of August 19, 2016.This information is current as

Capture by Marginal Zone MacrophagesMarginal Zone B Cells Regulate Antigen

John F. Kearney, Louis B. Justement and Robert H. CarterYuying You, Riley C. Myers, Larry Freeberg, Jeremy Foote,

http://www.jimmunol.org/content/186/4/2172doi: 10.4049/jimmunol.1002106January 2011;

2011; 186:2172-2181; Prepublished online 21J Immunol 

MaterialSupplementary

6.DC1.htmlhttp://www.jimmunol.org/content/suppl/2011/01/21/jimmunol.100210

Referenceshttp://www.jimmunol.org/content/186/4/2172.full#ref-list-1

, 15 of which you can access for free at: cites 33 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. All rights reserved.9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Marginal Zone B Cells Regulate Antigen Capture byMarginal Zone Macrophages

Yuying You,*,1 Riley C. Myers,* Larry Freeberg,* Jeremy Foote,*,2 John F. Kearney,*

Louis B. Justement,*,3 and Robert H. Carter†,3

The marginal zone (MZ) of the mouse spleen contains macrophages that express receptors that trap pathogens, including the sca-

venger receptor macrophage receptor with a collagenous structure and the C-type lectin specific intracellular adhesion mole-

cule-grabbing nonintegrin receptor 1 (SIGN-R1). We previously reported that expression of SIGN-R1 was decreased in CD19-

deficient mice. In this study, we demonstrate that SIGN-R1 is expressed on a subset of macrophage receptor with a collagenous struc-

ture (MARCO)+ macrophages. This subset is diminished when MZ B cells are absent due to either genetic developmental defects or

following transient migration of B cells out of the MZ. When B cells return to the MZ, there is a delay in recovery of SIGN-R1–

expressing macrophages. During this period, capture of Ficoll, which for the macrophages requires SIGN-R1, remains defective

not only by the macrophages, but also by the B cells. Thus, MZ B cells regulate expression of molecules on macrophages that

are important for trapping Ag, which, in turn, is required for Ag capture by the B cells. The Journal of Immunology, 2011, 186:

2172–2181.

The marginal zone (MZ) forms the outer boundary of thewhite pulp in the mouse spleen and contains macrophagesand B cells that surround blood sinuses (1). The MZ

surrounds the follicles that contain follicular B cells. The MZplays a critical role in defense against pathogens that have enteredthe circulation. A specialized macrophage, the MZ macrophage(MZM), expresses receptors such as the C-type lectin specificintracellular adhesion molecule-grabbing nonintegrin receptor 1(SIGN-R1), which binds polysaccharides such as those found onthe outer wall of bacteria (2, 3), and the scavenger receptormacrophage receptor with a collagenous structure (MARCO) (4,5). Without SIGN-R1, MZM fail to trap blood-borne Streptococ-cus pneumococcus, which results in increased mortality (6, 7). MZB cells, in turn, capture these Ags in transfer from MZM (8, 9) andrapidly produce protective Abs without T cell help (10). In ad-dition, MZ B cells transport Ags into the follicle, where they arepicked up by follicular dendritic cells (FDC) and trigger adaptiveimmunity (11, 12).

In our previous studies, we found that SIGN-R1 expression wasmarkedly diminished in mice that lacked CD19 (13). The resultssuggested that a primary defect in the differentiation of MZB cells, due to absence of CD19, resulted in a secondary defect inMZM. These results refined earlier observations that absence ofall B cells resulted in loss of MZM, as well as reports that lossof MZM produced a secondary defect in MZ B cells (14–16).However, the changes in the macrophages resulting from the ab-sence of MZ B cells and the functional consequences of thesechanges remained to be defined.In more recent studies to understand the effect of MZ B cells

on MZM, we find that MZM are not homogenous. In normalmice, there are at least two types of MZM (MARCO+SIGN-R12

and MARCO+SIGN-R1+) with different phenotypes and func-tions. Although MARCO+SIGN-R12 MZM remain in the absenceof MZ B cells, the MARCO+SIGN-R1+ subset disappears. Wedemonstrate that this also occurs in mice that lack MZ B cells forreasons other than CD19 deficiency. Even transient absence of MZB cells results in a decrease in the percentage of MARCO+ MZMthat express SIGN-R1. The functional consequence is that MZMthat lack expression of SIGN-R1 have reduced capacity to binda model Ag, Ficoll. Without SIGN-R1+ MZM, MZ B cells also areunable to capture Ficoll. Thus, the proper differentiation and lo-calization of MZ B cells is required for expression of receptors onmacrophages that capture Ag for presentation to the B cells.

Materials and MethodsMice

C57BL/6 mice were obtained from The Jackson Laboratory. The C57BL/6CD19-knockout (KO; CD19cre) was as previously described (13). Notch2conditional heterozygous KO mice (Notch2flox/+) were provided by Dr.John Kearney, with the permission of Dr. Tom Gridley (17). S1P1loxp/+

mice were provided by Dr. Proia (18). Both of these were crossed withCD19cre. For FTY720 treatment, B6 mice were injected i.p. with 1 mg/kgFTY720 (Cayman Chemical) or with an equivalent volume of saline. ForLPS treatment, B6 mice were injected i.p. with 25 mg LPS (Sigma-Aldrich). For trinitrophenyl (TNP)-Ficoll treatment, B6 mice were injec-ted i.v. with 100 mg TNP-Ficoll (Biosearch Technologies). To identifymacrophages, 200 mg heat-killed, Alexa Fluor 488-conjugated Staphylo-coccus aureus bioparticles (Invitrogen) were injected i.v. 30 min prior to

*Department of Microbiology, University of Alabama at Birmingham, Birmingham,AL 35294; and †National Institute of Arthritis and Musculoskeletal and Skin Dis-eases, Bethesda, MD 20892

1Current address: Stanford University School of Medicine, Stanford, CA.

2Current address: College of Veterinary Medicine, Auburn University, Auburn, AL.

3L.B.J. and R.H.C. contributed equally to this work.

Received for publication June 24, 2010. Accepted for publication December 9, 2010.

This work was supported in part by the National Institute of Arthritis and Musculo-skeletal and Skin Diseases and National Institutes of Health Grants R01 AI 46225 (toL.B.J.) and R01 AI 14782 (to J.F.K.).

