Pneumococcal Interaction with Human Dendritic Cells: Phagocytosis, Survival, and Induced Adaptive Immune Response Are Manipulated by PavA This information is current as of September 27, 2021. Nadja Noske, Ulrike Kämmerer, Manfred Rohde and Sven Hammerschmidt J Immunol 2009; 183:1952-1963; Prepublished online 1 July 2009; doi: 10.4049/jimmunol.0804383 Downloaded from http://www.jimmunol.org/content/183/3/1952

References This article cites 52 articles, 24 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/183/3/1952.full#ref-list-1

Why The JI? Submit online.

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

• No Triage! Every submission reviewed by practicing scientists by guest on September 27, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of 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 © 2009 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Pneumococcal Interaction with Human Dendritic Cells: Phagocytosis, Survival, and Induced Adaptive Immune Response Are Manipulated by PavA1

Nadja Noske,*†‡ Ulrike Ka¨mmerer,§ Manfred Rohde,¶ and Sven Hammerschmidt2*†‡

Dendritic cells (DCs) ingest and process for presenting their Ags to T cells. PavA (pneumococcal adherence and virulence factor A) is a key virulence determinant of pneumococci under in vivo conditions and was shown to modulate adherence of pneumococci to a variety of nonprofessional phagocytic host cells. Here, we demonstrated the role of PavA for the interaction of human DCs with live pneumococci and analyzed the induced host responses upon ingestion of viable pneumococci. Expression of PavA protected pneumococci against recognition and actin cytoskeleton-dependent phagocytosis by DCs compared with iso- genic pavA mutants. A major proportion of internalized pneumococci were found in membrane-bound phagosomes. Pneumococcal Downloaded from phagocytosis promotes maturation of DCs, and both wild-type pneumococci and PavA-deficient pneumococci triggered production of proinflammatory such as IL-1␤, IL-6, IL-8, IL-12, and TNF-␣ and antiinflammatory IL-10. However, production was delayed and reduced when DCs encounter pneumococci lacking PavA, which also results in a less efficient activation of the adaptive immune response. Strikingly, purified PavA reassociates to pneumococci but not DCs and reduced phagocytosis of the pavA mutant to levels similar to those of wild-type pneumococci. Additionally, pavA mutants covered with

exogenously provided PavA induced a DC cytokine profile similar to wild-type pneumococci. In conclusion, these results http://www.jimmunol.org/ suggest that PavA is key factor for live pneumococci to escape phagocytosis and to induce optimal cytokine productions by DCs and adaptive immune responses as well. The Journal of Immunology, 2009, 183: 1952–1963.

