Scavenger Receptor 1 As a Candidate Gene Response

A Quantitative Trait Loci Analysis to Map Genes Involved in Lipopolysaccharide-Induced Inflammatory Response: Identification of Macrophage This information is current as Scavenger Receptor 1 as a Candidate Gene of September 27, 2021. William B. Fulton, Roger H. Reeves, Motohiro Takeya and Antonio De Maio J Immunol 2006; 176:3767-3773; ; doi: 10.4049/jimmunol.176.6.3767 Downloaded from http://www.jimmunol.org/content/176/6/3767

References This article cites 37 articles, 12 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/176/6/3767.full#ref-list-1

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 September 27, 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 © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

A Quantitative Trait Loci Analysis to Map Genes Involved in Lipopolysaccharide-Induced Inﬂammatory Response: Identiﬁcation of Macrophage Scavenger Receptor 1 as a Candidate Gene1

William B. Fulton,* Roger H. Reeves,† Motohiro Takeya,‡ and Antonio De Maio2*†

Septic shock, which is a major complication observed after trauma and other human diseases, is likely the product of a prolonged and poorly controlled systemic inflammatory response. Symptoms of sepsis can be partially reproduced by injection of bacterial LPS in mice. Differences in mortality between C57BL/6Jhigh and A/Jlow mice after LPS injection have been previously observed and correlated with differences in the inflammatory response between these two inbred strains. In the present study, we have mapped four loci responsible for differences in levels of LPS-induced IL-10, named modifier of IL-10, between the two strains. A Downloaded from locus within mouse chromosome 8 was confirmed using chromosome 8 consomic mice. This locus was further reduced in size by haplotype analysis and evaluated by the presence of potential candidate genes. The macrophage scavenger receptor 1 (Msr1) within this locus emerged as a candidate gene based on differences at the expression and structural levels between C57BL/6J and A/J mice. In comparison with wild-type (C57BL/6J) mice, Msr1 knockout mice displayed reduced levels of LPS-induced IL-10, but not of TNF-␣ or IL-6, confirming a specific role for this gene in the regulation of IL-10. These results suggest that Msr1 is involved in the regulation of the anti-inflammatory process, thus offering a new perspective on the molecular mechanisms involved in http://www.jimmunol.org/ endotoxemia and sepsis. The Journal of Immunology, 2006, 176: 3767–3773.

epsis is a major health problem in the United States, with inflammatory reactants is likely to be responsible for the develop- an occurrence of ϳ750,000 cases per year (1). This con- ment of exaggerated inflammatory response conditions, such as S dition is triggered by an array of insults, including infec- those observed in sepsis. tion, trauma, burn injury, pancreatitis, and even elective surgery. Sepsis is likely modified by multiple factors, including the type The mortality rate for sepsis ranges from 35 to 50%, primarily due and extent of the initiating insult, the environment, and the genetic to the development of multiple organ dysfunction syndrome (1, 2). make-up, sex, and age of the patient (5). Although this assumption Despite tremendous advances in the care of critically ill patients, has been sustained by many clinical observations, it has not been by guest on September 27, 2021 therapy for sepsis remains supportive and extremely costly. A ma- experimentally tested in humans. Studies in experimental animal jor obstacle to the development of treatments for sepsis is the im- models have offered strong support for this hypothesis. For in- mense heterogeneity of the clinical profiles observed in septic pa- stance, a robust inflammatory response can be induced by injection tients. Moreover, there are no reliable predictors for this condition. of LPS. Indeed, mice maintained in identical environmental con- Although the etiology of sepsis is poorly understood, this condition ditions and injected with Escherichia coli LPS displayed an in- has been associated with a prolonged and inadequately controlled flammatory response that differed among the various inbred strains systemic inflammatory response. An important component of the (6, 7). The frequency of mortality was also different among inbred inflammatory process is secretion of cytokines by macrophages, mouse strains following injection of LPS (6). The genetic contri- monocytes, and polymorphic nuclear leukocytes. Proinflammatory bution to sepsis has been further supported by studies using a cytokines, such as TNF-␣ and IL-6, modulate gene expression in model of polymicrobial sepsis (8, 9). A direct correlation between a variety of organs (e.g., acute phase genes in the liver). Anti- extremely elevated anti-inflammatory (IL-10) cytokine levels and inflammatory cytokines, such as IL-10, are necessary to control the increased mortality has been observed after peritonitis (8, 9). Such inflammatory process. Specifically, IL-10 is involved in the down- studies, conducted on animals housed in identical environments regulation of TNF-␣ (3, 4). The imbalance between pro- and anti- and subjected to the same initiating insult, clearly illustrate that the genetic background is an important modifier of the inflammatory response and outcome. The question that emerges is whether these *Division of Pediatric Surgery, Department of Surgery, and †Department of Physi- genetic differences can be exploited to identify genes that modulate ology, Johns Hopkins University School of Medicine, Baltimore, MD, 21205; and the inflammatory response. The identification of such genes would ‡Second Department of Pathology, Kumamoto University School of Medicine, Ku- mamoto, Japan shed light on the possible pathways activated in response to sepsis. Received for publication October 5, 2005. Accepted for publication December In addition, this knowledge could result in the detection of genetic 23, 2005. markers to be used as predictors for the development of sepsis. 3 The costs of publication of this article were defrayed in part by the payment of page Mapping quantitative trait loci (QTL) is a direct approach to the charges. This article must therefore be hereby marked advertisement in accordance identification of genes that modify the inflammatory response (5). with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This study was supported by grants from the National Institutes of Health (GM50878 and GM62599). 3 Abbreviations used in this paper: QTL, quantitative trait loci; Chr, chromosome; 2 Address correspondence and reprint requests to the current address of Dr. Antonio Msr1, macrophage scavenger receptor 1; PM␾, peritoneal macrophage; LRS, likeli- De Maio, Department of Surgery, University of California, San Diego, 9500 Gilman hood ratio statistic; LOD, logarithm of the odds; RI, recombinant inbred; Milt, mod- Drive, #0739, La Jolla, CA 92093-0739. Email address: [email protected] ifier of IL-10; PVDF, polyvinylidene difluoride.

