IL-11 Activates Human Endothelial Cells to Resist Immune-Mediated Injury Keyvan Mahboubi, Barbara C. Biedermann, Joseph M. Carroll and Jordan S. Pober This information is current as of October 2, 2021. J Immunol 2000; 164:3837-3846; ; doi: 10.4049/jimmunol.164.7.3837 http://www.jimmunol.org/content/164/7/3837 Downloaded from References This article cites 53 articles, 28 of which you can access for free at: http://www.jimmunol.org/content/164/7/3837.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• 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

*average by guest on October 2, 2021 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. IL-11 Activates Human Endothelial Cells to Resist Immune-Mediated Injury

Keyvan Mahboubi,1* Barbara C. Biedermann,1,2* Joseph M. Carroll,† and Jordan S. Pober3*

IL-11, a gp130-signaling , is protective in several in vivo models of immune-mediated and inflammatory injury. HUVECs express IL-11 receptor ␣-chain and gp130. Human IL-11 causes rapid (2–10 min) tyrosine phosphorylation of gp130. IL-11 at 0.1 and 10 ng/ml induces tyrosine phosphorylation of STAT3 and STAT1, respectively, although maximal responses require 50 ng/ml. Phospho-STAT3 and phospho-STAT1 levels peak rapidly (2.5 min) and disappear by 60 min. The p42 and p44 mitogen-activated kinases (MAPKs) are phosphorylated in response to 0.3 ng/ml IL-11 with maximal activation at 30 ng/ml IL-11. Phos- phorylation of p42 and p44 MAPKs, which can be prevented by a mitogen-activated protein/extracellular signal-related kinase kinase-1 inhibitor, peaks by 15–20 min and largely disappears by 40 min. IL-11 does not activate NF-␬B nor does it inhibit NF-␬B activation by TNF. Similarly, IL-11 neither induces E-selectin or ICAM-1 nor blocks induction by TNF. Although IL-11 does not alter class I MHC complex molecule expression, pretreatment with 0.5 ng/ml IL-11 partially protects HUVECs against lysis by Downloaded from allospecific class I MHC-restricted cytolytic T or by anti-class I MHC Ab plus heterologous C. IL-11-induced cytoprotection is protein synthesis dependent and may depend on mitogen-activated protein/extracellular signal-related kinase kinase-1. Our results indicate that low (i.e., STAT3- and MAPK-activating) concentrations of IL-11 confer resistance to immune- mediated injury in cultured HUVECs without inhibiting proinflammatory responses. The Journal of Immunology, 2000, 164: 3837–3846. http://www.jimmunol.org/ nterleukin-11 is a 20-kDa stromal cell-derived pleiotropic cy- parameters of lung injury including alveolar-capillary protein leak, tokine that interacts with a variety of hemopoietic and non- endothelial cell (EC)4 and epithelial cell membrane injury, lipid I hemopoietic cell types. Recombinant human IL-11 stimulates peroxidation, pulmonary neutrophil recruitment, IL-12 and TNF megakaryocytopoiesis in vitro (1) and in vivo (2). It also stimulates production, and DNA fragmentation. erythropoiesis and regulates macrophage proliferation and differ- The mechanism(s) by which IL-11 protects mucosae are not entiation (3). Due to its thrombopoietic activities in vivo, IL-11 is completely understood. Both antiinflammatory and direct cytopro- used to treat chemotherapy-induced (4). tective effects of IL-11 could contribute to the reduction of injury. In addition to its hemopoietic activities, IL-11 protects against IL-11 may exert antiinflammatory effects by reducing cytokine by guest on October 2, 2021 various forms of mucosal injury. These effects are most exten- production by macrophages (12–14). It also may promote immune sively described in the gastrointestinal tract of rodents where IL-11 deviation from a TH1-like to a TH2-like phenotype, which may protects small intestinal cells from combined radiation, chemother- ameliorate some types of immune-mediated injury (9). To date, apy, and ischemia in mice (5–7); reduces experimental colitis in- there have been no reports of direct cytoprotection in cultured duced by trinitrobenzenesulfonic acid in rats (8); and ameliorates cells. inflammatory bowel disease in mice (6). In these studies, treatment Direct cytoprotection would likely involve the induction of spe- with IL-11 not only decreased mucosal damage but also acceler- cific products. IL-11 belongs to the IL-6 family of , ated healing and improved survival. IL-11 also reduces immune- all of which use gp130 as a critical component for signal trans- mediated small bowel injury in acute graft-vs-host disease after duction (15–17). IL-11 initiates signaling via binding to a unique murine allogeneic transplantation (9). IL-11-receptor-␣ (IL-11R␣) chain (18, 19). The IL-11/IL-11R␣ The protective effects of IL-11 are not restricted to the intestine complex is thought to bind to and induce clustering of gp130, because IL-11 has also been shown to improve survival after tho- leading to the activation, via transphosphorylation, of associated racic irradiation (10). Human IL-11, expressed as a transgene in Janus kinases (JAKs) (20, 21). Activated JAKs phosphorylate ty- bronchial mucosa, reduces mortality associated with hyperoxia in rosine residues within the cytoplasmic region of gp130 which then mice (11). This effect was associated with a reduction in multiple serve as docking sites for STAT3 and STAT1 (22, 23). The activated JAKs subsequently phosphorylate tyrosine residues within the bound STAT proteins, causing the STATs to dissociate *Molecular Cardiobiology Program, Boyer Center for Molecular Medicine, Yale University School of Medicine, New Haven, CT 06510; and †Genetics Institute, from gp130, dimerize (24), and enter the nucleus to act as tran- Andover, MA 01810 scriptional activators of target (25, 26). STAT dimers may Received for publication June 4, 1999. Accepted for publication January 27, 2000. be additionally phosphorylated on serine or threonine residues by The costs of publication of this article were defrayed in part by the payment of page mitogen-activated protein kinases (MAPKs) (27, 28) that are also charges. This article must therefore be hereby marked advertisement in accordance activated in response to cytokine binding to the receptor (29, 30). with 18 U.S.C. Section 1734 solely to indicate this fact. 1 K.M. and B.C.B. contributed equally to this work. 2 Current address: Medinische Universitaetsklinik, Bruderholzspital, CH-4101 Brud- erholz/Basel, Switzerland. 4 Abbreviations used in this paper: EC, endothelial cell(s); ECGS, endothelial cell 3 Address correspondence and reprint requests to Dr. Jordan S. Pober, Boyer Center growth supplement; IL-11R␣, IL-11-receptor ␣-chain; JAK, ; MAPKs, for Molecular Medicine, Room 454, 295 Congress Avenue, New Haven, CT 06510. mitogen-activated protein kinases; MEK, mitogen-activated protein/extracellular sig- E-mail address: [email protected] nal-related kinase kinase; ECL, enhanced chemiluminescence.