treptococcus pneumoniae (pneumococci) are commensals nation of immune defenses upon stimulation in response to micro- of the human respiratory tract and colonize up to 70% of bial signals (5). Immature DCs efficiently phagocytose or macropi- S the individuals without causing clinical symptoms. How- nocytose bacteria and process them into cell surface-presentable ever, these apparently harmless colonizers are also well known as Ags. During this process, the immature DCs mature and convert serious human that transmigrate into the lungs, enter the into potent APCs. Maturation of DCs is characterized by changes bloodstream, and cross the blood-brain barrier (1). As a conse- in surface expression of MHC, adhesion and costimulatory mole- by guest on September 27, 2021 quence, this versatile causes infections ranging from cules, and cytokine production as well (6). Upon maturation, DCs severe local infections, such as otitis media and sinusitis, to migrate from the place of Ag uptake into tissue-draining lymphoid life-threatening infections, including pneumonia, sepsis, and organs such as lymph nodes or spleen. Maturation of DCs dimin- meningitis (2). In healthy individuals, the mucosal surfaces with ishes their capacity to internalize and process Ags but greatly en- their epithelial cells and the secreted mucus constitute a physical hances their ability to prime naive T cells. As a consequence, DC barrier that prevents pathogens to gain access into deeper tissues. responses upon bacterial infections initiate adaptive immune re- Here, the pathogens are also faced by the production of antimi- sponses (7). However, pathogens have evolved various strategies crobial agents, such as defensins (3, 4). Additionally, mucosal tis- to escape host immune responses. Regarding DCs, their subversion sues are scattered with sentinel professional and APCs by pathogens and exploitation as a Trojan horse to disseminate including dendritic cells (DCs).3 DCs are involved in the coordi- within the host—as recently shown for HIV and Chlamydia—is under debate (8–10). Pneumococci are encased by a capsular polysaccharide (CPS), *Department of Microorganisms, Institute for Genetics and Functional Genomics, University of Greifswald, Greifswald, Germany; †Max von Pettenkofer and the CPS is recognized as a sine qua non for invasive diseases Institute, Ludwig-Maximilians University Mu¨nchen, Mu¨nchen, Germany; ‡Research (11). The capsule protects pneumococci against uptake into pro- Center for Infectious Diseases and §Department of Obstetrics and Gynecology, Uni- fessional phagocytes and complement-mediated opsonophagocy- versity of Wu¨rzburg, Wu¨rzburg, Germany; and ¶Department of Microbial Pathogen- esis, Helmholtz Center for Infection Research, Braunschweig, Germany tosis (12, 13). Another potent virulence factor of pneumococci Received for publication December 31, 2008. Accepted for publication May 28, 2009. interfering with both eukaryotic cell function and the immune sys- tem is the pore-forming cytolysin pneumolysin (14, 15). TLR4 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 recognizes pneumolysin as a pathogen-associated molecular pat- with 18 U.S.C. Section 1734 solely to indicate this fact. tern, thereby providing protection against pneumococcal infections 1 This work was supported in part by the Bundesministerium fu¨r Bildung und For- (16). Moreover, pneumolysin production enhances mucosal clear- schung (CAPNETZ C8; to S.H.) and the Deutsche Forschungsgemeinschaft (Sonder- ance of pneumococci by stimulating recruitment and forschungsbereich 479, Teilprojekt A7, DFG HA 3125/2-1 and 4-1). 2 Address correspondence and reprint requests to Dr. Sven Hammerschmidt, De- partment Genetics of Microorganisms, Institute for Genetics and Functional charide; FESEM, field emission electron microscopy; iDC, immature ; Genomics, Ernst Moritz Arndt University Greifswald, Friedrich Ludwig Jahn Lamp1, lysosomal-associated membrane protein 1; MBP, maltose-binding protein; Strasse 15a, D-17487 Greifswald, Germany. E-mail address: sven.hammerschmidt@ MOI, multiplicity of infection; OxMi, oxidative mitogenesis; PavA, pneumococcal uni-greifswald.de adherence and virulence factor A; TEM, transmission electron microscopy. 3 Abbreviations used in this paper: DC, dendritic cell; BMDC, bone marrow-derived dendritic cell; CLSM, confocal laser scanning microscopy; CPS, capsular polysac- Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.0804383 The Journal of Immunology 1953 promotes bacterial Ag delivery to the nasal-associated lymphoid mary Abs used for DCs were: human CD11c (IgG1), CD25 (IgG1), CD83 tissue (17, 18). The cell wall of pneumococci is decorated with (IgG2), CD86 (IgG2), or HLA-DR (IgG1) for MHC-class II. Nonspecific numerous virulence factors that execute their function in different binding of Abs was calculated by incubating DCs with the appropriate anti-IgG2 or anti-IgG1 (Caltag Laboratories, BD Pharmingen, or host niches (19, 20). Remarkably, atypical surface , lack- Miltenyi Biotec). The Abs were conjugated with PE or FITC. Surface ing a leader peptide or even a motif for membrane anchoring such marker expression was quantified using a FACSCalibur (BD Biosciences). as the PavA protein (pneumococcal adherence and virulence factor Infection of human DCs A), have been found to be key virulence factors of S. pneumoniae. Deficiency in PavA impairs adherence of pneumococci to host For our antibiotic protection assays DCs were seeded into 96-well cell ϫ 5 epithelial and endothelial cells and attenuates virulence of pneu- culture plates (Greiner Bio-One) at a density of 1 10 cells/well. To perform immunofluorescence staining or electron microscopy, DCs were mococci in mice infection models (21, 22). Although pneumococci seeded on glass coverslips (diameter, 12 mm) in 24-well cell culture plates lacking PavA produce a capsule and their cytotoxin pneumolysin, (Greiner Bio-One). For flow cytometric analysis and transmission electron these mutants were massively attenuated in a mouse sepsis model microscopy the cells were used in 6-well cell culture plates (Greiner Bio- and, additionally, intracranial infections of mice resulted in rapid One) at a density of 5 ϫ 105 cells/ml. The attachment of DCs to the surface of the dishes occurred within1hat37°C under 5% CO . The prepared DCs clearance from the CNS (21, 22). It is suggested that PavA func- 2 were infected with a multiplicity of infection (MOI) of 50 pneumococci per tion is not directly involved in adhesion or virulence, but is rather cell in RPMI 1640 supplemented with 1% FBS and IL-4/GM-CSF at 37°C modulating other, yet unidentified, important virulence determi- under 5% CO2. Synchronization of the infections was achieved by a cen- nants of S. pneumoniae (22). trifugation step at 100 ϫ g for 4 min. In naive mice immature bone marrow-derived myeloid DCs Quantification of pneumococcal phagocytosis by plating and (BMDCs) pulsed with heat-killed pneumococci stimulated hu- immune fluorescence microscopy Downloaded from moral responses specific for the surface proteins PspA and PsaA, and the surface-exposed polysaccharides phosphorylcholine and The number of recovered viable intracellular pneumococci after phagocy- tosis by DCs was quantified by the antibiotic protection assay as described CPS as well (23). Pneumococci induce apoptosis in BMDCs previously for nonprofessional host cells (26, 29). Briefly, DCs were through two mechanistically distinct pathways. Pneumolysin in- washed thoroughly with PBS after infection (in a standardized assay of 30 duces a rapid and caspase-independent apoptosis, while in a de- min) to remove unbound bacteria. To kill extracellular bacteria, DCs were