Using an intercross strategy, we previously identified loci on Immunostaining mouse chromosomes (Chr) 13 and 5 showing an epistatic relation- PM␾ were fixed with 4% paraformaldehyde for 10 min, then washed twice ship for the infiltration of polymorphonuclear leukocytes in the with PBS. Nonspecific binding was blocked by incubation with 10% FCS liver after LPS challenge (7). In the present study, we mapped in PBS for 30 min at 25°C. Fixed cells were incubated with anti-mouse several loci for plasma levels of IL-10 as part of the inflammatory MSR1 (1/200) in PBS-5% FCS for1hat25°C. Cells were washed three times with PBS and incubated with anti-rabbit IgG Cy3 Conjugate F(abЈ) response induced by LPS. A candidate gene associated with LPS- 2 (1/1000; Sigma-Aldrich) in PBS-5% FCS for 40 min at 25°C. Cells were induced IL-10 plasma levels was identified using a combination of washed three times with PBS, then mounted and visualized with a fluo- genetic mapping and informatics. rescent microscope. FACS analysis Materials and Methods PM␾ were isolated as described above, washed twice with PBS, and re-

Experimental animal model of endotoxemia suspended in PBS containing 5% FCS and 1 mm NaN3. Cells were incubated at 25°C with anti-mouse MSR1 (1 ␮g/1 ϫ 106 cells) for1hand Male mice (A/J, AKR/J, BALB/cJ, B6, DBA/2J, B6AF1/J (BXA), AXB/ washed twice with PBS. Cells were resuspended in PBS containing 5% BXA recombinant inbred, B6-Chr 8A/NaJ, and B6-Chr 9A/NaJ) were ob- FCS and 1 mm NaN3 and incubated with anti-rabbit IgG FITC Conjugate tained from The Jackson Laboratory. AB6F1/J (AXB) mice were bred in Ј F(ab )2 (1/1000; Sigma-Aldrich) for 40 min at 25°C. Cells were then our animal facility, and macrophage scavenger receptor (Msr1) knockout washed twice with PBS, resuspended in PBS-EDTA, and analyzed by flow mice were donated by Dr. M. Freeman (Massachusetts General Hospital, cytometry. Boston, MA). Mice were maintained under identical environmental conditions in a pathogen-free animal facility. All mice used in this study were Data analysis Helicobacter negative. All procedures were performed in accordance with Downloaded from the Guidelines for the Care and Use of Laboratory Animals of the National Statistical significance was determined by one-way ANOVA with Dunn’s Institutes of Health and were approved by the Institutional Animal Care correction or ANOVA on Ranks with Dunn’s correction or Student-New- and Use Committee of Johns Hopkins University School of Medicine. At man-Keuls method and Mann-Whitney U rank sum test or the Student t value of p Ͻ 0.05 was considered significant. Data are expressed as ,ء ;wk of age, animals were fasted for 16 h before injection of E. coli LPS test 8 serotype 026:B6 (15 mg/kg; Difco Laboratories) under aseptic conditions. mean Ϯ SEM. After injection, animals were given food and water ad libitum. At 1.5 h postinjection, mice were sacrificed under isoflurane anesthesia; blood was Results http://www.jimmunol.org/ drawn by cardiac puncture using a 22-gauge needle, placed into K2EDTA Several QTLs for LPS-induced IL-10 plasma levels were Microtainer tubes (BD Biosciences), and centrifuged for 5 min. Plasma (supernatant) was removed, aliquoted, and frozen at Ϫ80°C until use. identified using recombinant inbred mouse strains Plasma cytokine levels were measured by ELISA (BioSource We previously showed that the frequency of mortality of International). C57BL/6J (B6) mice was higher than that of A/J mice after injection with LPS. This difference in mortality was correlated with Linkage analysis differences in the inflammatory response between the two strains Linkage analysis was performed using MapManager QTXb20 (10). IL-10 (6). Because cytokines are excellent markers of the inflammatory plasma levels were averaged from five animals, log-transformed, and en- process, we evaluated the plasma levels of IL-10 in A/J and B6 tered as QTL. Mean values of IL-10 were weighted by SE (log-trans- mice 1.5 h after LPS injection. This time point corresponds to the by guest on September 27, 2021 formed) and compared against the AXB/BXA set for QTL mapping (11) maximal detectable levels of this cytokine after administration of that was corrected based on updated genomic mapping. Suggestive, sig- LPS. Significantly higher levels of LPS-induced IL-10 plasma lev- nificant, and highly significant threshold levels were determined by the permutation test function of MapManager, which is based on the statistical els were observed in B6 in comparison with A/J mice (Fig. 1). This methods developed by Churchill and Doerge (12). A likelihood ratio sta- analysis was performed several times over a period of a year to tistic (LRS) value of 4.6 is equivalent to one logarithm of the odds (LOD). A two-LOD support interval was defined for loci as the region at which the LRS value was within 9.2 U (2 LOD U) of the peak value.

