by the alkaline phosphatase-labeled antirabbit antibodies.

Ab8P(F)2 II

NRA

Ab8p(F;2

Ab4y(hF)

kDa 10367-

498

generated after PAP treatment of immuof in vitro translated RAR/32 (data not shown). However, interestingly, a phosphotyrosine residue(s) was present in the 51-kDa mRARP2 species translated in vitro, as shown by immunoprecipitation species was noprecipitates

with antiphosphotyrosine antibodies and subsequent autoradiography (Fig. 9, lane 6). No signal was seen in control reticulocyte lysates under these conditions (Fig. 9, lane 7).

4128-

. DISCUSSION

Fig. 7. In Vitro Phosphorylation of RAR Proteins by PKA Bacterially expressed hRARP2 was incubated under in vitro phosphorylation conditions with PKA and [y-32P]ATP and analyzed by immunoprecipitation with Ab8P(F)2 (lanes 1 and 2) or nonreactive ascite fluid (lanes 3 and 4). After electrophoresis and electrotransfer to NC filters, the immunoprecipitated phosphorylated proteins were visualized by autoradiography (lanes 2 and 4) and identified by incubation of the same NC filter with RPP(F)2 and peroxidase-labeled protein-A (lanes 1 and 3). The specific complexes were revealed by chemiluminescence. Bacterially expressed hRARr1 (lanes 7 and 8) was phosphorylated, immunoprecipitated with Ab4r(F) (30), analyzed by autoradiography (lane 7) and identified by immunoblotting with RPy(hF) (lane 8). As controls, assays were carried out with untransformed bacterial cells immunoprecipitated with Ab8P(F)2 (lanes 5 and 6).

The present study reports the production of monoclonal and polyclonal antibodies directed against mouse and human RARP proteins and their use to characterize endogenous RARP in mouse embryos and EC cells. Three mouse monoclonal antibodies and four rabbit polyclonal antibodies directed against region A2 [Ab6P2(A2), Ab7P2(A2), and RPP2(A2)], region D2 [RPP(D2)], and region F [Ab8P(F)2, RPP(F)l, and RPP(F)2] were characterized. They recognized mRARP2 produced in transfected COS-1 cells, using immunoblotting, immunoprecipitation, and gel shift assays. In addition, the nuclear localization of RARPP was confirmed by immunocytochemistry (29) (data not shown). As expected, the antibodies directed against the F region [Ab8P(F)2, RPP(F)2, and RPP(F)l] recognized not only RARP2, but also the other isoforms [RARPl , $3, and $41 in extracts of transfected cells. This was

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

MOL ENDO. 1992 2206

Vol6No.12

@

AntiPhT

kDa

kDa

103-

10367-

67-41-

41-

28-

28-

W(F)2

- + +

In

3%

kDa u1*

,1&

10367 41 -

Fig. 8. lmmunoblotting and lmmunoprecipitation with Antiphosphotyrosine Antibodies A, Extracts from COS-1 cells transfected with hRARfi2 (lane l), mRARfl2 (lane 2), mRARa1 (lane 3) mRARr1 (lane 4)‘ or the parental pSG5 (lane 5) expression vectors were fractionated by SDS-PAGE, electrotransferred onto NC filters, and then immunoprobed with a mouse monoclonal antibody to phosphotyrosine (Sigma), followed by peroxidase-labeled antimouse immunoglobulins. The specific complexes were revealed by chemiluminescence. B, Extracts from COS-1 cells transfected with mRARP2 (lane 2) or the parental pSG5 (lane 3) expression vectors were immunoprecipitated with a mouse monoclonal antibody to phosphotyrosine, and the immunoprecipitated material was immunoprobed with RPP(F)2 and peroxidase-labeled protein-A. As positive controls, extracts (20 pg) of mRARb2 (lane l)-transfected COS-1 cells were loaded on the gel without prior immunoprecipitation and immunoprobed.

