CD30 Shedding from Karpas 299 Lymphoma Cells Is Mediated by TNF- Α-Converting Enzyme

CD30 Shedding from Karpas 299 Lymphoma Cells Is Mediated by TNF- α-Converting Enzyme

This information is current as Hinrich P. Hansen, Sebastian Dietrich, Tatiana Kisseleva, of September 25, 2021. Thilo Mokros, Rolf Mentlein, Hans H. Lange, Gillian Murphy and Hilmar Lemke J Immunol 2000; 165:6703-6709; ; doi: 10.4049/jimmunol.165.12.6703

http://www.jimmunol.org/content/165/12/6703 Downloaded from

References This article cites 45 articles, 15 of which you can access for free at: http://www.jimmunol.org/content/165/12/6703.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 25, 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 © 2000 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. CD30 Shedding from Karpas 299 Lymphoma Cells Is Mediated by TNF-␣-Converting Enzyme1

Hinrich P. Hansen,2* Sebastian Dietrich,* Tatiana Kisseleva,3* Thilo Mokros,* Rolf Mentlein,† Hans H. Lange,* Gillian Murphy,‡ and Hilmar Lemke*

CD30 is a costimulatory receptor on activated lymphocytes and a number of human lymphoma cells. Specific ligation of membrane-bound CD30 or cellular stimulation by PMA results in a metalloproteinase-mediated down-regulation of CD30 and release of its soluble ectodomain (sCD30). In this report, it is demonstrated that PMA-induced CD30 cleavage from Karpas 299 cells was mediated by a membrane-anchored metalloproteinase which was active on intact cells following 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate extraction of membrane preparations. Moreover, CD30 shedding was blocked by the synthetic hydroxamic acid-based metalloproteinase inhibitor BB-2116 (IC50, 230 nM) and the natural tissue inhibitor of metalloproteinases Downloaded from (TIMP)-3 (IC50, 30 nM), but not by the matrix metalloproteinase inhibitors TIMP-1 and TIMP-2. This inhibition profile is similar to that of the TNF-␣- converting enzyme (TACE) and, indeed, mRNA transcripts of the membrane-bound metalloproteinase- disintegrin TACE could be detected in Karpas 299 cells. The ectodomain of TACE was expressed in bacteria as a GST fusion protein (GST-TACE) which cleaved CD30 from the surface of Karpas 299 cells and concomitantly increased the level of sCD30 in the cell supernatants. Hence, TACE does not only control the release of TNF-␣, but also that of sCD30. The Journal of Immunology, 2000, 165: 6703–6709. http://www.jimmunol.org/ he CD30 is a receptor that has been detected on a subset TNFRs, CD30, CD43, CD44, and the IL-6R, the endoproteolytic of activated T and B cells and a variety of human lym- conversion of membrane-anchored proteins is blocked by broad- T phomas (1, 2). Cloning of its cDNA characterized CD30 spectrum hydroxamate inhibitors of matrix metalloproteinases as a member of the TNFR superfamily (3). The CD30 ligand (MMP). This indicates that MMPs or enzymes related to MMPs (CD30L)4 is expressed on activated T cells, resting B cells, mono- are responsible (8–10). The TNF-␣-converting enzyme (TACE; cytes, and granulocytes and established as a member of the TNF ADAM 17), responsible for the release of TNF-␣ by proteolytic superfamily (4, 5). The CD30-CD30L interaction induces pleio- cleavage of the membrane-associated precursor form (pro-TNF-␣), ϩ tropic biological effects on CD30 cells, including activation, pro- was recently characterized as a membrane-anchored metallopro- by guest on September 25, 2021 liferation, differentiation, and cell death, depending on cell type teinase-disintegrin (ADAM, a metalloproteinase and disintegrin) and accessible other stimuli (6). Soluble CD30 (sCD30) is released (11, 12). Metalloproteinase-disintegrins belong to the metzincin from the cells by proteolytic cleavage of CD30. This CD30 shed- superfamily of metalloproteinases encompassing the physiologi- ϩ ding is enhanced by interaction with the CD30L cells and the cally important families of MMPs, reprolysins and astacins. Due to resulting sCD30 is proposed to reduce the CD30L-dependent ac- their x-ray crystal structure, metalloproteinase-disintegrins are ϩ tivation of CD30 cells (7). closely related to snake venom metalloproteinases (reprolysins) The ectodomains of many membrane proteins can be released and are therefore regarded as members of the reprolysin family of proteolytically from the cell by receptor stimulation or in response metalloproteinases (13). Both snake venom proteinases and met- to protein kinase C activators such as PMA. As demonstrated for alloproteinase-disintegrins exhibit membrane protein sheddase ac- the angiotensin-converting enzyme, the Alzheimer’s amyloid pre- tivity through their catalytic domain (11, 12, 14) and binding to ␣ cursor protein, pro-TNF- , the Fas ligand, CD16, L-selectin, the cellular disintegrin receptors mediated by the disintegrin domain (15–17). Besides pro-TNF-␣, TACE cleaves the amyloid precursor protein (18) and mediates the shedding of L-selectin, the p75 TNFR, and TGF-␣ (16). Other metalloproteinase-disintegrins such Departments of *Biochemistry and †Anatomy, University of Kiel, Kiel, Germany; and ‡School of Biological Science, University of East Anglia, Norwich, United Kingdom as the Drosophila ADAM 10 homologue KUZ and ADAM 9 have Received for publication July 12, 1999. Accepted for publication September 7, 2000. been found to cleave the Notch receptor or the membrane-an- The costs of publication of this article were defrayed in part by the payment of page chored heparin-binding EGF-like growth factor, respectively (19, charges. This article must therefore be hereby marked advertisement in accordance 20). Physiologically, the activity of metalloproteinases is regulated with 18 U.S.C. Section 1734 solely to indicate this fact. by tissue inhibitors of metalloproteinases (TIMPs). Although the 1 This research was supported by a grant from Haensel Stiftung (to H.H.) and a grant activity of MMPs is blocked by TIMP-1, -2, and -3 (21), TACE is from the Bundesministerium fu¨r Bildung und Forschung (to H.L.). only inhibited by TIMP-3 (22). Hence, the shedding of TNF-␣ and 2 Address correspondence and reprint requests to Dr. Hinrich Hansen, Department of Biochemistry, University of Kiel, Olshausenstrasse 40, D-24098 Kiel, Germany. E- L-selectin is inhibited by TIMP-3 and not by TIMP-1 and -2 mail address: [email protected] (23–25). 3 Current address: Department of Microbiology, Columbia University, New York, NY. In the present study, we characterized the CD30-cleaving en- 4 Abbreviations used in this paper: CD30L, CD30 ligand; ADAM, a metalloprotein- zyme found to be membrane-associated and inhibited by the ase and disintegrin; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propane- TACE inhibitor TIMP-3. Since TACE-specific mRNA was de- sulfonate; GSSG, oxidized glutathione; MFI, mean fluorescence intensity; MMP, ma- ϩ trix metalloproteinase; phOx, phenyloxazolone; sCD30, soluble CD30; TACE, TNF- tected in CD30 Karpas 299 cells, we describe the bacterial ex- ␣-converting enzyme; TIMP, tissue inhibitor of metalloproteinases. pression of the TACE ectodomain. Its potency to cleave CD30 was

