Signaling Events Inhibition on Lipopolysaccharide-Induced

The Proteasome as a Lipopolysaccharide-Binding Protein in Macrophages: Differential Effects of Proteasome Inhibition on Lipopolysaccharide-Induced This information is current as Signaling Events of October 1, 2021. Nilofer Qureshi, Pin-Yu Perera, Jing Shen, Guochi Zhang, Arnd Lenschat, Gary Splitter, David C. Morrison and Stefanie N. Vogel J Immunol 2003; 171:1515-1525; ; doi: 10.4049/jimmunol.171.3.1515 Downloaded from http://www.jimmunol.org/content/171/3/1515

References This article cites 52 articles, 38 of which you can access for free at: http://www.jimmunol.org/content/171/3/1515.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 by guest on October 1, 2021 • Fast Publication! 4 weeks from acceptance to publication

*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 © 2003 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

The Proteasome as a Lipopolysaccharide-Binding Protein in Macrophages: Differential Effects of Proteasome Inhibition on Lipopolysaccharide-Induced Signaling Events1

Nilofer Qureshi,2* Pin-Yu Perera,‡ Jing Shen,* Guochi Zhang,* Arnd Lenschat,‡ Gary Splitter,† David C. Morrison,*§ and Stefanie N. Vogel‡

We have developed a novel LPS probe using a highly purified and homogenous preparation of [3H] Escherichia coli LPS from the deep rough mutant, which contains a covalently linked, photoactivable 4-p-(azidosalicylamido)-butylamine group. This cross- linker was used to identify the LPS-binding proteins in membranes of the murine-macrophage-like cell line RAW 264.7. The ␣-subunit (PSMA1 C2, 29.5 kDa) and the ␤-subunit (PSMB4 N3, 24.36 kDa) of the 20S proteasome complex were identified as LPS-binding proteins. This is the first report demonstrating LPS binding to enzymes such as the proteasome subunits. Function- ally, LPS enhanced the chymotrypsin-like activity of the proteasome to degrade synthetic peptides in vitro and, conversely, Downloaded from the proteasome inhibitor lactacystin completely blocked the LPS-induced proteasome’s chymotrypsin activity as well as macrophage TNF-␣ secretion and the expression of multiple inflammatory mediator genes. Lactacystin also completely blocked the LPS-induced expression of Toll-like receptor 2 mRNA. In addition, lactacystin dysregulated mitogen-activated protein kinase phosphorylation in LPS-stimulated macrophages, but failed to inhibit IL-1 receptor-associated kinase-1 activity. Importantly, lactacystin also prevented LPS-induced shock in mice. These data strongly suggest that the proteasome complex regulates the LPS-induced signal transduction and that it may be an important therapeutic target in Gram-negative sepsis. The Journal of http://www.jimmunol.org/ Immunology, 2003, 171: 1515–1525.

ipopolysaccharide is the major constituent of the outer (LBP).3 Presently, it is thought that CD14 serves to “focus” the membrane of Gram-negative bacteria. The overproduc- LPS-LBP complex to a functional transducing receptor, which L tion of LPS-induced cytokines such as TNF, IL-1, and triggers the activation of the cells through a transmembrane sig- IL-6 has been considered central to the pathophysiologic derange- naling mechanism (4). ments associated with septic shock. It is generally accepted that the Over the past 15 years, many cellular proteins have been re-

first step in the activation of cells by LPS is the binding of this ported to bind LPS and/or lipid A (5–11). However, some of these by guest on October 1, 2021 chemically complex ligand to specific membrane receptors, which proteins have not been purified or characterized, and their biolog- ultimately leads to induction of gene expression and the release of ical functions remain to be fully elucidated. Recently, the type I cytokines and other mediators of inflammation (1, 2). CD14 has integral membrane protein Toll-like receptor 4 (TLR4) has been been identified as a cell surface, 55-kDa glycosylphosphatidyl- identified as the main signaling receptor for enterobacterial LPS inositol-linked protein expressed on the surface of macrophages, (12). A nonmembrane spanning molecule, MD-2, has also been monocytes, and neutrophils (3, 4). Other LPS-responding cell implicated as necessary for effective LPS signaling via TLR4 (13). types, i.e., lymphocytes, endothelial cells, and fibroblasts, express Several other proteins, such as the ␤-integrins (14), caveolin (15), very low levels of or are devoid of CD14, and they use a soluble CD55/decay accelerating factor (16), and moesin (17), have also form of this molecule. CD14 is now recognized to bind molecular been implicated in LPS-mediated signal transduction. Besides this complexes formed between LPS and plasma LPS binding protein pathway, LPS has been reported to also activate cells via the CD14-independent pathway (18). Despite this accumulated knowl- edge regarding effector molecules involved in LPS signaling, the mechanisms that facilitate the interaction of these various putative *Department of Basic Medical Science, School of Medicine and Shock/Trauma Re- coreceptors with TLR4 remain to be unraveled. search Center, University of Missouri, Kansas City, MO 64108; †Department of An- One approach to identify molecules within an “LPS signaling imal Health and Biomedical Sciences, University of Wisconsin, Madison, WI 53706; ‡Department of Microbiology and Immunology, University of Maryland, Baltimore, complex” was first presented nearly a decade ago. A radioiodi- MD 21201; and §Department of Surgical Science, University of Missouri, Kansas nated cross-linking agent, [125I]-sulfosuccinimidyl-2-[p-azidosali- City, MO 64108 cyl-amido] ethyl-1,3Ј-dithiopropionate (SASD), which reacts with Received for publication March 4, 2003. Accepted for publication May 21, 2003. free amino groups on phosphorylethanolamine residues of LPS, The costs of publication of this article were defrayed in part by the payment of page was first coupled to heterogenous preparations of either LPS from charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by National Institutes of Health Grants GM50870 (to 3 Abbreviations used in this paper: LBP, LPS binding protein; TLR, Toll-like recep- N.Q.), AI18797 (to S.N.V.), AI48490 (to G.S.), and AI23447 (to D.C.M). This manu- tor; SASD, sulfosuccinimidyl-2-[p-azidosalicyl-amido] ethyl-1,3Ј-dithiopropionate; script was presented in part at the Shock Society meeting (2002) and the International ReLPS, LPS from the deep rough mutant; ASBA, 4-[azidosalicylamido] butylamine; Endotoxin Society meeting 2002. MALDI-MS, matrix-assisted laser desorption ionization mass spectrometry; MAPK, 2 Address correspondence and reprint requests to Dr. Nilofer Qureshi, Department of mitogen-activated protein kinase; IRAK-1, IL-1 receptor-associated kinase-1; Basic Medical Science, School of Medicine, Shock/Trauma Research Center, 2411 RsDPLA, Rhodobacter sphaeroides diphosphoryl lipid A; AMC, amino-methyl-cou- Holmes Street, University of Missouri, Kansas City, MO 64108. E-mail address: marin; ERK, extracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; [email protected] 2-D, two dimensional; iNOS, inducible NO synthase; COX-2, cyclooxygenase 2.

