7506.Full.Pdf

The Virulence Function of Streptococcus pneumoniae Surface Protein A Involves Inhibition of Complement Activation and Impairment of Complement This information is current as Receptor-Mediated Protection of September 28, 2021. Bing Ren, Mark A. McCrory, Christina Pass, Daniel C. Bullard, Christie M. Ballantyne, Yuanyuan Xu, David E. Briles and Alexander J. Szalai

J Immunol 2004; 173:7506-7512; ; Downloaded from doi: 10.4049/jimmunol.173.12.7506 http://www.jimmunol.org/content/173/12/7506 http://www.jimmunol.org/ References This article cites 31 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/173/12/7506.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

by guest on September 28, 2021 • No Triage! Every submission reviewed by practicing scientists

• 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 © 2004 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

The Virulence Function of Streptococcus pneumoniae Surface Protein A Involves Inhibition of Complement Activation and Impairment of Complement Receptor-Mediated Protection1

Bing Ren,* Mark A. McCrory,† Christina Pass,† Daniel C. Bullard,‡ Christie M. Ballantyne,§ Yuanyuan Xu,† David E. Briles,* and Alexander J. Szalai2†

Complement is important for elimination of invasive microbes from the host, an action achieved largely through interaction of complement-decorated pathogens with various complement receptors (CR) on phagocytes. Pneumococcal surface protein A (PspA) has been shown to interfere with complement deposition onto pneumococci, but to date the impact of PspA on CR-mediated host defense is unknown. To gauge the contribution of CRs to host defense against pneumococci and to decipher the impact of ؊/؊ PspA on CR-dependent host defense, wild-type C57BL/6J mice and mutant mice lacking CR types 1 and 2 (CR1/2 ), CR3 Downloaded from CR3؊/؊), or CR4 (CR4؊/؊) were challenged with WU2, a PspA؉ capsular serotype 3 pneumococcus, and its PspA؊ mutant) JY1119. Pneumococci also were used to challenge factor D-deficient (FD؊/؊), LFA-1-deficient (LFA-1؊/؊), and CD18-deficient CD18؊/؊) mice. We found that FD؊/؊, CR3؊/؊, and CR4؊/؊ mice had significantly decreased longevity and survival rate upon) infection with WU2. In comparison, PspA؊ pneumococci were virulent only in FD؊/؊ and CR1/2؊/؊ mice. Normal mouse serum supported more C3 deposition on pneumococci than FD؊/؊ serum, and more iC3b was deposited onto the PspA؊ than the PspA؉ strain. The combined results confirm earlier conclusions that the alternative pathway of complement activation is indispensable http://www.jimmunol.org/ for innate immunity against pneumococcal infection and that PspA interferes with the protective role of the alternative pathway. Our new results suggest that complement receptors CR1/2, CR3, and CR4 all play important roles in host defense against pneumococcal infection. The Journal of Immunology, 2004, 173: 7506–7512.

treptococcus pneumoniae is a major cause of bacterial 3 (C3) (4–7). Ipso facto, PspAϪ pneumococci that are avirulent in pneumonia, otitis media, meningitis, and septicemia, espe- normal mice become virulent in C3-deficient (C3Ϫ/Ϫ) and factor S cially in infants, the elderly, and immunocompromised in- B-deficient (FBϪ/Ϫ) mice (5). dividuals. The thick and rigid cell wall of the Gram-positive pneu- It is currently accepted that three different pathways can lead to by guest on September 28, 2021 mococcus is protective against lysis by the complement membrane complement activation: the classical pathway, the mannose-bind- attack complex, so the main mechanism of complement-mediated ing lectin (MBL) pathway, and the alternative pathway (8). All elimination of pneumococci from infected hosts is by phagocyto- three pathways converge upon activation of C3, which is cleaved sis. Phagocytosis relies on the interaction of Ab and/or comple- into C3a and C3b. This process exposes a thioester group on the ment, deposited on the pathogen’s surface, with phagocyte recep- ␣-chain of C3b that allows C3b to covalently attach to any acti- tors for the Fc portion of Igs or various complement proteins. vating surface in close proximity (9). With the assistance of var- 3 Pneumococcal surface protein A (PspA) is an important viru- ious cofactors, surface-bound C3b is then sequentially processed, lence factor for S. pneumoniae (1–3). Accordingly, in a murine Ϫ via the enzymatic activity of factor I, into smaller fragments (iC3b model of bacteremia, PspA-deficient (PspA ) pneumococci have and C3d/C3dg). Importantly, C3b and its downstream cleavage significantly reduced virulence compared with pneumococci that ؉ products remain covalently bound to the target surface and act as express PspA (PspA ) (2, 3). Earlier studies from our laboratory ligands for different complement receptors (CRs) on various host and others’ have shown that PspA inhibits complement activation, phagocytes (10). Consequently, microbial pathogens sufficiently thereby limiting opsonization of pathogens by complement protein coated with C3 fragments are recognized by phagocytes. The phagocyte CRs include CR1 (also called CD35), CR3 (also Departments of *Microbiology, †Medicine, and ‡Genetics, University of Alabama, ␣ ␤ called M 2, Mac-1, or CD11b/CD18), and CR4 (also called Birmingham, AL 35294; and §Department of Medicine, Baylor College of Medicine, ␣ ␤ Houston, TX 77030 X 2, p150,95, or CD11c/CD18) (11). CR1 reportedly has high affinity for C3b and lower affinity for iC3b, and also binds C4b, Received for publication March 30, 2004. Accepted for publication October 15, 2004. C1q, and MBL (12–14). Likewise, murine CR1 has binding sites The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance for both C3b and C4b (15, 16). Because CR1 has cofactor activity, with 18 U.S.C. Section 1734 solely to indicate this fact. it also participates directly in complement activation by facilitating 1 This work was supported by National Institutes of Health Grants AI21548 and the factor I-mediated conversion of C3b to iC3b. This action of AI42183 (to A.J.S.) and the Carsten Cole Buckley Pediatric Meningitis Research Fund (to D.E.B.). CR1 can down-regulate further complement activation (17). A closely related, but nonphagocytic, receptor, CR2 (CD21), binds 2 Address correspondence and reprint requests to Dr. Alexander J. Szalai, Division of Clinical Immunology and Rheumatology, Department of Medicine, University of Al- C3dg/C3d and thereby influences B cell activation and develop- abama, 437B Tinsley Harrison Tower, 1530 Third Avenue South, Birmingham, AL ment (16, 18). In the mouse, CR1 and CR2 are alternatively spliced 35294-0006. E-mail address: [email protected] products of a single Cr2 gene (19). In contrast to CR1, CR3 and 3 Abbreviations used in this paper: PspA, pneumococcal surface protein A; CR, complement receptor; FB, factor B; FD, factor D; MBL, mannose-binding lectin; WT, CR4 have high affinity for iC3b and lower affinity for C3b. CR3 ␤ wild type. and CR4 are both members of the 2 integrin family, which also

