Reticulum Mainly Localized in the Endoplasmic Predominantly

Human Cathepsin W, a Cysteine Protease Predominantly Expressed in NK Cells, Is Mainly Localized in the Endoplasmic Reticulum This information is current as of September 24, 2021. Thomas Wex, Frank Bühling, Heike Wex, Dagmar Günther, Peter Malfertheiner, Ekkehard Weber and Dieter Brömme J Immunol 2001; 167:2172-2178; ; doi: 10.4049/jimmunol.167.4.2172

http://www.jimmunol.org/content/167/4/2172 Downloaded from

References This article cites 44 articles, 18 of which you can access for free at: http://www.jimmunol.org/content/167/4/2172.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 24, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Human Cathepsin W, a Cysteine Protease Predominantly Expressed in NK Cells, Is Mainly Localized in the Endoplasmic Reticulum1

Thomas Wex,2,3*† Frank Bu¨hling,2§ Heike Wex,*‡ Dagmar Gu¨nther,¶ Peter Malfertheiner,† Ekkehard Weber,¶ and Dieter Bro¨mme3*

Human cathepsin W (also called lymphopain) is a recently described papain-like cysteine protease of unknown function whose gene expression was found to be restricted to cytotoxic cells. Here we demonstrate that cathepsin W is expressed predominantly in NK cells and, to a lesser extent, in CTLs. Quantitative RT-PCR revealed that NK cells contained ϳ21 times more cathepsin W transcript than CTLs. The predominant expression of cathepsin W in NK cells was further confirmed by Western blot analysis and immunohistochemistry. IL-2-mediated stimulation of NK cells and CTLs revealed a stronger up-regulation of the cathepsin Downloaded from W gene and protein expression in NK cells (7-fold) than in CTLs (2-fold). Transfection experiments of HeLa cells and biochemical analyses revealed that cathepsin W is exclusively “high mannose-type” glycosylated and is mainly targeted to the endoplasmic reticulum (ER). Interestingly, the ER localization of cathepsin W was also found in NK cells, in which colocalization studies revealed an overlapping staining of cathepsin W and Con A, an ER-specific lectin. Furthermore, subcellular fractionation of cathepsin W-expressing cells confirmed the ER localization and showed that cathepsin W is membrane associated. Based on the results of this study, cathepsin W might represent a putative component of the ER-resident proteolytic machinery. The constitutive http://www.jimmunol.org/ expression in NK cells and the stronger up-regulation of cathepsin W by IL-2 in NK cells than CTLs suggest that cathepsin W is not just a marker of cytotoxic cells but is, rather, specifically expressed in NK cells. The Journal of Immunology, 2001, 167: 2172–2178.

atural killer cells and CD8ϩ CTLs represent two major found in both NK cells and CTLs. Here, binding of a target cell by cell populations involved in the initial immune response a CTL or NK cell stimulates a Ca2ϩ-dependent degranulation pro- N against tumors, viral infections, and allografts (1–3). cess in the effector cell that leads to a release of a “death cocktail,” However, despite the common ability of NK cells and CTLs to lyse causing the subsequent lysis of target cells (5, 8). by guest on September 24, 2021 target cells, there are differences between both cell types. Whereas The most prominent components of this death mixture are per- NK cells exist as preactivated cytotoxic cells capable of mediating forin and a family of serine proteases termed granzymes, which are their effector function without depending on presented antigenic both known to be specifically expressed in CTLs and NK cells (9). peptides (MHC class I independent) (2, 3), CTLs mostly have to be Besides the granzymes, there are also cysteine proteases involved activated by Ag-derived peptides bound to the MHC class I com- in the secretory pathway. Dipeptidyl peptidase I (synonymous with plex, and therefore, they are considered to act MHC class I de- cathepsin C) was shown to be the only enzyme capable of pro- pendently (3, 4). In general, two different cytotoxic pathways are cessing progranzymes into their active forms by removing a Glu- known (reviewed in Ref. 5). The first, the perforin-independent Glu dipeptide from the proforms (10, 11). Furthermore, the mem- 2ϩ lytic pathway, occurs in Ca -free conditions and was found pre- bers of the caspase family mediate the downstream events that dominantly in CTLs. This way is triggered through cell surface finally lead to the death of the target cell (12). Interestingly, some receptors such as Fas/Apo-1 (CD95) and TNFR (CD97) (6, 7). The other cysteine protease-related proteins such as the CTL Ag 2-␤ second, the secretory mechanism, is perforin dependent and was (13), the cysteine protease inhibitor leukocystatin (14), and cathepsin W (15) were found to be predominantly expressed in cytotoxic *Department of Human Genetics, Mount Sinai School of Medicine, New York, NY cells, suggesting specific functions of these proteins. The potential 10029; Departments of †Gastroenterology, Hepatology, and Infectious Diseases and involvement of other cysteine proteases in cytotoxic processes is ‡Pediatric Hematology and Oncology, and §Institute of Immunology, Otto-von-Gu- ericke University, Magdeburg, Germany; and ¶Medical Faculty, Institute of Physio- further supported by data showing the blocking of TCR-induced logical Chemistry, Martin Luther University Halle-Wittenberg, Wittenberg, Germany death of T cells by papain-like-specific protease inhibitors (16). Received for publication February 29, 2000. Accepted for publication June 7, 2001. Recently, the cDNA encoding a novel papain-like cysteine pro- The costs of publication of this article were defrayed in part by the payment of page tease, designated cathepsin W (also referred to as lymphopain), charges. This article must therefore be hereby marked advertisement in accordance was cloned and further characterized by our group and others (15, with 18 U.S.C. Section 1734 solely to indicate this fact. 17). Protein sequence alignments showed that this enzyme is a 1 This work was supported by a fellowship (We/2170/2-1) and in part by the SFB 387 (project B7) of the Deutsche Forschungsgemeinschaft (Germany). member of the papain superfamily, and that all major structural features such as the positions of the amino acids forming the cat- 2 T.W. and F.B. contributed equally to this work. alytic triad, the putative cleavage sites for the signal sequence, and 3 Address correspondence and reprint requests to Dr. Thomas Wex, Department of Gastroenterology, Hepatology, and Infectious Diseases, Otto-von-Guericke Univer- the propeptide are in accordance with other members of this pro- sity, Magdeburg, Leipziger Strasse 44, 39120 Magdeburg, Germany. E-mail address: tease family (15). The in vivo gene expression of cathepsin W [email protected] or Dr. Dieter Bro¨mme, Department of Hu- man Genetics, Mount Sinai School of Medicine, Box 1498, Fifth Avenue at 100th studied by Northern blot analyses revealed a high abundance of the Street, New York, NY 10029. E-mail address: [email protected] corresponding mRNA in peripheral cytotoxic cells and related cell

