Journal of Histochemistry & Cytochemistry

Journal of Histochemistry & Cytochemistry http://jhc.sagepub.com/

Nuclear Detection of Cellular Retinoic Acid Binding Proteins I and II with New Antibodies Marie-Pierre Gaub, Yves Lutz, Norbert B. Ghyselinck, Isabelle Scheuer, Véronique Pfister, Pierre Chambon and Cécile Rochette-Egly J Histochem Cytochem 1998 46: 1103 DOI: 10.1177/002215549804601002 The online version of this article can be found at: http://jhc.sagepub.com/content/46/10/1103

Published by: http://www.sagepublications.com

On behalf of:

Official Journal of The Histochemical Society

Additional services and information for Journal of Histochemistry & Cytochemistry can be found at: Email Alerts: http://jhc.sagepub.com/cgi/alerts Subscriptions: http://jhc.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Oct 1, 1998 What is This?

Downloaded from jhc.sagepub.com by guest on October 11, 2013

Volume 46(10): 1103–1111, 1998 The Journal of Histochemistry & Cytochemistry

http://www.jhc.org

ARTICLE

Nuclear Detection of Cellular Retinoic Acid Binding Proteins I and II with New Antibodies Marie-Pierre Gaub, Yves Lutz, Norbert B. Ghyselinck, Isabelle Scheuer,Véronique Pfister, Pierre Chambon, and Cécile Rochette–Egly Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP/Collège de France, France

Apart from the retinoic acid nuclear receptor family, there are two low molecular weight (15 kD) cellular retinoic acid binding proteins, named CRABPI and II. Mouse monoclonal and rabbit polyclonal antibodies were raised against these proteins by using as antigens either synthetic peptides corresponding to amino acid sequences unique to CRABPI or CRABPII, or purified CRABP proteins expressed in E. coli. Antibodies specific for mouse and/or human CRABPI and CRABPII were obtained and characterized by immunocytochemistry and immunoblotting. They allowed the detection not only of CRABPI but also of CRABPII in both nuclear and cytosolic extracts from transfected COS-1 cells, mouse embryos, and various cell lines. (J Histochem Cytochem 46:1103–1111, 1998)

SUMMARY

Retinoic acid (RA), the major active derivative of vitamin A, is indispensable for vertebrate development and homeostasis (Blomhoff 1994; Sporn et al. 1994). The RA signal is transduced by two families of nuclear receptors, the retinoic acid receptors (RARs) and the retinoid X receptors (RXRs), which act as ligandinducible transcriptional regulators (for reviews see Mangelsdorf and Evans 1995; Chambon 1994,1996). Mouse knockout mutants for either RARs or RXRs recapitulate the complete spectrum of defects previously associated with the fetal vitamin A deficiency syndrome, thus demonstrating the role of RARs and RXRs in the RA signaling pathway (Kastner et al. 1995,1997; Ghyselinck al. 1997 and references therein). In addition to RA nuclear receptors, there exist two low molecular mass (
15 kD) cytoplasmic RA binding proteins named cellular retinoic acid binding proteins I and II (CRABPI and II). These proteins are found in all vertebrates (for reviews see Napoli 1994; Ong et al. 1994) and are highly conserved (73%) across species. In addition, their structure is related to the fatty acid binding protein family (Kleywegt et al. Correspondence to: Dr. Cécile Rochette–Egly, Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), BP 163, 67404 Illkirch Cedex, C.U. de Strasbourg, France. Received for publication April 8, 1998; accepted June 9, 1998 (8A4644). © The Histochemical Society, Inc.

0022-1554/98/$3.30

KEY WORDS cellular retinoic acid binding proteins antibodies immunodetection nuclear localization

1994). Both CRABPI and II proteins bind all trans RA, although the affinity of CRABPII for RA is lower than that of CRABPI (Ong et al. 1994). In situ hybridization has shown that both CRABPs are expressed in mouse embryos but in a nonoverlapping pattern (Ruberte et al. 1992; Gustafson et al. 1993; Lyn and Giguère 1994; Maden 1994). However, in adult mice CRABPI is widely distributed, whereas CRABPII is restricted to the skin (Ong et al. 1994). CRABPs expression has been correlated with vulnerability to RA excess (Balling 1991; Gustafson et al. 1993; Lyn and Giguère 1994; Maden 1994). Therefore, it has been proposed that CRABPs could control the level of intracellular RA available for binding to the nuclear receptors, acting either by soaking up free retinoic acid and/or as modulators of RA metabolism, thus protecting the cell from RA excess (for reviews see Balling 1991; Napoli 1994; Ong et al. 1994). Furthermore, although it has been proposed that CRABPs may shuttle RA to the nuclear receptors (Takase et al. 1986), it is not yet clear whether CRABPs enter the nucleus. In fact, the role of CRABPs in the RA signaling pathway is far from being elucidated, since simple or combined knockout of CRABPI and II in mice resulted in normal, viable, and fertile adult animals (Gorry et al. 1994; Lampron et al. 1995). Up to now, the expression and distribution of CRABPs have been studied mainly at the mRNA level 1103

1104

Gaub, Lutz, Ghyselinck, Scheuer, Pfister, Chambon, Rochette–Egly

by in situ hybridization. Perhaps because of the difficulty in obtaining high-titer sensitive antisera to CRABPs, which are highly conserved among species, few studies report the expression of the corresponding proteins (Maden et al. 1992; Bucco et al. 1996; Gustafson et al. 1996; Venepally et al. 1996; Zheng et al. 1996). Moreover, these studies are in disagreement about the subcellular localization of CRABPI and CRABPII. Therefore, we raised specific antibodies to investigate the subcellular distribution of CRABPI and II proteins in several mouse and human cell lines and during mouse embryonic development.

