Cathepsin L2, a Novel Human Cysteine Proteinase Produced by Breast and Colorectal Carcinomas1

[CANCERRESEARCH58,1624-1630, April 15. 19981

Advances in Brief

Ifligo Santamarla, Gloria Velasco, Maite Cazorla, Antonio Fueyo, Elias Campo, and Carlos LÃ³pez-Otmn2

Departamento de Bioquimica y Biolog(a Molecular (I .S., G. V., C. L-O.J and Biolog(a Funcional (A F], Facultad de Medicina. Universidad de Oviedo, 33006 Oviedo. and Departamento de Anatom(a Patoldgica, Hospital Clmnicoâ€”Barcelona,08036Barcelona, [M. C., E. C.!, Spain

Abstract (3); (b) cathepsin L (4); (c) cathepsin H (5); (d) cathepsin S (6); (e) cathepsin C (7); (/) cathepsin 0 (8); (g) cathepsin K (9); and (h) We have identified and cloned a new member of the papain family of cathepsin W (10). Furthermore, several groups have described the cysteine proteinases from a human brain eDNA library. The isolated existence of additional cysteine proteinases including cathepsins M, cDNA codes for a polypeptide of 334 amino acids that exhibits all of the structural features characteristic of cysteine proteinases, including the N, P. and T, which were originally identified because of their degrad active site cysteine residue essential for peptide hydrolysis. PairWise corn ing activity on specific substrates such as aldolase, collagen, prom parisons of this amino acid sequence with the remaining human cysteine sulin, or tyrosine aminotransferase, but whose characterization at the proteinases identified to date showed a high percentage of identity (78%) molecular level has not yet been reported (11â€”14). with cathepsin L; the percentage of identity with all other members of the Structural comparisons between the different members of the cys family was much lower (<40%). On the basis of these structural charac teine proteinase family have shown that they are synthesized as teristics, we have tentatively called this novel protein cathepsin L2. The preproenzymes, which are processed to the corresponding proen eDNA encoding the mature cathepsin L2 was expressed In Escherichia zymes and targeted to the lysosomes by the mannose 6-phosphate coli, and after purification, the recombinant protein was able to degrade signal attached to them. However, in some cases, the precursors of the synthetic peptide benzyloxycarbonyl-L-phenylalanyl-L-arginine-7- these lysosomal enzymes escape from this processing pathway and amido-4-rnethylcoumarln, which is commonly used as a substrate for cysteine proteinases. Cathepsin L2 proteolytic activity on this substrate continue along the secretory route, entering storage granules and wasabolishedby trans-epoxysuccinyl-L-leucylamido-(4-guanidino)butane,finally being released into the extracellular space (15). Amino acid an inhibitor of cysteine proteinases, thus providing additional evidence sequence comparisons between all members of the family have re that the isolated eDNA encodes a functional cysteine proteinase. Northern vealed that they are not closely related; the percentage of identity blot analysis of polyadenylated RNAs isolated from a variety of human between them is less than 50%. Nevertheless, in their amino acid tissues demonstrated that cathepsin L2 is predominantly expressed in the sequences, all of them contain a series of amino acids that is abso thymus and testis. This finding is in marked contrast with the wide tissue lutely conserved and essential for their catalytic activity (1). distribution of most cysteine protelnases characterized to date, including Because it seems clear that cysteine proteinases play essential roles cathepsin L, and suggests that cathepsin L2 may play a specialized role in in both normal and pathological conditions, including tumor pro the thymus and testis. Expression analysis of cathepsin L2 in human cesses, over the last few years, we have been interested in examining tumors revealed a widespread expression in colorectal and breast carci nomas but not in normal colon or mammary gland or in peritumoral the possibility that additional, uncharactenzed members of this family tissues. Cathepsin L2 was also expressed by colorectal and breast cancer of proteolytic enzymes could be produced by human tumors. This cell lines as well as by some tumors of diverse origin, including ovarian search for new human cysteine proteinases led us to identify cathepsin and renal carcinomas. These results open the possibility that this novel 0, which was originallycloned from a breastcarcinomabut is widely enzyme may be involved in tumor processes, as already reported for other distributed in human tissues (8). Furthermore, we have recently re cysteine proteinases, including cathepsin L. ported the cloning and characterization of human bleomycin hydro lase, a cytosolic cysteine proteinase that is distantly related to other Introduction members of the papain family and is involved in chemotherapy The cysteine proteinases are a family of enzymes involved in many resistance (16). In this study, we describe the molecular cloning and normal cellular processes, including the turnover of intracellular pro complete nucleotide sequence3 of a cDNA encoding a novel member teins, prohormone activation, and bone remodeling (1). In addition, it of this family of enzymes, which has tentatively been called cathepsin has been suggested that these proteolytic enzymes play important L2. We also report the expression of the cDNA in Escherichia coli roles in a number of pathological conditions such as Alzheimer's and demonstrate that recombinant cathepsin L2 is a functional cys disease, pulmonary emphysema, rheumatoid arthritis, muscular dys teine proteinase. Furthermore, we analyze its expression in human trophy, and cancer invasion and metastasis (1, 2). At present, eight tissues, showing that it is mainly expressed in the thymus and testis. human cysteine proteinases of the papain family have been isolated Finally, we present evidence that cathepsin L2 is widely expressed in and characterized at the amino acid sequence level: (a) cathepsin B colorectal and breast carcinomas.

Received I 1/21/97; accepted 2/25/98. Materials and Methods The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with Materials. cDNA libraries from human brain, mouse embryo, and mouse 18 U.S.C. Section 1734 solely to indicate this fact. brain constructed in Agtll and Northern blots containing poly(A)@ RNAs t Supported by Grant 5AF97â€”0258 from the ComisiÃ³n Interministerial de Ciencia y Tecnologla, Grant BMH4-CT96-00l7 from European Union-BIOMED II, Maraton TV3 prepared from different human tissues and cancer cell lines were purchased Cancer (to E. C.), and Glaxo-Wellcome, Spain. I. S. is a recipient of a fellowship from from Clontech (Palo Alto, CA). Restriction endonucleases and other reagents Ministerio de EducaciÃ³ny Ciencia (Spain). used for molecular cloning were from Boehringer Mannheim (Mannheim, 2 To whom requests for reprints should be addressed, at Departamento de Bio quimica y BiologIa Molecular, Facultad de Medicina, Universidad de Oviedo, 33006 Oviedo, Spain. Phone: 34-85-104201; Fax: 34-85-103564/34-85-232255; E-mail: 3 The nucleotide sequence reported in this paper has been submitted to the GenBank/ [email protected]. European Molecular Biology Laboratory Data Bank with accession number Y14734. 1624

