of May 8, 2015.This information is current as

of Nox2 Activity in MacrophagesCutting Edge: Slamf8 Is a Negative Regulator

Detre and Cox TerhorstMaganto-Garca, Manuel Fresno, Ninghai Wang, Cynthia Rojas-Barros, Marta Aleman, Gongxian Liao, ElenaXavier Romero, Michael S. O'Keeffe, Domingo I. Guoxing Wang, Ana C. Abada-Molina, Scott B. Berger,

http://www.jimmunol.org/content/188/12/5829doi: 10.4049/jimmunol.1102620May 2012;

2012; 188:5829-5832; Prepublished online 16J Immunol

MaterialSupplementary

0.DC1.htmlhttp://www.jimmunol.org/content/suppl/2012/05/16/jimmunol.110262

Referenceshttp://www.jimmunol.org/content/188/12/5829.full#ref-list-1

, 5 of which you can access for free at: cites 11 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright 2012 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on May 8, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on May 8, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

http://www.jimmunol.org/cgi/adclick/?ad=44959&adclick=true&url=http%3A%2F%2Fwww.biolegend.com%2Frecombinant_proteinshttp://http://www.jimmunol.org/content/188/12/5829http://www.jimmunol.org/content/suppl/2012/05/16/jimmunol.1102620.DC1.htmlhttp://www.jimmunol.org/content/suppl/2012/05/16/jimmunol.1102620.DC1.htmlhttp://www.jimmunol.org/content/188/12/5829.full#ref-list-1http://jimmunol.org/site/subscriptions/http://www.aai.org/ji/copyright.htmlhttp://jimmunol.org/cgi/alerts/etochttp://www.jimmunol.org/http://www.jimmunol.org/

Cutting Edge: Slamf8 Is a Negative Regulator of Nox2Activity in MacrophagesGuoxing Wang,*,1 Ana C. Abada-Molina,,1 Scott B. Berger,*,1 Xavier Romero,*Michael S. OKeeffe,* Domingo I. Rojas-Barros, Marta Aleman, Gongxian Liao,*Elena Maganto-Garca,*, Manuel Fresno, Ninghai Wang,* Cynthia Detre,* andCox Terhorst*

Slamf8 (CD353) is a cell surface receptor that isexpressed upon activation of macrophages (MFs) byIFN-g or bacteria. In this article, we report that a veryhigh NADPH oxidase (Nox2) enzyme activity wasfound in Slamf82/2 MFs in response to Escherichiacoli or Staphylococcus aureus, as well as to PMA. Theelevated Nox2 activity in Slamf82/2 MFs was alsodemonstrated in E. coli or S. aureus phagosomes byusing a pH indicator system and was further confirmedby a reduction in the enzyme activity after transfectionof the receptor into Slamf8-deficient primary MFs orRAW 264.7 cells. Upon exposure to bacteria or PMA,protein kinase C activity in Slamf82/2 MFs is in-creased. This results in an enhanced phosphorylation ofp40phox, one key component of the Nox2 enzyme com-plex, which, in turn, leads to greater Nox2 activity. Takentogether, the data show that, in response to inflamma-tion-associated stimuli, the inducible receptor Slamf8negatively regulates inflammatory responses. TheJournal of Immunology, 2012, 188: 58295832.

Several members of the Slam family (Slamf) of cellsurface receptors modulate macrophage (MF) func-tions (13). For example, we recently found that on

the surface of MFs, Slamf1 acts as a microbial sensor andmodulates two bactericidal processes in the Escherichia coliphagosome: NADPH oxidase (Nox2) activity and phag-osomal maturation (1). Because human and mouse SLAMF8(BLAME, CD353) are expressed on the surface of myeloidcells (4, 5), we set out to determine whether this receptorwould also affect MF functions as Slamf1.We find that mouse Slamf8 is not expressed on the surface

of resident peritoneal MFs (pMFs), but it can be induced

by exposure to IFN-g and bacteria. Using MFs isolated fromSlamf8-deficient (Slamf82/2) mice or RAW 264.7 MFs,Slamf8 was found to be a negative regulator of Nox2 activityin bacterial phagosomes. In Slamf82/2 MFs, in contrast toSlamf8-sufficient cells, bacteria or PMA induce increasedprotein kinase C (PKC) activity, resulting in enhanced phos-phorylation of p40phox, which increases Nox2 activity.