compared with that of ADAM 10 and the related snake venom refolded by a 1:100 dilution in either 1) standard refolding buffer (20 mM ␮ metalloproteinase hemorrhargin. Tris-HCl (pH 7.6) containing 200 mM NaCl, 5 mM CaCl2, 100 M ZnCl2, 100 mM arginine, and 0.002% NaN3) or 2) in this buffer which was further supplemented with 1 mM DTT and 3 mM oxidized glutathione (GSSG) Materials and Methods (oxido-shuffling buffer). After incubation for 3 days at 4°C, the crude extract Materials was diluted with the same volume of ice-cold water and the recombinant proteins were purified on glutathione-Sepharose (Amersham Pharmacia Bio- The mAbs Ki-2 (␥1,␬) and Ki-3 (␥2b,␬) were used for detection of cell- tech). The elution of GST-TACE was conducted with 10 mM of reduced bound and sCD30 (26). The anti-2-phenyloxazolone (anti-phOx) Ab BH-1 glutathione in PBS. (␥2b,␬) served as an isotype-matched control in flow cytometry (27). OKT-3 (anti-CD3 mAb) was obtained from American Type Culture Col- Flow cytometry lection (Manassas, VA). The tissue inhibitors of metalloproteinases TIMP-1, TIMP-2, and TIMP-3 and the recombinant ADAM 10 were from Membrane-bound proteins were determined by direct immunofluorescence G.M. (Norwich, U.K.). BB-2116 (N1-(5-acetylamino-1S-methylcarbamoyl- analysis. Cells (3 ϫ 105) were labeled with saturating amounts of FITC- pentyl)-2S-(2,2-dimethyl-propyl)-N4-hydroxy-3R-(4-hydroxy-phenylsulfanyl- conjugated Abs Ki-3-FITC (anti-CD30), anti-CD25-FITC (Dako, Ham- methyl)-succinamide) was a kind gift from British Biotech Pharmaceuti- burg, Germany), or BH1-FITC (anti-phOx) in PBS containing 0.1% BSA cals (Oxford, U.K.). Hemorrhargin from Echis pyramidum laekeyi was and 0.02% sodium azide for 20 min on ice. Cells were analyzed using a donated by Gavin Laing (Liverpool, U.K.). The large cell anaplastic lym- FACScan flow cytometer (Becton Dickinson, Heidelberg, Germany). phoma cell line Karpas 299 was a kind gift from Abraham Karpas (Cam- bridge, U.K.). The cells were cultured at 37°C and 7% CO2 in RPMI 1640 Determination of sCD30 medium supplemented with 10% FCS, 2 mM glutamine, 100 U/ml peni- cillin, and 100 ␮g/ml streptomycin. Flexible 96-well microtiter plates (Integra Bioscience, Fernwald, Ger- many) were coated with 50 ␮l of Ki-2 mAb (50 ␮g/ml) by overnight Stimulation of human lymphocytes incubation at 4°C. The plates were washed and subsequently blocked with Downloaded from PBS/10% FCS for1hatroom temperature. After washing (three times PBLs were isolated from buffy coat preparations of normal donors. The with PBS/0.1% Tween 20), serial dilutions of both the CD30 standard lymphocytes were purified by Ficoll density centrifugation (1.077 g/ml; (1–300 U/ml) (29) and the sCD30-containing samples were added and Seromed, Berlin, Germany) and cultivated for1hincomplete RPMI 1640 incubated for1hatroom temperature. After washing,125I-labeled Ki-3 medium/10% FCS. Nonadherent lymphocytes were cultivated in complete mAb was added and incubated for 1 h and, after washing again, the ra- ϩ 6 RPMI 1640 medium/10% FCS 100 U/ml IL-2 (10 /ml cells) for 72 h in dioactivity of the single wells was determined and compared with the in- the absence or presence of PHA (0.1% phytohemagglutinin P; Difco, De-

