cdna cloning the human tumor motility-stimulating protein ... · production of anti-atx peptide...

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 269, No. 48, Issue of December 2, pp. 3047930484, 1994 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

cDNA Cloning of the Human Tumor Motility-stimulating Protein, Autotaxin, Reveals a Homology with Phosphodiesterases*

(Received for publication, August 2, 1994, and in revised form, September 16, 1994)

Jun Murata, Hoi Young Lee, Timothy Clair, Henry C. Krutzsch, Anders A. Lestad, Mark E. Sobel, Lance A. Liotta, and Mary L. StrackeS From the Laboratory of Pathology, National Cancer Institute, National Znstitutes of Health, Bethesda, Maryland 20892

A human cDNA clone encoding autotaxin, a tumor cell motility-stimulating protein, reveals that this protein is an ecto/exo-enzyme with significant homology to the plasma cell membrane differentiation antigen PC-1. ATX is a 125-kDa glycoprotein, previously isolated from a human melanoma cell line (A2058), which elicits chemotactic and chemokinetic responses at picomolar to nanomolar concentrations. Affinity-purified antipeptide antibodies to the ATX peptide, ATX-102, were employed to screen an A2058 cDNA expression library made in Agtll. The partial cDNA sequence which was obtained was then extended by utilizing reverse transcriptase on total cellular RNAfollowed by polymerase chain reaction am- plificatiop. The isolated cDNAclone contained 3251 base pairs, and the mRNA message size was approximately 3.3 kilobases. The deduced amino acid sequence of autotaxin matched 30 previously sequenced peptides and comprised a protein of 915 amino acids. Data base analysis of the ATX sequence revealed a 45% amino acid identity (including 30 out of 33 cysteines) with PC-1, a pyrophosphatase/type I phosphodiesterase expressed on the surface of activated B cells and plasma cells. ATX, like PC-1, was found to hydrolyze the type I phosphodiesterase substrate p-nitrophenyl thymidine-5‘-monophosphate. Autotaxin now defines a novel motility-regu- lating function for this class of ecto/exo-enzymes.

Active tumor cell motility is involved in many stages of the metastatic cascade, including the transition from in situ to invasive carcinoma (1). The regulation of this motile response is not well understood. However, multiple factors, both autocrine and paracrine in origin, appear to influence this tumor cell locomotion (2).

Recently, a potent new cytokine with molecular mass 125 kDa has been purified to homogeneity from the conditioned medium of the human melanoma cell line, A2058, utilizing sequential chromatographic methods (3). This new cytokine, termed autotaxin (ATX),l is a basic glycoprotein with PI - 7.8. ATX is active in the high picomolar to low nanomolar range, stimulating both chemotactic and chemokinetic responses in the ATX-producing A2058 cells as well as other tumor cells (4).

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

to the G’enBankmIEMBL Data Bank with accession number(s) L35594. The nucleotide sequence(s) reported in this paper has been submitted

Pathology, National Cancer Institute, National Institutes of Health, $To whom correspondence should be addressed: Laboratory of

Bldg. 10, Rm. 2A33, Bethesda, MD 20892. Tel.: 301-496-8906; Fax:

reaction; 5‘ RACE, rapid amplification of cDNA ends; Bistris, 2-[bis(2- The abbreviations used are: ATX, autotaxin; PCR, polymerase chain

hydroxyethyl~aminol-2-~hydroxymethyl)-propane-l,3-diol.

301-480-0853.

This motile response is abolished by pretreatment of the cells with pertussis toxin. ATX may therefore act through a G protein-linked cell surface receptor.

Sequence information, obtained initially on 19 purified tryp- tic peptides, confirmed that the protein is unique with no significant homology to growth factors or previously described motility factors. These peptide sequences have now been used as the basis for identifying and sequencing the cDNA clone for ATX.

EXPERIMENTAL PROCEDURES

Amp PCR Reagent kit with AmpliTaq and the GeneAmp RNA PCR kit Materials-Mi-Gel 10 affinity resin was from Bio-Rad. The Gene-

were purchased from Perkin-Elmer. The 5’ RACE kit was from Life Technologies, Inc. The p-nitrophenyl thymidine-5’-monophosphate was obtained from Calbiochem Biochemicals.

Cell Culture-The human melanoma cell line A2058, originally isolated by Todaro et al. (5), was maintained as described previously (6). The N-tera 2 (Dl clone) was a kind gift from Dr. Maxine Singer, Labo- ratory of Biochemistry, National Cancer Institute, National Institutes of Health, and was maintained as described (7).