Materials and MethodsMice

Slamf82/2 (Slamf8DE2-4) BALB/c mice were generated, as described inSupplemental Fig. 1A, and were generously donated by Drs. A. Coyle andJ.C. Guttierez-Ramos (Millennium Pharmaceuticals, Cambridge, MA).Slamf82/2 mice develop normally, and spontaneous disease or inflammationwas not observed. In Slamf82/2 mice, T and B cells and MFs developsimilarly as in wild-type (wt) mice (Supplemental Fig. 1). wt BALB/c micewere purchased from The Jackson Laboratory (Bar Harbor, ME); all micewere maintained in a specific pathogen-free facility. The Beth Israel Dea-coness Medical Center Institutional Animal Care and Use Committee ap-proved all animal care and experimental procedures.

Cells

F4/80+CD11b+ bone marrow-derived, resident peritoneal or thioglycollate(TGC)-elicited pMFs were obtained, as described (1). Slamf82/2 MFs andRAW 264.7 MFs were transfected with a Slamf8-mCherry vector usingAmaxa (Lonza Group, Basel, Switzerland).

Real-time PCR (TaqMan)

Cell RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA); 100ng RNA was used for TaqMan assay, 18S rRNA was used as an internalreference gene to normalize the transcripts level, and the relative gene mRNAexpression level was determined by the 22ct method.

Nox2 assays

Nox2 assays using lucigenin (Sigma-Aldrich, St. Louis, MO) were performedafter exposure to stimulated heat-killed Escherichia coli F18 or Staphylococcusaureus bacteria (multiplicity of infection [MOI] 100) or 1 mg/ml PMA usinga standard Glomax luminometer (Promega, Madison, WI) (1). PhagosomalpH was determined using pHrodo E. coli or S. aureus (MOI 15; Invitrogen),as described (1).

*Division of Immunology, Beth Israel Deaconess Medical Center, Harvard MedicalSchool, Boston, MA 02115; Departamento de Bioqumica y Biologa Molecular 3 eInmunologa, Centro de Investigacion Biomedica, Universidad de Granada, ParqueTecnologico de las Ciencias de la Salud, 18100 Armilla, Granada, Spain; and Centrode Biologa Molecular Severo Ochoa, Consejo Superior de Investigaciones Cientficas,Universidad Autonoma de Madrid, Cantoblanco, 28049 Madrid, Spain

1G.W., A.C.A.-M., and S.B.B. contributed equally to this work.

Received for publication September 14, 2011. Accepted for publication April 19, 2012.

This work was supported by grants from the National Institutes of Health (AI-15066to C.T.) and the Ministry of Education and Science of Spain (SAF2006-08529 andSAF2007-62562 to A.C.A.-M.). S.B.B. and X.R. were recipients of a fellowship from theCrohns and Colitis Foundation of America.

Address correspondence and reprint requests to Prof. Cox Terhorst, Division of Immu-nology, Beth Israel Deaconess Medical Center, Harvard Medical School, 3 BlackfanCircle, Boston, MA 02115. E-mail address: terhorst@bidmc.harvard.edu

The online version of this article contains supplemental material.

Abbreviations used in this article: MF, macrophage; MOI, multiplicity of infection;Nox2, NADPH oxidase; PKC, protein kinase C; pMF, peritoneal macrophage; Slamf,Slam family; TGC, thioglycollate; wt, wild-type.

Copyright 2012 by TheAmerican Association of Immunologists, Inc. 0022-1767/12/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1102620

by guest on May 8, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

mailto:terhorst@bidmc.harvard.eduhttp://www.jimmunol.org/

PKC activity

PKC activity was measured using the ADI-EKS-420A assay kit (Enzo LifeSciences, Plymouth Meeting, PA), as described in the manufacturers protocol.TGC-elicited MFs were stimulated with 1 mg/ml PMA or heat-killed E. coliF18 (MOI 100) for the indicated times and then lysed with 500 ml RIPAbuffer. Thirty microliters of lysate/sample was used to quantify PKC activity.

Western blotting

TGC-elicited MFs from wt and Slamf82/2 mice were stimulated with 1 mg/ml PMA or E. coli F18 (MOI 100) at the indicated time points. Phospho-p40phox (Thr154) Ab (no. 4311; Cell Signaling Technology, Beverly, MA)and p40-phox Ab (sc-30087; Santa Cruz Biotechnology, Santa Cruz, CA)were used to detect phosphorylation and protein expression of p40phox bystandard immunoblotting procedures.

Statistical analysis

Statistical significance was determined by the Student t test (two-tailed dis-tribution with a two sample equal variance). The p values , 0.05 wereconsidered significant.