Address correspondence and reprint requests to Dr. Robert H. Carter, National In-stitute of Arthritis and Musculoskeletal and Skin Diseases, Building 31, Room 4C32,31 Center Drive, MSC 2350, Bethesda, MD 20892-2350. E-mail address: carterrob@mail.nih.gov

The online version of this article contains supplemental material.

Abbreviations used in this article: FDC, follicular dendritic cell; KO, knockout; LT,lymphotoxin; MAdCAM-1, mucosal addressin cell adhesion molecule-1; MARCO,macrophage receptor with a collagenous structure; MZ, marginal zone; MZM, mar-ginal zone macrophage; S1P, sphingosine-1-phosphate; SIGN-R1, specific intracel-lular adhesion molecule-grabbing nonintegrin receptor 1; TNP, trinitrophenyl.

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1002106

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

sacrifice of the mice. The University of Alabama at Birmingham In-stitutional Animal Care and Use Committee approved all mouse protocols.Adoptive transfer experiments were as previously described (13).

Abs, flow cytometry, and immunofluorescence

Abs to the following targets were purchased and, in some cases, conjugatedto Alexa fluorochromes: C1qRp (AA4.1-allophycocyanin; eBioscience),IgM (II-41–FITC, BD Pharmingen; II-41–PE-Cy7, BD Biosciences), CD23(B3B4-PE; BD Pharmingen), B220 (RA3-6B2–PerCP and allophycocya-nin-Cy7, BD-Pharmingen), CD1d (1B1-FITC, BD Pharmingen), IgM (goatanti-mouse Alexa Fluor 555; Invitrogen), CD4 (Alexa Fluor 647; CaltagLaboratories), Siglec-1 (Moma-biotin or purified; BMABiomedicals), SIGN-

R1 (ER-TR9-biotin; BMA Biomedicals), mucosal addressin cell adhesionmolecule-1 (MAdCAM-1; MECA-367–purified; BD Pharmingen),MARCO (Santa Cruz Biotechnology), and TNP (G235-PE from BD Bio-sciences). Single-cell suspension of spleen were prepared, stained, andanalyzed by flow cytometry as previously described (13). For flow cyto-metric analysis of macrophages, 200 mg heat-killed, Alexa Fluor 488-conjugated S. aureus bioparticles (catalog number S-23371, Invitrogen)were injected i.v. 30 min prior to sacrifice of the mice. Spleens wereharvested, and splenocytes were stained with MARCO, followed by goatanti-rat IgG-Alexa 647 (Invitrogen) and SIGN-R1–biotin, followed bystreptavidin-Pacific blue (Invitrogen). Slides of spleen sections were pre-pared and analyzed as previously described (13). False colors are used toincrease contrast using Leica confocal software (Leica Microsystems).

FIGURE 1. Lack of SIGN-R1 on MARCO+ MZM in CD19ko mice. A, Spleen sections from wild-type and CD19ko mice were stained for MARCO

(green) and SIGN-R1 (blue) on MZM, for IgM (red) on B cells, and for CD4 (purple) on T cells. Original magnification 3200. B, SIGN-R1 cells per area

and MARCO cells per area in representative MZ from wild-type and CD19ko mice were measured using image analysis software. The ratio of SIGN-R1

cells and MARCO cells were calculated. C, Wild-type and CD19ko mice were injected i.v. with S. aureus bioparticles, and splenocytes were harvested and

stained for SIGN-R1. Bioparticle-positive cells were analyzed for SIGN-R1 expression.

The Journal of Immunology 2173

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

Except as stated otherwise, green represents signals for Alexa 488 staining,blue for Alexa 350, red for Alexa 555, and magenta for Alexa 647. ImageJ1.37V (National Institutes of Health) was used to count MARCO+ andSIGN-R1+ cells in follicles. Ten follicles of three representative mice werechosen in each group. Particles of MARCO and SIGN-R1 were counted byImageJ and divided by the area of follicle or perimeter of follicle (depictedby MAdCAM-1/MOMA-1 staining). Similarly, in FTY720 treatment ex-periments, there were three mice in each time group, and we analyzed 10follicles in images from the spleen of each mouse. Particles of SIGN-R1were counted by ImageJ and divided by perimeter of follicle. Every ex-periment was repeated twice.

Statistics

Statistical comparisons between groups were made using the two-tailedStudent t test. All experiments included at least three mice per group.All data shown are representative of at least three replicate experiments.

ResultsThe SIGN-R1+ subset of MZM is absent in CD19ko mice

SIGN-R1 is often used as a marker of MZM. However, the nearlycomplete absence of SIGN-R1, which we previously reported inCD19ko mice, could represent either the absence of MZM, due toapoptosis or migration, or failure of MZM to express this molecule.To address this question, we costained spleens for MARCO, whichis another marker for MZM, as well as for SIGN-R1. In wild-typemice, both MARCO and SIGN-R1 are present, and SIGN-R1colocalizes with MARCO, but SIGN-R1 is only expressed ona subset of the MARCO+ cells (Fig. 1A, top panels). Qualitatively,it appears that there is greater expression of SIGN-R1 in areaswhere there are more MZ B cells. Thus, SIGN-R1 appears to beexpressed on a subset of MZM and perhaps those in contact withMZ B cells.In the CD19ko mice, the near absence of SIGN-R1 expression is

as we reported previously (Fig. 1A, bottom panels). However,MARCO+ cells are still present. The double-positive SIGN-R1+

MARCO+ population is almost totally missing, whereas single-positive SIGN-R12MARCO+ MZM remain. By quantitative im-age analysis, SIGN-R1+ MZM are decreased by.85% in CD19komice. MARCO+ MZM are decreased by 35% (Fig. 1B). Thus, thepercentage of MARCO+ cells that also express SIGN-R1 is alsodramatically decreased in CD19ko mice to 20% of that in wild-type mice.To confirm the loss of SIGN-R1 expression on MZM in CD19ko

mice, we developed a strategy to study MZM by flow cytometry.Previous studies show that MARCO on MZM is required for thesecells to bind S. aureus (15). Mice were injected with labeled S.aureus bioparticles, sacrificed 30 min later, and spleen cells wereanalyzed for cells with bound or internalized S. aureus particles(Fig. 1C). With this technique, we again observe a decrease in themean level of SIGN-R1 expression on cells from CD19ko mice ascompared with wild-type mice, confirming the loss of SIGN-R1+

MZM in CD19ko mice. In addition, even though the frequency ofbioparticle-positive cells is comparable, the absolute number ofthis population is only half in CD19ko compared with the wild-type mice.