Materials and Methods Synthesis of Peptides, Preparation of Antisera, and Monoclonal Antibodies Synthetic peptides corresponding to amino acids 95–108 from mouse (m) CRABPI (SPB 63, CTQTLLEGDGPKTY) and CRABPII (SPB 64, CEQRLLKGEGPKTS) (Stoner and Gudas 1989; Giguère et al. 1990; Aström et al. 1991) were selected on the basis of their hydrophilicity and mobility and were synthesized as described (Rochette–Egly et al. 1991). These peptides are conserved between mouse and human except at one amino acid (position 98) for CRABPII (Aström et al. 1991). Purified human CRABPI and II produced in E. coli (Fogh et al. 1993) were also used as antigens. Rabbit antisera and mouse monoclonal antibody ascites fluids were prepared as previously described (Gaub et al. 1989; Rochette–Egly et al. 1991).

copy using a confocal laser scanning microscope. Controls were performed with untransfected cells.

Cells, Extracts, and Immunoblotting Mouse embryonal carcinoma cells (P19 cell line), human colorectal carcinoma cells (Caco-2 cell line), and human breast pleural metastasis cells (MCF7 cell line), were grown on petri dishes in appropriate medium (Pinto et al. 1983; Gaub et al. 1992; Widschwendter et al. 1995). Nuclear (N) and cytosolic extracts (C) were prepared from these cells as described (Rochette–Egly et al. 1991). Mouse embryos were collected at different days post coitum (dpc) and nuclear and cytosolic extracts were prepared (Rochette–Egly et al. 1991). Whole cell extracts (WCE) from transfected COS-1 cells were prepared as previously described (Rochette–Egly et al. 1991). Extracts were fractionated by SDS-PAGE (15% acrylamide) and electrotransferred onto nitrocellulose (NC) filters. After blocking in PBS–3% nonfat powdered milk, the filters were immunoprobed with the specific antibodies for 2 h at 37C. The polyclonal antibodies were diluted 1:200 and the monoclonal antibodies 1:1000. After extensive washing in PBS containing 0.05% Tween-20, the filters were then incubated for 30 min at RT with peroxidase-labeled anti-mouse immunoglobulins or protein A, diluted 1:10,000 (Amersham, Poole, UK). Specific complexes were revealed by chemiluminescence detection according to the manufacturer’s protocol (Amersham). The specificity of the reactions was checked by depleting the antisera from the specific antibodies by incubation with the corresponding peptide (Gaub et al. 1989; Rochette–Egly et al. 1991).

Expression Vectors and Transfections

Results

The mCRABPI and mCRABPII pSG5 expression vectors (Stoner and Gudas 1989; Giguère et al., 1990; Aström et al. 1991) were previously described (Gorry et al. 1994). COS-1 cells maintained in Dulbecco’s modified Eagle’s medium containing 5% fetal calf serum were transfected by the calcium phosphate technique as described (Rochette–Egly et al. 1991).

Preparation and Characterization of Mouse Monoclonal Antibodies Against CRABPI and II

Immunocytochemistry Transfected COS cells were grown on Leighton tubes (Costar; Cambridge, MA), washed in PBS, pH 7, and fixed with either paraformaldehyde (2% PFA, pH 7.4) or Bouin’s fixative (10% PFA, 5% acetic acid, 2% picric acid) for 4 min at room temperature (RT). Cells were incubated overnight at RT in a humidified chamber with the monoclonal antibodies against either CRABPI (4CRA3B1) or CRABPII (5CRA3B3) purified using caprylic acid and ammonium sulfate precipitation (Harlow and Lane 1988), diluted at 5
g/ml in culture medium, or with normal mouse IgG as controls. After extensive washing in PBS containing 0.1% Triton X-100, bound immunoglobulins were visualized by an antimouse indocarbocyanine (Cy3)-conjugated secondary antibody (Jackson Immunoresearch Laboratories; West Chester, PA) diluted 1:400 in PBS. The cells were then washed in PBS–Triton and the nuclei were counterstained with Hoechst 33258. After washings, the cells were analyzed by fluorescence micros-

