Role of MMPs in orchestrating inflammatory response in human monocytes via a TREM-1-PI3K-NF- B pathway

Transcription

1 Artile Role of MMPs in orhestrating inflammatory response in human monoytes via a TREM--PIK-NF- pathway Vanesa Gómez-Piña,, Eriel Martínez,, Irene Fernández-Ruíz, arlos del Fresno, Alessandra Soares-Shanoski,, Teresa Jurado, María Silieo, Vitor Toledano, Rosa Fernández-Palomares,, Franiso Garía-Rio, Franiso Arnalih, Subhra K. iswas, and Eduardo López-ollazo, Laboratory of Tumor Immunology, IdiPAZ, and Servies of Respiratory Diseases and Internal Mediine, La Paz Hospital, Madrid, Spain; EMPIREO Researh S.L., Madrid, Spain; and Singapore Immunology Network, iomedial Sienes Institutes, Singapore REEIVED JULY 8, ; REVISED FERUARY 9, ; AEPTED MARH,. DOI:.89/jlb.7 ASTRAT The MMPs onstitute a family of endopeptidases that an leavage extraellular proteins. They are involved in a number of events; some of these inlude inflammatory proesses. One of its targets is the TREM-, whih has emerged as an important modulator of innate immune responses in mammals. This transmembrane glyoprotein possesses an Ig-like etodomain readily shed by MMPs to generate strem-. Whereas membrane-anhored TREM- amplifies inflammatory responses, strem- exhibits anti-inflammatory properties. Here we show that sustained ell surfae expression of TREM- in human monoytes, through metalloproteinase inhibition, ounterats the well-haraterized down-regulation of several proinflammatory ytokines during the ET timeframe, also known as M or alternative ativation. In addition to the ytokines profile, other features of the ET phenotype were underdeveloped when TREM- was stabilized at the ell surfae. These events were mediated by the signal transduers PIKs and Syk. We also show that strem- ounterats the proinflammatory response obtained by membrane TREM- stabilization but failed to indue ET on naïve human monoytes. As the sustained TREM- expression at the ell surfae suffies to blok the progress of a refratory state in human monoytes, our data indiate that TREM- and MMPs orhestrate an Abbreviations: ADAM a disintegrin and metalloproteinase, AP allophyoyanin, AY AY -, F ysti fibrosis, DAP DNAXativating protein of moleular mass kda, EP Prostaglandin E Reeptor, ET endotoxin tolerane, IRAK IL-R-assoiated kinase, LAL Limulus amoeboyte lysate, MMP matrix metalloproteinase, p phosphorylated, PIK Phosphatidylinositol -kinases, Q-PR quantitative PR, s TREM- reombinant soluble triggering reeptor expressed on myeloid ells, st-inh standard protease inhibitor oktail, strem- soluble triggering reeptor expressed on myeloid ells, Syk spleen tyrosine kinase, t re time of reovery, TREM- triggering reeptor expressed on myeloid ells, t st time of stimulation, t tol time of tolerization The online version of this paper, found at inludes supplemental information. adaptive form of innate immunity by modulating the human monoytes response to endotoxin. J. Leuko. iol. 9: 9 95;. Introdution The MMPs belongs to the known family of endopeptidases impliated in inflammation. Through the use of models of human disease in mie with targeted deletions of individual MMPs, it has beome patent that MMPs at broadly in inflammation. It regulates barrier funtions, inflammatory ytokine [, ], as well as hemokine ativity and gradients [ 5], resolution of infetion, and tissue repair []. However, exessive release of these proteinases leads to severe tissue damage [7, 8]. A growing set of data suggests that different ell types an produe MMPs in response to endotoxins [], in partiular, inflammatory ells, suh as monoytes and neutrophils, whih express MMPs under a sepsis or after in vitro stimulation [9 ]. Interestingly, in human model of ET, marophages inreased their prodution of MMPs after a seond hallenge []; these data indiate a possible role of MMPs in the ontrol of this phenomenon. The term ET has been oined to desribe the state of redued responsiveness to an endotoxin hallenge after a primary baterial insult []; monoytes/marophages ritially mediate development of the ET phenotype in mie [5] and humans [, 7]. ET has been observed in several pathologies, suh as sepsis [8], aute oronary syndrome [9], and F []. This phenomenon, on one hand, has been assoiated with an inreased resistane and protetion from tissue injuries and even mortality []. On the other hand, some studies have suggested that the indution of ET. These authors ontributed equally.. orrespondene: Researh Unit, Laboratory of Tumor Immunology, IdiPAZ La Paz Hospital, 8 Madrid, Spain. elopez.hulp@ salud.madrid.org 7-5//9-9 Soiety for Leukoyte iology Volume 9, June Journal of Leukoyte iology 9

2 might impair resistane to infetion proess [,, ]. Moreover, it has been argued that the fatality of sepsis patients is, at least partially, attributed to their inability to respond to an unontrolled infetion. All of these data illustrate that ET is far more than a simple altered inflammatory response to []. In fat, ET is patent evidene that innate immunity may exhibit an adaptive response as a result of regulation of several innate immune reeptors (e.g., XR, D) [ 5], and it alters inflammatory signaling [, 8, 8]. In several previous studies, unresponsiveness has been assoiated with a dereased expression of NF- [9], an upregulation of IRAK-M [8, 7, ], as well as with deregulation of TLR []. In addition, we have demonstrated that a signifiant derease of membrane-bound TREM- in monoytes from F patients diretly orrelates with a refratory state []. TREM- belongs to a family of ell surfae moleules known as TREM. These are haraterized by an etodomain homologous to Ig-like modules. TREM-, whih is mainly expressed at the ell surfae of monoytes and neutrophils, is up-regulated by and other mirobial produts that strongly enhane leukoyte ativation [, ]. Although TREM- ligand(s) remain unknown, several lines of evidene indiate that reeptor engagement synergizes with the TLR pathway and up-regulates prodution of inflammatory ytokines [, ]. -indued reeptor expression appears to be at least partly mediated by endogenous PGE. The latter triggers EP and AMP/PKA-dependent mehanisms, whih are eventually followed by p8 MAPK ativation and PIK-mediated signaling [ ]. A strem variant has been deteted in mie and human serum [7, 8], as well as in AL fluids of pneumonia patients, plasma of septi patients, and in patients who suffer from systemi lupus erythematosus [9 ]. Other reports indiate that strem- is a useful marker of nosoomial sepsis []. In addition, the administration of rtrem-/f fusion proteins or antagonisti peptides resues mie from endotoxemia and polymirobial sepsis [, ]. We have shown previously that limited proteolysis of membrane-anhored TREM- is a relevant soure of strem- in -hallenged human monoytes and neutrophils []. We also demonstrated that MMPs are responsible for shedding the TREM- etodomain from the ell membrane, most likely through a single leavage within its long juxtamembrane linker []. A growing body of evidene indiates that