the deep rough mutant (ReLPS) of Salmonella minnesota R595 or bacterial products such as LPS can bind directly to and activate the smooth Escherichia coli LPS (5–7). Purification of smooth LPS or proteasome complex in vitro and that proteasome inhibitors like ReLPS (R595) to a single species is not possible, as has been lactacystin can selectively block this activation in vitro and prevent described for the ReLPS from E. coli D31 m4 (19). Nonetheless, LPS-induced shock in mice. Lei and Morrison (7) used this [125I]-ASD-LPS probe to label two major LPS-binding proteins in membranes of murine B cells and Materials and Methods macrophages (i.e., 70 kDa and 38 kDa). Unfortunately, they were Reagents unable to characterize these proteins by the relatively insensitive Highly purified hexaacyl ReLPS E. coli D31 m4 (19) and phenol-water- protein sequencing techniques that were available at the time. Sub- extracted E. coli K235 LPS (Ͻ0.008% protein) were prepared according to sequently, the purity and specificity of the [125I]-ASD-LPS probe the methods of Qureshi et al. (19) and McIntire et al. (23), respectively. came under intense scrutiny when Dziarski (20) demonstrated that Rhodobacter sphaeroides diphosphoryl lipid A (RsDPLA) was prepared as described by Qureshi et al. (24, 25). ASBA and 1-ethyl-3-[3-dimethyl- it preferentially cross-linked albumin that was bound to cells. Nev- aminopropyl]-carbodiimide hydrochloride were purchased from Pierce ertheless, this same probe was very recently used to cross-link (Rockford, IL). Chelex 100 (Naϩ) and Dowex 50 (Hϩ) were purchased TLR4, CD14, MD-2, and TLR2 in HEK 293 cells engineered to from Bio-Rad (Hercules, CA). Proteasome substrate III (Suc-Leu-Leu-Val- overexpress these proteins, suggesting that these molecules may be Tyr-amino-methyl-coumarin (AMC)) and MG-132 were purchased from organized within the membranes in sufficiently close proximity to Calbiochem (La Jolla, CA). Protease inhibitors were purchased from Roche (Mannheim, Germany). Lactacystin was purchased from Calbio- be cross-linked and radiolabeled (21). However, a major caveat chem, and was solubilized in DMSO to form a 2 mg/ml stock solution. For with such studies is that overexpressed proteins may be preferen- treatment of macrophages, further dilutions of the stock lactacystin were tially cross-linked, resulting in the failure to detect molecules made in supplemented RPMI 1640 medium or DMEM from BioWhittaker within the complex that are not overexpressed. In addition, another (Walkersville, MD). Heat-inactivated FCS was also purchased from Bio- Downloaded from 125 125 Whittaker. Rabbit polyclonal anti-phospho-extracellular signal-regulated potential flaw associated with using [ I]-ASD-LPS is that [ I]- kinase 1/2 (ERK-1/2) and anti-phospho-c-Jun N-terminal kinase 1/2 (JNK- SASD by itself (in the absence of LPS) was found to bind mem- 1/2) Abs were purchased from Promega (Madison, WI). Rabbit polyclonal brane proteins of 55- and 18-kDa molecular masses within the anti-phospho-p38 MAPK Ab was purchased from New England Biolabs RAW 264.7 macrophage cell line (B. Jarvis, D. Morrison, and N. (Beverly, MA), and rabbit anti-total-p38 Ab was purchased from Santa Qureshi, unpublished data). Thus, it would be important to con- Cruz Biotechnology (Santa Cruz, CA). tinue to exercise caution in the interpretation of such studies unless Preparation of the LPS-ASBA derivative of LPS http://www.jimmunol.org/ very rigorous attention is directed to the preparation and charac- [3H] E. coli ReLPS was prepared and purified using a DEAE-cellulose terization of LPS-cross-linking probes. column as described previously (19, 26). The specific activity of the As a consequence of the problems of specificity, incomplete [3H]ReLPS was 5.5 ϫ 108 dpm/mg LPS. ReLPS was weighed before der- derivatization, and LPS heterogeneity that were identified with the ivatization. Purified [3H]ReLPS (1 mg) was dissolved in 0.1 M MES buffer ␮ ␮ [125I]-ASD-LPS probe, we sought to generate a noniodinated, but (100 l; pH 4.5) and mixed with 0.6 mg of ASBA in 100 l of the same buffer. ASBA is a photoactivatable carboxyl-reactive cross-linker with a radiolabeled, homogenous LPS with a cross-linking group that spacer arm of 16.3 A. One hundred microliters of 1-ethyl-3-[3-dimethyl- would allow for greater specificity after covalent linkage to pro- aminopropyl]-carbodiimide hydrochloride (10 mg/ml), a dehydrating teins within macrophage membranes from primary macrophages agent, was added. The reaction mixture was incubated at 37°C for5hinthe by guest on October 1, 2021 or a similarly LPS-responsive macrophage cell line that has not dark, and was extracted with chloroform:methanol 2:1. The lower organic been transfected to overexpress likely components of the putative phase was filtered and evaporated to dryness under a stream of nitrogen. The ASBA cross-linker group (Pierce) that reacts with the carboxyl anions LPS signaling complex. ReLPS from E. coli contains charged was used in preparation of [3H]ReLPS with a representative specific ac- groups as two carboxyl anions in the two Kdo units, four anionic tivity of 5.5 ϫ 108 dpm/mg. The molecular mass of ReLPS-ASBA deriv- groups in the two phosphate groups, and no free amino groups ative is 2470 Da (N. Qureshi, S. Lin, and R. J. Cotter, manuscript in (19). The two phosphates are known to be very important for bi- preparation). ological activity because monophosphoryl lipid A is relatively Preparation of the solubilized membranes from RAW 264.7 cells nontoxic (22). Based upon these structural features, we developed RAW 264.7 cells (1 ϫ 109) grown in DMEM containing 10% FCS and an LPS probe using a highly homogenous, purified preparation of gentamicin were washed three times with saline. The cells were pelleted by [3H] E. coli ReLPS, which contains a covalently linked, novel, centrifugation at 700 ϫ g for 5 min. The pellet was suspended in lysis photoactivable 4-[azidosalicylamido] butylamine (ASBA) group. buffer (10 ml, 1 ϫ 108 cells/ml) containing 25 mM HEPES (pH 7.3), 0.5 ASBA was esterified to the carboxylic groups of the Kdo, and this mM EGTA, 0.5 mM sodium orthovanadate, 0.1 mM sodium molybdate, 1 mM sodium fluoride, protease inhibitors, and one tablet of Complete Mini cross-linker was then used to identify the LPS-binding proteins in EDTA-free (Roche, Indianapolis, IN) for every 50 ml of lysis buffer. Cells membranes of the murine-macrophage-like cell line, RAW 264.7 were lyzed by freeze-thawing three times. Cell debris was removed by cells. Using this derivatized LPS cross-linking agent, we have been centrifugation at 700 ϫ g for 5 min at 4°C. The supernatant was centri- able to detect binding of our radiolabeled probe to 13–18 proteins fuged at 5,800 ϫ g for 7 min at 4°C to remove aggregates of cytoskeletal ϫ in murine macrophage membranes. Two consistently labeled LPS- elements. The supernatant was finally centrifuged at 100,000 g for 1.5 h at 4°C to obtain the membrane pellet. binding proteins were identified by matrix-assisted laser desorp- The membrane pellet was suspended and solubilized in lysis buffer con- tion ionization-mass spectrometry (MALDI-MS) as the ␣-subunit taining 10 mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesul- (C2) and the ␤-subunit (N3) of the proteasome complex. Of rele- fonate (Sigma-Aldrich, St. Louis, MO). Protein determination was per- vance to this finding is the fact that underivatized LPS was able to formed by Bio-Rad. This method of membrane preparation is a modification of that reported by Bhat et al. (27). activate the chymotrypsin-like activity of bacterial and mammalian proteasomes in vitro in the absence of CD14 or LBP, further sup- Cross-linking of ReLPS-ASBA to solubilized membranes porting the notion that LPS can bind directly to the proteasome and [3H]ReLPS-ASBA was used for binding studies to the solubilized mem- activate it enzymatically. Finally, lactacystin, a functional inhibitor brane proteins from RAW 264.7 cells. Several concentrations of of the proteasome, blocked LPS-induced gene expression and dys- [3H]ReLPS-ASBA ranging from 10 to 714 ␮g/ml were tested. Typically, 3 6 regulated phosphorylation of mitogen-activated protein kinase [ H]ReLPS-ASBA (free acid form, 8 ␮g/8 ␮l, 4.4 ϫ 10 dpm, 286 ␮g/ml, or 16 ␮g/8 ␮l, 8.8 ϫ 106 dpm, 572 ␮g/ml) was dissolved in 0.5% trieth- (MAPK), but not IL-1 receptor-associated kinase-1 (IRAK-1) ac- ylamine solution in water and mixed with solubilized membrane proteins tivity, in macrophages. Importantly, lactacystin also prevented (100 ␮g/20 ␮l) in small glass tubes and incubated at 37°C for 1.5 h in the LPS-induced shock in mice. This is the first report establishing that dark, with occasional vortexing. Similar concentrations of [3H]ReLPS The Journal of Immunology 1517 were used in a competitive LPS binding study as reported previously SDS-PAGE and Western analysis ([3H]ReLPS (86.66 ␮g/ml), CD14, LBP, and RsDPLA (666 ␮g/ml–6.66 mg/ml)) (26). ReLPS-ASBA-protein complexes were irradiated with a Cytoplasmic extracts for analysis of MAPK phosphorylation were gener- long-wave UV lamp (model UVGL-55; Ultraviolet Products, Upland, CA) ated from macrophages pretreated with medium or lactacystin for1hand at room temperature for 30 min at a distance of 7 cm. then with medium or LPS for 15 min as described previously (30). The After irradiation, the samples (28 ␮l) were diluted 1:1 with SDS boiling resultant proteins were resolved on 10% SDS-PAGE gels and then trans- buffer and boiled for 5 min. Two-dimensional (2-D) electrophoresis was ferred onto Immobilon membranes. Western blot analysis was conducted performed according to the method of O’Farell (28). Isoelectric focusing for phospho-ERK-1/2, phospho-JNK-1/2, phospho-p38, and total p38. was conducted in a glass tube of inner diameter 2.0 mm, using 2.0% pH Proteasome assays 4–8 ampholines (Gallard Schlesinger, Long Island, NY) for 9600 volt- hours. One microgram of an isoelectric focusing internal standard protein, Proteasome activities of the 20S proteasomes of Methanosarcina ther- tropomysin, with lower spot of m.w. of 33,000 and pI 5.2, was added to the mophila (10 ␮g/ml) (32), rabbit muscle proteasome (0.4 ␮g/ml), and mac- sample. The tube gel pH gradient plot for these ampholines was measured rophage proteasome were assayed with synthetic peptide substrates in 0.02 by a surface pH electrode. After equilibration for 10 min in Buffer “O” M Tris-HCl buffer (pH 7.2). The substrates used for the chymotrypsin-like, (10% glycerol, 50 mM DTT, 2.3% SDS, and 0.625 M Tris buffer (pH 6.8)), trypsin-like, and postglutamase activity were 100 ␮M succinyl-Leu-Leu- the tube gel was sealed to the top of the stacking gel on a 10% acrylamide Val-Tyr-amino-methyl-coumarin, Z-ARR-AMC, and Abz-GPALA-NBA, slab gel (0.75 mm thick). SDS slab gel electrophoresis was conducted for respectively (Calbiochem). LPS concentrations ranging from 4 pg/ml to 4 ϳ4 h at 12.5 mA/gel. The following proteins (Sigma-Aldrich) were added ␮g/ml were used in a total volume of 250 ␮l. Fluorescence was measured as m.w. standards to the agarose, which sealed the tube gel to the slab gel: using an Flx 800 microplate fluorescence reader (Bio-Tek Instruments, myosin (220,000), phosphorylase A (94,000), catalase (60,000), actin Winooski, VT). (43,000), carbonic anhydrase (29,000), and lysozyme (14,000). These standards appear as horizontal lines on the Coomassie Brilliant Blue R 250- Purification of macrophage proteasomes stained 10% acrylamide slab gel. The gel was treated with En3Hance (NEN, Boston, MA) for 1 h, rehydrated in water for 30 min, and dried onto The RAW 264.7 cells were scraped from the plate using ice-cold PBS. Downloaded from filter paper with acid end to the left. Fluorography was conducted using After several washes with PBS, the cells were suspended in 50 mM Tris- X-OMAT AR film (Kodak, Rochester, NY) with exposures of 7–14 days HCl buffer (pH 7.5) containing 5 mM MgCl2 and 250 mM sucrose. The at Ϫ70°C. The films were developed using Kodak developer and fixer. A cells were broken using the freeze-thaw method, and the suspension was ϫ duplicate gel was run and stained with silver nitrate (Kendrick Laborato- centrifuged at 10,000 g for 10 min at 4°C. The supernatant was centri- ϫ ries, Madison, WI). The protein spots were excised from the silver nitrate fuged at 100,000 g for 1 h. Then the resulting supernatant was centri- ϫ stained gel, digested with endoproteinase Lys-C, and analyzed by fuged at 100,000 g for 5 h (33). The pellet was suspended in buffer and assayed for proteasome activity as described above. MALDI-MS. http://www.jimmunol.org/ IRAK assay Macrophage culture The assay for IRAK-1 activity was conducted essentially as described by Li et al. (34). Mouse macrophages were pretreated with lactacystin (25 RAW 264.7 macrophage-like cells were grown in DMEM supplemented ␮ with 10% FCS and gentamicin (10 mg/500 ml). Alternatively, 5- to 6-wk- M) and then exposed to LPS (10 or 100 ng/ml) at the times indicated. At old C3H/HeOuJ mice were purchased from The Jackson Laboratory (Bar each time point, cells were lysed in a buffer that contained several pro- Harbor, ME) and were injected i.p. with 3 ml of 3% fluid thioglycollate. teinase and phosphatase inhibitors, and the lysates were collected by cen- Four days later, macrophages were extracted by peritoneal lavage with trifugation. This kinase assay is dependent upon immunoprecipitation of pyrogen-free saline. Cells were then pelleted and resuspended in RPMI IRAK-1 with anti-IRAK-1 Abs, followed by an in vitro kinase assay on the immunoprecipitates. Ab to IRAK-1 (Upstate Biotechnology, Lake Placid, 1640 medium supplemented with 2% FBS, 2 mM glutamine, 30 mM by guest on October 1, 2021 ␮ NY) was added to each lysate, incubated, and followed by the addition of HEPES, 0.3% NaHCO3, 100 IU/ml penicillin, and 100 g/ml streptomy- cin. For RNA generation, 6.5 ϫ 106 cells per well were cultured overnight Protein G-agarose beads to bind specifically to the Abs and allow the and treated as indicated. separation of Ab-Ag complexes from other proteins in the lysate. The precipitate was analyzed by incubation with a substrate (myelin basic protein) in the presence of [32P]ATP. In the latter case, phosphorylation of the Extraction of total cellular RNA and RT-PCR substrate was visualized and quantified by running the samples on an SDS- PAGE gel and exposing the gel to a phosphor imager screen. This kinase For the extraction of total cellular RNA, macrophages were pretreated with activity is dose dependent and specific to IRAK precipitates because con- medium or medium containing lactacystin for 1 h and then with LPS for trol experiments with rabbit IgG used as precipitating Ab showed no in- 4 h. Macrophages were then solubilized in RNA-STAT 60, and total cel- crease in kinase activity. lular RNA was generated and reverse transcribed as detailed elsewhere (29). PCR amplifications were performed with the cDNA for the genes of Galactosamine-sensitized mouse model interest as described previously (30, 31). Amplified products were electro- Normal female BALB/c mice (6 wk) were obtained from The Jackson phoresed on 1% agarose gels and blotted overnight onto Nytran mem- Laboratory. Animals were housed in a constant day/night regime for 2 wk branes. The DNA was UV cross-linked onto the membranes and baked for before use. Mice were injected i.p. with 18 mg of galactosamine and 2 h at 80°C, and the amplified PCR products were detected by gene-specific ϩ ReLPS. Lactacystin was injected i.p. 1 h before the ReLPS galactosamine oligonucleotide probes. The sequences of the primers and probes used in challenge. LPS-induced lethality was measured and scored over a 24- to the detection of TNF-␣, IL-12 p40, IL-12 p35, inducible NO synthase 36-h period. All animal procedures were performed in accordance with the (iNOS), cyclooxygenase 2 (COX-2), TLR4, TLR2, and GAPDH have been National Institutes of Health guidelines for the use of experimental published previously (29). The sequences of primers and probes used in animals. this detection of CD14 and GAPDH are as follows: CD14 sense, 5Ј-CT GATCTCAGCCCTCTGTCC-3Ј; CD14 antisense, 5Ј-GCTTCAGCCCA GTGAAAGAC-3Ј; CD14 probe, 5Ј-ACCCTCCAAGTTTTAGCGCT-3Ј; Results GAPDH sense, 5Ј-CCATGGAGAAGGCCGGG-3Ј; GAPDH antisense, Synthesis of the [3H]ReLPS-ASBA 5Ј-CAAAGTTGTCATGGATGACC-3Ј; GAPDH probe, 5Ј-GGGTGTG AACCACGAGAAAT-3Ј. The housekeeping gene GAPDH served as a A novel tritiated E. coli ReLPS-ASBA photoactive cross-linker control for the amount of cDNA added to each amplification reaction. was synthesized as described in Materials and Methods and then used to detect the LPS binding proteins in a solubilized membrane preparation of RAW 264.7 cells. The final purified cross-linking TNF assay product was analyzed by soft ion mass spectrometry. These results RAW 264.7 cells were plated in eight-well plates (Corning, Corning, NY) confirmed that this coupling reaction is very efficient and that ap- 6 at a density of 0.6 ϫ 10 cells/well in a final volume of 0.5 ml. After 2 h proximately one ASBA is esterified to either one of the carboxyl of adherence at 37°C, RAW 264.7 cells were pretreated with medium or lactacystin for 1 h, and then with medium or 2 ng/ml LPS for 1.5 h. The groups of the Kdo units of the ReLPS. MALDI-MS revealed the supernatants obtained were assayed for TNF with the Quantikine M kit structure of the LPS-ASBA as presented in Fig. 1, which exhibited (R&D Systems, Minneapolis, MN). a molecular mass of 2470 Da (N. Qureshi, S. Lin, and R. J. Cotter, 1518 ROLE OF THE PROTEASOME IN LPS-INDUCED INFLAMMATION