ϩ lines only, whereas other hematopoetic cells such as CD4 T cells, Triton X-100 (three times for 20 min), the cells were treated with anti- B cells (CD19ϩ), and monocytes (CD14ϩ) did not contain signif- mouse IgG-FITC or anti-mouse IgG-tetramethylrhodamine isothiocyanate icant amounts of cathepsin W mRNA (15, 17). (TRITC), which were both obtained from Sigma (St. Louis, MO) and used as recommended by the manufacturer. After five washing steps with PBS/T In this study, we describe the cellular and subcellular localiza- containing 0.05% Triton X-100, cells were mounted using Fluoromount-G tion and regulation of endogenous cathepsin W, which is predom- (Southern Biotechnology Associates, Birmingham, AL) and viewed with a inantly expressed in NK cells and was found to be up-regulated by fluorescence microscope (Eclipse E-800; Nikon, Melville, NY). IL-2. Furthermore, we provided evidence that endogenous cathepsin W might be mainly localized in the endoplasmic reticulum Electrophoresis and Western blot analysis 4 (ER), a novel compartment for papain-like proteases. In general, SDS-PAGE was conducted by following the method of Lae- mmli (20). Briefly, samples were heated in reducing sample buffer at 95°C Materials and Methods for 5 min and subjected to SDS-PAGE (Fisher Scientific). Electrophoresis Expression of recombinant cathepsin W was performed at 50 V for 30 min and 120 V for another 1.5 h using the Mini Gel system (Bio-Rad, Hercules, CA). Prestained protein benchmarker Previously cloned cathepsin W cDNA (15) was used as a template for PCR (Life Technologies) was used as the molecular mass standard. Proteins amplifications with Pfu polymerase (Stratagene, La Jolla, CA) using the were electroblotted onto nitrocellulose membrane (Fisher Scientific) at 35 following primers that contained either a HindIII or XbaI restriction site for Vfor3h(4°C) using the buffer system that contained 25 mM Tris, 192 cloning (primer 1, 5Ј-AAA AAG CTT ACC GGC ATG GCA CTG ACT mM glycine, and 20% (v/v) methanol. Membranes were blocked in 7.5% GCC CAC-3Ј; and primer 2, 5Ј-AAA TCT AGA TCA GGG AGG GCA nonfat milk in TBST containing 0.05% Triton X-100 (TBST/T) for 1 h. GGA GAC TCG-3Ј). In addition, a similar PCR product was obtained After rinsing the membrane twice with TBST/T, the membrane was incu- using primers 1 and 3; the latter was extended by the T7 epitope (primer 3, bated with the primary Ab for 1–2 h at room temperature or at 4°C over- 5Ј-AAA TCT AGA GCC TAT CCC ATC TGC TGT CCT CCA GTC night, washed four times with TBST/T for 15 min, and incubated with the Downloaded from ATA CTG GCC ATG GGC AGG AGA CTC GGG GCT TC-3Ј). Both corresponding secondary Ab (anti-mouse- or anti-rabbit-HRP conjugate; PCR products were subcloned into the expression vector pcDNA 3.1 (In- Fisher Scientific) at room temperature for 1 h. After four washing steps, the vitrogen, Carlsbad, CA). The insert and flanking regions of the vector were IgG was detected with the Lumi-Light chemiluminescence detection sys- sequenced in both directions using an Applied Biosystems Model 377 Au- tem (Roche Molecular Biochemicals, Indianapolis, IN) as recommended tomated Sequencer (Applied Biosystems, Foster City, CA). by the manufacturer. Preparation of PBMC and purification of CTLs and NK cells Quantitative RT-PCR analysis http://www.jimmunol.org/ Human PBMC were isolated from healthy donors by Ficoll-Hypaque (Am- Total cellular RNA was isolated using the RNeasy kit (Qiagen, ersham Pharmacia Biotech, Piscataway, NJ) density gradient centrifuga- Hildesheim, Germany); 500 ng-1 ␮g total RNA were randomly primed and tion. Subpopulations of PBMC were enriched by anti-CD4, -8, -19, -56, reverse transcribed by 100 U Moloney murine leukemia virus reverse tran- and anti-CD16 beads, respectively, using the MACS system (Miltenyi Bio- scriptase (Amersham Pharmacia Biotech) using the standard protocol rec- tec, Bergisch Gladbach, Germany) as described previously (18). NK cells ommended by the manufacturer. An aliquot of the cDNA mixture was were directly isolated from PBMC using anti-CD16 microbeads. CTLs applied to enzymatic amplifications that were performed using primers 5 were prepared by a two-step protocol. T cells were isolated from PBMC by and 6 (primer 5, 5Ј-GCG GCC TGG CCA GTG AAA AG-3Ј; and primer anti-CD3 microbeads, and this cell population was subsequently used to 6, 5Ј-GTG CAG CCG GAA ATA GCC CTT CT-3Ј). The PCR mixture isolate Th