- Email: [email protected]
p198 protein with ErbB-3
Gene 229 (1999) 215–221
Interaction of the p23/p198 protein with ErbB-3 Joo-Yeon Yoo a,b,1, Anne W. Hamburger a,b,c,* a Molecular and Cellular Biology Program, USA b Greenebaum Cancer Center, University of Maryland at Baltimore, 655 W. Baltimore St. Baltimore, MD 21201, USA c Department of Pathology, University of Maryland, Baltimore, MD 21201, USA Received 15 July 1998; received in revised form 25 November 1998; accepted 30 November 1998; Received by I. Verma
Abstract The processes by which the kinase inactive receptor ErbB-3 transmits the signals of its ligand, heregulin (HRG), are incompletely understood. We used a yeast two-hybrid system to identify ErbB-3 interacting proteins that may participate in HRG signal transduction. We found that the protein p23, the human homolog of the mouse transplantation antigen P198, interacted with the cytoplasmic domain of ErbB-3 in the yeast two-hybrid system. P23 bound the 26-amino-acid juxtamembrane domain of ErbB-3 in vitro. The N-terminal end of p23 contained the ErbB-3 interacting region. P23 also bound to ErbB-3 in a human breast cell line. Two p23 mRNA transcripts were detected in normal human epithelial tissues including those of the heart, placenta, lung, brain, kidney, pancreas, skeletal muscle, and liver. These same transcripts were also detected in ErbB-3 overexpressing human tumor cell lines derived from breast and lung carcinomas, and a sarcoma. Transfection of p23 resulted in suppression of colony formation of the ErbB-3 overexpressing human breast cancer cell line, AU565, a decreased rate of cell growth, and induction of differentiation. The interaction of ErbB3 and p23 may play a role in regulation of proliferation of ErbB-3 expressing cells. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Breast cancer; ErbB-3; p23
1. Introduction C-erbB-3, a member of the epidermal growth factor receptor ( EGFR) or ErbB receptor family, is a 180-kDa transmembrane glycoprotein with a ligand-binding extracellular domain and an inactive cytoplasmic tyrosine kinase domain ( Kraus et al., 1989; Guy et al., 1994). Heregulin (HRG) treatment induces ErbB-3 tyrosine phosphorylation and recruitment of SH2-domain-containing proteins, such as Shc or phosphatidylinositol 3-kinase (PI 3-kinase), to the phosphorylated ErbB-3 receptor ( Fedi et al., 1994; Kim et al., 1994; Gamett et al., 1995). A complex program of cellular proliferation and differentiation then ensues * Corresponding author. Tel: +1 410-328-3911; Fax: +1 410-328-6559; e-mail: [email protected] 1 Present address: Department of Molecular Biology and Genetics, Johns Hopkins University School of Medicine, 725 N. Wolfe St. Baltimore, MD 21205, USA. Abbreviations: CTL, cytolytic T lymphocytes; EGFR, epidermal growth factor receptor; HRG, heregulin; LacZ, b-galactosidase; PI 3-kinase, phosphatidylinositol 3-kinase.
(Bacus et al., 1992; Marte et al., 1995; Jones et al., 1996). The molecular mechanisms of the activation of the signaling pathways of ErbB-3 are currently not clear. Increasing data suggest the importance of complexes of regulatory proteins with receptor tyrosine kinases before ligand stimulation ( Fazioli et al., 1993; GalchevaGargova et al., 1996). For example, the EGFR substrate, eps8, which directly binds to the juxtamembrane domain of EGFR, lacks demonstrable tyrosine phosphorylation sites, suggesting that eps8–EGFR interaction is phosphotyrosine- and SH2 (or PTB)-independent (Castagnino et al., 1995). The zinc-finger protein ZPR1, which interacts with the unstimulated EGFR, does not contain SH2 (or PTB) domains and associates with the EGFR in the absence of EGFR tyrosine phosphorylation (Galcheva-Gargova et al., 1996). Upon EGF stimulation, the ZPR1 protein becomes tyrosine-phosphorylated, dissociates from the EGFR, and accumulates in the nucleus. These studies suggest that interactions of receptor and substrates prior to growthfactor stimulation are important in the regulation of ligand-mediated signal transduction pathways. To better understand HRG-mediated signal transduc-
0378-1119/99/$ – see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S0 3 7 8 -1 1 1 9 ( 9 8 ) 0 0 60 4 - 0
216
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
tion pathways initiated by the kinase-inactive ErbB-3 receptor, we performed a yeast two-hybrid screen using the non-phosphorylated cytoplasmic domain of ErbB-3, and a human fetal brain cDNA library ( Fields and Song, 1989) to identify ErbB-3 interacting proteins. In this paper, we report that a previously cloned human 23-kDa protein with an unidentified function (Price et al., 1992) interacts with ErbB-3. This protein shares more than 88% nucleotide sequence identity with the mouse protein p198 transplantation antigen, which is recognized by clonal cytolytic T lymphocytes (CTL) (Sibille et al., 1990). In this study, we demonstrate that P23 interacted with the cytoplasmic domain of ErbB-3 both in the yeast two-hybrid system, in vitro, and in a human breast cell line. Transfection of p23 inhibited growth of human breast cancer cells, suggesting that p23 might be involved in the regulation of cellular proliferation through its interaction with ErbB-3.
2. Methods and materials 2.1. Yeast two-hybrid screening The cytoplasmic domain of rat ErbB-3 (amino acids 665–1339; nucleotides 2131–4155) (kindly provided by Dr John Koland, University of Iowa) (Hellyer et al., 1995) was directionally subcloned into the EcoRI and BamHI unique restriction endonuclease sites of the LexA DNA-binding domain vector pEG202 (pEG202/ErbB3) (provided by Dr Roger Brent, Harvard University). A human fetal brain derived cDNA library cloned into the transcription activator domain vector pJG4-5 (TRP1, ampr) (provided by Dr Brent) was cotransformed into the yeast strain EGY48 (ura3, his3, trp1, LexAop-leu2). ErbB-3 interacting clones were selected based on the acquisition of prototrophy for leucine and b-galactosidase (LacZ ) activity (Fields and Song, 1989). 2.2. Immunoprecipitation and immunoblotting Yeast transformed with pEG202/ErbB-3 (LexA DB-ErbB-3) were cultured overnight in Glu/SD-His-, Trp-, Ura-liquid media, and lysates were prepared in Breaking Buffer [100 mM Tris–HCl (pH 8.0), 100 mM NaCl, 1 mM DTT, 5% glycerol, aprotinin (1 mg/ml ), PMSF (1 mg/ml )] by vortexing in the presence of glass beads. For each immunoprecipitation, 200 mg of lysates were incubated with antibody against the cytoplasmic domain of ErbB-3 (C-17, Santa Cruz Laboratories, Santa Cruz, CA) and Protein A/G Agarose (Oncogene Science, Farmingdale, NY ) overnight at 4°C. Immunoprecipitates were analyzed on 7.5% SDS gels, and transferred on to Immobilon-P membranes (Millipore, Bedford, MA). Western blotting analysis was performed using antibodies against ErbB-3 or phos-