FIGURE 1. Blood clearance/growth of PspAϩ WU2 and PspAϪ JY1119 pneumococci in normal vs mutant mice. WT mice (solid lines) and mutant mice (dashed lines; deficient protein indicated) were infected i.v. with ϳ2 ϫ 105 CFU of WU2 (F) or JY1119 (E). Bacteremia was determined at different time points. Data are presented as the mean Ϯ SEM CFU per milliliter of blood for six to 12 mice (indicated) in each Downloaded from group. The asterisks indicate that bacteremia in deficient mice is significantly different (p Ͻ 0.05) from that in WT mice infected with the same strain of pneumococcus. http://www.jimmunol.org/ by guest on September 28, 2021

␣ ␤ ␣ ␤ includes LFA-1 (also called L/ 2 or CD11a/CD18) and D/ 2 diluted in lactated Ringer’s buffer to achieve the desired working concen- (CD11d/CD18) (20). CR3 and LFA-1 are important adhesion mol- tration. The number of bacteria injected into mice was confirmed by plating ecules involved in the migration of polymorphonuclear neutrophils residual inoculums on blood agar. To assess the net clearance/growth of pneumococci in the circulation, mice were challenged i.v. with 2 ϫ 105 and monocytes from the periphery to sites of inflammation (20). CFU/mouse, and 50 ␮l of blood was collected from the retro-orbital plexus In the present study the impact of deletion of various CRs on at 5 min, 1 h, 2 h, 4 h, and 6 h postinoculation. Blood was serially diluted pneumococcal virulence and on PspA’s virulence function was and plated on blood agar plates to determine the number of viable S. pneu- 2 assessed by comparing the outcome of infections with PspAϩ and moniae recovered. The limit of detection was 10 CFU of bacteria/ml PspAϪ pneumococci in wild-type and various CR-deficient mice. blood; thus, when no pneumococci grew on culture plates, a level of bacteremia of 102 CFU/ml was reported. The mortality of mice was monitored Our results suggest that in the mouse, CR1, CR3, and CR4 each for 10 days after challenge. contribute to host defense again pneumococcal infection.

Materials and Methods Complement deposition assays Animals Pneumococci were grown in Todd-Hewitt broth supplemented with 0.5% yeast extract until the OD was 0.45 (ϳ108 CFU/ml), and 150 ␮lof Wild-type (WT) C57BL/6J mice were obtained from The Jackson Labo- 600 nm Ϫ Ϫ bacterial culture was decanted. Bacteria were washed twice with PBS and ratory (Bar Harbor, ME). Factor D-deficient (FD / ), Cr2-deficient (CR1/ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ resuspended in 50 ␮l of 10% normal mouse serum (isolated and pooled 2 / ), CD11b-deficient (CR3 / ), CD11c-deficient (CR4 / ), CD11a-de- Ϫ/Ϫ Ϫ/Ϫ from five naive WT C57BL/6J mice) diluted in gelatin veronal buffer with ficient (LFA-1 ), and CD18-deficient (CD18 ; mice that lack ϩ calcium (0.15 mM) and magnesium (0.5 mM; GVB2 ; Sigma-Aldrich, St. functional CR3, CR4, and LFA-1) mice were from our own colonies. The Louis, MO). The bacteria were thus opsonized at 37°C for 5, 10, or 30 min, construction and phenotype of each targeted mutant have been fully de- after which ice-cold PBS buffer containing 10 mM EDTA was added to scribed previously (19, 21–23). All these mice are either C57BL/6J con- stop the reaction. Opsonized bacteria were washed, and the bacteria were genic or were backcrossed to C57BL/6J for at least six generations before resuspended in 50 ␮l of reducing SDS sample buffer, boiled for 10 min, use. All animals were housed in separate cages, and none exhibited any and loaded onto SDS-polyacrylamide gels for electrophoresis. Subsequent evidence of infection before pneumococcal challenge at 10–12 wk of age. immunoblots were prepared and developed using goat anti-mouse C3 pol- Bacteria and infections ysera (Bethyl Laboratories, Montgomery, TX) in concert with biotin-conjugated rabbit anti-goat IgG (H and L chains) and alkaline phosphatase- A capsular type 3 pneumococcal strain (WU2, PspAϩ) (24, 25) and its conjugated streptavidin. Immunoreactive bands of protein were quantitated isogenic mutant (JY1119, PspAϪ) (4, 7, 26) were used. Bacterial stocks by densitometry using the Bio-Rad gel documentation system (Bio-Rad, were frozen at Ϫ80°C in Todd-Hewitt broth supplemented with 0.5% yeast Hercules, CA). To validate the findings and determine their clinical rele- extract containing 10% glycerol. To maintain the mutant PspAϪ genotype, vance, the experiment was repeated using normal human serum (lot which carries an erythromycin resistance cassette, the culture medium for OM1013; Quidel, San Diego, CA). In this case, purified human C3, C3b, JY1119 always contained erythromycin (0.3 ␮g/ml). Frozen stocks of each and iC3b proteins were used as controls, and the immunoblot was devel- strain containing a known concentration of viable bacteria were thawed and oped using goat anti-human C3 polysera (both from Quidel). 7508 PspA INHIBITS COMPLEMENT AND COMPLEMENT RECEPTORS