Isolation of mouse peritoneal macrophages (PM␾) Mice were euthanized by cervical dislocation under isoflurane anesthesia, injected with 5 ml of cold RPMI 1640 medium supplemented with 1 U/ml heparin. The medium was gently flushed five times within the peritoneum and withdrawn. Cells were centrifuged and resuspended at a concentration of 1 ϫ 106 cells/ml in RPMI 1640 medium (Mediatech Cellgro), supplemented with 10% heat-inactivated FCS (Sigma-Aldrich), 50 U/ml penicil- lin, 50 ␮g/ml streptomycin (Invitrogen Life Technologies), plated in 10- 2 cm tissue culture dishes, and maintained at 37°C in 5% CO2, 95% incubator air for 1 h. Nonadherent cells were washed off, and PM␾ were lysed with cold RIPA buffer (20 mM HEPES (pH 7.4), 1% Nonidet P-40, 50 mM NaCl, 1 mM EGTA, 5 mM ␤-glycerophosphate, 30 mM sodium pyrophosphate, 100 mM sodium orthovanadate, 0.1 mM PMSF, 10 ␮g/ml leupeptin, and 10 ␮g/ml pepstatin A). The lysate was centrifuged at 2000 ϫ g for 2 min and the supernatant was used for SDS-PAGE and Western blotting. Proteins immobilized on a polyvinylidene difluoride (PVDF) membrane were quenched for 1 h with TBS-TBST (20 mM Tris, 500 mM NaCl, 0.1% Tween 20) containing 5% powdered milk, and incu- FIGURE 1. IL-10 plasma levels following LPS injection in A/J, B6, and bated with anti-mouse MSR1 (1/5,000) for 16 h at 4°C. This Ab was raised ϭ ϭ ϭ F1 generation. Male A/J (n 45), B6 (n 47), AXB (n 15), and BXA in rabbit against a synthetic peptide (KEEQAHVEQEVKQEVR) corre- (n ϭ 15) mice were fasted for 16 h and injected i.p. with LPS (15 mg/kg). sponding to the ␣ helical coiled-coil domain of mouse MSR1 common to Blood was drawn via cardiac puncture 1.5 h after the injection, and IL-10 various mouse strains (13). Membranes were washed three times with TBST and incubated with HRP-conjugated anti-rabbit-IgG (1/25,000; Am- levels were measured in plasma by ELISA. IL-10 plasma levels in B6, p Ͻ ,ء) ersham Pharmacia Biotech) in blocking buffer for 1 h. Membranes were AXB, and BXA mice were significantly higher than in A/J mice washed again with TBST and the signal was visualized using Super Signal 0.05 by ANOVA on Ranks, Dunn’s correction). Each group represents West Dura (Pierce Biotechnology). three or more experiments to account for seasonal variation. The Journal of Immunology 3769

FIGURE 2. QTL analysis for differences in IL-10 plasma levels in RI mouse strains. Eight- week-old, male mice from 26 AXB/BXA RI strains (n ϭ 5 per strain), and A/J (n ϭ 8) and B6 (n ϭ 9) mice, were fasted for 16 h and injected i.p. with LPS (15 mg/kg). Blood was drawn via cardiac puncture 1.5 h after the injection, and IL-10 levels were measured in plasma by ELISA. On average, five animals per strain were used for phenotyping and the experiments were performed over a 6-mo period. Downloaded from overcome any potential seasonal variability in LPS-induced cyto- The number of plausible candidate genes contained within the 95% kine levels. The difference in response between B6 and A/J mice confidence interval around this locus (40 cM) is very large, min- was observed regardless of any seasonal variability. Analysis of imizing the chance for successful systematic analysis. Conse- the first generation (F1) between B6 and A/J mice (AXB/BXA) quently, we attempted to reduce the number of candidate genes by ͗ showed that LPS-induced IL-10 levels were similar to those of B6 using a haplotype comparison among inbred mouse strains ( http:// http://www.jimmunol.org/ mice, suggesting that this trait is neither sex-linked nor imprinted snp.gnf.org͘) (15). We analyzed the IL-10 response in inbred and is likely B6 dominant. Recombinant inbred (RI) mouse strains strains (AKR/J, BALBc/J, DBA/2J, 129S1/SvImJ(129/SJ)) that were used to map the loci contributing to these phenotypes. Male shared haplotype information on Chr 8 with A/J but differed from mice from the AXB/BXA set (26 RI strains) were injected with B6 mice (15). LPS-induced IL-10 plasma levels were similar LPS and levels of IL-10 were measured (Fig. 2). The log-trans- among AKR/J, BALBc/J, DBA/2J, 129/SJ with respect to A/J formed data, weighted for variance, was analyzed by MapManager mice and significantly lower than B6 mice (Fig. 4A). Based on the QTXb20. Several contributing loci, named modifier of IL-10 haplotype comparison among “A/J-like” strains, we were able to (Milt), were obtained for the IL-10 plasma level phenotype. Two reduce the size of Milt3 from 40 cM to ϳ24 cM, decreasing the loci, one on Chr 8 (Milt3) and the other on Chr 13 (Milt4), dis- number of potential candidate genes from ϳ160 to 68 (Fig. 4B). by guest on September 27, 2021 played significant linkages, whereas loci on Chr 2 (Milt1) and 4 (Milt2) exhibited suggestive linkages (Table I). Msr1 as a candidate gene We focused on the two loci that displayed significant QTL Evaluation of possible candidate genes within the reduced 24 cM (Milt3 and Milt4). Consomic mice were used to confirm these loci region on Milt3 indicated only five genes with known polymor- for Chr 8 (B6-Chr8A/NaJ) and Chr 13 (B6-Chr13A/NaJ), respec- phisms between A/J and B6: Msr1, Rasl5–2, Nat1, Lpl, and Ucp1. tively. These consomic mice contained the chromosome of interest Msr1 (SR-A) was chosen as the primary candidate gene for several from A/J and all remaining chromosomes from B6 mice (14). Ob- reasons. Msr1 encodes for a cell surface glycoprotein that is spe- served LPS-induced IL-10 plasma levels in B6-Chr8A/NaJ mice cifically expressed in macrophages (16). These cells are central to were very similar to those of A/J mice (Fig. 3), whereas B6- the activation and regulation of the inflammatory response. MSR1 Chr13A/NaJ behaved as B6 mice (data not shown). An unrelated is apparently involved in the clearance of numerous substances, consomic mouse strain, B6-Chr9A/NaJ, also displayed a response including LPS, pathogens, and apoptotic cells (17). Polymor- similar to B6 mice. These observations confirm the influence of a phisms within the coding region of Msr1 have already been iden- gene within Milt3 on levels of LPS-induced IL-10. In contrast, tified between B6 and A/J mice, although no functional differences Milt4 may have an epistatic relationship with other genes. were previously attributed to these allelic variants (18). To inves- tigate the possible role of MSR1 in the differential response to Refinement of Milt3 position by haplotype/single nucleotide LPS, Msr1Ϫ/Ϫ mice bred onto a B6 background for at least six polymorphism analysis generations (provided by Dr. M. Freeman; Ref. 19) were injected Based on the consomic mouse results, Milt3 was further refined with LPS, and IL-10 plasma levels were measured 1.5 h postin- because this locus can be analyzed independently of other alleles. jection. LPS-induced IL-10 plasma levels were reduced by ϳ50%