also the case with the rabbit polyclonal antibodies directed against the D2 region [RPP(D2); data not shown]. In contrast, the antibodies raised against the

A2 region[Ab6P2(A2),Ab7P2(A2),and RPP2(A2)]failed to recognize the RARPl, $3, and -04 isoforms, in agreementwith the presenceof different A regions(Al, A3, and A4) in these isoforms. EndogenousRARPproteinswere detected in mouse embryos and EC cells (F9 and Pl9 cell lines) using either immunoblottingand/or immunoprecipitation.In mouseembryos, RARPwas presentat all stagesstudied, from 9.5-14.5 days pc (as already observed for mRARal), with maximumamounts around 10.5 days pc. The major51-kDa speciesrevealedcorrespondsto RAR/I2, sinceit was immunoprecipitatedby monoclonal antibodiesdirectedagainsteither the A2 or F region. In EC cells(F9, P19, END-2, and MES-1 cell lines)grown in the absence of RA, RARP was undetectable by immunoblottingor immunoprecipitationassays. However, after a 24-h treatment of these four cell lineswith RA, RARPproteinsbecamedetectable by immunoblotting and were identifiedas mRARP2.It must be pointed out that higher RARP2protein levelswere observed in RA-treated Pl9 and MES-1 cellsthan in RA-treated F9 or END-2 cells, where a weaker induction was observed. In this respect, it is worth recalling that Pl9

28 123456,7 I

Fig. 9. Characterization of in Vitro Translated mRARp2 Proteins Mouse RARP2 mRNA was translated in vitro in a rabbit reticulocyte lysate for 75 min at 30 C. Aliquots (2 ~1, lanes 2 and 5; 5 ~1, lane 3; 10 ~1,lane 4) were analyzed by SDS-PAGE, electrotransfer and autoradiography (lane 5). or immunoblotting with RP@(F)S together with peroxidase-labeled protein-A (lanes 2-4). [%]Met-labeled mRARP2 translated in vitro (lane 6) and control reticulocyte lysates (lane 7) were also immunoprecipitated with antiphosphotyrosine antibodies and analyzed by SDS-PAGE, electrotransfer, and autoradiography. As positive controls, extracts (20 pg protein) of mRARflP-transfected COS-1 cells were run in parallel (lane 1).

cells,when grown in monolayer,differentiateupon 1O-’ M RA treatment as mesodermal-likecells resembling the MES-1 cell line (33), whereas F9 cells under the same conditions differentiate as endodermal-likecells possessingdifferentiation markers similarto those of END-2 cells (32). It has been previously shown that RARa is involved in differentiation of EC cells, since a truncated RARa fails to transactivate the RARP gene in the RA-resistant RAC-65 (40, 41). Mutagenesis or disruption by homologousrecombinationof the RAR@ gene should be helpful to define its role in specifying mesodermalor endodermaldifferentiationtype. We also investigated whether the mRARP proteins could be phosphorylated, as previously observed for RARa and RARy (30, 31). By usingin vivo 32Plabeling of transfected COS-1 cells, we demonstratedthat the mRARP2, $1, and $33 proteins are phosphotylated.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