Copyright © 2000 by The American Association of Immunologists 0022-1767/00/$02.00 3838 IL-11 PROTECTS HUMAN EC FROM CTL AND Ab PLUS C

This additional phosphorylation may potentiate STAT function as Immunoprecipitation and immunoblotting an activator of transcription. To analyze protein by immunoblotting of cultured cell lysates, cells (ECs Immune and inflammatory processes depend on cytokine-medi- and K562) were washed twice with ice-cold PBS containing 1 mM sodium ated activation of vascular endothelium (e.g., new adhesion mol- orthovanadate and 1 mM sodium fluoride and were then lysed in 100 ␮l ecule expression via an NF-␬B-dependent process) (31) and often cold RIPA lysis buffer (PBS, 1% Nonidet P-40, 0.5% sodium deoxy- ␮ injure vascular endothelium, resulting in additional epithelial in- cholate, 0.1% SDS, 1 mM PMSF, 10 g/ml leupeptin, 1 mM sodium or- thovanadate). Cell lysates were clarified by centrifugation at 10,000 ϫ g jury as a consequence of ischemia which follows endothelial dam- for 15 min, and protein concentrations of the supernatant were determined age. Therefore, we investigated whether IL-11 could act on cul- by using a Bio-Rad assay kit (Bio-Rad, Hercules, CA). Lysates were pre- tured HUVECs to inhibit proinflammatory functions or to induce pared for SDS-PAGE by adding an equal volume of 2ϫ SDS-PAGE sam- cytoprotection. We report here that HUVECs express functional ple buffer (100 mM Tris-Cl (pH 6.8), 200 mM DTT, 4% SDS, 0.2% brom- phenol blue, 20% glycerol) and heating the mixture in a boiling water bath IL-11 receptors and respond to IL-11 by activating both STAT and for 3 min. Aliquots (10 ␮g) of cell lysate were resolved by SDS-PAGE MAPK signaling pathways. Pretreatment of HUVECs with low using 8% acrylamide gels and a Tris-glycine electrophoresis buffer system dose IL-11 can protect HUVECs from immune-mediated injury, (25 mM Tris, 250 mM glycine, 0.1% SDS, pH 8.3), and separated proteins measured as cytolytic CTL- or Ab plus C-mediated lysis, but did were transferred to a polyvinylidene difluoride membrane by electrophore- not inhibit TNF-induced activation of NF-␬B or expression of leu- sis (Immobilon P, Millipore, Bedford, MA). Membranes were incubated with blocking solution containing 5% nonfat dry milk in Tris-buffered sa- kocyte adhesion molecules. line-Tween (TBST) (10 mM Tris-HCl (pH 8.0), 0.150 mM NaCl, 0.05% Tween 20) at room temperature for 30 min followed by incubation with TBST containing the indicated Ab overnight at 4°C. Membranes were Materials and Methods washed and incubated with a suitable HRP-conjugated detecting reagent

Cytokines, drugs, and Abs and other reagents (Jackson ImmunoResearch, West Grove, PA), and HRP activity was de- Downloaded from tected using an enhanced chemiluminescence (ECL) kit according to the Recombinant human IL-11, neutralizing mouse mAb to IL-11, and three manufacturer’s instructions (Pierce, Rockford, IL). Autoradiograms were different mouse mAbs to IL-11R␣-chain were provided by Genetics Insti- scanned by laser densitometry. tute (Andover, MA). Recombinant human and recombinant For immunoprecipitation before immunoblotting of gp130, 500 ␮gof human IFN-␥ were purchased from R&D Systems (Minneapolis, MN). total cell lysate were precleared by incubation with 4 ␮g rabbit IgG for 90 Recombinant human TNF was a gift from Biogen (Cambridge, MA). Fi- min on rotator at room temperature, followed by addition of 50 ␮l Gam- broblast -1, commonly called endothelial cell growth supple- maBind G-Sepharose beads (Pharmacia, Piscataway, NJ) with continual ment (ECGS), was obtained from Collaborative Research/Becton Dickin- http://www.jimmunol.org/ incubation on a rotator at room temperature for an additional 90 min. The son (Bedford, MA) and used in conjunction with porcine intestinal heparin beads were removed from the precleared lysates by centrifugation. To form (Sigma, St. Louis, MO). The pharmacological inhibitor of mitogen-acti- specific immune complexes, 4 ␮g anti-gp130 Ab were added to the pre- vated protein/extracellular signal-related kinase kinase (MEK-1) cleared lysate which was then incubated for 90 min on a rotator at room (PD98059) was obtained from Calbiochem (La Jolla, CA), and the protein temperature. To collect specific immune complexes, 50 ␮l GammaBind synthesis inhibitor cycloheximide was obtained from Sigma. Rabbit poly- G-Sepharose beads were added, and the sample was incubated on a rotator clonal Abs reactive with STAT1, phosphotyrosine-STAT1, STAT3, phos- at 4°C overnight at which time beads were collected by centrifugation at photyrosine-STAT3, p42, and p44 MAPK and phosphothreonine/phospho- 13,000 ϫ g for 1 min. The beads containing immune complexes were tyrosine p42 and p44 MAPK were purchased from New England Biolabs washed five times with PBS. The immune complexes were solubilized (Beverly, MA). Rabbit polyclonal Ab to gp130 and mouse mAb to phos- from the beads by addition of 1ϫ SDS-PAGE sample buffer and heated in photyrosine were purchased from Upstate Biotechnology (Lake Placid,

a boiling water bath for 5 min. Aliquots were resolved by SDS-PAGE and by guest on October 2, 2021 NY). Rabbit polyclonal Ab to I-␬B␣ was obtained from Santa Cruz Bio- immunoblotted for total gp130 and for phosphotyrosine residues, as de- technology (Santa Cruz, CA). Anti-class I MHC mAb (W6/32) was pre- scribed above. pared as ascites in our laboratory from a clone provided by Dr. Jack Strominger (Harvard University, Cambridge, MA), and FITC-conjugated anti-class I MHC mAb (W6/32) was purchased from Serotech (Oxford, Indirect immunofluorescence and FACS analysis U.K.). Mouse anti-E-selectin mAb (clone H14/18) and nonbinding (K16/ 16) Ab were made as ascites in our laboratory. Mouse anti-ICAM-1 mAb After treatment with cytokines, HUVECs were washed with HBSS and (2D5) was a gift from Dr. Dario Altieri (Yale University, New Haven, CT). incubated for 1 min with trypsin-EDTA. Detached cells were collected and FITC-conjugated polyclonal goat anti-mouse Ab was purchased from washed twice with ice-cold PBS containing 1% BSA and were incubated Boehringer Mannheim (Indianapolis, IN). Baby rabbit C was purchased with specific primary mAb (either anti-E-selectin, anti-ICAM-1, or FITC- from Pel-Freez (Brown Dear, WI). conjugated anti-class I MHC) for 30 min at 4°C. Replicate aliquots were incubated with nonbinding isotype control mAb. Labeled cells were Cell isolation and culture washed twice with PBS-1% BSA and were fixed with 2% paraformalde- hyde before being analyzed. In the case of E-selectin and ICAM-1, cells Human EC were isolated from umbilical veins as previously described (32, were incubated with a FITC-conjugated goat anti-mouse Ab for 30 min on 33) and cultured on gelatin (J. T. Baker, Phillipsburg, NJ)-coated tissue ice followed by washing twice with PBS-1% BSA before fixation. After culture plastic at 37°C in 5% CO2, humidified air in Medium 199 contain- fixation, cells were analyzed by FACS using a FACSort and Lysis II soft- ␮ ing 20% FBS, 2 mM L-glutamine, 100 U/ml penicillin, 100 g/ml strep- ware (Becton Dickinson, San Jose, CA). Corrected mean fluorescence val- ␮ tomycin (all from Life Technologies, Grand Island, NY), 50 g/ml ECGS ues was calculated as follows. For each treatment, the mean fluorescence ␮ and 100 g/ml porcine intestinal heparin. K562 and CACO-2 cells were value for the isotype-matched nonbinding control Ab was subtracted from obtained from the American Type Culture Collection (Rockville, MD; ac- the mean fluorescence value for the specific Ab. cession number for K562, CCL-243) and cultured at 37°C in 5% CO2- humidified air in RPMI 1640 (Life Technologies) containing 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin, and 100 ␮g/ml streptomycin. Transfection and promoter reporter gene assays RNase protection assay Transient transfection of HUVECs were performed using a DEAE-dextran protocol as described previously (34). Cells were transfected with both a RNA was harvested from HUVECs, K562 cells, and CACO-2 cells using ␬B-luciferase promoter reporter gene, which contains two ␬B sites from an RNeasy kit (Qiagen, Santa Clarita, CA), and 4 ␮g of each RNA were Ig␬ enhancer (35), and a constitutively active ␤-galactosidase expression incubated with a 32P-labeled probe cocktail against human-IL-11R␣-chain construct (Promega, Madison, WI). Cell lysates were assayed for luciferase and GAPDH as loading control (Riboquant kit and custom template, and ␤-galactosidase activities using the Promega reporter assay system PharMingen, San Diego, CA). Hybridization reactions were incubated (Promega). Luciferase activity was measured using a Berthold (Schwar- overnight at 56°C and then digested with RNAs A/T1 and proteinase K. zwald, Germany) model LB9501 luminometer, and ␤-galactosidase activ- Protected fragments were precipitated and separated using a 6% acryl- ity was determined by measuring absorbance at 420 nm using a spectro- amide Tris-borate EDTA (TBE)-urea gel and then visualized by autora- photometer (Dynatech Laboratories, Easton, MA). Luciferase values in diography. Autoradiograms were scanned with a laser densitometer (Mo- relative light units were normalized to ␤-galactosidase units to control for lecular Dynamics, Sunnyvale, CA) transfection efficiency. The Journal of Immunology 3839