To generate specific mouse monoclonal antibodies, two peptides localized in the central part of mCRABPI (SPB 63) and mCRABPII (SPB 64) proteins (Stoner and Gudas 1989; Giguère et al. 1990; Aström et al. 1991) were selected and synthesized. Two hybridomas were obtained after immunization with SPB 63 (3CRA10F5 and 3CRA10G2) and one with SPB64 (1CRA4C9). Purified human CRABPI or CRABPII proteins overexpressed in E. coli were also used as antigens and gave rise to the hybridomas 4CRA3B1 and 5CRA3B3, respectively (see Table 1). The monoclonal antibodies produced by these hybridomas recognized specifically the cognate peptide or purified protein as checked by ELISA (data not shown). Note that neither 5CRA3B3 or 4CRA3B1 recognized the synthetic peptides SPB 64 and SPB 63, respectively (data not shown). These monoclonal antibodies were identified as IgG1 except for 5CRA3B3, which was IgG2a. The monoclonal antibodies were tested for their ability to detect by Western blotting the recombinant mCRABP proteins overexpressed in COS-1 cells. Figure 1A shows that 1CRA4C9 and 5CRA3B3 recog-

1105

Nuclear Localization of Cellular Retinoic Acid Binding Proteins Table 1 Characteristics of anti-CRABP antibodies Immunocytochemistry CRABPI

Western blotting

CRABPII

CRABPI

CRABPII

Antibody

Antigen

Mouse

Human

Mouse

Human

Mouse

Human

Mouse

Human

204

SPB 63 (mCRABPI)



ND



ND



ND



ND

206

SPB63 (mCRABPI)

















207

SPB64 (mCRABPII)



ND



ND



ND



ND

208

SPB 64 (mCRABPII)

















3CRA10F5

SPB 63 (mCRABPI)

















3CRA10G2

SPB 63 (mCRABPI)

















4CRA3B1

hCRABPI

















1CRA4C9

SPB 64 (mCRABPII)

















5CRA3B3

hCRABPII

















nize mCRABPII (Figure 1, Lanes 2 and 4) as a single band migrating close to 15 kD, the expected molecular weight predicted from the cDNA sequence. No crossreactions were seen with extracts from mCRABPItransfected COS-1 cells (Figure 1, Lanes 1 and 3), confirming that these antibodies are specific for CRABPII. In contrast, 3CRA10F5 and 4CRA3B1 recognize specifically mCRABPI (Figure 1, Lanes 5–8). Note that 3CRA10G2 recognize both mCRABPI and II proteins (Figure 1, Lanes 9 and 10). The fact that the 16 aminoacid sequence of mCRABPI, between amino acids 95 and 108 (SPB 63), differs from that of mCRABPII (SPB 64) by eight amino acids may explain such a crossreactivity, thus suggesting the existence of a common epitope between both proteins. The specificity of the signals was corroborated by using either extracts from COS-1 cells transfected with a control expression vector or antisera immunoabsorbed with the corresponding peptide (data not shown). Immunoblotting Detection of CRABPs in Mouse Embryos and Cultured Cell Lines Using the Monoclonal Antibodies

The monoclonal antibodies were used for detection of endogenous CRABP proteins by immunoblotting. First, to check the specificity of these antibodies, cytosolic extracts from 13.5 dpc CRABPI//CRABPII/ double mutant mice (Gorry et al. 1994; Lampron et al. 1995) were analyzed by Western blotting. As shown in Figure 2A, no signal was detected in double mutant extracts (Figure 2, Lanes 2 and 6) with 1CRA4C9 and 3CRA10F5 antibodies, whereas CRABPs were de-

tected in cytosolic extracts from wild-type embryos (Figure 2, Lanes 1 and 5). Interestingly, both antibodies revealed a strong signal corresponding to mCRABPI or II, not only in the cytosol but also in the nucleus of 13.5 dpc mouse embryos (Figure 2B, upper panels; Lanes 6, 7, 13, and 14). Such a nuclear signal is specific because it was absent from nuclear extracts of CRABPI//CRABPII/ double mutant mice (data not shown). In addition, these signals disappeared when the monoclonal antibodies were immunodepleted with the cognate peptide (Figure 2B, lower panels). 1CRA4C9 and 3CRA10F5 also detected CRABPII and CRABPI proteins, respectively, in both cytoplasmic and nuclear extracts from mouse P19 cells (Figure 2B, Lanes 2–5 and 9–12). The signal corresponding to CRABPII was increased on treatment with RA (107 M) in both the cytosolic (Figure 3B, Lanes 2 and 4) and nuclear extracts (Figure 2B, Lanes 3 and 5), in agreement with RNA studies and the characterization of an RA response element in the promoter region of the mCRABPII gene (Durand et al. 1992; Aström et al. 1994). In contrast, the signal corresponding to CRABPI was not affected (Figure 2B, Lanes 9–12). Then we checked whether these monoclonal antibodies were also able to recognize human CRABPs by using extracts from the human Caco-2 (Figures 3A and 3B) and MCF7 (Figure 3C) cell lines. Unfortunately, 1CRA4C9 did not recognize human CRABPII in these cells (Figure 3A, Lane 2; and data not shown). However, 5CRA3B3, also raised against CRABPII, was able to detect the human protein in cytoplasmic extracts from Caco-2 cells (Figures 3A, Lane 4, and