layed onset and associated with maturation of DCs, a caspase- then incubated with RPMI 1640, 1% FBS containing 100 IU penicillin G, http://www.jimmunol.org/ and 100 ␮g of gentamicin for 1 h at 37°C under 5% CO . Intracellular dependent apoptosis is induced that does not require uptake of 2 viable bacteria were recovered after washing with PBS by a saponin-me- bacteria (24). diated (1% (w/v)) for 10 min at 37°C. The amount of released pneu- However, the mechanisms of recognition, ingestion, and intra- mococci per well were enumerated by plating serial dilutions on blood agar cellular fate of live pneumococci by DCs and the contribution of plates. For immune fluorescence microscopy pneumococci-infected DCs specific bacterial components of pneumococci on induced adaptive were fixed with 3.7% paraformaldehyde. To distinguish between adherent (extracellular) and intracellular (phagocytosed) pneumococci the bacteria immune responses have not been studied in great detail. In the were labeled with Alexa Fluor 488 and/or Alexa Fluor 568 (Invitrogen) as present study we demonstrate the key role of PavA for internal- previously reported in detail (26, 30). Image acquisition of representative ization and processing of viable pneumococci by immature human experiments was performed with confocal laser scanning microscopy blood-derived DCs. Additionally, we show that the presence of (CLSM; Leica TCS SP5) and the CLSM software. by guest on September 27, 2021 pneumococcal PavA is required to stimulate optimal cytokine pro- Electron microscopy duction by DCs and adaptive immune responses. Preparations of samples for field emission electron microscopy (FESEM) and transmission electron microscopy (TEM) were performed identical to Materials and Methods the procedures described recently (31). Bacterial strains, culture conditions, and protein purification Real-time RT-PCR S. pneumoniae strains ATCC 11733 (serotype 2), R800 (nonencapsulated), D39 (serotype 2), D39⌬cps, NCTC 10319 (serotype 35A), and the pneu- RNA of DCs was purified with the RNAeasy purification kit and DNA was molysin (ply) as well as the PavA-deficient mutants D39⌬pavA, S.p. eliminated by RNase-Free DNase treatment (Qiagen). Equal amounts of 35A⌬ply, S.p. 35A⌬pavA, and S.p. 35A⌬ply⌬pavA, respectively, were re- purified RNA (calculated using the NanoDrop ND-1000), the SuperScript III reverse transcriptase (Invitrogen), and oligo(dT) primers (Invitrogen) ported recently (22, 25, 26). Bacteria were grown to mid-log phase (OD600 of 0.3) in Todd-Hewitt broth (Oxoid) supplemented with 0.5% yeast ex- were used to synthesize cDNA. PCR reactions were done with IL-specific tract (THY) and antibiotics when appropriate, or cultured on blood agar. oligonucleotides and hsRPS9 primers (Table I), and real-time PCR was Full-length pavA (derived from R800) was subcloned from clone performed on an ABI PRISM 7000 (Applied Biosystems) using SYBR UB1155 (21) by digestion with BamHI and PstI into the similarly digested Green quantitative PCR (Roche Diagnostics) according to the manufac- expression vector pMALc2x (New England Biolabs). Maltose-binding pro- turers’ instructions. tein (MBP) and MBP-PavA were produced in JM109 and Cytokine assays purified according to the manufacturer’s guidelines by affinity chromatog- raphy using an amylose resin (New England Biolabs) followed by anion Cytokine concentrations in the DC cell culture supernatants were measured exchange column (HiTrap Q HP) purification. with the Luminex technology (BioSource Europe) according to the man- ufacturer’s instructions. A human inflammatory five-plex kit (GM-CSF, Isolation of human DCs from peripheral blood cells and TNF-␣, IL-6, IL-8, and IL-1␤) and beads for IL-12 (IL-12p40 and IL- maturation of DCs 12p70), IL-10, and IFN-␥ were used for quantification (BioSource Europe), and 50 ␮l of a 1/10 dilution was used in the assays. Evaluation of the data Human -derived DCs were isolated and cultured from PBMCs by was performed with the MasterPlexQT software (MiraiBio). a standard protocol (27). were either extracted from human peripheral blood, from buffy coat (PBMC) suspensions (German Red activation assay of oxidative mitogenesis variant (OxMi) Cross, Wiesentheid, Germany), or from an highly enriched monocyte con- of the classical MLR centrate that was obtained after employing a combination of leukapheresis of blood cells and centrifugal elutriation as previously described in detail To obtain strong signals in a nonradioactive detection assay, allogeneic T (28). DC precursors were cultured for 5–8 days in RPMI 1640 medium cells used as reporter cells in the mixed leukocyte reaction were prepared (PAA Laboratories) supplemented with 10% FBS, 2 mM glutamine and following the method of Hill et al. (32). T cells were isolated from PBMCs penicillin/streptomycin (100 IU/ml and 100 ␮g/ml; PAA Laboratories), 20 by resetting with sheep RBCs and washed twice with ice-cold PBS. T cells IU/ml purified recombinant human GM-CSF, and 16 IU/ml IL-4 (Strath- (1 ϫ 107) were then resuspended in 975 ␮l of ice-cold PBS, carefully man Biotech). Cytokines and medium were replaced every second day. mixed with 25 ␮l of filtered sodium periodate (5 ␮g/ml; Sigma-Aldrich; Purity of immature DCs was Ͼ90% as indicated by FACS analyses. Pri- dissolved in ice-cold PBS), and incubated on ice for 20 min. After washing 1954 PavA INFLUENCES PHAGOCYTOSIS AND HOST IMMUNE RESPONSE