SIGN-R1 expression of MZM reflects MZ B cells inNotch2-heterozygous mice

In CD19ko mice, not only are MZ B cells absent, but follicularB cells are decreased in number, and germinal centers are alsodefective.We previously showed, by adoptive transfer experiments,that when CD19ko mice are reconstituted by wild-type B cells,which restores the MZ B cells, then SIGN-R1+ MZM are recon-stituted, indicating that the effect was due to an intrinsic defect inthe B cells. However, this also leads to a reconstitution of normal

follicular cells and germinal centers (data not shown). Thus, theprevious results do not rule out effects of CD19 on cells other thanMZ B cells, or effects on the MZ B cells that are specific toa defect in CD19, that lead to failure of differentiation of MZM.In contrast, Notch22/2 mice have normal follicular B cells andgerminal centers. To determine whether the loss of SIGN-R1+

MZM represents a specific feature related to CD19 deficiency or isa more general defect related to the absence of MZ B cells, westudied Notch2 heterozygous mice. Conditionally targeted de-letion of Notch2 in B cells results in the absence of MZ B cellsand their precursors (19). Notch22CD192 double-deficient micelack MZ B cells, and, as expected, SIGN-R1+ MZM are alsoabsent (Fig. 2A, top panels). Notch2 heterozygous mice have amore variable phenotype: some have few MZ B cells, whereassome have a frequency comparable to wild-type mice (19).Splenocytes from mice that were genetically conditionally het-erozygous for Notch2 in B cells were screened by flow cytometryfor the presence or absence of MZ B cells (representative resultsare shown for both types of mice in Fig. 2B) (CD19cre/+ micehave normal MZ). Splenic sections from mice that were soidentified were then analyzed histologically for expression ofSIGN-R1 on MZM. When MZ B cells are present, SIGN-R1 is

FIGURE 2. Correlation of SIGN-R1+ MZM and MZ B cells in Notch2

heterozygous mice. A, Spleen sections from Notch2loxp/loxpCD19cre/cre and

Notch2loxP/+CD19cre/+ mice were stained for SIGN-R1 (blue) and MARCO

(green) on MZM and for IgM (red) on B cells. Original magnification

3200. B, Splenocytes from representative wild-type and indicated mice in

A were analyzed by flow cytometry. IgM+B220+AA4.12 mature B cells

were analyzed for IgMhiCD23low MZ B cells. Images in the bottom two

panels of A are from the same mice illustrated in the bottom two panels of

B and are representative of mice that either have or lack MZ B cells.

2174 MZ B CELLS REGULATE Ag CAPTURE BY MACROPHAGES

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

expressed on MZM (Fig. 2A, middle panels). Mice that lack MZB cells as a result of conditional heterozygous deficiency of Notch2 lack expression of SIGN-R1 (Fig. 2A, bottom panels), similar towhat is observed in CD19ko mice (Fig. 2A, top panels). Thissuggests that the reduced expression of SIGN-R1 on MZM isa general defect when MZ B cells are absent.

Transient migration of MZ B cells results in transient loss ofSIGN-R1 MZM

The two models presented above lack MZ B cells on the basisof a genetically determined developmental deficiency. We askedwhat would happen in developmentally normal mice if MZ B cellswere induced to migrate out of the MZ. Sphingosine-1-phosphate(S1P) 1 and S1P3, two receptors for S1P, are expressed at highlevels on MZ B cells and control their localization in the MZ (20).Treating mice with FTY720, an antagonist to S1P, releases the MZB cells from the MZ so that MZ B cells migrate into the follicles(20). B cells move out of the MZ as early as 4 h posttreatment. Atthat time point, SIGN-R1 expression is already significantly de-creased compared with control mice, although the defect is not asprofound as in mice with genetically determined absence of MZB cells (Fig. 3A). Although MZ B cells are again present in theMZ at 24 h, SIGN-R1 expression is not fully restored until 96 hlater, as demonstrated by quantitative image analysis (Fig. 3B).Thus, transient migration of MZ B cells out of the MZ results intransient loss of SIGN-R1 expression on MZM. In these mice, MZB cells and MZM develop normally (in contrast to the mice withgenetically determined absence of MZ B cells), but the continuedpresence of MZ B cells is required to maintain SIGN-R1 expres-sion. The restored expression of SIGN-R1 on MARCO+ MZM ap-pears in proximity to IgM-bright MZ B cells, again suggesting thatMZ B cells probably trigger SIGN-R1 expression on macrophagesvia a cell–cell contact (replicating this in vitro is hampered bythe rapid apoptosis of the MZ B cells in culture).

Expression of SIGN-R1 on MZM requires the presence of MZB cells in the MZ

S1P receptors are not only expressed on MZ B cells, but also onother lymphocytes, such as T cells. To exclude the possible non-specific effects by FTY720, we developed S1P1 conditional KOmice (S1P1loxP/loxPCD19cre/+), in which only B cells lack S1P1. Insuch mice, mature MZ B cells and precursors are both present.However, they are located in the follicle and are largely absentfrom the MZ (Fig. 4A). Flow cytometry confirms that S1P1 con-ditional KO mice have a larger proportion of AA4.12CD23low

IgMhi MZ B cells than wild-type mice (Fig. 4B). Nevertheless,SIGN-R1+ MZM are absent in these mice (Fig. 4A, 4C), similar tothe CD19ko mice and Notch2 conditional mice, which lack ma-ture MZ B cells. These results suggest that MZ B cells need to bephysically located in the MZ to trigger and maintain SIGN-R1expression on MZM.The molecular mechanism by which MZ B cells regulate

SIGN-R1 expression onMZM remains to be elucidated. The role ofCD19 is likely to be indirect, related to survival of B cells in theMZ. Indeed, as CD19ko mice age beyond 10–12 wk, some beginto accumulate MZ B cells and recover expression of SIGN-R1(Supplemental Fig. 1). In preliminary studies, we have foundthat expression of SIGN-R1 on MZM is equivalent to wild-typein mice deficient in either IL-10 or C3, suggesting that these arenot the necessary factors (Supplemental Fig. 1). Furthermore, Bcells from mice lacking lymphotoxin (LT)-b were able to restoreexpression of SIGN-R1 on MZM after adoptive transfer intoCD19ko mice (Supplemental Fig. 2). Therefore, CD19, C3, IL-10,and LT do not appear to be required for induction of SIGN-R1expression on MZM.

