Hideomi WatanabeS, Pnina CarmiP, Victor Hogan, Tirza Raz, Steve Silletti, Ivan R. Nabin, and. Avraham RazII. From the Cancer Metastasis Program, Michigan ...
VOl. 266, No . 20, Issue of July 15, PP. 13442-13448, 1991 Printed in U.S.A.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 62 1991 by The American Society for Biochemistry and Molecular Biology, Inc.
Purification of Human Tumor Cell Autocrine Motility Factor and Molecular Cloningof Its Receptor* (Received for publication, January 7, 1991)
Hideomi WatanabeS,Pnina CarmiP, Victor Hogan, Tirza Raz, Steve Silletti,Ivan R. Nabin, and Avraham RazII From the Cancer Metastasis Program, Michigan Cancer Foundation, Detroit, Michigan 48201
Tumor autocrine motility factor (AMF) has been detected in and purified from serum-free conditioned medium of humanHT-1080 fibrosarcoma cells. Under nonreducing conditions, AMF migrates in sodium dodecyl sulfate-polyacrylamide gel electrophoresis as a single band of 55 kDa but under reducing conditions as a band of 6 4 kDa. Two-dimensional polyacrylamide gel electrophoresis of the purified AMF resolved two groups of polypeptides with isoelectric points of 6.1 and 6.2 (majors), 6.35 and 6.4 (minors). Purified AMF stimulated HT-1080 cell migration in a dose-dependent fashion. The motility stimulation of the fibrosarcoma cells withAMF is associated with the phosphorylation of the AMF receptor, a 78-kDa cell surface glycoprotein (gp78),suggesting protein kinase participation in migratory signal transduction. The gene encoding gp78 was cloned from an HT- 1080 fibrosarcoma complementary DNA library. The deduced sequence encodes a polypeptide of 323 amino acids. The nucleotide and predicted amino acid sequence of the gp78 revealssignificant homology with the human suppressor/oncogenep53 protein.
invasiveness of epithelial tumor cells (2-4). Migration stimulating factor is produced by human fetal and tumor-derived fibroblasts and enhances cell migration into collagen gel matrices (5, 6). AMF’ is produced by and enhances the motility of the producing cells (7, 8). These factors and others may represent a family of cytokines whose regulated expression induces cell motility in normal situations such as wound healing (scatter factor) andembryogenesis (migration stimulatingfactor)and whose constitutive autocrine expression (AMF) may confer metastatic capabilities on neoplastic cells. AMF-stimulated tumor motility is coupled to pseudopodial protrusion (9), with an increased rate of phospholipid methylation ( 7 7 , andit was suggested that AMF initiates cell motility by interacting with a cell surface receptor that is coupled to a pertussis toxin-sensitive G protein (10). We have demonstrated previously that a monoclonal antibody (mAb) against a 78-kDa B16-Fl melanoma cell surface glycoprotein (gp78) and B16-F1 melanoma serum-free medium containing AMF stimulate cell motility through a GTP-binding protein pathway and designated gp78 as thereceptor for B16-Fl AMF (ll).’AMF was purified previously from the human A2058 (7) and from the mouse B16-F1 melanoma serum-free conditioned media.* Inbothpreparations AMF migrates under nonreducing conditions in SDS-PAGE as asingle band of 55 Kinesis of eukaryotic cells plays afundamental role in kDa whereas under reducing conditions it migrates as asingle diverse biological phenomena such as spermatogenesis, mor- polypeptide of 64 kDa. Purified AMF stimulates B16-F1 cell phogenesis, wound healing, chemotaxis, and metastasis (1). migration in a dose-dependent fashion and binds directly in However, despite many studies the mechanisms underlying a protein-protein binding assay to the gp78-AMF receptor.’ cell movement and motility stimulation are not fully underIn thispaper we demonstrate that the55-kDa (nonreduced) stood. Recently a group of cytokines has been described which and 64-kDa (reduced) forms of AMF are notmelanoma unique are distinguished by their ability to stimulate motility in a as asimilar AMF moleculewas purified from the human HTvariety of eukaryotic cells. The specificity of their expression 1080 fibrosarcoma serum-free culture conditioned media, and and action may be a mechanism by which cellular motility is the present data show that the motility stimulation of HTcontrolled and regulated within the organism. A fibroblast- 1080 cells with AMF is associated with gp78 phosphorylation. derived scatter factor has been shown to induce the dispersion Furthermore, here we report on the molecular cloning of the of Madin-Darby canine kidney epithelial cells and to induce human gp78-AMFR. Unexpectedly, the nucleotide and predicted amino acid sequences of this clone reveal significant * This work wassupported in part by the Paul Zuckerman Support homology with the human p53 protein, and this is the only Foundation for Cancer Research (to A. R.). The costs of publication of this article were defrayed in part by the payment of page charges. polypeptide found to date to be homologous to p53. This article must therefore be hereby marked ‘‘advertisement” in accordance with 18 U.S.C. Section 1734 solelyto indicate this fact. The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBLDataBankwith accession number(s) M63175. $ Present address: Dept. of Orthopaedic Surgery, Gunma University School of Medicine, 3-39-22 Showa-Machi, Maebashi-Shi, Gunma-Ken, Japan. § Present address: Dept. of Cell Biology, The Weizmann Institute of Science, Rehovot, Israel. T Present address: Dept. of CellBiology and Anatomy, Cornel1 University Medical College, New York, NY 10021. (1 To whom correspondence should be addressed Michigan Cancer Foundation, 110 E. Warren Ave., Detroit, MI 48201. Tel.: 313-8330710.