Germany). Oligonucleotides were synthesized by the phosphoramidite method volume of PBS, lysed by using a French press, and centrifuged at 20,000 X g in an Applied Biosystems DNA synthesizer (model 392 A) and used directly for 20 mm at 4Â°C.The soluble extract was treated with glutathione-Sepharose after synthesis. Double-stranded DNA probes were radiolabeled with 4B and eluted with glutathione elution buffer [10 mMreduced glutathione in 50 [cs-32P]dCTP (3000 Ci/mmol) from Amersham (Amersham, Buckinghamshire, mM Tris-HC1 (pH 8.0)]. United Kingdom) using a commercial random-priming kit purchased from Enzyme Assays. The enzymatic activity of purified cathepsin L2 produced Pharmacia LKB Biotechnology AB (Uppsala, Sweden). Synthetic peptides in E. coli was measured by using 20 @.tMZ-Phe-Arg-AMC or Z-Arg-Arg-AMC Z-Phe-Arg-AMC4 and Z-Arg-Arg-AMC were from BACHEM, Inc. (Buben as the substrate and following the procedure described by Barrett and Kirschke dorf, Switzerland), and proteinase inhibitor E-64 was from Sigma (St. Louis, (17) with minor modifications. Assays were performed at 30Â°C in 100 mM MO). sodium acetate buffer (pH 4.5) containing 10 mM DTT and 2 mM EDTA. Screening of a Human Brain cDNA Library. After searching the Gen Substrate hydrolysis was monitored in a Cytofluor 2350 fluorometer (Milli Bank database of human ESTs for sequences with homology to human cys pore, Bedford, MA) at excitation and emission wavelengths of 360 and 460 teine proteinases, we identified a mouse sequence (AA013726; deposited by nm, respectively. For inhibition assays, the reaction mixture was preincubated M. Marra et aL, WashingtonUniversity-HowardHughesMedicalInstitute with 20 @.LME-64at 30Â°Cfor 15 mmn, and the remaining activity was deter Mouse EST Project) that, when translated, showed a significant similarity to mined using the fluorogenic substrate Z-Phe-Arg-AMC as described previ amino acid sequences previously determined for other cathepsins. This DNA ously. As a positive control of enzyme assays, we used recombinant cathepsin fragment was obtained by PCR amplification of cDNA from mouse embryo 0 producedin E. coli,as describedpreviously(8). and brain cDNA libraries. Total A-phage DNA from two commercially avail Northern Blot Analysis. Northern blots containing 2 @gofpoly(A)@ able mouse libraries (total embryo and brain) was screened for the presence of RNA from different human tissue specimens and cancer cell lines or 20 @.tg the EST using two specific primers, 5'-ATGTGGCTCVFGTFGGGCTF of total RNA isolated from paired human colorectal carcinomas and adja (primer ml) and 5'-GCCTCAGCTTCAAGACCI'T (primer m2), derived from cent normal tissue were prehybridized at 42Â°C for 3 h in 50% formamide, the AA013726 sequence. The PCR reaction was carried out in a GeneAmp 5 X saline-sodium phosphate-EDTA [1 X saline-sodium phosphate 2400 PCR system from Perkin-Elmer Corp. for 35 cycles of denaturation EDTA = 150 mMNaC1, 10 mMNaH2PO4,and 1 mMEDTA (pH 7.4)], lOX (94Â°C, 15 s), annealing (56Â°C, 15 s), and extension (72Â°C,30 s). The PCR Denhardt's solution, 2% SDS, and 100 @.tg/ml denatured herring sperm product of approximately 200 bp was phosphorylated with T4 polynucleotide DNA. After prehybrsdization, filters were hybridized with a 270-bp radio kinase and cloned into an EcoRV-cut pBS vector. The cloned cDNA was labeled probe corresponding to the 3' untranslated region of the cathepsin sequenced and found to be identical to the AA013726 sequence. This cDNA L2 cDNA obtained from digestion of the full-length cDNA with FokI and was then excised from the vector, radiolabeled, and used to screen a human EcoRI. The same filters were also hybridized with a 797-bp fragment brain cDNA library according to standard procedures. Hybridization to the corresponding to positions 550-1347 of cathepsin L cDNA (4). After radiolabeledprobewas carriedout for 18h in 6X SSC [1X SSC = 150mt@i hybridization, the filters were washed with 0.1 X SSC and 0.1% SDS for 2 h NaC1 and 15 mM sodium citrate (pH 7.0)], 5X Denhardt's solution (1 X at 50Â°Candexposed to autoradiography. RNA integrity and equal loading Denhardt's solution = 0.02% BSA, 0.02% polyvinyl pirrolidone, and 0.02% were assessed by hybridization with an actin probe. Ficoll), 0.1% SDS, and 100 pg/mi denatured herring sperm DNA at 55Â°C.The Reverse Transcription and PCR Amplification of RNA from Human membranes were washed twice for 1 h at 55Â°Cin 2X SSC and 0.1% SDS and Carcinomas and Cancer Cell Lines. Total RNA was isolated from a variety exposed to XAR-5 ifim (Kodak) at â€”70Â°Cwith intensifying screens. After of human carcinomas and breast cancer cell lines by guanidinium thiocyanate plaque purification, several cloned inserts were excised by EcoRI digestion, phenol chloroform extraction and used for cDNA synthesis with the RNA PCR and the resulting fragments were subcloned into the EcoRI site of pUC18. kit from Perkin-Elmer Corp. After reverse transcription using 4 .tg of total Nudeotide Sequence Analysis. DNA fragments selected for nucleotide RNA and random hexamers as primer, the whole mixture was used for PCR sequencing were inserted in the polylinker region of phage vector M13mp19 with two oligonucleotides specific for human cathepsin L2 (5'-CTTAAGGA and sequenced by the dideoxy chain termination method using either M13 CAGCATGTCTGGGGAA and 5'-AGTClTFGATATCATAAAGCTGTG). universal primer or cDNA-specific primers and the Sequenase Version 2.0 kit The PCR reaction was carried out in a GeneAmp 2400 PCR system (Perkin (United States Biochemical Corp., Cleveland, OH). All nucleotides were Elmer Corp.) for 40 cycles of denaturation (94Â°C, 15 s), annealing (60Â°C, identified in both strands. Sequence ambiguities were solved by substituting 15 s), and extension (72Â°C,15 s). The PCR products were analyzed in 2.5% drip fordGTPinthesequencingreactions.ComputeranalysisofDNAand agarose gels. cDNA quality was verified by performing control reactions with protein sequences was perfonned with the Genetics Computer Group software primers derived from the sequence of actin. Negative controls were also packageof the Universityof Wisconsin GeneticsComputerGroup. performed in all cases by omitting the template or reverse transcriptase. Construction ofExpression Vectors and Expression in E. coli. A 669-bp DNA fragmentcontainingthecodingsequenceformaturehumancathepsinL2 Results and Discussion was obtained by PCR amplification of the isolated full-length cDNA with primers 5'-ATGClâ€•FCCCAAATCTGTGGATFGG and 5'-TCACACATF Identification, Cloning, and Sequencing of a cDNA Encoding GGGGTAGCTGG. Because of the design of the 5' primer, the amplified Human Cathepsin L2. To identify new cysteine proteinases that fragment contained the coding information for a methionine residue at the could be produced by human tissues, we performed an extensive @ end. The PCR reaction was carried out for 30 cycles of dena search of the GenBank database of ESTs, looking for sequences with turation (95Â°C,30 s), annealing (60Â°C,30 s), and extension (72Â°C, 1 mm) homology to previously described cathepsins. This analysis led to the using the Expand Long Template PCR System (Boehringer Mannheim) to identification of a short mouse DNA fragment that, when translated, reduce error frequency. The PCR product was phosphorylated with T4 polynu generated an open reading frame with extensive sequence similarity to cleotide kinase, repaired with Klenow fragment, and ligated to expression a region found in the catalytic domain of all cysteine proteinases of vector pOEX-3X (Pharmacia LKB Biotechnology AB), which was previously the papain family characterized to date. The sequence similarities treated with SinaI and alkaline phosphatase. The resulting plasmid, called were especially relevant with mouse cathepsin L (>80%), although pGEX-3X CathL2, was transformed into E. coli strain BL21(DE3), and the transformed cells were grown in LB broth diluted 1:100 with the same medium the differences were enough to assume that the identified sequence and grown to an A@ of 1.0. Then, isopropyl-l-thio-f3-D-galactopyranoside corresponded to a new gene rather than to a polymorphic form of was added to a final concentration of 1 nms, and the incubation was continued cathepsin L. A cDNA containing part of this EST was obtained by for3 h. Cells werecollectedby centrifugation,washedandresuspendedin0.05 PCR amplificationof total A-phageDNA preparedfrom mouse em bryo and brain cDNA libraries. Positive PCR amplification results