cells or CTLs by anti-CD4 or anti-CD8 microbeads, respec- contained 0.2 mM dNTP (Roche Diagnostices, Mannheim, Germany), 0.5 tively. All preparations were done according to the manufacturer’s protocol ␮M of each primer, reaction buffer with BSA containing 3 mM MgCl , by guest on September 24, 2021 system (Miltenyi Biotec). The purity of cell populations was analyzed by 2 SybrGreen (both obtained from Idaho Technology, Salt Lake City, UT), flow cytometry and was, in general, between 95 and 99% (relative cell and 1 U Goldstar DNA Polymerase (Eurogentec, Brussels, Belgium). The number) for the corresponding Ag, whereas the content of “contaminating ϩ Ϫ amplification and online monitoring of the PCR was performed using the cells” (e.g., CD16 cells in the CD16 cell population) was usually Ͻ2%. LightCycler (Idaho Technology), and an aliquot was analyzed by agarose PBMC, CTLs, or NK cells were grown in complete IMDM supple- gel electrophoresis. The initial amount of template was calculated using the mented with 10% FCS, 20 mM glutamine, 100 U penicillin, and 100 U LightCycler software. For calibration of the PCR, the PCR product was streptomycin (all obtained from Fisher Scientific, Fair Lawn, NJ). Purified cloned into the PCRII-plasmid (Invitrogen) that was transformed and ex- cells were stimulated either with 1000 U/ml IL-2 (Strathmann, Hannover, panded in Escherichia coli using standard procedures and later used as Germany) or 200 ng/ml IL-12 (R&D Systems, Minneapolis, MN) for 1–3 standard. days, respectively. Cell lines, culture, transfection, and selection of stably Subcellular fractionation of cell lysates transfected clones Cells were resuspended in 250 mM sucrose-20 mM HEPES (pH 7.2) and HeLa cells were maintained in DMEM supplemented with 20 mM glu- homogenized in an ice bath using a syringe with a 27-gauge needle. After tamine, 100 U penicillin, 100 U streptomycin (all obtained from Fisher centrifugation (500 ϫ g at 4°C), 5 ml postnuclear supernatant was mixed Scientific), and 10% FCS (Gemini Bio-Products, Calabasas, CA) at 37°C with 3 ml 80% Percoll and subjected to ultracentrifugation (68.000 ϫ g for in a humidified 5% CO2 atmosphere. HeLa cells were transfected by Li- 45 min at 4°C, fixed rotor). Finally, 18–20 fractions were collected from pofectAMINE (Life Technologies, Rockville, MD) and subsequently se- the top (light density) to the bottom (high density). An aliquot was removed lected by G418 as described earlier (19). for the analysis of the lysosomal marker enzyme (cathepsin B-like activity), which was detected using 10 ␮M Z-Arg-Arg-7-amido-4-methylcou- Immunofluorescence marin in acetate buffer (pH 5.5) as previously described (21). The fractions were stored at Ϫ80°C for subsequent Western analysis of cathepsin W and Cells were either grown on positively charged three-field slides (BioGenex the 58K-9 Golgi protein (G2404; Sigma). Laboratories, San Ramon, CA) or resuspended in PBS (pH 7.4) and then subsequently transferred onto the slides. After an incubation of 10 min, unbound cells were removed by two washing steps using PBS (pH 7.4), Results and Discussion and the attached cells were immediately fixed with fresh 4% (w/v) para- Characterization of anti-cathepsin W Abs formaldehyde (Fisher Scientific) in PBS containing 0.1% Tween 20 (pH 7.4) (PBS/T) for 10 min. After three washes with PBS/T, cells were di- A fusion protein containing prepro-cathepsin W and an N-terminal rectly used for staining or subsequently stored in PBS/T containing 0.4% located vector-encoded T7 epitope (42 kDa) was expressed in E. paraformaldehyde at 4°C for up to 4 wk. After blocking cells using PBS/T containing 10% FCS and 5% BSA for 45 min, cells were rinsed once with coli, gel-purified, and used to generate polyclonal and mAbs. Poly- PBS/T and then incubated with primary Abs for 2–4 h at room temperature clonal anti-cathepsin W Abs (R-W47/48) were further purified by or at 4°C overnight. After washing the cells with PBS/T containing 0.05% affinity chromatography using nitrocellulose-blotted recombinant E.coli-expressed cathepsin W as bait. As shown in Fig. 1B, the 4 Abbreviations used in this paper: ER, endoplasmic reticulum; PBS/T, PBS contain- E.coli-derived T7-cathepsin W fusion product was recognized by ing 0.1% Tween 20 (pH 7.4); TRITC, tetramethylrhodamine isothiocyanate. either the mAb against the T7 epitope (mAb-T7) and the purified 2174 CATHEPSIN W EXPRESSION IN NK CELLS