ternal standard. http://www.jimmunol.org/ troit, MI) and immobilized OKT3 (anti-CD3-mAb), respectively. For the OKT3-induced activation, dishes were coated with 10 ␮g/ml OKT3 for 1 h Gelatin zymography at room temperature and washed thoroughly before lymphocyte stimulation. TACE was analyzed by SDS-polyacrylamide gel zymography. Samples were prepared in conventional nonreducing loading buffer without boiling Membrane extraction and run on 10% polyacrylamide gels containing 0.1% gelatin. The gels Karpas 299 cells were cultivated with or without PMA (30 ng/ml, 30 min) were then washed twice for 15 min in 2.5% Triton X-100 to remove SDS, followed by a 20-min wash in metalloproteinase buffer (50 mM Tris-HCl and lysed by ultrasonication (5 min) on ice. A crude membrane fraction ␮ was obtained by centrifugation (50,000 ϫ g, 2 h, 4°C) which was resus- (pH 7.6), 5 mM CaCl2, 100 M ZnCl2, and 0.02% NaN3). After a further pended by ultrasonication (2 min on ice) in 2 ml of 0.14 M NaCl and 20 incubation period for 20 h at 37°C in this buffer, the gels were stained with mM HEPES (pH 7.4). Aliquots were then incubated under shaking with 1% Coomassie brilliant blue. A clear zone indicated the presence of a protein- by guest on September 25, 2021 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS; Sigma, ase with gelatinolytic activity. St. Louis, MO) for2hat4°C. The mixture was then centrifuged (50,000 ϫ g, 3 h, 4°C) and supernatants were intensively dialyzed against 140 mM NaCl Results and 20 mM HEPES (pH 7.4) and then collected as solubilizate. Hydroxamate inhibitor blocks PMA-induced CD30 shedding on Cloning of the human TACE ectodomain and its expression as normal and tumor cells GST-TACE fusion protein CD30 is expressed on CD30ϩ lymphoma cell lines, e.g., the Total RNA from Karpas 299 cells was extracted with Trizol reagent (Life Hodgkin’s disease derived cell line L540 and the large cell ana- Technologies, Karlsruhe, Germany), digested with RNase-free DNase I plastic lymphoma cell line Karpas 299 as well as on normal PBL (Roche Diagnostics, Mannheim, Germany), and cDNA synthesis was per- after stimulation with PHA or anti-CD3 Ab (OKT-3). Constitutive formed using a SuperScript Preamplification System and oligo(dT)12–18 shedding of CD30 was detected from all cells. This endoproteo- primers (Life Technologies). For an amplification of the cDNA fragment encoding the catalytic, disintegrin, and cysteine-rich domain of TACE, lytic release of the CD30 ectodomain (sCD30) was strongly acti- total cDNA was subjected to 30 cycles of PCR using proofreading poly- vated by PMA, resulting in a loss of membrane-bound CD30 de- merase (Elongase Enzyme mix; Life Technologies). As specific primers, tected by flow cytometry and a corresponding increase of sCD30 which were 5Ј extended to allow restriction enzyme cleavage, we used the in the supernatant detected by an CD30 Ab-based sandwich RIA Ј BamHI-TACE sense primer (5 -GTGAAAAGGATCCTTGCTGACCCA (Fig. 1, A and C). The FITC-labeled 2-phOx Ab BH1 (BH1-FITC), GATCCCATGAAGAACACG and the EcoRI-TACE antisense primer (5Ј-AGTGCTAGAATTCTACATCCTGTACTCGTTTCTCACATTTGCC. which served as an isotype control in flow cytometry, showed no The sequences were derived from GenBank accession number U69611. The binding to any of the tested cells (Fig. 1B). PMA-induced shedding PCR product was purified by UV band elution (28) and restricted by BamHI of CD30 was effectively blocked by the hydroxamic acid-based and EcoRI enzymes (Roche Diagnostics) before the in-frame ligation with T4 broad-spectrum MMP inhibitor BB-2116 (20 ␮M), whereas inhi- DNA ligase (New England Biolabs, Beverly, MA) into the BamHI-EcoRI- restricted multiple cloning site of the pGEX-3X expression vector (Amersham bition of constitutive shedding varied depending on the cell type. Pharmacia Biotech, Freiburg, Germany). The insert of one clone was subse- Although BB-2116 blocked the constitutive sCD30 release from quently sequenced using an Abi Prism 310 Genetic Analyzer (Applied Bio- tumor cells, the release from activated normal cells was hardly systems, Weiterstadt, Germany). Then the recombinant expression vector was affected. These data showed that PMA-induced shedding was me- transformed into competent Escherichia coli cell strain XL1-blue and grown diated by metalloproteinases; however other proteinases might be on ampicillin-containing agar plates. PCR-tested colonies were used to inoculate 2-ml cultures of Luria-Bertani medium containing 100 ␮g/ml ampicillin involved in constitutive shedding from normal cells. and 1 ml of this culture was used to inoculate an ampicillin-containing 500-ml

culture. When the culture reached an OD (A600) of 0.5, the expression was The CD30-cleaving enzyme is localized at the plasma membrane induced by addition of isopropyl-1-thio-␤-D-galactopyranoside (1.5 mM ﬁnal concentration) and the cells were incubated for 5 h more at 37°C. GST-TACE To decide whether PMA-induced CD30 cleavage was mediated by was obtained from the inclusion bodies and solubilized in 20 mM Tris-HCl a secreted or a membrane-anchored metalloproteinase, the CD30- (pH 8) containing 8 M urea and 100 mM 2-ME. The solubilized proteins were cleaving activities of supernatant ﬂuids and membrane extracts of The Journal of Immunology 6705 Downloaded from

FIGURE 2. Localization of CD30 sheddase. Karpas 299 cells (1 ϫ 106/ ml) were incubated for 30 min with 30 ng/ml PMA (PMA Karpas) or not stimulated (Karpas). Cell supernatant and membranes were prepared as described in Materials and Methods. Membranes were extracted with 1%