Flow cytometry was used to further quantitate the amount of C3 depos- mice infected with PspAϩ and PspAϪ S. pneumoniae, Mann-Whitney U ited onto bacteria. Pneumococci were incubated with 30% FDϪ/Ϫ or WT tests were used. serum in GVB2ϩ at 37°C for 30 min. Bacteria incubated with only GVB2ϩ served as a negative control. After washing, bacteria were stained with Results FITC-conjugated goat IgG to mouse C3, and the fluorescence intensity of Ϫ each sample was determined using a FACScan instrument (BD Bio- PspA pneumococci persist in the blood and cause fatal Ϫ/Ϫ sciences, Mountain View, CA) bacteremia in FD mice ϩ Ϫ Surface expression of CRs on peripheral blood cells We reported previously that when PspA vs PspA pneumococci were used to challenge FBϪ/Ϫ mice, the PspAϪ bacteria were as The expression of CR1/2 and CR3 on peripheral blood cells was deter- virulent as PspAϩ bacteria (5). However, because the mouse factor mined by staining freshly isolated cells with fluorochrome-conjugated goat anti-mouse CD21/35 or anti-mouse CD11b, respectively. Briefly, heparin- B-encoding gene, Bf, resides within the H-2 locus, we could not ized blood from WT, CR1/2Ϫ/Ϫ, and CD18Ϫ/Ϫ mice was washed with formally rule out the possibility that disruption of H-2, rather than HBSS containing 0.1% BSA and stained with FITC-conjugated rat anti- absence of FB, was responsible for the increased susceptibility of mouse mAb 7G6 (anti-CD21/35) or R-PE-conjugated mAb M1/70 (anti- FBϪ/Ϫ mice to PspAϪ pneumococci. FB and FD are both essential CD11b), both obtained from BD Pharmingen (San Diego, CA). After one wash, each blood sample was incubated with FACS lysing solution (BD for formation of the alternative pathway C3 convertase. C3b, the Biosciences) to deplete erythrocytes. By flow cytometry, total leukocytes activated form of C3, binds FB and presents it for cleavage by FD, were gated based on forward and side scatter properties, and the fluores- yielding C3bBb, the alternative pathway C3 convertase. C3bBb cence intensity profiles of the gated cell populations were determined. sequentially activates more C3 and triggers the alternative pathway Ϫ/Ϫ Ϫ/Ϫ Measurement of serum Abs amplification loop. Consequently, like FB mice, FD mice have no operable alternative pathway and have impaired classical Downloaded from Ϫ/Ϫ Ϫ/Ϫ Sera from CD1/2 , CD18 , and WT mice were collected, and the pathway activity due to the absence of an amplification loop. Im- levels of Abs directed against heat-killed pneumococci and phosphocholine Ϫ Ϫ portantly, FD / mice are not likely to have H-2 impairment. Ag therein were determined by ELISAs. Microtiter plates were coated with Ϫ/Ϫ heat-killed WU2 (108 CFU/well in PBS) or phosphocholine-conjugated We determined the ability of FD to resist infection with the ϩ Ϫ BSA (5 ␮g/well) overnight at 4°C. Mouse serum was serially diluted and PspA and PspA strains. The clearance/growth of pneumococci added to the plates. ELISA plates were washed, sequentially developed in WT and FDϪ/Ϫ mice was not different for the first 4 h after ϩ with biotin-conjugated, goat anti-mouse Ig and streptavidin-alkaline phos- infection with PspA WU2, but pneumococcal burden was sig- http://www.jimmunol.org/ phatase (Southern Biotechnology Associates, Birmingham, AL), and ϳ Ϫ/Ϫ scanned at 405 nm. nificantly higher ( 5-fold higher) in FD than WT by 6 h (Fig. 1A, F). As a consequence, the longevity and survival rate of Statistical analysis FDϪ/Ϫ mice challenged with WU2 were significantly lower than To compare levels of bacteremia, Student’s t tests were applied to log- those in WT mice ( p ϭ 0.013; Fig. 2, left panel). In contrast, the Ϫ transformed data. To compare survival times (longevity) of WT and mutant PspA JY1119 strain was completely cleared from the blood of by guest on September 28, 2021