Purification of Autotaxin-The purification of ATX has been described previously in detail (3) and involved successive fractionations of A2058 serum-free conditioned medium using phenyl-Sepharose CL-4B (Pharmacia Biotech Inc.), agarose-bound C o d (Vector Laboratories Inc.), ZORBAX BioSeries-WAX (weak anion exchange, Mac Mod), Spherogel-TSK 4000SW and 3000SW (TosoHaas), and Pro-Pac PA1 (strong anion exchange, Dionex Corp.) chromatographic separations. In some cases, chromatographic fractionation with ZORBAX BioSeries WCX (weak cation exchange, Mac Mod) was utilized between the weak anion exchange and the gel filtration steps. For this chromatographic procedure, the motility-stimulating peak from the weak anion exchange column was pooled and dialyzed into 25 mM Bistris, 20% (v/v) ethylene glycol (pH 6.5). Proteins were eluted with a linear gradient of sodium chloride (0-0.3 M).

Production of Anti-ATX Peptide Antibodies-Anti-ATX peptide antibodies were generated as described previously with slight modification (8, 9). In brief, the peptide termed ATX-102 (3) was synthesized on a Biosearch 9600 peptide synthesizer and conjugated to bovine serum albumin using glutaraldehyde. For the first injection into New Zealand White rabbits, the bovine serum albumin-peptide conjugate was emulsified with complete Freund’s adjuvant and injected subcutaneously. For subsequent injections, the bovine serum albumin-peptide conjugate was emulsified with incomplete Freund’s adjuvant. The resultant antiserum was heat-inactivated at 56 “C for 30 min. Immunoglobulins were precipitated in 47% saturated ammonium sulfate, then redissolved and dialyzed into phosphate-buffered saline. Antibodies were adsorbed onto peptide-conjugated Affi-Gel 10 resin, eluted with 0.1 N acetic acid, and neutralized with 2 M Tris-HC1 (pH 8). The resulting affinity-purified antibodies were dialyzed into Dulbecco’s phosphate-buffered saline, concentrated, and stored in aliquots at -20 “C.

Gel EZectrophoresis-Protein samples were analyzed by SDS-polyacrylamide gel electrophoresis in a Tris glycine buffer system as described by Laemmli (lo), using prepared 8-16% gradient gels (Novex).

Western Blot Analysis-Immunoblots were performed as described previously (11, 12) with 4-chloro-1-naphthol as developing agent. Pri- mary antibody was antipeptide 102 antibody diluted as described. Sec- ondary antibody was horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin (Pierce) diluted 1:2500.

Preparation of a Agtll cDNA Expression Library-Oligo(dT)-selected poly(dA) mRNA was prepared from A2058 cells using standard tech-

30479

30480 cDNA Clone for the Motility Factor Autotaxin niques. The mRNA >lo00 base pairs were then purified from an agarose gel and used to prepare double-stranded cDNA using the commercially available cDNA synthesis system from Promega. The cDNA inserts were synthesized using Not1 adapters and modified with EcoRI linkers. This allowed directional ligation into EcoRIINotI-digested Agtll Sfi-Not vector (Promega).

Antibody Screening of Agtll cDNA Expression Library-Approx- imately 5 x lo5 phage were screened by infecting LE 392 cells with the recombinant Agtll followed by growth on LB agar plates (13). Phage were transferred onto isopropyl-P-D-thiogalactoside-impregnated nitro- cellulose membranes by overnight incubation at 37 "C. The antibody screening was performed using a modification of previously described immunoblotting techniques (11, 12). In brief, the membranes were blocked for 30 min in blocking buffer (1 M glycine, 5% (w/v) nonfat dry milk, 5% (v/v) fetal bovine serum, and 1% (w/v) ovalbumin). The affinity purified anti-ATX-102 antibody was incubated with the membranes in blocking buffer for 2 h at room temperature, using a concentration of antibody which was double that which gave a strong response on West- ern blot analysis. Secondary antibody was horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin, and the blot was developed colorimetrically with 4-chloro-1-naphthol. Positive phage were selected and purified repeatedly using the same detection techniques until ho- mogeneously pure. The positive reactivities of these clones were confirmed by competition for antibody by specific versus unrelated peptides. cDNA inserts were subcloned into pBluescript (Stratagene) for further analysis.

5' Extension of the ATX cDNA Clone-A reverse transcriptase reaction was performed using total or oligo(dT) purified RNA from A2058 or N-tera 2D1 cells as template and an antisense primer from the 5' end of 4Cll (GCTCAGATAAGGAGGAAAGAG). This was followed by one or two PCR amplifications of the resultant cDNA using the commercially available kit from Perkin Elmer and following the manufacturer's directions. These PCR reactions utilized nested antisense primers from 4Cl l (GAATCCGTAGGACATCTGCTT and TGTAGGCCAAACAGT- TCTGAC) as well as degenerate, nested sense primers deduced from ATX peptides: ATX-101 (AAYTCIATGCARACIGTI'ITYGTIG and

GTIGAYGAYAT and GAYGAYATIACICTIGTICCIGGIAC), or ATX-224 (TGYTTYGARYTICARGARGCIGGICCICC). The amplified DNA was then purified from a polyacrylamide gel using standard procedures and ligated into the pCRTM plasmid using the TA cloning kit (Invitrogen Corp.) according to manufacturer's directions.