Ag capture by MZM is impaired without MZ B cells

SIGN-R1 is a C-type lectin that plays an important role in uptake ofpolysaccharides, which are major T-independent Ags on various

FIGURE 3. Transient relocation of MZ B cells

induces loss of SIGN-R1 on MZM. A, Total of 1

mg/kg FTY720 was injected into wild-type mice.

At 4, 24, 48, and 96 h, spleen sections were

stained for SIGN-R1 (green) on MZM, IgM (red)

on B cells, and MAdCAM-1 (blue) on marginal

sinus lining cells. The middle and bottom panels

contain higher magnification images of the areas

in the purple boxes in the upper panels. The

arrows indicate the MZ. Original magnification

3100. B, SIGN-R1+ cells in the perimeters of the

marginal sinus were measured using image anal-

ysis software. *p , 0.01, **p , 0.005, ***p ,0.001.

The Journal of Immunology 2175

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

bacteria (6, 21). To test the hypothesis that without MZ B cells,

and hence secondarily without SIGN-R1+ macrophages, the

trapping of certain types of Ag would be impaired, we compared

CD19ko and wild-type mice after i.v. injection of TNP-Ficoll. In

spleen of wild-type mice, TNP-Ficoll is trapped in the MZ (Fig.

5A). In contrast, in CD19ko mice, although both MARCO+ and

MOMA-1+ macrophages are present, trapping of TNP-Ficoll is

significantly reduced. After recovery of SIGN-R1 expression in

CD19ko mice following adoptive transfer of wild-type B cells,

TNP-Ficoll capture is restored (Fig. 5B). In the CD19ko mice in

Fig. 5B, note that the S. aureus bioparticles were trapped in the

MZ, which indicates that functional MZM were still present, but

these MZM could not capture TNP-Ficoll efficiently. In S1P1

conditional KO mice, which contain mature MZ B cells that are

mislocated in the follicle but that lack SIGN-R1+ MZM, TNP-

Ficoll could not be efficiently trapped in the MZ at short time

periods (Fig. 5C). Thus, SIGN-R1–expressing MZM, which re-

quire the presence of MZ B cells, play a crucial role in capturing

certain types of Ags, such as TNP-Ficoll, in the spleen. When MZ

B cells are absent or have migrated out of the MZ, MZM lose the

ability to take up these Ags.

Capture of Ag by B cells is defective in mice that lackSIGN-R1+ MZM

Ag trapped byMZM is then captured byMZB cells and transportedinto follicles (12). We tested whether conditions that depleteB cells from the MZ and secondarily reduce expression of SIGN-R1 would alter capture of Ag by B cells. LPS, which rapidlyinduces migration of MZ B cells out of the MZ, was injected i.p.into wild-type mice, and spleen sections were analyzed for ex-pression of SIGN-R1 and trapping of Ficoll by macrophages (Fig.6) and for capture of Ficoll by B cells (Fig. 7). At day 1 after LPSadministration, MZ B cells were absent from the MZ, and SIGN-R1 expression was greatly reduced. At day 2, MZ B cells began torepopulate the MZ, whereas SIGN-R1+ MZM were still absent.SIGN-R1 expression on MZM was only restored by days 7–9 (Fig.6A, 6B). Thus, at day 2 after LPS injection, B cells have repo-pulated the MZ, but SIGN-R1 expression is still absent. To ana-lyze Ag trapping when MZ B cells were present but SIGN-R1+

MZM were not, mice were injected with either LPS or PBS, asa control, and then 2 d later injected with TNP-Ficoll. Mice weresacrificed 30 min postinjection with TNP-Ficoll, and splenicB cells were analyzed by flow cytometry for captured TNP. Robust

FIGURE 4. MZ B cells located in the follicle

are unable to induce SIGN-R1 on MZM. A,

Spleen sections from wild-type, S1P1loxP/loxP

CD19cre/+, and S1P1loxP/loxPCD19cre/cre mice

were stained for SIGN-R1 (red) and MARCO

(green) onMZM, IgM (blue) on B cells, and CD4

(gray) on T cells in the leftmost three lanes and

IgM (red) and MARCO (green) in the rightmost

two lanes. The rightmost column contains higher

magnification images of the areas in the purple

boxes in the adjacent column. The arrows indicate

theMZ. Original magnification3100.B, IgM+

B220+AA4.12 mature spleen B cells from wild-

type and S1P1loxP/loxPCD19cre/+ mice were

stained with anti-IgM and anti-CD23 to measure

IgMhiCD23low MZ B cells. C, Wild-type and

S1P1loxP/loxPCD19cre/+ mice were injected i.v.

with S. aureus bioparticles. Splenocytes were

harvested and stained for SIGN-R1. Bioparticle+

cells were analyzed for SIGN-R1 expression.

2176 MZ B CELLS REGULATE Ag CAPTURE BY MACROPHAGES

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

uptake of Ag was observed in MZ B cells of wild-type mice thathad been treated with PBS 2 d before. In contrast, despite the factthat MZ B cells were present, as determined by histology (noteIgM+ cells outside Siglec-1+ marginal metallophilic macrophagesin Fig. 7) and by flow cytometry (gated as B220+AA4.12IgMhi

CD23low), little TNP-Ficoll was captured by MZ B cells in micethat had received LPS 2 d previously (Fig. 7A). Thus, despite thepresence of B cells in the MZ, these cells fail to capture Ag whenSIGN-R1+ MZ are absent, most likely due to prior interruption ofthe contact of the MZM with MZ B cells. We also tested uptake ofAg by B cells in CD19ko mice and S1P1 conditional KO mice,which lack SIGN-R1+ MZM, as above. Although MZ B cellprecursors (gated as B220+CD1dhiIgMhiCD23hi) or mature MZB cells (gated as B220+CD1dhiIgMhiAA42CD23low), respectively,are present in these mice, they failed to capture TNP-Ficoll (Fig.7B, 7C). Interestingly, Ag capture was reduced not only on the MZB cells or their precursors, but apparently also on follicular B cells(presumably through noncognate mechanisms). The impaired Agcapture is consistent with the decreased short-term Ab responsereported earlier (22, 23).