Results and DiscussionSlamf8 is upregulated in MFs upon stimulation

To evaluate whether expression of Slamf8 was induced byinflammatory agents, resident pMFs isolated from BALB/cmice were incubated with LPS, IFN-g, E. coli, S. aureus, orPMA, followed by TaqMan analysis. Slamf8 mRNA wasvirtually undetectable in the resting cells, but it increased upto 100-fold after exposure to IFN-g (Fig. 1A, 1B). E. coli orPMA caused a 10-fold induction of expression, whereas LPSand S. aureus each had a lesser effect (Fig. 1B). IFN-g alsosignificantly induced Slamf8 mRNA expression in bonemarrow-derived MFs (Fig. 1C). Slamf8 expression was up to200 times higher in TGC-elicited MFs than in resident cells(Fig. 1D). Nonetheless, in response to IFN-g, Slamf8 ex-pression was still upregulated further in cell sorter-purifiedTGC-elicited monocytes (CD11b+Ly6ClowF4/80low) andMFs (CD11b+F4/80+) (Fig. 1E). This upregulation ofSlamf8 in MFs, particularly by IFN-g, could indicate thatthis receptor functions in late responses to innate stimuli.Thus, Slamf8 expression in MFs depends upon the presenceof stimuli associated with bacterial infections.

Slamf8 is a negative regulator of Nox2 activity

To assess whether Slamf8 would have an effect on MFfunctions, Nox2 enzyme activity in Slamf82/2 BALB/c MFswas tested. Unexpectedly, in response to E. coli and S. aureus,a very significantly elevated Nox2 activity was found inSlamf82/2 bone marrow-derived MFs (Fig. 2A). Surpris-ingly, this phenotype was also observed in response to thereceptor-independent PMA stimulus. Importantly, Slamf82/2

TGC-elicited MFs had a phenotype similar to that of thebone marrow-derived MFs (Fig. 2B), in contrast to residentresting MFs (data not shown).To determine whether the increased Nox2 activity in

Slamf82/2 MFs occurs in the bacterial phagosome, we useda fluorescence-based assay based on the principle that con-version of the reactive oxygen species that are produced byNox2 to HO2 and H2O2 consumes protons, which enter thephagosome via a proton pump. Thus, a defective Nox2 ac-tivity results in less proton consumption in the bacterialphagosome, and, hence, the pH is decreased more rapidly (1).Therefore, we assumed that a hyperactive Nox2 enzyme inthe bacterial phagosome would result in increased protonconsumption and impaired acidification. To test this, a pH-

sensitive dye, pHrodo, coupled to either E. coli or S. aureus,was used in conjunction with flow cytometry. Consistent withthe Nox2 phenotype, Slamf82/2-resident MFs had no de-tectable changes in phagosomal acidification in comparisonwith wt cells (Fig. 2C). As predicted, Slamf82/2 TGC-elicitedMFs displayed subnormal phagosomal acidification com-pared with Slamf8+/+ cells (Fig. 2D). We conclude that Slamf8negatively modulates phagosomal Nox2 activity and acidifi-cation in a cargo-independent manner.To confirm that Slamf8 negatively modulates Nox2 activity,

RAW 264.7 MFs were transfected with Slamf8, which greatlyreduced Nox2 activity in response to bacteria and PMA (Fig.3A). Furthermore, transfection of Slamf8 into Slamf82/2

primary MFs also resulted in a significant reduction in theproduction of reactive oxygen species by Nox2 (Fig. 3B).Because we found that IFN-g significantly enhances Slamf8

expression in RAW 264.7 MFs (Figs. 1F, 3C), we tested thedependence of Nox2 activity in these MFs. Indeed, thepresence of Slamf8 suppressed Nox2 activity in these cells(Fig. 3D). Taken together, these studies demonstrate that

FIGURE 1. Upregulation of expression of Slamf8. (A) Time course of

induction of Slamf8 expression by resident pMFs in response to IFN-g (10

ng/ml), as judged by TaqMan analysis. (B) Relative expression of Slamf8 in

resident pMFs upon 12 h stimulation with LPS (100 ng/ml), IFN-g (10 ng/

ml), E. coli F18 (MOI 100), S. aureus (MOI 100), or PMA (1 mg/ml). (C)Slamf8 expression in bone marrow-derived MFs upon IFN-g (10 ng/ml)

stimulation. (D) Comparison of Slamf8 expression in resting and TGC-eli-

cited peritoneal cells using TaqMan analysis. (E) Expression of Slamf8 before

and after IFN-g (10 ng/ml) stimulation in FACS-sorted TGC pMFs

(CD11b+F4/80+) and monocytes (CD11b+Ly6ClowF4/80low). The data are

representative of at least three independent experiments.