FIGURE 1. Characterization of recombinant cathepsin W and anti-cathepsin W Abs. A, Western blot of several human cathepsins detected either by polyclonal R-W47/48 or monoclonal M-W401B1 anti-cathepsin W Ab. Cathepsins F, K, L, S, and V were expressed in Pichia pastoris, and the cathepsin W sample was a cell lysate from HeLa-G12 cells. B, Isopropyl-␤-D-thiogalactopyranoside-induced expression of cathepsin W in E. coli BL21. Samples were harvested as indicated, separated on a 10% polyacrylamide gel, and total protein was stained by Coomassie R250. The induced 42-kDa band was identiﬁed as the cathepsin W-T7 fusion protein by the anti-T7 mAb or the R-W47/48 Ab, respectively. C, Exclusive asparagine-linked glycosylation of cathepsin W. Left panel, Puriﬁed cathepsin W protein was in vitro treated with endoglucanase H for 30 min–4 h. The deglycosylation was completed after 30 min; no differences were found in aliquots obtained after 60 min, and the same pattern was obtained using a mixture of endoglucanase H and F (data not shown).

Right panel, HeLa cells transiently transfected with the cathepsin W cDNA and control cells were incubated for 24 h without (Ϫ) and in the presence of Downloaded from 10 ␮g/ml tunicamycin (T). The cathepsin W was detected by Western blot analysis using the Ab R-W47/48. anti-cathepsin W Abs (R-W47/48), respectively, whereas the pre- COS-7 cells (19). All together, the Ab characterization proved that serum did not show speciﬁc signals. Both polyclonal and mono- both the polyclonal Ab R-W47/48 and the mAb M-W401B1 are

clonal anti-cathepsin W Abs were analyzed regarding their poten- equally suitable for analyzing cathepsin W expression. http://www.jimmunol.org/ tial cross-reactivity toward other human cathepsins. Only one of nine monoclonal supernatants tested revealed a cross-reactivity to- Characterization of endogenous cathepsin W gene and protein ward human cathepsin L (data not shown). Neither the purified expression in NK cells and CTLs and induction of cathepsin W polyclonal rabbit Abs (R-W47/48) nor the hybridoma clone by IL-2 M-W401B1 revealed any cross-reactivity against human cathep- Previous Northern blot analyses revealed a cytotoxic cell-specific sins F, K, S, L, and V, but both of them strongly recognized re- gene expression of cathepsin W (15, 17). To characterize the ex- combinant cathepsin W (Fig. 1A). To analyze the specificity of the pression of cathepsin W in more detail, we used magnetic sepa- anti-cathepsin W Abs in human cells, cathepsin W expression was ration techniques to obtain leukocyte subpopulations from PBMC analyzed in cathepsin W cDNA-transfected HeLa cells. As shown samples. RT-PCR analysis confirmed the killer cell-specific ca- by guest on September 24, 2021 in Fig. 1C, the Ab R-W47/48 detected a major double band at thepsin W gene expression (partially presented in Fig. 2A). Quan- 50–55 kDa as well as a weak minor band at 42 kDa in HeLa cells titative RT-PCR revealed that NK cells (CD16ϩ) contained an transfected with the cathepsin W cDNA. The same results were average of ϳ21 times more cathepsin W transcript than CTLs obtained using monoclonal anti-cathepsin W Abs (data not (CD3ϩCD8ϩ cells), whereas Th cells (CD3ϩCD4ϩ cells) con- shown). When stably transfected HeLa cells were subjected to im- tained only traces of cathepsin W mRNA (Fig. 2B). Notably, the munofluorescence using R-W47/48 and M-W401B1, a reticular distribution of cathepsin W transcript and its concentration varied staining was observed, suggesting the staining of the ER compart- remarkably among different probands (e.g., 3–25 times enrichment ment as described previously in transiently transfected HeLa or of cathepsin W transcript in CD16ϩ vs CD16Ϫ cells). To confirm

FIGURE 2. Detection of endogenous cathepsin W and quantitative RT-PCR analysis of T cells and NK cells. A, PBMC were either enriched or depleted using anti-CD4, -8, -19, or -56 Abs, respectively, and then subjected to RT-PCR and cytofluorometric analysis. To prove the integrity of the total RNA, the ␤-actin transcript was detected in all samples. M, Molecular mass marker. B, Quantitative RT-PCR analysis was performed using total RNA from cells as indicated (n ϭ 4–8). On average, NK cells (CD16ϩ), CTLs (CD3ϩCD8ϩ) or Th cells (CD3ϩCD4ϩ) contained ϳ3.2 ϫ 106 copies, 1.2 ϫ 105 copies, or 2.8 ϫ 103 copies per micorgram of total RNA, respectively. C, PBMC or subpopulations were isolated as described in Materials and Methods and subjected to Western blot analysis using the Ab R-W47/48 (PBMC) or M-W401B1 (subpopulations). The left panel documents the presence of endogenous cathepsin W in three different PBMC samples, whereas the distribution of cathepsin W in NK cells vs different T cell subpopulations from two probands is presented on the right panel. The Journal of Immunology 2175 the gene expression studies, the endogenous cathepsin W protein hypothesis that cathepsin W is not just a marker of cytotoxic lym- expression was analyzed by Western blot analysis and immuno- phocytes but is, rather, specifically expressed in NK cells. histochemistry. As shown in Fig. 2C, three independent PBMC Although IL-12 in general is capable of stimulating NK cells in samples tested, as well as isolated NK cells and CTLs, contained vitro (22), in similar stimulation experiments, it did not alter the endogenous cathepsin W that exhibited a protein pattern in SDS- expression levels of either the cathepsin W mRNA or the protein PAGE similar to that of the HeLa-derived transfectants (Fig. 1). (data not shown). The up-regulation of cathepsin W by IL-2 may Generally, the Western blot analysis confirmed quantitative RT- be mediated by the intermediate affinity IL-2R, whose ␤ and ␥ PCR analyses by showing that NK cells contain more cathepsin W subunits are known to be constitutively expressed on NK cells and protein than CTLs. CTLs. The binding of IL-2 to its receptor triggers downstream Furthermore, endogenous cathepsin W was detected in immu- events organized in three major pathways, the ras-signaling path- nohistochemical studies. Purified PBMC were simultaneously way, the phosphatidylinositol-3 kinase pathway, and the STAT/Jak stained with anti-CD16 and anti-cathepsin W Abs. The double- pathway, that finally result in transcriptional activation and cellular staining experiments revealed an almost complete overlap between proliferation (23, 24). Cathepsin W is not the only protein in cy- the expression of CD16 and cathepsin W (Fig. 3A–C), indicating totoxic cells whose transcript level is up-regulated by IL-2. Similar that NK cells represent the major population of cathepsin W-ex- results were described for the dipeptidyl peptidase I (25) and the pressing cells among human leukocytes. NK lytic-associated molecule (26), which are both known to be Although NK cells are known to be preactivated, the cytolytic critical for the perforin-dependent pathway in the lysis of cells. activity of NK cells can be further enhanced by stimuli such as Obviously, the cathepsin W gene expression is regulated on the IL-2 or IL-12 (22). To study the gene regulation of the cathepsin transcriptional level, but the underlying mechanisms for its regu- Downloaded from W gene in response to these cytokines, transcript and protein levels lation still need to be elucidated. Future studies, particularly pro- of cathepsin W were analyzed in NK cells stimulated either by moter studies and those that assess cathepsin W in different sub- IL-2 or IL-12 and in CTLs stimulated with IL-2. In NK cells, a populations of cytotoxic cells during the course of the cytotoxic 12.8-fold induction of the cathepsin W mRNA level ranging from attack, will shed light on the transcriptional regulation of the ca- 3.3-fold to 34-fold (n ϭ 8) was detected in the presence of IL-2 thepsin W gene.