The 5' RACE kit was utilized to extend the 5' end ofATX cDNAusing total RNA from N-tera 2D1 as template and a previously obtained sequence as primer (GCTGTCTTCAAACACAGC). The 5' end of the A2058-synthesized protein was obtained by using a previously obtained sequence as primer (CTGGTGGCTGTAATCCATAGC) in a reverse transcriptase reaction with total A2058 RNA as template, followed by PCR amplification utilizing the 5' end of the N-tera 2D1 sequence (see "Re- sults") as sense primer (CGTGAAGGCAAAGAGAACACG) and a nested antisense primer (GCTGTCTTCAAACACAGC).

DNA Sequencing-DNA sequencing was performed using dideoxy methodology (14) and P5S1dATP (DuPont NEN).

Assay for Fype I Phosphodiesterase Enzymatic Activity-Phospho- diesterase activity was measured according to Razzell (15) with minor modifications. Samples were assayed in a 100-pl volume containing 50 mM Tris-HC1, pH 8.9, and 5 mM p-nitrophenyl thymidine-5'-monophosphate. After a 30-min incubation at 37 "C the reactions were terminated by addition of 900 1.11 of 0.1 N NaOH and the amount of product formed was determined by reading the absorbance at 410 nm.

TTYGTIGGITAYGGICCIACITTYAA), ATX-103 (AAYTAYCTIACMY-

RESULTS

Peptide Sequences of Purified ATX-Three separate homoge- neously pure protein preparations were digested with cyanogen bromide followed by trypsin (ATX-18 to ATX-591, trypsin alone (ATX-100 to ATX-104), or with endo Lys-C (ATX-204 to ATX- 244). Purification of the fragments and sequence analysis were performed as described previously (3). A total of 30 peptides containing 315 amino acid residues have now been sequenced (Table I). The partial amino acid sequence of ATX did not ex- hibit homology to known growth factors or previously described motility factors. Several peptides were sequenced from two of the three purifications (shown by double numbers in the peptide name). In addition, 2 peptides (ATX 212 and ATX-2131 217A) had apparent asparagine-linked glycosylation sites

Peptide

ATX- 18

ATX-20 ATX- 19

ATX-29 ATX-34B ATX-4 1 ATX-47 ATX-48 ATX-59 ATX-100 ATX-l01/223A ATX-102 ATX-103

ATX-104 ATX-204 ATX-205 ATX-209 ATX-210 ATX-212

ATX-215/34A

ATX-216

ATX-223B/24

ATX-214

ATX-213/217A

ATX-218/44

ATX-224 ATX-229 ATX-230 ATX-239 ATX-244/53

TABLE I Peptide sequences for autotaxin

Amino acid sequence

WHVAXN PXLDVYK

YPAFK PEEVTRPNYL

RVWNYFQR HLLYGRPAVLY

YDVPFDAT SNPPFENINLY

TFPNLYTFLATGGLYW

VNSMQTVFVGYGF'TFK XGGQPLWITATK

DIEHLTSLDFFR TEFLSNYLTNVD-

DITLVPGTLGR VNVISGPIFDYDYDGLHDTEDK

MHTARVRD FSNNAKYD

VMPNIEK TARGWECT

XDSPWT(N)ISGS LRSCGTHXPYM

TYLHTYES AIIANLTCKKPDQ

IVGQLMDG TSRSYPEILTPL

QAEVSSVPD RCFELQEAGPPDDK

SYTSCCHDFDEL XFNHQWGGQQP

AAECVPA QMSYGFLFPPYLSSSP

which contained the motif: (N)XT/S. The NH, terminus of the 125-kDa ATX band was successfully

sequenced from Immobilonm to yield the sequence: DSP- WT(N)I. This appears to be identical to ATX-212, now the presumed amino terminus of the secreted protein. Apparently, the NH,-terminal aspartic acid is oxidized in the electrophoretic procedures, since yields from the sequencing were <lo%.

Characterization of Antipeptide Am-102-Antibodies were raised in rabbits to the peptide designated ATX-102 and affln- ity-purified on a peptide affinity column. The resulting anti- ATX-102 antibody recognized a 120-kDa protein on Western blot analysis of partially purified ATX (Fig. 1, Eane 2). When a 1000-fold molar excess of peptide ATX-102 was preincubated with the antibody for 1 h prior to reacting with the membrane, detection of the protein band disappeared, indicating antibody specificity (Fig. 1, lane 3). This antibody appeared to recognize only denatured protein; it neither neutralized motility-stimulating activity nor immunoprecipitated native ATX.