1106

Gaub, Lutz, Ghyselinck, Scheuer, Pfister, Chambon, Rochette–Egly

Figure 1 Characterization of CRABPI and CRABPII antibodies by immunoblotting. Whole cell extracts (5
g) from COS-1 cells transfected with either mCRABPII or mCRABPI expression vectors were fractionated by SDS-PAGE, electrotransferred onto NC filters and immunoprobed. (A) With the mouse monoclonal antibodies 1CRA4C9 (Lanes 1 and 2), 5CRA3B3 (Lanes 3 and 4), 3CRA10F5 (Lanes 5 and 6), 4CRA3B1 (Lanes 7 and 8), and 3CRA10G2 (Lanes 9 and 10), followed by peroxidase-labeled anti-mouse immunoglobulins and chemiluminescence detection. (B) With antisera 204 (Lanes 1 and 5), 206 (Lanes 2 and 6), 207 (Lanes 3 and 7), and 208 (Lanes 4 and 8), followed by peroxidase-labeled protein A and chemiluminescence detection. The position of the prestained molecular weight standards (Biorad; Hercules, CA) is indicated in kilodaltons.

3B, Lane 2). The monoclonal antibodies raised against CRABPI, 3CRA10F5, and 4CRA3B1 also recognized the human protein in Caco-2 cells (Figures 3A, Lane 6, and 3B, Lane 5). Finally, 3CRA10G2, which was found to recognize both mouse CRABPI and II, was also able to detect endogenous human CRABPs in these cells (Figure 3A, Lane 8). It must be stressed that, with our antibodies, both CRABPI and CRABPII proteins were confined to the cytoplasm of Caco-2 cells and were undetectable in the nuclei of these cells (Figure 3B, Lanes 2, 3, 5, and 6). In contrast, CRABPII was present in both cytoplasmic and nuclear extracts from MCF7 cells (Figure 3C, Lanes 2 and 3). However, no CRABPI was detectable in MCF7 cells with either 3CRA10F5 (Figure 3C, Lanes 5 and 6) or 4CRA3B1 (data not shown).

Figure 2 Immunoblotting of CRABP proteins in mouse embryos and P19 cells extracts with monoclonal antibodies. (A) Western blot analysis of CRABPs in cytosolic extracts (50
g) from wild-type (Lanes 1 and 5) and double CRABPI//CRABPII/ mutant (Lanes 2 and 6) mouse embryos (13.5 dpc). Lanes 3, 4, 7, and 8 correspond to control WCE from COS cells transfected with mCRABPI and mCRABPII expression vectors. Immunodetection was performed with 1CRA4C9 (Lanes 1–4) and 3CRA10F5 (Lanes 5–8) diluted 1:1000. B. Nuclear (N) and cytosolic (C) extracts (50
g) from P19 cells, with (Lanes 4, 5, 11, and 12) or without (Lanes 2, 3, 9, and 10) a 24-hr treatment with 107M RA, and from 13.5 dpc mouse embryos (Lanes 6, 7, 13, and 14) were fractionated by SDS-PAGE and electrotransferred onto NC filters. Immunodetection was performed with 1CRA4C9 (Lanes 1–7) or 3CRA10F5 (Lanes 8–14) diluted 1:1000. The specificity of the signals was checked with the same antibodies immunodepleted with the corresponding peptide (lower panels). Lanes 1 and 8 show WCE (5
g) from COS-1 cells transfected with either mCRABPII or mCRABPI expression vectors as controls.

Immunolocalization of CRABPs in Transfected COS-1 Cells Using the Monoclonal Antibodies

The monoclonal antibodies were also tested for their ability to recognize, by immunocytochemistry, mouse CRABPs overexpressed in COS-1 cells (Figure 4). When fixation was performed with PFA (2%, pH 7.4), 4CRA3B1 and 5CRA3B3 revealed a strong signal in nuclei of cells transfected with CRABPI and CRABPII, respectively (Figures 4A1 and 4A4). In the positive nuclei, CRABPs were evenly localized, excluding nucleoli. It must be stressed that about 10% of the labeled

Nuclear Localization of Cellular Retinoic Acid Binding Proteins

1107

4B1, and 4B4). This staining was homogeneous and excluded nuclei. With both fixation procedures, none of these antibodies showed any crossreaction towards the noncorresponding CRABP (Figure 4A2, 4A3, 4B2, and 4B3). Similar results were obtained with 3CRA10F5 and 1CRA4C9 (data not shown). According to these results, one can speculate that PFA and Bouin’s fixative preferentially reveal the epitopes of CRABPs confined in the nuclei or the cytoplasm, respectively. Such a hypothesis is corroborated by the observation that, in transfected COS cells, CRABPs are detected in both the cytosolic and nuclear compartments by Western blot with the same antibodies (data not shown). Nevertheless, it cannot be excluded that Bouin’s fixation (acidic conditions) promotes the diffusion of CRABPs from nuclei into the cytoplasm. Rabbit Polyclonal Antibodies Against mCRABPI and II