4 The abbreviations used are: Z-Phe-Arg-AMC, benzyloxycarbonyl-L-phenylalanyl were obtained with both templates, indicating that the gene was L-arginifle-7-amido-4-methylcoumadn; Z-Arg-Arg-AMC, benzyloxycarbonyl-L-arginyl represented in both cDNA libraries. The amplified DNA fragment L-argiflifle-7-amido-4-methylcoumarin; EM, trans-epoxysuccinyl-L-leucylamido-(4- guaaidino)butane; EST, expressed sequence tag; RT-PCR, reverse transcription-PCR; (approximately 200 bp), the identity of which was confirmed by poly(A)@RNA, polyadenylated RNA. nucleotide sequencing analysis, was cloned and used as a probe to 1625

screen a human brain cDNA library. Upon screening of approximately The deduced amino acid sequence for cathepsin L2 exhibits all of 1 X 106 plaque-forming units, DNA was isolated from 15 independent the structural characteristics typical of the different cysteine protein clones showing positive hybridization with the probe, and their flu ases of the papain family (Fig. 2). Thus, the finding of a stretch of cleotide sequences were determined. An analysis of the resulting hydrophobic amino acids close to the initial methionine suggests the sequences revealed that all but one clone corresponded to human presence of the signal peptide that is also present in the remaining cathepsin L, whereas the remaining clone (called 5.1.1) had a Se cysteine proteinases of the papain family (1). Upon cleavage at quence similar to but distinct from that reported for this cysteine position 17â€”18ofthis signal sequence, after the consensus sequence proteinase (4). The sequence was also different from those derived Ala-X-Ala, the resulting protein would be composed of 317 amino from genomic clones for three cathepsin L-like genes or pseudogenes acid residues with a predicted Mr of 35,693. The alignment of cathep identified in chromosome 10 (18). The insert of the 5.1.1-positive sin L2 with the remaining cysteine proteinases also allows us to clone, of approximately 1.3 kb, was subcloned in a plasmid vector, identify a proregion and a mature proteinase sequence and to define and its complete nucleotide sequence was determined in both strands. the putative cleavage site between both protein domains. In fact, as Computer analysis of the obtained sequence (Fig. 1) revealed an open can be deduced from Fig. 2, the NH2 terminus of most mature cysteine reading frame coding a protein of 334 amino acids and a predicted Mr proteinases can be precisely aligned, because the second amino acid is of 37,369. An amino acid sequence comparison between the identified proline in all cases. By analogy with them, the active processed form amino acid sequence and those from other human cysteine proteinases of cathepsin L2 would start at the leucine residue located at position showed that the percentage of identity ranged from 78% with proca 114, with a predicted Mr of 24,130. The multiple alignment shown in thepsin L to 20% with procathepsin C. On the basis of this high degree Fig. 2 also allows us to identify the putative cathepsin L2 active site of sequence similarity with cathepsin L, we propose calling this new Cys residue (at position 138) as well as other residues proposed to be cysteine proteinase family member cathepsin L2. important in the catalytic mechanism of cysteine proteinases, includ ing the His-277 and Asn-301 residues, which would complete the catalytic triad of these enzymes (1). The amino acid sequences sur

1 60 rounding these three residues are also very well conserved in cathep sin L2 when compared with the remaining cathepsins. Thus, the 61 TTTTG5AAC@T@ATCTTTCGCTCGTCC?GGCTGCCTTTTGC??OGG&kTkGCCTCCGCT 120 WNL S LVLAATCLG IASA NH2-terminal region contains the glutamine residue (at position 132)