(Fig. 4A). The elevated transcript levels in IL-2-stimulated NK http://www.jimmunol.org/ cells were accompanied by the increase of the protein level as well Biochemical characterization of cathepsin W (Fig. 4B). IL-2-stimulated CTLs revealed a 2.1-fold increase of Western blot analyses of HeLa-derived transfectants and immune their cathepsin W mRNA content (1.3- to 6.4-fold; n ϭ 3) and cells revealed an identical protein pattern (Figs. 1, 2, and 4). In- slightly elevated protein levels (Fig. 4B). Although Th cells terestingly, the molecular mass of the major double band (50–55 (CD3ϩCD4ϩ) showed a similar relative increase of cathepsin W kDa) was higher than the 42.1 kDa calculated from the cDNA mRNA levels at much lower concentrations, no clear signal was sequence and detected in the in vitro transcription/translation assay detected in Western blot analysis (Fig. 4B), suggesting that the (17). Sequence analyses revealed two potential N-linked glycosyl- traces of cathepsin W mRNA in these cells probably results from ation sites (Asn78 and Asn156) and three putative O-linked glyco- contaminating CTLs, which might be present in the isolated sylation sites (Ser313, Ser316, and Thr324). To study the N-linked by guest on September 24, 2021 CD3ϩCD4ϩ cells. The fact that NK cells contained signiﬁcantly glycosylation, transfected HeLa cells were cultivated in the pres- more cathepsin W transcript and protein than CTLs supports the ence of tunicamycin, which prevents any N-linked glycosylation.

FIGURE 3. Immunoﬂuorescence localization of endogenous cathepsin W in CD16ϩ cells. Indirect immunoﬂuorescence detection of cathepsin W was performed using the Ab R-W47/48 as described in Materials and Methods. The colocalization of cathepsin W and CD16 in PBMC (C) was shown by using anti-CD16-FITC (A) and anti-cathepsin W detected by anti-rabbit-TRITC (B). The second panel documents the colocalization of cathepsin W detected by anti-rabbit-FITC (D) and the ER (E) detected by rhodamine-succinylated Con A (Vector Laboratories, Burlingame, CA). Separate images were taken in the corresponding channels and later merged using Adobe Photoshop software (Adobe Systems, San Jose, CA) (C and F). The lower panel represents the corresponding negative controls. G, Anti-rabbit-TRITC; H, anti-rabbit-FITC; and I, anti-mouse-FITC. 2176 CATHEPSIN W EXPRESSION IN NK CELLS