refratoriness is modulated, at least in part, by ell surfae expression of TREM- and/or strem- generation. In this regard, urrent data suggest a dual role for the reeptor: on one hand, TREM- peptides exhibit a signifiant anti-inflammatory role [5]; on the other, the reeptor expression per se appears to amplify proinflammatory responses [, 5]. These findings, together with our previous observation of low TREM- levels in F monoytes after hallenge [], whih are loked into a refratory state [], prompted us to study in more detail the impliations of reeptor expression for ET. Here, we show that inhibition of MMPs with a syntheti metalloproteinase inhibitor ounterats the onset of a -refratory state in human monoytes. In addition, our data strongly suggest that membrane-anhored TREM- and the kinase PIK play a role in this proinflammatory mehanism. Further, we demonstrate that this blokade impairs rapid up-regulation of IRAK-M and deregulates other ruial elements, suh as I - and Akt. Our findings indiate that stabilization of TREM- at the ell surfae prevents the ET phenotype and suggest alternatives to modulate the endotoxin-tolerant state of human monoytes via inhibition of MMPs and/or PIK. MATERIALS AND METHODS Antibodies and reagents The following antibodies were used: anti-d-fit, anti-il-p5/p7, and anti-il- (Miltenyi iote, Auburn, A, USA); anti-trem--pe (R&D Systems, Minneapolis, MN, USA); anti-da-fit, anti-db-fit, and anti-d89-fit (Serote, Oxford, UK); anti-d-perp (Immunostep, Spain); anti-d9 (R&D Systems); anti-akt and anti-p-akt (ell Signaling Tehnology, Danvers, MA, USA); and anti-tnf-, anti-irak-m, anti-i, anti-p-i, and anti- -atin (Santa ruz iotehnology, Santa ruz, A, USA). The medium used for ell ulture was DMEM from Invitrogen (Paisley, UK). from Salmonella abortus was a kind gift from Dr. hris Galanos (Max-Plank-Insitut für Immunobiologie, Freiburg, Germany). The st-inh was from Rohe Diagnostis (Germany), and the general MMP inhibitor GM (also known as Ilomastat) was purhased from hemion International (Temeula, A, USA). The PIK inhibitor LY9 was obtained from ell Signaling Tehnology. s TREM-was provided by Abnova (Walnut, A, USA). The Syk inhibitor AY was purhased from albiohem (Darmstadt, Germany). All other reagents were obtained from Sigma- Aldrih (St. Louis, MO, USA), unless stated otherwise. PM isolation and ell ulture PMs were isolated from buffy oats by entrifugation on Fioll-Hypaque Plus (Amersham iosienes, Pisataway, NJ, USA), as we desribed before [8 ]. All reagents used for ell ulture were endotoxin-free, as assayed with the LAL test (ambrex, North runswik, NJ, USA). One seeded, adherent ells were treated or not with ng/ml for 8 h (t tol ; Supplemental Fig. A). After that, ells were washed three times with PS and kept in omplete medium for h (t re ). Then, ells were restimulated or not with ng/ml for periods of time ranging from to 8 h (t st ). Throughout this work, nontreated ells are denoted. The treatment termed tol orresponds to t tol 8h,t re h, and the indiated restimulation times, as illustrated in Supplemental Fig. A (see also Results). Note that this model was established before in our laboratory []. In other assays, strem-, at or 5 ng/ml, was added during the t re. The omposition of this adherent population of ells was analyzed by FAS (Supplemental Fig. ). In addition, ytokine prodution was tested to hek the model (Supplemental Fig. and D). For some experiments, a st-inh or the general MMP and ADAM family inhibitor, GM, were used. st-inh and GM were endotoxin-free, as assayed with the LAL test. Other speifi inhibitors, suh as LY9 and AY, were used in some experiments. RNA isolation and DNA synthesis ells were washed one with PS, and their RNA was isolated using the High Pure RNA isolation kit from Rohe Diagnostis, whih inludes a DNase I treatment. DNA was obtained by reverse transription of g RNA using the High apaity DNA reverse transription kit from Applied iosystems (Foster ity, A, USA). mrna quantitation Gene expression levels were analyzed by real-time Q-PR using the Lightyler system from Rohe Diagnostis and DNA obtained as desribed above. Q-PR was performed using a QuantiMix Easy SYG kit from iotools (Madrid, Spain) 9 Journal of Leukoyte iology Volume 9, June

3 Gómez-Piña et al. TREM- ounterats indution of tolerane and speifi primers (see below). Results were normalized to the expression of -atin, and the DNA opy number of eah gene of interest was determined using a seven-point standard urve as we desribed before [8 ]. Primers The sequenes of oligonuleotides used and their annealing temperatures are given in Table. All primers were synthesized, desalted, and purified by onsai ioteh (Madrid, Spain). ELISA TNF-, IL-, and TREM- onentrations of human TNF- and IL- in supernatants were determined using the ELISA development kit, supplied by PeproTeh (Roky Hill, NJ, USA), and a ommerial ELISA purhased from ender MedSystems (urlingame, A, USA), respetively. strem- levels were measured with an ELISA, supplied by R&D Systems. Protein array The amount of several hemokines in supernatants of ultured monoytes was measured using the Rayio Human hemokine Antibody Array from Rayioteh (Norross, GA, USA), following the manufaturer s instrutions. Eah film was sanned into an image-proessing and analysis program (Phoretix array v, TotalLab, Newastle, UK), and spots were digitalized. The densities were exported into Rayio Analysis Tool (Rayioteh). Data were normalized to % of positive ontrol. eton Dikinson A Monoytes were ultured at a density of 5 /ml in -well plates and treated following our model. The amount of several hemokines in supernatants of ultured monoytes was measured using the eton Diksinon A human inflammation kit (eton Dikinson, San Jose, A, USA). Data were normalized by a living ell. ell apoptosis analysis ell apoptosis was analyzed with the PE Anenexin V apoptosis detetion kit I (eton Dikinson), following the manufaturer s instrutions. For the apoptosis experiments, ells were then resuspended ( /ml) in annexin-binding buffer with annexin V and 7-amino-atinomyin D and inubated on ie for 5 min. Samples were analyzed by flow ytometry using a FASalibur (eton Dikinson), and results were analyzed with ellquestpro. Flow ytometri analysis Several surfae proteins were deteted using the speifi antibodies indiated above, as desribed previously [9, ]. For measurement of intraellular IL- and IL-, monoytes were stimulated for h in the presene of. nmol/l brefeldin A (iolegend, San Diego, A, USA), washed with PS, and finally, fixed at room temperature with FASlysis (eton Dikinson). ell membranes were permeabilized with PS ontaining % Triton X- and 5% FS. ells were then inubated for min in PS, supplemented with.% saponin, % FS, and PE-onjugated anti-il- and AP-onjugated anti-d or AP-onjugated anti-il- and FIT-onjugated anti-d-fit, washed, and analyzed by flow ytometry. Western blot analysis ell extrats were boiled in Laemmli buffer, resolved on 5% SDS-polyarylamide gels in 5 mm Tris, 5 mm glyine, and.