EXPERIMENTALPROCEDURES
Cells and Culture Conditions-The cell lines used in this study were the human HT-1080 fibrosarcoma (obtained from the American Type Culture Collection (ATCCCCLlSl)), the mouse B16-Fl melaI The abbreviations used are: AMF, autocrine motility factor; gp78, glycoprotein of78 kDa; AMFR, autocrine motility factor receptor; mAb, monoclonal antibody; SDS, sodium dodecyl sulfate, PAGE, polyacrylamide gel electrophoresis; CMF, Ca2’- and M$’-free; PBS, phosphate-buffered saline (pH 7.2); BES, 2-[(2-hydroxyethyl)amino] ethanesulfonic acid. Silletti, S., Watanabe, H., Hogan, V., Nabi, R. I., and Raz, A. (1991) Cancer Res., in press.
13442
and Motility Receptor Factor noma (obtained from Dr. I. J. Fidler, M. D. Anderson Cancer Center, Houston) and thehuman HeLa-S3cervical carcomina (obtained from Dr. R. Lotan, M. D. Anderson Cancer Center). All the cell lines were grown on plastic in Dulbecco's modified Eagle's minimal essential medium supplemented with glutamine, essential and nonessential amino acids, vitamins, antibiotics, and 10% heat-inactivated fetal bovine serum. The cells were maintained at 37 "C in a humidified atmosphere of 7% COz,93% air. To ensure reproducibility, all experiments were performed with cultures grown for no longer than 6 weeks after recovery from frozen stocks. HT-1080 AMF Purification-Semiconfluent 3-day monolayer cultures of HT-1080 fibrosarcoma cells were washed three times inCa'+and M$+-free phosphate buffered saline (CMF-PBS) and then cultured in serum-free medium. Twenty-four h later the medium was collected and replaced by fresh serum-free medium which was also collected 24 h later; no loss of cell viability was observed during this culturing as determined by dye exclusion. The conditioned medium (1,000 ml) was clarified by centrifugation a t 10,000 X g for 10 min at 4 "C and contained 7.5 pg/ml protein (13) with a specific activity (the concentration required to yield 100% motility stimulation of cells as compared with anti-gp78 mAb stimulation (11)) of 6 pg/ml. Samples were assayed for dose-dependent migratory stimulation according to Albreicht-Buehler (14) using serial dilution insterilePBS. The remaining mediumwas dialized extensively (at least 40 volumes) against double-distilled sterilized water at 4 "C and concentrated by lyophilization (VirTis model 10-145MR-BA). The freeze-dried material was resuspended in 10 mlof PBS buffer (pH 7.2). redialized against 40 volumes of PBS (4 "C), and assayed for protein content and migration stimulating activity as above (protein concentration, 720 pglml; activity, 6 pglrnl). This was followed by washing through a 30,000 NMWL filter unit (Ultrafree-MC, Millipore) to remove any small molecular weight macromolecules. The resultant (10 ml) was assayed for migration stimulating activity as above (5.5 pg/ml) and applied to a 123-ml bed volume Sephacryl S-200 (Pharmacia LKB Biotechnology Inc.) molecular sieve using CMF-PBSas eluting buffer. One-ml fractions were collected at a flow rate of 5 ml/h at 4 "C. Each fraction was assayed for migration stimulatory activity, and a single major peak of activity was identified and contained 2.5 pg/ml purified AMF. The optimal concentration necessary to elicit a 100% motility stimulus was determined by serial dilution of the purified AMF in sterile PBS as above. The peak was '261-labeled by the chloramine-T method (15) and analyzed by one- and two-dimensional polyacrylamide gel electrophoresis. One-dimensional 12.5% slab SDS-PAGE was performed according to Laemmli (16). Twodimensional PAGE was performed according to themethod of O'Farre11 (17) as described (18), and iodinated proteins were detected by autoradiography. Molecular Cloning of the Human gp78-A human cervical carcinoma, HeLa cells Xgtll cDNA library (provided by Dr. J. Christman, Michigan Cancer Foundation, Detroit) was screened a t a density of 15,000 plaque-forming units/l50-mm plate, according to Young and Davis (19) as described (20). Briefly, after 3 h at 42 "C the plates were transferred to 37 "C, and nitrocellulose filters presaturated with were overlaid on the 10 mM isopropyl 1-thio-@-D-galactopyranoside agar for 16 h. After washing with TBS buffer (150 mM NaCl, 50 mM Tris-HC1, pH 8, containing10% low fat milk and 0.05% sodium azide) the filters were incubated with rat anti-gp78 antibody ( l l ) , preabsorbed with lysate Escherichia coli strain Y1088 bound to nitrocellulose and diluted 1500 in TBS buffer for 2 h at 24 "C. After washing in TBS buffer containing 0.02% Tween 20, the filters were reincubated with 'ZSI-labeledsheep anti-rat Ig for 30 min (21) at 24 "C (8 X lo6 pM/fiker) and washed extensively, as above. Those plaques containing bacteriophages that generated positive signals were processed through successive rounds of antibody screening until 100% positive plaques were obtained. One of the clones, HL6, was processed as follows. Plaque-purified