FIGURE 1. Structure of the ReLPS-ASBA. Only one ASBA group was cross-linked to the Kdo of ReLPS, as demonstrated by mass spectrometry. ASBA may be linked to either one of the Kdo and may represent a mixture of the two derivatives. Downloaded from manuscript in preparation). Mass spectrometry also revealed the stained protein spots delineated A and B, as observed on the silver- absence of detectable free underivatized LPS in the sample, sug- stained gel. Some of the radioactive signals seen in Fig. 2B were gesting that most of the ReLPS had been chemically derivatized. in spots that did not coincide with any detectable protein spots in The ReLPS-ASBA and the underivatized ReLPS had similar TNF- the silver-stained gel, and these could not be analyzed. The major ␣-inducing activity in cultures of RAW 264.7 cells (data not radiolabeled spots, A and B, were excised, digested with endopro- http://www.jimmunol.org/ shown), thus suggesting that the derivatization did not adversely teinase Lys-C, and analyzed by MALDI-MS at the Protein Chem- affect the biological activity of ReLPS. istry Core Facility at the Howard Hughes Medical Institute/Co- lumbia University College of Physicians and Surgeons (New York, 3 Use of [ H]ReLPS-ASBA to identify novel LPS-binding NY). Based upon this analysis, the protein spots A and B were membrane proteins positively identiﬁed as proteasome components: spot A correlated Initial studies were conducted using this probe to detect novel LPS with the PSMA1 C2 (␣ subunit, 29.5 kDa) and spot B with PSMB4 binding proteins in murine macrophage membrane preparations. N3 (␤ subunit, 24.4 kDa) of the proteasome. This experiment was