% SDS, and transferred to Immun- lot PVDF membranes (io-rad, Herules, A, USA) at ma for.5 h at. After bloking for hintts ontaining 5% nonfat milk, membranes were washed three times in TTS alone and probed for h with anti-trem- antibody diluted in TTS. Following further washes in TTS, membranes were inubated for 5 min with a seondary, HRP-onjugated antibody (diluted :) and finally washed three times with TTS. The bound antibodies were deteted using EL Plus reagents, aording to the manufaturer s instrutions (Amersham Pharmaia ioteh, The Netherlands). Data analysis The number of experiments analyzed is given in eah figure legend. Data were olleted from a minimum of two experiments to alulate the mean and sd. The statistial signifiane was assessed using the unpaired Student s test, and differenes were onsidered signifiant at P.5. RESULTS Metalloproteinase inhibition abolishes the ET phenotype Monoytes are one of the primary effetors of innate immunity during infetion. However, after a first enounter with, these ells enter a refratory state, termed ET [, 7]. As we demonstrated that irulating monoytes isolated from F patients are loked into an endotoxin refratory state [, ] and express very low levels of TREM- at the ell surfae [], we deided to analyze a possible orrelation between the expression of the membrane-bound reeptor and the modulation of ET in human monoytes. In addition, we took into onsideration other authors who had also indiated a potential role of MMPs in ET [5,, ]. Previously, we had established a robust and reproduible model of ET, in whih human monoytes are pretreated for 8 h with (t tol ) and left to reover for a further h (t re ) before a seond endotoxin stimulation [] (see Materials and Methods and Supplemental Fig. A). Here, we used the ET model with essentially pure human monoytes (note that 9% of ells were monoytes; Supplemental Fig. ) and monitored the model, analyzing the expression of ytokines after a seond hallenge (Supplemental Fig. and D). TALE. Sequenes and Annealing Temperatures ( ) of Primers Used Sequene sense Sequene antisense Tm ( ) TNF- AAGAGTAGAGTGTGAG TAATTGTGGAAGGTGG IL- AAAGAATTAGATGAATAA GATTAAAAAAAATTG 5 IL-p GAATTAGTGTAAAGAGA TTGTTGTTTGATT 5 IL-p9 GTTATATAGTGTGG GAGGTTGGAATTGTGAG IL- AAGTGAGAAA TTAAGGGGTGGGTAG 8 IRAK-M TTTGAATGAGAGTTGA GATTGTTATGGAGAAT TREM- GGAGGAGAGATGAAGA AAGAGGAGAATGAAAT Atin GTATTAGAGATGT ATTTAGTTAGTTT Tm, Temperature. Volume 9, June Journal of Leukoyte iology 95

4 To study the possible role of TREM- and MMPs in ET, tolerant monoytes were exposed to for 8 h in the presene or absene of a speifi pan-metalloproteinase inhibitor, GM [9, ]. As expeted, GM inhibits the membrane-anhored TREM- shedding in both ontrol [] and tolerant human monoytes (Fig. ). Whereas the TREM- level was stabilizing at the ell surfae in the presene of GM (Fig. A and ), the generation of strem- was bloked (Fig. and D). As we tested before, other general proteinase inhibitors were unable to impede TREM- shedding (data not shown) []. Surprisingly, the tolerant phenotype was fully prevented by MMPs inhibition, as revealed by the mrna and protein quantifiation of two major proinflammatory ytokines: TNF- (Fig. A and D) and IL- (Fig. and E). The expression of these two ytokines was restored in ultures of tolerant human monoytes treated with GM (note that in the ase of TNF-, both sereted and intraellular onentrations were analyzed by ELISA and Western blot, respetively; Fig. D). Similar results were obtained from other ytokines (e.g., IL- and IL-; Supplemental Fig. ). Furthermore, tolerant ells, whih were restimulated with for,, or h in the presene of inreasing amounts of GM, restored their inflammatory phenotype in a dose-dependent manner (data not shown). esides, high levels of the anti-inflammatory ytokine IL- were maintained in tolerant monoytes, regardless of the presene or absene of the metalloproteinase inhibitor at early times of restimulation (Fig. and F). Note that the MMPs inhibitor did not affet the inflammatory response when ontrol ells were hallenged with in the presene of GM (data not shown). We stress that in all experiments, the modulation of expression of the ell surfae-anhored reeptor was verified by FAS analysis. Altogether, our data establish that a pan-metalloproteinase inhibitor is able to prevent the endotoxin-tolerant phenotype in human monoytes and thus, suggest that the presene of TREM- at the ell surfae ould play a role in this ontext. GM restores inflammatory ellular signaling in tolerant monoytes Several signaling pathways, suh as TLR/IRAK, I /NF-, MAPK, and PIK/Akt, are notably influened by an endotoxin refratory state. In partiular, IRAK-M expression and phosphorylation of Akt and I - are affeted dramatially during the ET time-frame [, 7]. We and others have shown previously that the pseudokinase IRAK-M is up-regulated rapidly in tolerant monoytes after hallenge and plays an important role in the ontrol of ET [8, 7,, 8]. Here, a time-ourse study suggested that tolerant ells stimulated with reovered their normal IRAK-M expression kinetis in the presene of GM (Fig. A). Most notably, the fast up-regulation of IRAK-M in tolerant monoytes hallenged with is almost abolished ompletely in the presene of the metalloproteinase inhibitor. These mrna expression data were orroborated at the protein level (Fig. ). The kinetis of phosphorylation of Akt and I - after hallenge have been explored extensively; both proteins and in partiular, I -, are phosphorylated rapidly in monoytes following treatment [7]. y ontrast, our results indiate that phosphorylation of Akt is abolished almost ompletely during ET, and I - phosphorylation is retarded ( min; Fig. ). However, the presene of GM almost fully restored the kinetis desribed for -hallenged, naïve monoytes, as indiated by the appearane of a strong band orresponding to p-akt after 5 min of stimulation (Fig. ). In line with this finding, in those ultures of tolerant monoytes, where GM was added, p-i - was deteted after just 5 min of treatment. We stress that in all experiments, the expression of ell surfae-anhored TREM- was established by FAS analysis (data not shown). From the results of these assays, we onlude that the metalloproteinase inhibitor, GM, essentially restores the normal kinetis of phosphorylation of several ritial monoyte-signaling fators upon endotoxin stimulation. A TREM- positive ells (%) D TREM- RP GM GM (nm).5. tol tol tol tol 8 h tol tol GM 8 8 h ounts 5 5 tol (8h) tol GM (8h) TREM- strem- (ng/ml) tol tol GM GM Figure. GM stabilizes membrane-anhored TREM- and inhibits strem- generation in tolerant ells. (A and ) irulating human monoytes were isolated from healthy volunteers, and then ultures were pretreated (tol) or not [ontrol ()] with ng/ml following the model established (see Supplemental Fig. A). Then, ontrol and -pretreated monoytes were hallenged with ng/ml for indiated times in the presene or absene of GM. (A) ells were harvested and stained with anti- TREM- antibody followed by flow ytometri analysis; the fration of ells stained is given. Perent of positive ells is showed (n ). P.5; tol versus tol GM. () A typial histogram obtained from flow ytometri analysis is showed. GM, GM. () strem- onentrations were measured in ulture supernatants using a ommerial ELISA (n ). P.5; tol versus tol GM. (D) The presene of strem- in supernatants of ultures of tolerant (tol) and ontrol monoytes exposed to, in the presene or not of different onentrations of GM, was also studied by Western blot analysis using an anti-trem- antibody. A Poneau red-stained membrane (RP) is shown as a loading ontrol. (The results of a typial experiment are shown; n ). Note that in those onditions that inlude GM, the inhibitor was kept in the ell ulture during the whole experiment. 8 8 h 9 Journal of Leukoyte iology Volume 9, June