121 G?TCCAAAAT?TGACCAAAATTTGGATACAAAGTGGTA.CCACTGGAAGGCAACACACAGA 180 of the putative oxyanion hole as well as the tryptophan residue and the @@@ D Q N L D T K S I Q it K A T H S hydrophobic segment immediately adjacent to the cysteine active site 181 A@TTATATGCCGCGTGGSAGGATGGAGGAGSGCAGTGTGGGSAAAGAATATGMA 240 characteristic of these proteolytic enzymes. In addition, the COOH @ R L T G A N S K GW S S AVW KN K K terminal region contains a series of conserved aromatic residues and 211 A?GAT?GAACTGCACA.ATGGGGSATACAGCCAAGGG.w.cATGGcT?cACA@ATQQCCATG 300 NI I L UNGS! SQGKHGFTWAM the tripeptide Asn-Ser-Trp located in the active site. Human cathepsin L2 contains three potential sites of N-glycosylation located at posi 301 AATGCTTTTCCTGACATGACCAATGAAGAATTCAGGCAGATGATGGGTTGCTT?CGA@C 360 NATPDMTNEK VRQNNGCFRN tions 2, 221, and 292. It is noteworthy that only the glycosylation site

361. 420 at position 221 is conserved in cathepsin L. Because all mammalian Q K P R K C K V V R I P L P L D@,P KS cysteine proteinases characterized to date are glycosylated, it is pre 421 GTGG&TTGG@AA&GLAAGGCTACGTGACGCCAGTGAkGAATCAGAA&CAGTGTGGTTCT 480 sumed that at least one of these residues has attached the phosphom VD SR K K G Y VT PVKNQ KQC CS annosyl moieties required for lysosomal targeting via the mannose 481 TGTTGQGCTTTTAGTGCGACTGGTGCTCTTGAAGGACAGSTGTTCCGGAAAACTGGGSAA 540 CWAT SA T GAL S GQMFRKT GK 6-phosphate receptor system. Finally, the amino acid sequence de duced for cathepsin L2 contains six cysteine residues that are highly 541 CT?GTCTCACTGAGCGAGCAGAATCTGGTGG&CTGTTCGCGTCCTCA&GGCAATCAGGGC 600 LVS L SE QNLVD C SRPQGNQG conserved in members of the cysteine proteinase family and have been

601 TGCAATGGTGGCTTCATGGCThGGGCCTTCCAGT&TGTCAAGGAGA&CGGAGGCCTGGAC 660 found to form three disulfide bridges in papain and cathepsin B (1). In CNGG?MARA@QYVKENGGLD summary, considering all these structural characteristics together, we 661 TCTGAGGSATCCTATCCATATGTAGCAGTGGATGAAATCTGTAAGTACAGACCTG&GAAT 720 can conclude that cathepsin L2 is a novel human cysteine proteinase S S K S T P YVAVD I I C KY API N of the papain family that exhibits an amino acid sequence markedly 721 TCTGTTGCTAATG&CACTGGCTTCAC.ft.GTGGTCGCACCTGGSAAGGSGA&GGCCCTG&TG 780 SVANDTGFTVVAPGKEKALN similar to that of cathepsin L. Nevertheless, the finding of some structural differences between both enzymes opens the possibility of 781 AAAGCAGTCGCAACTGTGGGCCCCATCTCCGTTGCTATGGATGCAGGCCATTCGTCCTTC 810 KAVATVGP I SVAMDAGH SSV analyzing whether they can also differ in their enzymatic character

841 CAGTTCTACAAATCAGGCATTTATTTTGA&CCAGACTGCAGCAGCASAAACCTGGATCAT 900 istics and expression pattern in human tissues. QTYKS G I T FE PD C 8 S KNLDH Production of Cathepsin L2 In E. coli and Analysis of the 901 GGTGTTCTGGTGGTTGGCTACGGCTTTGAAGGAGCAAATTCGAATAACAGCAAGTATTGG960 EnzymatIc Activity of the PUrified Recombinant Protein. To GVLVVGT GFZ GANSNNSKTW investigate whether the isolated cathepsin L2 cDNA codes for a 961 CTCGTCAAASS.CAGCTGGGGTCCAGAATGGGGCTCGAATGGCTATGTAAMATAGCC&AA 1020 functional cysteine proteinase, we expressed the cloned cDNA in a LVK!SWGPEWGSNGTVKIAKG&CAAGSACAACCACTGTGGAATCGCCACAGCAGCCAGCTACCCCAATGTGTG&GCTG&TDKNNHC 1080 bacterial system by using the strategy described for other cysteine 1021 IATAASYPNV*GGATGGTGAGGAGGAAGGACTTA&GGA.CAGCATCTCTGGGGMATTTTATCTTGSAACTGC proteinases (19). To do that, a 669-bp fragment containing the corn

1081 1140 plete coding sequence for the predicted mature cathepsin L2 was

@2OO inserted in the polylinker region of the expression vector pGEX-3X. 1141 1260 The resulting plasmid, pGEX-3X CathL2, as well as the original 1201 vector was transformed into E. coli BL21(DE3), and the transformed 1261 CAGCTTTATGATATCAAAGACTCATTGCTThATTCTAAGA.CTTTTGAATTTTCATTTTTT 1320 bacteria were induced with isopropyl-1-thio-@3-o-galactopyranoside. 1321 AAAAAGATGTACAAAACAGTTT 1342 Extracts were then prepared from the induced bacteria, and their Fig. I. Nucleotide sequence and deduced amino acid sequence of human cathepsin L2 protein composition was analyzed by SDS-PAGE. As seen in Fig. 3A, cloned from a brain eDNA library. The amino acid sequence is shown in single-letter code the bacteria transformed with the recombinant plasmid contained a below the nucleotide sequence.The active site residues characteristic of cysteine protein ases are underlined. Arrows indicate the putative cleavage sites between the signal fusion protein of approximately 52 kDa that was not present in the sequence and the propeptide as well as between the propeptide and the mature enzyme. control extracts. Fig. 3 also shows a band of 29 kDa in the control 1626