photyrosine (PY-20, Santa Cruz) and an Enhanced Chemiluminescence Kit (Amersham, Arlington Heights, IL). 2.3. Sequence analysis Plasmids for sequencing were prepared using Qiagen columns as described by the manufacturer. Sequencing was performing using suitable fluorescent labeled primers and an automated ABI 373 sequencer (Applied Biosystems, Foster City, CA). 2.4. Northern blotting Total RNA was isolated from the ErbB-3 expressing human mammary carcinoma cell lines AU565, EP62.1, MDA-MB231, and MDA-MB468, a human lung epithelial cell line E6, and a sarcoma cell line SK-LMS-1. Total RNA (20 mg) was analyzed on 1.2% formaldehyde agarose gels. After transfer of RNA onto nitrocellulose membranes, the p23 cDNA isolated in the two-hybrid system was radiolabelled by random priming and hybridized to blots at 65°C overnight. A human multiple tissue Northern blot (Clontech, Palo Alto, CA) was also hybridized with the radiolabelled P23 cDNA probes as recommended by the manufacturer. 2.5. Construction of deletion mutants and binding assays A construct encoding GST-fused wt ErbB-3 was prepared by subcloning ErbB-3 (aa 665–1339) from pEG202/ErbB-3 into the EcoRI and Sal1 restriction sites of pGEX4T-1 (Pharmacia, Piscataway, NJ ). Unique restriction sites (NcoI, SacII, BglII, XhoI, NdeI ) in the pGEX-4T-1/ErbB-3 were used to create deletion mutants of ErbB-3 (aa 665–1241, 665–1120, 665–931, 665–822, 665–756, respectively). Further deletion constructs of ErbB-3 (aa 665–732, 665–711, 665–690) were created by PCR using the appropriate restriction enzyme adaptors. ErbB-3 (aa 756–1339) was generated by bluntend ligation after EcoRI and NdeI restriction-enzyme digestion. The sequences of these constructs were confirmed by automated DNA sequencing. A deletion construct of P23 (amino acids 1–105) was created by bluntend ligation using the unique restriction-enzyme site in pcDNA3.1/P23, BsmBI. GST-full length or truncated ErbB3 fusion proteins were expressed in the BL21 bacterial strain, purified on glutathione sepharose beads, and examined by Coomassie Blue staining of SDS–polyacrylamide gels. For in-vitro binding experiments, 35S-labeled P23 proteins were obtained by in-vitro transcription/translation using a TNT/T7-coupled reticulocyte lysate system (Promega, Madison, WI ). Equal amounts of GST or GST-ErbB3 fusion proteins were incubated with full-length or truncated P23 at 4°C for 2 h in NP-40 buffer (50 mM Tris–HCl, pH 8.0, 120 mM
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
NaCl, 0.5% NP-40) and analyzed by SDS–PAGE and autoradiography. 2.6. Construction of P23 expression vectors and in-vivo binding experiments The P23 expression plasmids were generated by inserting a P23 PCR product inframe into the BamH1–EcoR1 sites of the mammalian expression plasmid pcDNA3.1/HisA (Invitrogen). The coding region of P23 (amino acids 1–203; 0.6 kb) was PCR-amplified using sense (5∞CGCGGATCCATGGCGGAGGTGCAGGTCCT-3∞) and antisense (5∞-CGCGAATTCGCAGGCAACGCATGAGGAAT-3∞) primers from pEG202/P23. For construction of myc tagged proteins, the antisense primer CCGGAATTCTCACAGATCCTCTTCTGAGATGAGTTTTTGTTCGACCAGGAGTCCG was used. The sequence of all the inserts was confirmed by automated DNA sequencing. To determine binding of p23 to ErbB-3 and EGFR, MCF10A cells (American Type Culture Collection, Rockville, MD) were transfected with 10 mg of the expression vector encoding a 3∞ myc-tagged p23. Cell lysates were harvested 48 h later in lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 1 mg/ml of leupeptin, 1 mM PMSF ). Lysates were immunoprecipitated with an antiErbB-3 (C-17) or an anti-EGFR antibody (Ab 528, Oncogene Science) overnight, followed by incubation with Protein A/G agarose for 1 h. Immunoprecipitated proteins were analyzed as described above. Filters were cut horizontally at the 81-kDa marker prior to immunoblotting. Filters were immunoblotted with an antimyc antibody (Invitrogen, Carlsbad, CA), the C-17 anti-ErbB-3 antibody, or an anti-EGFR antibody (sc003, Santa Cruz Laboratories). 2.7. Stable transfection, colony and cell proliferation assays AU565 cells were transfected with 10 mg of pcDNA3.1/HisA, or pcDNA3.1/P23 using lipofectin reagent (Gibco BRL, Bethesda, MD) and selected in G418 (1000 mg/ml ) for 4 weeks. Individual colonies were pooled to create mass transfectants. For colony assays, AU565 (1×105 cells) were seeded in 60-mm dishes, and 2 or 5 mg of DNA were stably transfected using lipofectin. The number of colonies was determined after 4 weeks of G418 selection (1000 mg/ml ). For proliferation assays, the mass-transfected cells (5×103) were seeded in triplicate in individual wells of 12 well tissue culture plates (Corning, Corning, NY ) in complete media and cultured for 12 days. Cells were trypsinized at the times indicated, stained with Trypan Blue (0.4%) and counted with a hemacytometer. For Oil Red O assays, AU565 cells (5×103), were
217
seeded in Lab-Tek 8 chamber slides (Nunc, Naperville, IL) and incubated with or without HRG (10 ng/ml ) for 6 days. Cells were fixed in 10% phosphate-buffered formalin solution (Sigma, St. Louis, MO) and stained with Oil Red O working solution as previously described (Bacus et al., 1992). 2.8. Statistical analysis The results were analyzed using a two-tailed Student’s t-test. Significance was established at p≤0.05.
3. Results 3.1. Yeast two-hybrid screening The cytoplasmic domain of ErbB3 was subcloned into the pEG202 LexA DNA-binding domain vector, and the expression of a LexA-ErbB-3 fusion protein was determined in lysates of pEG202/ErbB-3 transformants by immunoprecipitation and Western blotting with an antibody against ErbB-3. Tyrosine phosphorylation blots of ErbB-3 immunoprecipitates from pEG202/ ErbB-3 transformants revealed that ErbB-3 was not tyrosine-phosphorylated in yeast (Fig. 1). This screen was therefore biased towards the identification of proteins that bind to ErbB-3 in the absence of tyrosine phosphorylation. A human fetal brain cDNA library was screened for proteins interacting with the cytoplasmic domain of ErbB-3. A total of 1.4×106 yeast cells were transformed with the cDNA library, the bait pEG202/ErbB-3, and the reporter plasmid pSH18-34. Eighty-six yeast clones were selected, based on growth in the absence of leucine. Among 86 Leu+ clones, 22 were selected after the first 8 days of incubation and tested for LacZ gene expression in galactose or glucose containing media. b-galactosidase assays indicated that 13 out of 22 yeast clones interacted with ErbB-3 in this assay. Of the 13 clones sequenced, two identical clones, with 0.8-kb inserts, were isolated. Automated DNA sequence analysis of these inserts revealed a 100% nucleotide identity with the previously isolated human p23 protein (GenBank No. X56932). The 0.8-kb p23 cDNA isolated from the yeast twohybrid screen contained both translation initiation and termination codons. This cDNA encodes a protein of 203 amino acid residues with a calculated molecular mass of 23 kDa. Database comparisons of this clone showed a 100% amino acid identity to the previously isolated human highly basic p23 protein (Price et al., 1992). The predicted amino acid sequence also showed an 88% identity with a mouse tumor antigen p198 recognized by clonal cytolytic T lymphocytes (CTL) (Sibille et al., 1990). Amino acid sequence analysis revealed one putative casein kinase II phosphorylation
218
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
A
Fig. 1. Expression and tyrosine phosphorylation of the LexA DB-ErbB-3 fusion protein in yeast lysates. Yeast lysates from EGY48 transformants ( lane 1) or pEG202/ErbB-3 transformants ( lanes 2 and 3) were immunoprecipitated (IP) with an antibody against ErbB-3 and immunoblotted ( WB) with either antibodies against phosphotyrosine or ErbB-3 as indicated. Molecular weight markers are shown on the left side. The arrow indicates the 100-kDa LexA DB-ErbB-3 fusion protein.
site (S/T–X–X–D/E), three putative protein kinase C sites (S/T–X–R/K ), and one N-glycosylation site. In addition, at positions 2–43, p23 has six heptad repeats of a leucine zipper-like motif.
B
Fig. 2. mRNA expression of p23. (A) Tissue distribution of p23 mRNA expression. RNA from the indicated human tissues was probed with radiolabeled p23 cDNA. The arrow indicates the 1.3-kb molecular weight marker. (B) Northern blot analysis of p23 expression in human cell lines. RNA from AU565 ( lane 1), MDA-MB231 ( lane 2), and MDA-MB468 ( lane 3), breast carcinoma cell lines; EP62.1 ( lane 4), an ErbB-2 transfected human mammary epithelial cell line; E6 ( lane 5), SV-40 T antigen transformed ErbB-2 overexpressing lung epithelial cell line; SK ( lane 6) the sarcoma cell line SK-LMS-1; were probed with radiolabeled p23 cDNA.