FIGURE 2. Survival times of WT and mutant mice infected i.v. with ϳ2 ϫ 105 CFU of WU2 (left panel) or JY1119 (right panel). Each dot or cross represents the death of a single mouse, and the mutant gene carried by mice is indicated on the abscissa. For animals infected with each kind of bacteria, the median time to death (horizontal lines) is shown. The longevity of mutants was compared with that of WT mice, and mutants whose survival time was signiﬁcantly less than that of WT mice infected with the same bacteria (p Ͻ 0.05) are indicated with an asterisk. The Journal of Immunology 7509

WT in Ͻ4 h, but persisted in FDϪ/Ϫ (Fig. 1A, E). WT mice chal- tion with JY1119. These data indicate that CR3 and CR4 also are lenged with JY1119 all survived at least 10 days, whereas FDϪ/Ϫ required for full protection from pneumococci. mice challenged with JY1119 had significantly shortened longevity (3.5 days; p ϭ 0.004 compared with WT) and only 14% sur- Clearance/killing of blood-borne pneumococci is marginally vived (Fig. 2, right panel). In vitro, WT serum supported more C3 improved in LFA-1Ϫ/Ϫ and CD18Ϫ/Ϫ mice Ϫ/Ϫ deposition onto pneumococci than did FD serum regardless of Both PspAϩ and PspAϪ bacteria disappeared from the blood of whether WU2 or JY1119 was being opsonized, and the amount of LFA-1Ϫ/Ϫ marginally faster than from WT mice (Fig. 1E, dashed C3 deposited onto JY1119 was consistently greater than that de- lines), and the longevity of WU2-infected LFA-1Ϫ/Ϫ mice was posited onto WU2 (Fig. 3). Together, these results reinforce those somewhat prolonged (Fig. 2, left panel). Similarly, and despite Ϫ/Ϫ we reported previously using FB mice, confirming the impor- their inherited absence of functional CR3 and CR4, WU2 was tance of an intact alternative pathway for the efficient clearance/ removed more quickly from the blood of CD18Ϫ/Ϫ than from WT killing of pneumococci in the mouse. They also support our hy- mice (Fig. 1F, dashed lines), and no difference was observed in the pothesis that PspA circumvents the role of this pathway in survival time of CD18Ϫ/Ϫ and WT mice infected with either pneu- protection. mococcus (Fig. 2). Clearance/killing of blood-borne PspAϪ pneumococci is slower Ϫ Ϫ Low expression of CR1/2 and high anti-pneumococcal Ab titers in CR1/2 / mice, leading to fatal infection in CD18Ϫ/Ϫ mice For up to 6 h postinfection, no significant difference in blood clear- Ϫ Ϫ ␤ CD18 is the common -chain shared by CR3, CR4, and LFA-1; Downloaded from ance/killing of WU2 was observed in CR1/2 / mice compared F thus, deficiency of CD18 results in a lack of all three functional with WT mice (Fig. 1B, ); however, the clearance of JY1119 was Ϫ/Ϫ Ϫ/Ϫ molecules; yet CD18 mice resisted infection. A compensatory significantly delayed in CR1/2 compared with WT mice (Fig. Ϫ/Ϫ E Ϫ increase in the expression of CR1/2 in CD18 mice could ex- 1B, ). Thus, at 4 h postinfection, 10-fold more PspA pneumo- Ϫ/Ϫ Ϫ Ϫ plain this seeming paradox. However, CD18 mice had Ͼ2-fold cocci were recovered from the blood of CR1/2 / mice than from less expression of CR1/2 on their PBL than WT mice (Fig. 4A, contemporaneous samples from WT mice. In WU2-infected CR1/ Ϫ/Ϫ compare bottom left and bottom right panels). Alternatively, a 2 mice, the 10-day survival rate (12%; one of eight) was lower http://www.jimmunol.org/ compensatory heightened level of naturally occurring Abs in than that in WU2-infected WT (42%; five of 12), although median CD18Ϫ/Ϫ could also explain our observation. Indeed, we observed longevity was not significantly different (Fig. 2, left panel). All that CD18Ϫ/Ϫ did have much higher levels of Igs reactive with WT mice challenged with JY1119 survived infection, but only Ϫ Ϫ heat-killed pneumococci (Fig. 4B, left panel). It is therefore likely 20% (one of five) of CR1/2 / mice did ( p ϭ 0.018; Fig. 2, right that bacteremia is lessened, and survival rate is unaffected in panel). These data reveal that, like FB and FD, CR1/2 is required CD18Ϫ/Ϫ mice compared with WT mice, because they make for full protection from pneumococci. Ϫ Ϫ/Ϫ higher than normal levels of protective anti-pneumococcal Abs. Clearance/killing of blood-borne PspA pneumococci in CR3 Ϫ/Ϫ Ϫ Ϫ Likewise, although CR1/2 mice had normal expression of CR3 or CR4 / mice is slower, but does not lead to fatal infection