CHAPS. Supernatant (SN30) and membrane preparations of nonstimulated http://www.jimmunol.org/ and PMA-activated cells were applied to untreated Karpas 299 cells with or without 20 ␮M BB-2116 and incubated for1hat37°C. A, CD30 expression was determined by FACS analysis using FITC-labeled Ki-3 mAb. B, FITC-labeled anti-phOx mAb was used as an isotype control. The results are given as means of the MFI for triplicate determinations.

reduction of CD30 to 77% of the untreated control. As with PMA,

these membrane effects were dependent on metalloproteinases, by guest on September 25, 2021 since BB-2116 (20 ␮M) abolished the CD30-cleaving activity. The CHAPS-extracted membrane preparation of PMA-activated Kar- pas 299 cells also induced a loss of CD30 on PHA-activated normal PBLs, but had no influence on the expression of CD25 on both cells (data not shown). These data suggest that PMA-stimulated CD30 shedding was mediated by endogenous membrane-anchored metalloproteinases that could be activated or up-regulated by FIGURE 1. Inhibition of CD30 shedding by hydroxamate inhibitor. PMA. Moreover, membrane anchoring via the metalloproteinase Cells (106/ml) were incubated for 90 min in the absence or presence of transmembrane domain is not essential for activity in this system. PMA (30 ng/ml) with or without BB-2116 (20 ␮M). Cells used were: The CD30ϩ cell lines Karpas 299 and L540, CD30Ϫ normal PBL, PHA-stim- TIMP-3 inhibits shedding of CD30 ulated PBL, and OKT3-stimulated PBL (anti-CD3 PBL). A, Membrane- For a characterization of the responsible enzyme(s), the sensitivity bound CD30 was determined by FACS analysis using FITC-labeled Ki-3 of the PMA-induced CD30 shedding for a variety of metallopro- mAb, B, FITC-labeled anti-phOx mAb was used as an isotype control. The teinase inhibitors was tested: 1) The efficacy of the broad-spectrum results are given as means of the MFI for triplicate determinations. C, hydroxamate MMP inhibitor BB-2116 (20 ␮M) to prevent CD30 sCD30 was determined in the cell supernatants. The results show units per cleavage (10) (Figs. 1–3) suggested the involvement of MMPs or milliliter as means Ϯ SD for triplicate determinations. related enzymes in the shedding of CD30. 2) In contrast, the natural inhibitors TIMP-1 (10 ␮g/ml) and TIMP-2 (10 ␮g/ml), which PMA-stimulated Karpas 299 cells were compared. Although a di- inhibit soluble and membrane-type MMPs (21, 30), could not pre- rect incubation of Karpas 299 cells with PMA (30 ng/ml, 30 min) vent the PMA-induced CD30 shedding (Fig. 3A). This does not induced a loss of membrane-bound CD30 to 65% of the untreated support an involvement of MMPs. 3) Phosphoramidon (10 ␮M) control, an incubation of the cells for 1 h with the supernatant of and captopril (10 ␮M), which are known inhibitors of thermolysin, PMA-activated Karpas cells (30 ng/ml, 30 min) had no influence neprilysin, and angiotensin-converting enzyme, respectively, also on the CD30 expression (Fig. 2). For the determination of CD30- failed to influence CD30 cleavage. 4) Like BB-2116, TIMP-3 in- cleaving activity in cell membranes, membrane fractions of PMA- hibited the PMA-induced CD30 shedding from Karpas 299 cells in activated Karpas 299 cells were extracted with the solubilizing a dose-dependent manner (Fig. 3B). The inhibitory potency of reagent CHAPS. The extract induced a loss of CD30 to 57% of the TIMP-3 (IC50, 30 nM) was 8-fold higher than that of BB-2116 dialysis buffer-incubated control (buffer). Interestingly, membrane (IC50, 230 nM). A similar inhibition by TIMP-3 was obtained extracts of nonstimulated cells exhibited CD30 sheddase activity when we analyzed CD30 shedding from CD30ϩ normal cells (data as well. This activity was weaker but still capable to induce a not shown). Since the MMP inhibitor TIMP-3 potently blocks the 6706 CD30 CLEAVAGE BY TACE

FIGURE 4. RT-PCR of TACE mRNA. RT-PCR ampliﬁcation of a cDNA fragment coding for the TACE ectodomain using mRNA from Kar- pas 299 cells. The arrow indicates the 1299-bp PCR-product. Downloaded from http://www.jimmunol.org/

FIGURE 3. Inhibition of PMA-induced CD30 down-regulation. A, Kar- pas 299 cells (1 ϫ 106/ml) were incubated for 1 h with 30 ng/ml PMA in the presence of the following metalloproteinase inhibitors: BB-2116 (20 ␮M), phosphoramidon (10 ␮M), captopril (10 ␮M), TIMP-1 (10 ␮g/ml), TIMP-2 (10 ␮g/ml), and TIMP-3 (800 nM). CD30 expression was determined by FACS analysis. sCD30 was determined in the cell supernatants.

The results show units per milliliter as means Ϯ SD for triplicate deter- by guest on September 25, 2021 minations. The sCD30 concentrations in the supernatants of untreated or only PMA-stimulated cells are indicated by dotted lines. B, The IC50 of TIMP-3 and BB-2116 was determined.

membrane-anchored metalloproteinase-disintegrin TACE (22), our data provide evidence for a role of TACE or a TACE-like proteinase in CD30 shedding.

TACE mRNA is found in Karpas 299 cells TACE-specific mRNA transcripts were detected by RT-PCR from PMA-activated Karpas 299 cells and CD30ϩ PBLs. Using TACE- specific primers, a 1299-bp cDNA fragment was amplified from both sources as shown for Karpas 299 (Fig. 4). The purified 1299-bp fragment from Karpas 299 cells, encoding the catalytic, disintegrin, and cysteine-rich domain of TACE, was cloned into the pGEX-3X GST gene fusion vector. Sequencing confirmed that it was identical with the published mRNA sequence of TACE derived from the GenBank accession number U69611 (data not shown).