FIGURE 2. IL-11 induces tyrosine phosphorylation of gp130 in HU- VECs. HUVECs were either untreated (control) or treated with IL-11 (100 ng/ml) or oncostatin M (20 ng/ml) for 2 and 10 min. Cell lysates were immunoprecipitated with specific Ab to gp130. Immune complexes were FIGURE 1. A, RNase protection assay for IL-11R␣. Total RNA from resolved on SDS-PAGE and immunoblotted with a phosphotyrosine-spe- cultured HUVECs, CACO-2, or K562 (4 ␮g) cells was incubated overnight cific Ab. Results are quantitated by densitometry. One of two independent with probes for human IL-11R␣ and GAPDH genes. Samples were di-

experiments with similar results. Downloaded from gested with RNase A/T1 and protected fragments (321 bp for IL-11R␣ and 96 bp for GAPDH) were resolved on a 6% acrylamide/TBE-urea gel. One of two independent experiments with similar results. B, Immunoblot of ␣ IL-11R -chain protein. Lysates from either HUVECs or K562 cells were 2mML-glutamine, 100 U/ml penicillin, and 100 ␮g/ml streptomycin. resolved on SDS-PAGE and immunoblotted with specific Ab to IL-11R␣. Where indicated, 20 ␮g/ml cycloheximide were included during the incu- One of three independent experiments with similar results. Results are bation with IL-11. To initiate cytotoxicity, cells were incubated with anti- quantitated by densitometry. MW, molecular weight. class I MHC mAb (W6/32) for 30 min at 37°C. Baby rabbit C was then added at the indicated dilutions. To measure the extent of lysis after 1 h, http://www.jimmunol.org/ supernatant was removed and transferred into a flat-bottom 96-well plate C-mediated lysis and released calcein was quantitated using a fluorescence multiwell plate reader (Cytofluor 2, PerSeptive Biosystems, Cambridge, MA; excitation Target HUVECs were grown in 96-well plates to confluence and incubated wavelength, 485 nm; emission wavelength, 530 nm). Replicate wells were with 20 ␮M calcein-AM (Molecular Probes, Eugene, OR) in Medium 199 incubated with lysis buffer (50 mM sodium borate, 0.1% Triton X-100, pH and 5 mM HEPES for 30 min at 37°C. The medium was replaced by 9.0) to determine maximum release or without C treatment to determine complete EC growth medium, and cells were rested overnight. IL-11 in spontaneous release. Percent specific lysis was calculated as [(sample re- media or media alone were added for an addition of 6 h, after which cells lease Ϫ spontaneous release)/(maximal release Ϫ spontaneous release)] ϫ were washed twice with Medium 199 containing 5% FBS, 5 mM HEPES, 100%. Spontaneous release was generally Ͻ25%. by guest on October 2, 2021

FIGURE 3. Dose-dependent phosphorylation of STAT3 and STAT1 by IL-11 in HUVECs. HUVECs were untreated (control), treated with increasing concentration of IL-11, or treated with increasing concentrations of oncostatin M for 10 min. Lysates were resolved on SDS-PAGE and immunoblotted with specific Ab to either phospho-STAT3 (P-STAT3), STAT3, phospho-STAT1 (P-STAT1), or STAT1. Results are quantitated by densitometry. One of two independent experiments with similar results. 3840 IL-11 PROTECTS HUMAN EC FROM CTL AND Ab PLUS C Downloaded from

FIGURE 4. Time-dependent phosphorylation of STAT3 and STAT1 by IL-11 in HUVECs. HUVECs were either untreated (control), treated with IL-11 (100 ng/ml), or treated with oncostatin M (200 pg/ml) for various time as indicated. Lysates were resolved on SDS-PAGE and immunoblotted with specific

Ab to either phospho-STAT3 (P-STAT3) or phospho-STAT1 (P-STAT1). Results are quantitated by densitometry. One of two independent experiments http://www.jimmunol.org/ with similar results.

Generation of allospecific CTLs and CTL killing assay described above. Percent specific lysis was calculated as described above. Spontaneous release was generally Ͻ25%. Alloreactive class I-restricted CD8ϩ T cells clones were produced as de- scribed elsewhere from peripheral blood CD8ϩ T cells (36, 37). To deter- mine percent lysis, target HUVECs were loaded with calcein as described above and incubated overnight. IL-11 in media or media alone were added Results either overnight or 6 h before assay, after which cells were washed twice HUVECs express IL-11R␣ by guest on October 2, 2021 with Medium 199 containing 5% FBS, 5 mM HEPES, 2 mM L-glutamine, 100 U/ml penicillin, 100 ␮g/ml streptomycin. Where indicated, the MEK-1 The functional receptor for IL-11 is a heterodimer composed of a inhibitor PD98059 or vehicle control was added 90 min before adding signal transduction subunit, gp130, and a specific ligand-binding IL-11. Effector CTL cells were added in a total volume of 150 ␮l/well at receptor component called IL-11R␣ (18). HUVECs have previ- varying E:T ratios. In some experiments, 1 ␮g/ml PHA (Sigma) was added ously been reported to express gp130 and respond to leptin (38) to potentiate killing by lectin bridging. Replicate wells were incubated with lysis buffer (50 mM sodium borate, 0.1% Triton X-100, pH 9.0) to deter- and oncostatin M (39–42), two cytokines that utilize the gp130 mine maximum release or with media alone to determine spontaneous re- signal transducer. We used an RNase protection assay to assess the lease. After 4 h incubation at 37°C, released calcein was measured as presence of IL-11R␣ mRNA in HUVECs. A human myeloid cell