MNLSLVLAAFMjPTLILAAFMWGLKVLLMKRÃˆVCVLLMWATL CathL2CathLCathKCathSCathHCathOCathW KFDHPLFA GANEHNRLY MAMNAFPDMTYGINQFSYLFFGVTPFSDLTYETLTLGDM.* QPEF.TEDT GMNEYGKQYKEKNEHRKTY LYVCSSAVÃ‚QLHGVPVCG|AELLLCRGGGDADALLVAGLAQGKVTTYCNETMFR.H"PPTIfH "SVNStEKFHSRAPFTPTWPIRGPL. PLLCAGAWLL.MDVR STEEYLNSLFPSENFNRSYLSPEETAHLKNSQEK* ALPWtPWLLW. RHRKEAFKLFQIQGTASENVYVNt TAF| ...MALTAHPSCLLDYKWF ..RA ...DLGTAEKSWTATTYMEKL. CathBCathCCathL2 MWQLWATGWVHBVLGRKGKVFREPLF PSFHPLSDYSHRXYKYDH*ELVNYVNKRNNFVKAINAIQLEGQMFRKTG AFFKYKEEGS

85FMIKHKYL.EEEFGQLYGY *.LDLPKSVDW. * 173QNLVDCSRPQ ..NQKFR RKK. ..G.YV TPVKNQKQCG SCWAFSATGA .-VSLSE CathL IQ. ..NRKPR KGKVFQEPLF REK. . G.YV TPVKNQGQCG SCWAFSATGA LEGQMFRKTG RL. ..ISLSE QNLVDCSGPQ CathK LKVPLSHSRS NDTLYIPEWE RKK. . G.YV TPVKNQGQCG SCWAFSSVGA LEGQLKKKTG KL. ..LNLSP QNLVDCVSE . CathS LRVP .SQWQ.R NIT .YKSNPN REK. . G.CV TEVKYQGSCG ACWAFSAVGA LEAQLKLKTG KL. ..VSLSA QNLVDCSTEK CathH WSEPQHCSAT KSNYLRGT. . Kg ...GNFV SPVKNQGACG SCWTFSTTGA LESAIAIATG KM. ..LSLAE QQLVDCA .QD CathO RSKPSKFPRY SAEVHMSIPN fc. ...KQW T2VRNQ.QMCG GCWAFSWGA VESAYAIKGK PL ...EDLSV QQVIDCS ... CathW RRAAGGVPSM GREIRSBEPE ESVPFSCDW Â«V. ..AGAIREQWPQCPTISPIKDQKNCN CCWAMAAAGN IKTLWRISFW DF. ..VDVSV QELLDCG. .. CathB K ...RLC .GT FLGGPKPPQR VMFTED....APLTAEIQQKM|KGLVSGGLYE.LKLPASFDA KEIRDQGSCG SCWAFGAVEA ISDRICIHT. NAHVSVEVSA EDLLTCCGSM CathCCathLZCathLCathKCathSCathHCathOCathWCathBGG174I ...RRS . HSRKIPRPKPMARAFQYVKEMDYAFQYVQDMTNAFQYVQKMTTAFQYIIDPSQAFEYILYTLNALNWLNKVWDAFITVLNPAEAWNFWTRILHLPTSWDWSHVGCRPYSICRNVH .GINFyi^T.oflp.FGGLDSEERGIDSEDKGIDSDAKGIMGEDQVKLVKDSSGLASEKPPCEHHVNGSSPVRNQASCG* SCYSFASMGMCKYNPKYCMYNPTGCQYDSKYCKFQPGKCHÃ•FSGSLEARIRILT.SVANKAAKRAATAIGF8GFS"r*NNSQTPILSPDTGFQE WSC ..SQ* ** ** 240KQEKALMKAVGNEKALKRAVGREDVLKEAVYDEEAMVEAVNSEKDIMAEINGYVKIAKDKGGYVKMAKDRNGYFLIERGKDGYAHVKMGKGYFRLHRG .GNQGCNGGF. ISYPYEATEESAYPYVGQEESSYPYKAMDLKTYPYQGKDGYEYPFKAQNGLDYPFQGKVRARPPCTGEGDTSYPYVAVDE GNEGCNGGL. PCRGYREIPVDI .NDGCGGGYYGNKGCNGGFFNNYGCQGGLYNNYGCNGGSRCGDGCHGGFCGD ECSKYTELP YVKDVANIT IIKGYSAYDFSIQDFIM HHRCHPKKYQPKCSKICEPGdÃ¶iKEDCFRYETVSSQSQPQNMKMZELHNGNMKMZELHNQNUCY1SIHNLNWRKINAHNNLAQAQRLQEÂ°KVAWY LQHYGYNSYSVSKNSWGPEWGSKNSWGEEWGMKNSWGENWGNKNSWGHNFGEKNSWGPQWGMRNSWGSSWGVKNSWGAQWGE .GCNGGY .SPTYKQDK CathCCathL2CathLCathKCathSCathHCathOCathWYAO..GCEGGFATVGPISVAMATVGPISVAIARVGPVSVAIANKGPVSVGVALYNPVSFAFLTFGPLWIVATYGPITVTIPYLIAGKYAQDAGHSSFQFYDAGHESFLFYDASLTSFQFYDARHPSFFLYEVTQDDKSGIYFEPDKEGIYFEPDSKGVYYDESRSGVYYEPSRTGIYSSTSLGGIIQHH.RKGVIKATPTFGLVEEADHGVLWGYGDHGVLWGYGNHAVLAVGYGNHGVLWGYGNHAVLAVGYGNHAVLITGFDDHSVLLVGFGCFPYTGTDSPFEGANSFESTEjIQDLEKY SSEYNNSKYWLVDNN

SSKNLC SSEDMC KYWLVKGNKHWIINGKEYWLVNGIPYWIVTPYWIVPPHPTPYWIL NSDNLC TQNV. .FMMYDAV .CHKTPDKVâ€¢ ...KTGSSVKSEEGIWAvjTDSQEGWRRAVWEKEGWRRAVWEKEAVRRLIWEKYHHRLQTFASHAHRLDIFAH ..SWQDYNMK. -8SGEA.Tfâ€¢â€¢â€¢ .PLQLY ..DPQLVRDPFNPFELTWYQWKATWTKWKAMWELWKKTWHLWKKTFKSWMSKRESLN . CathBCathCCathL2CathLCathKCathSCathHCathOCathWYKNGPVEGAFVHHGPMAVAF320*SVYSDFLLYEV. KSGVYQHVTKKGIYHHTGLKPRVSCPPGHAIRILGWGNHAVLLVGYGHRRLY NGTPYWLVSGMDYWIVEYSQGKHGFTEYREGKHSFTEHSMGMHSYDNGNHTFKSANSWNTDWGDKNSWGTGWGE84NGFFKILRGNGYFRIRRG YDDFLHY334

NNHCGIATAARNHCGIASAANNACGIANLAGNHCGIASFP.SYPNV....SYPTV ....SFPKM ....SYPEISYPIPLV.