phages were incubated in LB medium with E. coli Y1088, grownfor 4 hat 42 "C, and induced by adding isopropyl 1-thio-@-D-galactopyranosideto a final concentration of 10 mM at 37 "C for 3 h. Fused protein was extracted by adding 0.01 volume of 10 X extraction buffer (10 mM NaHPO,, pH 7.5, 68 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 0.4 mM phenylmethylsulfonyl fluoride). Extracted fused proteins were separated by 8%SDS-PAGE and transferred to a nitrocellulose filter. The filter was incubated with mouse anti-@-galactosidase antibody (Promega Biotec, Madison, WI) or the anti-gp78 mAb (11).Positive bands were identified with Iz5I-labeledsheep anti-rat Ig or '251-labeledgoat antimouse Ig (Amersham Corp.) for gp78 or @-galactosidase,respectively. used The HL6 cDNA clone was 5' end-labeled with [T-~*P]ATP and
p53
13443
as a probe to screen 1.6 X lo6 Xgtll recombinant phages from a human HT-1080 fibrosarcoma cDNA library (Clontech). Briefly, plaques were transferred to nitrocellulose and were prehybridized at 65 "C for 3 h in a solution containing 5 X Denhardt's solution (6 X SSC, 0.1% SDS, 0.1 M NaPO,, pH 7.0), and 150 pg/ml denatured herring sperm DNA. Hybridization was done overnight at 65 "C in prehybridization solution containing 106-107cpm/ml probe. After hybridization, filters were washed once in 3 X SSC, 0.1% SDS for 30 min, once in 0.5 X SSC, 0.1% SDS for 30 min, and once in 0.1% SDS for 15 min. All washes were at 65 "C. cDNA restriction fragments were subcloned into PGEM-3Zf(-) (Promega) vector, and the nucleotide sequence of the cDNA was determined in both strands, in opposite directions by the dideoxy chain termination method. The T7 DNA polymerase Sequenase version 2.0 system was used for sequencing according to the instructions of the manufacturer (U. S. Biochemical Corp.). Northern Blot Analysis-Five pg ofpoly(A)+RNA wasfractionated by electrophoresis on 1%formaldehyde agarose gels and blotted onto nitrocellulose. The filters were probed with [32P]dATP-and 13'P] dCTP-labelednick-translated probes (specific activity, 2-5 X 10' cpm/pg; 3 X lo7 cpm/filter). The filters were washed twice in 2 X SSC, 0.2% SDS for 30 min at room temperature and twice in 0.1 X SSC, 0.1% SDS for 30 min a t 50 "C before autoradiography. Biosynthetic Labeling and Analysis of gp78-B16-F1, HeLa, and HT-1080 cells were cultured for 3 days to -70% confluence in 60mm culture plates. The plates were washed twice with PBS and incubated for 1h in L-cysteine-free Eagle's medium (GIBCO) supplemented with 5% dialyzed fetal bovine serum. ~ - [ ~ ~ S ] c y s t e(100 ine pCi/ml) was added, and the cultures were reincubated for an additional 4 hat 37 "C. Cellextraction and immunoprecipitation with rat anti-mouse anti-gp78 mAb (11) were performed as described previously (21). Radiophosphate Labeling and Immunoprecipitation of the Human gp78-HT1080 human fibrosarcoma cells were plated at equal numbers and grown for 3 days to -70% confluence in 100-mm culture plates. The plates were washed two times with ABS buffer (6.4 g/ liter NaCl, 0.3 g/liter CaCl X 2 HzO, 0.4 g/liter KCl, 4.0 g/liter NaHC03, 0.2 g/liter MgS0, X 7 H20, 1.0 g/liter glucose, 0.1 g/liter sodium pyruvate, 2.13 g/liter BES (Sigma), pH 7.4) prewarmed to 37 "C. After washing, the cells were incubated for 30 min at 37 "C with 4 ml of ABS buffer. At this time 400 pCi of [y3'P]H3P04was added to each plate, and the plates were incubated for 1 h a t 37 "C. AMF was then added for 4 min at 37 "C in a volume of 1 ml with CMF-PBS as buffer, and 1 ml CMF-PBS was added to thecontrol a t this time. After the 4-min incubation with stimulus sodium azide was added to a final concentration of 0.2% to stop cell metabolism. The plates were washed two times with ABS uffer 0.2% azide (room temperature) and then harvested with 2 mM EDTA + 0.2% azide prewarmed to 37 "C (2 ml, 2-min incubation a t 37 "C). Plates were washed once with an equal volume of ice-cold ABS buffer 0.2% azide (to remove any remaining cells), and the cells were pelleted by centrifugation at 9,000 X g for 3 min a t 4 "C and the supernatants discarded. Pellets were resuspended in 500 pl of lysis buffer (0.5% Nonidet P-40, 1 mM EDTA, 2 mM phenylmethylsulfonyl fluoride in CMF-PBS, pH 7.51, incubated for 30 min on ice, and processed for immunoprecipitation as above. Phagokinetic Track Motility Assay-Uniform carpets of gold particles were prepared on glass coverslips coated with bovine serum albumin, as described by Albrecht-Buehler (14). Colloidal gold-coated coverslips were placed in 35-mm tissue culture dishes, and 2,000 HT1080 cells in suspension culture were added to each plate. After 24 h the phagokinetic tracks were visualized using dark-field illumination in aNikon inverted microscope at a magnification of 200 X. The area cleared of gold particles by a t least 50 cells was measured and the standard errorreflecting a 95% confidence level calculated.