5 Gómez-Piña et al. TREM- ounterats indution of tolerane A D [TNF / -atin] x tol tol GM 8 8 h TNF (ng/ml) x tol tol GM hours GM TNF-α β-atin tol tol (GM) [ILp/ -atin] x tol tol GM 8 8 h E IL p5/p7 (MFI) x tol tol GM [IL/ -atin] x h F IL (MFI) x tol tol GM. tol tol GM Figure. A speifi MMP inhibitor (GM) reinstates the inflammatory phenotype in tolerant human monoytes. Human monoytes were isolated from healthy volunteers and pretreated or not with ng/ml. Then, ontrol and tolerant monoytes were hallenged with ng/ml for indiated times in the presene or absene of M GM. (A ) ells were harvested, total RNA isolated, and DNA synthesized. (A) TNF-, () IL-p, and () IL- mrna levels were quantitated by real-time Q-PR using speifi primers. The ratio [gene/ -atin] is given (n ). P.5; tolerant (tol ) versus the same ondition in the presene of GM (tol GM). (D) -pretreated and ontrol monoytes were hallenged with ng/ml for and h in the presene or absene of GM. (Left) TNF- onentrations in the supernatant of eah ulture, h after stimulation, were evaluated using a ommerial ELISA (n ). P.5; tolerant (tol ) versus the same ondition in the presene of GM (tol GM). (Right) Total intraellular proteins were reovered and Western blots performed with anti-tnf- and anti- -atin antibodies. The results of one experiment are shown (n ). (E and F) -pretreated and ontrol monoytes were hallenged with ng/ml for hinthepresene or absene of GM. Intraellular IL-p5/p7 (E) and IL- onentrations (F) were evaluated by staining with speifi antibodies, followed by flow ytometri analysis; the fration of ells stained is given (n ). P.5; tolerant (tol ) versus the same ondition in the presene of the metalloproteinase inhibitor (tol GM). Note that in onditions where GM was used, it was kept in the ell ulture during the entire assay period. MFI, Mean fluorese intensity. A PIK inhibitor restores the ET phenotype in tolerant ells treated with GM We have demonstrated previously that the general PIK inhibitor, LY9, aelerates IRAK-M expression in human monoytes treated with [8]. In addition, other authors have reported that TREM--mediated ytokine prodution requires PIK signaling [5] and that PIK ativity influenes the phosphorylation state of Akt and I - [7, 9]. Our urrent findings suggest that metalloproteinase inhibition and subsequently, stabilization of membrane-anhored Volume 9, June Journal of Leukoyte iology 97

6 A [IRAK-M/ -atin] x - tol tolgm hours - tol GM tol GM IRAK-M β-atin - (h) tol tol GM GM min GM p-akt AKT p-i-κα I-κα β-atin Figure. GM reversed the tolerant phenotype in human monoytes, whih were isolated from healthy volunteers and pretreated or not with ng/ml in the presene or absene of GM ( M). (A) ultures were hallenged with ng/ml for indiated times. Next, ells were harvested, total RNA isolated, and DNA synthesized. IRAK-M mrna levels were evaluated by real-time Q-PR, and the ratio [IRAK-M/ atin] is shown (n ). () Tolerant and ontrol monoytes were hallenged with ng/ml for or h in the presene ( ) or absene ( ) of GM. Then, total proteins were isolated, and IRAK-M and -atin expression were analyzed by Western blotting. The results of a representative experiment are shown (n ). () Tolerant and ontrol monoytes were exposed ( ) or not ( ) to ng/ml for indiated times in the presene or absene of GM ( M). Next, ells were harvested, and total protein was isolated. A Western blot analysis was performed using speifi antibodies; -atin was used as a loading ontrol. Data shown are representative of four independent experiments. Note that in those onditions where GM was added, the inhibitor was kept in the ell ulture during the entire assay. TREM- oppose rapid IRAK-M up-regulation in tolerant monoytes, whereas upon hallenge, Akt and I - are phosphorylated aording to normal kinetis in these ells (Fig. ). These observations prompted us to study in more detail the role of PIK in modulating the ET phenotype. ultures of ontrol and tolerant monoytes were exposed to in the presene or absene of GM and LY9. As expeted, the expression of proinflammatory ytokines (TNF-, IL-, IL-p, and IL-p9) was inhibited signifiantly in tolerant monoytes treated with in the presene of both inhibitors, as ompared with the same ondition without LY9 (Fig. ). Moreover, extra- and intraellular TNF- and intraellular IL- protein onentrations exhibited a profile aording to our findings at the mrna level (Fig. 5A and ). We also evaluated the kinetis of phosphorylation of Akt and I - in our model, in the presene or absene of the PIK inhibitor. As Fig. 5 shows, LY9 re-established the tolerant phenotype in those ultures where metalloproteinase ativity was inhibited by GM. In all ases, the presene of TREM- at the ell surfae was evaluated by flow ytometry analysis (Fig. 5D, left panel). In addition, we tested that strem- generation was not affeted by the presene of LY9 in tolerant monoytes (Fig. 5D, right panel). A Syk inhibitor also restores the ET phenotype in tolerant ells treated with GM Several authors have desribed that Syk is involved in the downstream pathway of TREM- DAP [,, 5]. In addition, there are published data demonstrating a diret impliation of Syk in this pathway, ontrolling PIK ativity [5 5]. Our findings indiate that after MMPs inhibition, the ET phenotype does not take plae in human monoytes when TREM- shedding is inhibited. To onnet these two events, we analyzed the ET phenotype in tolerant monoytes exposed to GM in the presene of the very speifi Syk inhibitor, AY. Interestingly, the tolerant phenotype was reinstated in those ultures treated with GM but in the presene 98 Journal of Leukoyte iology Volume 9, June