RAR$ on Tyrosine Residues

The phosphorylation of RARP2 occurred regardless of the presence of RA, as previously observed for RARrl (30) and RARal (31). In contrast, RARPl and RARP3 proteins exhibited a basal phosphorylation that was markedly increased in the presence of RA. Putative phosphorylation sites for several kinases are present in the RARB proteins. Although phosphoacceptor residues have not as yet been mapped, we note that the Al and A3 domains of RARPl and RAR/33 contain a greater number of putative phosphorylation sites than the A2 domain of RARP2, including a calmodulin-dependent protein kinase site which is absent from the A2 domain of RARP2. Whether these putative sites are involved in the RA-dependent increase in phosphorylation of RARBI and RARP3 remains to be investigated by phosphopeptide mapping and mutagenesis. In any event, this difference in the RA dependency of RARP isoform phosphorylation supports our previous suggestion that the various isoforms may perform unique functions (9). It is well established that CAMP, in concert with RA, is implicated in F9 cell differentiation (37-39). In fact, CAMP regulates a variety of processes in mammalian cells by activating PKA, which, in turn, “regulates” gene expression by phosphorylating nuclear transcription factors (42). Interestingly, putative sites for PKA are mainly localized in the C and E regions of all three isotypes, RARul , $2, and -71. However, our present results show that bacterially expressed RARP2 appears to be only weakly phosphorylated by PKA in vitro, in contrast to RARcul and RARyl proteins, which are strongly phosphorylated by this kinase. This difference may be due to the presence of additional putative PKA sites in the F and A domains of RARLuI and RARyl or to differences in the accessibility of the phosphoacceptor serines/threonines. Several other putative phosphorylation sites for casein kinases I and II, prolinedependent protein kinase, and glycogen synthase kinase3 (43-45) are present in RARP proteins. Whether these kinases are able to phosphorylate RARP2 remains to be investigated. The phosphorylation of RARP2 appears to be different from that of RARal and RARrl. Indeed, the 51kDa RARotl and RARrl proteins synthesized in vivo in transfected COS-1 cells are sensitive to alkaline phosphatase treatment, resulting in their conversion into a dephosphorylated 47-kDa species (30, 31) which migrates at the same level as RARcvl and RARrl proteins synthesized in vitro in a reticulocyte lysate (data not shown). In contrast, the 51-kDa RARP2 species synthesized in vivo is resistant to alkaline phosphatase treatment and weakly sensitive to acidic phosphatase treatment. Furthermore, in vitro translation of mRARP2 mRNA results in the synthesis of a single 51-kDa protein without any 47-kDa species. In fact, by using an antiphosphotyrosine monoclonal antibody, we have shown that the 51-kDa RARP2 proteins produced either in vivo in transfected COS-1 cells or in vitro in reticulocyte lysates contain phosphotyrosine residues, whereas RARal ad RARyl do not. Phosphorylation on

2207

tyrosines concomitantly with the weak phosphorylation by serine-threonine kinases might explain the low sensitivity of RARP to phosphatases. Phosphopeptide mapping and phosphoaminoacid analysis will allow us to determine which residues are phosphorylated in RARP. The existence of a tyrosine kinase-signaling pathway in P19 EC cells (46) and the observation that RARP2 is phosphorylated in these cells upon RA treatment (our unpublished results) argue in favor of a specific role of tyrosine phosphorylation in the function of RARP2, with no counterpart in RARa and -y functions. Whether phosphorylation of RARs affects their DNA-binding activity, ligand-binding activity, and transactivation properties is under investigation.

MATERIALS

AND

Expression

Vectors

METHODS and Reporter

Genes

The plasmids containing the mouse and human RAR genecoding sequences RARnl, RARP2, and RARyl were previously described (30, 31). The construction of the isoform mRARD1, mRARp3, and mRARP4 (mRARP4 CTG) expression vectors has been reported (19, 26). The prokaryotic expression vector PET RARrl encoding hRARr1 was constructed as previously described (14). The PET RARP2 expression vector was constructed according to the same protocol (14) and was a generous gift from Jian-Yang Chen (manuscript in preparation). These isopropyl-thio galactoside G-inducible prokaryotic expression vectors encode fusion proteins with a higher mol wt than endogenous RARs, since they contain 15 amino acids from the parental vector PET-3a (47) fused to the N-terminal amino acid of full-length hRARB2 or hRARr1. Synthesis Monoclonal

of Peptides, Antibodies

and Preparation

of Antisera

and

The synthetic peptides SP248 (A2 region of human and mouse RARP2), SPA39 (D2 region of human and mouse RARP), SP172 (F region of hRARp; see Table 1) deduced from the cDNA sequences of human and mouse RARB2 were synthesized in solid phase, as previously described (29, 30). Fp (DHFR) is a fusion protein between the F region of hRARP2 and human dihydrofolate reductase (gift of Hoffman LaRoche, Nutley, NJ). Rabbit antisera and monoclonal antibodies were generated as previously described (29, 30). Cell Cultures