Ϫ Ϫ (Fig. 4A) and WT-like levels of anti-pneumococcal Abs (Fig. 4B, by guest on September 28, 2021 The residence time of WU2 in the bloodstream of CR3 / and Ϫ Ϫ bottom left panel), their lower levels of anti-phosphocholine Abs CR4 / mice was not substantially different from that observed in (Fig. 4B, lower right panel) might account for their extreme WT mice; bacteremia was fairly constant throughout the experi- susceptibility. ment in all cases (Fig. 1, C and D, F). Despite this, both kinds of receptor-deficient mice showed increased death rates; 88% of Ϫ Ϫ Ϫ Ϫ Degradation of C3 on the pneumococcal surface CR3 / and 75% of CR4 / died within 72 h compared with 58% of WT mice (Fig. 2, left panel). The median longevity of CR3Ϫ/Ϫ Because CR1/2, CR3, and CR4 have variable affinities for similar and CR4Ϫ/Ϫ mice was significantly reduced compared with that of C3-derived ligands, we sought to determine which C3-derived WT mice ( p ϭ 0.003 for CR3Ϫ/Ϫ; p ϭ 0.007 for CR4Ϫ/Ϫ). In fragments are present on the surface of serum-opsonized pneumo- contrast to WU2, clearance/killing of JY1119 was significantly cocci. To this end, WU2 and JY1119 were incubated in 10% nor- slower in both CR3Ϫ/Ϫ and CR4Ϫ/Ϫ compared with WT mice; mal serum to allow opsonization to proceed. The incubation was thus, at 4 h postinfection, 100-fold more PspAϪ pneumococci terminated at different time points by the addition of cold PBS were recovered from CR3Ϫ/Ϫ than WT mice (Fig. 1C), and 10-fold buffer containing EDTA. C3 retained on the surface of the opso- more was recovered from CR3Ϫ/Ϫ (Fig. 1D). However, unlike nized bacteria was then revealed by immunoblot analysis (Fig. CR1/2Ϫ/Ϫ mice, all CR3Ϫ/Ϫ and all CR4Ϫ/Ϫ mice survived infec- 5A). Based on comparison with purified human C3, C3b, and iC3b ␣ controls and the presence of the 2 fragment (Fig. 5B), the major form of C3 deposited on WU2 and JY1119 appears to be iC3b, although C3b was also present in some samples. As incubation time was increased, iC3b accumulated on both types of bacteria (Fig. 5A). Importantly, however, at each time point comparatively more iC3b was present on JY1119 than on WU2. With the Bio- ␣ Rad gel documentation system, we measured the densities of 2 and ␤ subunits in each sample. At 5, 30, and 60 min, the ␤-chain signal was 1.44-, 1.47-, and 1.32-fold greater, respectively, for FIGURE 3. C3 deposition on pneumococci in the presence of WT and Ϫ/Ϫ JY1119 than for WU2. This indicates that deposition of C3 per se FD mouse sera. Pneumococci were incubated in 30% pooled WT or Ϫ Ϫ/Ϫ was consistently greater for the PspA strain. Contemporaneously, FD serum at 37°C for 30 min, and C3 deposition was detected by flow ␣ cytometry using FITC-conjugated goat anti-mouse C3 IgG. For WU2 in- the 2 fragment was 2.03-, 1.91-, and 1.09-fold more intense, re- cubated with WT serum, FDϪ/Ϫ serum, and buffer (no serum), the mean spectively, for JY1119 than for WU2. This indicates that process- fluorescence intensities achieved were 71.2, 36.8, and 4.1 U, respectively. ing of deposited C3 is more rapid on JY1119 than on WU2. Based Ϫ/Ϫ ␣ ␤ For JY1119 incubated with WT serum, FD serum, or buffer, the mean on the 2/ ratio, processing of C3b to iC3b was estimated to be fluorescence intensities were 168.9, 64.5, and 3.6, respectively. 1.4- and 1.3-fold greater for JY1119 than WU2 at 5 and 30 min, 7510 PspA INHIBITS COMPLEMENT AND COMPLEMENT RECEPTORS