FIGURE 5. Dose- and time-dependent phosphorylation of p42 and p44 MAPK by IL-11 in HUVECs. HUVECs were cultured in Medium 199 containing 1% FCS for 17 h, after which the medium was removed and cells were rested for2hintheMedium 199 containing no FCS. A, HUVECs were either untreated (control) or treated with increasing concentrations of IL-11 for 10 min. One of two independent experiments with similar results. B, HUVECs were either untreated (control), treated with IL-11 (200 ng/ml) for various times as indicated, or treated with oncostatin M (OnM) (1 ng/ml) for 40 min. Lysates were resolved on SDS-PAGE and immunoblotted with specific Ab to either phospho-p44/p44 Ab (P-42/44) or p42/p44 Ab (p42/p44). Results are quantitated by densitometry. One of two independent experiments with similar results. The Journal of Immunology 3841 line, K562, was used as a positive control because these cells ex- press IL-11R␣ (43). As shown in Fig. 1A, both HUVECs and K562 cells express mRNA for IL-11R␣. In contrast, IL-11R␣ mRNA was not detected in CACO-2 cells which are unresponsive to IL-11. To assess the presence of IL-11R␣-chain protein in HUVECs we performed an immunoblot using three newly generated anti- IL-11R␣ mAbs. As shown in Fig. 1B, we detected a single major immunoreactive species in K562 cells and in HUVECs with ap- parent molecular mass of ϳ80 kDa. The same results were ob- tained with the other two mAbs (data not shown). The RNase protection assay and immunoblot indicated that HUVECs consti- tutively express IL-11R␣ mRNA and protein, respectively.

IL-11 causes tyrosine phosphorylation of gp130 in HUVECs Binding of IL-11 to IL-11R␣ is necessary, although not sufficient for signal transduction. Previous studies have indicated that IL-11

utilizes JAK-induced tyrosine phosphorylation of gp130 to initiate Downloaded from signal transduction (18, 20). HUVECs express gp130 and use it to transduce signals in response to an oncostatin M (40, 44, 45) or leptin (38), two other IL-6 families of cytokines. We were unable to detect IL-11-induced tyrosine phosphorylation of gp130 by im- munoblot of HUVECs lysates, although oncostatin M response could be detected by this approach (data not shown). To increase http://www.jimmunol.org/ the sensitivity of the assay, we immunoprecipitated gp130 before immunoblotting with phosphotyrosine Ab. In this experiment, HU- VECs were stimulated with either IL-11 or oncostatin M for 2 or 10 min, and lysed. The lysates were immunoprecipitated with gp130 Ab, and the anti-gp130 immunoprecipitates were subjected to SDS-PAGE followed by immunoblotting with Abs to phospho- tyrosine (Fig. 2). In the absence of stimulation, HUVECs express no detectable tyrosine phosphorylation of gp130. However, ty- rosine phosphorylation of gp130 was reproducibly detected after 2 by guest on October 2, 2021 or 10 min of stimulation with IL-11. Oncostatin M, used as a positive control, was a more potent inducer of gp130 phosphory- lation than IL-11. FIGURE 6. Inhibition of IL-11-induced phosphorylation of MAPK by the MEK-1 inhibitor PD98059. HUVECs were pretreated with the indi- cated concentration of MEK-1 inhibitor for 90 min before treatment with IL-11 induces tyrosine phosphorylation of STAT3 and STAT1 IL-11 or ECGS. Treatments and analyses are the same as described in Fig. 5. Values are results of two independent experiment with similar results. Cytokines that signal through gp130 typically cause JAK-mediated Results are quantitated by densitometry. tyrosine phosphorylation of STAT3 and, to a lesser extent, of STAT1 (22, 28). Therefore, we used immunoblotting to determine whether IL-11 induces tyrosine phosphorylation of STAT3 and STAT1 in HUVECs. In these experiments, the specific anti- phospho-STAT Abs we used only detect phosphorylation that oc- munoblot analysis of the same samples with anti-STAT3 and anti- curs on the specific tyrosine residues involved in JAK-mediated STAT1 Abs (Fig. 3) indicated that the total levels of STAT3 and activation of STAT dimerization, and thus immunoreactivity can STAT1 proteins, respectively, were not changed by IL-11 treat- be used to infer STAT activation, although this cannot be directly ment (i.e., phosphorylated plus nonphosphorylated). In parallel tested without knowledge (and analysis) of a STAT3-dependent control experiments, oncostatin M induced detectable tyrosine gene. No tyrosine phosphorylation of either STAT3 or STAT1 was phosphorylation of STAT3 and STAT1 at 60 pg/ml and 1.5 ng/ml, detected in unstimulated HUVECs by immunoblot analysis using respectively (Fig. 3). These data show that IL-11 is a less potent Ab to phospho-STAT3 or phospho-STAT1 (Fig. 3, control). After inducer of STAT3 and STAT1 phosphorylation than is oncostatin 10 min of cytokine treatment, tyrosine phosphorylation of STAT3 M. Moreover, even maximal concentrations of IL-11 failed to in- was induced by 0.1 ng/ml IL-11 and increased, in a dose-depen- duce levels of STAT phosphorylation observed at optimal oncosta- dent manner, to maximal levels at ϳ50 ng/ml IL-11 (Fig. 3). IL-11 tin M levels. In other words, IL-11 is also less efficacious than induced tyrosine phosphorylation of STAT1 only at concentrations oncostatin as an activator of the JAK-STAT signaling pathway in of Ͼ10 ng/ml (Fig. 3). Tyrosine phosphorylation of STAT1 was HUVECs. Nevertheless, IL-11 clearly does activate gp130/JAK/ increased at higher concentrations of IL-11, peaking at ϳ50 ng/ml STAT3 and STAT1 pathways in this cell type. IL-11 (Fig. 3). These experiments suggest that the threshold con- We next determined the time course of IL-11 actions. Increased centration for an IL-11 effect on STAT1 phosphorylation was tyrosine phosphorylation of STAT3 after addition of IL-11 was Ͼ100-fold higher than for phosphorylation of STAT3, although seen as early as 2.5 min and largely disappeared by 60 min (Fig. maximal phosphorylation occurred at similar concentrations. Im- 4). Similarly, phospho-STAT1 was detected after 2.5 min stimu- 3842 IL-11 PROTECTS HUMAN EC FROM CTL AND Ab PLUS C