3.2. Distribution of p23 mRNA
3.3. Interactions of ErbB-3 and p23
The distribution of p23 mRNA in various human tissues was determined by Northern blot analysis (Fig. 2A). Two transcripts of approximately 0.8 and 1.2 kb were observed in all tissues examined, including heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas. The 0.8-kb P23 transcript was more highly expressed than the 1.2-kb transcript in every tissue examined. Both the 0.8- and 1.2-kb mRNA transcripts were also detected in various ErbB-3 expressing carcinoma cell lines by Northern blot analysis ( Fig. 2B). The cell lines examined included the human breast carcinoma cell lines, AU565, MDA-MB-231, and MDAMB468, a human mammary epithelial cell line EP62.1, a lung carcinoma cell line E6, and a sarcoma cell line SK-LMS-1. Ethidium bromide staining of gels prior to transfer revealed that equal amounts of RNA had been loaded (data not shown).
We next investigated the interaction of p23 with ErbB-3 in vitro. Purified GST or a GST-ErbB-3 (aa 665–1339) fusion protein was incubated with in-vitrotranslated 35S-labeled P23 ( Fig. 3). GST-ErbB3, but not GST alone, bound in-vitro-translated p23. To identify the region of ErbB-3 required for binding to p23, deletion constructs of GST-ErbB-3 were prepared and examined for binding with in-vitro-translated p23 ( Fig. 3). GST-ErbB3 , which contains the first 26 665–690 amino acids of the cytoplasmic domain of ErbB-3, retained the ability to bind p23. GST-ErbB3 , 756–1339 which contained the full cytoplasmic domain of ErbB3, except aa 665–756, failed to bind to p23. These data indicate that the 26 amino acids of the juxtamembrane domain of ErbB-3 (aa 665–690) were required for binding. To identify the region of p23 that is required for
219
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221 A
B
Fig. 3. In-vitro interactions between p23 and ErbB-3. Equal amounts of in-vitro-translated 35S radiolabelled p23 were incubated with GST alone ( lanes 2 and 10), or GST-ErbB-3 fusion proteins containing amino acids 665–1339 ( lane 3), 665–1241 ( lane 4), 665–1120 ( lane 5), 665–931 ( lane 6), 665–822 ( lane 7), 665–756 ( lanes 8 and 11), 665–732 ( lane 12), 665–711 ( lane 13, 665–690 ( lane 14) or 756–1339 ( lane 15). Lanes 1 and 9, IVT products alone (20% of the input for the binding assays). Bound proteins were collected on GSH-agarose beads and proteins resolved by 12% SDS–PAGE and fluorography. The amounts of GST or GST-ErbB-3 fusion proteins were determined by SDS–PAGE and Coomassie Blue staining.
binding to ErbB-3, a deletion construct of p23 was prepared. In-vitro-translated full-length or truncated p23 were incubated with GST-ErbB3 (Fig. 4) or 665–690 GST-alone. The P23 truncated protein bound the 1–105 26-amino-acid juxtamembrane binding domain of ErbB-3, suggesting that the 105-amino-acid N-terminal end of p23 contains the binding site for ErbB3. GST alone did not bind either full-length or truncated p23 (data not shown). The interaction between p23 and ErbB-3 was further investigated in vivo in the MCF-10A mammary cell line transiently transfected with an expression vector coding for a myc tagged p23 protein. Protein immunoblot analysis with an antibody against the myc epitope demonstrated the presence of an approximately 25-kDa myc-reactive protein in ErbB-3 immunoprecipitates (Fig. 5). To determine the ability of other endogenously expressed erbB receptors to bind p23, these lysates of MCF-10A cells transiently transfected with p23 tagged myc were also immunoprecipitated with antibody to the EGFR. Myc-tagged p23 did not coimmunoprecipitate with the EGFR ( Fig. 5B).
Fig. 4. Mapping of the p23 protein binding domain for interaction with ErbB-3. Equal amounts of in-vitro-translated 35S-radiolabelled full-length ( lanes 1) or truncated p23 proteins ( lane 2=p23 ) were 1–105 incubated with GST-ErbB-3 , as described in Section 2. Bound 665–690 proteins were collected on GSH-agarose beads and analyzed by 15% SDS–PAGE and fluorography. Full-length or deleted p23 products are marked by arrows.
Fig. 5. Coimmunoprecipitation of p23 and ErbB-3 in MCF-10A cells. (A) Lysates of MCF-10A cells, transiently transfected with an expression vector encoding a myc tagged p23 protein, were immunoprecipitated (IP) with an anti-ErbB-3 rabbit antibody ( lane 1) or preimmune rabbit IgG ( lane 2). Cell lysates were resolved by SDS–PAGE and analyzed by immunoblotting (IB) with either a murine anti-myc antibody (top) or a rabbit anti-ErbB-3 antibody (bottom). (B) Aliquots of the MCF10A cell lysates used above were immunoprecipitated with a murine anti-EGFR antibody or a preimmune murine IgG, as indicated. Cell lysates were resolved by SDS–PAGE and analyzed by immunoblotting (IB) with either a murine anti-myc antibody (top) or a rabbit anti-EGFR antibody (bottom). The large arrow in the upper panel of the blot indicates the position of the murine IgG band.
3.4. Effect of p23 expression on cell growth and differentiation We investigated the effect of p23 overexpression on cell growth. AU565 human breast carcinoma cells, which express p23 mRNA and relatively high levels of ErbB-3 ( Yoo and Hamburger, 1998), were transfected with 2 or 5 mg of pcDNA3.1 alone, or pcDNA3.1/P23, and selected for growth in G418. The numbers of colonies of stable transfectants were counted after 4 weeks. The growth of P23 transfectants was significantly ( p≤0.05) reduced 65 or 92% when 2 or 5 mg of the p23 expression vector, respectively, were transfected as compared to vector controls (Fig. 6). To determine whether the decrease in the number of colonies obtained after G418 selection was due to changes in the growth rate of transfected cells, we obtained clones of AU565 cells stably transfected with p23 cDNA. The growth rate of the transfected clones was compared to that of clones transfected with the empty vector. The growth of the p23 transfectants was
220
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
resulted in an increase in the number of Oil Red O staining cells to 57±1% after 6 days of treatment. In contrast, 59±2% of p23 transfectants were Oil Red O-positive in the absence of HRG. The number of Oil Red O-positive p23 transfectants increased to 73±4% after HRG incubation.
4. Discussion
Fig. 6. P23 inhibition of colony growth. AU565 cells were transfected with 2 or 5 mg of p23 pcDNA3.1 or pcDNA3.1 alone. The number of colonies surviving 4 weeks of G418 selection (1000 mg/ml ) were compared to the number of surviving colonies transfected with vector alone. Each point represents the mean ±SE of five plates, representative of two independent experiments.
significantly decreased as compared to that of the vector control ( Fig. 7). HRG, the ligand for ErbB-3, induces differentiation of AU565 cells (Bacus et al., 1992). We therefore investigated whether the growth inhibitory effects of p23 overexpression were due to differentiation, using P23 stable transfectants of AU565 cells. Oil Red O staining, which detects the presence of neutral lipid droplets and is a marker of differentiation of mammary epithelial cells (Bacus et al., 1992), indicated that 20±9% of the unstimulated vector control transfectants stained positive in the absence of HRG. HRG (10 ng/ml ) treatment
Fig. 7. Transfection of p23 decreases the growth rate of AU565 cells. Equal numbers of vector control transfected AU565 cells or p23 stable mass transfectants were plated at day 0. At the indicated time points, the viable cell number was determined as described in Section 2. Each point represents the mean of three wells±SE.