Cloning of the ATX cDNA-Affinity-purified antipeptide ATX-102 antibody was used to screen an A2058 cDNA expression library. Eight purified cDNA clones were obtained that reacted with antibody in a peptide-specific manner. These clones were each sequenced and one clone, designated 4Cl1, appeared to contain the 3' terminus of ATX (Fig. 2 and Fig. 3). The 4Cll clone contained 1084 base pairs, including the polya- denylated tail and the AATAAA polyadenylation signal motif. The open reading frame region was 628 base pairs long and coded for 209 amino acids. Within this coding region were matches for 8 previously identified ATX peptides: ATX-20, ATX- 34, ATX-102, ATX-104, ATX-204, ATX-215, ATX-218144, and ATX-244. Northern blot analysis gave a weak band at -3.3 kilobase pairs, indicating that this clone represented approximately one-third of the total mRNA length. Data base analysis of the 4Cll clone revealed a 45% amino acid identity and a 57% nucleotide identity with a human nucleotide pyrophosphatase and kinase, PC-1, found on the surface of activated B cells and plasma cells. This homology was utilized to estimate the rela-

cDNA Clone for the Motility Factor Autotaxin 30481

281,300 +

100,500 +

71,800 +

43,200 +

28,500

1 2 3 FIG. 1. Immunoblots of ATX with antipeptide ATX-102 anti-

body. Antipeptide antibodies were produced in rabbits by injecting peptide ATX-102 conjugated to the carrier, bovine serum albumin. The resultant antiserum was purified on an peptide affinity column. Par- tially purified ATX was then separated by SDS-polyacrylamide gel electrophoresis run under nonreducing conditions and transferred electro- phoretically to an ImmobilonT" membrane. The affinity-purified antipeptide antibody recognized a broad 120-kDa protein band on immunoblot (lane 2). When the antibody was preincubated with 1000- fold molar excess of peptide for 1 h prior to reacting with the immunoblot, the protein band was not detectable (lane 3 ). Prestained molecular weight markers are shown for comparison in lane 1.

2 2 4 1 0 3 1 0 1 4C11

I I P

W A ?. V V V S"UTR 3'-UTR

I I I

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

FIG. 2. Cloning strategy for ATX cDNA. The cDNA clone, 4Cl1, was selected from an A2058 expression library using the antipeptide ATX-102 antibody as a probe. This sequence was then utilized to syn- thesize primers for a reverse transcriptase reaction usingA2058 total or oligo(dT)-purified RNA as template. Nested primers from 4Cll and previously sequenced ATX peptides (ATX-224, ATX-103, and ATX-101) were utilized for PCR amplification of the reverse transcriptase reaction product (as described under "Experimental Procedures"). 5' RACE was utilized to obtain the far 5' end in N-tera 2D1 cells and this sequence was confirmed in A2058 cDNA. The arrows show the direction of DNA sequencing. The scale indicates nucleotide positions in kilobases (hb). UTR, untranslated region.

tive positions of several of the larger ATX peptides. The 4Cl l clone was extended using reverse transcriptase/

PCR methodology. The reverse transcriptase reaction was performed using the 5' end of 4Cll as primer and total or oligo(dT)-purified RNA from A2058 or N-tera 2D1 as template.

This was followed by one or two PCR amplifications using nested primers deduced from ATX-101, ATX-103, and ATX-224 (Fig. 2). In order to obtain the cDNA sequence encoding the amino terminus of the protein, 5' RACE methodology was utilized and succeeded in extending the 5' end of the cDNA product from N-tera 2D12 but not from A2058 ATX. An oligonucleo- tide with sequence matching the 5' terminus of this N-tera 2D1 cDNA clone was then utilized in a reverse transcriptasePCR amplification to obtain the 5' end sequence of A2058 ATX cDNA. The cDNA and deduced protein sequence is shown in Fig. 3. The total length of the cDNA is 3251 base pairs, roughly equivalent to the length of the mRNA seen on Northern blot analysis. I t includes an apparent initiating methionine sur- rounded by a partial Kozak consensus sequence (16).

This deduced sequence has been shown to match all 30 of the previously sequenced peptides (Table I, Fig. 3). For 25 of these peptides, the match is exact. In the four peptides ATX-230, ATX-19, ATX-48, and ATX-18, technical difficulties caused some misreadings of the sequence data. For peptide ATX-59, the glutamine in the 10th amino acid position appeared to be correct and could represent a variation in splicing. In summary, the peptide data provides strong evidence that we have sequenced the correct cDNA for purified ATX.

Protein Domains and Homologies-Searches of protein data bases (17) confirmed that the homology between ATX and PC-1 was present throughout the length of the extracellular portion of the molecules (18, 19). There is a 45% amino acid identity and a 64% similarity between the 2 protein sequences (Fig. 4). For the cDNA sequence, the identity is -57%.