+
+
RESULTS
Purification of HT-1080AMF-Tumor AMF is a cytokine that stimulates both random and directed migration by selfproducing cells2 (7, 8, 11). HT-1080, a human fibrosarcoma cell line, was utilized for these studies because it was found to secrete into the culture medium potent AMF activity. For AMF purification, serum-free conditioned medium was concentrated and subjected to molecular sieve chromatography. Each fraction was analyzed for its effect on HT-1080 cell motility by the phagokinetic track motility assay; the activity
Motility Receptor Factor
13444
and p53
was found to be eluted as a single peak as found for B16-Fl AMF.2 This material was collected, iodinated, and separated by 12.5% SDS-PAGE in the presence or absence of mercaptoethanol and found by autoradiography to comprise a single polypeptide (Fig. 1). Under reducing conditions the band migrates with an apparent molecular mass of 64 kDa and under nonreducing conditions as 55 kDa,suggesting that it is rich in cysteines forming intrapeptide disulfide bonds. Similar molecular sizes (under reducing and nonreducing conditions) were reported previously for the humanA2058 and themouse B16-Fl melanoma AMFs (7),' implying that the human and the mouse AMFs are likely to be similar and that the55-kDa AMF is not melanoma specific. The purified HT-1080 AMF stimulates HT-1080 cell motility in a dose-dependent fashion and exerts maximal stimulating activity between 10 and 20 pg/ml (Fig. 2 A ) . Fig. 2, B
and C, is a pictorial demonstration of the motility assay and the stimulatory effect of the AMF. HT-1080 cells were plated on gold particle-coated substrate in the absence (Fig. 223) or presence (Fig. 2C) of 10 pg/ml AMF. The AMF stimulates cell motility from 8.2 f 1.3 i2/h (Fig. 2B) to 17.8 f 1.5 i'/h (Fig. 2C). The 2-fold stimulation of the HT-1080 AMF is similar in magnitude to theeffect of the A2058 and the B16F1 AMFs on their respective cells' (7, 11). To testwhether the 55-kDa AMF band is compromised of one or more polypeptides, the '*'I-labeled AMF was subjected to two-dimensional PAGE. The "'I-labeled AMF resolved into two groups of polypeptides exhibiting PI values of 6.1, 6.2 (major) and 6.35, 6.4 (minor), respectively (Fig. 3). This observation raised the possibility that theHT-1080 AMF is a glycoprotein and thatsubtle changesin thecarbohydrate side chain composition may be responsible for the appearance of the four polypeptides with different isoelectric focusing points. Treatment of the '"I-labeled AMF with neuraminidase (0.1 units for 30 min at 37 "C)to remove terminal sialic a b acid residues did not affect its apparent molecular weight on SDS-PAGE (not shown). In addition, a nonradioactive AMF was separated by SDS-PAGE, electroblottedto nitrocellulose, and subjected to a "'I-labeled wheat germ agglutinin overlay procedure (21). No radioactive band was detected after pro64Om-55 longed (2 weeks) autoradiography (not shown). Based on these two criteria we have concluded that HT-1080 AMF is most probably not a glycoprotein. The AMF polypeptide lost its migratory stimulation activity after heat inactivation (1 min a t 100 "C) and/or reduction with 2 mM mercapthoethanol. Demonstration That Human Tumor Cells Express a Phosphorylated gp78"We have shown recently that a 78-kDa surface glycoprotein (gp78) of murine B16-F1 melanoma cells is involved in the stimulation of cell motility in vitro and FIG. 1. HT-1080 AMF was purified from HT-1080 serum- metastasis in vivo2 (11, 21). We have designated gp78 as the free culture-conditioned medium by molecular sieve chroma- B16-F1 melanoma AMF receptor (AMFR). In later studies tography (as described under "Experimental Procedures"). The peak of activity was "'I-labeled and separated by 12.5% SDS- gp78 was found not to exhibit melanoma specificity, as it is mediates cell migration of other cells (22, PAGE under reducing ( a ) and nonreducing ( b ) conditions followed both present on and by autoradiography.Arrows mark the migration of the 64-kDa (left) 23). Immunoprecipitation of ["S]cysteine-labeled cells was and the 55-kDa (right)polypeptides. performed to establish the identity of the human target antigen reactive with mouse anti-gp78 antibodies. As shown in Fig. 4, a rat mAb raised against affinity-purified mouse gp78 (11) recognizes its antigen out of total cell extract proteins 100 "
6.2
-.' 5
8.4
80
60
I4 -0
a 40
tn
5 5 KO-
*
20
0 0
10
20
30
40
50
CONCENTRATION(pprml)
FIG. 2. Dose-dependent motility stimulation of HT-1080 cells by purified AMF. HT-1080 cells were plated on colloidal gold cover slides in culture medium in the absence or presence of various AMF concentrations. After 24 h cells were photographed under darkfield illumination, and the particle-free area of at least 50 individual cells was measured. Motility stimulation is plotted as a percent of the maximal stimulation in the presence of 50 pg/ml AMF ( A ) .Representative photomicrographs of cell plated in the absence ( B ) or presence (C) of 10 pg/ml AMF are shown (magnification X 350). The average area cleared of gold particles and the standard error are given under "Phagokinetic Track Motility Assay."
FIG. 3. Two-dimensional gel electrophoresis analysis of purified HT-1080 AMF. A sample of "'I-labeled AMF was subjected to isoelectric focusing in the first dimension (left to right) and to SDS-PAGE (nonreducing) in the second dimension (top to bottom). The pH values of the gel slices corresponding to the migration of the AMF in the isoelectric gels are indicated on the upper part of the figure. I4C molecular weight standards were electrophoresed in the second dimension alongside the AMF sample. The gels were processed for autoradiography.