FIGURE 7. IL-11 neither induces the degrada- tion of I␬B␣ nor inhibits the effect of TNF on this response. A, HUVECs were treated with IL-11 (10 ng/ml), IL-11 (100 ng/ml), or TNF (10 U/ml) for 5 and 10 min. B, HUVECs were pretreated with media alone (control) or media containing various doses of IL-11 (as indicated). After 4 h, cells were treated with media alone or media containing 10 U/ml TNF for 15 min. Lysates were resolved on 10% SDS- PAGE and immunoblotted with specific Ab to I-␬B␣. Cell lysates from HUVECs pretreated with 500 and 5000 ng/ml IL-11 were run on a separate gel. Values are the results of two independent ex- periment with similar results. Results are quantitated by densitometry Downloaded from http://www.jimmunol.org/ lation with IL-11 and disappeared at 40 min (Fig. 4). Replicate gels IL-11R␣ Ab have been identified to date, so we could not assess probed for STAT3 (data not shown) and STAT1 (data not shown) the role of the receptor in this response. indicated that there were no changes in expression of these pro- teins after stimulation with IL-11 over the same time interval. In- IL-11 activates p42 and p44 MAPK in a dose- and time- terestingly, this kinetic study indicated that IL-11 acted more rap- dependent manner in HUVECs idly than oncostatin M, which did not induce detectable STAT In addition to activation of STAT3 and STAT1, the MAPK sig- phosphorylation until 5 min. naling pathway has also been reported to be activated by cytokines by guest on October 2, 2021 To be certain that IL-11-induced STAT signaling in HUVECs that signal through gp130 (15). Specifically, IL-11 has been re- was due to IL-11, we assessed the phosphorylation of STAT3 by ported to activate both p42 and p44 MAPK in K562 cells, in mono- IL-11 in the presence of neutralizing Ab to IL-11. Preincubation of cytic leukemia U937 cells, and in 3T3-L1 adipocytes (29, 30). To IL-11 with mouse mAb anti-IL-11 Ab, but not with control Ab, assess the potential involvement of MAPK signaling pathway in completely abolished IL-11-induced phosphorylation of STAT3. the IL-11 signal transduction in HUVECs, we determined whether The same Ab had no effect on oncostatin M-induced STAT3 phos- IL-11 increases threonine/tyrosine phosphorylation of p42 and p44 phorylation (data not shown). Unfortunately, no inhibitory anti- using Ab specific for the phosphorylated forms of these enzymes. These MAPKs are continuously activated in our cultured HUVECs through the actions of serum and growth factors necessary for cell culture. Therefore, to assess an effect of IL-11 on MAPK activa- tion, we first made the cells quiescent by removing growth factors and lowering serum levels to 1%. Under these conditions, HU- VECs contain minimal phospho-p42/p44 (Fig. 5, control). Addi- tion of 0.3 ng/ml IL-11 caused detectable phosphorylation of p42/ p44 above background, and the level of phosphorylation increased with increasing concentration of IL-11 (Fig. 5A). Lower concen- trations of IL-11 (Ͻ0.3 ng/ml) did not cause significant phosphor- ylation of p42/p44 (data not shown). In time course experiments, phospho-p42/p44 was detected by 10–15 min but decreased at 30 min and largely disappeared after 40 min of stimulation with IL-11 (Fig. 5B). However, significant levels of phospho-p42/p44 were still detected after 40 min of stimulation with oncostatin M (1 FIGURE 8. IL-11 does not inhibit TNF-mediated ␬B-luciferase pro- ng/ml) (Fig. 5B). No change in the total p42/p44 protein levels moter-reporter gene activity. HUVECs were transiently cotransfected with were observed on stimulation with IL-11 over this time period ␬B-luciferase promoter-reporter gene and a ␤-galactosidase expression (Fig. 5). These data suggest that IL-11 can activate MAPK signal- construct. Cell were treated with different doses of IL-11 as indicated. After 4 h, cells were left untreated or treated with 3 U/ml TNF for 18 h. Lucif- ing in HUVECs. In growth factor responses, the threonine/tyrosine erase activity was expressed as light units normalized to ␤-galactosidase phosphorylation of p42 and p44 MAPK is mediated by MEK-1. activity (RLU). Data are presented as the mean Ϯ SEM of triplicates in Both IL-11- and ECGS-induced phosphorylation of p42/p44 each group from one experiment. Shown is one of the three different ex- MAPK are comparably reduced by the MEK-1 inhibitor PD98059 periments with similar outcome. in a dose-dependent manner (Fig. 6). The Journal of Immunology 3843

Table I. E-selectin and ICAM-1 expression on HUVECs after 6 h treatment with TNF, IL-11, or TNF plus IL-11a

Expt. 1 Expt. 2

E-selectin ICAM-1 E-selectin ICAM-1 Treatment (cMFI) (cMFI) (cMFI) (cMFI)

None 5 9 12 19 TNF (1 U/ml) 186 81 310 154 IL-11 (100 ng/ml) 9 10 11 23 IL-11 (1000 ng/ml) 12 11 12.8 15 TNF (1 U/ml) ϩ IL-11 (100 ng/ml) 341 105 340 304 TNF (1 U/ml) ϩ IL-11 (1000 ng/ml) 211 109 330 142

a HUVECs were treated with cytokines as indicated. Surface levels of E-selectin and ICAM-1 were quantitated by indirect immunofluorescence using H4/18 and 2D5 mAb, respectively. The numbers are expressed as corrected mean fluorescence intensity (cMFI) calculated by subtracting the mean fluorescence value for isotype-matched control mAb (K16/16) from the mean fluorescence value for either H4/18 or 2D5 mAb. Results from two independent experiments are shown.

IL-11 does not influence the proinflammatory NF-␬B signaling IL-11 at various doses did not increase the transcription of a tran- pathway in HUVECs siently transfected ␬B-promoter-reporter gene construct (Fig. 8). Downloaded from Many proinflammatory cytokines activate the transcription factor Moreover, pretreatment with various doses of IL-11 did not inhibit ␬ NF-␬B (46). The activation of NF-␬B has also been associated activation of the B-promoter-reporter gene induced by a sub- with cytoprotection against TNF-induced apoptosis (47, 48). On maximal dose of TNF (3 U/ml) (Fig. 8). ␬ the other hand, part of the antiinflammatory actions of IL-11 have The TNF-induced NF- B-dependent responses in HUVECs in- been attributed to inhibition of NF-␬B activation in LPS-treated clude the up-regulation of leukocyte adhesion molecules such as macrophages (12). TNF is the prototypic activator of the NF-␬B E-selectin and ICAM-1 (31). To assess whether IL-11 modulates http://www.jimmunol.org/ pathway in HUVECs, causing rapid degradation of I-␬B proteins inflammatory functions of ECs, we examined the effects of IL-11 (49, 50). We initially assessed the effects of IL-11 on NF-␬B ac- on E-selectin and ICAM-1 expression on HUVECs using indirect tivation in the presence or absence of TNF by measuring the dis- immunofluorescence. IL-11 had no direct effect on E-selectin or appearance of I-␬B␣ by immunoblotting. As shown in Fig. 7A, ICAM-1 expression (Table I). In contrast, a submaximal concen- IL-11 by itself, at either 10 or 100 ng/ml, did not cause degradation tration of TNF (1 U/ml) significantly induced both E-selectin and of I-␬B␣. Moreover, pretreatment of HUVECs cultures for 4 h ICAM-1 expression at 5 h (Table I). Neither 6 h pretreatment (data with various doses of IL-11 did not inhibit I-␬B␣ degradation in- not shown) nor cotreatment (Table I) with IL-11 inhibited E-se-

duced by TNF (Fig. 7B). Consistent with these biochemical data, lectin or ICAM-1 expression induced by TNF. by guest on October 2, 2021

FIGURE 9. IL-11-treated HUVECs acquire re- sistance against CTL-mediated cytolysis. Allospe- cific CTL clones were generated as described (see Materials and Methods). Confluent target HUVECs were pretreated overnight with either media (con- trol) or IL-11 at indicated concentrations. Calcein release-based CTL killing assays was performed (see Materials and Methods).Results are from four different combinations of CTL and HUVEC at var- ious E:T ratios. Data are presented as the mean Ϯ p Ͻ ,ء p Ͻ 0.01 and ,ءء .(SEM (n ϭ 3 in each group 0.05 control vs IL-11 as determined by t test. 3844 IL-11 PROTECTS HUMAN EC FROM CTL AND Ab PLUS C

Table II. Class I MHC expression on HUVECs after 18 h treatment with IFN-␥, IL-11, or IFN␥ plus IL-11a

Class I MHC Expression (cMFI)