Table I. Oligonucleotides used for PCR and real-time PCR

Primer Sequence

hsIL-1␤ 5Ј-CAGGGACAGGATATGGAGCAACAA-3Ј 3Ј-CATCTTTCAACACGCAGGACAGGT-5Ј hsIL-6 5Ј-CACCCCTGACCCAACCACAAAT-3Ј 3Ј-TCCTTAAAGCTGCGCAGAATGAGA-5Ј hsIL-8 5Ј-GACCACACTGCGCCAACACA-3Ј 3Ј-CAGCCCTCTTCAAAAACTTCTCCA-5Ј hsIL-10 5Ј-CCGCCTCAGCCTCCCAAAGT-3Ј 3Ј-CCCTAACCTCATTCCCCAACCAC-5Ј hsIL-12p35 5Ј-GCCGCGGCCACAGGTCT-3Ј 3Ј-GTGGCCACGGGGAGGTTTCT-5Ј hsIL-12p40 5Ј-ATTGAGGTCATGGTGGATGC-3Ј 3Ј-AATGCTGGCATTTTTGCGGC-5Ј hsIFN-␥ 5Ј-TAGCAACAAAAAGAAACGAGATGACT-3Ј 3Ј-CCTTTTTCGCTTCCCTGTTTTAG-5Ј hsTNF-␣ 5Ј-AGGCCAAGCCCTGGTATG AGC-3Ј 3Ј-CACAGGGCAATGATCCCAAAGTAG-5Ј hsRPS9 5Ј-CTTAGGCGCAGACGGGGAAGCG-3Ј 3Ј-CGAAGGGTCTCCGCGGGGTCACAT-5Ј Downloaded from

the cell solution with 4 ml of FBS per 1 ml of cells and two washes in RPMI 1640 with 10% FBS, the treated responder T cells (9 ϫ 104 per well) were cocultured with graded numbers of 9000 to 111 cells per well of DCs for 3 days in U-bottom 96-well plates in 150 ␮l of RPMI 1640/10% FBS.