Receptor Factor Motility
and p53
13445
insert was designated gp78-hAMFR (human autocrine motility factor receptor). The isolated gp78-hAMFR cDNA was used as a probe to analyze the mRNA expressed in human tumor cell lines (Fig. . ) 6). A single size transcript of 3.5 kilobases was detected in ~ ~ 7 8 both S3-HeLa and HT-1080 fibrosarcoma cells. The 1.9 kilobases of hAMFR cDNA were subcloned and sequenced from both ends (Fig. 7). The HL6 insertwas sequenced in parallel, and the two cDNA cloneswere found to be identical over the overlapping regions that encompassed the first490 base pairs of the 5' end. Fig. 7 depicts the nucleotide sequence of the 176-base 5'-flanking region which is high in G + C content 0 (62%), 969 coding, and 665 3'-flanking nucleotides including FIG.4. Immunoprecipitation analysis of tumor cell extracts the consensus polyadenylation signal (AATAAA). The disby anti-gp78 mAb. a, B16-Fl mouse melanoma; b, HT-1080 human fibrosarcoma; c, S3-HeLa cervical carcinoma. a', b', and c', control crepancy between the size of the isolated cDNA clone and immunoprecipitations with preimmune serum.Arrow a t /eft points to that of the transcript estimatedfrom Northern analysismay the migration of 78 kDa. Subconfluent cultures were washed twice be a result of a deficiency of 5' sequence, lack of poly(A) tail, with PBS and incubated for 1 h in L-cysteine-free Eagle's medium or both. The nucleotide sequence of gp78 contains an open supplemented with 5% diaylzed fetal bovine serum. ~-[:''S]Cysteine reading frameencoding for 323 amino acids with an initiating (100 pCi/ml) was added, and the cultures were reincubated for an methionine at nucleotide 177 (Fig. 7). The mature protein is additional 4 h a t 37 "C. Cell extraction and immunoprecipitationwere performed as described under "Experimental Procedures." The pro- predicted to begin at Ala-18 which follows the putativeleader teins were analyzed by 8% SDS-PAGE and visualized by fluorogra- sequence with a peptide cleavage signal site that obeys the phy. Lane M , "C-protein markers from top, 92.5,69,46, and 30 kDa. (-3,-1) rule (24). The gp78-hMFR has the structuralfeatures of an integral membrane protein,with a hydrophobic stretch of 25 amino acidslocatedbetweenresidues 111 and 137, M a h c d consistent witha transmembrane segment that makes asingle helical span. The extracellular NH, terminus includesone 200potential N-linked andseveral potential 0'-linked glycosylation sites, supporting theprevious finding thatgp78 is glycoc? sylated with both N- and 0'-linked oligosaccharides (21). 92.5* 9P7(5 A computer-assisted searchof data bases for gene homology x 89(data basesfrom GenBank, EuropeanMolecular Biology Laboratory, and the National Biomedical Research Foundation 48were analyzed with software from the University of Wisconsin Genetics ComputerGroup and theWORDSEARCH, ALIGN " and FASTA programs, last test screen, December 6, 1990) -mrevealed that the gp78-hAMFR cDNA has not been cloned FIG. 5. Phosphorylation of gp78 from resting and migrapreviously. Surprisingly, however, the gp78-hMFR cDNA nution-stimulated HT-1080 fibrosarcomacells. Resting HT-1080 cleotide sequence was found to share 50.1% homology with cells ( a and b ) and AMF-stimulated cells ( c and d ) were labeled with '9' and immunoprecipitated with preimmune serum ( a and c) or anti- the human p53 cDNA nucleotide sequence (not shown). Opgp78 mAb ( b and d ) (for details see "ExperimentalProcedures." The timal alignment of the deduced amino acid sequences of the
a
a'
---
b
b' c
-
C'
--
M
--
-
0,
2
phosphoproteins were analyzed by 8% SDS-PAGE andvisualized by autoradiography. Lane M , 'IC-protein markers.
a
b
from B16-F1 cells and from two human cell lines, i.e. HT1080 fibrosarcoma and S3-HeLacervical carcinoma (Fig. 4). It was shown previously that the motility stimulation pathway of AMF and of anti-gp78 mAb, which mimics the physiological effect of AMF, involves a GTP-binding protein(s) (7, 11). To study further the role of gp78 in AMF motility stimulation, gp78 was examined by labeling HT-1080 cells with [:"P]H3P04 and then stimulating with AMF (Fig. 5). The gp78 from '"P-labeled HT-1080 was not constitutively phos28s phorylated (lane b ) and was detected after the cells were - 0 stimulated for 4 min at 37 "C with AMF (lune d ) . Molecular Cloning of the HT-1080 gp78"Screening of a Xgtll cDNA library from human HeLa carcinoma cells was 18Sperformed using ratmAb against affinity-purified mouse gp78 (11) and '*'I-labeled goat anti-rat IgG. Six clones were obFIG. 6. RNA blot analysis of gp78-hMFR a, HT-1080 hutained after four rounds of purification. The largest clone man fibrosarcoma andb, S3-HeLa human cervical carcinoma. (HL6) -500 base pairs, produced a /3-galactoside-fused protein Positions of 18 and 28 S rRNA are indicated at the left. Five pg of recognized by anti-gp78 (not shown). Since no larger cDNA polyadenylated RNAwas fractionated by electrophoresis on 1%formnitrocellulose. The filters were clones were obtained, the HL6 insertwas used as a probe for aldehyde-agarose gels and blotted onto with ['"PIdATP- and ['"PldCTP-labeled nick-translatedgp78screening a second human Xgtll cDNA library constructed probed hMFR (specific activity, 2-5 X 10' cpm/pg; 3 X 10' cpm/filter). The from HT-1080 fibrosarcoma cells(Clontech). Restriction map filters were washed twice in 2 X SSC, 0.2% SDS for 30 min a t room analysis revealed that clone HT-19 (-1,900 base pairs) shared temperature and twice in 0.1 X SSC, 0.1% SDS for 30 min a t 50 "C a common restriction site (PuuII) with the HL6 clone. This and visualized by autoradiography.
.