Figure 3 Immunoblotting of CRABP proteins in human cell lines with monoclonal antibodies. (A) Cytosolic extracts (50
g) from Caco-2 cells (Lanes 2, 4, 6 and 8) were fractionated by SDS-PAGE, electrotransferred onto NC filters, and immunoprobed with 1CRA4C9 (Lanes 1 and 2), 5CRA3B3 (Lanes 3 and 4), 3CRA10F5 (Lanes 5 and 6), or 3CRA10G2 (Lanes 7 and 8). Each antibody was diluted 1:1000. In parallel, WCE (5
g) from COS-1 cells transfected with either mCRABPI (Lanes 5 and 7) or mCRABPII (Lanes 1 and 3) expression vectors were immunoblotted as controls. The specific signals are indicated by an arrow and the nonspecific ones by an arrowhead. (B) Nuclear (N; Lanes 3 and 6) and cytosolic (C; Lanes 2 and 5) extracts (50
g) from Caco-2 cells and WCE (5
g) from COS-1 cells transfected with either mCRABPII (Lane 1) or mCRABPI (Lane 4) expression vectors were immunoblotted with 5CRA3B3 (Lanes 1–3) or 4CRA3B1 (Lanes 4–6) diluted 1:1000. (C) Nuclear extracts (N; Lanes 3 and 6) and cytosolic extracts (C; Lanes 2 and 5) from MCF7 cells were immunoblotted with 5CRA3B3 (Lanes 1–3) and 3CRA10F5 (Lanes 4–6) as in B. No CRABPI was revealed in MCF7 cells (B, Lanes 5 and 6) even when the filter was overexposed (data not shown).

cells also displayed a weak punctate staining in the cytoplasmic compartment. Cells with a cytosolic signal always displayed nuclear labeling. In other words, among the labeled cells examined here, none excluded CRABPs from the nucleus. In contrast, when Bouin’s fixative was used (acidic pH), these antibodies revealed a strong signal in the cytoplasm only (Figures

Rabbits were also immunized with the synthetic peptides SPB63 and SPB64. Two rabbit polyclonal antisera were obtained with SPB 63 (antisera 204 and 206) and two with SPB 64 (antisera 207 and 208). As described above for the monoclonal antibodies, these antisera were tested by immunoblotting for their ability to specifically detect recombinant mCRABPI and/or mCRABPII proteins overexpressed in transfected COS-1 cells. Figure 1B shows that antisera 207 and 208 recognize mCRABPII (Figure 1B, Lanes 3 and 4) and not mCRABPI (Figure 1B, Lanes 7 and 8), whereas antiserum 206 recognizes specifically CRABPI (Figure 1B, Lanes 2 and 6). However, antiserum 204 recognized both CRABPI and CRABPII (Figure 1B, Lanes 1 and 5), as did the monoclonal antibodies 3CRA10G2 (also raised against SPB 63), suggesting again that there is a common epitope in mCRABPI and mCRABPII. Antisera 206 and 208 also recognized the human CRABPI and CRABPII proteins respectively (data not shown). By immunocytochemistry, antisera 206 and 208 recognized mouse CRABPI and CRABPII, respectively, in transfected COS cells (data not shown). As described above for the monoclonal antibodies, both CRABPs could be detected in the nuclei after PFA (pH 7.4) fixation, whereas they were excluded from this cell compartment when Bouin’s fixative was used (acidic pH) (data not shown). Note that antisera 206 and 208 gave weaker signals than the monoclonal antibodies. The polyclonal antibodies 206 and 208 were also able to specifically detect endogenous CRABPI and II proteins, respectively, in mouse embryos by immunoblotting. Figure 5A shows that CRABPs were readily detected in extracts from wild-type embryos (13.5 dpc; Figure 5A, Lanes 1 and 5), whereas no signal was

1108

Gaub, Lutz, Ghyselinck, Scheuer, Pfister, Chambon, Rochette–Egly

Figure 4 Immunofluorescence of CRABPI- and CRABPII-transfected COS-1 cells with monoclonal antibodies. COS-1 cells transfected with either mCRABPI (1, 2, 5, and 6) or mCRABPII (3, 4, 7, and 8) expression vectors were grown on Leighton tubes and fixed with either paraformaldehyde (A) or Bouin’s fixative (B) for 4 min at room temperature. The purified monoclonal antibodies 4CRA3B1 (1 and 3) and 5CRA3B3 (2 and 4), diluted at 5
g/ml, were used to reveal CRABPI and CRABPII, respectively. Panels 5–8 correspond to Hoetchst controls. Controls were performed with untransfected COS cells and with nonimmune mouse immunoglobulins. Bars  10
m.

detected in the double CRABPI//CRABPII/ mutants (Figure 5A, Lanes 2 and 6), thus corroborating their specificity. Then cytosolic and nuclear extracts from mouse embryos at different days post coitum (10.5–17.5 dpc) were immunoprobed with antiserum 208 or 206 (Figure 5B). According to our results, both CRABPs were detectable in cytosolic extracts as early

as 10.5 dpc. However, CRABPII was detectable only up to 13.5 dpc and CRABPI up to 16.5 dpc. Nevertheless, both proteins were barely detectable at 17.5 dpc. These results which are in agreement with the in situ hybridization studies (Giguère et al. 1990), suggest that CRABPs are much less necessary as differentiation progresses and that CRABPI is predominant. In-