Table I. QTLs that modify the LPS-induced IL-10 response

Relative Size of Support Locus Chr Support Interval (2-LOD)a Interval (cM) Signiﬁcance/LRS

Milt1 2 D2mit31-Iapls2–3 3 Suggestive 16.6 Milt2 4 D4mit31-D4mit251 14.7 Suggestive 14.9 Milt3 8 Telomere-D8mit45 37 Signiﬁcant/20.8 Milt4 13 D13mit122-D13Mit11 4 Signiﬁcant/25.7

a Interval mapping for modiﬁers of IL-10 trait (Milt) QTL were obtained from Map Manager QTX analysis of the different RI strains. One LOD is equivalent to a 4.6 LRS, and 2-LOD support interval is within 9.2 U of the peak LRS value. 3770 ROLE OF Msr1 IN INFLAMMATION

FIGURE 3. IL-10 plasma levels in consomic mice. Eight-week-old, B6- Chr8A/NaJ (n ϭ 18), and B6-Chr9A/NaJ (n ϭ 4), A/J (n ϭ 35), and B6 (n ϭ 37) male mice were fasted for 16 h and injected i.p. with LPS (15 Downloaded from mg/kg). Blood was drawn via cardiac puncture 1.5 h after the injection, and IL-10 levels were measured in plasma by ELISA. IL-10 plasma levels in B6 and B6-Chr9A/NaJ mice were signiﬁcantly higher than A/J and B6- .(p Ͻ 0.05 by ANOVA on Ranks, Dunn’s correction ,ء) Chr8A/NaJ

FIGURE 4. Refining the Milt3 locus by haplotype analysis. A, Eight- http://www.jimmunol.org/ in the knockout mice compared with B6 mice (wild type). More- week-old male mice from the inbred strains AKR/J (n ϭ 8), BALB/cJ (n ϭ over, LPS-induced IL-10 levels in the null mice were similar to 8), DBA/2J (n ϭ 7), and 129S1/SvImJ (n ϭ 6), which share haplotypes in those observed in A/J mice (Fig. 5A). In contrast, LPS-induced a portion of the Milt3 region with A/J, A/J (n ϭ 8), and B6 (n ϭ 8) were TNF-␣ and IL-6 plasma levels were similar between Msr1Ϫ/Ϫ and fasted for 16 h and injected i.p. with LPS (15 mg/kg). Blood was drawn via B6 mice (Fig. 5, B and C). These results specifically segregate the cardiac puncture 1.5 h after the injection, and cytokine levels were measured in plasma by ELISA. IL-10 plasma levels in B6 mice were signifi- effect of MSR1 on LPS-induced IL-10 from that on other ,p Ͻ 0.05 by ANOVA on Ranks ,ء) cantly higher than the other strains cytokines. Dunn’s correction). B, Markers were chosen at sites where B6 differed from all A/J-like haplotypes. If a region contained at least one A/J haplo-