7 Gómez-Piña et al. TREM- ounterats indution of tolerane A [TNF / -atin] x GM LY9 D [ILp9/ -atin] x - 8 GM LY9 h of tol tol tol tol h of tol - tol tol - tol [IL/ -atin] GM LY9 h of tol tol tol tol [ILp/ -atin] x GM LY9 h of tol tol tol tol Figure. PIK inhibition reinstates the tolerant phenotype in human monoytes treated with GM. Human monoytes were isolated from healthy volunteers and pretreated or not with ng/ml in the presene or absene of GM ( M) and LY9 (5 M). Next, ultures were hallenged with ng/ml for,, or h. ells were harvested, total RNA isolated, and DNA synthesized. The mrna levels of TNF- (A), IL- (), IL-p (), and IL-p9 (D) were analyzed by real-time Q-PR, and ratios [gene/ -atin] are given (n ). P.5; tol GM versus tol GM LY9. of AY (Fig. A). Protein levels of IL-, IL-, and IL- were not restored when the signaling downstream of TREM- was bloked by the Syk inhibitor. However, strem- generation was not affeted by Syk inhibition, as shown in Fig.. strem- ounterats the proinflammatory response obtained by membrane TREM- stabilization Findings presented above suggest that membrane-bound TREM- plays an important role in the ontrol of ET in human monoytes, as interferene with its metalloproteinasemediated shedding impairs its refratory state. In addition, our data indiated that although kineti of expression of membrane-anhored TREM- after stimulation was not signifiantly affeted in tolerant monoytes, the strem- is aumulated at similarly high levels and at all times tested, only on supernatants of -tolerant ultures (Fig. ). Equivalent onentrations of strem- are observed in ultures of ontrol monoytes with a delay of 8 h after hallenge. Thus, we deided to study whether the soluble reeptor form influenes the ET program in our system. To test this hypothesis, we used s TREM-, whih omprises almost the entire reeptor etodomain. Aording to previous data, the presene of s TREM- impaired TNF- and IL- mrna expression indued by hallenge and up-regulated IL- mrna levels in naïve human monoytes in a dose-dependent manner (see Supplemental Fig. ). In addition, s TREM- redues onentrations of TNF- and IL- proteins in the ulture supernatant of monoytes treated with in the presene of GM and up-regulated the IL- prodution (Fig. 7A ). Note that s TREM- and were added simultaneously. The anti-inflammatory properties of s TREM- were also evidened by the analysis of a set of hemokines, XL9, XL, and XL5, previously established as M and M markers []. In ontrast, despite other authors having lassified L as an M marker [5, 55], this hemokine exhibited a patent up-regulation in tolerant ells stimulated with. In addition, the MMPs inhibitor redued L levels, and when strem- was added, we observed a signifiant inrease of L prodution (Fig. 7D). Note that although in mie, L expression dereases in tolerant marophages hallenged with [5, 55], in a tumor environment, L promotes a M phenotype [5]. esides that, levels of L7 and L expression were impaired in ativated () and tolerant (tol ) monoytes. The analysis of several proinflammatory pathways also revealed that s TREM- reindued an anti-inflammatory profile in tolerant monoytes treated with and GM (Fig. 7E). strem- failed to indue an ET phenotype in human monoytes Results showed above indiate that reversion of the ET phenotype aused by metalloproteinase inhibition is nullified by the presene of strem-. In this ontext, we deided to test whether strem- has a speifi role in ET indution, or its ativity is limited to an anti-inflammatory response in a similar way as IL- [, 57]. Then, naïve, human monoytes were exposed to s TREM-, following the model desribed in Fig. 8A. We added a new ondition, where ontrol mono- Volume 9, June Journal of Leukoyte iology 99

8 A TNF (ng/ml) GM LY9 tol tol tol hours GM TNF-α β-atin tol none tol (GM) LY9 IL p5/p7 (MIF) x tol tolgm tolgmly tol GM D 8 none min GM LY9 p-akt AKT p-iκ-α tol none LY TREM-positive ells (%) 8 h strem- (ng/ml) 5 GM LY GMLY Iκ-α β- atin tol tolgm tolly tolgm LY tol Figure 5. A PIK inhibitor restores the expression pattern of tolerant ells in human monoytes treated with GM. Human monoytes were pretreated or not with ng/ml for indiated times in the presene or absene of M GM and 5 M of LY9 (LY). Next, ultures were hallenged with ng/ml for indiated times. (A, left) Total intraellular proteins were reovered and Western blots performed with anti-tnf- and anti- -atin antibodies. The results of one experiment are shown (n ). (Right) TNF- onentrations in the supernatants of eah ulture, h after stimulation, were evaluated using a ommerial ELISA (n ). P.5; tolerant (tol ) versus the same ondition in the presene of GM (tol GM). () Intraellular IL-p5/p7 was evaluated hafter hallenge by staining with a speifi antibody, followed by flow ytometri analysis; the fration of ells stained is given (n ). P.5; tol GM versus the same ondition in the presene of the PIK inhibitor [tol GM LY9 (LY)]. () Time ourse of I - and AKT phosphorylation analyzed via Western blot of patterns with speifi antibodies; -atin was used as a loading ontrol. Data shown are representative of three independent experiments. (D) TREM- expression and strem- generation were evaluated. (Left) TREM- expression was analyzed by staining with a speifi antibody, followed by flow ytometri analysis; the fration of ells stained is given (n 5). Note that in experiments where GM was used, it was kept in the ell ulture during the entire assay period. (Right) Human monoytes isolated from healthy volunteers were pretreated with ng/ml. Next, levels of strem- were quantified by ELISA at the supernatant of these tolerant human monoytes hallenged with ng/ml for 8 h in the presene of M GM, 5 M LY9, and ombinations. P.5; all onditions versus tol. ytes were exposed to s TREM- for h to simulate the strem- generation after the pretreatment in tolerant monoytes (s TREM- ). Interestingly, pretreatment of monoytes with s TREM- for h failed to indue a tolerant phenotype, as was observed by the standard proinflammatory ytokine profile obtained after a hallenge (Fig. 8 D). s TREM- was unable to reprodue the ET phenotype. Note that strem- was tested in two different onentrations ( ng/ml and 5 ng/ml), and no signifiant effet was observed. In addition, several periods of inubation were tested with similar results, and protein data onfirmed our findings (data not shown). olletively, these demonstrate that despite its anti-inflammatory ativity, strem- fails to indue a refratory status in human monoytes. DISUSSION The MMPs were initially believed to play a role solely in aner. However, many pathologial onditions are haraterized by the overexpression of these enzymes (e.g., sepsis and septi shok) []. Aording to several linial trials, the disappointing results obtained with MMP inhibitors in the aner field were mainly a result of severe side-effets in long-term treatments. However, aute diseases, suh as sepsis and septi shok, usually require only short-term treatment. Here, we show that a general MMP inhibitor affets the levels of membrane TREM- and subsequently, the inflammation and the ET framework of human innate immune ells. It is well known that MMPs play a ruial role 9 Journal of Leukoyte iology Volume 9, June