The mechanisms by which the kinase inactive ErbB-3 receptor transmits signals is currently unknown. Here, we report that a previously identified human basic protein p23 interacts with the ErbB-3 receptor in a yeast two-hybrid assay, in vitro, and in vivo in human breast epithelial cells. Transfection of p23 into ErbB-3 overexpressing human breast cancer cells led to inhibition of colony formation. P23 is the human homolog of the murine tumor antigen p198. P198 was identified as a 23.5-kDa protein whose gene contained a point mutation that resulted in the substitution of a threonine for an alanine at position 154. The deduced amino acid sequence of human and bovine p23 and mouse P198 exhibits 94% identity. The p198 protein is present on the cell surface of p815 tumor cells. This amino acid substitution made the protein antigenic and led to tumor rejection by cytolytic T lymphocytes (Sibille et al., 1990). From a physiological standpoint, the p23 protein appears to play an important role in cell function based on the fact that a single point mutation in the mouse gene resulted in a protein capable of eliciting a cytotoxic T lymphocyte response. Human p23 protein also shows an 89% identity with the rat L13A ribosomal protein. Whether p23 also functions as a ribosomal protein in AU565 cells is not known. However, our data strongly suggest that the association of ErbB-3 and p23 in vivo, as detected by coimmunoprecipitation assays, reflects biological interactions, rather than non-specific coprecipitation of ribosomes with nascent transmembrane proteins. First, only a small defined region of ErbB-3 (the first 26 amino acids of the juxtamembrane domain) was needed to bind p23 in vitro. Larger ErbB-3-GST constructs (aa 756–1339) were unable to bind p23 in vitro. Second, p23 was not present in immunoprecipitates of EGFR of MCF-10A cells. EGFR and ErbB-3 share a 40% amino acid identity within the juxtamembrane domain that binds p23 ( Kraus et al., 1989). If binding were purely non-specific, ErbB-3 would be expected to be found in these EGFR immunoprecipitates. The biological relevance of ErbB-3 interactions with p23 and the mechanisms of p23 dissociation from ErbB-3 remain to be elucidated. The p23 protein has putative protein kinase C and casein kinase II phosphorylation sites that might be important in its regulation. For example, after binding to ErbB-3, HRG activates
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
PKC in human breast cancer cells (Bacus et al., 1993). It is possible that HRG-ErbB-3 binding results in PKCmediated phosphorylation of p23 and subsequent activation. In addition, at its NH terminus at positions 2–43, 2 p23 has six heptad repeats of a leucine zipper-like motif. The heptads have hydrophobic residues in the d position, only two of which are leucine, the others being isoleucine and valine. Generally, leucine zippers have an NH 2 adjacent domain containing clusters of basic amino acids that function in DNA binding. No such basic regions are found in p23. The zipper-like motifs may, however, still function in protein dimerization. The fact that the amino terminal domain of p23 still binds ErbB-3 is consistent with such a hypothesis. The role of p23 as a possible mediator of heregulin function has not yet been defined. Overexpression of p23 resulted in inhibition of AU565 colony growth, as does heregulin treatment (Bacus et al., 1993). However, we have not yet directly demonstrated that this growth inhibition is related to the heregulin-ErbB-3 signaling pathway. P23 overexpression did mimic heregulin-mediated differentiation in the absence of exogenous ligand, suggesting that the p23 protein can mediate biological events initiated by heregulin. In conclusion, we report here that p23 binds to the juxtamembrane domain of ErbB-3. Downstream p23-mediated events that occur after ErbB-3 binding must be investigated further.
Acknowledgements We wish to thank Dr Roger Brent (Harvard University, Cambridge, MA) for the plasmids and the cDNA library used for the yeast two hybrid screening. We also thank Dr John Koland ( University of Iowa, Iowa, City, IA) for the rat erbB-3 plasmid. This work was supported in part by a gift from the Mildred Mindell Cancer Foundation (to J.Y.Y.), the Susan Komen Foundation and NIH R01 CA76047 (to A.W.H.).
References Bacus, S.S., Huberman, E., Chin, D., Kiguchi, K., Simpson, S., Lippman, M., Lupu, R., 1992. A ligand for the ErbB-2 oncogene product (gp30) induces differentiation of human breast cancer cells. Cell Growth Differ. 3, 401–411.
221
Bacus, S.S., Gudkov, A.V., Zelnick, C.R., Chin, D., Stern, R., Stancovski, I., Peles, E., Ben-Baruch, N., Farbstein, H., Lupu, R., Wen, D., Sela, M., Yarden, Y., 1993. Neu Differentiation Factor (heregulin) induces expression of intercellular adhesion molecule:1: Implications for mammary tumors. Cancer Res. 54, 5251–5261. Castagnino, P., Biesova, Z., Wong, W.T., Fazioli, F., Gill, G.N., Di Fiore, P.P., 1995. Direct binding of eps8 to the juxtamembrane domain of EGFR is phosphotyrosine and SH2 independent. Oncogene 19, 723–729. Fazioli, F., Minichiello, L., Matoska, V., Castagnino, P., Miki, T., Wong, W.T., Di Fiore, P.P., 1993. Eps8, a substrate for the epidermal growth factor receptor kinase, enhances EGF-dependent mitogenic signals. EMBO J. 12, 3799–3808. Fedi, P., Pierce, J.P., Di Fiore, P.P., Kraus, M.H., 1994. Efficient coupling with phosphatidylinositol 3-kinase, but not phospholipase C-c or GTPase activating protein, distinguished erbB-3 signaling from that of other Erb/EGFR family members. Mol. Cell. Biol. 14, 492–500. Fields, S., Song, O., 1989. A novel genetic system to detect protein–protein interactios. Nature 340, 245–247. Galcheva-Gargova, Z., Konstantinov, K.N., Wu, I.-H., Klier, F.G., Barrett, T., Davis, R.J., 1996. Binding of zinc protein ZPR1 to the epidermal growth factor receptor. Science 272, 1797–1802. Gamett, D.C., Greene, T., Wagreich, A.R., Kim, H.-H., Koland, J.G., Cerione, R.A., 1995. Heregulin-stimulated signaling in rat pheochromocytoma cells. J. Biol. Chem. 270, 19022–19027. Guy, P.M., Platko, J.V., Cantley, L.C., Cerione, R.A., Carraway III, K.L., 1994. Insect cell-expressed p180erbB3 possesses an impaired tyrosine kinase activity. Proc. Natl. Acad. Sci. USA 91, 8132–8136. Hellyer, N.J., Kim, H.-H., Greaves, C.H., Sierke, S., , LKoland, J.G., 1995. Cloning of the rat ErbB3 cDNA and characterization of the recombinant protein. Gene 165, 279–284. Jones, F.E., Jerry, D.J., Guarino, B.C., Andrews, G.C., Stern, D.F., 1996. Heregulin induces in vivo proliferation and differentiation of mammary epithelium into secretory lobulo-alveoli. Cell Growth Differ. 7, 1031–1038. Kim, H., Sierke, S.I., Koland, J.G., 1994. Epidermal growth factordependent association of phosphadtidylinositol 3-kinase with the erbB-3 gene product. J. Biol. Chem 269, 24747–24755. Kraus, M.H., Issing, W., Miki, T., Popescu, N.C., Aaronson, S.A., 1989. Isolation and characterization of ERBB3, a third member of the ERBB1 epidermal growth factor receptor family: evidence for overexpression in a subset of human mammary tumors. Proc. Natl. Acad. Sci. USA 86, 9193–9197. Marte, B.M., Graus-Porta, D., Jeschke, M., Fabbro, D., Hynes, N.E., Taverna, D., 1995. NDF/heregulin activates MAP kinase and p70/ p85 S6 kinase during proliferation or differentiation of mammary epithelial cells. Oncogene 10, 167–175. Price, S.R., Nightingale, M.S., Bobak, D.A., Tsuchiya, M., Moss, J., Vaughan, M., 1992. Conservation of a 23-kDa human transplantation antigen in mammalian species. Genomics 14, 959–964. Sibille, C., Chomez, P., Wildmann, C., Van Pel, A., De Plaen, E., Maryanski, J.L., de Bergeyck, V., Boon, T., 1990. Structure of the gene of tum-transplantation antigen P198: A point mutation generates a new antigenic peptide. J. Exp. Med. 172, 35–45. Yoo, J.-Y., Hamburger, A.W., 1998. Changes in Heregulin b1 signaling after inhbition of ErbB-2 expression in a human breast carcinoma cell line. Mol. Cell. Endocrinol. 138, 163–171.