The expression of Msr1 is different between B6 and A/J mice by guest on September 27, 2021 type similar to B6, that area was discarded. The black bar denotes areas that We compared the presence of MSR1 in isolated naive PM␾ from differed between B6 and all the A/J haplotypes. This combined region A/J and B6 mice. Western blot analyses of protein lysates from spans 24 cM. PM␾ isolated from these mouse strains were performed using Abs specific for mouse MSR1 that did not differentiate between genetic variants (13). Higher levels of MSR1 were consistently observed (Fig. 7B). These observations provide another potential explana- in B6 with respect to A/J-isolated PM␾ (1.6-fold, p Ͻ 0.005; Fig. tion for a difference in functional ability of the allelic variants. 6, A and B). In addition to differences in expression level, we observed a difference in the electrophoretic mobility of MSR1 be- Discussion tween B6 and A/J mice in which the glycoprotein derived from A/J We used QTL mapping in RI mouse strains as a direct approach to mice migrated faster (Fig. 6A). This observation raises the possi- the identification of modifier genes of the inflammatory process, in bility of differential glycosylation or some other posttranslational particular, LPS-induced IL-10 plasma levels. IL-10 is an important modification between the two strains. Polymorphisms within the component of the regulation of the inflammatory response. Four coding region of Msr1 (18) result in a different number of hypo- loci were identified by this approach, with only two, Milt3 (Chr 8) thetical N-linked glycosylation sites between B6 and A/J MSR1, in and Milt4 (Chr 13), displaying statistically significant levels. Milt3 particular, a substitution of 198H for N. Protein extracts derived was confirmed to be involved in the regulation of LPS-induced from B6 and A/J PM␾ were treated with endoglycosydase F, IL-10 levels by using recently developed consomic mice (14). In which cleaves N-linked carbohydrate moieties of glycoproteins contrast, Milt4 could not be confirmed using the respective con- (Fig. 6C). This treatment resulted in a decrease in the apparent somic mice, probably because it is involved in an epistatic rela- m.w. of MSR1, as well as sharper appearance of the band, con- tionship with other genes. Previous studies have mapped QTL to sistent with the presence of oligosaccharides in this molecule. In the same region on mouse Chr 8 for susceptibility to arthritis (20), addition, the difference in electrophoretic mobility between MSR1 colitis (21, 22), and experimental allergic encephalomyelitis (23). derived from B6 and A/J mice disappeared, while the difference in It is likely that these diseases share an inflammatory component the amount of MSR1 was maintained between the two strains. with endotoxemia. The boundaries of Milt3 were delimited based Finally, we investigated the presence of MSR1 by immunostain- on SNP haplotypes (͗http://snp.gnf.org͘; Ref. 15) and only five ing. The cell surface localization was the same in PM␾ isolated genes within this region were found to display polymorphisms from A/J and B6 mice, but B6 displayed apparently higher levels (͗www.informatics.jax.org͘) between A/J and B6 mice. Of these, of MSR1 than A/J-derived PM␾ (Fig. 7A). This observation was Msr1 was selected as a candidate gene because it is specifically further corroborated by FACS analysis of isolated A/J or B6 de- expressed in cells involved in the inflammatory process (i.e., mac- rived PM␾ using the Ab specific for MSR1. We found 78% of B6 rophages). In addition, previous studies have suggested that Msr1 cells positive for MSR1 in contrast to 31% for A/J-derived PM␾ plays a role in the inflammatory response (24–26). The Journal of Immunology 3771 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 6. Msr1 protein levels in PM␾ from A/J and B6 mice. PM␾ from A/J and B6 mice were obtained by peritoneal lavage. Macrophages were allowed to attach for 1 h, washed with PBS, and then lysed in RIPA buffer containing PMSF and complete antiprotease mixture. A, Protein was separated on a 10% polyacrylamide gel, transferred onto PVDF membranes, and incubated with an Ab against mouse MSR1 that recognized B6 FIGURE 5. Cytokine plasma levels following LPS injection in A/J, B6, and A/J variants. B6 MSR1 migrated slower than A/J MSR1. The same and Msr1Ϫ/Ϫ mice. Male mice (8-wk old) were fasted for 16 h and injected blot was stripped and reprobed using Abs against ␤-actin. The blot is rep- i.p. with LPS (15 mg/kg). Blood was drawn via cardiac puncture 1.5 h resentative of at least three experiments. B, MSR1 signal was quantitated (TNF-␣ and IL-10) or 6 h (IL-6) after the injection, and cytokine levels and normalized to the corresponding signal for ␤-actin. B6 protein levels were measured in plasma by ELISA. A, IL-10 plasma levels in B6 mice p ϭ 0.003 by Mann-Whitney ,ء) were significantly higher than A/J levels p Ͻ ,ء) were significantly higher as compared with A/J and Msr1Ϫ/Ϫ mice rank sum test, n ϭ 10 per strain). C, Proteins (40 ␮g) were treated with 0.05 by ANOVA on Ranks, Dunn’s correction, n ϭ 10 per strain). For PNGaseF (1000 U) in the presence of 10% Nonidet P-40 and G7 buffer for TNF-␣ (B) and IL-6 (C) plasma levels, A/J mice were significantly differ- 75 min at 37°C and separated on a 7.5% polyacrylamide gel, transferred ent compared with B6 and Msr1Ϫ/Ϫ mice (#, p Ͻ 0.05 by one-way onto PVDF membranes, and incubated with an Ab against mouse MSR1. ANOVA by Student-Newman-Keuls Method for TNF-␣, n ϭ 5 per strain; Notice that the difference in electrophoretic migration between A/J and B6 and ‡, p Ͻ 0.001 for IL-6, n ϭ 8 per strain). MSR1 disappeared whereas the difference in intensity was preserved.