9 Gómez-Piña et al. TREM- ounterats indution of tolerane A [Protein] (ng/ml) strem- (ng/ml) 5 5 IL IL x - IL tol tol none GM AY GM AY in the generation of soluble forms of several membraneanhored proteins [5, ], and its expression is up-regulating in the presene of endotoxins, suh as [, 58, 59]. We have shown previously that membrane-anhored TREM- is shed after a proteolyti event atalyzed by MMPs to generate strem- []. Note that these data have been verified in the urrent artile (Fig. ). In addition, we have learned that monoytes from patients with various onditions, leading to or assoiated with impaired immune response, present low levels of TREM- [], and several groups have independently demonstrated that the reeptor is overexpressed on the monoyte surfae after TLR engagement by [9,,, ]. Interestingly, whereas membrane-anhored TREM- amplifies inflammatory responses, strem- exhibits anti-inflammatory properties [8, ]. These observations, among others, prompted a thorough study of the interation among MMP inhibition, membrane TREM- levels, and ET. Our urrent findings provide strong evidene that the standard ET phenotype, haraterized by a signifiant downregulation of proinflammatory ytokines, suh as TNF-, IL-, IL-p, and IL-p9, is prevented by GM, a none GM AY GM AY Figure. A Syk inhibitor (AY) restores the expression pattern of tolerant ells in human monoytes treated with GM. Human monoytes isolated from healthy volunteers were pretreated with ng/ml. Then, these tolerant monoytes were hallenged with ng/ml for the indiated h in the presene or absene of GM ( M), AY ( M), or ombinations. (A) Extraellular proinflammatory ytokines IL-, IL-, and IL- were quantified by A (n 5). P.5; tol Syk-I and tol GM Syk-i versus tol GM. () Levels of strem- were evaluated at the supernatant of these ultures by ELISA (n 5). P.5; all onditions versus tol. general MMP and ADAM family inhibitor able to blok the TREM- shedding [, ]. In ontrast, the anti-inflammatory ytokine IL- is not signifiantly affeted at the early stages of hallenge. This is onsistent with the previous observation that despite its anti-inflammatory effet, IL- does not indue tolerane in human monoytes per se [, 57]. Moreover, the role of IL- as a key regulator in tolerane has been diretly refuted in mie, as several studies demonstrate that ET takes plae in IL-( / ) animals [57, ]. However, at late points, levels of IL- were signifiantly higher in tolerant ultures stimulated with in the presene of GM than in the absene of the inhibitor. In line with our findings, others have demonstrated that partial silening of TREM- mrna was assoiated with a redution of IL- prodution during peritonitis-indued sepsis []. In addition, the proinflammatory intraellular signaling was restored in the presene of GM in tolerant monoytes stimulated with. We and others have reported an essential role for IRAK-M in the ontrol of ET [8, 9,, 8]. Our previous data onfirmed that this pseudokinase was up-regulated rapidly in tolerant monoytes exposed to, thus bloking the TLR/IRAK pathway [7]. y ontrast, in the presene of GM, the kinetis of IRAK-M expression in tolerant monoytes was similar to that observed in naïve ells. These findings suggest a new role for TREM- in the ontrol of inflammation; its presene at the ell surfae modulates the expression of ruial negative fators, suh as IRAK-M. Analysis of other pathways also onfirmed reversion of the ET phenotype in the presene of the pan-metalloproteinase inhibitor (GM). It is well known that the NF- pathway is affeted in ET, and one of the main events bloked is the early phosphorylation of I - upon hallenge []. y ontrast, the NF- inhibitory protein is normally phosphorylated in tolerant monoytes in whih TREM- is permanently expressed at the ell surfae. Moreover, we have observed that the presene of GM indued p-akt in tolerant ultures exposed to. Therefore, TREM- leads to ativation of several transription omplexes that synergize with NF- toindue transription of proinflammatory genes []. It has been established that TLR signaling involves the ativation of PIK and MAPKs, all of whih influene the -indued inflammatory responses [5,, 5]. Moreover, PIK plays an important role in the ontrol of several pathways impliated in ET, and its expression is inreased during a refratory state [7, ]. Our previous results showed that the PIK inhibitor, LY9, aelerates IRAK-M expression in human monoytes treated with [8], thus affeting the ET proess. In addition, it is well known that PIK promotes signal amplifiation indued by TREM- [5]. As disussed above, sustained expression of TREM- at the ell surfae abolishes the ET phenotype in human monoytes. However, when a PIK inhibitor was added to ultures of tolerant monoytes treated with GM, all ET features, suh as down-regulation of proinflammatory ytokines and delayed I - phosphorylation, were re-established. Volume 9, June Journal of Leukoyte iology 9

10 A D TNF (ng/ml) GM s TREM- tol tol tol IL p5/p7 positive ells(%) GM s TREM- tol tol tol E min of s TREM p-akt AKT a.u. p-i-κα XL9 XL XL7 L L5 L tol GM tol GM s TREM- tol tol GM tol GM s TREM- IL positive ells(%) GM s TREM- tol tol tol I-κα β-atin Figure 7. strem- inhibits the inflammation in tolerant human monoytes treated with GM. Human monoytes were pretreated or not with ng/ml in the presene or absene of GM. Next, ultures were hallenged with ng/ml for indiated times and in the presene or not of 5 nm s TREM-. (A) TNF- onentrations were evaluated in the supernatant of eah ulture, h after addition, using a ommerial ELISA (n ). ( and ) Intraellular levels of IL-p5/p7 and IL- were determined, h after treatment, by staining with speifi antibodies, followed by flow ytometri analysis; the fration of ells stained is given (n ). P.5; tol GM versus the same ondition in the presene of s TREM-(tol GM s TREM-). (D) The onentrations of several hemokines were measured in the supernatant of eah ulture, h after addition, using the Rayio Human hemokine Antibody Array (n ). P.5; tol GM versus the same ondition in the presene of strem- (tol GM s TREM-). (E) A Western blot analysis was performed using speifi antibodies; -atin was used as a loading ontrol. Data shown are representative of three independent experiments. As the metalloproteinase inhibitor ould interfere with proteolysis of other membrane proteins, the atual influene of TREM- in the observed phenomenon was studied. The inhibition of another well-known fator of the downstream TREM- pathway, the kinase Syk, indiated the diret impliation of TREM- in the phenomenon observed. As other authors have demonstrated, Syk plays an important role in those pathways, where DAP is involved [5, 5]. In our system, Syk inhibition restores the ET phenotype but as well as the PIK inhibitor LY9, did not affet the TREM- shedding. In addition, with the help of s TREM-, we demonstrate that despite its evident anti-inflammatory ativity, strem- does not indue a refratory state in theses ells (Fig. 8). strem- plays a ruial role in the ontrol of inflammation, as its presene redued the -indued response and reversed the effet of GM on tolerant monoytes. This observation was substantiated by following its ytokine prodution and its hemokine expression and by analyzing its NF- and Akt signaling pathways. In this regard, previous reports indiate that strem- exhibits anti-inflammatory ativity [8,, 5]. However, our data also indiate that the strem- is unable to indue an ET phenotype in human monoytes. As well as IL-, strem- only affets the inflammatory response but never provokes a re-eduation of human monoytes in this ontext. In full agreement with our urrent data, we had demonstrated previously that patients who suffer from F are loked in an ET status [, 9 Journal of Leukoyte iology Volume 9, June