Recombinant soluble TACE cleaves CD30 The TACE ectodomain was expressed in bacteria as a GST fusion FIGURE 5. Bacterial expression of TACE. A, Schematic representation protein (GST-TACE) (Fig. 5A). Crude extracts from the insoluble of the GST-TACE fusion protein. B, SDS-PAGE analysis of the bacterial expressed GST-TACE fusion protein that was purified on glutathione- fraction of the transfected bacteria were either refolded by 1) a Sepharose, after refolding by dilution in the presence of GSSG/DTT (lane standard dilution protocol that allowed the renaturation of the GST 1, GST-TACE(ox)) or in the absence of GSSG/DTT (lane 2, GST-TAC- tag, but not of the cysteine-rich TACE ectodomain (GST-TAC- E(st)). Control, Wild-type GST was refolded and purified (lane 3). C, Gel- E(st)) or 2) dilution into the oxido-shuffling system which facili- atinolytic activity of GST-TACE. Ten nanograms of affinity-purified GST- tates the formation of proper disulfide bonds after denaturing ex- TACE(ox) (lane 1), 1 ␮g of GST-TACE(st), and 1 ␮g of wild-type GST traction (GST-TACE(ox)). After affinity purification on reduced (lane 3) were analyzed by gelatin zymography. The Journal of Immunology 6707 glutathione-Sepharose, homogeneous GST-TACE was obtained F). To elucidate the physiological relevance of TACE, we applied from both preparations (Fig. 5B). From the oxido-shuffling buffer, recombinant ADAM 10, another membrane-anchored metallopro- a lower amount of material could be recovered, yielding about teinase-disintegrin. Although ADAM 10 was active degrading bo- 20–30% of that from the standard refolded preparation. This might vine myelin basic protein (data not shown), its CD30-cleaving ac- be due to the glutathione (GSSG) in the oxido-shuffling buffer tivity was weak in comparison to that of TACE. No loss of CD30 which might interact with the subsequent purification procedure. was visible in flow cytometry. Only the more sensitive sandwich However, the proteinase activity of GST-TACE(ox) recovered RIA detected an enhanced sCD30 level in the supernatant of Kar- from the oxido-shuffling buffer was 100-fold higher, since 10 ng pas 299 cells after treatment with 3000 ng/ml ADAM 10 (Fig. 6B). elicited a gelatinolytic activity equivalent to 1 ␮g of GST-TAC- These findings suggest TACE being the more relevant CD30 E(st) in gelatin zymography (Fig. 5C). Hence, the two refolding sheddase. As a control, GST-TACE-, ADAM 10, and snake ven- procedures allowed the preparation of an active and inactive con- om-mediated CD30 shedding was blocked by BB-2116 (20 ␮M), formation of the TACE ectodomain. indicating that CD30 cleavage was only effected by metallopro- Next, the recombinant metalloproteinase-disintegrin GST- teinase action. TACE was tested for its potency to cleave CD30 from Karpas 299 cells. Indeed, GST-TACE(ox) caused a dose-dependent decrease Discussion of membrane-bound CD30 and a concomitant increase of sCD30 The objective of the present study was to determine the hitherto in the cell supernatant. As a control, inactive GST-TACE(st) had unknown enzyme(s) responsible for the shedding of the lympho- no effect in this respect (Figs. 6, A and E). The CD30 cleavage was cytic activation marker CD30. The membrane-bound molecule of detectable at 300 ng/ml GST-TACE, comparable in its activity 120 kDa is released from the cells in vitro as well as in vivo, giving Downloaded from with the purified snake venom metalloproteinase-disintegrin hem- rise to a soluble form of 90 kDa. Elevated concentrations of this orrhargin from Echis pyramidum laekeyi, which also caused a de- sCD30 turned out to be a reliable serum marker for monitoring the tectable CD30 cleavage at the same concentration (Fig. 6, B and activity of a variety of immunological disorders such as CD30ϩ http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 6. CD30 shedding by GST-TACE. Karpas 299 cells (106/ml) were incubated for 1 h with active GST-TACE(ox), inactive GST-TACE(st), ADAM 10 or hemorrhargin in concentrations, as indicated, with or without 20 ␮M BB-2116. A and B, Cell surface expression of CD30 was determined by flow cytometry using FITC-labeled Ki-3 mAb. C and D, FITC-labeled anti-phOx mAb was used as an isotype control. The results are given as means of the MFI Ϯ SD for triplicate determinations. E and F, sCD30 was determined in the cell supernatants. The results show units per milliliter as means Ϯ SD for triplicate determinations. 6708 CD30 CLEAVAGE BY TACE lymphomas (31), systemic lupus erythematosus (32), rheumatoid has also been described for two other metalloproteinase-disinte- arthritis (33), and Wegener’s granulomatosis (34). CD30 shedding grins, i.e., ADAM 12 and ADAM 9 (MDC 9) (43, 44). The inhi- could be inhibited by the synthetic broad-spectrum hydroxamate bition profile of ADAM 12 appears to be completely different from MMP inhibitor BB-2116 on activated normal lymphocytes as well that of TACE since neither the hydroxamate inhibitor BB-94 nor as on CD30ϩ lymphoma cells. This MMP inhibitor showed a TIMP-1, -2, and -3 were effective (45). Inhibition data on MDC 9 wider spectrum of inhibition including MMPs and reprolysins (34) are not yet available. Hence, although inhibition data suggest a role (Fig. 6), and hence referred to as broad-spectrum MMP inhibitor. of TACE in CD30 shedding, the involvement of so far uncharac- Unlike BB-2116, the physiological MMP inhibitors TIMP-1 and -2 terized TACE-related enzymes cannot be excluded. failed to inhibit PMA-induced CD30 shedding. We concluded that CD30 is released by a MMP-related metalloproteinase but neither by soluble MMPs nor by any of the hitherto characterized mem- Acknowledgments brane-type MMPs, since these are all sensitive to TIMP-2 or We thank M. Burmester for excellent assistance with membrane preparations. TIMP-1 (21, 30, 35–39). In contrast to TIMP-1 and -2, CD30 shedding was potently inhibited by TIMP-3. This inhibition profile has also been described for the shedding of other membrane pro- References teins, e.g., pro-TNF-␣, L-selectin, the p75 TNFR, and the IL-6R 1. Schwab, U., H. Stein, H. Gerdes, H. Lemke, H. Kirchner, M. Schaadt, and (22–25), suggesting a common releasing mechanism. V. Diehl. 1982. Production of a monoclonal antibody specific for Hodgkin and Sternberg-Reed cells of Hodgkin’s disease and a sub-set of normal lymphoid Pro-TNF-␣, L-selectin, and the p75 TNFR are released by cells. Nature 299:65.