Treatment Expt. 1 Expt. 2

None 93 90 IL-11 (50 ng/ml) 92 93 IFN-␥ (100 ng/ml) 130 126 IL-11 (50 ng/ml) ϩ IFN-␥ (100 ng/ml) 127 130

a After overnight treatment with cytokines as indicated, HUVECs were FIGURE 10. IL-11-treated HUVECs acquire resistance against C-me- trypsinized and stained for class I MHC. The numbers are expressed as corrected diated cytolysis. Confluent pooled HUVECs were loaded with calcein, mean fluorescence intensity (cMFI) calculated by subtracting the mean fluorescence chased, and pretreated for 6 h without (control) or with 0.5 ng/ml IL-11 in value for isotype-matched control Ab. the presence (B) or absence (A) of cycloheximide (20 ␮g/ml). HUVECs were washed and incubated with anti-W6/32 Ab (2.5 ␮g/ml) followed by IL-11-pretreated HUVECs are partially resistant to CTL- and addition of baby rabbit C at the indicated dilutions. After 1 h, supernatant C-mediated cytolysis was harvested and released calcein was measured. Percent specific lysis was calculated (see Materials and Methods). Maximal and spontaneous We next assessed whether IL-11 could protect HUVECs from im- release was significantly different for cycloheximide-treated cells and con- mune-mediated injury in two assays: killing by class I MHC-re- trols. These values were therefore determined for the two groups sepa- Downloaded from stricted allospecific CTL clones; and killing by anti-class I MHC rately, and sample release was normalized to these values. Data are pre- mAb plus C. Pretreatment with IL-11 at 0.5 ng/ml was capable of sented as the mean Ϯ SEM of triplicates in each group from one partially protecting HUVECs from lysis by CTL, but pretreatment experiment. Shown is one of the three different experiments with similar .p Ͻ 0.01 control vs IL-11 as determined by t test ,ءء .with a higher concentration of IL-11 (50 ng/ml) failed to do so outcome (Fig. 9). Preincubation of the IL-11 (0.5 ng/ml) with anti-IL-11 mAb, but not with control K16/16 Ab, completely inhibited the ␣

toprotection. We found that cultured HUVECs express IL-11R http://www.jimmunol.org/ cytoprotective effect of IL-11 in this assay (data not shown). Cy- mRNA and protein as well as gp130 and that stimulation of HU- toprotection is not mediated by reduction in the expression of tar- VECs with IL-11 induced rapid phosphorylation of gp130, get molecule (i.e., class I MHC) because expression of these mol- STAT3, STAT1, and p42/p44 MAPKs. We also found that IL-11 ecule is not altered (Table II). IL-11 pretreatment also reduced pretreatment can induce HUVECs to become partially resistant to cytolysis by CTL that has been potentiated by PHA lectin bridging injury mediated by CTL or mAb plus C. Both p42/p44 MAPK (Table III). Pretreatment of HUVEC targets with the MEK-1 in- phosphorylation and protection from CTL were reduced by the hibitor PD98059 also reduced CTL-mediated killing. However, in MEK-1 inhibitor PD98059. However, this drug independently re- this case, no further protection could be conferred by IL-11. These duced killing; therefore, this connection must be regarded with data support a link between IL-11 cytoprotection and p42/p44 caution. In contrast, IL-11 does not inhibit proinflammatory re- by guest on October 2, 2021 MAPK activation, but this should be interpreted with caution be- sponses of HUVECs to TNF, such as NF-␬B activation or adhe- cause of the independent actions of the MEK-1 inhibitor. sion molecule expression. These data provide the first in vitro ev- Finally, we determined whether IL-11 can protect HUVECs idence of a direct cytoprotective action of IL-11 and do not support against C-mediated lysis initiated by an anti-MHC class I mAb and an antiinflammatory effect of this cytokine on endothelium. A sim- baby rabbit C. Pretreatment with IL-11 at 0.5 ng/ml significantly ilar lack of antiinflammatory effect of IL-11 was observed when reduced percent specific lysis induced by mAb plus 3 and 10% C LPS rather than TNF was used as stimulus for endothelial activa- (Fig. 10A). The action of IL-11 quantitated by dye release at 1 h tion (K.M. and J.S.P., unpublished observations). seems to be truly protective because it correlated with cell survival Functional receptor complexes for the IL-6 family of cytokines 24 h after mAb plus C treatment determined by a Hoechst staining including IL-11, oncostatin M, and IL-6 share gp130 as a compo- assay (data not shown). The protective effect of IL-11 pretreatment nent critical for signal transduction (16, 17). IL-11 binds to IL- is dependent on protein synthesis because it was not observed 11R␣ on the cell surface, and the IL-11/IL-11R␣ complex then when cycloheximide was present during the pretreatment period associates with gp130, causing it to cluster. This is essentially the (Fig. 10B). same mechanism of action that has been observed for IL-6/IL-6R signaling. Oncostatin M differs from IL-11 and IL-6 in that it di- Discussion rectly binds to gp130 and signals through either gp130/leukemia- The present study was undertaken to determine whether IL-11 can inhibitory factor receptor ␤ or gp130/oncostatin M receptor het- directly act on vascular EC to inhibit inflammation or induce cy- erodimers (51, 52). Because gp130 is ubiquitously expressed, the

Table III. Effect of MEK-1 inhibition on IL-11-induced resistance to CTLa

% Specific Lysis (mean Ϯ SEM)

IL-11 ϩ Pretreatment CTL Clone E:T Ratio Mock IL-11 PD98059 PD98059

Expt. 1 (no PHA) 8.2 15 60 Ϯ 15 39 Ϯ 841Ϯ 958Ϯ 8 2.11 25 5 Ϯ 21Ϯ 11Ϯ 110Ϯ 2

Expt. 2 (1 ␮g/ml PHA) 8.11 45 35 Ϯ 723Ϯ 626Ϯ 128Ϯ 6 8.1 25 47 Ϯ 436Ϯ 230Ϯ 827Ϯ 8

a All determinations were performed in triplicate. IL-11 (0.5 ng/ml) was added 6 h before assay and PD98059 (30 ␮M) was added 7.5 h before assay. The Journal of Immunology 3845 responsiveness of cells to a particular cytokine of the IL-6 family 8. Qiu, B. S., C. J. Pfeiffer, and J. C. Keith, Jr. 1996. Protection by recombinant is determined by the relative expression of other receptor compo- human -11 against experimental TNB- induced colitis in rats. Dig. Dis. Sci. 41:1625. nents. Among receptors for IL-6 cytokines, endothelial cells lack 9. Hill, G. R., K. R. Cooke, T. Teshima, J. M. Crawford, J. C. Keith, Jr., IL-6R␣ (53) but express receptors for leptin (38) and oncostatin M Y. S. Brinson, D. Bungard, and J. L. Ferrara. 1998. Interleukin-11 promotes T (40, 44, 45). EC respond to oncostatin M with activation of MAPK cell polarization and prevents acute graft-versus-host disease after allogeneic bone marrow transplantation. J. Clin. Invest. 102:115. activity (39), IL-6 secretion (40), P-selectin (42), ICAM-1, and 10. Redlich, C. A., X. Gao, S. Rockwell, M. Kelley, and J. A. Elias. 1996. IL-11 E-selectin synthesis (41), and increased growth (39). In the present enhances survival and decreases TNF production after radiation-induced thoracic study, we showed that, compared with IL-11, oncostatin M is a injury. J. Immunol. 157:1705. much stronger inducer of gp130, STAT3, STAT1, and MAPK 11. Waxman, A. B., O. Einarsson, T. Seres, R. G. Knickelbein, J. B. Warshaw, R. Johnston, R. J. Homer, and J. A. Elias. 1998. Targeted lung expression of phosphorylation in HUVECs. Oncostatin M but not IL-11 appears interleukin-11 enhances murine tolerance of 100% oxygen and diminishes hy- to activate proinflammatory functions of HUVECs (41, 42, 44). peroxia-induced DNA fragmentation. J. Clin. Invest. 101:1970. Thus, not all gp130-signaling cytokines induce the same biological 12. Leng, S. X., and J. A. Elias. 1997. Interleukin-11 inhibits macrophage interleu- kin-12 production. J. Immunol. 159:2161. responses in EC, and some differences may relate to signal 13. Trepicchio, W. L., L. Wang, M. Bozza, and A. J. Dorner. 1997. IL-11 regulates strength. We have not been able to observe cytoprotection using macrophage effector function through the inhibition of nuclear factor-␬B. J. Im- oncostatin M (our unpublished observation), but this may be an munol. 159:5661. issue of finding a suitable concentration. 14. Trepicchio, W. L., M. Bozza, G. Pedneault, and A. J. Dorner. 1996. Recombinant human IL-11 attenuates the inflammatory response through down-regulation of The present study raises the question of how IL-11 treatment proinflammatory cytokine release and nitric oxide production. J. Immunol. 157: results in an injury-resistant state. The concentrations of cytokine 3627. that are protective correlate with those that activate STAT3 and 15. Taga, T., and T. Kishimoto. 1997. Gp130 and the interleukin-6 family of cyto-