11 Gómez-Piña et al. TREM- ounterats indution of tolerane A D [ILp9]/[ -Atin] x - t tol = 8 h s TREM- t re = h t re = h t st t st t st tol s TREM-() s TREM-(5) Analysis ( ) Analysis (tol) Analysis (s TREM-) h h [TNF ]/[ -Atin] tol s TREM-() s TREM-(5) h h [ILp/ β-atin] x tol s TREM-() s TREM-(5) Figure 8. strem- failed to indue an ET phenotype in human monoytes. (A) Shemati representation of the model used in the urrent study (t tol,t re, and t st indiate times of tolerane indution, reovery, and restimulation with endotoxin, respetively). ( D) ontrol and tolerant and pretreated with s TREM- ( and 5 ng/ml) human monoytes were hallenged with ng/ml for and h following the model represented in A. Then, ells were harvested, total RNA isolated, and DNA synthesized. () TNF-, () IL-p, and (D) IL-p9 mrna levels were quantitated by real-time Q-PR using speifi primers. The ratio [gene/ -atin] is given (n ). h h ], and their irulating monoytes did not express membrane-anhored TREM- after an hallenge []. Moreover, strem- was not deteted in the serum of these patients, and it supports our present results that indiated that the soluble form is not neessary for ET indution []. How does the sustained presene of membrane-bound TREM- blok development of the ET phenotype in monoytes? It is an open question. A potential explanation ould be that the TREM- DAP Syk omplex might be entral to a larger signaling omplex that onstitutively promotes proinflammatory responses; thus, only reeptor removal after metalloproteinase leavage would allow the development of the unresponsiveness state. This and others hypotheses will be tested in future researh. In summary, as sustained membrane-anhored expression of TREM- suffies to blok the progress of a refratory state in human monoytes; our data indiate that TREM- and MMPs orhestrate an adaptive form of innate immunity by modulating the human monoyte response to endotoxin. This event is mediated, at least in part, by PIKs. Disruption of the balane between inflammation and ET plays an important role in a number of diverse linial situations (e.g., sepsis, septi shok, F, and aute oronary syndrome). Therefore, the regulatory mehanism desribed in the urrent manusript and in partiular, more speifially, the use of MMP inhibitors as potential modulators of -refratory states might be taken into onsideration when designing new, linial strategies for treating onditions haraterized by unbalaned inflammatory/tolerant responses. In this regard, more seleted MMP inhibitors will help unravel the important roles that individual MMPs play in TREM- versus strem- expression and subsequently, in the ontrol of infetion, inflammation, and septi shok. AUTHORSHIP E.L-. oneived of and designed the experiments. V.G-P., E.M., I.F-R.,.d.F., A.S-S., T.J., M.S., V.T., R.F-P., F.G-R., and F.A. performed the experiments. E.L-., V.G-P., E.M.,.d.F., and S.K.. analyzed the data. E.L-. wrote the paper. AKNOWLEDGMENTS This study was supported by grants from the former Ministerio de Eduaión y ienia, from the former Ministerio de ienia e Innovaión, from Fondos de Investigaión Sanitaria, and from Fundaión Média de la Mutua Madrileña de Automovilístia to E.L-. We thank Dr. Sylvia Hottinger-raig for her proofreading. DISLOSURES The authors delare no ompeting finanial interests. REFERENES. Haro, H., rawford, H.., Fingleton,., Shinomiya, K., Spengler, D. M., Matrisian, L. M. () Matrix metalloproteinase-7-dependent release of tumor nerosis fator- in a model of herniated dis resorption. J. lin. Invest. 5, 5.. Shonbek, U., Mah, F., Libby, P. (998) Generation of biologially ative IL- by matrix metalloproteinases: a novel aspase--independent pathway of IL- proessing. J. Immunol.,.. Dean, R. A., ox, J. H., ella,. L., Douet, A., Starr, A. E., Overall,. M. (8) Marophage-speifi metalloelastase (MMP-) trunates and inativates ELR X hemokines and generates L, -7, -8, and - antagonists: potential role of the marophage in terminating polymorphonulear leukoyte influx. lood, 55.. Van den Steen, P. E., Husson, S. J., Proost, P., Van Damme, J., Opdenakker, G. () arboxyterminal leavage of the hemokines MIG and IP- by gelatinase and neutrophil ollagenase. iohem. iophys. Res. ommun., Garton, K. J., Gough, P. J., Raines, E. W. () Emerging roles for etodomain shedding in the regulation of inflammatory responses. J. Leuko. iol. 79, 5. Volume 9, June Journal of Leukoyte iology 9

12 . Vanlaere, I., Libert,. (9) Matrix metalloproteinases as drug targets in infetions aused by gram-negative bateria and in septi shok. lin. Mirobiol. Rev., Lagente, V., oihot, E. () Role of matrix metalloproteinases in the inflammatory proess of respiratory diseases. J. Mol. ell. ardiol. 8,. 8. Mott, J. D., Werb, Z. () Regulation of matrix biology by matrix metalloproteinases. urr. Opin. ell iol., usiek, D. F., aragi, V., Nehring, L.., Parks, W.., Welgus, H. G. (995) Matrilysin expression by human mononulear phagoytes and its regulation by ytokines and hormones. J. Immunol. 5, Lai, W.., Zhou, M., Shankavaram, U., Peng, G., Wahl, L. M. () Differential regulation of lipopolysaharide-indued monoyte matrix metalloproteinase (MMP)- and MMP-9 by p8 and extraellular signalregulated kinase / mitogen-ativated protein kinases. J. Immunol. 7, 9.. Rhee, J. W., Lee, K. W., Kim, D., Lee, Y., Jeon, O. H., Kwon, H. J., Kim, D. S. (7) NF- -dependent regulation of matrix metalloproteinase-9 gene expression by lipopolysaharide in a marophage ell line RAW.7. J. iohem. Mol. iol., Pugin, J., Widmer, M.., Kossodo, S., Liang,. M., Preas II, H. L., Suffredini, A. F. (999) Human neutrophils serete gelatinase in vitro and in vivo in response to endotoxin and proinflammatory mediators. Am. J. Respir. ell Mol. iol., 58.. Manione, A. M., MGuire, J. K. (8) Matrix metalloproteinases as modulators of inflammation. Semin. ell. Dev. iol. 9,.. iswas, S. K., Lopez-ollazo, E. (9) Endotoxin tolerane: new mehanisms, moleules and linial signifiane. Trends Immunol., Freudenberg, M. A., Galanos,. (988) Indution of tolerane to lipopolysaharide ()-D-galatosamine lethality by pretreatment with is mediated by marophages. Infet. Immun. 5, availlon, J. M., Adib-onquy, M. () enh-to-bedside review: endotoxin tolerane as a model of leukoyte reprogramming in sepsis. rit. are,. 7. Granowitz, E. V., Porat, R., Mier, J. W., Orenole, S. F., Kaplanski, G., Lynh, E. A., Ye, K., Vannier, E., Wolff, S. M., Dinarello,. A. (99) Intravenous endotoxin suppresses the ytokine response of peripheral blood mononulear ells of healthy humans. J. Immunol. 