FIGURE 4. A, Surface expression of CR3 (top row) and CR1/2 (bottom row) on peripheral blood cells of WT, CR1/2Ϫ/Ϫ, and CD18Ϫ/Ϫ mice. PBL were stained with anti-CD11b (anti-CR3) mAb or anti-CD21/35 (anti-CR1/2) mAb (shaded areas) or were left unstained (solid line). The data shown are representative of results obtained from groups of three to five mice. B, Serum Abs to heat-killed pneumococci and phosphocholine Ag. Serum was col- Downloaded from lected from WT, CR1/2Ϫ/Ϫ, and CD18Ϫ/Ϫ mice (three to five mice per group), and Ab titers (IgG plus IgM) determined by ELISA. The results are presented as the mean Ϯ SE for each titration. http://www.jimmunol.org/ by guest on September 28, 2021 respectively, but by 60 min, the advantage was negated. The dif- cause the latter pathway amplifies C3 deposition initiated by the ference in the amount of iC3b deposited onto the two strains of former. pneumococci was more obvious when mouse, rather than human, Immunoblots provided indirect evidence that one of the major serum was used as the complement source, but the reaction was forms of C3 on opsonized WU2 and JY1119 is iC3b, a fragment slower. These findings are consistent with the proposal that more known to serve as an efficient ligand for CR3, CR4, and perhaps C3 is deposited on JY1119 than WU2, the majority of C3 depos- also CR1/2. For mice infected with WU2, the strain on which C3 ited is iC3b, and the rate of conversion of C3b to iC3b is increased deposition and blood clearance/killing is limited by the presence of on JY1119. The fact that similar observations were made using PspA, deficiency of any one of these CRs is thus sufficient to mouse vs human serum suggests that this has clinical relevance. increase the mortality rates of mice. Notably, the negative impact on survival rate and longevity achieves statistical significance in Discussion CR3Ϫ/Ϫ and CR4Ϫ/Ϫ mice, an outcome consistent with the re- In this study and previous ones (4–7), we showed that substantially ϩ ported high affinity binding of iC3b to these receptors. In contrast, less C3 is deposited onto the PspA S. pneumoniae strain WU2 in mice challenged with JY1119, to which more C3 binds and than onto its isogenic PspAϪ counterpart JY1119 regardless of the blood clearance/killing is more efficient because PspA is absent, serum source, and that WT mouse serum supports more deposition Ϫ/Ϫ Ϫ/Ϫ deficiency of CR3 or CR4 did not impact survival, but deficiency of C3 onto the pneumococcal surface than FB (7) or FD Ϫ Ϫ of CR1/2 significantly decreased it. Furthermore, CR1/2 / , serum (Fig. 3). This decrease in deposition of C3 in vitro correlates Ϫ/Ϫ Ϫ/Ϫ Ϫ Ϫ CR3 , and CR4 mice all showed retarded removal of with impaired host resistance to infection. Thus, in FD / mice (Figs. 1 and 2), JY1119 becomes as virulent and as lethal as WU2 JY1119 from the bloodstream. is in normal mice. Together with observations made previously by CR1/2 also plays an important role in the humoral immune re- others (27), our initial findings (5), and the new data we report in sponse. Haas et al. (19) reported that serum IgM, IgG1, IgG2b, and Ϫ/Ϫ this study clearly indicate that in mice an intact alternative path- IgG3 are all significantly reduced in CR1/2 mice, whereas way is vital for proper host defense against S. pneumoniae infec- IgG2a and IgA are normal. In our hands, the titer of anti- Ϫ/Ϫ tion, and that PspA impairs this pathway of protection. Recently, pneumococcal Abs in CR1/2 did not differ from WT mice, Brown et al. (28) argued that the classical pathway is the dominant whereas the titer of anti-phosphocholine Abs was indeed substan- pathway for host innate immunity against pneumococcal infection tially lower. Anti-phosphocholine Abs are known to be protective Ϫ Ϫ in mice. Our current findings are not at odds with this proposition. against pneumococcal infection (29, 30); thus, CR1/2 / mice are Indeed, the classical and alternative pathways are probably both probably highly susceptible to the normally avirulent PspAϪ pneu- required for full protection against pneumococcal infection, be- mococci, because, on the one hand, they lack CR1/2 and, on the The Journal of Immunology 7511 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 5. A, Western blot analysis of C3 deposition on PspAϩ (WU2) and PspAϪ JY1119 (JY) pneumococci. A, Pneumococci were opsonized with human serum (left panel) or mouse serum (right panel) for the times indicated, washed with PBS, boiled in reducing buffer, and electrophoresed, then transferred to nitrocellulose. Membranes were developed with anti-human or anti-mouse C3 Abs as required. Purified human (Hu) C3, C3b, and iC3b were included as controls. The identity of each chain or degraded fragment of C3 is indicated. B, Schematic view of C3 processing during complement activation. cleavage of the ␣-chain of C3 by the cofactor-dependent action of ,(ء) After covalent attachment of C3b to pneumococci via the ␣-chain thioester bond factor I (FI) generates iC3b, which remains bound to the opsonized bacterium. other, they have reduced titers of this (and perhaps other) protec- PspAϩ or PspAϪ pneumococci. The fact that CR3Ϫ/Ϫ and tive Abs. CR1 probably also interacts with C1q and MBL (12–14), CR4Ϫ/Ϫ mice express normal amounts of CR1/2 (data not shown) and CR1 and Fc receptors exert synergetic effects to enhance partially explains why these animals are still resistant to JY1119. phagocytosis (31), so the host defense role of CR1/2 in mice in- LFA1Ϫ/Ϫ and CD18Ϫ/Ϫ mice cleared WU2 somewhat more rap- fected with JY1119 probably has additional complexity and im- idly than WT, and the survival time of WU2- and JY1119-infected portance. The relative importance of factor H-mediated clearance animals was not adversely affected by deficiency of LFA-1. This of opsonized bacteria by platelets, which presumably direct uptake difference between our observations and those of others might be of the organism by the liver and spleen (32), also remains to be due to the different outcome parameters and routes of infection determined. used; i.e., we focused on blood clearance/killing after i.v. infec- Previous studies showed that CD18-deficient animals have in- tion, whereas others used an i.p. infection model to emphasize the creased spontaneous infections and earlier mortality compared migration ability of phagocytes into the sites of infection. More- with WT mice (33). Others reported that LFA-1Ϫ/Ϫ mice have over, in an earlier i.p. experiment a serotype 6 pneumococcal strain increased mortality and a higher incidence of otitis media. These was used (34), whereas we used a serotype 3 strain. Notably, LFA- mice are more likely to develop meningitis after being challenged 1Ϫ/Ϫ animals reportedly have a higher number of neutrophils in with S. pneumoniae i.p., and the LFA-1-ligand interaction is the the peripheral blood than WT (33, 34), which could also account stimulatory signal for neutrophils to express full activation. In our for their improved resistance to i.v. infection. hands, there was no significant difference in mortality between The seeming paradox that CD18Ϫ/Ϫ mice remain fully resistant CR3Ϫ/Ϫ, CR4Ϫ/Ϫ, and CD18Ϫ/Ϫ mice after infection with either can be explained. CD18Ϫ/Ϫ mice lack CR3 and CR4 and have 7512 PspA INHIBITS COMPLEMENT AND COMPLEMENT RECEPTORS lower expression of CR1/2, which tends to make them more sus- 12. Birmingham, D. J., and L. A. Hebert. 2001. CR1 and CR1-like: the primate ceptible. In contrast, these mice are predicted to have neutrophilia immune adherence receptors. Immunol. Rev. 180:100. 13. Krych-Goldberg, M., and J. P. Atkinson. 2001. Structure-function relationships associated with their lack of LFA-1 (33, 34), and we showed that of complement receptor type 1. Immunol. Rev. 180:112. they express elevated levels of anti-pneumococcal Abs. We pro- 14. Klickstein, L. B., S. F. Barbashov, T. Liu, R. M. Jack, and A. Nicholson-Weller. Ϫ/Ϫ 1997. Complement receptor type 1 (CR1, CD35) is a receptor for C1q. Immunity pose that in CD18 mice, the combined benefits of neutrophilia 7:345. and high protective Ab titers overcomes the detrimental effects of 15. Kalli, K. R., and D. T. Fearon. 1994. Binding of C3b and C4b by the CR1-like low CR1/2 expression and the absence of CR3 and CR4. site in murine CR1. J. Immunol. 152:2899. 16. Molina, H., T. Kinoshita, C. B. Webster, and V. M. Holers. 1994. Analysis of As in this study, others have infected mutant mice with S. pneu- C3b/C3d binding sites and factor I cofactor regions within mouse complement moniae to study the contributions to protection made by CR1/2, receptors 1 and 2. J. Immunol. 153:789. CR3, CD18, and LFA-1 (19, 33, 34). These earlier studies focused 17. Hourcade, D., M. K. Liszewski, M. Krych-Goldberg, and J. P. Atkinson. 2000. Functional domains, structural variations and pathogen interactions of MCP, on the impact of CR1/2 deficiency on the development of protec- DAF and CR1. Immunopharmacology 49:103. tive adaptive immunity or on the effects of CD18, LFA-1, and CR3 18. Fearon, D. T., and M. C. Carroll. 2000. Regulation of B lymphocyte responses to on long term survival after i.p. inoculation of pneumococci. No- foreign and self-antigens by the CD19/CD21 complex. Annu. Rev. Immunol. 18:393. tably, our study is the first comprehensive investigation of the con- 19. Haas, K. M., M. Hasegawa, D. A. Steeber, J. C. Poe, M. D. Zabel, C. B. Bock, tribution of these receptors to protection against PspAϩ and D. R. Karp, D. E. Briles, J. H. Weis, and T. F. Tedder. 2002. Complement Ϫ receptors CD21/35 link innate and protective immunity during Streptococcus PspA pneumococcal strains and to date is the only one to study pneumoniae infection by regulating IgG3 antibody responses. Immunity 17:713. the impact of CR4 deletion on host defense. In summation, we 20. Harris, E. S., T. M. McIntyre, S. M. Prescott, and G. A. Zimmerman. 2000. The conclude that PspA confers virulence onto S. pneumoniae mainly leukocyte integrins. J. Biol. Chem. 275:23409. 21. Ahearn, J. M., M. B. Fischer, D. Croix, S. Goerg, M. Ma, J. Xia, X. Zhou, Downloaded from by interfering with the protective role of the alternative pathway of R. G. Howard, T. L. Rothstein, and M. C. Carroll. 1996. Disruption of the Cr2 complement activation, thus minimizing C3b deposition and slow- locus results in a reduction in B-1a cells and in an impaired B cell response to ing iC3b generation. The pneumococcus is thereby shielded from T-dependent antigen. Immunity 4:251. 22. Ding, Z. M., J. E. Babensee, S. I. Simon, H. Lu, J. L. Perrard, D. C. Bullard, host CR-mediated pathways of host defense. X. Y. Dai, S. K. Bromley, M. L. Dustin, M. L. Entman, et al. 1999. Relative contribution of LFA-1 and Mac-1 to neutrophil adhesion and migration. J. Im- Acknowledgments munol. 163:5029. 23. Berlin-Rufenach, C., F. Otto, M. Mathies, J. Westermann, M. J. Owen,