Nuclear Localization of Cellular Retinoic Acid Binding Proteins

1109

terestingly, from 12.5 to 15.5 dpc a weak signal corresponding to CRABPI could be seen in nuclear extracts with antiserum 206 (Figures 5B and 5C, Lane 7, lower panels). These results corroborate those obtained with the monoclonal antibodies, although the observed signals were less intense. However, no CRABPII could be detected in nuclear extracts with antiserum 208 (Figures 5B and (5C, Lane 7, upper panels). Similar observations were made with P19 cells. As shown in Figure 5C, CRABPI was detectable in both cytosolic and nuclear extracts from P19 cells with antiserum 206 (Figure 5C, Lanes 2–5, lower panel), whereas CRABPII remained confined to the cytoplasm of these cells when antiserum 208 was used (Figure 5C, Lanes 2–5, upper panel). However, an increase in CRABPII was observed on RA treatment (Figure 5C; compare Lanes 2 and 4).

Figure 5 Immunoblotting of CRABPs in mouse embryos and P19 cell extracts with rabbit polyclonal antibodies. (A) Western blot analysis of CRABPs in cytosolic extracts (50
g) from wild-type (Lanes 1 and 5) and double CRABPI//CRABPII/ mutant (Lanes 2 and 6) mouse embryos (13.5 dpc). Lanes 3, 4, 7, and 8 correspond to control WCE prepared from COS-1 cells transfected with mCRABPI and mCRABPII expression vectors. Immunodetection was performed with antisera 208 (Lanes 1–4) and 206 (Lanes 5–8), diluted 1:200. (B) Cytosolic (C) and nuclear (N) extracts (50
g) from mouse embryos (10.5–17.5 dpc) were subjected to SDS-PAGE, electrotransferred, and immunoblotted with antiserum 208 or 206 (diluted 1:200). Controls correspond to WCE (5
g) from COS-1 cells transfected with either mCRABPII (upper panel) or mCRABPI (lower panel) expression vectors. (C) Nuclear (N) and cytosolic (C) extracts (50
g) from P19 cells, with (Lanes 4 and 5) or without (Lanes 2 and 3) a 24-hr treatment with RA (10 7M), and from 13.5 dpc mouse embryos (Lanes 6 and 7), were fractionated by SDS-PAGE and immunoblotted with antiserum 208 (upper panel) or 206 (lower panel) diluted 1:200. Lanes 1 and 8 show WCE (5
g) from COS-1 cells transfected with either mCRABPI or mCRABPII expression vectors as controls. The specific signals are indicated by an arrow and the nonspecific ones by an arrowhead.

Discussion In this study we have described the preparation and the characterization of mouse monoclonal and rabbit polyclonal antibodies specific for CRABPI and CRABPII proteins. These antibodies recognize their cognate protein either by immunocytochemistry or by Western blotting analysis. In addition, their mouse and/or human specificity has been determined. These results are summarized in Table 1. Interestingly, the present antibodies enabled us to demonstrate that both CRABPI and CRABPII can be detected not only in the cytoplasm but also in nuclei of transfected COS cells, mouse embryos, and various cell lines by Western blotting and immunocytochemistry. The nuclear localization of CRABPI could be seen with any of our polyclonal or monoclonal antibodies either by immunoblotting or by immunocytochemistry, thus confirming previous reports (Gustafson et al. 1996; Venepally et al. 1996). The interesting discovery from our work presented here is the evidence for a nuclear localization of CRABPII. This nuclear localization of CRABPII has been seen by immunoblotting in mouse embryos and in the various cell lines described herein. A similar nuclear localization has been also observed with our antibodies in hematopoietic cells (Delva et al., unpublished observations). Nevertheless, such a nuclear localization appears to depend on the cell type, CRABPII accumulating in the nucleus of some cells and being excluded from the nuclei of other cells, as recently reported by Gustafson et al. (1996) for CRABPI. However, it must be stressed that, by immunoblotting, endogenous CRABPII was detectable in nuclear extracts only with our monoclonal antibody 1CRA4C9 or 5CRA3B3, whereas it was not with antiserum 208. In fact, our monoclonal antibodies (1CRA4C9 or 5CRA3B3, diluted 1:1000) gave stronger signals than