The involvement of Msr1 in LPS-induced IL-10 plasma levels was sustained by using homozygote knockout mice, which display congenic region is less likely because 129 mice showed a different LPS-induced IL-10 levels similar to A/J and different from B6 LPS-induced cytokine proﬁle (TNF-␣ and IL6) than B6, A/J, and (wild-type) mice. These knockout mice were generated using 129- Msr1Ϫ/Ϫ mice. derived ES cells (A3-1), and were backcrossed to B6 for at least MSR1 is an integral plasma membrane glycoprotein composed six generations, providing a 99.2% B6 genomic background. Al- of three identical chains forming a collagenase-like structure. This though we cannot disregard the possibility that other genes within glycoprotein is involved in the clearance of polyanionic macro- the congenic area are involved in the regulation of LPS-induced molecules, including acetylated or oxidized low density lipopro- IL-10 levels, the potential contribution of genes within this 129 tein, lipoteichoic acid, lipopolysaccharides, polynucleotides, and 3772 ROLE OF Msr1 IN INFLAMMATION

LPS. This observation is consistent with the decrease in IL-10 levels observed in Msr1-deﬁcient mice with respect to B6 wild- type mice. In contrast, Haworth et al. (31) showed that Msr1 knockout mice were more resistant to LPS after they were primed with bacillus Calmette-Gue´rin. The discrepancy between the studies could stem from the fact that the insults used were not identical, and that the genetic backgrounds of the null mice used in the two studies differed. In addition, these knockout mice have mixed genetic backgrounds of 129 and ICR due to a limited backcrossing strategy. In contrast, we used knockout mice that had been backcrossed to B6 mice for six generations, resulting in a more ho- mogenous genetic background (19). Moreover, we did not ﬁnd a correlation between TNF-␣ or IL-6 and Msr1, as proposed by oth- ers (31, 32). We are, however, in agreement with the results of studies that used the same knockout mice (33), indicating that MSR1 does not play a direct role in the TNF-␣-signaling pathway. Internalization of Brucella spp. was reported to be mediated by MSR1 and blocked by the addition of LPS from the same species

or from Salmonella, but not from E. coli LPS (24). Msr1 knockout Downloaded from mice displayed decreased protection from infection with Listeria monocytogenes and herpes virus (26), which correlates with A/J (susceptible) and B6 (resistant) phenotypes for these two infec- tious agents (34). These ﬁndings, and our own observations, strongly support a role for MSR1 in host defense and innate im-

munity. MSR1 has also been implicated in the pathological dep- http://www.jimmunol.org/ osition of cholesterol during atherogenesis (35). Consistent with FIGURE 7. Comparison of MSR1 surface expression on PM␾ isolated these results, A/J mice have been found to be more resistant than from A/J and B6 mice. A,PM␾ isolated from A/J and B6 mice were B6 mice in experimental models of atherosclerosis, diabetes, and harvested by peritoneal lavage and allowed to attach to glass coverslips. obesity (34). These observations also suggest a possible link be- Cells were immunostained with an anti-mouse MSR1 Ab and visualized by tween MSR1 and IL-10 in many diseases besides sepsis. Ј ϫ Cy3 IgG F(ab )2. Figure depicts a representative ﬁeld under 100 objec- MSR1 has also been implicated in the clearance of apoptotic tive. B,PM␾ isolated from A/J and B6 mice were harvested by peritoneal cells (29), the appearance of which in immune organs (i.e., spleen lavage and analyzed by FACS. Top row, Cells stained with anti-mouse and thymus), intestinal epithelial, and liver is a hallmark of sepsis

MSR1 and as a secondary anti-rabbit IgG FITC conjugate F(abЈ) ; bottom by guest on September 27, 2021 2 (36). Therefore, it could be speculated that cells expressing differ- row, with secondary alone. Numbers in the top right quadrant represent ent Msr1 genetic variants could clear apoptotic cells with different number of cells gated that were positive for MSR1. efficiencies. The prolonged presence of apoptotic cells may result in postapoptotic necrosis that induces a secondary inflammatory sulfated polysaccharides (17). In addition, MSR1 has been impli- response, which may lead to the hyperinflammatory stage charac- cated in cell adhesion (27, 28) and the clearance of bacteria (24) teristic of septic shock. Recently, polymorphisms have been de- and apoptotic cells (29). MSR1 is almost exclusively expressed on tected within the human MSRI gene and correlated with the inci- macrophages (16), which are the primary mediators of the inflam- dence of prostate cancer. The role of MSR1 in cancer may be matory process. Polymorphisms within the Msr1 coding region related to observations indicating the infiltration of macrophages have been reported between B6 and many other mouse strains, within the tumor (37). Thus, MSR1 appears to be involved in the including A/J (18). We found that one of these polymorphisms inflammatory response associated with cancer. The involvement of resulted in a different number of N-glycosylation sites between the MSR1 in sepsis introduces a possible new pathway in the septic B6 and A/J variants, which may modify the function of this re- process and contributes to a better understanding of the molecular ceptor. For example, an extra glycosylation site could affect the mechanisms involved in this condition. Future research into the affinity of MSR1 for one or more of its several ligands or change pathway involving MSR1 may reveal innovative therapeutic ap- the stability of the protein. The number of glycosylation sites ob- proaches to ameliorate the detrimental consequences of sepsis. served in the B6 variant is conserved within other species, includ- Moreover, Msr1 may emerge as a genetic marker for the incidence ing humans (18). In addition, the content of MSR1 is higher in of endotoxemia, sepsis, and related conditions. Finally, our study B6-derived PM␾ than in A/J cells. Thus, the IL-10 hyperresponse demonstrates the feasibility of using QTL to identify new compo- observed in B6 may be related to elevated abundance of MSR1. nents (genes) that are involved in the inflammatory process, which This explanation is consistent with decreased LPS-induced IL-10 may be used as genetic markers to recognize a human population levels in the knockout mice. at risk for developing sepsis after trauma and infection. We have correlated a hyperinflammatory response, in particular IL-10 levels, with an increase in mortality of B6 mice after endotoxic (6) and septic shock (8, 9), respectively. Other investigations Acknowledgments have shown a correlation between overexpression of IL-10 and a We thank Dr. Mason Freeman for his generous provision of Msr1 detrimental outcome after infection (30). However, exogenous ad- knockout mice. ministration of IL-10 has been shown to be protective against LPS by down-regulating proinflammatory cytokines (3, 4). Kobayashi et al. (25) showed that the deletion of Msr1 resulted in protection Disclosures from endotoxic shock, which was correlated with direct binding to The authors have no financial conflict of interest. The Journal of Immunology 3773