DiscussionMZ B cells provide a critical connection between innate andadaptive immunity. On one hand, they are a defense against blood-borne Ags, such as bacteria, that bind both the Ag receptor andTLRs and thereby trigger rapid, T-independent differentiation of

MZ B cells into short-lived plasmablasts. In addition, MZ B cellsefficiently transport Ag into the follicle, where the Ag is transferredto FDC, which play a role in the germinal center, the heart ofthe adaptive humoral response. We had found that SIGN-R1–expressing macrophages are lost in the MZ of the spleen ofCD19ko mice (13). We report in this study that this is not due toabsence of CD19 per se, but rather reflects a requirement for thephysical presence of MZ B cells in the MZ. Other genetic defectsthat result in an absence of MZ B cells (Notch2 deficiency), oreven just displacement of MZ B cells out of the MZ (S1P1loxP/loxP

CD19cre/+ mice), have an effect comparable to that observed in theabsence of CD19. Furthermore, even transient migration of MZB cells out of the MZ following administration of LPS or FTY720results in a rapid and substantial loss of SIGN-R1.Data from this study, in conjunction with findings from others

(15, 22), suggest that there are bidirectional physical and func-tional interactions that take place between MZ B cells and MZMthat have important implications for the formation and mainte-nance of the MZ itself, as well as the function of individual cellpopulations within the MZ. Studies have demonstrated that B cellsare critical for the formation and maintenance of the MZ, in-cluding the sinus lining metallophilic macrophages and MZM (14,24). Conversely, it has been shown that MARCO+ MZM play animportant role in the retention of MZ B cells via a process thatinvolves direct contact between MARCO expressed on MZMand MZ B cells (15). The present study demonstrates that the

FIGURE 5. Defective Ag uptake in mice that lack

SIGN-R1+ MZM. A, TNP-Ficoll was injected i.v. into

wild-type and CD19ko mice, and 30 min later, mice

were sacrificed, and sequential spleen sections were

stained for TNP (blue) and either IgM (red) on B cells

and for Siglec-1 (green) on marginal metallophilic

macrophages or for MARCO (green) or SIGN-R1

(green) on MZM. B, Seventeen days posttransfer of

wild-type B cells into CD19ko mice, host mice were

injected i.v. with TNP-Ficoll and S. aureus bioparticles

and sacrificed 30 min later. Spleen sections were ana-

lyzed for TNP (blue), S. aureus bioparticles (green),

and IgM (red). C, Wild-type and S1P1loxP/loxPCD19cre/+

mice were injected with TNP-Ficoll as in A. Spleen

sections were stained for TNP (red), MAdCAM-1

(green) on marginal sinus lining cells, and IgM (blue)

on B cells. Original magnification 3200 (A, B) and

3100 (C).

The Journal of Immunology 2177

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

maintenance of SIGN-R1+ MZM requires the physical presence ofMZ B cells in the MZ itself. This finding is significant becauseSIGN-R1–expressing MZM perform multiple, important functionsthat are not compensated for by other MZM populations lackingSIGN-R1 expression.Most importantly, in CD19ko and S1P1loxP/loxPCD19cre/+ mice

that lack B cells in the MZ, minimal amounts of the classicpolysaccharide model Ag, Ficoll, are retained in the MZ after i.v.injection. Indeed, the disappearance of SIGN-R1 expression thatfollows even transient migration of MZ B cells out of the MZresults in a reduction in capture of Ficoll by any cell in the MZ.Reconstitution of MZ B cells by adoptive transfer of wild-typeB cells into CD19ko mice rescues the ability of cells in the MZ totrap polysaccharide Ags. However, this effect does not involve thedirect capture of Ag by the MZ B cells themselves, but it requiresthe presence of SIGN-R1–expressing MZM, which are maintainedby the MZ B cells, to first trap the Ag. This conclusion is sup-ported by the finding that initially after reconstitution of MZB cells following adoptive transfer of WT B cells into CD19ko

mice, or repopulation of MZ B cells after acute depletion, whenMZ B cells are present but SIGN-R1–expressing macrophages arenot, the ability to capture Ficoll remains defective. Indeed, Ficolltrapping only returns after SIGN-R1+ MZM reappear several dayslater. This is consistent with earlier reports in which blockingSIGN-R1 also prevents the capture of Ficoll in the MZ (25, 26).Studies have now shown that SIGN-R1–expressing MZM per-

form critical functions in addition to the capture of polysaccharideAgs. SIGN-R1 binds C1q, leading to the formation of a classicalC3 convertase that promotes complement activation. Indeed, it hasbeen proposed that SIGN-R1 on MZM plays a dominant role information of the C3 convertase that functions as the primarycomplement fixation pathway for pneumococcal polysaccharides(7). Thus, SIGN-R1 expressed on MZM is likely to be importantfor generating complement components that promote bacterialopsonization and uptake in addition to playing a direct role inbinding and uptake of bacteria. More recently, SIGN-R1+ MZMhave been shown to bind sialylated IgG present in i.v. Ig, leadingto the production of anti-inflammatory factors that promote the

FIGURE 6. Decreased SIGN-R1 expression on

MARCO+ MZM after exposure to LPS. A, Mice were

injected with either LPS or PBS and were sacrificed at

days 0, 1, 2, or 9. Spleen sections were stained for IgM

(red) on B cells, SIGN-R1 (blue) on MZM, and Siglec-

1 (green) on marginal metallophilic macrophages.

Original magnification 3200. B, Wild-type mice were

injected with PBS or LPS on day 0 and sacrificed 1, 2,

or 10 d later. Mice were injected with fluorescent

bioparticles and TNP-Ficoll 30 min prior to sacrifice.

Phagocytic cells (leftmost two columns) from the spleen

were analyzed by flow cytometry for expression of

MARCO and SIGN-R1 and for bound TNP (days 2 and

10).