1110

Gaub, Lutz, Ghyselinck, Scheuer, Pfister, Chambon, Rochette–Egly

antiserum 208 (diluted 1:200), suggesting that their titer and/or affinity for CRABPII are higher than that of our polyclonal antibodies. Therefore, high-titer and high-affinity antibodies, such as the monoclonal antibodies 1CRA4C9 and 5CRA3B3, are required for the detection of CRABPII by immunoblotting in the nuclear compartment. By immunocytochemistry, both CRABPI and CRABPII can be detected in the nuclei of transfected COS cells with our antibodies. Again, stronger signals were obtained with the monoclonals. However, according to our results, the fixation procedure appears to be crucial for the detection of nuclear CRABPs. Indeed, in the present study, PFA fixation (at neutral pH) allows the detection of both CRABPs (CRABPI and CRABPII) in nuclei, as recently reported by Gustafson et al. (1996) and Venepally et al. (1996) for CRABPI, with similar fixation conditions. Only 10% of the labeled cells displayed a weak, punctate cytosolic staining. Interestingly, a nuclear localization of endogenous CRABPII was also observed by immunofluorescence in PFA-fixed hematopoietic cells (Delva et al. unpublished observations), thus corroborating these results. However, Bouin’s fixation (acidic conditions) appears to prevent nuclear localization because CRABPs are excluded from the nuclear compartment and are evenly confined to the cytoplasm. Similar results were obtained with the acidic fixative (4% PFA containing 2% tricloroacetic acid) described by Bucco et al. (1996) and by Zheng et al. (1996) (data not shown). These observations point out the importance of the fixation procedure for detection of CRABPs in nuclei with a given antibody. Fixation type and pH may also explain the heterogeneity of the cytosolic labeling (punctate or homogeneous) observed under both fixation conditions. However, the origin and the titer of the antibodies, as well as the incubation conditions, may also affect the distribution of labelings and should be taken into account. Although the low molecular mass of CRABPs suggests that these proteins would freely enter the nucleus, it cannot be excluded that accessory factors, expressed in a cell-specific fashion, may regulate the subcellular compartmentalization of these proteins (Gustafson et al. 1996). In any event, all these results open a new area of investigation concerning the role of nuclear CRABPs in the RA signaling pathway. Acknowledgments Supported by funds from the Centre National de la Recherche Scientifique (CNRS), the Institut National de la Santé et de la Recherche Médicale (INSERM), the Hôpital Universitaire de Strasbourg (HUS), the Collège de France, The Association pour la Recherche sur le Cancer (ARC), the Fondation pour la Recherche Médicale (FRM), the Ligue Nationale contre le Cancer, the Human Frontier Science Program, and Bristol Myers Squibb.

We are grateful to V. Schultz and N. Young for their help in the preparation of the monoclonal antibodies, to G. Duval for the rabbit injections, and to P. Oberling for preparation of the synthetic peptides. We are indebted to J.L. Vonesch for confocal analysis. We also wish to thank the cell culture group for maintaining and providing cells, M. Lemeur for providing embryos, and C. Werlé and B. Boulay for art work. We thank Dr J.J. Voorhees for the gift of purified human CRABPI and CRABPII proteins overexpressed in E. coli.

Literature Cited Aström A, Petterson U, Chambon P, Voorhees JJ (1994) Retinoic acid induction of human cellular retinoic acid-binding protein II gene transcription is mediated by retinoic acid receptor-retinoid X receptor heterodimers bound to one far upstream retinoic acid-responsive element with 5-base pair spacing. J Biol Chem 269:22334–22339 Aström A, Tavakkol A, Petterson U, Cromie M, Elder JT, Voorhees JJ (1991) Molecular cloning of two human cellular retinoic acidbinding proteins (CRABP). Retinoic acid-induced expression of CRABP-II in adult human skin in vivo and in skin fibroblasts in vitro. J Biol Chem 266:17662–17666 Balling B (1991) CRABP and the teratogenic effects of retinoids. Trends Genet 7:35–36 Blomhoff R (1994) Vitamin A in Health and Disease. New York, Basel, Hong Kong, Marcel Dekker Bucco RA, Zheng WL, Wardlaw SA, Davis JT, Sierra–Riviera E, Osteen KG, Melner MH, Kakkad BP, Ong DE (1996) Regulation and localization of cellular retinol-binding protein, cellular retinoic acid binding protein (CRABP), and CRABPII in the uterus of the pseudopregnant rat. Endocrinology 137:3111–3122 Chambon P (1994) The retinoid signalling pathway: molecular and genetic analyses. Semin Cell Biol 5:115–125 Chambon P (1996) A decade of molecular biology of retinoic acid receptors. FASEB J 10:940–954 Durand B, Saunders M, Leroy P, Leid M, Chambon P (1992) Alltrans and 9-cis retinoic acid induction of CRABPII transcription is mediated by RAR-RXR heterodimers bound to DR1 and DR2 repeated motifs. Cell 71:13–85 Fogh K, Voorhees JJ, Aström A (1993) Expression, purification and binding properties of human cellular retinoic acid binding protein (CRABP) type I and type II. Arch Biochem Biophys 300:751–755 Gaub MP, Lutz Y, Ruberté E, Petkovich M, Brand N, Chambon P (1989) Antibodies specific to the retinoic acid human nuclear receptors  and . Proc Natl Acad Sci USA 86:3089–3093 Gaub MP, Rochette–Egly C, Lutz Y, Matthes H, Scheuer I and Chambon P (1992) Immunodetection of multiple species of retinoic acid receptor : evidence for phosphorylation. Exp Cell Res 201:335–346 Ghyselinck N, Dupé V, Dierich A, Messadecq N, Garnier JM, Rochette–Egly C, Chambon P, Mark M (1997) Role of the retinoic acid receptor beta (RAR) during mouse development. Int J Dev Biol 41:425–447 Giguère V, Lyn S, Yip P, Siu CH, Amin S (1990) Molecular cloning of cDNA encoding a second cellular retinoic acid-binding protein. Proc Natl Acad Sci USA 87:6233–6237 Gorry P, Lufkin T, Dierich A, Rochette–Egly C, Decimo D, Dollé P, Mark M, Durand B, Chambon P (1994) The cellular retinoic acid binding protein I is dispensable. Proc Natl Acad Sci USA 90:932–936 Gustafson AL, Dencker L, Eriksson U (1993) Non-overlapping expression of CRBP I and CRABP I during pattern formation of limbs and craniofacial structures in the early mouse embryo. Development 117:451–460 Gustafson AL, Donovan M, Annerwall E, Dencker L, Eriksson U