Motility Factor Receptor and p53
13446
GGGGGGMGGCCMGCAGTGACCAGGMGAGGGAG~CTTCTGCTCAGACCGAGC
GTGTGCCACTGGACCTCAGTCCTCGCCTGGAGGAGACGCTGGACTTCffiCGAGGTGGMG TGGILGCCCAGTGAGGTGGAGACTTCGAGGCTCGTffiGAGCCGCTTCTCCMGTCTGCTQ ATGAGAGACAGCGCATGCTGGTCGCAGCGTMGGACGMCTCCTCCAGCMGCTCGCAAR
1
M
R
D
S
A
C
Y
S
Q
R
K
D
E
L
L
Q
Q
A
R
K
21
CGTTTCTTGAACMMGTTCTGMGATGATGCGGCCTCAGAGAGCTTCCTCCCCTCGGAA R F L - S E D D A A S E S F L P S E
41
GGTGCGTCCTCTGILCCCCGTGACCCTGCGTCGMGGATGCTGGCTGCCGCGCGGMCGGA G A S S D P V T L R R R M L A A A R N G
81
GGCTTCAGMGCAGCAGACCTCCTAGCGCTCCCTTGCCTTCCTCAGCT~CTCCTGCGCC G F R S S R P P S A P L P S S A A S C A
b1
CTGTGCCCGACTGACTGGAGGAGGCCTGTCCCMTTCTGCCGCTCCATGG~GCGGGC L C P T D W R R P V P I L P L H G K A G
TTGACTGCATTGCCGCTGTATAAI\GCATGTGGTCTTATAGTGTTTGGACAGCTGATRRAT L T A L P L Y K A C G L I V F G Q L I N 101
121
TTMTCCTTCTTTGTMTACTTTCTATGTGACATTTCTCTTCCCCTTAG~CACTGCM L I L L C N T F Y V T F L F P L E T L Q 0
, ATTTTMCTGTAGGTATGATCTCTTCTGGTGTTGACTGGACTGCTTG~TGGGGGACGA "I L T V G M I S S G V D W T A Y G G G R 141
TCAGGAGGAAGTGAGCCAGTCGCCTGCCTGCAGCAGGCAGCTTCTACTCCTGCCTCATGC 181 S G G S E P V A C L Q Q A A S T P A S C ATACGTCCCACAAATGCAGGTGTCCTGAGCACCACACCCAGTGG~GAGTGTGGGGGAG
161
I
R
P
T
N
A
G
V
L
S
T
T
P
S
G
K
S
V
G
E
201
GCGCACAGTGTGAGCCCGCCCCCACGTCGTGGGGTAACATCTGTTATC~ACT~TGTCG A H S V S P P P R R G V T S V I K L L S
ra 1
TTGTTGTGGMGCATGTAGACTGTGCCAGAGCCAGACCCACG~CT~TG~CCCCTGAG L L W K H V D C A R A R P T G S C T P E
¶4 1
CAGCAGGGCATCTTGGMAAGGMCTCTTGGTTCGATACCTGGILGCAGAGGILGG~AAAG Q Q G I L E K E L L V R Y L E Q R R G K
281
TCCAGGGCTATAGGGTGTGATGMGTCACCCCTTTCACCCCTTTCTGTCCCACTACATCTGGGACTGAC S R A I G C D E V T P F C P T T S G T D
re1
TTTCCGAGCCTCCAGTCCRCCGGCTTGATTTCCGTGMCTCTGGTGCTCCT~ATCT F P S L Q S K A G L I S V N S G A P A S
Wl
CATGAGTGTGCCCCATffiGTCCCCTCCCCTCTCAGCATTTCCTTGTCCCGTCTGGACCTG H E C A P W V P S P L S I S L S R L D L GGGAGTGGTTAGGCAGCAAGCTTTGGTTTTATGGTTTTCATTCATTGGTGMGTAMTTAG
a21
G
S
G
*
GCAGTGCTAAAGCCTGTGGGTTTGGTCCTTGMCMGATGTGGGCCTTGCMGATffiGAG
AGTAAACCTTG~GGGCTTTATTA~GRRAT~~GMCTTTTGTATCTTTTATCCTG GGAGCACTGCGTTTTCCTAGCTGTGTTATTCCTGGTTTMTTCAGCAGAGMGGTMGGT
GTGAACCTACCTGCCTTGGAGAGGCCCAGGTCCCMATCTCTTC~TTCTTCA~TGTT TARCTTTAAGGATTTGAACCATGAAGTCATAGG~ACAGACCTCAGTTTTATGCCCCATT
GGATTACTTTTTTTTTTTTTTTTTTTTTTTTACTCTTTGAAAGCTTTGTTTTGTGGTAGT
CGCTTTTGGGMGAATCCAGTATTATCTACMTTATTGGC~GTTT~TGTATTTTAC ATMCGGRRAGTTTTTAGAATGTTGMAAGTMTTG-GGTGATAGGTAAATTTTTA GGCRMGATMTTTATTTCMTAAATCTTTC~AGCCTTACCTT~TGCTGTTAGTA AAT~CTGTGCATTTTTTTTTTTTMTTTGTTTTGCTGAGILGCATAGCTATTTGTTTTTA TTGTAMCCCGCCC
FIG. 7. The cDNA sequence and predicted amino acid sequence of gp78-hAMFR. Nucleotides are numbered positively in the 5' to 3' direction, beginning with the initiator methionine and are numbered negatively in the 3' to 5' direction (right). The predicted amino acid sequence is shown below the nucleotide sequence in one-letter aminoacids (left).The transmembrane domain is singly underlined, the potential N-linkedglycosylation site isdoubly underlined, and cysteines are identified by asterisks.
two cDNA clones (ALIGN program) revealed 27.2% identity over 296 amino acid residues, and when conserved amino substitutes are considered the degree of homology increases t o 44.5% (Fig. 8). DISCUSSION
The data presented in this report establish that HT-1080 fibrosarcoma cells secrete an AMF with an apparent molec-
W R hpS3
59 MGGFRSS 80 PTPAAPA
@ t TLP
gf: ED^^^^^^^^^^ ~ ~ ~ ~ ~
h"R
151
hp53 172
30": &qgg$g!gEgjl FIG. 8. Homology between thegp78-hAMFR and the human p53 protein. Identical residues are boned. 0, conserved amino acid substitution. Numbers indicate position of amino acid from the first methionine. Dashes indicated gaps for optimal alignment.