We thank Drs. Gunnar Lindahl and Susan K. Hollingshead for reviewing http://www.jimmunol.org/ A. Hamann, and N. Hogg. 1999. Lymphocyte migration in lymphocyte function- our work and for their many helpful suggestions. associated antigen (LFA)-1-deﬁcient mice. J. Exp. Med. 189:1467. 24. Briles, D. E., C. Forman, S. Hudak, and J. L. Claﬂin. 1982. Anti-PC antibodies References of the T15 idiotype are optimally protective against Streptococcus pneumoniae. 1. Briles, D. E., S. K. Hollingshead, E. Swiatlo, A. Brooks-Walter, A. Szalai, J. Exp. Med. 156:1177. A. Virolainen, L. S. McDaniel, K. A. Benton, P. White, K. Prellner, et al. 1997. 25. Briles, D. E., C. Forman, J. C. Horowitz, J. E. Volanakis, W. H. Benjamin, Jr., PspA and PspC: their potential for use as pneumococcal vaccines. Microb. Drug L. S. McDaniel, J. Eldridge, and J. Brooks. 1989. Antipneumococcal effects of Resist. 3:401. C-reactive protein and monoclonal antibodies to pneumococcal cell wall and 2. Briles, D. E., J. Yother, and L. S. McDaniel. 1988. Role of pneumococcal surface capsular antigens. Infect. Immun. 57:1457. protein A in the virulence of Streptococcus pneumoniae. Rev. Infect. Dis. 26. Yother, J., and D. E. Briles. 1992. Structural properties and evolutionary rela- 10(Suppl. 2):S372. tionships of PspA, a surface protein of Streptococcus pneumoniae, as revealed by sequence analysis. J. Bacteriol. 174:601.