and Extract

Preparations

WCE were prepared from COS-1 cells routinely grown and transfected with mRARP2, $1, $3, or $34 expression vectors (30). NE were prepared from confluent EC, F9, and P19,6 cells grown as monolayers on petri dishes and treated or not for 48 h with lo-’ M RA (30, 31). The same protocol was used for PlS-derived clonal cell lines (MES-1 and END-2 cells lines), which were routinely cultured according to the method of Mummery et al. (32, 33). Mouse embryos were collected at 9.5, 10.5, 11.5, 12.5, 13.5, and 14.5 days pc, and NE were prepared as described for EC cells, except that the crude nuclear pellets were further purified on a sucrose cushion (30, 31). Production of bacterially expressed hRARr1 and hRARP2 has been previously described (14). Immunoblotting, Assays

Immunoprecipitation,

and Gel Shift

WCE from RARP-transfected COS-1 cells (50 pg protein) or NE from mouse embryos or EC cells (300 pg protein) were

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

MOL 2208

ENDO.

Vol6No.12

1992

immunoprecipitated according to the protocol described by Rochette-Egly et al. (30). Proteins (with or without previous immunoprecipitation) were fractionated by SDS-PAGE (10% polyacrylamide), electrotransferred onto a NC filter, and immunoprobed using appropriate antibodies (generally diluted 1:500), followed by [‘?]protein-A, [1251]goat immunoglobulins, antimouse peroxidase-labeled protein-A, or peroxidase-labeled antimouse immunoglobulins. In the two latter cases, specific complexes were revealed by chemiluminescence according to the manufacturer’s protocol (Amersham, Aylesbury, Buckinghamshire, United Kingdom). The specificity of the reaction was checked by depleting the antisera from the specific antibodies, as previously described (30). Mobility shift assays were performed with WCE (5 Kg) from RARB-transfected COS-1 cells, incubated as previously reported (30, 48) with wild-type or mutated double stranded oligonucleotides (RAREP and RAREPm, respectively) corresponding to the RARE of the RARP2 promoter (19, 25, 27). Alkaline Phosphatase Vitro and in Vivo

Treatment

and Phosphorylation

in

Calf intestinal alkaline phosphatase or PAP (Boehringer, Mannheim, Germany) treatment of RARP immunoprecipitates from transfected cell WCE was performed as previously reported (30). For in vitro phosphorylation, bacterially expressed hRAR/32 was incubated in a final reaction volume of 100 ~1 with PKA buffer [lo mM Tris-HCI (pH 7.5), 5 mM MgCl*, 1 mM CaCI,, and 100 PM unlabeled ATP] and 1 &i [y--3’P]ATP. The reaction was initiated by the addition of 1 U PKA catalytic subunit (Sigma) and incubated at 30 C for 1 h. Phosphorylated hRARP2 was isolated by immunoprecipitation and analyzed by SDS-PAGE. In viva 32P phosphorylation of mRAR@ isoforms was performed in transfected COS-1 cells, following the protocol described by Rochette-Egly et al. (30). In Vitro Transcription