Interaction of the p23/p198 protein with ErbB-3 Joo-Yeon Yoo a,b,1, Anne W. Hamburger a,b,c,* a Molecular and Cellular Biology Program, USA b Greenebaum Cancer Center, University of Maryland at Baltimore, 655 W. Baltimore St. Baltimore, MD 21201, USA c Department of Pathology, University of Maryland, Baltimore, MD 21201, USA Received 15 July 1998; received in revised form 25 November 1998; accepted 30 November 1998; Received by I. Verma
Abstract The processes by which the kinase inactive receptor ErbB-3 transmits the signals of its ligand, heregulin (HRG), are incompletely understood. We used a yeast two-hybrid system to identify ErbB-3 interacting proteins that may participate in HRG signal transduction. We found that the protein p23, the human homolog of the mouse transplantation antigen P198, interacted with the cytoplasmic domain of ErbB-3 in the yeast two-hybrid system. P23 bound the 26-amino-acid juxtamembrane domain of ErbB-3 in vitro. The N-terminal end of p23 contained the ErbB-3 interacting region. P23 also bound to ErbB-3 in a human breast cell line. Two p23 mRNA transcripts were detected in normal human epithelial tissues including those of the heart, placenta, lung, brain, kidney, pancreas, skeletal muscle, and liver. These same transcripts were also detected in ErbB-3 overexpressing human tumor cell lines derived from breast and lung carcinomas, and a sarcoma. Transfection of p23 resulted in suppression of colony formation of the ErbB-3 overexpressing human breast cancer cell line, AU565, a decreased rate of cell growth, and induction of differentiation. The interaction of ErbB3 and p23 may play a role in regulation of proliferation of ErbB-3 expressing cells. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Breast cancer; ErbB-3; p23
1. Introduction C-erbB-3, a member of the epidermal growth factor receptor ( EGFR) or ErbB receptor family, is a 180-kDa transmembrane glycoprotein with a ligand-binding extracellular domain and an inactive cytoplasmic tyrosine kinase domain ( Kraus et al., 1989; Guy et al., 1994). Heregulin (HRG) treatment induces ErbB-3 tyrosine phosphorylation and recruitment of SH2-domain-containing proteins, such as Shc or phosphatidylinositol 3-kinase (PI 3-kinase), to the phosphorylated ErbB-3 receptor ( Fedi et al., 1994; Kim et al., 1994; Gamett et al., 1995). A complex program of cellular proliferation and differentiation then ensues * Corresponding author. Tel: +1 410-328-3911; Fax: +1 410-328-6559; e-mail: [email protected] 1 Present address: Department of Molecular Biology and Genetics, Johns Hopkins University School of Medicine, 725 N. Wolfe St. Baltimore, MD 21205, USA. Abbreviations: CTL, cytolytic T lymphocytes; EGFR, epidermal growth factor receptor; HRG, heregulin; LacZ, b-galactosidase; PI 3-kinase, phosphatidylinositol 3-kinase.
(Bacus et al., 1992; Marte et al., 1995; Jones et al., 1996). The molecular mechanisms of the activation of the signaling pathways of ErbB-3 are currently not clear. Increasing data suggest the importance of complexes of regulatory proteins with receptor tyrosine kinases before ligand stimulation ( Fazioli et al., 1993; GalchevaGargova et al., 1996). For example, the EGFR substrate, eps8, which directly binds to the juxtamembrane domain of EGFR, lacks demonstrable tyrosine phosphorylation sites, suggesting that eps8–EGFR interaction is phosphotyrosine- and SH2 (or PTB)-independent (Castagnino et al., 1995). The zinc-finger protein ZPR1, which interacts with the unstimulated EGFR, does not contain SH2 (or PTB) domains and associates with the EGFR in the absence of EGFR tyrosine phosphorylation (Galcheva-Gargova et al., 1996). Upon EGF stimulation, the ZPR1 protein becomes tyrosine-phosphorylated, dissociates from the EGFR, and accumulates in the nucleus. These studies suggest that interactions of receptor and substrates prior to growthfactor stimulation are important in the regulation of ligand-mediated signal transduction pathways. To better understand HRG-mediated signal transduc-
0378-1119/99/$ – see front matter © 1999 Elsevier Science B.V. All rights reserved. PII: S0 3 7 8 -1 1 1 9 ( 9 8 ) 0 0 60 4 - 0
216
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
tion pathways initiated by the kinase-inactive ErbB-3 receptor, we performed a yeast two-hybrid screen using the non-phosphorylated cytoplasmic domain of ErbB-3, and a human fetal brain cDNA library ( Fields and Song, 1989) to identify ErbB-3 interacting proteins. In this paper, we report that a previously cloned human 23-kDa protein with an unidentified function (Price et al., 1992) interacts with ErbB-3. This protein shares more than 88% nucleotide sequence identity with the mouse protein p198 transplantation antigen, which is recognized by clonal cytolytic T lymphocytes (CTL) (Sibille et al., 1990). In this study, we demonstrate that P23 interacted with the cytoplasmic domain of ErbB-3 both in the yeast two-hybrid system, in vitro, and in a human breast cell line. Transfection of p23 inhibited growth of human breast cancer cells, suggesting that p23 might be involved in the regulation of cellular proliferation through its interaction with ErbB-3.
2. Methods and materials 2.1. Yeast two-hybrid screening The cytoplasmic domain of rat ErbB-3 (amino acids 665–1339; nucleotides 2131–4155) (kindly provided by Dr John Koland, University of Iowa) (Hellyer et al., 1995) was directionally subcloned into the EcoRI and BamHI unique restriction endonuclease sites of the LexA DNA-binding domain vector pEG202 (pEG202/ErbB3) (provided by Dr Roger Brent, Harvard University). A human fetal brain derived cDNA library cloned into the transcription activator domain vector pJG4-5 (TRP1, ampr) (provided by Dr Brent) was cotransformed into the yeast strain EGY48 (ura3, his3, trp1, LexAop-leu2). ErbB-3 interacting clones were selected based on the acquisition of prototrophy for leucine and b-galactosidase (LacZ ) activity (Fields and Song, 1989). 2.2. Immunoprecipitation and immunoblotting Yeast transformed with pEG202/ErbB-3 (LexA DB-ErbB-3) were cultured overnight in Glu/SD-His-, Trp-, Ura-liquid media, and lysates were prepared in Breaking Buffer [100 mM Tris–HCl (pH 8.0), 100 mM NaCl, 1 mM DTT, 5% glycerol, aprotinin (1 mg/ml ), PMSF (1 mg/ml )] by vortexing in the presence of glass beads. For each immunoprecipitation, 200 mg of lysates were incubated with antibody against the cytoplasmic domain of ErbB-3 (C-17, Santa Cruz Laboratories, Santa Cruz, CA) and Protein A/G Agarose (Oncogene Science, Farmingdale, NY ) overnight at 4°C. Immunoprecipitates were analyzed on 7.5% SDS gels, and transferred on to Immobilon-P membranes (Millipore, Bedford, MA). Western blotting analysis was performed using antibodies against ErbB-3 or phos-