Nuclear Localization of Cellular Retinoic Acid Binding Proteins (1996) Nuclear import of cellular retinoic acid-binding protein type I in mouse embryonic cells. Mechanisms Dev 58:27–38 Harlow E, Lane D (1988) Antibodies, a Laboratory Manual. Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press. Kastner P, Mark M, Chambon P (1995) Nonsteroid nuclear receptors: what are genetic studies telling us about their role in real life? Cell 83:859–869 Kastner P, Mark M, Ghyselinck N, Krezel W, Dupé V, Grondona JM, Chambon P (1997) Genetic evidence that the retinoid signal is transduced by heterodimeric RXR/RAR functional units during mouse development. Development 124:313–326 Kleywegt GJ, Bergfors T, Senn H, Le Motte P, Gsell B, Shudo K, Jones TA (1994) Crystal structures of cellular retinoic acid binding proteins I and II in complex with all-trans-retinoic acid and a synthetic retinoid. Structure 2:1241–1250 Lampron C, Rochette–Egly C, Gorry P, Dollé P, Mark M, Lufkin T, Lemeur M, Chambon P (1995) Mice deficient in cellular retinoic acid binding protein II (CRABP-II) or in both CRABP-I and CRABP-II are essentially normal. Development 121:539–548 Lyn S, Giguère V (1994) Localization of CRABP-I and CRABP-II mRNA in the early mouse embryo by whole-mount in situ hybridization: implications for teratogenesis and neural development. Dev Dyn 199:280–291 Maden M (1994) Role of retinoids in embryonic development. In Blomhoff R, ed. Vitamin A in Health and Disease. New York, Marcel Dekker, 289–322 Maden M, Horton C, Graham A, Leonard L, Pizzey J, Siegenthaler G, Lumsden A, Eriksson U (1992) Domains of cellular retinoic acid binding protein I (CRABP I) expression in the hindbrain and neural crest of the mouse embryo. Mechanisms Dev 37:13–23 Mangelsdorf DJ, Evans RM (1995) The RXR heterodimers and orphan receptors. Cell 83:841–850 Napoli JL (1994) Retinoic acid homeostasis. Perspective roles of -carotene, retinol, CRBP and CRABP. In Blomhoff R, ed. Vitamin A in Health and Disease. New York, Marcel Dekker, 135–188 Ong D E, Newcomer ME, Chytil F (1994) Cellular retinoid-binding

1111

proteins. In Sporn MB, Roberts AB, Goodman DS, eds. The Retinoids: Biology, Chemistry and Medicine. 2nd ed. New York, Raven Press, 288–317 Pinto M, Robine–Leon S, Appay MD, Kedinger M, Triadou N, Dussaulx E, Lacroix B, Simon–Assmann P, Haffen K, Fogh J, Zweibaum A (1983) Enterocyte-like differentiation and polarization of the human colon carcinoma cell line Caco-2 in culture. Biol Cell 47:323–330 Rochette-Egly C, Lutz Y, Saunders M, Scheuer I, Gaub MP, Chambon P (1991) Retinoic acid receptor : specific immunodetection and phosphorylation. J Cell Biol 115:535–545 Ruberte E, Friederich V, Morris–Kay G, Chambon P (1992) Differential distribution patterns of CRABP I and CRABP II transcripts during mouse embryogenesis. Development 115:973–987 Sporn MB, Roberts AB, Goodman DS (1994) The Retinoids: Biology, Chemistry, and Medicine. 2nd ed. New York, Raven Press Stoner CM, Gudas LJ (1989) Mouse cellular retinoic acid binding protein: cloning, complementary DNA sequence, and messenger RNA expression during the retinoic acid-induced differentiation of F9 wild type and RA-3-10 mutant teratocarcinoma cells. Cancer Res 49:1497–1504 Takase S, Ong DE, Chytil F (1986) Transfer of retinoic acid from its complex with cellular retinoic acid-binding protein to the nucleus. Arch Biochem Biophys 247:328–334 Venepally P, Reddy LG, Sani BP (1996) Analysis of the effects of CRABP I expression on the RA-induced transcription mediated by retinoid receptors. Biochemistry 35:9974–9982 Widschwendter M, Daxenblicher G, Dapunt O, Marth C (1995) Effect of retinoic acid and gamma-interferon on expression of retinoic acid receptor and cellular retinoic acid-binding protein in breast cancer cells. Cancer Res 55:2135–2139 Zheng WL, Bucco RA, Schmitt C, Wardlaw SA, Ong DE (1996) Localization of cellular retinoic acid-binding protein (CRABP) II and CRABP in developing rat testis. Endocrinology 137:5028– 5035