kines. Annu. Rev. Immunol. 15:797. Downloaded from p42/p44 MAPK. Because cytoprotection depends on new protein 16. Zhang, X. G., J. J. Gu, Z. Y. Lu, K. Yasukawa, G. D. Yancopoulos, K. Turner, synthesis, it is reasonable to suppose that STAT3 and/or MAPK M. Shoyab, T. Taga, T. Kishimoto, R. Bataille, et al. 1994. Ciliary neurotropic lead to transcription of a cytoprotective protein. These two signals factor, interleukin 11, leukemia inhibitory factor, and oncostatin M are growth factors for human myeloma cell lines using the signal transducer may act coordinately; i.e., MAPK may phosphorylate STAT3 to gp130. J. Exp. Med. 179:1337. increase its transcription factor activity. However, we have not 17. Yang, Y. C., and T. Yin. 1995. Interleukin (IL)-11-mediated signal transduction. rigorously established a causal link between these signals and cy- Ann. NY Acad. Sci. 762:31. toprotection. There are no available agents to block STAT3 sig- 18. Nandurkar, H. H., D. J. Hilton, P. Nathan, T. Willson, N. Nicola, and http://www.jimmunol.org/ C. G. Begley. 1996. The human IL-11 receptor requires gp130 for signalling: naling. The best inhibitor of MAPK activation, the MEK-1 inhib- demonstration by molecular cloning of the receptor. Oncogene 12:585. itor PD98059, does block IL-11 effects but seems to have 19. Miyatake, T., K. Sato, K. Takigami, N. Koyamada, W. W. Hancock, H. Bazin, independent actions unrelated to IL-11 signaling. It is also unclear D. Latinne, F. H. Bach, and M. P. Soares. 1998. Complement-fixing elicited why higher concentrations of IL-11 are not protective. Such con- antibodies are a major component in the pathogenesis of xenograft rejection. J. Immunol. 160:4114. centrations activate STAT1, and it has been observed that STAT1 20. Yin, T., K. Yasukawa, T. Taga, T. Kishimoto, and Y. C. Yang. 1994. Identifi- may heterodimerize with STAT3, preventing STAT3 homodimer- cation of a 130-kilodalton tyrosine-phosphorylated protein induced by interleu- ization. Again, we cannot causally connect STAT1 activation to kin-11 as JAK2 tyrosine kinase, which associates with gp130 signal transducer. Exp. Hematol. 22:467. loss of cytoprotection. Transgenic and knockout mouse studies 21. Wang, X. Y., D. K. Fuhrer, M. S. Marshall, and Y. C. Yang. 1995. Interleukin-11 by guest on October 2, 2021 may be helpful in testing these hypotheses. induces complex formation of Grb2, Fyn, and JAK2 in 3T3L1 cells. J. Biol. We have provided the first evidence of a direct cytoprotective Chem. 270:27999. effect of IL-11 in vitro. The responsive target cell type, namely 22. Lutticken, C., U. M. Wegenka, J. Yuan, J. Buschmann, C. Schindler, A. Ziemiecki, A. G. Harpur, A. F. Wilks, K. Yasukawa, T. Taga, et al. 1994. vascular endothelium, could well be a primary site for injury in Association of transcription factor APRF and protein kinase Jak1 with the inter- many of the processes for which IL-11 is cytoprotective. Although leukin-6 signal transducer gp130. Science 263:89. our focus has been on immune-mediated injury, a primary mem- 23. Hemmann, U., C. Gerhartz, B. Heesel, J. Sasse, G. Kurapkat, J. Grotzinger, A. Wollmer, Z. Zhong, J. E. Darnell, Jr., L. Graeve, et al. 1996. Differential brane action, as we have proposed to explain our data, could ex- activation of acute phase response factor/Stat3 and Stat1 via the cytoplasmic tend to nonimmune models of cell injury as well. domain of the interleukin 6 signal transducer gp130. II. Src homology SH2 do- mains define the specificity of stat factor activation. J. Biol. Chem. 271:12999. 24. Zhong, Z., Z. Wen, and J. E. Darnell, Jr. 1994. Stat3: a STAT family member Acknowledgments activated by tyrosine phosphorylation in response to epidermal growth factor and We thank Louise Benson, Gwen Davis, and Lisa Gras for excellent tech- interleukin-6. Science 264:95. nical assistance in cell culture. 25. Ihle, J. N. 1996. STATs: signal transducers and activators of transcription. Cell 84:331. 26. Akira, S. 1997. IL-6-regulated transcription factors. Int. J. Biochem. Cell Biol. References 29:1401. 1. Weich, N. S., A. Wang, M. Fitzgerald, T. Y. Neben, D. Donaldson, J. Giannotti, 27. Zhang, X., J. Blenis, H. C. Li, C. Schindler, and S. Chen-Kiang. 1995. Require- J. Yetz-Aldape, R. M. Leven, and K. J. Turner. 1997. Recombinant human in- ment of serine phosphorylation for formation of STAT-promoter complexes. Sci- terleukin-11 directly promotes megakaryocytopoiesis in vitro. Blood 90:3893. ence 267:1990. 2. Orazi, A., R. J. Cooper, J. Tong, M. S. Gordon, L. Battiato, G. W. Sledge, Jr., 28. Boulton, T. G., Z. Zhong, Z. Wen, J. E. Darnell, Jr., N. Stahl, and J. A. Kaye, M. Kahsai, and R. Hoffman. 1996. Effects of recombinant human G. D. Yancopoulos. 1995. STAT3 activation by cytokines utilizing gp130 and interleukin-11 (Neumega rhIL-11 growth factor) on megakaryocytopoiesis in hu- related transducers involves a secondary modification requiring an H7-sensitive man bone marrow. Exp. Hematol. 24:1289. kinase. Proc. Natl. Acad. Sci. USA 92:6915. 3. de Haan, G., B. Dontje, C. Engel, M. Loeffler, and W. Nijhof. 1995. In vivo 29. Adunyah, S. E., G. C. Spencer, R. S. Cooper, J. A. Rivero, and K. Ceesay. 1995. effects of interleukin-11 and in combination with erythropoietin Interleukin-11 induces tyrosine phosphorylation, and c-jun and c-fos mRNA ex- in the regulation of erythropoiesis. Br. J. Haematol. 90:783. pression in human K562 and U937 cells. Ann. NY Acad. Sci. 766:296. 4. Kaye, J. A. 1996. Clinical development of recombinant human interleukin-11 to 30. Yin, T., and Y. C. Yang. 1994. Mitogen-activated protein kinases and ribosomal treat chemotherapy-induced thrombocytopenia. Curr. Opin. Hematol. 3:209. S6 protein kinases are involved in signaling pathways shared by interleukin-11, 5. Du, X., Q. Liu, Z. Yang, A. Orazi, F. J. Rescorla, J. L. Grosfeld, and interleukin-6, leukemia inhibitory factor, and oncostatin M in mouse 3T3–L1 D. A. Williams. 1997. Protective effects of interleukin-11 in a murine model of cells. J. Biol. Chem. 269:3731. ischemic bowel necrosis. Am. J. Physiol. 272:G545. 31. Pober, J. S., and R. S. Cotran. 1990. The role of endothelial cells in inflammation. 6. Orazi, A., X. Du, Z. Yang, M. Kashai, and D. A. Williams. 1996. Interleukin-11 Transplantation 50:537. prevents apoptosis and accelerates recovery of small intestinal mucosa in mice treated with combined chemotherapy and radiation. Lab. Invest. 75:33. 32. Gimbrone, M. A., Jr. 1976. Culture of vascular endothelium. Prog. Hemost. 7. Keith, J. C., Jr., L. Albert, S. T. Sonis, C. J. Pfeiffer, and R. G. Schaub. 1994. Thromb. 3:1. IL-11, a pleiotropic cytokine: exciting new effects of IL-11 on gastrointestinal 33. Thornton, S. C., S. N. Mueller, and E. M. Levine. 1983. Human endothelial cells: mucosal biology. Stem Cells 12(Suppl. 1):79. use of heparin in cloning and long-term serial cultivation. Science 222:623. 3846 IL-11 PROTECTS HUMAN EC FROM CTL AND Ab PLUS C