ular mass of 55 kDa on SDS-PAGE. Under reducing conditions, AMF exhibits an apparent molecular mass of 64 kDa. The purified AMF induces motility of the secreting HT-1080 cells in a dose-dependent fashion. AMFs have been purified from A2058 human and B16-F1 mouse melanoma conditioned media and migrated on nonreduced and reduced SDS-PAGE with molecular weights identical to the HT-1080 AMF' (7, Fig. 1).It was proposed that the higher migration of AMF in the presence of a reducing agent indicates the reduction of existing intrachain disulfide bonds. The HT-1080 fibrosarcoma AMF wasresolved by isoelectric focusing into two groups of four polypeptides with PI values of 6.1,6.2 (majors) and6.35,6.4 (minors). In contrast the mouse AMF was resolved by isoelectric focusing into two polypeptides with PI values identical to the minor human group, 6.35 (major)and 6.4 (minor).' The reason for the appearance of AMFs with the same molecular weight but with four different isoelectric points is unknown. Post-translational modification such as glycosylation could explain this behavior; however, attempts to demonstrate the presence of a covalently bound carbohydrate in purified HT-1080 AMF by neuraminidase cleavage and lectin binding were unsuccessful. It is possible that there ismore than one homologous gene coding for AMFs of the same molecular weight or that the polypeptides undergo slight post-translational modification that may lead to theappearance of the differently charged AMF isoforms. Similarly no carbohydrate side chains were detected on the B16-F1 AMF.' These properties distinguish AMFfrom other motility-inducing factors, i.e. the 70-kDa migration stimulating factor (5, 6), the glycoslyated 32-92kDa scatter factors (2-4), the acidic and basic fibroblast growth factors (25, 26), and insulin-like growth factor (26). AMF, like scatter factor, migrates on SDS-PAGE differently in the presence of a reducing agent (2-4). However, unlike scatter factor, which dissociates under reducing conditions into several polypeptides (2-4), the mouse B16-Fl; the human A2058 melanoma (7), and the HT-1080 fibrosarcoma (Fig. 1)AMFs remain as intactpolypeptides regardless of the reducing conditions. Previously, a mAb directed against the AMFR-gp78 was used to study its surface distribution and possible function in cell locomotion (11).On motile cells, gp78 was localized by immunofluorescence to the leading lamella as well as to the trailing edge, suggesting shuffling of gp78 during cell migration. Anti-gp78 mAb was found to mimic the physiological effect of AMF and the enhanced motility induced by either anti-gp78 mAb or AMF were both mediated by a PT-sensitive G protein pathway(s) as has been described for other motility
O
Motility Factor Receptor and p53
13447
factors (7, 28). The binding of anti-gp78 mAb to its antigen residues with internal homology from 47.1% (human-Xenowas inhibited (10-fold) by preincubation with B16-F1 AMF- pus) to 95.7% (human-monkey). All of the p53 polypeptides share five highly conserveddomains (I-V). Although it is containing serum-free conditioned medium. Based on such functional properties it was suggested that gp78 is the AMF evident that gp78-hAMFR is nota p53-like molecule since it receptor of B16-F1 melanoma cells (11). More recently we lacks the five distinct conserved domains and isa membrane between showed that immunopurified gp78, immobilized on nitrocel- protein, there are several similar structural features lulose, binds purified AMF directly in a protein-protein bind- p53 and gp78 which permit the speculation that both genes ing assay; this in conjunction with the competitive inhibitionmay have diverged from a common ancestral gene, probably of osteichthyes: 1)the nucleotide and of AMF-stimulated cell motility by soluble gp78 has demon- prior to the appearance strated that AMF is the natural ligand of gp78.’ In this report amino acid sequence homology; 2) both are phosphoproteins; glycosylation site; 4) both the human homologue of the murine gp78 was identified and 3) both have one potential N-linked have proteinbinding sites;5) the amino andcarboxyl termini found tobe of identical molecular weight in SDS-PAGE. The motility stimulating pathway of AMF and of a che- of both molecules are helical and hydrophylic whereas the motactic peptide involves a GTP-binding protein (7, 11, 28). internal aminoacid sequence domain ishydrophobic and high Here we show that gp78-hAMFR undergoes phosphorylation in P-sheet. The last parameter is of special interest since it conserved after interaction with its ligand, which adds a new insight hasbeenpointedoutthatstructureisbetter into the poorly defined molecular pathway by which AMF through evolution than sequence (12). The availability of a cDNA clone containing thecomplete manifests itseffect on cells. The gp78-hMFR cDNA contains coding region for a human motility factor receptor will now the Ser/Thr-Xaa-Lys/Agr motif in the cytoplasmic (Ser-Glymolecular mechanisms by Lys 194-196) domain, which fits the consensus sequence for permit detailed studies into the which ligand-receptor interactions induce cell migration. Sitea phosphorylation site (29). The cytoplasmic domainof gp78of modified gp78 in hMFR also includes anucleotide-binding consensus sequence directedmutagenesisandexpression (Gly-Xaa-Gly-Xaa-Xaa-Gly),which is found in nucleotide- mammalian cells will make it possible to study thebiochembinding proteins including serine/threonine kinases (30) and ical mechanism of gp78 function. In addition, these results is located at the157-162 amino acid domain. Thus, activation may lead t o a better understanding