FIGURE 4. Induction of human cathepsin W in NK cells and CTL by IL-2. A, A strong increase of the cathepsin W mRNA level was observed in NK cells (12.8-fold; n ϭ 8), whereas CTLs exhibited an increase of 2.1-fold (n ϭ 3) and Th cells of 1.7-fold (n ϭ 3). Cells were incubated with 1.000 U/ml IL-2 for 2 days (left panel). Quantitative RT-PCR analysis was performed using total RNA from NK cells activated with different concentrations of IL-2 for the time periods indicated (right panel). B, The induction of cathepsin W transcript levels (1.000 U/ml, 2 days) was accompanied by elevated protein levels in NK cells and CTLs (Ab M-W401B1), as shown by three independent experiments, whereas no cathepsin W was detected in Th cells (two experiments). Downloaded from http://www.jimmunol.org/ Furthermore, purified recombinant cathepsin W was treated with or autoactivation at acidic pH, and the enzymatic activities of ma- either endoglucanase H alone or with a mixture of endoglucanase ture cathepsins are determined by the cleavage of peptide-derived H and F, which that both remove different types of glycosylation substrates (21). But so far, similar studies performed on recombi- from asparagine residues. As presented in Fig. 1C, these experi- nant cathepsin W have failed to reveal the proteolytic character- ments revealed a similar shift of 3–4 kDa in the molecular mass of istics of cathepsin W. Considering the unique structural features of cathepsin W, suggesting the usage of both putative N-linked gly- cathepsin W, it might be possible that the requirements for its cosylation sites and that only high mannose-type glycosylation is enzymatic activity, such as pH and substrate specificity, com- present. Because it is known that O-linked glycosylation is less pletely differ from those of other cathepsins, as it was shown re- complex (27), the usage of all potential N- and O-linked glyco- cently for cathepsin X (Z). Cathepsin X exhibits a unique car- by guest on September 24, 2021 sylation sites might account for just 4–6 kDa, resulting in molec- boxypeptidase activity, but only a very weak endopeptidase ular mass of ϳ46–48 kDa. It is known that some cathepsins ex- activity (28). hibit a slightly higher apparent molecular mass in Western blot analysis (E. Weber, Martin Luther University Halle-Wittenberg, Subcellular localization of cathepsin W Wittenberg, Germany, unpublished observation). Currently, it is Taking into consideration the ER-like distribution of cathepsin W not known whether the higher apparent molecular mass of cathep- in the HeLa-derived transfectants (19), the subcellular localization sin W is caused by similar “electrophoretic artifacts” or whether it of cathepsin W was further studied. Keeping in mind that HeLa reflects the presence of additional posttranslational modifications cells do not usually express cathepsin W, it is notable that other in the protein. cathepsins, such as cathepsin F, expressed in HeLa cells were lo- Because the cathepsin W cDNA-encoded protein exhibits all calized in vesicles (21). Therefore, it seems unlikely that the ER- structural features of an active cysteine protease such as the cata- like distribution of cathepsin W in transfected HeLa cells is just an lytic triad (Cys, His, Asn), recombinant cathepsin W was analyzed artifact. To study the subcellular localization of cathepsin W in regarding its enzymatic activity. Similar to other cathepsins, ca- immune cells, PBMC were double-stained with Con A and anti- thepsin W is translated as a zymogen and requires the proteolytic cathepsin W Abs. As shown in Fig. 3, D–F, the majority of ca- processing of the polypeptide for gaining enzymatic activity. In thepsin W protein was found to be colocalized with the ER marker general, cathepsin propeptides can be removed by pepsin treatment Con A, a plant-derived lectin that preferentially binds ER-resident

FIGURE 5. Distribution of cathepsin W by subcellular fractionation. Western blot analysis of subcellular cell fractions (for de- tails see Materials and Methods). Cathepsin W was detected using the R-W47/48 Ab. The distribution of the 58K9 protein exclusively localized in the Golgi was analyzed using the mAb 58K-9 (G2404; Sigma). As a lysosomal marker, cathepsin B-like activity was detected and found to be predominantly present in the high density fractions. The distribution of the 58K-9 and the cathepsin B was similar for all experiments. The Journal of Immunology 2177 high mannose-type-glycosylated proteins (29). To confirm the ER- insight into the regulation of ER-resident proteolysis in NK cells like distribution of cathepsin W with an alternative method, sub- and will enable us to develop a functional model for cathepsin W. cellular fractionation of cathepsin W-containing cells was performed in a self-generating Percoll gradient. As shown in Fig. 5, Acknowledgments cathepsin W was found in the lighter fractions of the gradient of We are very grateful to Ruth Hilde Ha¨dicke, Marianne Blichmann (Otto- transfected HeLa cells as well as in two PBMC samples, whereas von-Guericke University, Magdeburg, Germany), and Rita Medek (Martin wild-type HeLa cells did not reveal any signal. To assess the qual- Luther University Halle-Wittenberg) for their excellent technical ity of the subcellular fractionation, the distribution of Golgi marker assistance. 58K-9 (30) and cathepsin B as a lysosomal marker (31) were studied. The 58K-protein was predominantly found in fractions 7–11, References whereas the cathepsin B-like activity was mostly identified in the 1. Herberman, R. B., J. Djeu, H. D. Kay, J. R. Ortaldo, C. Riccardi, G. D. Bonnard, H. T. Holden, R. Fagnani, A. Santoni, and P. Puccetti. 1979. Natural killer cells: high density fractions (fractions 16–18) of the gradient (Fig. 5). characteristics and regulation of activity. Immunol. Rev. 44:43. Notably, these fractions did not contain any or contained a very 2. Lotzova, E. 1993. Definition and functions of natural killer cells. Nat. Immun. little amount of cathepsin W. To prove the membrane association 12:169. 3. Berke, G. 1995. Unlocking the secrets of CTL and NK cells. Immunol. Today of cathepsin W, postnuclear supernatants were subjected to ultra- 16:343. centrifugation (100.000 ϫ g for 2 h), and the particulate fraction 4. Kos, F. J., and E. G. Engleman. 1996. Immune regulation: a critical link between NK cells and CTLs. Immunol. Today 17:174. containing all membrane components and the cytosolic superna- 5. Page, L. J., A. J. Darmon, R. Uellner, and G. M. Griffiths. 1998. L is for lytic tant were recovered. Subsequent Western blot analysis revealed an granules: lysosomes that kill. Biochim. Biophys. Acta 1401:146. almost exclusive signal in the particulate fraction, whereas the su- 6. Kagi, D., F. Vignaux, B. Ledermann, K. Burki, V. Depraetere, S. Nagata, Downloaded from H. Hengartner, and P. Golstein. 1994. Fas and perforin pathways as major mech- pernatant contained only traces of cathepsin W (data not shown). anisms of T cell-mediated cytotoxicity. Science 265:528. Because ER-resident proteins have to be actively retained from 7. Strasser, A., A. W. Harris, D. C. Huang, P. H. Krammer, and S. Cory. 1995. Bcl-2 further transportation or retrieved from the Golgi compartment, the and Fas/APO-1 regulate distinct pathways to lymphocyte apoptosis. EMBO J. 14:6136. cathepsin W cDNA was screened for known protein signals me- 8. Henkart, P. A. 1994. Lymphocyte-mediated cytotoxicity: two pathways and mul- diating ER retention, but none of these motifs, such as the KDEL tiple effector molecules. Immunity 1:343. 9. Jenne, D. E., and J. Tschopp. 1988. Granzymes, a family of serine proteases motif (32), KDEL-like motifs (33), C-terminal cysteine residues http://www.jimmunol.org/ released from granules of cytolytic T lymphocytes upon T cell receptor stimu- (34, 35), basic dipeptides within KKXX motifs (36), or internal lation. Immunol. Rev. 103:53. signal sequences (37) were detected. Because, cathepsin W con- 10. McGuire, M. J., P. E. Lipsky, and D. L. Thiele. 1993. Generation of active myeloid and lymphoid granule serine proteases requires processing by the gran- tains two unique motifs, a 21-aa insertion and an 8-aa C-terminal ule thiol protease dipeptidyl peptidase I. J. Biol. Chem. 268:2458. extension, which were not found in any other cathepsin (15), it will 11. Pham, C. T., and T. J. Ley. 1999. Dipeptidyl peptidase I is required for the be necessary to characterize these unique motifs regarding their processing and activation of granzymes A and B in vivo. Proc. Natl. Acad. Sci. USA 96:8627. ability to mediate targeting of cathepsin W to the ER compartment. 12. Zhivotovsky, B., D. H. Burgess, and S. Orrenius. 1996. Proteases in apoptosis. Based on the results of the subcellular localization and the almost- Experientia 52:968. exclusive presence of cathepsin W in the particulate fraction, a 13. Delaria, K., L. Fiorentino, L. Wallace, P. Tamburini, E. Brownell, and D. Muller.