5830 CUTTING EDGE: SLAMF8 REGULATES NOX2 ACTIVITY

by guest on May 8, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

http://www.jimmunol.org/

Slamf8 is a negative regulator of Nox2 activity in MFs.Therefore, this function of Slamf8 contrasts with Slamf1,which is a positive regulator of the same enzyme activity inresponse to E. coli but not to S. aureus or PMA (1).

Slamf8 negatively controls phosphorylation of p40 upon induction byE. coli or PMA

To elucidate how Nox2 activity in Slamf82/2 MFs could beincreased by bacteria and PMA, we reasoned that PKC ac-tivity in these cells might be increased, because this enzyme isknown to phosphorylate p40phox (6). To this end, we firstdetermined that the protein expression of the key componentsof Nox2 (i.e., gp91phox, p22phox, p40phox, p47phox, andp67phox) was similar in wt and Slamf82/2 MFs. Westernblotting of wt and Slamf82/2 MF lysates showed that thebasal expressions of all subunits are comparable (Supple-

mental Fig. 2A). Furthermore, after LPS, E. coli F18, S. au-reus, or PMA activation, the mRNA expression level of themembrane-bound enzyme components gp91phox and p22phoxwas identical in wt and Slamf82/2 MFs (Supplemental Fig.2B).The isoforms of the serine/threonine kinase PKC in mouse

MFs, which are regulated by PMA, are PKC-b, -d, -, and -h(5). PKC-d is known to phosphorylate p40phox on threonine154 and serine 315 (7), Phospho-p40phox stabilizes the as-sembled Nox2 enzyme complex (8). We first determined thekinetics of overall PKC activity in response to various stimuliin wt and Slamf82/2 MFs (Fig. 4A). The PKC activityreached a peak at 5 min after PMA stimulation and wassignificantly higher in Slamf82/2 MFs than in wt cells, whichcorrelates with Nox2 activity (Fig. 2). Similarly, upon expo-sure to E. coli, enzyme activation was higher in Slamf82/2

FIGURE 2. Slamf8 is a negative regulator of Nox2

activity in MFs. Bone marrow-derived (A) and TGC-

elicited (B) MFs were stimulated with E. coli F18 (MOI100), S. aureus (MOI 100), or PMA (1 mg/ml) andassessed for Nox2 activity using lucigenin. Resident

peritoneal MFs (C) or TGC-elicited pMFs (D) were

examined for phagosomal acidification using E. coli orS. aureus (MOI 15) pHrodo bacteria. The data are rep-resentative of three independent experiments.

FIGURE 3. Introduction of Slamf8 into cells by

transfection or IFN-g stimulation rescues the Nox2-as-

sociated phenotype. (A) RAW 264.7 cells were trans-

fected with Slamf8-mCherry or a mock construct,

stimulated with E. coli F18 (MOI 100), S. aureus (MOI100), or PMA (1 mg/ml), and assessed for Nox2 oxidase

activity. (B) Slamf82/2 primary MFs were transfectedwith Slamf8-mCherry, stimulated with E. coli F18 (MOI100), and assessed for Nox2 activity. (C) RAW 264.7

cells were incubated for 12 h with IFN-g (10 ng/ml) and

assessed for Slamf8 expression, as in Fig. 1. (D) Nox2

activity in IFN-gstimulated RAW 264.7 cells was de-

termined as in (A). The data are representative of three

independent experiments.

The Journal of Immunology 5831

by guest on May 8, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

http://www.jimmunol.org/

MFs than in Slamf8+/+ MFs, but it peaked at 15 min (Fig.4A). Consistently with enhanced PKC activity in Slamf82/2

MFs, the level of p40phox phosphorylation was significantlyhigher in Slamf82/2 MFs, and phosphorylation in responseto PMA preceded that of E. coli (Fig. 4B, 4C).These data indicated that Slamf8 negatively controls PKC

activity in MFs, which is known to phosphorylate p40phox,a key component of Nox2 in MFs. Therefore, in the absenceof Slamf8, Nox2 activity increases. Further biochemical andcell biology experiments are required to understand howSlamf8 is coupled to the PKC activated by bacteria and PMAin MFs and phagosomes, in particular.Taken together, the outcomes of these studies show that

Slamf8 is a novel negative regulator of Nox2 activity in MFs.Because Slamf8 expression is increased in monocytes andMFs upon IFN-g stimulation and/or exposure to bacteria,the data suggest that Slamf8 signaling could become opera-tional to dampen an ongoing innate immune response. Thisnotion is supported by the previous observation that the de-gree of MF activation has a differential effect on phagosomeactivity (911). Because Slamf8 is a negative regulator ofNox2 activity in MFs, Slamf8-based therapeutic strategiescould be designed to mitigate inflammatory conditions.