These proteins share several interesting properties and domains (Fig. 5). Both have a number of potential N-linked glycosylation sites: four for ATX (Asd4, Asn8") and nine for PC-1. Both have adjacent somatomedin B domains near the amino end of the extracellular domain. This somatomedin B domain is a cysteine-rich region containing 3 presumed cystine cross-linkages. ATX has 33 Cys residues and PC-1 has 37; 30 of these Cys residues are identical in place- ment. Both proteins also contain the loop region of an EF-hand (18, 20). In addition, both proteins have a transmembrane/ signal peptide region with a short intracellular peptide, com- mon in ectoenzymes (21). However, the amino acid identity between ATX and PC-1 in the intracellular and transmembrane regions is only 11%.

Finally, both proteins have a region homologous to the bovine intestinal phosphodiesterase enzymatic domain with conserva- tion of the threonine that is thought to act as the intermediate phosphate binding site (22). PC-1 has been demonstrated to have phosphodiesterase type I, nucleotide pyrophosphatase, and threonine-specific kinase enzymatic activities (23, 24). In order to test whether purified ATX had type I phosphodiesterase activity, samples were incubated with p-nitrophenyl thymidine-5'-monophosphate a t pH 8.9 for 30 min. ATX was found to hydrolyze the p-nitrophenyl thymidine-5'-monophosphate (15) a t a rate of 10 pmol/ng/min, a reaction rate similar to that reported for PC-1 (25).

DISCUSSION

We have sequenced the cDNA and deduced the primary amino acid structure of ATX, a tumor cell motility-stimulating protein. Protein data base searches of this sequence revealed a 45% amino acid identity with the plasma cell membrane marker protein, PC-1. ATX and PC-1 appear to share a number of domains, including two somatomedin B domains, the loop region of an EF-hand, and the enzymatic site of type I

* Stracke, M. L., Arestad, A. A., Levine, M. D., Krutzsch, H. C. and Liotta, L. A., Melanoma Res., in press.

30482 cDNA Clone for the Motility Factor Autotaxin

8 74

33 149

58 224

83 299

108 374

133 449

158 524

183 599

208 674

233 749

258 824

283 899

308 974

333 1049

358 1124

383 1199

408 1274

433 1349

458 1424

483 1499

508 1574

533 1649

558 1724

583 1799

608 1874

633 1949

658 2024

683 2099

708 2174

733 2249

758 2324

783 2399

808 2474

833 2549

858 2624

883 2699

908 2774

915 2849

2924

2999

3074

3149

3224

FIG. 3. Nucleotide and deduced amino acid sequence of ATX The combined nucleotide sequence of four cDNA clones and their deduced amino acid sequence is shown. The solid tine above the amino acid sequence indicates regions that match peptides previously sequenced by enzymatic digestion and Edman degradation of purified autotaxin. Regions of peptide overlap are indicated by a heavier line.

phosphodiesteraselnucleotide pyrophosphatase. Like PC-1, ATX hydrolyzes p-nitrophenyl thymidine-5'-monophosphate, a type I phosphodiesterase substrate. This enzymatic function suggests a newly identified function for ectolexo-enzymes in cellular motility.

ATX has been previously purified from human melanoma cell conditioned medium by testing fractions of sequential chromatographic separations for their capacity to stimulate cellular motility (3). ATX was demonstrated to be a 125-kDa glycoprotein whose molecular mass reduced to 100-105 kDa after de- glycosylation with N-glycosidase F.' The calculated molecular mass of the cloned protein is 100 kDa (secreted form) or 105 kDa (full-length protein). Based on amino acid composition, the estimated PI is 9.0 which is higher than the PI determined by two-dimensional gel electrophoretic analysis (7.7-8.0) of purified ATX. This could perhaps be explained by the presence of sialic acid residues on the sugar moieties. In addition, 25 out of 30 peptides, which had been sequenced from purified ATX by Edman degradation, perfectly matched the deduced amino acid sequence of the clone; the other 5 peptides had reasonable matches. Taken together, these data provide good evidence that we have sequenced the appropriate clone for ATX.

The homology with PC-1 was unexpected but continued throughout the extracellular portion of the proteins (Fig. 5). Both have adjacent somatomedin B domains near the amino terminus of their extracellular portions. Somatomedin B, derived from the amino terminus of vitronectin, forms the presumed binding site for type 1 plasminogen activator inhibitor (26, 27). In extracellular matrix and in plasma, vitronectin is thought to be the primary binding protein of activated plasminogen activator inhibitor (26). Although PC-1 has not been

shown to bind to plasminogen activator inhibitor, this homology does suggest a kinship with extracellular matrix proteins. In addition, both ATX and PC-1 have regions homologous to the active site of bovine type I phosphodiesterase (22) and both hydrolyze the type I phosphodiesterase substrate,p-nitrophenyl thymidine-5'-monophosphate. ATX and PC-1 also both contain the loop region of an EF-hand, which is a calcium binding domain that structurally forms a helix-loop-helix (20). The loop region is the actual calcium binding site. Other proteins which lack one or both helical regions are variably capable of binding calcium (20); however, this binding has not been demonstrated for either PC-1 or ATX.