and Translation

Before in vitro transcription, full-length RARP2, -yl, or -0tl expression vectors (30, 31) were linearized with Xbal. Two micrograms of each linearized cDNA were transcribed in vitro using T7 RNA polymerase in a final reaction volume of 100 II (49). In vitro translation was performed using a rabbit reticuiocite lysate according to the.manufacturer’s-(Promega, Madison, WI) instructions. Translation products labeled with [%I methionine were analyzed by SDS-PAGE, electrotransfer, and subsequent autoradiography or immunoblotting with RP@(F)2. Acknowledgments We are grateful to J. Y. Chen, P. Kastner, A. Krust, P. Leroy, and S. Nagpal for gifts of RAR expression vectors, Hoffman LaRoche (Basel, Switzerland) for the gift of the fusion protein FP-DHFR, and C. L. Mummery for the gift of MES-1 and END2 cell lines. We wish to thank C. Kbdinger for critically reading the manuscript. We also thank V. Schultz and N. Yung for their help in the preparation of the monoclonal antibodies, G. Duval for the rabbit injections, A. Chevalier for the preparation of the synthetic peptides, the cell culture group for maintaining and providing cells, and M. Lemeur for providing mouse embryos. We thank the illustration and secretarial staff for their help in preparing the manuscript.

Received August 4, 1992. Revision received October 7, 1992. Accepted October 7, 1992. Address requests for reprints to: Dr. Pierre Chambon, Unite 184 de Biologie Mol&ulaire et de GEnie GBnBtique, I’INSERM, lnstitut de Chimie Biologique, Faculti, de Mgdecine, 11 rue Humann, 67085 Strasbourg Cedex, France. This work was supported by funds from the lnstitut National de la Sante et de la Recherche MBdicale, the Centre National

de la Recherche Scientifique, the Centre Hospitalier Universitaire RBgional, the Association pour la Recherche sur le Cancer, the Fondation pour la Recherche MBdicale, and the Human Frontier Science Program.

REFERENCES

1. Tabin CJ 1991 Retinoids, homeoboxes and growth factors; toward molecular models for limb development. Cell 66:199-217 2. Thaller C, Smith SM, Eichele G 1991 Retinoids in vertebrate development: pattern formation in limbs and in the central nervous system. In: Saurat JH (ed) Retinoids: 10 Years On. Basel, Karger, pp 89-l 08 3. Tickle C, Brickell PM 1991 Retinoic acid and limb development. Semin Dev Biol 2:189-l 97 4. Ragsdale Jr CW, Brockes JP 1991 Retinoids and their targets in vertebrate development. Curr Biol3:928-934 5. Jesse1 TM, Melton DA 1992 Diffusible factors in vertebrate embryonic induction. Cell 68:257-270 6. Mendelsohn C, Ruberte E, Chambon P 1992 Retinoids in vertebrate limb development. Dev Biol 152:50-61 7. Maden M 1991 Retinoid-binding proteins in the embryo. Semin Dev Biol 2:161-l 70 8. Blomhoff R, Green MH, Green JB, Berg T, Norum KR 1991 Vitamin A metabolism: new perspectives on absorption, transport and storage. Physiol Rev 71:951-989 9. Chambon P, Zelent A, Petkovich M, Mendelsohn C, Leroy P, Krust A, Kastner P, Brand N 1991 The family of retinoic acid nuclear receptors. In: Saurat JH (ed) Retinoids: 10 Years On. Basel, Karger, pp 1 O-27 10. DeLuca LM 1991 Retinoids and their receptors in differentiation embryogenesis and neoplasia. FASEB J 5:29242933 11. Mangelsdorf DJ, Ong ES, Dyck JA, Evans RM 1990 Nuclear receptor that identifies a novel retinoic acid response pathway. Nature 345:224-229 12. Mangelsdorf DJ, Borgmeyer U, Heyman RA, Zhou JY, Ong ES, Oro AE, Kakizuka A, Evans RM 1992 Characterization of three RXR genes that mediate the action of 9-cis retinoic acid. Genes Dev 6:329-344 13. Rowe A, Eager NSC, Brickell PM 1991 A member of the RXR nuclear receptor family is expressed in neural-crestderived cells of the developing chick peripheral nervous system. Development 111:771-778 14. Leid M, Kastner P, Lyons R, Nakshatri H, Saunders M, Zacharewski T, Chen JY, Staub A, Garnier JM, Mader S, Chambon P 1992 Purification, cloning, and RXR identity of the HeLa cell factor with which RAR or TR heterodimerizes to bind target sequences efficiently. Cell 68:377395 15. Evans RM 1988 The steroid and thyroid hormone receptor superfamily. Science 240:889-895 16. Green S, Chambon P 1988 Nuclear receptors enhance our understanding of transcription regulation. Trends Genet 4:309-314 17. Beato M 1989 Gene regulation by steroid hormones. Cell 56:335-344 18. Leroy P, Krust A, Zelent A, Mendelsohn C, Garnier JM, Kastner P, Dierich A, Chambon P 1991 Multiple isoforms of the mouse retinoic acid receptor LY are generated by alternative splicing and differential induction by retinoic acid. EMBO J 10:59-69 19. Zelent A, Mendelsohn C, Kastner P, Krust A, Garnier JM, Ruffenach F, Leroy P, Chambon P 1991 Differentially expressed isoforms of the mouse retinoic acid receptor /3 are generated by usage of two promoters and alternative splicing. EMBO J 10:71-81 20. Kastner P, Krust A, Mendelsohn C, Garnier JM, Zelent A, Leroy P, Staub A, Chambon P 1990 Murine isoforms of