To determine the proliferation index, the cultures were then incubated with http://www.jimmunol.org/ WST-8 cell counting substrate (Alexis) for 2 h. OD was analyzed at 450 nm using a microplate reader. Six to eight independent assays were per- formed for each condition. PavA protein binding assays Binding of FITC-labeled pneumococci to immobilized purified MBP or MBP-PavA, which were immobilized in wells of a 96-well microtiter plate (polystyrene surface), was measured as described previously (26). Binding of soluble MBP or MBP-PavA to pneumococci or DCs was analyzed by

flow cytometry using the following Abs: polyclonal anti-MBP antiserum by guest on September 27, 2021 (1/1000; New England Biolabs), polyclonal anti-PavA IgG (22) (1/200), and secondary anti-rabbit FITC conjugate (MoBiTec). Statistical analysis FIGURE 1. Internalization of S. pneumoniae by DCs and effects of CPS production. Human monocyte-derived immature DCs (Ͼ95% CD11cϩ All data are reported as mean Ϯ SD unless otherwise noted. Results were cells) were infected for 30 min with a MOI of 50 viable pneumococci per statistically analyzed using the paired two-tailed Student’s test, and a value DC. A, Internalization of encapsulated NCTC 10319 (serotype 35A) and of p Ͻ 0.05 was accepted as indicating significance. ATCC 11733 (serotype 2), D39 (serotype 2), and nonencapsulated R800 and D39⌬cps pneumococci was determined by applying the antibiotic pro- Results tection and quantitative plating. The results are expressed as CFU recov- Impact of the capsular polysaccharide on phagocytosis of ered per 96-well plate (mean Ϯ SD) obtained from triplicate experiments. -p Ͻ 0.05. B, Immunofluorescence microscopy of adherent and internal ,ء pneumococci by human DCs To elucidate the interference of the amount of surface-expressed ized pneumococcal strains after 30 min of DC infection. Pneumococci attached to DCs were labeled with Alexa Fluor 488 and 568 (red and green capsule on phagocytosis by DCs, we studied the phagocytosis of fluorescence, which appears yellow when merged), whereas internalized various pneumococcal strains. A recent study indicated that pneu- pneumococci were labeled only with Alexa Fluor 568 and show only red mococcal strains NCTC 10319 and ATCC 11733 produce lower fluorescence. Bar represents 10 ␮m. amounts of CPS compared with the mouse virulent strain D39 (25). Immature DCs (iDCs) were infected for 30 min with the encapsulated pneumococcal strains ATCC 11733 (serotype 2), by DCs. This pneumococcal strain is further characterized for its NCTC 10319 (serotype 35A), D39 (serotype 2), its nonencapsu- interaction with nonprofessional host cells and produces all known lated mutant D39⌬cps, and with the nonencapsulated strain R800. pneumococcal virulence factors (22, 29, 33). The results of the antibiotic protection assay and double immuno- fluorescence microscopy demonstrated that pneumococcal uptake PavA impedes phagocytosis of pneumococci by immature DCs by DCs negatively correlates with the expression capsular poly- To investigate the effect of PavA on phagocytosis of live pneu- saccharide (Fig. 1). Impaired phagocytosis was observed for D39, mococci by human DCs, infections were conducted with wild- while internalization of nonencapsulated R800 and strains NCTC type S.p. 35A, isogenic pavA mutant S.p. 35A⌬pavA, pneumo- 10319 and ATCC 11733 was highly efficient (Fig. 1). Phagocytosis lysin-deficient mutant S.p. 35A⌬ply, or the double mutant S.p. of the nonencapsulated mutant D39⌬cps was significantly en- 35A⌬ply⌬pavA (21, 22). First, pneumococcal uptake by DCs was hanced compared with its isogenic wild-type D39 (Fig. 1). Due to scored over time by double immunofluorescence microscopy (Fig. a high phagocytosis rate, we selected the pneumococcal strain 2, A and B). In general, the number of intracellular bacteria was NCTC 10319 (referred to as S.p. 35A) to analyze the impact of the clearly dependent on the period of infection and the MOI. After 30 virulence factor PavA on immunomodulation after phagocytosis min a maximum of intracellular bacteria was reached and extended The Journal of Immunology 1955 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 2. Expression of PavA interferes with recognition and phagocytosis of pneumococci by DCs. DCs were infected with a MOI of 50 viable pneumococci of wild-type strain S.p. serotype 2 (ATCC 11733), its isogenic mutant S.p. 35A⌬pavA, S.p. 35A (NCTC 10319), its isogenic mutants S.p. 35A⌬ply and S.p. 35A⌬pavA, or its double mutant S.p. 35A⌬ply⌬pavA. Infections with serotype 2 strains D39 and D39⌬pavA were performed with a MOI of 200 per DC. A, The number of intracellular pneumococci was determined by immunofluorescence microscopy 15, 30, and 45 min after infection of DCs. p Ͻ 0.05. B, Double immunofluorescence microscopy of pneumococci attached ,ء .Results present means Ϯ SD of at least three independent experiments to DCs (yellow (green plus red) fluorescence) and intracellular pneumococci (red fluorescence only) after infecting DCs for 30 min. C and D, Phagocytosis of wild-type pneumococci and their isogenic mutants as determined by the antibiotic protection assay and quantitative plating. The results are expressed p Ͻ 0.05. E, Intracellular fate of pneumococci as determined ,ء .as CFU recovered per 96-well plate (mean Ϯ SD) obtained from triplicate experiments by enumeration of recovered intracellular bacteria. DCs were infected for 30 min, and after killing extracellular bacteria by antibiotic treatment, infections were continued up to the indicated time points to determine the number of internalized and viable pneumococci by quantitative plating. The results are expressed as CFU recovered per 96-well plate (mean Ϯ SD) obtained from triplicate experiments. infections resulted in similar (45 min) or lower numbers (Ͼ60 min; say after infecting the DCs with a MOI of 50 bacteria per cell (Fig. data not shown) of intracellular bacteria (Fig. 2A). When using a 2, C and D). After 30 min of DC infection the rates of internalized MOI of 50 bacteria per DC, which was explored for further anal- and recovered pneumococci were at least 2-fold higher for the ysis, the number of bacteria present in each DC varied between 4 pavA mutant S.p. 35A⌬pavA or S.p. 35A⌬ply⌬pavA compared and 9 after 30 min for the wild-type and ply mutant (data not with the isogenic wild-type or ply mutant S.p. 35A⌬ply (Fig. 2, C shown). Enumeration of intracellular bacteria by immunofluores- and D). Similar to the immunofluorescence microscopic analysis, cence microscopy showed no significant differences between wild- the antibiotic protection assay showed no significant differences type and pneumolysin-negative pneumococci when DCs were in- between wild-type and pneumolysin-deficient bacteria (Fig. 2C). fected for 15, 30, or 45 min (Fig. 2A). In contrast, 30 and 45 min Strikingly, the enhanced ingestion of PavA-deficient pneumococci after infection the number of intracellular PavA- and pneumolysin- was also observed for the serotype 2 strain ATCC 11733 and the deficient pneumococci was significantly higher compared with highly encapsulated and mouse virulent pneumococcal strain D39 wild-type S.p. 35A and S.p. 35A⌬ply (Fig. 2, A and B). These (Fig. 2D). In conclusion, these data suggest that the expression of results were confirmed when applying the antibiotic protection as- PavA impedes uptake of S. pneumoniae by human DCs. Additionally, 1956 PavA INFLUENCES PHAGOCYTOSIS AND HOST IMMUNE RESPONSE