Despite the similarities, there are a number of important differences between ATX and PC-1. First, the intracellular region ofATX is only 11 amino acids long and is different from the 26 intracellular amino acids found in PC-1. Likewise, the transmembrane domains are dissimilar. In addition, PC-1 exists in the plasma membrane as a homodimer. Soluble monomeric forms of PC-1 have been demonstrated in supernatants of plas- macytoma cells, in transfected mouse L cells, and in normal mouse serum (28). However, the presumed cleavage site of PC-1, when the soluble form is generated, is between Arg16' and Ala170, i.e. between the second somatomedin B domain and the phosphodiesterase active site (28). In contrast, ATX appears to be cleaved between Ser48 and Asp4', proximal to both somatomedin B domains. Therefore, the soluble forms of the two molecules appear to be substantially different.

It remains unanswered how the domain structure of ATX affects its capacity to stimulate motility. Other ectolexo- enzymes have been shown to affect cellular motility or cellular interactions with the extracellular matrix. For example, the

cDNA Clone for the Motility Factor Autotaxin 30483

hATX

h P C l

hATX

h P C l

hATX

h P C l

hATX

hPC 1

hATX

h P C l

hATX

h P C l

hATX

hPCl

hATX

h P C l

hATX

h P C l

hATX

h P C l

JI M A R R S S F Q S C Q I I S L F * P F A V O V S I C L G F * P A H R I K R A E O W E E C P P T V L ~ D ~ P W D F 9 0

""""

I I I I I I 1 I I I I I I I I I I I I I I I I MDVGEEPLEKAARARTAKDPNTYKVLSLVLSVCVLTTIL ........ OCIFG .... LKP8CAKEVK.SCKGRCF...ERTFONCRCDMCVEU3NCCLDY 84 "

"""_. D E L C L ~ ~ G ~ T ~ ~ ~ G E E ~ A C H C S ~ ~ G ~ ~ ~ C K G E S ~ D ~ E E I ~ P A G W R P P L I I F S M O F R A ~ Y ~ I I ( K G S K V W P N I E 1 9 0

" QE~:IEPEHIWTCNKFRCGERRLTRSLCACSDDCKDRODCCINYS~VCQOEK~WVEEPCE~INEPQCP~FETPPT~LFSLDCP YLHTWOOLLWIS 184

I I I l l 1111l1l11111 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I

""""_ I I I I I I I I I I I I I I I I I I I I1 I I I I I I I I I I I I I I I I I I I I I I I I I I

Y K L R S C G T H S P Y M R P W P T K T F P N L Y T L A T G L Y P E S H G I V G N ~ D ~ D A T F H L R G R E K ~ ~ P L W I T A T K ~ T F ~ S . . . . . . . . . . . . . 272

K L K K C G m T K N M R r m Y P T K T F P N H Y S I V T G L Y P E S H G I I D N l f m D P K M N A 8 F S L K S K E K ~ P ~ K G E P I ~ ~ ~ L K S G T F ~ S ~ I N G I F P D I 284

..... WIPHERRILTILRWLTLPDHERPSVYAFYSEQPDFSGHKYGPFGPEESSYGSPPPPAltRPRRKVAPKRRQERWAPPRKR 372

YKMYNGSVPFEERILAVLQQLPKDERPHPYTLYLEEPDSSGH~GWSSE ................................................ 336

RQDKMTNPLREIDKIVGQLMDGLKQLKLRIFVGDHGMEDVTCDRTEFLSL~DITLVPOTLGRIR.SKF~.AKYDPKAIIANLTCKKPD 470

.... VIKALQRVDGWVCMLMDGLKEHRCL~ILISDHGUEQOSCKKYIYLNKYLGWKNIKVIYGPMRLRPSDVPDRIYSFNYeOIARNLSCREPN 432

QHFKPYLKQHLPKRLHYANNRRIEDIHLLVERRWHVARKPLDVYKKPSGKCFFQGDHGFDN~~~GYGPTFKYKTKVPPF~I~~L~ 570

QHFKPYLKHFLPKRLHFAKSDRIEPLTFYLDPQWQLALNPSE .. RKYCGSGF .... H G S D N V P S N H Q A L F V G Y G P G F K H G I ~ F ~ I ~ D L ~ 526

L K P A P N N G T H G S L N H L L R T N F R P T M P E E V T R P N Y P C I M Y D . E L N K R L H T K G S T E E R H L L Y G R P A V L Y R T R . Y D I L Y H T 668

LTPAPNNGTHGSLNHLLKNPTPKHPKEV.HPLVQCPFTVDRNPRD~GCSCNPSILPIEDFQTQ~L~~EKIIKH~LPYGRPR~K~1CLLSQH 625

D F E S G Y S E I F L M L L ~ S Y ~ S K Q A E V S S V P D H L T S ~ P D ~ V S P S F S Q N C L A Y K N D K Q M S Y G F L F P P Y L S S S P ~ . D ~ L ~ P ~ P ~ K R ~ 767