photyrosine (PY-20, Santa Cruz) and an Enhanced Chemiluminescence Kit (Amersham, Arlington Heights, IL). 2.3. Sequence analysis Plasmids for sequencing were prepared using Qiagen columns as described by the manufacturer. Sequencing was performing using suitable fluorescent labeled primers and an automated ABI 373 sequencer (Applied Biosystems, Foster City, CA). 2.4. Northern blotting Total RNA was isolated from the ErbB-3 expressing human mammary carcinoma cell lines AU565, EP62.1, MDA-MB231, and MDA-MB468, a human lung epithelial cell line E6, and a sarcoma cell line SK-LMS-1. Total RNA (20 mg) was analyzed on 1.2% formaldehyde agarose gels. After transfer of RNA onto nitrocellulose membranes, the p23 cDNA isolated in the two-hybrid system was radiolabelled by random priming and hybridized to blots at 65°C overnight. A human multiple tissue Northern blot (Clontech, Palo Alto, CA) was also hybridized with the radiolabelled P23 cDNA probes as recommended by the manufacturer. 2.5. Construction of deletion mutants and binding assays A construct encoding GST-fused wt ErbB-3 was prepared by subcloning ErbB-3 (aa 665–1339) from pEG202/ErbB-3 into the EcoRI and Sal1 restriction sites of pGEX4T-1 (Pharmacia, Piscataway, NJ ). Unique restriction sites (NcoI, SacII, BglII, XhoI, NdeI ) in the pGEX-4T-1/ErbB-3 were used to create deletion mutants of ErbB-3 (aa 665–1241, 665–1120, 665–931, 665–822, 665–756, respectively). Further deletion constructs of ErbB-3 (aa 665–732, 665–711, 665–690) were created by PCR using the appropriate restriction enzyme adaptors. ErbB-3 (aa 756–1339) was generated by bluntend ligation after EcoRI and NdeI restriction-enzyme digestion. The sequences of these constructs were confirmed by automated DNA sequencing. A deletion construct of P23 (amino acids 1–105) was created by bluntend ligation using the unique restriction-enzyme site in pcDNA3.1/P23, BsmBI. GST-full length or truncated ErbB3 fusion proteins were expressed in the BL21 bacterial strain, purified on glutathione sepharose beads, and examined by Coomassie Blue staining of SDS–polyacrylamide gels. For in-vitro binding experiments, 35S-labeled P23 proteins were obtained by in-vitro transcription/translation using a TNT/T7-coupled reticulocyte lysate system (Promega, Madison, WI ). Equal amounts of GST or GST-ErbB3 fusion proteins were incubated with full-length or truncated P23 at 4°C for 2 h in NP-40 buffer (50 mM Tris–HCl, pH 8.0, 120 mM
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
NaCl, 0.5% NP-40) and analyzed by SDS–PAGE and autoradiography. 2.6. Construction of P23 expression vectors and in-vivo binding experiments The P23 expression plasmids were generated by inserting a P23 PCR product inframe into the BamH1–EcoR1 sites of the mammalian expression plasmid pcDNA3.1/HisA (Invitrogen). The coding region of P23 (amino acids 1–203; 0.6 kb) was PCR-amplified using sense (5∞CGCGGATCCATGGCGGAGGTGCAGGTCCT-3∞) and antisense (5∞-CGCGAATTCGCAGGCAACGCATGAGGAAT-3∞) primers from pEG202/P23. For construction of myc tagged proteins, the antisense primer CCGGAATTCTCACAGATCCTCTTCTGAGATGAGTTTTTGTTCGACCAGGAGTCCG was used. The sequence of all the inserts was confirmed by automated DNA sequencing. To determine binding of p23 to ErbB-3 and EGFR, MCF10A cells (American Type Culture Collection, Rockville, MD) were transfected with 10 mg of the expression vector encoding a 3∞ myc-tagged p23. Cell lysates were harvested 48 h later in lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 1 mg/ml of leupeptin, 1 mM PMSF ). Lysates were immunoprecipitated with an antiErbB-3 (C-17) or an anti-EGFR antibody (Ab 528, Oncogene Science) overnight, followed by incubation with Protein A/G agarose for 1 h. Immunoprecipitated proteins were analyzed as described above. Filters were cut horizontally at the 81-kDa marker prior to immunoblotting. Filters were immunoblotted with an antimyc antibody (Invitrogen, Carlsbad, CA), the C-17 anti-ErbB-3 antibody, or an anti-EGFR antibody (sc003, Santa Cruz Laboratories). 2.7. Stable transfection, colony and cell proliferation assays AU565 cells were transfected with 10 mg of pcDNA3.1/HisA, or pcDNA3.1/P23 using lipofectin reagent (Gibco BRL, Bethesda, MD) and selected in G418 (1000 mg/ml ) for 4 weeks. Individual colonies were pooled to create mass transfectants. For colony assays, AU565 (1×105 cells) were seeded in 60-mm dishes, and 2 or 5 mg of DNA were stably transfected using lipofectin. The number of colonies was determined after 4 weeks of G418 selection (1000 mg/ml ). For proliferation assays, the mass-transfected cells (5×103) were seeded in triplicate in individual wells of 12 well tissue culture plates (Corning, Corning, NY ) in complete media and cultured for 12 days. Cells were trypsinized at the times indicated, stained with Trypan Blue (0.4%) and counted with a hemacytometer. For Oil Red O assays, AU565 cells (5×103), were
217
seeded in Lab-Tek 8 chamber slides (Nunc, Naperville, IL) and incubated with or without HRG (10 ng/ml ) for 6 days. Cells were fixed in 10% phosphate-buffered formalin solution (Sigma, St. Louis, MO) and stained with Oil Red O working solution as previously described (Bacus et al., 1992). 2.8. Statistical analysis The results were analyzed using a two-tailed Student’s t-test. Significance was established at p≤0.05.
3. Results 3.1. Yeast two-hybrid screening The cytoplasmic domain of ErbB3 was subcloned into the pEG202 LexA DNA-binding domain vector, and the expression of a LexA-ErbB-3 fusion protein was determined in lysates of pEG202/ErbB-3 transformants by immunoprecipitation and Western blotting with an antibody against ErbB-3. Tyrosine phosphorylation blots of ErbB-3 immunoprecipitates from pEG202/ ErbB-3 transformants revealed that ErbB-3 was not tyrosine-phosphorylated in yeast (Fig. 1). This screen was therefore biased towards the identification of proteins that bind to ErbB-3 in the absence of tyrosine phosphorylation. A human fetal brain cDNA library was screened for proteins interacting with the cytoplasmic domain of ErbB-3. A total of 1.4×106 yeast cells were transformed with the cDNA library, the bait pEG202/ErbB-3, and the reporter plasmid pSH18-34. Eighty-six yeast clones were selected, based on growth in the absence of leucine. Among 86 Leu+ clones, 22 were selected after the first 8 days of incubation and tested for LacZ gene expression in galactose or glucose containing media. b-galactosidase assays indicated that 13 out of 22 yeast clones interacted with ErbB-3 in this assay. Of the 13 clones sequenced, two identical clones, with 0.8-kb inserts, were isolated. Automated DNA sequence analysis of these inserts revealed a 100% nucleotide identity with the previously isolated human p23 protein (GenBank No. X56932). The 0.8-kb p23 cDNA isolated from the yeast twohybrid screen contained both translation initiation and termination codons. This cDNA encodes a protein of 203 amino acid residues with a calculated molecular mass of 23 kDa. Database comparisons of this clone showed a 100% amino acid identity to the previously isolated human highly basic p23 protein (Price et al., 1992). The predicted amino acid sequence also showed an 88% identity with a mouse tumor antigen p198 recognized by clonal cytolytic T lymphocytes (CTL) (Sibille et al., 1990). Amino acid sequence analysis revealed one putative casein kinase II phosphorylation
218
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
A
Fig. 1. Expression and tyrosine phosphorylation of the LexA DB-ErbB-3 fusion protein in yeast lysates. Yeast lysates from EGY48 transformants ( lane 1) or pEG202/ErbB-3 transformants ( lanes 2 and 3) were immunoprecipitated (IP) with an antibody against ErbB-3 and immunoblotted ( WB) with either antibodies against phosphotyrosine or ErbB-3 as indicated. Molecular weight markers are shown on the left side. The arrow indicates the 100-kDa LexA DB-ErbB-3 fusion protein.
site (S/T–X–X–D/E), three putative protein kinase C sites (S/T–X–R/K ), and one N-glycosylation site. In addition, at positions 2–43, p23 has six heptad repeats of a leucine zipper-like motif.
B
Fig. 2. mRNA expression of p23. (A) Tissue distribution of p23 mRNA expression. RNA from the indicated human tissues was probed with radiolabeled p23 cDNA. The arrow indicates the 1.3-kb molecular weight marker. (B) Northern blot analysis of p23 expression in human cell lines. RNA from AU565 ( lane 1), MDA-MB231 ( lane 2), and MDA-MB468 ( lane 3), breast carcinoma cell lines; EP62.1 ( lane 4), an ErbB-2 transfected human mammary epithelial cell line; E6 ( lane 5), SV-40 T antigen transformed ErbB-2 overexpressing lung epithelial cell line; SK ( lane 6) the sarcoma cell line SK-LMS-1; were probed with radiolabeled p23 cDNA.