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.

RAP@ on Tyrosine

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Residues

retinoic acid receptor y with specific patterns of expression. Proc Nat1 Acad Sci USA 87:2700-2704 Dolle P, Ruberte E, Kastner P, Petkovich M, Stoner CM, Gudas LJ, Chambon P 1989 Differential expression of genes encoding N, p and y retinoic acid receptors and CRABP in the developing limbs of the mouse. Nature 342:702-705 Dolle P, Ruberte E, Leroy P, Morriss-Kay G, Chambon P 1990 Retinoic acid receptors and cellular retinoid binding proteins. I. A systematic study of their differential pattern of transcription during mouse embryogenesis. Development 110:1133-1151 Ruberte E, Dolle P, Krust A, Zelent A, Morriss-Kay G, Chambon P 1990 Specific spatial and temporal distribution of retinoic acid receptor gamma transcripts during mouse embryogenesis. Development 108:213-222 Ruberte E, Dolle P, Chambon P, Morriss-Kay G 1991 Retinoic acid receptors and cellular retinoid binding proteins. II. Their differential pattern of transcription during early morphogenesis in mouse embryos. Development 111:45-60 Mendelsohn C, Ruberte E, LeMeur M, Morriss-Kay G, Chambon P 1991 Developmental analysis of the retinoic acid-inducible RARP2 promoter in transgenic animals. Development 113:723-734 Nagpal S, Zelent A, Chambon P 1992 RARP4, a novel retinoic acid receptor isoform, is generated from RARP2 by alternative splicing and usage of a CUG initiator codon. Proc Natl Acad Sci USA 89:2718-2722 The H (de), Vivanco-Ruiz Mhdel Mar, Tiollais P, Stunnenberg H, Dejean A 1990 Identification of a retinoic acid responsive element in the retinoic acid receptor fl gene. Nature 343:177-l 80 Sucov HM, Murakami KK, Evans RM 1990 Characterization of an autoregulated response element in the mouse retinoic acid receptor type p gene. Proc Natl Acad Sci USA 87:5392-5396 Gaub MP, Lutz Y, Ruberte E, Petkovich M, Brand N, Chambon P 1989 Antibodies specific to the retinoic acid human nuclear receptors (Y and p. Proc Natl Acad Sci USA 86:3089-3093 Rochette-Egly C, Lutz Y, Saunders M, Scheuer I, Gaub MP, Chambon P 1991 Retinoic acid receptor y: specific immunodetection and phosphorylation. J Cell Biol 115X535-545 Gaub MP, Rochette-Egly C, Lutz Y, Ali S, Matthes H, Scheuer I, Chambon P 1992 lmmunodetection of multiple species of retinoic acid receptor alpha; evidence for phosphorylation. Exp Cell Res 201:335-346 Mummery CL, Feijen A, Van der Saag PT, Van den Brink CE, DeLaat SW 1985 Clonal variants of differentiated P19 embryonal carcinoma cells exhibit epidermal growth factor receptor kinase activity. Dev Biol 109:402-410 Mummery CL, Feijen A, Moolenaar WH, Van den Brink CE, DeLaat SW 1986 Establishment of a differentiated

34.