For these experiments, solubilized membrane proteins obtained repeated six times and in each experiment these two proteasome by guest on October 1, 2021 from RAW 264.7 cells were incubated with [3H]ReLPS-ASBA subunits were consistently labeled. Proteins that were not consis- and subjected to photo-cross-linking by exposure to UV light. The tently labeled were not analyzed. In competition experiments, a photo-cross-linked complexes were boiled with SDS boiling buffer 5-fold or 10-fold excess of unlabeled ReLPS or RsDPLA was at 100°C for 5 min and then fractionated by two-dimensional elec- bound to the membrane proteins, followed by [3H]ReLPS-ASBA trophoresis (as described in Materials and Methods). Duplicate (286 ␮g/ml–572 ␮g/ml). In these experiments, the unlabeled gels were run under identical conditions; one gel was stained with ReLPS competed for binding sites, and binding of [3H]ReLPS- silver nitrate, whereas the other gel was stained with Coomassie ASBA to macrophage proteasome subunits was not observed. 3 blue, treated with En Hance, dried, and analyzed by ﬂuorography. 3 As anticipated, more than 100 protein species were resolved on the Binding of [ H]ReLPS-ASBA to the proteasome of M. 2-D silver-stained gel (Fig. 2A). The replicate gel, subjected to thermophila ﬂuorography as shown in Fig. 2B, revealed the presence of 15–18 The results mentioned above strongly suggested that LPS was tritium-labeled spots, indicating the presence of the cross-linked bound to subunits of the macrophage proteasome. To pursue this probe. For this analysis, we selected two major radioactive bands observation further, therefore, we evaluated the ability of the that in multiple experiments coincided with readily detectable [3H]ReLPS-ASBA probe to bind to the commercially available

FIGURE 2. Two-dimensional SDS-PAGE analysis of the membrane proteins of RAW 264.7 (100 ␮g, 8.3 mg/ ml) cells complexed to [3H]ReLPS-ASBA (16 ␮g, 8.8 ϫ 106 dpm, 572 ␮g/ml) in a total volume of 28 ␮l. The complexes were irradiated with long-wave UV and analyzed by 2-D gel electrophoresis as described in Materials and Methods. Isoelectric focusing was conducted at pH 4–8, and a 10% acrylamide slab gel was used for electrophoresis. Duplicate gels were run: one was stained with the silver nitrate stain (A), and the other was treated with Coomassie blue, En3Hance for 1 h, and then it was rehydrated in water for 30 min, dried, and ﬂuorographed (B). Major bands A and B were labeled with [3H]ReLPS- ASBA. Several other proteins were also labeled. The ar- rowhead represents the internal standard tropomysin, which has a molecular mass of 33 kDa and a pI of 5.2. The Journal of Immunology 1519

FIGURE 3. Two-dimensional SDS-PAGE analysis of the proteasome of M. thermophila (25 ␮g, 0.830 mg/ml) complexed to [3H]ReLPS-ASBA (5 ␮g, 2.75 ϫ 106 dpm, 167 ␮g/ml) in a total volume of 30 ␮l. The complexes were irradiated with long-wave UV, lyophilized, and boiled with SDS boiling buffer as described in Materials and Methods. Isoelectric focusing was conducted at pH 4–8, and a 10% acrylamide slab gel was used for electrophoresis. Slab gel was stained with silver nitrate, revealing the ␣ (24 kDa) (arrow 1) and ␤ subunits (22 kDa) (arrow 2) of the proteasome of M. thermophila (A). The 3 gel was treated with En Hance for 1 h, rehydrated in water for 30 min, dried, and ﬂuorographed (B). Downloaded from

20S proteasome of M. thermophila (Calbiochem). In this experi- lyzed as described above for the macrophage membrane prepara- ment, [3H]ReLPS-ASBA was incubated with the 20S proteasome tions. Both the ␣ and ␤ subunits of the proteasome were clearly derived from M. thermophila, cross-linked by UV, and then ana- identiﬁed in silver-stained 2-D gels after electrophoresis (Fig. 3A). Importantly, however, and consistent with the data shown in Fig. http://www.jimmunol.org/ 2, [3H]ReLPS-ASBA bound predominantly to the ␣ subunit (three bands, 24 kDa) of the proteasome and not to the ␤ subunit (22 kDa) (Fig. 3B). These data provide direct evidence that ReLPS- ASBA is capable of binding to the ␣ subunit of this simpler Archae proteasome.

LPS activates the chymotrypsin-like activity and the postglutamase activity of the proteasome

One of the primary cellular functions of the proteasome is prote- by guest on October 1, 2021 olysis of proteins. Therefore, we examined the effect of LPS on the chymotrypsin-like, trypsin-like, and postglutamase activity of the proteasome complex. For these studies, either rabbit muscle proteasome or macrophage proteasome preparations were incubated with various concentrations of LPS, and enzymatic cleavage of the

FIGURE 4. Rabbit muscle proteasome (P) activity in the presence of ReLPS. A, LPS (L) or RsDPLA (R) (40 ng–4 ␮g) was incubated with the rabbit muscle proteasome (0.4 ␮g/ml) and AMC substrate III succinyl- Leu-Leu-Val-Tyr-7-amido-4-methyl coumarin (100 ␮M) in 96-well plates for 20 min. B, Rabbit muscle proteasome was incubated with RsDPLA (5 FIGURE 5. Induction of TNF-␣ in RAW 264.7 cells pretreated with the min), LPS (4 ␮g/ml), and AMC substrate III for 20 min. C, Macrophage proteasome inhibitor lactacystin. RAW 264.7 cells were pretreated with the proteasome was incubated either with LPS (black bars) or with lactacystin indicated concentration of lactacystin for 1 h, followed by addition of 2 (30 min) and LPS (gray bars) and with AMC substrate III for 20 min. The ng/ml LPS for 1.5 h. The supernatants obtained were assayed for TNF-␣ plates were read at an absorption/emission of 360/460 nm using a micro- with the Quantikine M kit (R&D Systems) as described in Materials and plate fluorescence reader. The data are representative of three identical Methods. The data are representative of two identical experiments per- experiments performed. formed. M, Medium; L, LPS; and LC, lactacystin. 1520 ROLE OF THE PROTEASOME IN LPS-INDUCED INFLAMMATION proteasome-specific substrate III (succinyl-Leu-Leu-Val-Tyr-7- cells to lactacystin completely blocked the LPS-induced chymo- amido-4-methyl-coumarin) was determined. LPS showed a dose- trypsin activity of this proteasome in in vitro experiments. dependent activation of the rabbit muscle proteasome’s (Biomol, Plymouth Meeting, PA) chymotrypsin activity (as shown by the Proteasome enzymatic inhibitors lactacystin and MG-132 block data in Fig. 4A) and postglutamase activity, but no activation of the LPS-induced TNF secretion trypsin-like activity (data not shown) was observed. In addition, no To probe further the role of the proteasome in the LPS-induced activation of the proteasome’s chymotrypsin-like activity was ob- signal transduction, we next queried whether inhibitors of the pro- served with the LPS antagonist (24) from RsDPLA, even at con- teasome would block the activating effects of LPS on macro- centrations up to 4 ␮g/ml (Fig. 4A). These results suggest that the phages. To do these studies, RAW 264.7 cells were pretreated with proteasome can be activated only by toxic LPS. RsDPLA blocked specific proteasome inhibitors, lactacystin or MG-132 (carboben- the LPS-induced proteasome activity (Fig. 4B) over a concentra- zooxy-L-leucyl-L-leucyl-L-leucinal), before addition of underivat- tion range consistent with its ability to block LPS signaling in ized ReLPS. The supernatants were collected 1.5 h after LPS stim- macrophages. Similar results were observed with the M. ther- ulation and were assayed for TNF-␣. Pretreatment of RAW 264.7 mophila proteasome (data not shown) and the macrophage protea- cells with lactacystin (Fig. 5) or MG-132 (data not shown) resulted some as observed in Fig. 4C, where LPS-activated chymotrypsin in a dose-dependent inhibition of TNF-␣ secretion with concen- activity was inhibited by RsDPLA. Prior exposure of RAW 264.7 trations of 25 ␮M lactacystin or 3 ␮M MG-132, completely block- ing secretion of LPS-induced TNF-␣ in these cells. Cell death, as assessed by dye exclusion test, was not observed with cells treated with these concentrations of the inhibitors. In addition, neither lac- Downloaded from tacystin nor MG-132 alone induced any TNF-␣ secretion.