2178 MZ B CELLS REGULATE Ag CAPTURE BY MACROPHAGES

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

generation of FcgRIIb+ effector macrophages as opposed to in-flammatory macrophages (27). Thus, maintenance of SIGN-R1expression by MZ B cells may be important for preventing the ini-tiation of inflammatory responses until such time as bacterialpathogens are encountered. The binding sites on SIGN-R1 formannan and 2,6-sialylated Fc fragments of IgG are overlapping,and thus, there may be competition between different ligands forbinding to SIGN-R1 that leads to distinct MZM responses. In sucha situation, MZ B cells may enforce SIGN-R1 expression and ananti-inflammatory steady state that can be shifted to an active in-flammatory state upon encounterwith bacterial polysaccharideAgs.These findings indicate that MZ B cells not only provide a

link between the innate and adaptive immune response, they alsofunction as regulatory cells that control the phenotype and func-tion of other cells involved in the acute innate immune response.Capture of Ag is important for survival after bacteremia, even apartfrom any effect on Ab responses (6, 7). Thus, it appears that MZB cells are regulating a function of MZM that is important to pro-tect against lethal infection but that is independent of lymphocyte-mediated immunity.The impaired Ag trapping due to loss of SIGN-R1 expression on

MZM also leads to decreased ability of MZ B cells to bindpolysaccharide Ag. As shown in Fig. 7, even when MZ B cells arepresent in the MZ, they capture little or no Ficoll when SIGN-R1+

MZM are absent. Thus, the SIGN-R1+ MZM are critical for theinitial uptake of polysaccharide Ags, which can then be trans-ferred to MZ B cells, thereby promoting the early innate-likehumoral response associated with differentiation of MZ B cellsinto short-lived plasma cells. Interestingly, this applies not only tothose B cells with a MZ B cell phenotype, but also to B cells in thesame mice with a follicular phenotype. Thus, the regulation of

SIGN-R1+ MZM by MZ B cells alters the (noncognate) capture ofAg by follicular B cells, presumably as a result of diminished Agavailability due to impaired trafficking of Ficoll by MZ B cellsinto the follicles, either directly or after transfer first to FDC andthen to follicular B cells.MZ B cells can acquire Ag from blood DC (28) and MZM;

however, they seem not able to acquire Ag directly from bloodpathogens, which indicates that MZ B play a role as a second lineAPC, following Ag trapping and transfer by macrophages or DC.This is also consistent with earlier reports in which blockingSIGN-R1 prevents the capture of Ficoll in the MZ (23). Our cur-rent results emphasize the bidirectional nature of the interaction,as the MZ B cells induce the expression of molecules requiredfor trapping and transfer of Ag by the MZM.Previous reports, from us and others, have addressed the question

of Ab responses. The short-term Ab response (day 4), initiatedpredominantly by MZ B cells, is decreased in SIGN-R1ko mice ormice injected with anti–SIGN-R1 Ab (22, 23). The decreasedshort-term Ab response is consistent with the impaired Ab capturein the MZ in the absence of SIGN-R1+ MZM demonstrated in thisstudy. Similarly, the titer of intermediate-term Ab responses (days10–14) is comparable in CD19ko and SIGN-R1ko mice (6, 29).Additionally, we previously demonstrated that the distribution of

CD11c+ dendritic cells is also altered in CD19ko mice that lackMZ B cells (13). In the absence of MZ B cells, the dendritic cellsno longer concentrate in the bridging channels and instead aredistributed circumferentially around the MZ. Thus, although thefunctional effect associated with the redistribution of CD11c+ DChas not been fully elucidated, this may represent a secondmechanism by which MZ B cells regulate APC and thereby alterthe adaptive immune response.

FIGURE 7. Defective capture of Ag by B cells

in mice that lack SIGN-R1+ MZM. A, At day 2

after LPS administration, mice were injected with

TNP-Ficoll, and spleens were harvested 30 min

later. Sections were stained for IgM (red) on B

cells and Siglec-1 (green) on marginal metal-

lophilic macrophages to demonstrate the presence

of MZ B cells in the spleen at this time. Follicular

B cells and MZ B cells from the same spleens were

analyzed by flow cytometry for bound TNP. Orig-

inal magnification 3200. B, TNP-Ficoll was

injected into wild-type and CD19ko mice, and the

bound TNP was compared among total B cells,

mature MZ B cells, and MZ B cell precursors. C,

TNP-Ficoll was injected into wild-type and

S1P1loxP/loxPCD19cre/+ mice, and the level of bound

TNP was compared among B cells, non-B cells,

and MZ B cells (CD1dhiIgMhi B cells).

The Journal of Immunology 2179

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

Although MZM were first characterized in the 1980s, they arestill not well understood due to the lack of characteristic markersand the lack of appropriate in vitro assays to assess their function.MARCO, a scavenger receptor, and SIGN-R1, a C-type lectinreceptor, are to date the most well-characterized surface proteinsexpressed by MZM (4, 30–32). Whereas most previous reportshave assumed that MZM are a homogeneous cell population, ourfindings demonstrate that MZM are not homogeneous; instead,SIGN-R1+MZMconstitute a subset ofMARCO+MZM.Using bothimmunofluorescence microscopy and flow cytometry, MARCO+

MZM could be clearly divided into two populations based on theexpression of SIGN-R1: SIGN-R1+MARCO+ macrophages andSIGN-R12MARCO+ macrophages. SIGN-R1+ MZM exhibit adistinct functional role that cannot be compensated for by SIGN-R12MARCO+ MZM. In mice lacking SIGN-R1+ MZM, eventhoughMARCO+MZM are still present and capable of taking up S.aureus bioparticles, the ability to capture Ficoll in an acute manneris lost. Whereas MZ B cells are critical for maintenance of SIGN-R1+ MZM, the loss of MZ B cells has only a modest effect on theoverall MARCO+SIGN-R12 MZM population that is likely a re-flection of the loss of the SIGN-R1+ subset.The type of signal thatMZB cells provide to maintain SIGN-R1+

MZM could be in the form of a differentiation, retention, or sur-vival signal, although it should be noted that these distinct types ofsignals are not mutually exclusive. With respect to delivery ofa differentiation signal, MZ B cells may induce the expressionof SIGN-R1 on any given MARCO+ MZM. Alternatively, SIGN-R1+ cells might represent a subset of MARCO+ MZM that con-stitute a unique macrophage lineage within the MZ that is dif-ferentially regulated by MZ B cells. In support of the latterhypothesis, it was observed that in mice lacking MZ B cells, andwhich lack SIGN-R1+ MZM, there is a diminution in the totalnumber of MARCO+ cells that is proportional to the reduction inthe number of SIGN-R1+ cells. This suggests that MZ B cells mayin fact control the fate of a subset of MZM, because if MZ B cellswere simply regulating the expression of SIGN-R1 on MARCO+