3.2. Distribution of p23 mRNA
3.3. Interactions of ErbB-3 and p23
The distribution of p23 mRNA in various human tissues was determined by Northern blot analysis (Fig. 2A). Two transcripts of approximately 0.8 and 1.2 kb were observed in all tissues examined, including heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas. The 0.8-kb P23 transcript was more highly expressed than the 1.2-kb transcript in every tissue examined. Both the 0.8- and 1.2-kb mRNA transcripts were also detected in various ErbB-3 expressing carcinoma cell lines by Northern blot analysis ( Fig. 2B). The cell lines examined included the human breast carcinoma cell lines, AU565, MDA-MB-231, and MDAMB468, a human mammary epithelial cell line EP62.1, a lung carcinoma cell line E6, and a sarcoma cell line SK-LMS-1. Ethidium bromide staining of gels prior to transfer revealed that equal amounts of RNA had been loaded (data not shown).
We next investigated the interaction of p23 with ErbB-3 in vitro. Purified GST or a GST-ErbB-3 (aa 665–1339) fusion protein was incubated with in-vitrotranslated 35S-labeled P23 ( Fig. 3). GST-ErbB3, but not GST alone, bound in-vitro-translated p23. To identify the region of ErbB-3 required for binding to p23, deletion constructs of GST-ErbB-3 were prepared and examined for binding with in-vitro-translated p23 ( Fig. 3). GST-ErbB3 , which contains the first 26 665–690 amino acids of the cytoplasmic domain of ErbB-3, retained the ability to bind p23. GST-ErbB3 , 756–1339 which contained the full cytoplasmic domain of ErbB3, except aa 665–756, failed to bind to p23. These data indicate that the 26 amino acids of the juxtamembrane domain of ErbB-3 (aa 665–690) were required for binding. To identify the region of p23 that is required for
219
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221 A
B
Fig. 3. In-vitro interactions between p23 and ErbB-3. Equal amounts of in-vitro-translated 35S radiolabelled p23 were incubated with GST alone ( lanes 2 and 10), or GST-ErbB-3 fusion proteins containing amino acids 665–1339 ( lane 3), 665–1241 ( lane 4), 665–1120 ( lane 5), 665–931 ( lane 6), 665–822 ( lane 7), 665–756 ( lanes 8 and 11), 665–732 ( lane 12), 665–711 ( lane 13, 665–690 ( lane 14) or 756–1339 ( lane 15). Lanes 1 and 9, IVT products alone (20% of the input for the binding assays). Bound proteins were collected on GSH-agarose beads and proteins resolved by 12% SDS–PAGE and fluorography. The amounts of GST or GST-ErbB-3 fusion proteins were determined by SDS–PAGE and Coomassie Blue staining.
binding to ErbB-3, a deletion construct of p23 was prepared. In-vitro-translated full-length or truncated p23 were incubated with GST-ErbB3 (Fig. 4) or 665–690 GST-alone. The P23 truncated protein bound the 1–105 26-amino-acid juxtamembrane binding domain of ErbB-3, suggesting that the 105-amino-acid N-terminal end of p23 contains the binding site for ErbB3. GST alone did not bind either full-length or truncated p23 (data not shown). The interaction between p23 and ErbB-3 was further investigated in vivo in the MCF-10A mammary cell line transiently transfected with an expression vector coding for a myc tagged p23 protein. Protein immunoblot analysis with an antibody against the myc epitope demonstrated the presence of an approximately 25-kDa myc-reactive protein in ErbB-3 immunoprecipitates (Fig. 5). To determine the ability of other endogenously expressed erbB receptors to bind p23, these lysates of MCF-10A cells transiently transfected with p23 tagged myc were also immunoprecipitated with antibody to the EGFR. Myc-tagged p23 did not coimmunoprecipitate with the EGFR ( Fig. 5B).
Fig. 4. Mapping of the p23 protein binding domain for interaction with ErbB-3. Equal amounts of in-vitro-translated 35S-radiolabelled full-length ( lanes 1) or truncated p23 proteins ( lane 2=p23 ) were 1–105 incubated with GST-ErbB-3 , as described in Section 2. Bound 665–690 proteins were collected on GSH-agarose beads and analyzed by 15% SDS–PAGE and fluorography. Full-length or deleted p23 products are marked by arrows.
Fig. 5. Coimmunoprecipitation of p23 and ErbB-3 in MCF-10A cells. (A) Lysates of MCF-10A cells, transiently transfected with an expression vector encoding a myc tagged p23 protein, were immunoprecipitated (IP) with an anti-ErbB-3 rabbit antibody ( lane 1) or preimmune rabbit IgG ( lane 2). Cell lysates were resolved by SDS–PAGE and analyzed by immunoblotting (IB) with either a murine anti-myc antibody (top) or a rabbit anti-ErbB-3 antibody (bottom). (B) Aliquots of the MCF10A cell lysates used above were immunoprecipitated with a murine anti-EGFR antibody or a preimmune murine IgG, as indicated. Cell lysates were resolved by SDS–PAGE and analyzed by immunoblotting (IB) with either a murine anti-myc antibody (top) or a rabbit anti-EGFR antibody (bottom). The large arrow in the upper panel of the blot indicates the position of the murine IgG band.
3.4. Effect of p23 expression on cell growth and differentiation We investigated the effect of p23 overexpression on cell growth. AU565 human breast carcinoma cells, which express p23 mRNA and relatively high levels of ErbB-3 ( Yoo and Hamburger, 1998), were transfected with 2 or 5 mg of pcDNA3.1 alone, or pcDNA3.1/P23, and selected for growth in G418. The numbers of colonies of stable transfectants were counted after 4 weeks. The growth of P23 transfectants was significantly ( p≤0.05) reduced 65 or 92% when 2 or 5 mg of the p23 expression vector, respectively, were transfected as compared to vector controls (Fig. 6). To determine whether the decrease in the number of colonies obtained after G418 selection was due to changes in the growth rate of transfected cells, we obtained clones of AU565 cells stably transfected with p23 cDNA. The growth rate of the transfected clones was compared to that of clones transfected with the empty vector. The growth of the p23 transfectants was
220
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
resulted in an increase in the number of Oil Red O staining cells to 57±1% after 6 days of treatment. In contrast, 59±2% of p23 transfectants were Oil Red O-positive in the absence of HRG. The number of Oil Red O-positive p23 transfectants increased to 73±4% after HRG incubation.
4. Discussion
Fig. 6. P23 inhibition of colony growth. AU565 cells were transfected with 2 or 5 mg of p23 pcDNA3.1 or pcDNA3.1 alone. The number of colonies surviving 4 weeks of G418 selection (1000 mg/ml ) were compared to the number of surviving colonies transfected with vector alone. Each point represents the mean ±SE of five plates, representative of two independent experiments.
significantly decreased as compared to that of the vector control ( Fig. 7). HRG, the ligand for ErbB-3, induces differentiation of AU565 cells (Bacus et al., 1992). We therefore investigated whether the growth inhibitory effects of p23 overexpression were due to differentiation, using P23 stable transfectants of AU565 cells. Oil Red O staining, which detects the presence of neutral lipid droplets and is a marker of differentiation of mammary epithelial cells (Bacus et al., 1992), indicated that 20±9% of the unstimulated vector control transfectants stained positive in the absence of HRG. HRG (10 ng/ml ) treatment
Fig. 7. Transfection of p23 decreases the growth rate of AU565 cells. Equal numbers of vector control transfected AU565 cells or p23 stable mass transfectants were plated at day 0. At the indicated time points, the viable cell number was determined as described in Section 2. Each point represents the mean of three wells±SE.