34. Karmann, K., W. Min, W. C. Fanslow, and J. S. Pober. 1996. Activation and 44. Frasca, D., C. Pioli, F. Guidi, S. Pucci, M. Arbitrio, G. Leter, and G. Doria. 1996. homologous desensitization of human endothelial cells by CD40 ligand, tumor IL-11 synergizes with IL-3 in promoting the recovery of the immune system after necrosis factor, and interleukin 1. J. Exp. Med. 184:173. irradiation [see comments]. Int. Immunol. 8:1651. 35. Min, W., S. Ghosh, and P. Lengyel. 1996. The -inducible p202 protein 45. Schieven, G. L., J. C. Kallestad, T. J. Brown, J. A. Ledbetter, and P. S. Linsley. as a modulator of transcription: inhibition of NF-␬B, c-Fos, and c-Jun activities. 1992. Oncostatin M induces tyrosine phosphorylation in endothelial cells and Mol. Cell. Biol. 16:359. activation of p62yes tyrosine kinase. J. Immunol. 149:1676. 36. Biedermann, B. C., and J. S. Pober. 1998. Human endothelial cells induce and 46. Ghosh, S., M. J. May, and E. B. Kopp. 1998. NF-␬B and Rel proteins: evolu- regulate cytolytic T cell differentiation. J. Immunol. 161:4679. tionarily conserved mediators of immune responses. Annu. Rev. Immunol. 16: 37. Biedermann, B. C., and J. S. Pober. 1999. Human vascular endothelial cells favor 225. clonal expansion of unusual alloreactive cytolytic T lymphocytes. J. Immunol. 47. Wang, C. Y., M. W. Mayo, R. G. Korneluk, D. V. Goeddel, and A. S. Baldwin, 162:7022. Jr. 1998. NF-␬B antiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and 38. Sierra-Honigmann, M. R., A. K. Nath, C. Murakami, G. Garcia-Cardena, c- IAP2 to suppress caspase-8 activation. Science 281:1680. A. Papapetropoulos, W. C. Sessa, L. A. Madge, J. S. Schechner, M. B. Schwabb, 48. Van Antwerp, D. J., S. J. Martin, I. M. Verma, and D. R. Green. 1998. Inhibition P. J. Polverini, et al. 1998. Biological action of leptin as an angiogenic factor. of TNF-induced apoptosis by NF-␬B. Trends Cell Biol. 8:107. Science 281:1683. 49. Collins, T., M. A. Read, A. S. Neish, M. Z. Whitley, D. Thanos, and T. Maniatis. 39. Faris, M., B. Ensoli, N. Kokot, and A. E. Nel. 1998. Inflammatory cytokines 1995. Transcriptional regulation of endothelial cell adhesion molecules: NF-␬B induce the expression of basic fibroblast growth factor (bFGF) isoforms required and cytokine-inducible enhancers. FASEB J. 9:899. for the growth of Kaposi’s sarcoma and endothelial cells through the activation of AP-1 response elements in the bFGF promoter. AIDS 12:19. 50. Johnson, D. R., I. Douglas, A. Jahnke, S. Ghosh, and J. S. Pober. 1996. A sus- ␬ ␤ ␬ 40. Brown, T. J., J. M. Rowe, J. W. Liu, and M. Shoyab. 1991. Regulation of IL-6 tained reduction in I B- may contribute to persistent NF- B activation in hu- expression by oncostatin M. J. Immunol. 147:2175. man endothelial cells. J. Biol. Chem. 271:16317. 41. Modur, V., M. J. Feldhaus, A. S. Weyrich, D. L. Jicha, S. M. Prescott, 51. Renne, C., K. J. Kallen, J. Mullberg, T. Jostock, J. Grotzinger, and S. Rose-John. G. A. Zimmerman, and T. M. McIntyre. 1997. Oncostatin M is a proinflammatory 1998. A new type of cytokine receptor antagonist directly targeting gp130. mediator: in vivo effects correlate with endothelial cell expression of inflamma- J. Biol. Chem. 273:27213. tory cytokines and adhesion molecules. J. Clin. Invest. 100:158. 52. Auguste, P., C. Guillet, M. Fourcin, C. Olivier, J. Veziers, 42. Yao, L., J. Pan, H. Setiadi, K. D. Patel, and R. P. McEver. 1996. or A. Pouplard-Barthelaix, and H. Gascan. 1997. Signaling of type II oncostatin M oncostatin M induces a prolonged increase in P-selectin mRNA and protein in receptor. J. Biol. Chem. 272:15760. Downloaded from human endothelial cells. J. Exp. Med. 184:81. 53. Romano, M., M. Sironi, C. Toniatti, N. Polentarutti, P. Fruscella, P. Ghezzi, 43. Cherel, M., M. Sorel, B. Lebeau, S. Dubois, J. F. Moreau, R. Bataille, R. Faggioni, W. Luini, V. van Hinsbergh, S. Sozzani, et al. 1997. Role of IL-6 S. Minvielle, and Y. Jacques. 1995. Molecular cloning of two isoforms of a and its soluble receptor in induction of and leukocyte recruitment. receptor for the human hematopoietic cytokine interleukin-11. Blood 86:2534. Immunity 6:315. http://www.jimmunol.org/ by guest on October 2, 2021