MZM in a stochastic manner, then one would not expect a changein the total number MARCO+ cells when MZ B cells are absent,which is not the case (Fig. 1B). In terms of the possibility that MZB cells deliver a retention signal that keeps SIGN-R1+ MZM frommigrating out of the MZ, SIGN-R1 expression appears to belimited to the MZ in the spleen. We have not been able to detectthe migration of SIGN-R1+ MZM into other areas of the spleen orinto the blood after depletion of B cells from the MZ, suggestingthat a retention signal is less likely. In contrast, the available ev-idence suggests that MZ B cells may indeed provide a survivalsignal that is critical for maintenance of SIGN-R1+ MZM becausethe ability to detect SIGN-R1 expression is rapidly lost within 2 to3 h after acute migration of MZ B cells, and as noted above, thereis a decrease in the total number of MARCO+ MZM that is pro-portional to the loss of SIGN-R1+ cells after MZ B cells are in-duced to migrate out of the MZ. Although this could in theory bedue to acute downregulation of SIGN-R1 expression on MZM,this is less likely because one would expect that repopulation ofMZ B cells in the MZ would lead to an equally rapid upregulationof SIGN-R1 expression on MZM, which is not the case. Indeed, itwas observed that reacquisition of SIGN-R1+ MZM requiredanywhere from 2–10 d after B cells repopulated the MZ dependingon the experimental system being studied. This finding also sup-ports the conclusion that MZ B cells may control the differenti-ation of a unique subset of MZM.In summary, our findings demonstrate that MZ B cells are es-

sential for maintenance of SIGN-R1+ MZM in the MZ of thespleen. Loss of SIGN-R1–expressing MZM dramatically affects

the acute uptake of Ficoll-type polysaccharides not only by MZM,but also MZ B cells. Thus, we have established a cross-talkpathway between B cells and certain macrophages: on the onehand, MZ B cells are required to trigger the expression of SIGN-R1 on MZM; on the other hand, B cells can only capture certainAg efficiently when MZM express SIGN-R1. This implies thatB cells, through regulation of SIGN-R1 on macrophages, havea feedback effect on Ag presentation, similar to the effect of thetwo-way interaction of T cells and the dendritic cells that are theAPC for T cells. Considering this cross-talk between MZ B cellsand MZM, impaired Ag capture in mice without MZ B cells inearlier reports (9, 33) could be due to a defect of SIGN-R1+ MZM.Because SIGN-R1 expression is important for other functionsassociated with complement activation and regulation of the in-flammatory state in the animal, understanding the molecularmechanisms by which MZ B cells regulate SIGN-R1 expression inthe MZ is significant. Moreover, studies will need to be extendedto humans to elucidate the role that MZ B cell populations play interms of regulating the function of macrophage and dendritic cellpopulations, as there are structural and phenotypic differences thatmay alter the nature of the physical and functional interactionsbetween these cell populations.

AcknowledgmentsWe thank R.L. Proia for the S1P1loxp/+ mice, A. Szalai for the C3ko mice,

C. Weaver for the IL-10ko mice, D. Chaplin for the LT-a/LT-b/TNF triple

KO mice, and J. Slides for review of the manuscript.

DisclosuresThe authors have no financial conflicts of interest.

References1. Mebius, R. E., and G. Kraal. 2005. Structure and function of the spleen. Nat. Rev.

Immunol. 5: 606–616.2. Kang, Y. S., J. Y. Kim, S. A. Bruening, M. Pack, A. Charalambous, A. Pritsker,

T. M. Moran, J. M. Loeffler, R. M. Steinman, and C. G. Park. 2004. The C-type

lectin SIGN-R1 mediates uptake of the capsular polysaccharide of Streptococcus

pneumoniae in the marginal zone of mouse spleen. Proc. Natl. Acad. Sci. USA

101: 215–220.3. Taylor, P. R., L. Martinez-Pomares, M. Stacey, H. H. Lin, G. D. Brown, and

S. Gordon. 2005. Macrophage receptors and immune recognition. Annu. Rev.

Immunol. 23: 901–944.4. Elomaa, O., M. Kangas, C. Sahlberg, J. Tuukkanen, R. Sormunen, A. Liakka,

I. Thesleff, G. Kraal, and K. Tryggvason. 1995. Cloning of a novel bacteria-

binding receptor structurally related to scavenger receptors and expressed in

a subset of macrophages. Cell 80: 603–609.5. Ito, S., M. Naito, Y. Kobayashi, H. Takatsuka, S. Jiang, H. Usuda, H. Umezu,

G. Hasegawa, M. Arakawa, L. D. Shultz, et al. 1999. Roles of a macrophage

receptor with collagenous structure (MARCO) in host defense and heterogeneity

of splenic marginal zone macrophages. Arch. Histol. Cytol. 62: 83–95.6. Lanoue, A., M. R. Clatworthy, P. Smith, S. Green, M. J. Townsend, H. E. Jolin,

K. G. Smith, P. G. Fallon, and A. N. McKenzie. 2004. SIGN-R1 contributes to

protection against lethal pneumococcal infection in mice. J. Exp. Med. 200:

1383–1393.7. Kang, Y. S., Y. Do, H. K. Lee, S. H. Park, C. Cheong, R. M. Lynch,

J. M. Loeffler, R. M. Steinman, and C. G. Park. 2006. A dominant complement

fixation pathway for pneumococcal polysaccharides initiated by SIGN-R1

interacting with C1q. Cell 125: 47–58.8. Pillai, S., A. Cariappa, and S. T. Moran. 2005. Marginal zone B cells. Annu. Rev.

Immunol. 23: 161–196.9. Guinamard, R., M. Okigaki, J. Schlessinger, and J. V. Ravetch. 2000. Absence of

marginal zone B cells in Pyk-2-deficient mice defines their role in the humoral

response. Nat. Immunol. 1: 31–36.10. Martin, F., A. M. Oliver, and J. F. Kearney. 2001. Marginal zone and B1 B cells

unite in the early response against T-independent blood-borne particulate anti-

gens. Immunity 14: 617–629.11. Ferguson, A. R., M. E. Youd, and R. B. Corley. 2004. Marginal zone B cells

transport and deposit IgM-containing immune complexes onto follicular den-

dritic cells. Int. Immunol. 16: 1411–1422.12. Cinamon, G., M. A. Zachariah, O. M. Lam, F. W. Foss, Jr., and J. G. Cyster.