QFMSGYSQDILMPL~SYTVDRNDSFS..TEDFSNCLYQDFRIPLSPVHKCSFYKNNTKVmGFLSPPQLNKNSSGIYSEALLT 723

FQRVLVKKYASERNGVNVISGPIFDYDYDGLHDTEDKIKQ.. .WEGSSIPVPTHYYSIITSCLDFTQPADKCDGPLSVSSFILPHRPDNEE~SSEDE 875

FHDTLLRKY~ERNGVNWSGPVFDFDYDCRCDSLENLRQKRRVI~QEILIPTHFFIVLTSCKDTSQTPLHCPl.LDTLAFILPHRTDNSE8CVHCKHD 822

I I I l l I I I I I I l l I I I I I I l l I l l I

I I I I I I I I I I I I I1 I I I I I I I I I I I I I I I l l I I I I I

I I I I I I I I I I I I I I I I l l I l l I I I I I I I I I I I I I I I I I I I I I I I I I I I I I

I I I I I I I I I I I I I I I I I I I I I I l l I I I I I I I I I I I

I I I I I I I I I I I I I I I I I I I I I l l I I I I I I I I I I I I I I l l 1 I I I

I I I l l I I I I I I I I l l I I I l l 1 I I I I l l I l l I I I I I I I I I I I I I l l

SKWVEELMKMHTARVRDIEHLTSLDFFRKTSRSYPEILTLKTYLHTYESEI 915 I I I I I I I I I I I I I I I I I I I I I I I SSWVEELLMLHRARITDVEHITGLSFYQQRKEPVSDILKLKTHLPTFSQED 873

FIG. 4. Comparison of amino acid sequences of ATX and PC-1. The amino acid sequences of ATX and PC-1 are compared. Amino acid identity is indicated by a vertical line between the sequences. The location of the putative transmembranehignal sequence is shown by solid lines. The two somatomedin B domains are identified by dashed lines. The putative phosphodiesterase active site is indicated by emboldened lines. The loop region of a single EF-hand loop region is identified with double lines. The presumed cleavage site for each protein is indicated with arrows.

Extracellular Cleavage Slter

/ intracellular

region \ \ Phosphodleslerase Active Site

Somatomedin E domains

Tranmembrarm dmah

"Loop"

indicated for the two homologous proteins, ATX and PC-1. FIG. 5. Domain structure of ATX and PC-1. Putative domains are

cell surface glycoprotein, dipeptidyl peptidase IV, binds to fi- bronectin and affects hepatocyte interaction with the extracellular matrix (29), and amphibian pronephric duct cells have been found to require cell surface alkaline phosphatase for proper guidance of their migration pathways (30). Platelet- derived endothelial cell growth factor, which has thymidine phosphorylase activity, is both mitogenic and chemotactic for endothelial cells (31,32). However, platelet-derived endothelial cell growth factor is normally an intracellular protein that lacks a secretory signal sequence. In addition, human thymidine phosphorylase, with a 120-amino acid sequence identical to platelet-derived endothelial cell growth factor (33), also has mitogenic and chemotactic activity for endothelial cells (34).

Thus, specific intracellular enzymes have been shown to be chemotactic when placed in an extracellular environment.

If ATX is an extracellular kinase/phosphodiesterase like PC-1, its capacity to modulate protein phosphorylation could affect multiple cell surface proteins as well as components of the extracellular matrix. Such a capacity would suggest that ATX could play a regulatory role in the interaction of the mi- grating cell with its microenvironment as well as a direct role in the receptor-mediated stimulation of motility.

Acknowledgments-We thank Dr. Elliott Schiffmann for his constant moral and intellectual support. We also thank Dr. Richard E. Manrow for his help and technical suggestions in the cloning of the autotaxin cDNA.

REFERENCES 1. Boyer, B., VallBs, A,"., Tucker, G. C., DelouvBe, A., and Thiery, J. P. (1993) in

Cell Behaui0ur:Adhesion and Motility (Jones, G., Wigley, C., and Warn, R., eds) pp. 183-195, The Company of Biologists Limited, Cambridge

2. Stracke, M. L., Liotta, L. A,, and Schifiinann, E. (1993) in Cell Behauiour: Adhesion and Motility (Jones, G., Wigley, C., and Warn, R., eds) pp. 197-

3. Stracke, M. L., Krutzsch, H. C., Unsworth, E. J., Arestad, A,, Cioce, V., Schiff- 214, The Company of Biologists Limited, Cambridge

4. Kahn, E. C., Francis, E. A,, Liotta, L. A,, and Schiffmann, E. (1990) Int. J. mann, E., and Liotta, L. A. (1992) J. Biol. Chem. 267, 2524-2529

Cancer 48. 287-292 "