5, Esoll, P., del Fresno,., Garia, L., Valles, G., Lendinez, M. J., Arnalih, F., Lopez-ollazo, E. () Rapid up-regulation of IRAK-M expression following a seond endotoxin hallenge in human monoytes and in monoytes isolated from septi patients. iohem. iophys. Res. ommun., Del Fresno,., Soler-Rangel, L., Soares-Shanoski, A., Gomez-Pina, V., Gonzalez-Leon, M.., Gomez-Garia, L., Mendoza-arbera, E., Rodriguez-Rojas, A., Garia, F., Fuentes-Prior, P., Arnalih, F., Lopez-ollazo, E. (7) Inflammatory responses assoiated with aute oronary syndrome up-regulate IRAK-M and indue endotoxin tolerane in irulating monoytes. J. Endotoxin. Res., Del Fresno,., Gomez-Pina, V., Lores, V., Soares-Shanoski, A., Fernandez-Ruiz, I., Rojo,., Alvarez-Sala, R., aballero-garrido, E., Garia, F., Veliz, T., Arnalih, F., Fuentes-Prior, P., Garia-Rio, F., Lopez-ollazo, E. (8) Monoytes from ysti fibrosis patients are loked in an tolerane state: down-regulation of TREM- as putative underlying mehanism. PLoS ONE, e7.. availlon, J. M., Adrie,., Fitting,., Adib-onquy, M. () Endotoxin tolerane: is there a linial relevane? J. Endotoxin. Res. 9, 7.. Kawasaki, T., Ogata, M., Kawasaki,., Tomihisa, T., Okamoto, K., Shigematsu, A. () Surgial stress indues endotoxin hyporesponsiveness and an early derease of monoyte md and HLA-DR expression during surgery. Anesth. Analg. 9,.. Pahot, A., azalis, M. A., Venet, F., Turrel, F., Faudot,., Voirin, N., Diasparra, J., ourgoin, N., Poitevin, F., Mougin,., Lepape, A., Monneret, G. (8) Dereased expression of the fratalkine reeptor XR on irulating monoytes as new feature of sepsis-indued immunosuppression. J. Immunol. 8, 9.. ontin,., Pitard, V., Itai, T., Nagata, S., Moreau, J. F., Dehanet-Merville, J. () Membrane-anhored D is proessed by the tumor nerosis fator- -onverting enzyme. Impliations for D signaling. J. iol. hem. 78, Gough, P. J., Garton, K. J., Wille, P. T., Ryhlewski, M., Dempsey, P. J., Raines, E. W. () A disintegrin and metalloproteinase -mediated leavage and shedding regulates the ell surfae expression of X hemokine ligand. J. Immunol. 7, Del Fresno,., Garia-Rio, F., Gomez-Pina, V., Soares-Shanoski, A., Fernandez-Ruiz, I., Jurado, T., Kajiji, T., Shu,., Marin, E., Gutierrez del Arroyo, A., Prados,., Arnalih, F., Fuentes-Prior, P., iswas, S.K., Lopez-ollazo, E. (9) Potent phagoyti ativity with impaired antigen presentation identifying lipopolysaharide-tolerant human monoytes: demonstration in isolated monoytes from ysti fibrosis patients. J. Immunol. 8, Lopez-ollazo, E., Fuentes-Prior, P., Arnalih, F., del Fresno,. () Pathophysiology of interleukin- reeptor-assoiated kinase-m: impliations in refratory state. urr. Opin. Infet. Dis. 9, Meijer, J., Ogink, J., Kreike,., Nuyten, D., de Visser, K. E., Roos, E. (8) The hemokine reeptor XR and its ligand XL are expressed in arinomas and inhibit proliferation. aner Res. 8, ohrer, H., Qiu, F., Zimmermann, T., Zhang, Y., Jllmer, T., Mannel, D., ottiger,. W., Stern, D. M., Waldherr, R., Saeger, H. D., Ziegler, R., ierhaus, A., Martin, E., Nawroth, P. P. (997) Role of NF inthe mortality of sepsis. J. lin. Invest., Van t Veer,., van den Pangaart, P. S., van Zoelen, M. A., de Kruif, M., irjmohun, R. S., Stroes, E. S., de Vos, A. F., van der Poll, T. (7) Indution of IRAK-M is assoiated with lipopolysaharide tolerane in a human endotoxemia model. J. Immunol. 79, Zhong,., Ma, H. Y., Yang, Q., Gu, F. R., Yin, G. Q., Xia,. M. (8) Derease in Toll-like reeptors and in the spleen of mouse with endotoxi tolerane. Inflamm. Res. 57, ouhon, A., Dietrih, J., olonna, M. () utting edge: inflammatory responses an be triggered by TREM-, a novel reeptor expressed on neutrophils and monoytes. J. Immunol., Klesney-Tait, J., Turnbull, I. R., olonna, M. () The TREM reeptor family and signal integration. Nat. Immunol. 7, 7.. Murakami, Y., Kohsaka, H., Kitasato, H., Akahoshi, T. (7) Lipopolysaharide-indued up-regulation of triggering reeptor expressed on myeloid ells- expression on marophages is regulated by endogenous prostaglandin E. J. Immunol. 78, Dower, K., Ellis, D. K., Saraf, K., Jelinsky, S. A., Lin, L. L. (8) Innate immune responses to TREM- ativation: overlap, divergene, and positive and negative ross-talk with baterial lipopolysaharide. J. Immunol. 8, olonna, M. () TREMs in the immune system and beyond. Nat. Rev. Immunol., Muller,., Genay, M. M., Gibot, S., Stolz, D., Hunziker, L., Tamm, M., hrist-rain, M. (7) irulating levels of soluble triggering reeptor expressed on myeloid ells (strem)- in ommunity-aquired pneumonia. rit. are Med. 5, Gibot, S., Kolopp-Sarda, M. N., ene, M.., ollaert, P. E., Lozniewski, A., Mory, F., Levy,., Faure, G.. () A soluble form of the triggering reeptor expressed on myeloid ells- modulates the inflammatory response in murine sepsis. J. Exp. Med., Determann, R. M., Millo, J. L., Gibot, S., Korevaar, J.., Vroom, M.., van der Poll, T., Garrard,. S., Shultz, M. J. (5) Serial hanges in soluble triggering reeptor expressed on myeloid ells in the lung during development of ventilator-assoiated pneumonia. Intensive are Med., Gibot, S., ravoisy, A. () Soluble form of the triggering reeptor expressed on myeloid ells- as a marker of mirobial infetion. lin. Med. Res., Phua, J., Koay, E. S., Zhang, D., Tai, L. K., oo, X. L., Lim, K.., Lim, T. K. () Soluble triggering reeptor expressed on myeloid ells- in aute respiratory infetions. Eur. Respir. J. 8, Gibot, S. () The therapeuti potential of TREM- modulation in the treatment of sepsis and beyond. urr. Opin. Investig. Drugs 7, 8.. ouhon, A., Fahetti, F., Weigand, M. A., olonna, M. () TREM- amplifies inflammation and is a ruial mediator of septi shok. Nature, 7.. Gomez-Pina, V., Soares-Shanoski, A., Rodriguez-Rojas, A., Del Fresno,., Garia, F., Vallejo-remades, M. T., Fernandez-Ruiz, I., Arnalih, F., Fuentes-Prior, P., Lopez-ollazo, E. (7) Metalloproteinases shed TREM- etodomain from lipopolysaharide-stimulated human monoytes. J. Immunol. 79, Gibot, S., uonsanti,., Massin, F., Romano, M., Kolopp-Sarda, M. N., enigni, F., Faure, G.., ene, M.., Panina-ordignon, P., Passini, N., Levy,. () Modulation of the triggering reeptor expressed on the myeloid ell type pathway in murine septi shok. Infet. Immun. 7, Gupta, G. P., Nguyen, D. X., hiang, A.., os, P. D., Kim, J. Y., Nadal,., Gomis, R. R., Manova-Todorova, K., Massague, J. (7) Mediators of vasular remodelling o-opted for sequential steps in lung metastasis. Nature, Fan, H., ook, J. A. () Moleular mehanisms of endotoxin tolerane. J. Endotoxin. Res., Kobayashi, K., Hernandez, L.D., Galan, J.E., Janeway.A., Jr., Medzhitov, R., Flavell, R.A. () IRAK-M is a negative regulator of Toll-like reeptor signaling. ell, Shimazu, R., Akashi, S., Ogata, H., Nagai, Y., Fukudome, K., Miyake, K., Kimoto, M. (999) MD-, a moleule that onfers lipopolysaharide responsiveness on Toll-like reeptor. J. Exp. Med. 89, Turnbull, I. R., olonna, M. (7) Ativating and inhibitory funtions of DAP. Nat. Rev. Immunol. 7, Lee, Y. G., hain,. M., ho, J. Y. (9) Distint role of spleen tyrosine kinase in the early phosphorylation of inhibitor of via ati- 9 Journal of Leukoyte iology Volume 9, June