35.

36.

37.

38.

39. 40.

41.

42.

43. 44. 45.

46.

47.

48.

49.

mesodermal line from P19 EC cells expressing functional PDGF and EGF receptors. Exp Cell Res 165:229-242 Hoeck W, Rusconi S, Groner B 1989 Down-regulation and phosphorylation of glucocorticoid receptors in cultured cells. Investigations with a monospecific antiserum against a bacterially expressed receptor fragment. J Biol Chem 264:14396-l 4402 Mylin LM, Bhat JP, Hopper JE 1989 Regulated phosphorylation and dephosphorylation of GAL4, a transcriptional activator. Genes Dev 3:1157-l 165 Denner LA, Schrader WT, O’Malley BW, Weigel NL 1990 Hormonal regulation and identification of chicken progesterone receptor phosphorylation sites. J Biol Chem 265:16548-l 6555 Strickland S, Smith KK, Marotti KR 1980 Hormonal induction of differentiation in teratocarcinoma stem cells; generation of partial endoderm by retinoic acid and dibutyryl CAMP. Cell 211347-355 Hu L, Gudas LJ 1990 Cvclic AMP analoqs and retinoic acid influence the expression of retinoic acid receptor (Y, B and y mRNAs in F9 teratocarcinoma cells. Mol Cell Biol i 0:391-396 Gudas LJ 1991 Retinoic acid and teratocarcinoma stem cells. Semin Dev Biol 2:171-l 79 Kruyt FAE, Van den Brink CE, Defize LHK, Donath MJ, Kastner P, Kruijer W, Chambon P, Van der Saag PT 1991 Transcriptional regulation of retinoic acid receptor @ in retinoic acid-sensitive and -resistant P19 embrvocarcinoma cells. Mech Dev 33:171-l 78 Kruvt FAE. Van der Veer LJ. Mader S. Van den Brink CE, Feijkn A, Jonk LJC, Kruijer W, Van der Saag PT 1992 Retinoic acid resistance of the variant embryonal carcinoma cell line RAC65 is caused by expression of a truncated RARn. Differentiation 49:27-37 Gonzalez GA, Montminv MR 1989 Cvclic AMP stimulates somatostatin gene transcription by- phosphorylation of CREB at serine 133. Cell 59:675-680 Kemp BE, Pearson RB 1990 Protein kinase recognition sequence motifs. Trends Biochem Sci 15:342-346 Roach PJ 1991 Multisite and hierarchal protein phosphorylation. J Biol Chem 266:14139-l 4142 Kennelly PJ, Krebs EG 1991 Consensus sequences as substrate specificity determinants for protein kinases and protein phosphatases. J Biol Chem 266:15555-l 5558 Den Hertog J, Eman R, Tertoolen LGJ, de Laat SW, Kruijer W 1991 Characterization of a tyrosine kinase signaling pathway in undifferentiated P19 embryonal carcinoma cells. Exp Cell Res 196:226-232 Studier WF, Rosenberg AH, Dunn JJ, Dubendorff JW 1991 Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymoll85:60-89 Nicholson RC, Mader S, Nagpal S, Leid M, Rochette-Egly C, Chambon P 1990 Negative regulation of the rat stromelysin gene promoter by retinoic acid is mediated by an API binding site. EMBO J 9:4443-4454 Kumar V, Green S, Staub A, Chambon P 1986 Localisation of the oestradiol and putative DNA-binding domains of the human oestrogen receptor. EMBO J 5:2231-2236

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 January 2015. at 18:08 For personal use only. No other uses without permission. . All rights reserved.