3. McDaniel, L. S., J. Yother, M. Vijayakumar, L. McGarry, W. R. Guild, and by guest on September 28, 2021 D. E. Briles. 1987. Use of insertional inactivation to facilitate studies of biolog- 27. Hosea, S. W., E. J. Brown, and M. M. Frank. 1980. The critical role of comple- ical properties of pneumococcal surface protein A (PspA). J. Exp. Med. 165:381. ment in experimental pneumococcal sepsis. J. Infect. Dis. 142:903. 4. Ren, B., A. J. Szalai, O. Thomas, S. K. Hollingshead, and D. E. Briles. 2003. 28. Brown, J. S., T. Hussell, S. M. Gilliland, D. W. Holden, J. C. Paton, Both family 1 and family 2 PspA proteins can inhibit complement deposition and M. R. Ehrenstein, M. J. Walport, and M. Botto. 2002. The classical pathway is the confer virulence to a capsular serotype 3 strain of Streptococcus pneumoniae. dominant complement pathway required for innate immunity to Streptococcus Infect. Immun. 71:75. pneumoniae infection in mice. Proc. Natl. Acad. Sci. USA 99:16969. 5. Tu, A. H., R. L. Fulgham, M. A. McCrory, D. E. Briles, and A. J. Szalai. 1999. 29. Briles, D. E., C. Forman, and M. Crain. 1992. Mouse antibody to phosphocholine Pneumococcal surface protein A inhibits complement activation by Streptococcus can protect mice from infection with mouse-virulent human isolates of Strepto- pneumoniae. Infect. Immun. 67:4720. coccus pneumoniae. Infect. Immun. 60:1957. 6. Abeyta, M. 1999. Pneumococcal surface protein A and capsular polysaccharide 30. Yother, J., C. Forman, B. M. Gray, and D. E. Briles. 1982. Protection of mice in virulence of Streptococcus pneumoniae. Ph D. thesis. University of Alabama, from infection with Streptococcus pneumoniae by anti-phosphocholine antibody. Birmingham, AL. Infect. Immun. 36:184. 7. Ren, B., A. J. Szalai, S. K. Hollingshead, and D. E. Briles. 2004. Effects of PspA 31. Waytes, A. T., A. Malbran, D. A. Bobak, and L. F. Fries. 1991. Pre-ligation of and antibodies to PspA on activation and deposition of complement on the pneu- CR1 enhances IgG-dependent phagocytosis by cultured human monocytes. J. Im- mococcal surface. Infect. Immun. 72:114. munol. 146:2694. 8. Prodinger, W. M., R. Wurzner, A. Erdei, and M. P. Dierich. 1999. Complement. 32. Alexander, J. J., B. K. Hack, P. N. Cunningham, and R. J. Quigg. 2001. A protein In Fundamental Immunology. W. E. Paul, ed. Lippincott-Raven, Philadelphia, with characteristics of factor H is present on rodent platelets and functions as the p. 967. immune adherence receptor. J. Biol. Chem. 276:32129. 9. Law, S. K., and A. W. Dodds. 1997. The internal thioester and the covalent 33. Scharffetter-Kochanek, K., H. Lu, K. Norman, N. van Nood, F. Munoz, binding properties of the complement proteins C3 and C4. Protein Sci. 6:263. S. Grabbe, M. McArthur, I. Lorenzo, S. Kaplan, K. Ley, et al. 1998. Spontaneous 10. Agramonte-Hevia, J., A. Gonzalez-Arenas, D. Barrera, and M. Velasco-Velazquez. skin ulceration and defective T cell function in CD18 null mice. J. Exp. Med. 2002. Gram-negative bacteria and phagocytic cell interaction mediated by comple- 188:119. ment receptor 3. FEMS Immunol. Med. Microbiol. 34:255. 34. Prince, J. E., C. F. Brayton, M. C. Fossett, J. A. Durand, S. L. Kaplan, 11. Wright, S. D. 1992. Receptors for complement and the biology of phagocytosis. C. W. Smith, and C. M. Ballantyne. 2001. The differential roles of LFA-1 and In Basic Principles and Clinical Correlates. J. I. Gallin, I. M. Goldstein, and Mac-1 in host defense against systemic infection with Streptococcus pneumoniae. R. Snyderman, eds. Raves Press, New York, p. 477. J. Immunol. 166:7362.