The mechanisms by which the kinase inactive ErbB-3 receptor transmits signals is currently unknown. Here, we report that a previously identified human basic protein p23 interacts with the ErbB-3 receptor in a yeast two-hybrid assay, in vitro, and in vivo in human breast epithelial cells. Transfection of p23 into ErbB-3 overexpressing human breast cancer cells led to inhibition of colony formation. P23 is the human homolog of the murine tumor antigen p198. P198 was identified as a 23.5-kDa protein whose gene contained a point mutation that resulted in the substitution of a threonine for an alanine at position 154. The deduced amino acid sequence of human and bovine p23 and mouse P198 exhibits 94% identity. The p198 protein is present on the cell surface of p815 tumor cells. This amino acid substitution made the protein antigenic and led to tumor rejection by cytolytic T lymphocytes (Sibille et al., 1990). From a physiological standpoint, the p23 protein appears to play an important role in cell function based on the fact that a single point mutation in the mouse gene resulted in a protein capable of eliciting a cytotoxic T lymphocyte response. Human p23 protein also shows an 89% identity with the rat L13A ribosomal protein. Whether p23 also functions as a ribosomal protein in AU565 cells is not known. However, our data strongly suggest that the association of ErbB-3 and p23 in vivo, as detected by coimmunoprecipitation assays, reflects biological interactions, rather than non-specific coprecipitation of ribosomes with nascent transmembrane proteins. First, only a small defined region of ErbB-3 (the first 26 amino acids of the juxtamembrane domain) was needed to bind p23 in vitro. Larger ErbB-3-GST constructs (aa 756–1339) were unable to bind p23 in vitro. Second, p23 was not present in immunoprecipitates of EGFR of MCF-10A cells. EGFR and ErbB-3 share a 40% amino acid identity within the juxtamembrane domain that binds p23 ( Kraus et al., 1989). If binding were purely non-specific, ErbB-3 would be expected to be found in these EGFR immunoprecipitates. The biological relevance of ErbB-3 interactions with p23 and the mechanisms of p23 dissociation from ErbB-3 remain to be elucidated. The p23 protein has putative protein kinase C and casein kinase II phosphorylation sites that might be important in its regulation. For example, after binding to ErbB-3, HRG activates
J.-Y. Yoo, A.W. Hamburger / Gene 229 (1999) 215–221
PKC in human breast cancer cells (Bacus et al., 1993). It is possible that HRG-ErbB-3 binding results in PKCmediated phosphorylation of p23 and subsequent activation. In addition, at its NH terminus at positions 2–43, 2 p23 has six heptad repeats of a leucine zipper-like motif. The heptads have hydrophobic residues in the d position, only two of which are leucine, the others being isoleucine and valine. Generally, leucine zippers have an NH 2 adjacent domain containing clusters of basic amino acids that function in DNA binding. No such basic regions are found in p23. The zipper-like motifs may, however, still function in protein dimerization. The fact that the amino terminal domain of p23 still binds ErbB-3 is consistent with such a hypothesis. The role of p23 as a possible mediator of heregulin function has not yet been defined. Overexpression of p23 resulted in inhibition of AU565 colony growth, as does heregulin treatment (Bacus et al., 1993). However, we have not yet directly demonstrated that this growth inhibition is related to the heregulin-ErbB-3 signaling pathway. P23 overexpression did mimic heregulin-mediated differentiation in the absence of exogenous ligand, suggesting that the p23 protein can mediate biological events initiated by heregulin. In conclusion, we report here that p23 binds to the juxtamembrane domain of ErbB-3. Downstream p23-mediated events that occur after ErbB-3 binding must be investigated further.
Acknowledgements We wish to thank Dr Roger Brent (Harvard University, Cambridge, MA) for the plasmids and the cDNA library used for the yeast two hybrid screening. We also thank Dr John Koland ( University of Iowa, Iowa, City, IA) for the rat erbB-3 plasmid. This work was supported in part by a gift from the Mildred Mindell Cancer Foundation (to J.Y.Y.), the Susan Komen Foundation and NIH R01 CA76047 (to A.W.H.).
References Bacus, S.S., Huberman, E., Chin, D., Kiguchi, K., Simpson, S., Lippman, M., Lupu, R., 1992. A ligand for the ErbB-2 oncogene product (gp30) induces differentiation of human breast cancer cells. Cell Growth Differ. 3, 401–411.
221
Bacus, S.S., Gudkov, A.V., Zelnick, C.R., Chin, D., Stern, R., Stancovski, I., Peles, E., Ben-Baruch, N., Farbstein, H., Lupu, R., Wen, D., Sela, M., Yarden, Y., 1993. Neu Differentiation Factor (heregulin) induces expression of intercellular adhesion molecule:1: Implications for mammary tumors. Cancer Res. 54, 5251–5261. Castagnino, P., Biesova, Z., Wong, W.T., Fazioli, F., Gill, G.N., Di Fiore, P.P., 1995. Direct binding of eps8 to the juxtamembrane domain of EGFR is phosphotyrosine and SH2 independent. Oncogene 19, 723–729. Fazioli, F., Minichiello, L., Matoska, V., Castagnino, P., Miki, T., Wong, W.T., Di Fiore, P.P., 1993. Eps8, a substrate for the epidermal growth factor receptor kinase, enhances EGF-dependent mitogenic signals. EMBO J. 12, 3799–3808. Fedi, P., Pierce, J.P., Di Fiore, P.P., Kraus, M.H., 1994. Efficient coupling with phosphatidylinositol 3-kinase, but not phospholipase C-c or GTPase activating protein, distinguished erbB-3 signaling from that of other Erb/EGFR family members. Mol. Cell. Biol. 14, 492–500. Fields, S., Song, O., 1989. A novel genetic system to detect protein–protein interactios. Nature 340, 245–247. Galcheva-Gargova, Z., Konstantinov, K.N., Wu, I.-H., Klier, F.G., Barrett, T., Davis, R.J., 1996. Binding of zinc protein ZPR1 to the epidermal growth factor receptor. Science 272, 1797–1802. Gamett, D.C., Greene, T., Wagreich, A.R., Kim, H.-H., Koland, J.G., Cerione, R.A., 1995. Heregulin-stimulated signaling in rat pheochromocytoma cells. J. Biol. Chem. 270, 19022–19027. Guy, P.M., Platko, J.V., Cantley, L.C., Cerione, R.A., Carraway III, K.L., 1994. Insect cell-expressed p180erbB3 possesses an impaired tyrosine kinase activity. Proc. Natl. Acad. Sci. USA 91, 8132–8136. Hellyer, N.J., Kim, H.-H., Greaves, C.H., Sierke, S., , LKoland, J.G., 1995. Cloning of the rat ErbB3 cDNA and characterization of the recombinant protein. Gene 165, 279–284. Jones, F.E., Jerry, D.J., Guarino, B.C., Andrews, G.C., Stern, D.F., 1996. Heregulin induces in vivo proliferation and differentiation of mammary epithelium into secretory lobulo-alveoli. Cell Growth Differ. 7, 1031–1038. Kim, H., Sierke, S.I., Koland, J.G., 1994. Epidermal growth factordependent association of phosphadtidylinositol 3-kinase with the erbB-3 gene product. J. Biol. Chem 269, 24747–24755. Kraus, M.H., Issing, W., Miki, T., Popescu, N.C., Aaronson, S.A., 1989. Isolation and characterization of ERBB3, a third member of the ERBB1 epidermal growth factor receptor family: evidence for overexpression in a subset of human mammary tumors. Proc. Natl. Acad. Sci. USA 86, 9193–9197. Marte, B.M., Graus-Porta, D., Jeschke, M., Fabbro, D., Hynes, N.E., Taverna, D., 1995. NDF/heregulin activates MAP kinase and p70/ p85 S6 kinase during proliferation or differentiation of mammary epithelial cells. Oncogene 10, 167–175. Price, S.R., Nightingale, M.S., Bobak, D.A., Tsuchiya, M., Moss, J., Vaughan, M., 1992. Conservation of a 23-kDa human transplantation antigen in mammalian species. Genomics 14, 959–964. Sibille, C., Chomez, P., Wildmann, C., Van Pel, A., De Plaen, E., Maryanski, J.L., de Bergeyck, V., Boon, T., 1990. Structure of the gene of tum-transplantation antigen P198: A point mutation generates a new antigenic peptide. J. Exp. Med. 172, 35–45. Yoo, J.-Y., Hamburger, A.W., 1998. Changes in Heregulin b1 signaling after inhbition of ErbB-2 expression in a human breast carcinoma cell line. Mol. Cell. Endocrinol. 138, 163–171.