Lactacystin blocks LPS-induced gene expression Although the data provided thus far suggest that inhibitors of proteasome activation inhibit LPS-induced cytokine secretion, they do not provide information on either the speciﬁcity of inhibition or the point in the activation pathway at which inhibition occurs. To de- http://www.jimmunol.org/ ﬁne the role of the proteasome in LPS-mediated gene expression, we pretreated primary cultures of murine peritoneal macrophages isolated from LPS-responsive C3H/HeOuJ mice with lactacystin before the addition of LPS. Total cellular RNA was then extracted and reverse transcribed, and gene expression analyses were performed by RT-PCR and Southern blot analysis. The panel of genes selected included a number of important macrophage activities and ␣ included cytokine genes TNF- , IL-6, IL-12 p35, and IL-12 p40, by guest on October 1, 2021 as well as COX-2 and iNOS, two genes that encode enzymes that regulate key functions of macrophages. Expression of genes that encode LPS receptors TLR4, TLR2, and CD14 were also examined. As shown by the data in Fig. 6, LPS markedly induced the mRNA levels of all of the genes examined, with the exception of TLR4, for which the constitutive level of TLR4 mRNA was

FIGURE 7. Effect of lactacystin pretreatment on LPS-induced MAPK FIGURE 6. RT-PCR analysis of LPS-induced gene expression in mac- in macrophages. Macrophages from C3H/HeOuJ mice were pretreated with rophages pretreated with the proteasome inhibitor lactacystin. Thioglycol- indicated concentrations of lactacystin for 1 h followed by 1 ng/ml LPS for late-elicited macrophages derived from C3H/HeOuJ mice were pretreated 15 min. Cytoplasmic extracts were generated, and proteins were resolved with lactacystin for1hasindicated followed by 1 ng/ml LPS for 4 h. Total on SDS-PAGE gels and immunoblotted for phospho-speciﬁc ERKs, JNKs, cellular RNA was generated, reverse transcribed, ampliﬁed, and detected as and p38 as described in Materials and Methods. After phospho-ERK de- described in Materials and Methods. The data are representative of two tection, membrane was stripped and reprobed for total p38. The data are identical experiments performed. representative of two identical experiments performed. The Journal of Immunology 1521

Table I. Effect of treatment with lactacystin on lethality of galactosamine-sensitized BALB/c mice challenged with toxic ReLPS

Experimental Challengea Group Pretreatmenta (Ϫ30 min) (0 h) Survival/Total

A Vehicle ReLPS 0/5 B Vehicle Vehicle 5/5 C Lactacystin (0.08 ␮mol) ReLPS 0/5 D Lactacystin (0.4 ␮mol) ReLPS 4/5 E Lactacystin (0.4 ␮mol) Vehicle 5/5

a At Ϫ30 min, mice were treated with either vehicle or lactacystin. Then at 0 h, the animals were challenged with vehicle alone or with ReLPS (1 ␮g/mouse) plus galactosamine (18 mg) i.p. This experiment was repeated twice with similar results. The mice were monitored for death for 24 h. p Ͻ 0.01.

vation is required for IRAK-1 activity, IRAK-1 was ﬁrst immunoprecipitated from lysates of RAW 264.7 cells stimulated with LPS in the absence or presence of lactacystin. Kinase activity was measured by the incorporation of 32P into MBPs (Fig. 8A). Fig. 8B

illustrates that, although the lactacystin was effective at inhibiting Downloaded from ERK-1,2 phosphorylation in these same cells, it failed to inhibit IRAK-1-associated kinase activity signiﬁcantly (Fig. 8, A and C).

Lactacystin prevents LPS-induced shock in mice To further investigate the effect of lactacystin on LPS-induced

shock, we examined its effect in galactosamine-sensitized mice http://www.jimmunol.org/ challenged with toxic ReLPS (Table I). In this experiment, ReLPS FIGURE 8. RAW 264.7 cells were treated with LPS in the absence or challenge (0.5 ␮g) caused 100% lethality, whereas pretreatment ␮ presence of lactacystin (25 M). Cells were lysed at the indicated times with lactacystin (0.4 ␮moles) 30 min before challenge reduced after LPS (100 ng/ml) stimulation, and they were assayed for IRAK-1 lethality to 20%. Lactacystin alone in the absence of LPS challenge associated kinase activity as described in Materials and Methods. A, was not toxic at a dose of 0.4 ␮moles. This experiment was re- IRAK-1 activity from a single representative experiment. B, ERK-1 phosphorylation detected by Western blot analysis in the same experiment. C, peated within a 2-wk time interval with similar results. Mean Ϯ SEM of four separate experiments in which kinase activity was measured densitometrically. Discussion The most significant finding from this study is the novel observa- by guest on October 1, 2021 tion that LPS has the capacity to bind to selected subunits of the slightly down-regulated. It is of importance that the LPS-induced macrophage proteasome and that the consequences of that binding expression of TLR2 mRNA was completely blocked by high con- include the selective activation of specific proteases of the protea- centrations of lactacystin. Interestingly, pretreatment with lacta- some. Evidence that the proteasome has functional significance in cystin blocked expression of most of the LPS-inducible genes in a the pathway of LPS-mediated macrophage signal transduction trig- dose-dependent manner, as seen in Fig. 6, whereas lactacystin gered by LPS derives from the observation that pretreatment of alone had minimal effect on gene expression. The failure of lac- macrophages with lactacystin, a well-characterized inhibitor of tacystin to block expression of the housekeeping gene GAPDH proteasome activity, inhibits LPS-induced up-regulation of TLR2 suggests that the observed inhibition is not due to a toxic effect of and a number of proinflammatory genes in a dose-dependent fash- lactacystin on macrophages. ion. Because our studies also document that lactacystin differentially regulates several of the kinases known to be involved in Lactacystin blocks the LPS-induced phosphorylation of ERK-1 or ERK-2 To further explore the role of the proteasome in LPS-mediated signaling, we also analyzed the effect of the proteasome inhibitor lactacystin on LPS-induced phosphorylation of the MAPK superfamily. As illustrated in Fig. 7, LPS induced phosphorylation of ERK-1,2, JNK-1,2, and p38 MAPK. Although lactacystin alone had no effect on ERK-1 or ERK-2 phosphorylation, lactacystin pretreatment resulted in a significant dose-dependent inhibition of LPS-induced ERK phosphorylation. In contrast, lactacystin itself induced phosphorylation of JNK, and pretreatment of macrophages with lactacystin before stimulation with LPS resulted in increased JNK phosphorylation. Lactacystin also increased phosphorylation of p38, particularly in the absence of LPS. IRAK-1 is a key component of the TLR4- and IL-1R-mediated signaling FIGURE 9. Subunits of the human proteasome (39, 40). The mouse pathways, and it is recruited and activated after the interaction of proteasome has essentially the same subunits. X, Y, and Z subunits (IFN- the adaptor molecule MyD88 with the intracytoplasmic regions of inducible subunits) exhibit chymotrypsin, peptidyl glutamase, and trypsin TLRs or the IL-1R (35). To determine whether proteasome acti- activities, respectively. LPS binds to the C2 and N3 subunits. 1522 ROLE OF THE PROTEASOME IN LPS-INDUCED INFLAMMATION

Table II. Spot A: proteasome component PSMA1 C2 (EC 3.4.99.46; macropain subunit C2; proteasome NU chain; multicatalytic endopeptidase complex subunit C2; molecular mass-29.5 kDa). (This structure was obtained from Protein prospector website).

m/z Position Peptide Sequence

3114 218–243 DLEFTTYDDDDVSPFLDGLEERPQRK 2150 190–208 HGLRALRETLPAEQDLTTK 1342 244–256 AQPSQAADEPAEK 1239 51–61 RAQSELAAHQK

1 11213141 MFRNQYDNDV TVWSPQGRIH QIEYAMEAVK QGSATVGLKS KTHAVLVALK 51 61 71 81 91 RAQSELAAHQ KKILHVDNHI GISIAGLTAD ARLLCNFMRQ ECLDSRFVFD 101 111 121 131 141 RPLPVSRLVS LIGSKTQIPT QRYGRRPYGV GLLIAGYDDM GPHIFQTCPS 151 161 171 181 191 ANYFDCRAMS IGARSQSART YLERHMSEFM ECNLDELVKH GLRALRETLP 201 211 221 231 241 AEQDLTTKNV SIGIVGKDLE FTTYDDDDVS PFLDGLEERP QRKAQPSQAA 251 261 DEPAEKADEP MEH Downloaded from