AcknowledgmentsWe thank Glenn E. Brown and Michael B. Yaffe (Massachusetts Institute of

Technology, Cambridge, MA) for help with the initial Nox2 analyses. We are

grateful to Jacky Flipse for helpful discussions and Drs. Pablo Engel, Chun-

yan Ma, and Marton Keszei for critical review of the manuscript.

DisclosuresThe authors have no financial conflicts of interest.

References1. Berger, S. B., X. Romero, C. Ma, G. Wang, W. A. Faubion, G. Liao, E. Compeer,

M. Keszei, L. Rameh, N. Wang, et al. 2010. SLAM is a microbial sensor thatregulates bacterial phagosome functions in macrophages. Nat. Immunol. 11: 920927.

2. Calpe, S., N. Wang, X. Romero, S. B. Berger, A. Lanyi, P. Engel, and C. Terhorst.2008. The SLAM and SAP gene families control innate and adaptive immuneresponses. Adv. Immunol. 97: 177250.

3. Cannons, J. L., S. G. Tangye, and P. L. Schwartzberg. 2011. SLAM family receptorsand SAP adaptors in immunity. Annu. Rev. Immunol. 29: 665705.

4. Kingsbury, G. A., L. A. Feeney, Y. Nong, S. A. Calandra, C. J. Murphy,J. M. Corcoran, Y. Wang, M. R. Prabhu Das, S. J. Busfield, C. C. Fraser, andJ. L. Villeval. 2001. Cloning, expression, and function of BLAME, a novel memberof the CD2 family. J. Immunol. 166: 56755680.

5. Heng, T. S., M. W. Painter; Immunological Genome Project Consortium. 2008.The Immunological Genome Project: networks of gene expression in immune cells.Nat. Immunol. 9: 10911094.

6. Bertram, A., and K. Ley. 2011. Protein kinase C isoforms in neutrophil adhesionand activation. Arch. Immunol. Ther. Exp. (Warsz.) 59: 7987.

7. Bouin, A. P., N. Grandvaux, P. V. Vignais, and A. Fuchs. 1998. p40(phox) isphosphorylated on threonine 154 and serine 315 during activation of the phagocyteNADPH oxidase. Implication of a protein kinase c-type kinase in the phosphory-lation process. J. Biol. Chem. 273: 3009730103.

8. Bedard, K., and K. H. Krause. 2007. The NOX family of ROS-generating NADPHoxidases: physiology and pathophysiology. Physiol. Rev. 87: 245313.

9. Geissmann, F., M. G. Manz, S. Jung, M. H. Sieweke, M. Merad, and K. Ley. 2010.Development of monocytes, macrophages, and dendritic cells. Science 327: 656661.

10. Yates, R. M., A. Hermetter, G. A. Taylor, and D. G. Russell. 2007. Macrophageactivation downregulates the degradative capacity of the phagosome. Traffic 8: 241250.

11. Rybicka, J. M., D. R. Balce, M. F. Khan, R. M. Krohn, and R. M. Yates. 2010.NADPH oxidase activity controls phagosomal proteolysis in macrophages throughmodulation of the lumenal redox environment of phagosomes. Proc. Natl. Acad. Sci.USA 107: 1049610501.

FIGURE 4. PKC activity and phosphorylation of p40phox in Slamf82/2 and Slamf8+/+ MFs upon stimulation with PMA and E. coli. (A) PKC kinase activityin MFs after stimulation with 1 mg/ml PMA and E. coli F18 (MOI 100). The data are representative of two independent experiments, each consisting of threeSlamf8+/+ and Slamf82/2 mice. Phosphorylated p40phox was determined by Western blotting in response to 1 mg/ml PMA (B) or E. coli F18 (MOI 100) (C).Relative p40phox phosphorylation level was normalized to total p40phox (B) based on three independent experiments, or to b-actin (C), which is representative of

three independent experiments.

5832 CUTTING EDGE: SLAMF8 REGULATES NOX2 ACTIVITY

by guest on May 8, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

http://www.jimmunol.org/