NMCGLAACASNVCGIADSVSNTCGITKFP.SSIFVLTARVQKPDM.

.. CathBCathC1 QDHCGIESEVTDECAIESIACLGIASAVPKCLGIASATLTLPWSFAVAGIPRTDO.YVAATPIPKL.FDQNLDTWEKIRSREREAAAFQDLGPQPLELSLCCLLVNWACFTGKKV4

Fig. 2. Multiple amino acid sequence alignment of cathepsin L2 with other human eysteine protetnases. The amino acid sequences of previously known human cysteine proleinases were extracted from the SwissProt database, and multiple alignment was performed with the P1LEUP program from the Genetics Computer Group package. Numhering corresponds to Ihe sequence of cathepsin L2. The NH2-terminal extension characteristic of cathepsin C has not been included in the multiple alignment. *. residues that are common to all sequences. Common residues between any human cathepsin and cathepsin L2 are shaded. Gaps are indicated by dots. extracts corresponding to the parental glutathione S-transferase. Re- (phenylmethylsulfonyl fluoride), and aspartyl-proteinases (pepstatin combinant cathepsin L2 was purified by affinity chromatography A; Fig. 3ÃŸ:data not shown). Taken together, these preliminary enzy using glutathione-Sepharose 4B, leading to a single band of the matic analyses indicate that cathepsin L2 is a functional cysteine expected size as assessed by SDS-PAGE (Fig. 3A). To examine the proteinase with a substrate specificity similar to that of cathepsin L enzymatic activity of affinity-purified cathepsin L2, we investigated and an inhibitor profile characteristic of this family of proteolytic its degrading activity against fluorescent substrates commonly used enzymes. Nevertheless, it is remarkable that although cathepsin L2 for the analysis of cysteine proteinases, including Z-Phe-Arg-AMC activity against Z-Phe-Arg-AMC is similar to that obtained for re- and Z-Arg-Arg-AMC. As could be anticipated from the similarity of combinant cathepsin L produced in Â£.coli, it is considerably lower its sequence to that of cathepsin L, recombinant cathepsin L2 dis than the activity against the same substrate observed for cathepsin L played a significant proteolytic activity against the synthetic peptide purified from different tissues or produced in eukaryotic cells (20). Z-Phe-Arg-AMC. which is an optimal substrate for cathepsin L (Ref. Consequently, it is likely that the recombinant cathepsin L2 produced 17; Fig. 3ÃŸ).This proteolytic activity was also similar to that corre in bacteria is only partially functional, and therefore, additional stud sponding to human cathepsin O produced in E. coli (Ref. 8; Fig. 3ÃŸ). ies will be required to produce this novel enzyme in eukaryotic In contrast, similar to the case of cathepsin L, the activity of recom expression systems, which could result in a fully active enzyme. binant cathepsin L2 against the fluorogenic substrate Z-Arg-Arg- These studies will be also helpful in examining the putative cathepsin AMC was very low. Furthermore, the proteolytic activity of cathepsin L2 proteolytic activity against different substrates that are degraded L2 against Z-Phe-Arg-AMC was abolished by E-64, a widely used by cathepsin L as well as in searching for hypothetical differences in inhibitor of this subclass of proteolytic enzymes, but was not abol terms of substrate specificity and sensitivity to inhibitors between ished by inhibitors of metalloproteinases (EDTA), serin-proteinases these two structurally related enzymes. 1627

A U B 300 L@ @I U

@ 94

â€” @67 I $43

@30

50 @2O 0 14.4 30 60 90 120 @@ .,@ Time (miii)

Fig. 3. Purification and enzymatic analysis of recombinant cathepsin L2 produced in E. coil. A, 5-gd aliquots of bacterial extracts (pGEX-3X and pGEX-3X Cath 12) as well as I @.dofpurified fusion protein (Cat/i L2) were analyzed by SDS-PAGE. Arrows indicate parental glutathione S-transferase (29 kDa) and fusion protein (approximately 52 Wa). Right, numbers represent the sizes (in kDa) of the molecular size markers (MWM). B, recombinant cathepsin Ii (O, 0) and cathepsm 0 (U, 0) were incubated with 20 pi@iZ-Phe-Arg-AMC, and the substrate hydrolysis at 30Â°Cwasmonitored in the presence (0. 0) or in the absence (R, â€¢)of20 @e@iE-64at the indicated times.

Analysis of Cathepsin L2 Expression in Normal and Tumor and regulation of T-cell cytolytic activity (cathepsin W). In this Tissues. To analyze the expression of the cathepsin L2 gene in human context, the present fmdings showing that cathepsin L2 is predomi tissues, Northern blots containing poly(A)@ RNAs prepared from a nantly expressed in the thymus and testis strongly suggest that this number of different tissues were hybridized with a 270-bp FokVEcoPJ enzyme must be included in the growing group of nonubiquitous probe corresponding to the 3' untranslated region of the cDNA cysteine protemases that seem to play tissue-specific roles distinct encoding cathepsin L2. This Northern blot analysis revealed the from their common functions in the normal intracellular protein presence of a single transcript of approximately 1.8 kb (Fig. 4) that turnover. At present, the specific function of cathepsin L2 remains was strongly expressed in the thymus and testis. Longer exposure of speculative, but its predominant expression in thymus suggests that it membranes led to the appearance of very low levels of the same has a role related to the immune system, whereas its high level of cathepsin L2 transcript in other tissues, including the brain and intes expression in the testis suggests that it has a putative role in fertili tine. By contrast, hybridization of the same membranes with a probe zation processes, as suggested for other lysosomal enzymes (22). specific for cathepsin L confirmed the widespread distribution of this Regardless of the putative normal function ofcathepsin L2, its close enzyme in human tissues (Fig. 4). The finding of a preferential structural similarity to cathepsin L, which is also known as the major expression of cathepsin L2 in the thymus and testis is a unique feature excreted protein of malignantly transformed fibroblasts (23), indicated of this novel enzyme when compared with all human cysteine pro the possibility that this novel cysteine proteinase could be involved in teinases of the papain family described thus far. In fact, most members tumor processes as reported for cathepsin L (2, 24). As a preliminary of this gene family including cathepsins B, L, H, C, and 0 are widely step in this regard, we hybridized a Northern blot containing poly(A)@ expressed in human tissues, which is consistent with a role for these RNAs extracted from different cancer cell lines (leukemia HL-60, enzymes in the intracellular protein degradation that takes place in HeLa cell S3, leukemia K-562, leukemia MOLT-4, Burkkit's lysosomes of all cell types. Interestingly, recent findings have sug lymphoma Raji, colorectal adenocarcinoma SW480, lung carcinoma gested that in addition to their housekeeping role as lysosomal en A549, and melanoma G361) with a probe specific for cathepsin L2. zymes, some cathepsins may play highly specific roles in those tissues As seen in Fig. 4, a transcript of the expected size (approximately 1.8 in which they are overexpressed or even exclusively expressed. This kb) was detected exclusively in colorectal adenocarcinoma SW480 is the case with the recently described cathepsins K, S, and W, which cells. Fig. 4 also shows that cathepsin L is expressed in most of the are produced by osteoclasts (9, 21), lymphatic tissues (6), and T tumor cell lines that were analyzed. The finding of cathepsin L2 lymphocytes (10), respectively and have been proposed to be involved expression in colorectal cancer cells, together with its absence in in bone remodeling (cathepsin K), antigen presentation (cathepsin S), normal colon, suggests that cathepsin L2 may be linked somewhat to 1628