1994. Inhibition of cathepsin L-like cysteine proteases by cytotoxic T-lympho- by guest on September 24, 2021 potential membrane attachment of cathepsin W would be an in- cyte antigen-2␤. J. Biol. Chem. 269:25172. teresting hypothesis that would explain the higher apparent mo- 14. Halfon, S., J. Ford, J. Foster, L. Dowling, L. Lucian, M. Sterling, Y. Xu, M. Weiss, M. Ikeda, D. Liggett, et al. 1998. Leukocystatin, a new class II cystatin lecular mass of cathepsin W in the SDS-PAGE and possibly rep- expressed selectively by hematopoietic cells. J. Biol. Chem. 273:16400. resent the mechanism for the ER retention. Current studies are 15. Linnevers, C., S. P. Smeekens, and D. Bromme. 1997. Human cathepsin W, a ϩ aimed to determine whether cathepsin W is attached to the mem- putative cysteine protease predominantly expressed in CD8 T-lymphocytes. FEBS Lett. 405:253. brane and to elucidate the underlying mechanism of its attachment. 16. Sarin, A., D. H. Adams, and P. A. Henkart. 1993. Protease inhibitors selectively block T cell receptor-triggered programmed cell death in a murine T cell hybrid- Functional implications oma and activated peripheral T cells. J. Exp. Med. 178:1693. 17. Brown, J., E. Matutes, A. Singleton, C. Price, H. Molgaard, D. Buttle, and Based on the predominant ER localization, cathepsin W might play T. Enver. 1998. Lymphopain, a cytotoxic T and natural killer cell-associated a functional role in the proteolytic machinery of this compartment. cysteine proteinase. Leukemia 12:1771. 18. Buhling, F., D. Kunz, D. Reinhold, A. J. Ulmer, M. Ernst, H. D. Flad, and For some years, it has been known that the ER contains a variety S. Ansorge. 1994. Expression and functional role of dipeptidyl peptidase IV of proteolytic enzymes including cysteine proteases (ER-60) either (CD26) on human natural killer cells. Nat. Immun. 13:270. involved in the ER-associated protein degradation or the process- 19. Wex, T., B. Levy, S. P. Smeekens, S. Ansorge, R. J. Desnick, and D. Bromme. 1998. Genomic structure, chromosomal localization, and expression of human ing of polypeptides by “limited proteolysis” (38–41). In addition, cathepsin W. Biochem. Biophys. Res. Commun. 248:255. several studies provided evidence for regulatory functions of ER- 20. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the resident proteases in the processing of other proteins. The signal head of bacteriophage T4. Nature 227:680. 21. Wang, B., G. P. Shi, P. M. Yao, Z. Li, H. A. Chapman, and D. Bromme. 1998. peptidase located in the ER membrane is the sole enzyme respon- Human cathepsin F: molecular cloning, functional expression, tissue localization, sible for cleaving off the signal peptide from all proteins translated and enzymatic characterization. J. Biol. Chem. 273:32000. 22. Uharek, L., M. Zeis, B. Glass, J. Steinmann, P. Dreger, W. Gassmann, into the ER (42). The yeast protease Ste24p and presenilin 1 are N. Schmitz, and W. Muller-Ruchholtz. 1996. High lytic activity against human other ER-resident proteases involved in the processing of C-ter- leukemia cells after activation of allogeneic NK cells by IL-12 and IL-2. Leuke- minal CAAX motif and the amyloid precursor protein, respectively mia 10:1758. 23. Waldmann, T., Y. Tagaya, and R. Bamford. 1998. Interleukin-2, interleukin-15, (43, 44). Considering the predominant expression of cathepsin W and their receptors. Int. Rev. Immunol. 16:205. in NK cells and its subcellular localization in terms of a potential 24. Karnitz, L. M., and R. T. Abraham. 1996. Interleukin-2 receptor signaling mech- involvement in cytotoxic processes, it is tempting to assume that anisms. Adv. Immunol. 61:147. 25. Pham, C. N., R. J. Armstrong, D. B. Zimonjic, N. C. Popescu, D. G. Payan, and its substrates should be NK cell specific. Studies focused on the T. J. Ley. 1997. Molecular cloning, chromosomal localization, and expression of enzymatic activity of cathepsin W are currently underway. So far, murine dipeptidyl peptidase I. J. Biol. Chem. 272:10695. cathepsin W can be characterized as a cysteine protease that might 26. Kozlowski, M., J. Schorey, T. Portis, V. Grigoriev, and J. Kornbluth. 1999. NK lytic-associated molecule: a novel gene selectively expressed in cells with cyto- be a component of the ER-resident proteolytic machinery that is lytic function. J. Immunol. 163:1775. involved in the processing of NK (CTL) cell-specific proteins. The 27. Gadroy, P., M. Stridsberg, C. Capon, J. C. Michalski, J. M. Strub, identification of these substrates as well as the full characterization A. Van Dorsselaer, D. Aunis, and M. H. Metz-Boutigue. 1998. Phosphorylation and O-glycosylation sites of human chromogranin A (CGA79–439) from urine of of the intrinsic enzymatic activity of cathepsin W will give more patients with carcinoid tumors. J. Biol. Chem. 273:34087. 2178 CATHEPSIN W EXPRESSION IN NK CELLS