References 21. Mahler, M., C. Most, S. Schmidtke, J. P. Sundberg, R. Li, H. J. Hedrich, and G. A. Churchill. 2002. Genetics of colitis susceptibility in IL-10-deficient mice: 1. Angus, D. C., W. T. Linde-Zwirble, J. Lidicker, G. Clermont, J. Carcillo, and backcross versus F2 results contrasted by principal component analysis. Genom- M. R. Pinsky. 2001. Epidemiology of severe sepsis in the United States: analysis ics 80: 274–282. of incidence, outcome, and associated costs of care. Crit. Care Med. 29: 22. Farmer, M. A., J. P. Sundberg, I. J. Bristol, G. A. Churchill, R. Li, C. O. Elson, 1303–1310. and E. H. Leiter. 2001. A major quantitative trait locus on chromosome 3 controls 2. Meakins, J. L. 1990. Etiology of multiple organ failure. J. Trauma 30: colitis severity in IL-10-deficient mice. Proc. Natl. Acad. Sci. USA 98: S165–S168. 13820–13825. 3. Howard, M., T. Muchamuel, S. Andrade, and S. Menon. 1993. Interleukin 10 23. Blankenhorn, E. P., R. J. Butterfield, R. Rigby, L. Cort, D. Giambrone, protects mice from lethal endotoxemia. J. Exp. Med. 177: 1205–1208. P. McDermott, K. McEntee, N. Solowski, N. D. Meeker, J. F. Zachary, et al. 4. de Waal Malefyt, R., J. Abrams, B. Bennett, C. G. Figdor, and J. E. de Vries. 2000. Genetic analysis of the influence of pertussis toxin on experimental allergic 1991. Interleukin 10 (IL-10) inhibits cytokine synthesis by human monocytes: an encephalomyelitis susceptibility: an environmental agent can override genetic autoregulatory role of IL-10 produced by monocytes. J. Exp. Med. 174: checkpoints. J. Immunol. 164: 3420–3425. 1209–1220. 24. Kim, S., M. Watarai, H. Suzuki, S. Makino, T. Kodama, and T. Shirahata. 2004. 5. De Maio, A., M. B. Torres, and R. H. Reeves. 2005. Genetic determinants in- Lipid raft microdomains mediate class A scavenger receptor-dependent infection fluencing the response to injury, inflammation, and sepsis. Shock 23: 11–17. of Brucella abortus. Microb. Pathog. 37: 11–19. 6. De Maio, A., M. L. Mooney, L. E. Matesic, C. N. Paidas, and R. H. Reeves. 1998. 25. Kobayashi, Y., C. Miyaji, H. Watanabe, H. Umezu, G. Hasegawa, T. Abo, Genetic component in the inflammatory response induced by bacterial lipopoly- M. Arakawa, N. Kamata, H. Suzuki, T. Kodama, and M. Naito. 2000. Role of saccharide. Shock 10: 319–323. macrophage scavenger receptor in endotoxin shock. J. Pathol. 192: 263–272. 7. Matesic, L. E., E. L. Niemitz, A. De Maio, and R. H. Reeves. 2000. Quantitative trait loci modulate neutrophil infiltration in the liver during LPS-induced inflam- 26. Suzuki, H., Y. Kurihara, M. Takeya, N. Kamada, M. Kataoka, K. Jishage, mation. FASEB J. 14: 2247–2254. O. Ueda, H. Sakaguchi, T. Higashi, T. Suzuki, et al. 1997. A role for macrophage Nature 8. Stewart, D., W. B. Fulton, C. Wilson, C. L. Monitto, C. N. Paidas, R. H. Reeves, scavenger receptors in atherosclerosis and susceptibility to infection. 386: and A. De Maio. 2002. Genetic contribution to the septic response in a mouse 292–296. model. Shock 18: 342–347. 27. Fraser, I., D. Hughes, and S. Gordon. 1993. Divalent cation-independent macrophage adhesion inhibited by monoclonal antibody to murine scavenger receptor. 9. Torres, M. B., and A. De Maio. 2005. An exaggerated inflammatory response Downloaded from after CLP correlates with a negative outcome. J. Surg. Res. 125: 88–93. Nature 364: 343–346. 10. Manly, K. F., and J. M. Olson. 1999. Overview of QTL mapping software and 28. Kosswig, N., S. Rice, A. Daugherty, and S. R. Post. 2003. Class A scavenger introduction to map manager QT. Mamm. Genome 10: 327–334. receptor-mediated adhesion and internalization require distinct cytoplasmic do- 11. Sampson, S. B., D. C. Higgins, R. W. Elliot, B. A. Taylor, K. K. Lueders, mains. J. Biol. Chem. 278: 34219–34225. R. A. Koza, and B. Paigen. 1998. An edited linkage map for the AXB and BXA 29. Platt, N., H. Suzuki, Y. Kurihara, T. Kodama, and S. Gordon. 1996. Role for the recombinant inbred mouse strains. Mamm. Genome 9: 688–694. class A macrophage scavenger receptor in the phagocytosis of apoptotic thymo- 12. Churchill, G. A., and R. W. Doerge. 