LPS-initiated signaling pathways, these findings collectively sup- proteins indicates several basic amino acids in the C2 and N3 port the concept that the proteasome may well play an important subunits that we have shown to bind LPS. Given the anionic char- and, as yet not fully appreciated, regulatory role in LPS-mediated acteristics of LPS, it is attractive to consider that LPS may be induction of inflammation. associating selectively with these residues (Tables II and III). http://www.jimmunol.org/ The original objective of these studies was to identify important However, additional experiments will be required to confirm or and novel molecular targets in the macrophage with which LPS refute this hypothesis. We also found that the LPS probe would interacts, which might contribute to activation and secretion of bind to the ␣ subunit of the structurally simplified M. thermophilia inflammatory mediator molecules. The basic approach that was proteasome. In more recent experiments, we have examined by used was to synthesize a highly specific and chemically pure pho- photo-cross-linking the presence of LBPs in a human urethral cell toactivatable radiolabeled LPS probe and to examine its binding to line. In addition to selective labeling of proteasome subunit N3, constituents of partially purified macrophage fractions. Our expec- two heat shock proteins (heat shock cognate 71 kDa and 27-kDa tation was that this radiolabeled LPS probe would associate with at heat shock protein) were also identified (N. Qureshi and D. by guest on October 1, 2021 least some of the now reasonably well-characterized membrane Bjorling, unpublished data). Thus, it would appear reasonable to constituents recognized as critical to LPS signaling, including conclude that LPS interaction with cellular proteasomes is not CD14, TLR4, and MD-2. Interestingly, there has been one study unique to the macrophage, but may well be a general characteristic published in the literature that reports evidence to suggest that LPS 125 of LPS interactions with mammalian cells. specifically binds to TLR4 (21). However, in that study [ I]- Importantly, the consequences of interaction of LPS with pro- ASD-LPS was used to assess binding in cells in which TLR4 was teasome subunits appear not to be inconsequential. In this regard, overexpressed. Furthermore, to detect binding, TLR4 was tagged binding of the LPS to the M. thermophilia proteasome serves to and immunoprecipitated to concentrate this cellular constituent be- activate its postglutamase activity and chymotrypsin-like enzy- fore detection. It is noteworthy that the binding of LPS to TLR4 matic activity. The latter activity is also observed in partially pu- did not initiate signaling (21). The studies reported in this manu- rified rabbit muscle proteasome and murine macrophage 20S pro- script involve a direct labeling of cell membrane components using teasome after addition of LPS. It is of note that binding of LPS to cells in which membrane constituents have been overexpressed. the macrophage proteasome did not result in its activation of tryp- Surprisingly, the results of our studies consistently showed specific binding to PSMA1 C2 and PSMB4 N3 subunits of the murine sin-like activity, another enzyme activity closely associated with macrophage 20S proteasome. Proteasomes are known to be present the activated proteasome. These findings would be consistent with in the cytoplasm, nucleus, and endoplasmic reticulum of macro- the selective binding of the LPS to specific subunits as discussed phages (36), and our identification of proteasome subunits in the above. Furthermore, the fact that the enzymatic activity generated membrane fraction of macrophages would suggest that the mem- by interaction of LPS with the macrophage proteasome could be branes of the endoplasmic reticulum likely copurify with the cy- readily blocked by lactacystin (37, 38) suggests strongly that pro- toplasmic membrane fractions. If, as we suspect, endogenous lev- teolytic activity was correctly associated with the proteasome, els of TLR4 or CD14 are substantially lower than those of the rather than some unrelated protease activity. However, it might be proteasome subunits in macrophage subfractions, this would ac- noted that the available evidence would suggest that LPS does not count for preferential binding of LPS to proteasome subunits. Nev- bind to the proteasome exactly at the same site as does the peptide ertheless, the data clearly support interaction of LPS with selective substrate and lactacystin (Fig. 9). We postulate that LPS activates proteasome subunit constituents. the enzyme activity by opening up the proteasome channel to fa- LPS is known to bind to a variety of different mammalian-de- cilitate entry of the peptide substrate. In this regard, RsDPLA, the rived proteins, including LBP, bacterial permeability inducing pro- biologically inactive lipid A antagonist, may be binding at the tein, lysozyme, lactoferrin, myelin basic protein, and even albu- same site as LPS, but it does not facilitate the entry of the peptide min. Therefore, it is of interest that LPS binds selectively to the C2 substrate. The activation of chymotrypsin-like activity of the 20S and N3 subunits of the proteasome. Analysis of the proteasomal proteasome has been shown to be important in signaling because The Journal of Immunology 1523

Table III. Spot B: proteasome ␤ chain precursor PSMB4 N3 (macropain ␤ chain; multicatalytic endopeptidase complex ␤ chain; proteasome chain 3; molecular mass ϭ 24.36 kDa); propeptide in positions 1–45 has been removed. (This structure was obtained from Protein prospector website).

m/z Position Peptide Sequence

2603 241–264 GVEIEGPLSAQTNWDIAHMISGFE 1558 228–240 SYNRFQIATVTEK 1210 81–89 FRNISRIMR (1Met-ox) 1015 140–147 AMYSRRSK

111 21 3141 MEAFWESRAG HWAGGPAPGQ FYRIPATPSG LMDPASAPCE GPITRTQNPM 51 61 71 81 91 VTGTSVLGVK FDGGVVIAAD MLGSYGSLAR FRNISRIMRV NDSTMLGASG 101 111 121 131 141 DYADFQYLKQ VLGQMVIDEE LLGDGHSYSP RAIHSWLTRA MYSRRSKMNP 151 161 171 181 191 LWNTMVIGGY ADGESFLGYV DMLGVAYEAP SLATGYGAYL AQPLLREVLE 201 211 221 231 241 KQPVLSQTEA RELVERCMRV LYYRDARSYN RFQIATVTEK GVEIEGPLSA 251 261 QTNWDIAHMI SGFE Downloaded from