TACE which is a membrane-anchored metalloproteinase-disinte- 2. Stein, H., D. Y. Mason, J. Gerdes, N. O’Connor, J. Wainscoat, G. Pallesen, Downloaded from grin (ADAM) and belongs, along with soluble snake venom met- K. Gatter, B. Fallini, G. Delsol, H. Lemke, and K. Lennert. 1985. The expression of the Hodgkin’s-disease-associated antigen Ki-1 in reactive and neoplastic lym- alloproteinases, to the reprolysin family of metalloproteinases (11, phoid tissue: evidence that Sternberg-Reed cells and histiocytic malignancies are 12, 16). TACE is strongly inhibited by TIMP-3, but not affected by derived from activated lymphoid cells. Blood 66:848. the MMP inhibitors TIMP-1 and -2 (22). The strong inhibition of 3. Durkop, H., U. Latza, M. Hummel, F. Eitelbach, B. Seed, and H. Stein. 1992. Molecular cloning and expression of a new member of the nerve growth factor CD30 shedding from Karpas 299 cells by TIMP-3 and the inability receptor family that is characteristic for Hodgkin’s disease. Cell 68:421.

of TIMP-1 and -2 suggested that TACE might be responsible for 4. Smith, C. A., H. J. Gruss, T. Davis, D. Anderson, T. Farrah, E. Baker, http://www.jimmunol.org/ the cleavage of CD30 as well. This assumption was further sub- G. R. Sutherland, C. I. Brannan, N. G. Copeland, N. A. Jenkins, et al. 1993. CD30 antigen, a marker for Hodgkin’s lymphoma, is a receptor whose ligand defines an stantiated: 1) It could be demonstrated that CD30 shedding was emerging family of cytokines with homology to TNF. Cell 73:1349. mediated by a membrane-associated metalloproteinase since sol- 5. Gruss, H. J., A. Pinto, A. Gloghini, E. Wehnes, B. Wright, N. Boiani, D. Aldinucci, V. Gattei, V. Zagonel, C. A. Smith, et al. 1996. CD30 ligand ubilized membrane extracts of Karpas 299 cells were able to expression in nonmalignant and Hodgkin’s disease-involved lymphoid tissues. cleave CD30 at the cell surface of intact cells. 2) TACE-specific Am. J. Pathol. 149:469. mRNA transcripts were detected in Karpas 299 cells by means of 6. Gruss, H. J., N. Boiani, D. E. Williams, R. J. Armitage, C. A. Smith, and R. G. Goodwin. 1994. Pleiotropic effects of the CD30 ligand on CD30-expressing RT-PCR. 3) It could be demonstrated that the recombinant fusion cells and lymphoma cell lines. Blood 83:2045. protein containing GST and the TACE ectodomain (GST-TACE) 7. Gruss, H. J., A. Pinto, J. Duyster, S. Poppema, and F. Herrmann. 1997. Hodgkin’s elicited CD30 sheddase activity in a dose-dependent manner. disease: a tumor with disturbed immunological pathways. Immunol. Today by guest on September 25, 2021 These findings demonstrated that TACE is expressed by Karpas 18:156. 8. Arribas, J., L. Coodly, P. Vollmer, T. K. Kishimoto, S. Rose-John, and 299 cells and is able to cleave membrane-bound CD30. The CD30- J. Massague. 1996. Diverse cell surface protein ectodomains are shed by a system cleaving activity of the construct was comparable to that of the sensitive to metalloprotease inhibitors. J. Biol. Chem. 271:11376. snake venom metalloproteinase hemorrhargin, both belong to the 9. Hooper, N. M., E. H. Karran, and A. J. Turner. 1997. Membrane protein secre- tases. Biochem. J. 321:265. reprolysin family of metalloproteinases. Since hemorrhargin is a 10. Hansen, H. P., T. Kisseleva, J. Kobarg, O. Horn Lohrens, B. Havsteen, and TNF-␣-releasing enzyme (14), the cleaving of CD30 serves to cor- H. Lemke. 