vÂ» tu t UL LU COLONCARCINOMAS

iÂ¡ NTNTNT kb

9.5 - - 9.5 7.5 - - 7.5

4.4 - - 4.4

2.4 - 2.4 â€¢ I 1.35 - 1.35

CathL CathL

4CTIN *CTIN

Fig. 4. Northern blot analysis of cathepsin L2 mRNA in human tissues, cancer cell lines, and colorectal carcinomas. Two jig of poly(A)* prepared from the indicated tissues and cell lines or 20 ^.g of total RNA from paired colorectal carcinomas IT")and adjacent normal tissue (N) were analysed by hybridi/ation with a 270-bp probe corresponding to the 3' end of the cDNA for human cathepsin L2. The positions of RNA si/.e markers are shown. Fillers were subsequently hybridized with a human cathepsin L probe (Calli Li and finally with a human actin probe. malignant transformation and adds a new dimension to the study of superimposable pattern with cathepsin L2, and in at least one case, this novel human cysteine proteinase. To provide more insight into cathepsin L expression was higher in normal mucosa than in its paired this observation, we examined the expression of cathepsin L2 in a tumor tissue (Fig. 4). We also tried to extend these observations to panel of paired primary colorectal carcinomas and adjacent normal tumors from other sources, such as breast carcinomas, which have mucosa. Northern blot analysis revealed that cathepsin L2 transcripts been reported to produce significant amounts of cathepsin L (2, 25). were detected in a significant number of colorectal carcinomas (6 of To do that, we performed RT-PCR analysis with RNAs obtained from 10) but not in paired adjacent normal tissue. Fig. 4 shows a repre normal resting mammary gland, breast carcinomas, and breast cancer sentative Northern blot displaying one negative case and two colo cells with different characteristics, including T-47D, ZR-75-1, rectal carcinomas expressing cathepsin L2 in the tumor tissue but not HS-578T and MDA-MB435. Oligonucleotides 5'-CTTAAGGAC- in the paired adjacent normal colonie mucosa. Interestingly, an anal AGCATGTCTGGGGAA and 5'-AGTCTTTGATATCATAAAGCT- ysis of cathepsin L expression in the same samples did not reveal a GTG were used as the cathepsin L2-specific primer pair to amplify a

Normal Breastcancer breast Breastcarcinomas cells Othercarcinomas H 8 9 10 11 12 13 14 15 16 17 N

-Actin

-Cat L2

Fig. 5. RT-PCR expression analysis of cathepsin L2 in breast carcinomas, breast cancer cell lines, and tumors from other sources. RT-PCR was performed on 4 /ig of total RNA from the samples. A 184-bp fragment corresponding to a segment of cathepsin L2 cDNA (Cal L2}was amplified in a volume of 100 /Â¿I.and20 /Â¿Iofthe reaction were separated on a 2.5% agarose gel run in Tris-borate-EDTA. A fragment of aclin was amplified in all cases as a positive control. pBR322 digested with Haelll was used as a si/e marker (M). Lanes I and 2. normal breast;Lanes 3-X. breast carcinomas;Lanes 9-12. T-47D, ZR-75-1. Hs-578T. and MDA-MB435 breast cancer cells; Lanes 13-17. renal, bladder, prostate, stomach, and ovarian carcinomas. N. negative control. 1629