2008. Follicular shuttling of marginal zone B cells facilitates antigen transport.

Nat. Immunol. 9: 54–62.

2180 MZ B CELLS REGULATE Ag CAPTURE BY MACROPHAGES

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

13. You, Y., H. Zhao, Y. Wang, and R. H. Carter. 2009. Cutting edge: Primary andsecondary effects of CD19 deficiency on cells of the marginal zone. J. Immunol.182: 7343–7347.

14. Crowley, M. T., C. R. Reilly, and D. Lo. 1999. Influence of lymphocytes on thepresence and organization of dendritic cell subsets in the spleen. J. Immunol.163: 4894–4900.

15. Karlsson, M. C., R. Guinamard, S. Bolland, M. Sankala, R. M. Steinman, andJ. V. Ravetch. 2003. Macrophages control the retention and trafficking ofB lymphocytes in the splenic marginal zone. J. Exp. Med. 198: 333–340.

16. Nolte, M. A., R. Arens, M. Kraus, M. H. van Oers, G. Kraal, R. A. van Lier, andR. E. Mebius. 2004. B cells are crucial for both development and maintenance ofthe splenic marginal zone. J. Immunol. 172: 3620–3627.

17. McCright, B., X. Gao, L. Shen, J. Lozier, Y. Lan, M. Maguire, D. Herzlinger,G. Weinmaster, R. Jiang, and T. Gridley. 2001. Defects in development of thekidney, heart and eye vasculature in mice homozygous for a hypomorphicNotch2 mutation. Development 128: 491–502.

18. Allende, M. L., T. Yamashita, and R. L. Proia. 2003. G-protein-coupled receptorS1P1 acts within endothelial cells to regulate vascular maturation. Blood 102:3665–3667.

19. Saito, T., S. Chiba, M. Ichikawa, A. Kunisato, T. Asai, K. Shimizu,T. Yamaguchi, G. Yamamoto, S. Seo, K. Kumano, et al. 2003. Notch2 is pref-erentially expressed in mature B cells and indispensable for marginal zone Blineage development. Immunity 18: 675–685.

20. Cinamon, G., M. Matloubian, M. J. Lesneski, Y. Xu, C. Low, T. Lu, R. L. Proia,and J. G. Cyster. 2004. Sphingosine 1-phosphate receptor 1 promotes B celllocalization in the splenic marginal zone. Nat. Immunol. 5: 713–720.

21. Geijtenbeek, T. B., P. C. Groot, M. A. Nolte, S. J. van Vliet, S. T. Gangaram-Panday, G. C. van Duijnhoven, G. Kraal, A. J. van Oosterhout, and Y. vanKooyk. 2002. Marginal zone macrophages express a murine homologue of DC-SIGN that captures blood-borne antigens in vivo. Blood 100: 2908–2916.

22. Koppel, E. A., M. Litjens, V. C. van den Berg, Y. van Kooyk, andT. B. Geijtenbeek. 2008. Interaction of SIGNR1 expressed by marginal zonemacrophages with marginal zone B cells is essential to early IgM responsesagainst Streptococcus pneumoniae. Mol. Immunol. 45: 2881–2887.

23. Moens, L., A. Jeurissen, G. Wuyts, P. G. Fallon, B. Louis, J. L. Ceuppens, andX. Bossuyt. 2007. Specific intracellular adhesion molecule-grabbing nonintegrin

R1 is not involved in the murine antibody response to pneumococcal poly-saccharides. Infect. Immun. 75: 5748–5752.

24. Chen, Y., T. Pikkarainen, O. Elomaa, R. Soininen, T. Kodama, G. Kraal, andK. Tryggvason. 2005. Defective microarchitecture of the spleen marginal zoneand impaired response to a thymus-independent type 2 antigen in mice lackingscavenger receptors MARCO and SR-A. J. Immunol. 175: 8173–8180.

25. Claassen, E., N. Kors, and N. Van Rooijen. 1986. Influence of carriers on thedevelopment and localization of anti-2,4,6-trinitrophenyl (TNP) antibody-forming cells in the murine spleen. II. Suppressed antibody response to TNP-Ficoll after elimination of marginal zone cells. Eur. J. Immunol. 16: 492–497.

26. Kraal, G., H. Ter Hart, C. Meelhuizen, G. Venneker, and E. Claassen. 1989.Marginal zone macrophages and their role in the immune response against T-independent type 2 antigens: modulation of the cells with specific antibody. Eur.J. Immunol. 19: 675–680.

27. Anthony, R. M., F. Wermeling, M. C. Karlsson, and J. V. Ravetch. 2008. Iden-tification of a receptor required for the anti-inflammatory activity of IVIG. Proc.Natl. Acad. Sci. USA 105: 19571–19578.

28. Balazs, M., F. Martin, T. Zhou, and J. Kearney. 2002. Blood dendritic cells in-teract with splenic marginal zone B cells to initiate T-independent immuneresponses. Immunity 17: 341–352.

29. Rickert, R. C., K. Rajewsky, and J. Roes. 1995. Impairment of T-cell-dependentB-cell responses and B-1 cell development in CD19-deficient mice. Nature 376:352–355.

30. van Vliet, E., M. Melis, and W. van Ewijk. 1985. Marginal zone macrophages inthe mouse spleen identified by a monoclonal antibody. Anatomical correlationwith a B cell subpopulation. J. Histochem. Cytochem. 33: 40–44.

31. Dijkstra, C. D., E. Van Vliet, E. A. Dopp, A. A. van der Lelij, and G. Kraal.1985. Marginal zone macrophages identified by a monoclonal antibody: char-acterization of immuno- and enzyme-histochemical properties and functionalcapacities. Immunology 55: 23–30.

32. van der Laan, L. J., E. A. Dopp, R. Haworth, T. Pikkarainen, M. Kangas,O. Elomaa, C. D. Dijkstra, S. Gordon, K. Tryggvason, and G. Kraal. 1999.Regulation and functional involvement of macrophage scavenger receptorMARCO in clearance of bacteria in vivo. J. Immunol. 162: 939–947.

33. Xue, L., S. W. Morris, C. Orihuela, E. Tuomanen, X. Cui, R. Wen, and D. Wang.2003. Defective development and function of Bcl10-deficient follicular, marginalzone and B1 B cells. Nat. Immunol. 4: 857–865.

The Journal of Immunology 2181

by guest on August 19, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from