1995. A zinc metalloproteinase is responsible for the release of CD30 roborate the common releasing mechanism. on human tumor cell lines. Int. J. Cancer 63:750. ␣ 11. Black, R. A., C. T. Rauch, C. J. Kozlosky, J. J. Peschon, S. L. Slack, Regarding the related shedding of TNF- , it has not been clar- M. F. Wolfson, B. J. Castner, K. L. Stocking, P. Reddy, et al. 1997. A metallo- ified whether pro-TNF-␣ is exclusively cleaved by TACE on all proteinase disintegrin that releases tumour-necrosis factor-␣ from cells. Nature cell types since inactivation of the TACE gene in mouse T cells 385:729. ␣ 12. Moss, M. L., S. L. Jin, S. L., M. E. Milla, W. Burkhart, H. L. Carter, W. J. Chen, caused a marked decrease but not complete inhibition of TNF- W. C. Clay, J. R. Didsbury, D. Hassler, C. R. Hoffman, et al. 1997. Cloning of release (11). ADAM 10, another membrane-anchored metallopro- a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-␣. teinase-disintegrin closely related to TACE, was also able to Nature 385:733. ␣ 13. Maskos, K., C. Fernandez-Catalan, R. Huber, G. P. Bourenkov, H. Bartunik, cleave pro-TNF- (40, 41) from the cell surface, indicating that G. A. Ellestad, P. Reddy, M. F. Wolfson, C. T. Rauch, B. J. Castner, et al. 1998. other metalloproteinases might play a role. Similarly, sCD30 was Crystal structure of the catalytic domain of human tumor necrosis factor-␣-con- not only released by TACE but also by ADAM 10, although less verting enzyme. Proc. Natl. Acad. Sci. USA 95:3408. 14. Moura-da-Silva, A. M., G. D. Laing, M. J. Paine, J. M. Dennison, V. Politi, effectively. Since we detected mRNA of both TACE and ADAM J. M. Crampton, and R. D. Theakston. 1996. Processing of pro-tumor necrosis ϩ 10 (data not shown) in CD30 Karpas 299 cells, the question factor-␣ by venom metalloproteinases: a hypothesis explaining local tissue dam- arises whether TACE represents the only physiologically relevant age following snake bite. Eur. J. Immunol. 26:2000. 15. Huovila, A. P. J., E. A. Almeida, and J. M. White. 1996. ADAMs and cell fusion. CD30-cleaving enzyme. The data from direct application of the Curr. Opin. Cell Biol. 8:692. recombinant ADAMs do not clearly answer this question. The 16. Peschon, J. J., J. L. Slack, P. Reddy, K. L. Stocking, S. W. Sunnarborg, D. C. Lee, evaluation of their inhibition profile appears to be more promising. W. E. Russell, B. J. Castner, R. S. Johnson, J. N. Fitzner, et al. 1998. An essential role for ectodomain shedding in mammalian development. Science 282:1281. Although inhibition data on ADAMs are incomplete, TIMPs play 17. Kamiguti, A. S., M. Zuzel, and R. D. Theakston. 1998. Snake venom metallo- a role in distinguishing different ADAMs. TACE is inhibited by proteinases and disintegrins: interactions with cells. Braz. J. Med. Biol. Res. TIMP-3 and not by TIMP-1 and 2, whereas both TIMP-1 and -3 31:853. 18. Buxbaum, J. D., K. N. Liu, Y. Luo, J. L. Slack, K. L. Stocking, J. J. Peschon, inhibit ADAM 10 (22, 42). Thus, the inhibition of CD30 shedding R. S. Johnson, B. J. Castner, D. P. Cerretti, and R. A. Black. 1998. Evidence that by TIMP-3 and not by TIMP-1 and -2 argues in favor of TACE as tumor necrosis factor ␣ converting enzyme is involved in regulated ␣-secretase the physiologically relevant CD30-cleaving metalloproteinase. cleavage of the Alzheimer amyloid protein precursor. J. Biol. Chem. 273:27765. 19. Pan, D., and G. M. Rubin. 1997. Kuzbanian controls proteolytic processing of TACE and ADAM 10 are structurally very similar and differ from Notch and mediates lateral inhibition during Drosophila and vertebrate neuro- other known ADAMs (13, 40, 41). Nonetheless, catalytic activity genesis. Cell 90:271. The Journal of Immunology 6709