lactacystin and MG-132, which inhibit this activity in vitro, can then degraded by the proteasome. This leaves NF-␬B available to block LPS signaling. translocate to the nucleus, where it can serve to promote transac- These observations regarding the interaction of LPS with selec- tivation of NF-␬B-dependent genes. Lactacystin and other protea- tive subunits of the macrophage proteasome might be considered some inhibitors have been shown to block the degradation of I␬B. http://www.jimmunol.org/ within the framework of other proteasome activators. Other studies In addition, members of the MAPK superfamily can also modulate have documented that PA28, a known proteasome activator, also the activity of NF-␬B, and thus influence transcription of NF-␬B- binds to the C2 subunit, similar to what we have demonstrated for dependent genes (47). Recently, we have shown that MAPKs are LPS (39, 40). This activator has recently been shown to have im- important for LPS-induced TNF (48). Our findings reported here portant consequences for Ag processing (41). Mutant mice, who demonstrating that lactacystin will inhibit LPS-dependent activa- fail to express PA28, are characterized by impaired processing of tion of macrophages confirm that the proteasome may well be antigenic epitopes derived from either exogenous or endogenous involved in the activation of ERK-1 and -2 in response to LPS. Ags, consistent with the concept that compounds that function by We have also noted that lactacystin alone results in measurable by guest on October 1, 2021 activating the chymotrypsin activity of the proteasome are likely to increases in phosphorylation of p38 kinase and JNK, indicating be important for Ag processing. Given the fact that LPS is well recognized as a potent immunologic adjuvant that promotes im- that basal levels of proteasome proteolytic activity may negatively mune responses to unrelated protein Ags through a macrophage- regulate basic levels of activation of these two important signaling dependent pathway, it is attractive to consider LPS-dependent ac- molecules. Importantly, our studies also indicate that lactacystin is tivation of the proteasome as a potential pathway by which not able to diminish significantly LPS-induced levels of phosphor- potentiation of immune responses occurs. ylation of IRAK-1, even though proteasome activity has been re- One of the more interesting findings reported in this manuscript ported to be necessary for degradation of phosphorylated IRAK-1 is the fact that LPS-induced activation of the proteasome chymot- (49). Taken together, these experiments would support the concept rypsin-like activity in vitro can be totally blocked by lactacystin, a that the proteasome complex may be involved in regulation of potent proteasome inhibitor. Moreover, treatment of in vitro-cul- several of the very early steps that contribute to LPS-induced sig- tured macrophages with lactacystin for short periods of time before nal transduction pathways, specifically at the levels of MAPK and stimulation with LPS resulted in a dose-dependent inhibition of I␬B, resulting in inhibition of LPS-induced macrophage activation LPS-induced TNF-␣ secretion. In addition, lactacystin also inhib- and gene expression of proinflammatory cytokines. This would ited essentially all LPS-inducible genes measured (e.g., TNF-␣, further suggest that activation of IRAK-1 can be dissociated from IL-6, IL-12 p40 and p35, COX-2, and iNOS). Finally, LPS treat- the proteasome activity thought to be upstream and required for ment of macrophages leading to up-regulation of TLR2 mRNA normal activation of MAPK and I␬B. was also blocked by lactacystin, and this inhibitor also reduced We recognize that, given the intracellular location of the pro- constitutive levels of TLR4 mRNA expression. Collectively, these teasome, it would be necessary to postulate that LPS must effec- studies would be consistent with an important functional regula- tively be internalized into the macrophage in order for effective tory role for the proteasome in dictating pathways of LPS-depen- interaction with and activation of the proteasome to occur. It is dent signaling. possible that proteasomes may be present in the outer membrane of It is generally recognized that one of the key pathways up-reg- the cell, as has been previously reported for certain T cell sub- ulated by LPS involves activation of NF-␬B through dissociation populations (50). Nevertheless, there is ample evidence in the lit- and degradation of I␬B. It is of interest that proteasomes have previously been shown to be involved in LPS-induced degradation erature, both from our own studies and from those of others, that of I␬B (42–46). Current understanding of this pathway suggests LPS is rapidly internalized into macrophages. In our own pub- that, after LPS activation of macrophages, a cascade of adaptors lished research, LPS was found to be internalized and readily de- and kinases interacts with the intracytoplasmic domain of TLR4, tectable in the cytoplasm of macrophages within seconds of addi- leading to the phosphorylation and ubiquitination of I␬B, which is tion to macrophages (51). The internalized LPS therefore could be 1524 ROLE OF THE PROTEASOME IN LPS-INDUCED INFLAMMATION rapidly made available to the proteasome by a phagosome-to-cy- 14. Ingalls, R. R., B. G. Monks, R. Savedra, Jr., W. J. Christ, R. L. Delude, tosol transfer pathway as originally described by Kovascovics- A. E. Medvedev, T. Espevik, and D. T. Golenbock. 1998. CD11/CD18 and CD14 share a common lipid A signaling pathway. J. Immunol. 161:5413. Bankowski and Rock (52). Through this pathway, the proteasome 15. Lei, M. G., and D. C. Morrison. 2000. Differential expression of caveolin-1 in complex might well serve as a cytoplasmic “receptor” for LPS lipopolysaccharide-activated murine macrophages. Infect. Immun. 68:5084. secondary to TLR4-mediated signaling. 16. El-Samalouti, V. T., J. Schletter, I. Chyla, A. Lenschat, U. Mamat, L. Brade, H. D. Flad, A. J. Ulmer, and L. Hamann. 1999. Identification of the 80-kDa The concept of LPS internalization and interaction with cyto- LPS-binding protein (LMP80) as decay-accelerating factor (DAF/CD55). FEMS plasmic constituents would readily encompass a scenario in which Immunol. Med. Microbiol. 23:259. 17. Tohme, Z. N., S. Amar, and T. E. Van Dyke. 1999. Moesin functions as a lipo- Gram-negative bacteria might be engulfed by phagocytic cells polysaccharide receptor on human monocytes. Infect. Immun. 67:3215. such as macrophages, enter the cytosol, and interact directly with 18. Perera, P. Y., S. N. Vogel, G. R. Detore, A. Haziot, and S. M. Goyert. 1997. the 20S proteasome and thus facilitate increased proteolytic activ- CD14-dependent and CD14-independent signaling pathways in murine macrophages from normal and CD14 knockout mice stimulated with lipopolysaccharide ity of the proteasome. This would be expected, at a minimum, to or taxol. J. Immunol. 158:4422. lead to initiation of Ag presentation and triggering of a specific 19. Qureshi, N., K. Takayama, P. Mascagni, J. Honowich, R. Wang, and R. J. Cotter. acquired immune response, in addition to the potent innate im- 1988. Complete structural determination of lipopolysaccharides obtained from deep rough mutant of Escherichia coli: purification by high performance liquid mune response characterized by induction of proinflammatory cy- chromatography and direct analysis by plasma desorption mass spectrometry. tokines. Support for this concept derives from the recent studies of J. Biol. Chem. 263:11971. Maksymowych et al. (53), who have documented that invasion of 20. Dziarski, R. 1994. Cell-bound albumin is the 70-kDa peptidoglycan-, lipopolysaccharide-, and lipotechoic-acid-binding protein on lymphocytes and macro- macrophages by Salmonella typhimurium rapidly induces in- phages. J. Biol. Chem. 269:20431. creased expression of LMP, MECL, and PA28 proteasome genes. 21. Silva Correia, J. D, K. Soldau, U. Christen, P. S. Tobias, and R. J. Ulevitch. 2001. Finally, the data presented indicate that pretreatment of mice with Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex. J. Biol. Chem. 276:21129. Downloaded from lactacystin prevents LPS-induced shock in vivo. The galac- 22. Qureshi, N., K. Takayama, and E. Ribi. 1982. Purification and structural deter- tosamine-sensitized mouse model is a well-established model for mination of nontoxic lipid A obtained from the lipopolysaccharide of Salmonella LPS-induced shock, and the fact that lactacystin blocks mortality typhimurium. J. Biol. Chem. 257:11808. 23. McIntire, F. C., H. W. Sievert, G. H. Barlow, R. A. Finley, and A. Y. Lee. 1967. in this model suggests that it might be important as a potential Chemical, physical, and biological properties of a lipopolysaccharide from Esch- therapeutic approach for LPS-induced shock. Although additional erichia coli K-235. Biochemistry 6:2363. studies will be required to fully elucidate the role of LPS-protea- 24. Qureshi, N., B. Jarvis, and K. Takayama. 1999. Nontoxic RsDPLA as a potent Endotoxin in Health and Disease. antagonist of toxic lipopolysaccharide. In http://www.jimmunol.org/ some interactions in the host response to Gram-negative microor- H. Brade, S. M. Opal, S. N. Vogel, and D. C. Morrison, eds. Marcel Dekker, USA ganisms, the results of the studies presented here provide support p. 687. for the conclusion that such studies would be highly likely to be 25. Qureshi, N., K. Takayama, K. C. Meyer, T. N. Kirkland, C. A. Bush, L. Chen, R. Wang, and R. J. Cotter. 1991. Chemical reduction of 3-oxo and unsaturated informative. groups in fatty acids of diphosphoryl lipid A from the lipopolysaccharide of Rhodopseudomonas sphaeroides: comparison of biological properties before and after reduction. J. Biol. Chem. 266:6532. Acknowledgments 26. Jarvis, B., H. Lichenstein, and N. Qureshi. 1997. Diphosphoryl lipid A from We thank John J. Monaco for helpful discussions, Bruce Jarvis for his help Rhodobacter sphaeroides inhibits the complexes that form in vitro between li- with another probe during the initial stages of this research, Jon Johansen popolysaccharide (LPS)-binding protein, LBP, soluble CD14 and spectral-pure Infect. Immun. 65:3011. and Nancy Kendrick (Kendrick Laboratories) for 2-D-SDS-PAGE, and lipopolysaccharide. 27. Bhat, N., P. Y. Perera, J. M. Carboni, J. Blanco, D. T. Golenbock, T. N. Mayadas, by guest on October 1, 2021 Todd Sievert for preparing tritium-labeled E. coli for the initial studies. and S. N. Vogel. 1999. Use of a photoactivable taxol analogue to identify unique cellular targets in murine macrophages: identification of murine CD18 as a major taxol-binding protein and a role for Mac-1 in taxol-induced gene expression. References J. Immunol. 162:7335. 1. Morrison, D. C., and J. L. Ryan. 1987. Endotoxin and disease mechanisms. Annu. 28. O’Farell, P. H. 1975. High resolution two-dimensional electrophoresis of pro- Rev. Med. 38:417. teins. J. Biol. Chem. 250:4007. 2. Wright, S. D. 1988. Multiple receptors for endotoxin. Curr. Opin. Immunol. 3:83. 29. Manthey, C. L., P.-Y. Perera, C. A. Salkowski, and S. N. Vogel. 1994. Taxol 3. Schumann, R. R., S. R. Leonh, G. W. Flaggs, P. W. Gray, S. D. Wright, provides a second signal for murine macrophage tumoricidal activity. J. Immunol. J. C. Mathison, P. S. Tobias, and R. J. Ulevitch. 1990. Structure and function of 152:825. lipopolysaccharide-binding protein. Science 249:1429. 30. Perera, P.-Y., T. N. Mayadas, O. Takeuchi, S. Akira, M. Zaks-Zilberman, 4. Wright, S. D., R. A. Ramos, P. S. Tobias, R. J. Ulevitch, and J. C. Mathison. S. M. Goyert, and S. N. Vogel. 2001. CD11b/Cd18 acts in concert with CD14 and 1990. CD14, a receptor for complexes of lipopolysaccharides (LPS) and LPS Toll-like receptor (TLR-4) to elicit full lipopolysaccharide and taxol-inducible binding protein. Science 249:1431. gene expression. J. Immunol. 166:574. 5. Tobias, P. S., K. Soldau, L Kline, J. D. Lee, K. Kato, T. P. Martin, and 31. Medvedev, A. E., K. M. Kopydlowski, and S. N. Vogel. 2000. Inhibition of R. J. Ulevitch. 1993. Cross-linking of lipopolysaccharide (LPS) to CD14 on lipopolysaccharide-induced signal transduction in endotoxin-tolerized mouse THP-1 cells mediated by LPS-binding protein. J. Immunol. 150:3011. macrophages: dysregulation of cytokine, chemokine, and toll-like receptor 2 and 6. Kirkland, T. N., G. D. Virca, T. Kuus-Reichel, F. K. Multer, S. Y. Kim, and 4 gene expression. J. Immunol. 164:5564. R. J. Ulevitch. 1990. Identification of lipopolysaccharide-binding proteins in 70.Z/3 cells by photoaffinity cross-linking. J. Biol. Chem. 265:9520. 32. Maupin-Furlow, J. A., and J. G. Ferry. 1995. A proteasome from the methano- 7. Lei, M. G., and D. C. Morrison. 1988. Specific endotoxic lipopolysaccharide- genic Archaeon Methanosarcina thermophila. J. Biol. Chem. 270:28617. binding receptors on murine splenocytes. I. Detection of lipopolysaccharide-bind- 33. Hobler, S. C., A. Williams, D. Fischer, J. J. Wang, X. Sun, J. E. Fischer, ing sites on splenocytes and splenocyte subpopulations. J. Immunol. 141:996. J. J. Monaco, and P. O. Hasselgren. 1999. Activity and expression of the 20 S 8. Vita, N. S. Lefort, P. Sozzani, R. Reeb, S. Richards, L. K. Borysiewicz, proteasome are increased in skeletal muscle during sepsis. Am. J. Physiol. 277: P. Ferrara, and M. O. Labeta. 1997. Detection and biochemical characteristics of R434. the receptor for complexes of soluble CD14 and bacterial lipopolysaccharide. 34. Li, L., S. Cousart, J. Hu, and C. E. McCall. 2000. Characterization of interleu- J. Immunol. 158:3457. kin-1 receptor-associated kinase in normal and endotoxin-tolerant cells. J. Biol. 9. Hampton, R. Y., D. T. Golenbock, and C. R. H. Raetz. 1988. Lipid A binding Chem. 275:23340. sites in membranes of macrophage tumor cells. J. Biol. Chem. 263:14802. 35. Zhang, F. X., C. J. Kirchning, R. Mancinelli, X.-P. Xu, Y. Jin, E. Faure, 10. Schletter, J., H. Brade, L. Brade, C. Kruger, H. Loppnow, S. Kusumoto, A. Mantovani, M. Rothe, M. Muzio, and M. Arditi. 1999. Bacterial lipopolysac- E. T. Rietschel, H. D. Flad, and A. J. Ulmer. 1995. Binding of lipopolysaccharide charide activates nuclear factor-␬B through interleukin-1 signaling mediators in (LPS) to an 80-kilodalton membrane protein of human cells is mediated by sol- cultured human dermal endothelial cells and mononuclear phagocytes. J. Biol. uble CD14 and LPS-binding protein. Infect. Immun. 63:2576. Chem. 274:7611. 11. Triantifilou, K., M. Triantifilou, and R. L. Dedrick. 2001. A CD14-independent 36. Hirsch, C., and H. L. Pleogh. 2000. Intracellular targeting of the proteasome. LPS receptor cluster. Nat. Immunol. 2:338. Trends Cell Biol. 10:268. 12. Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. V. Huffel, X. Du, D. Birdswell, 37. Dick, L. R., A. A. Cruikshank, A. T. Destree, L. Grenier, T. A. McCormack, E. Alejos, M. Silva, C. Galanos, et al. 1998. Defective LPS signaling in C3H/HeJ F. D. Melandri, S. L. Nunes, V. J. Palombella, L. A. Parent, L. Plamondon, and and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282:2085. R. L. Stein. 1997. Mechanistic studies on the inactivation of the proteasome by 13. Akashi, S., R. Shimizu, H. Ogata, Y. Nagai, K. Takeda, M. Kimoto, and lactacystin in cultured cells. J. Biol. Chem. 272:182. K. Miyake. 2000. Cutting edge: cell surface expression and lipopolysaccharide 38. Geier, E., G. Peifer, M. Wilm, M. Lucchiari-Hatz, W. Baumeister, K. Eichmann, signaling via the Toll-like receptor4-MD-2 complex on mouse peritoneal mac- and G. Niedermann. 1999. A giant protease with potential to substitute for some rophages. J. Immunol. 164:3471. functions of the proteasome. Science 283:978. The Journal of Immunology 1525