28. Nagler, D. K., R. Zhang, W. Tam, T. Sulea, E. O. Purisima, and R. Menard. 1999. 37. Ouzzine, M., J. Magdalou, B. Burchell, and S. Fournel-Gigleux. 1999. An inter- Human cathepsin X: a cysteine protease with unique carboxypeptidase activity. nal signal sequence mediates the targeting and retention of the human UDP- Biochemistry 38:12648. glucuronosyltransferase 1A6 to the endoplasmic reticulum. J. Biol. Chem. 274: 29. Atoji, Y., Y. Hori, Y. Suzuki, and M. Sugimura. 1988. Concanavalin A binding 31401. sites of rough endoplasmic reticulum containing intracisternal microtubules in 38. Brodsky, J. L., and A. A. McCracken. 1999. ER protein quality control and canine neurones. J. Comp. Pathol. 99:55. proteasome-mediated protein degradation. Semin. Cell Dev. Biol. 10:507. 30. Bloom, G. S., and T. A. Brashear. 1989. A novel 58-kDa protein associates with 39. Gardner, A. M., S. Aviel, and Y. Argon. 1993. Rapid degradation of an unas- the Golgi apparatus and microtubules. J. Biol. Chem. 264:16083. sembled immunoglobulin light chain is mediated by a serine protease and occurs 31. Kirschke, H., A. J. Barrett, and N. D. Rawlings. 1995. Proteinases 1: lysosomal in a pre-Golgi compartment. J. Biol. Chem. 268:25940. cysteine proteinases. Protein Proﬁle 2:1581. 40. Otsu, M., R. Urade, M. Kito, F. Omura, and M. Kikuchi. 1995. A possible role of ER-60 protease in the degradation of misfolded proteins in the endoplasmic 32. Pelham, H. R. 1990. The retention signal for soluble proteins of the endoplasmic reticulum. J. Biol. Chem. 270:14958. reticulum. Trends Biochem. Sci. 15:483. 41. Loo, T. W., and D. M. Clarke. 1998. Quality control by proteases in the endo- 33. Andres, D. A., I. M. Dickerson, and J. E. Dixon. 1990. Variants of the carboxyl- plasmic reticulum: removal of a protease-sensitive site enhances expression of terminal KDEL sequence direct intracellular retention. J. Biol. Chem. 265:5952. human P-glycoprotein. J. Biol. Chem. 273:32373. 34. Fra, A. M., C. Fagioli, D. Finazzi, R. Sitia, and C. M. Alberini. 1993. Quality 42. Dalbey, R. E., M. O. Lively, S. Bron, and J.M. van Dijl. 1997. The chemistry and control of ER synthesized proteins: an exposed thiol group as a three-way switch enzymology of the type I signal peptidases. Protein Sci. 6:1129. mediating assembly, retention and degradation. EMBO J. 12:4755. 43. Schmidt, W. K., A. Tam, K. Fujimura-Kamada, and S. Michaelis. 1998. Endo- 35. Isidoro, C., C. Maggioni, M. Demoz, A. Pizzagalli, A. M. Fra, and R. Sitia. 1996. plasmic reticulum membrane localization of Rce1p and Ste24p, yeast proteases Exposed thiols confer localization in the endoplasmic reticulum by retention involved in carboxyl-terminal CAAX protein processing and amino-terminal a- rather than retrieval. J. Biol. Chem. 271:26138. factor cleavage. Proc. Natl. Acad. Sci. USA 95:11175. 36. Townsley, F. M., and H. R. Pelham. 1994. The KKXX signal mediates retrieval 44. Annaert, W., and B. De Strooper. 1999. Presenilins: molecular switches between of membrane proteins from the Golgi to the ER in yeast. Eur. J. Cell Biol. 64:211. proteolysis and signal transduction. Trends Neurosci. 22:439. Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021