20. Izumi, Y., M. Hirata, H. Hasuwa, R. Iwamoto, T. Umata, K. Miyado, Y. Tamai, 33. Gerli, R., C. Muscat, O. Bistoni, B. Falini, C. Tomassini, E. Agea, R. Tognellini, T. Kurisaki, A. Sehara-Fujisawa, S. Ohno, and E. Mekada. 1998. A metallopro- P. Biagini, and A. Bertotto. 1995. High levels of the soluble form of CD30 tease-disintegrin, MDC9/meltrin-␥/ADAM9 and PKC␦ are involved in TPA-in- molecule in rheumatoid arthritis (RA) are expression of CD30ϩ T cell involve- duced ectodomain shedding of membrane-anchored heparin-binding EGF-like ment in the inflamed joints. Clin. Exp. Immunol. 102:547. growth factor. EMBO J. 17:7260. 34. Wang, G., H. Hansen, E. Tatsis, E. Csernok, H. Lemke, and W. L. Gross. 1997. 21. Murphy, G., and F. Willenbrock. 1995. Tissue inhibitors of matrix metalloen- High plasma levels of the soluble form of CD30 activation molecule reflect dis- dopeptidases. Methods Enzymol. 248:496. ease activity in patients with Wegener’s granulomatosis. Am. J. Med. 102:517. 22. Amour, A., P. M. Slocombe, A. Webster, M. Butler, C. G. Knight, B. J. Smith, 35. Butler, G. S., H. Will, S. J. Atkinson, and G. Murphy. 1997. Membrane-type-2 P. E. Stephens, C. Shelley, M. Hutton, V. Knauper, et al. 1998. TNF-␣ converting matrix metalloproteinase can initiate the processing of progelatinase A and is enzyme (TACE) is inhibited by TIMP-3. FEBS Lett. 435:39. regulated by the tissue inhibitors of metalloproteinases. Eur. J. Biochem. 23. Borland, G., G. Murphy, and A. Ager. 1999. Tissue inhibitor of metalloprotein- 244:653. ases-3 inhibits shedding of L-selectin from leukocytes. J. Biol. Chem. 274:2810. 36. English, W. R., X. S. Puente, J. M. Freije, V. Knauper, A. Amour, 24. Hargreaves, P. G., F. Wang, J. Antcliff, G. Murphy, J. Lawry, R. G. Russell, and A. Merryweather, C. Lopez-Otin, and G. Murphy. 2000. Membrane type 4 matrix P. I. Croucher. 1998. Human myeloma cells shed the interleukin-6 receptor: metalloproteinase (MMP17) has tumor necrosis factor-␣ convertase activity but inhibition by tissue inhibitor of metalloproteinase-3 and a hydroxamate-based does not activate pro-MMP2. J. Biol. Chem. 275:14046. metalloproteinase inhibitor. Br. J. Haematol. 101:694. 37. Will, H., S. J. Atkinson, G. S. Butler, B. Smith, and G. Murphy. 1996. The 25. Smith, M. R., H. Kung, S. K. Durum, N. H. Colburn, and Y. Sun. 1997. TIMP-3 soluble catalytic domain of membrane type 1 matrix metalloproteinase cleaves induces cell death by stabilizing TNF-␣ receptors on the surface of human colon the propeptide of progelatinase A and initiates autoproteolytic activation: regu- carcinoma cells. Cytokine 9:770. lation by TIMP-2 and TIMP-3. J. Biol. Chem. 271:17119. 26. Horn-Lohrens, O., M. Tiemann, H. Lange, J. Kobarg, M. Hafner, H. Hansen, 38. Shimada, T., H. Nakamura, E. Ohuchi, Y. Fujii, Y. Murakami, H. Sato, M. Seiki, W. Sterry, R. M. Parwaresch, and H. Lemke. 1995. Shedding of the soluble form and Y. Okada. 1999. Characterization of a truncated recombinant form of human of CD30 from the Hodgkin-analogous cell line L540 is strongly inhibited by a membrane type 3 matrix metalloproteinase. Eur. J. Biochem. 262:907. new CD30-specific antibody (Ki-4). Int. J. Cancer 60:539. 39. Wang, X., J. Yi, J. Lei, and D. Pei. 1999. Expression, purification and charac- 27. Lange, H., M. Solterbeck, C. Berek, and H. Lemke. 1996. Correlation between terization of recombinant mouse MT5-MMP protein products. FEBS Lett. immune maturation and idiotypic network recognition. Eur. J. Immunol. 26:2234. 462:261. 28. Hansen, H., H. Lemke, and U. Bodner. 1993. Rapid and simple purification of 40. Rosendahl, M. S., S. C. Ko, D. L. Long, M. T. Brewer, B. Rosenzweig, E. Hedl, Downloaded from PCR products by direct band elution during agarose gel electrophoresis. Bio- L. Anderson, S. M. Pyle, J. Moreland, M. A. Meyers, et al. 1997. Identification Techniques 14:28. and characterization of a pro-tumor necrosis factor-␣-processing enzyme from 29. Josimovic-Alasevic, O., H. Durkop, R. Schwarting, E. Backe, H. Stein, and the ADAM family of zinc metalloproteases. J. Biol. Chem. 272:24588. T. Diamantstein. 1989. Ki-1 (CD30) antigen is released by Ki-1-positive tumor 41. Lunn, C. A., X. Fan, B. Dalie, K. Miller, P. J. Zavodny, S. K. Narula, and cells in vitro and in vivo. I. Partial characterization of soluble Ki-1 antigen and D. Lundell. 1997. Purification of ADAM 10 from bovine spleen as a TNF␣ detection of the antigen in cell culture supernatants and in serum by an enzyme- convertase. FEBS Lett. 400:333. linked immunosorbent assay. Eur. J. Immunol. 19:157. 42. Amour, A., C. G. Knight, A. Webster, P. M. Slocombe, P. E. Stephens,

30. Murphy, G., H. Stanton, S. Cowell, G. Butler, V. Knauper, S. Atkinson, and V. Knauper, A. J. Docherty, and G. Murphy. 2000. The in vitro activity of http://www.jimmunol.org/ J. Gavrilovic. 1999. Mechanisms for pro matrix metalloproteinase activation. ADAM-10 is inhibited by TIMP-1 and TIMP-3. FEBS Lett. 473:275. APMIS 107:38. 43. Loechel, F., B. J. Gilpin, E. Engvall, R. Albrechtsen, and U. M. Wewer. 1998. 31. Pizzolo, G., F. Vinante, M. Chilosi, F. Dallenbach, O. Josimovic Alasevic, Human ADAM 12 (meltrin ␣) is an active metalloprotease. J. Biol. Chem. T. Diamantstein, and H. Stein. 1990. Serum levels of soluble CD30 molecule 273:16993. (Ki-1 antigen) in Hodgkin’s disease: relationship with disease activity and clin- 44. Roghani, M., J. D. Becherer, M. L. Moss, R. E. Atherton, ical stage. Br. J. Haematol 75:282. H. Erdjument-Bromage, J. Arribas, R. K. Blackburn, G. Weskamp, P. Tempst, 32. Caligaris Cappio, F., M. T. Bertero, M. Converso, A. Stacchini, F. Vinante, and C. P. Blobel. 1999. Metalloprotease-disintegrin MDC9: intracellular matu- S. Romagnani, and G. Pizzolo. 1995. Circulating levels of soluble CD30, a ration and catalytic activity. J. Biol. Chem. 274:3531. marker of cells producing Th2-type cytokines, are increased in patients with 45. Loechel, F., M. T. Overgaard, C. Oxvig, R. Albrechtsen, and U. M. Wewer. 1999. systemic lupus erythematosus and correlate with disease activity. Clin. Exp. Regulation of human ADAM 12 protease by the prodomain: evidence for a func- Rheumatol. 13:339. tional cysteine switch. J. Biol. Chem. 274:13427. by guest on September 25, 2021