1994. Empirical threshold values for quan- cytes in vitro. Proc. Natl. Acad. Sci. USA 93: 12456–12460. titative trait mapping. Genetics 138: 963–971. 30. Hoffmann, K. F., A. W. Cheever, and T. A. Wynn. 2000. IL-10 and the dangers 13. Tomokiyo, R., K. Jinnouchi, M. Honda, Y. Wada, N. Hanada, T. Hiraoka, of immune polarization: excessive type 1 and type 2 cytokine responses induce http://www.jimmunol.org/ H. Suzuki, T. Kodama, K. Takahashi, and M. Takeya. 2002. Production, char- distinct forms of lethal immunopathology in murine schistosomiasis. J. Immunol. acterization, and interspecies reactivities of monoclonal antibodies against human 164: 6406–6416. class A macrophage scavenger receptors. Atherosclerosis 161: 123–132. 31. Haworth, R., N. Platt, S. Keshav, D. Hughes, E. Darley, H. Suzuki, Y. Kurihara, 14. Nadeau, J. H., J. B. Singer, A. Matin, and E. S. Lander. 2000. Analysing complex T. Kodama, and S. Gordon. 1997. The macrophage scavenger receptor type A is genetic traits with chromosome substitution strains. Nat. Genet. 24: 221–225. expressed by activated macrophages and protects the host against lethal endotoxic 15. Wiltshire, T., M. T. Pletcher, S. Batalov, S. W. Barnes, L. M. Tarantino, shock. J. Exp. Med. 186: 1431–1439. M. P. Cooke, H. Wu, K. Smylie, A. Santrosyan, N. G. Copeland, et al. 2003. 32. Hsu, H. Y., S. L. Chiu, M. H. Wen, K. Y. Chen, and K. F. Hua. 2001. Ligands Genome-wide single-nucleotide polymorphism analysis defines haplotype pat- of macrophage scavenger receptor induce cytokine expression via differential terns in mouse. Proc. Natl. Acad. Sci. USA 100: 3380–3385. modulation of protein kinase signaling pathways. J. Biol. Chem. 276: 16. Hughes, D. A., I. P. Fraser, and S. Gordon. 1995. Murine macrophage scavenger 28719–28730. receptor: in vivo expression and function as receptor for macrophage adhesion in 33. Kim, W. S., C. M. Ordija, and M. W. Freeman. 2003. Activation of signaling by guest on September 27, 2021 lymphoid and non-lymphoid organs. Eur. J. Immunol. 25: 466–473. pathways by putative scavenger receptor class A (SR-A) ligands requires CD14 17. Krieger, M., and J. Herz. 1994. Structures and functions of multiligand lipopro- but not SR-A. Biochem. Biophys. Res. Commun. 310: 542–549. tein receptors: macrophage scavenger receptors and LDL receptor-related protein 34. Nesbitt, M. N., and E. Skamene. 1984. Recombinant inbred mouse strains derived (LRP). Annu. Rev. Biochem. 63: 601–637. from A/J and C57BL/6J: a tool for the study of genetic mechanisms in host 18. Fortin, A., M. Penman, M. M. Stevenson, M. Krieger, and P. Gros. 2000. Iden- resistance to infection and malignancy. J. Leukocyte Biol. 36: 357–364. tification and characterization of naturally occurring variants of the macrophage 35. Brown, M. S., and J. L. Goldstein. 1983. Lipoprotein metabolism in the macro- scavenger receptor (SR-A). Mamm. Genome 11: 779–785. phage: implications for cholesterol deposition in atherosclerosis. Annu. Rev. Bio- 19. Kunjathoor, V. V., M. Febbraio, E. A. Podrez, K. J. Moore, L. Andersson, chem. 52: 223–261. S. Koehn, J. S. Rhee, R. Silverstein, H. F. Hoff, and M. W. Freeman. 2002. 36. Hotchkiss, R. S., P. E. Swanson, J. P. Cobb, A. Jacobson, T. G. Buchman, and Scavenger receptors class A-I/II and CD36 are the principal receptors responsible I. E. Karl. 1997. Apoptosis in lymphoid and parenchymal cells during sepsis: for the uptake of modified low density lipoprotein leading to lipid loading in findings in normal and T- and B-cell-deficient mice. Crit. Care Med. 25: macrophages. J. Biol. Chem. 277: 49982–49988. 1298–1307. 20. Otto, J. M., G. Cs-Szabo, J. Gallagher, S. Velins, K. Mikecz, E. I. Buzas, 37. Xu, J., S. L. Zheng, A. Komiya, J. C. Mychaleckyj, S. D. Isaacs, J. J. Hu, J. T. Enders, Y. Li, B. R. Olsen, and T. T. Glant. 1999. Identification of multiple D. Sterling, E. M. Lange, G. A. Hawkins, A. Turner, et al. 2002. Germline loci linked to inflammation and autoantibody production by a genome scan of a mutations and sequence variants of the macrophage scavenger receptor 1 gene are murine model of rheumatoid arthritis. Arthritis Rheum. 42: 2524–2531. associated with prostate cancer risk. Nat. Genet. 32: 321–325.