5. Todaro, G. J., Fryling, C., and DeLarco, J. E. (1980) Proc. Natl. Acad. Sci. ", ---

U. S. A. 77. 5258-5262 6. Liotta, L.A., Mandler, R., Murano, G., Katz, D. A,, Gordon, R. K., Chiang, P. K.,

7. Andrews, P. W., Goodfellow, P. N., and Bronson, D. L. (1983) Cold Spring and Schifhann, E. (1986) Proc. Nutl. Acad. Sci. U. S. A. 83 ,33023306

Harbor Conf. Cell Prolif. 10, 579-590

30484 cDNA Clone for the Motility Factor Autotaxin 8. Hliyhtyli, M., Thrpeenniemi-Hujanen, T., Stetler-Stevenson, W., Krutzsch, H.,

9. Wacher, M. P., Krutzsch, H. C., Liotta, L. A., and Stetler-Stevenson, W. G. Tryggvason, K., and Liotta, L. A. (1988) FEES Lett. 853, 109-113

(1990) J. Immunol. Methods 186,239-246 10. Laemmli, U. K. (1970) Nature 227,680485 11. Brown, P. D., Levy, A. T., Margulies, I. M. K. , Liotta, L. A,, and Stetler-

12. Margulies, I. M. K., Hliyhtyli, M., Evans, C., Stracke, M. L., Liotta, L. A,. and Stevenson, W. G. (1990) Cancer Res. SO, 6184-6191

Stetler-Stevenson, W. G. (1992) Cancer Epidemiology, Bwmarkers, Prevention 1,467-474

13. Young, R. A., and Davis, R. W. (1983) Proc. Natl. Amd. Sci. U. S. A. 80,1194- 1198

14. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U. S. A. 74,54634467

15. Razzell, W. E. (1963) Methods Enzymol. 6, 236-258 16. Kozak, M. (1987) J. Mol. Bid. lW, 947-950 17. Pearson, W. R., and Lipman, D. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,

2444-2448 18. Buckley, M. F., Loveland, K A., McKinstry, W. J., Garson, 0. M., and Goding,

J. W. (1990) J. Bid. Chem. 26S, 17506-17511 19. Funakoshi, I., Kato, H., Hone, K., Yano, T., Hori, Y., Kobayaahi, H., Inoue, T.,

Suzuki, H., Fukui, S., Taukahara, M., Kajii, T., and Yamashina, I. (1992)

20. Kretsinger, R. H. (1987) Cold Spring Harbor Symp. Quant. Biol. 52,499-510 Arch. Biochem. Biophys. 295, 180-187

21. Mamu, S. (1987) in Mammalian Ectoenzymes (Kenny, A. J., and Turner, A. J., eds) pp. 15-45, Elsevier Science Publishers, Amsterdam

22. Culp, J. S., Blytt, H. J., Hermodson, M., and Butler, L. G. (1985) J. Biol. Chem.

23. Rebbe, N. F., Tong, B. D., Finley, E. M., and Hickman, S. (1991) Proc. Natl.

24. Oda, Y., Kuo, M.-D., Huang, S. S., and Huang, J. S. (1991) J. Biol. Chem. 266,

25. Oda, Y., Kuo, M.-D., Huang, S. S., and Huang, J. S. (1993) J. Biol. Chem. 288,

26. Seiffert, D., and Loskutoff, D. J. (1991) J. Biol. Chem. 266,2824-2830 27. Seiffert, D., Ciambrone, G., Wagner, N. V., Binder, B. R., and Loskutoff, D. J.

28. Belli, S. I., van Driel, I. R., and Goding. J. W. (1993) Eur. J. Biochem. 217, (1994) J. Bwl. Chem. 269,2659-2666

29. Piarra, G. A,, Callanan, H. M., Mowery, J., and Hixson, D. C. (1989) Biochem. 421-428

30. Thibaudeau, G., Drawbridge, J., Dollarhide, A. W., Haque, T., and Steinberg, J. 262,327434

31. Ishikawa, F., Miyazono, K. , Hellman, U., Drexler, H., Wernstedt, C., Hagiwara, M. S. (1993) Deu. Bwl. 169,657468

IC, Usuki, K., Takaku, F., Risau, W., and Heldin, C.-H. (1989) Nature 338, 557462

32. Finnis, C., Dodsworth, N., Pollitt, C. E., Carr, G., and Sleep, D. (1993) Bur: J. Biochem. 212,201-210

33. Furukawa, T., Yoshimura, A., Sumizawa, T., Haraguchi, M., Akiyama, %I., Fukui, K., Ishizawa, M., and Yamada, Y. (1992) Nature 356,668

34. Haraguchi, M., Miyadera, K. , Uemura, K., Sumizawa, T., Furukawa, T., Yamada, K, Akiyama, SA., and Yamada, Y. (1994) Nature S88,198

260,8320-8324

Acad. Sci. U. S . A. 88,5192-5196

16791-16795

27318-27326

cdna cloning the human tumor motility-stimulating protein ... · production of anti-atx peptide...

Documents