of the structure-function of gp78 may result from its autophosphorylation afterligand- relationships of the p53 protein. binding or from a GTP-binding “coupling” protein(s) that is Acknowledgments-We thank J. Jenkins for typing; Dr. G. associated with the receptor on the cytoplasmic face of the Heppner, for her critical evaluation of the manuscript; and Dr. L. membrane, as describedfor rhodopsin, P-adenergic musca- Baker,Director of the Meyer L. Prentis Comprehensive Cancer rinic acetycholine receptors and the mating receptor of yeast Center of Metropolitan Detroit, for his support. (for review see Ref. 27). Alternatively gp78 may be coupled t o REFERENCES a differentintegralmembraneprotein by a GTP-binding protein as was described for the formyl-chemotactic peptide 1. Singer, S. J., and Kupfer, A. (1986) Annu. Reu. Cell Biol. 2, 337365 receptor (28). Most GTP-binding proteins are activated fol2. Gheradi, E., Gray, M., Stoker, M., Perryman, M., and Fulong, R. lowing G T P binding and become deactivated upon G T P hy(1989) Proc. Natl. Acad. Sci. U. S. A . 86, 5844-5848 drolysis. This may also provide a mechanism for coupling the 3. Rosen, E. M., Meromsky, L., Setter, E.,Vinter, D. W., and energy of G T P hydrolysis to the motilityprocess. Goldberg, I. D. (1990) Exp. Cell Res. 186, 22-31 Gp78-hAMFR wascloned from a humanfibrosarcoma 4. Weidner, K. M., Behrens, J., Vanderkerchove, J., and Birchmeier, cDNA library, and a cDNA clone containing the complete 2097-2108 W. (1990) -I. Cell Biol. 11 1, 5. Schor, S. L., Schor, A. M., Grey, A. M., and Ruston, G. (1988) J . coding region was characterized. The sequence contains a Cell Sci. 90, 391-399 single openreadingframeencoding 323 amino acids with 6. Grey, A. M., Schor, A. M., Ruston, G., Ellis, I., and Schor, S. L. features typical of an integral membrane protein. A hydro(1989) Proc. Natl. Acad. Sci. U. S. A . 86, 2438-2492 phobic region of 25 amino acids is located between residues 7. Liotta, L. A., Mandler, R., Murano, G., Katz, D. A,, Gordon, R. 111 and 137, consistent with a transmembrane segment that K., Chaning, P. K., and Schiffman, E. (1986) Proc. Natl. Acad. Sci. U. S. A . 83, 3302-3306 makes a single helical span. The sequences upstreamof this 8. Atnip, K. D., Carter, L. M., Nicolson, G. L., and Dabbous, M. K. element includes one potential N-linked and several potential (1987) Biochem. Biophy. Res. Commun. 146,996-1002 0’-linked glycosylation sites, supporting the previous finding 9. Guirguis, R., Margulies, I., Taralodetti, G., Schiffman,E., and that gp78 is glycosylated with both N- and 0’-linked oligoLiotta, L. A. (1987) Nature 329, 261-263 saccharides (21). Based on the above and on the facts that 10. Stracke, M.L., Kohn, E. C., Aznavoorian, S. A., Wilson, L. L., the antigenic recognition site of the anti-gp78 mAb AMF Salomon, D., Krutzsch, H. C., Liotta, L. A., and Schiffman, E. (1988) Biochem. Biophys. Res. Commun. 153, 1076-1083 binding site on gp78 and the N’-linked glycosylation site all reside on the extracellular exposed domain of gp78-hAMFR, 11. Nabi, I. R., Watanabe, H., and Raz, A. (1990) Cancer Res. 5 0 , 409-414 we conclude that the NH, terminusof gp78 is extracellularly 12. Soussi, T., de Fromental, C. C., and May, P. (1990) Oncogene 5 , exposed. 945-952 A computer search of several sequence data bases queried 13. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J . (1951) J. Biol. Chem. 193,265-275 with the entire1,810 base pairsof the gp78-hAMFR revealed the most surprising finding of this report: the homology of 14. Albreicht-Buehler, G. (1977) Cell 11,359-404 F. C., Hunter, W.M., and Glover, J. S. (1963) gp78 to the human p53 cDNA. p53 codes for a cellular serine- 15. Greenwood, Biochem. J . 89, 114-123 phosphoprotein, which normally acts as a tumor suppressor 16. Laemmli, U. K. (1970) Nature 227,680-685 gene, whereas its mutatedform acts as anoncogene. The gene 17. O’Farrell, P. H. (1975) J . Biol. Chem. 250,4007-4021 18. Paine, P. L. (1984) J. Cell Biol. 99, 188s-195s product is thought to be involved in cell growth regulation and was shown to form protein-protein complexes mainly 19. Young, R. A., and Davis, R. W. (1983) Science 222, 778-782 with viral antigens (for review see Ref. 12). The available data 20. Raz, A., Avivi,A., Pazerini, G., and Carmi, P. (1987) Exp. Cell Res. 173, 109-116 concerning the evolutionary considerationof p53 reveal that 21. Nabi, I. R., and Raz, A. (1987) Znt. J . Cancer 40, 396-402 this gene is restricted to vertebrates. The gene encodes a 22. Hendrix, M. J. C., Wood, R. W., Seftor, E. A., Lotan, D., Nakagroup of molecules ranging in size from 362 to 396 amino acid jima, M., Misiorowski, R. L., Seftor, R. E. B., Stetler-Stevenson,
13448
23. 24. 25. 26.
Motility Factor Receptor and p53
W. G., Bevacqua, S.J., Liotta, L. A., Sobel, M. E., Raz, A., and Lotan, R. (1990) Cancer Res. 50,4121-4130 Ochieng, J., Basolo, F., Albini, A., Melchior, A., Watanabe, H., Elliott, J., Raz, A., Parodi, S., and Russo, J. (1991) Znuasion Metastasis., in press von Heijne, G . (1986) Nucleic Acids Res. 14,4683-4690 Sato, Y.,and Rifkin, D. B. (1988) J. Cell Biol. 1 0 7 , 1199-1205 VallBs, A. M.,Boyer, B., Badet, J., Tucker, G. L., Barritault, D.,
27. 28. 29. 30.
and Thiery, J. P. (1990) Proc.Natl. Acad. Sci. U. S. A . 8 7 , 1124-1128 Herskowitz, I., and Marsh, L. (1987) Cell 60, 995-996 Brandt, S. J., Dougherty, R. W., Lapetina, G. E., and Niedel, J. E. (1985) Proc. Natl. Acad. Sci. U. S. A . 8 2 , 3272-3280 Woodgett, J. R., Gould, K. L., and Hunter, T. (1986) Eur. J. Biochern. 161,177-184 Hank, S.K., Quinn, A. M., and Hunter, T. (1988) Science 2 4 1 , 42-52