39. DeMartino, G. N., and C. A. Slaughter. 1999. The proteasome, a novel protease 47. Means, T. K., R. P. Pavlovich, D. Roca, M. W. Vermeulen, and M. J. Fenton. regulated by multiple mechanisms. J. Biol. Chem. 274:22123. 2000. Activation of TNF-␣ transcription utilizes distinct MAP kinase pathways in 40. Rechsteiner, M., C. Realini, and V. Ustrell. 2000. The proteasome activator 11 S different macrophage populations. J. Leukocyte Biol. 67:885. Reg (PA28) and class I antigen presentation. Biochem. J. 345:1. 48. Jarvis, B., T. H. Harris, N. Qureshi, and G. A. Splitter. 2002. Rough lipopoly- 41. Preckel, T., W.-P. Fung-Leung, Z. Cai, A. Vitiello, L. Salter-Cid, O. Winquist, saccharide from Brucella abortus and Escherichia coli differentially activates the T. G. Wolfe, M. V. Herrath, A. Angulo, P. Ghazal, et al. 1999. Impaired immu- same mitogen-activated protein kinase signaling pathways for tumor necrosis noproteasome assembly and immune responses in PA28Ϫ/Ϫ mice. Science 286: factor ␣ in RAW 264.7 macrophage-like cells. Infect. Immun. 70:7165. 2162. 49. Yamin, T.-T., and D. K. Miller. 1997. The interleukin-1 receptor-associated ki- 42. Wang, X. C., R. Saban, J. H. Kaysen, M. R. Saban, P. L. Allen, E. N. Benes, and nase is degraded by proteasomes following its phosphorylation. J. Biol. Chem. T. G. Hammond. 2000. Nuclear factor ␬B mediates lipopolysaccharide-induced 272:21540. inﬂammation in the urinary bladder. J. Urol. 163:993. 50. Bureau, J. P., M. Olink-Coux, N. Brouard, S. Bayle-Julien, M. Huesca, 43. Stasiolek M., V. Gavrilyuk, A. Sharp, P. Horvath, K. Selmaj, and D. L. Feinstein. M. Herzberg, and K. Scherrer. 1995. Characterization of prosomes in human 2000. Inhibitory and stimulatory effects of lactacystin on expression of nitric lymphocyte subpopulations and their subpopulations and their presence as sur- oxide synthase type 2 in brain glial cells: the role of I␬B-␤. J. Biol. Chem. face antigens. Exp. Cell Res. 231:50. 275:24847. 51. Kutuzova, G. D., R. M. Albrecht, C. M. Erickson, and N. Qureshi. 2001. Diphos- 44. Wright, G., I. S. Singh, J. D. Hasday, I. K. Farrance, G. Hall, A. S. Cross, and phoryl lipid A from R. sphaeroides blocks the binding and internalization of T. B. Rogers. 2002. Endotoxin stress-response in cardiomyocytes: NF-␬B acti- lipopolysaccharide in RAW 264.7 cells. J. Immunol. 167:482. vation and tumor necrosis factor-␣ expression. Am. J. Physiol. Heart Circ. 52. Kovascovics-Bankowski, M., and K. L. Rock. 1995. A phagosome-to-cytosol Physiol. 282:H872. pathway for exogenous antigens presented on MHC class I molecules. Science 45. Moors, M. A., and S. B. Mizel. 2000. Proteasome-mediated regulation of 267:236. interleukin-1␤ turnover and export in human monocytes. J. Leukocyte Biol. 53. Maksymowych, W. P., T. Ikawa, A. Yamaguchi, M. Ikeda, D. McDonald, 68:131. L. Laouar, R. Lahesmaa, N. Tamura, A. Khuong, D. T. Y. Yu, and K. P. Kane. 46. Frost, R. A, G. J. Nystrom, and C. H. Lang. 2002. Lipopolysaccharide regulates 1998. Invasion by Salmonella typhimurium induces increased expression of the proinﬂammatory cytokine expression in mouse myoblasts and skeletal muscle. LMP, MECL, and PA28 proteasome genes and changes in the peptide repertoire Am. J. Physiol. Regul. Integr. Comp. Physiol. 283:R698. of HLA-B27. Infect. Immun. 66:4624. Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021