184-bp fragment corresponding to nucleotides 1100â€”1283of cathep 6. Shi, G. P., Munger, J. S., Meara, J. P., Rich, D. H., and Chapman, H. A. Molecular sin L2 cDNA. As seen in Fig. 5, a band of the expected size confirmed cloning and expression of human alveolar macrophage cathepsin 5, an elastinolytic cysteine protease. J. Biol. Chem.. 267: 7258â€”7262, 1992. to be cathepsin L2 by nucleotide sequencing was amplified from most 7. Paris, A., Strukelj, B., Pungercar, J., Renko, M., Dolenc, I., and Turk, V. Molecular breast carcinomas and breast cancer cell lines (ZR-iS-i, Hs-578T, and cloning and sequence analysis of human preprocathepsin C. FEBS Left., 369: 326â€” MDA-MB435). By contrast, no amplification product was detected in 330, 1995. 8. Velasco, G., Ferrando, A. A., Puente, X. A., Sanchez, L. M., and Ldpez-OtIn, C. any sample of normal resting mammary gland or in noninvolved Human cathepsin 0: molecular cloning from a breast carcinoma, production of the breast tissues adjacent to breast carcinomas (Fig. 5; data not shown). active enzyme in Escherkhia cv@li,andexpression analysis in human tissues. 3. Biol. Finally, preliminary analysis of cathepsin L2 expression in human Chem., 269: 27136â€”27142, 1994. carcinomas from other sources revealed that some renal and ovarian 9. Inaoka, T., Bilbe, G., Ishibashi, 0., Tezuka, K., Kumegawa, M.. and Kokubo, T. Molecular cloning of human eDNA for cathepsin K: novel cysteine proteinase carcinomas also have the ability to produce this novel cysteine pro predominantly expressed in bone. Biochem. Biophys. Res. Common., 206: 89â€”96, teinase (Fig. 5). Taken together, these results strongly suggest that 1995. cathepsin L2 can be included in the growing list of proteolytic 10. Linnevers, C., Smeekens, S. P., and Bromme, D. Human cathepsin W, a putative cysteine protease predominantly expressed in CD8@ T-lymphocytes. FEBS Left., enzymes potentially involved in tumor processes. These results may 405: 253â€”259,1997. be of special interest in light of the striking similarity at both the 11. Pontremoli,S.,Melloni,E.,Salamino,F.,Sparatore,B.,Micheni,M.,andHorecker, nucleotide and amino acid sequence levels between cathepsin L2 and B. L Cathepsin M: a lysosomal proteinase with aldolase-inactivating activity. Arch. Biochem. Biophys., 214: 376â€”385,1982. cathepsin L, another cysteine proteinase produced by a number of 12. Maciewicz, R., and Etherington, D. A comparison of four cathepsins (B, L, N and S) tumors from different sources (2, 23â€”25).In fact, this structural with collagenolytic activity from rabbit spleen. Biochem. J., 25: 433â€”440, 1988. similarity strongly suggests that all previous studies could have de 13. Docherty, K., Carroll, R. J., and Steiner, D. F. Conversion of proinsulin to insulin: termined a hypothetical mixture of both proteases present in tumor involvement of a 31,500 molecular weight thiol protease. Proc. Natl. Aced. Sci. USA, 79: 4613â€”4617,1982. specimens. Consistent with this proposal, probes commonly used for 14. Gohda, E., and Pitot, H. C. A new thiol proteinase from rat liver. J. Biol. Chem., 256: expression analysis of cathepsin L in tumor tissues cross-hybridize 2567â€”2572,1981. with cathepsin L2 (data not shown). Similarly, it is tempting to 15. Sloane, B. F., Dunn, J. R., and Honn, K. V. Lysosomal cathepsin B: correlation with metastatic potential. Science (Washington DC), 212: 1151â€”1153,1981. speculate that polyclonal antibodies raised against cathepsin L and 16. Ferrando, A. A., Velasco, G., Campo, E., and LÃ³pez-Otmn,C.Cloning and expression used in most of these previous studies will likely recognize cathepsin analysis of human bleomycin hydrolase, a cysteine proteinase involved in chemo L2. The availability of specific reagents for cathepsin L2 generated in therapy resistance. Cancer Res., 56: 1746â€”1750,1996. 17. Barrett, A. J., and Kirschke, H. Cathepsin B, cathepsin H, and cathepsin L. Methods this work will be very helpful in differentiating both cysteine protein Enzymol., 80: 535â€”561,1981. ases and studying the potential role of cathepsin L2 in tumor pro 18. Bryce,S. D., Lindsay,S., Gladstone,A.J., Braithwaite,K.,Chapman,C., Spurr, cesses. N. K., and Lunec, J. A novel family ofcathepsin L-like (CFSLL) sequences on human chromosomelOqandrelatedtranscripts.Genomics,24:568â€”576,1994. Acknowledgments 19. Petanceska, S., and Dcvi, L. Sequence analysis, tissue distribution, and expression of rat cathepsin S. J. Biol. Chem., 267: 26038â€”26043, 1992. 20. Smith, S. M., and Gottesman, M. M. Activity and deletion analysis of recombinant We thank Drs. M. Balbin, J. P. Freije, and A. M. PendÃ¡sfor helpful human cathepsin L expressed in Eschthchia coli. J. Biol. Chem., 264: 20487â€”20495, comments and S. Alvarez for excellent technical assistance. 1989. 21. Gelb, B. D., Shi, G. P., Chapman, H. A., and Desnick, R. J. Pycnodysostosis, a References lysosomal disease caused by cathepsin K deficiency. Science (Washington DC), 273: 1236â€”1238,1996. 1. Bern, P. J., and Storer A. C. Alignment/phylogeny of the papain superfamily of 22. McDonald, i. K., Culbertson, J. T., and Owers, N. 0. Identification and localization cysteme proteases. J. Mol. Biol., 246: 273â€”283,1995. of a novel cathepsin S-like proteinase in guinea pig spermatozoa. Arch. Biochem. 2. Berquin, I. M., and Sloane, B. F. Cysteine proteases and tumor progression. Perspect. Drug Discov. Des., 2: 371â€”388,1994. Biophys., 305: 1â€”8,1993. 3. Chan, S. J., San Segundo, B., McCormick, M. B., and Steiner, D. F. Nucleotide and 23. Gal, S., and Gottesman, M. M. The major excreted protein of transformed fibroblasts predicted amino acid sequences of cloned human and mouse preprocathepsin B is an activable acid-protease. J. Biol. Chem., 261: 1760â€”1765,1986. cDNAs. Proc. Nati. Acad. Sci. USA, 83: 7721â€”7725,1986. 24. ChaUhan, S. S., Goldstein, L 3., and Gottesman, M. M. Expression ofcathepsin L in 4. Gal, S., and Gottesman, M. M. Isolation and sequence of a eDNA for human human tumors. Cancer Res., 51: 1478â€”1481,1991. pro-(cathepsin L). Biochem. J., 253: 303â€”306,1988. 25. Lab, T. T., Calaf, G., Kalman, E., Sinde, B. G., Somers, R., Estrada, S., Salero, E., 5. Ritonja, A., Popovic, T., Kotnik, M., Machleidt, W., and Turk, V. Amino acid Russo, J., and Daskal, I. Cathepsins D, B, and L in transformed human breast sequences ofthe human kidney cathepsins H and L. FEBS tart., 228: 341â€”345,1988. epitheial cells. Breast Cancer Res. Treat., 39: 221â€”233,1996.

1630

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1998 American Association for Cancer Research. Cathepsin L2, a Novel Human Cysteine Proteinase Produced by Breast and Colorectal Carcinomas

Iñigo Santamaría, Gloria Velasco, Maite Cazorla, et al.

Cancer Res 1998;58:1624-1630.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/8/1624

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/58/8/1624. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.