these data revealed that pneumococcal uptake by DCs is not in- fluenced by pneumolysin. To avoid pneumolysin-induced cytolytic or cytotoxic effects, pneumococci devoid of pneumolysin (⌬ply) were used for long-term infection experiments.

The intracellular fate of S. pneumoniae in DCs is not influenced by PavA expression To assess survival of internalized pneumococci postinfection and the role of PavA in this scenario, the intracellular fate of S.p. 35A⌬ply and its isogenic pavA mutant were compared. DCs were infected with pneumococci and 30 min postinfection extracellular pneumococci were killed by antibiotic treatment and eliminated. After removing the antibiotics the viability of intracellular pneu- mococci was investigated for the indicated time points (up to 6 h postinfection) by plating the intracellular bacteria on blood agar plates. Similar to our other experiments, enumeration of CFU on blood agar plates revealed significantly higher numbers of intra- cellular viable bacteria for the pavA mutant at early times points compared with S.p. 35A⌬ply (Fig. 2, C and E). Herein, we were Downloaded from especially interested in the ability of pneumococci to survive within the DC. The results revealed killing of both strains by DCs. Apparently, the time-dependent decrease in recovered bacteria was not significantly different when comparing PavA-deficient pneu- mococci with the PavA-expressing bacteria (Fig. 2E). The major

proportion of phagocytosed bacteria was processed 3 h postinfec- http://www.jimmunol.org/ tion by DCs. In conclusion, these data suggest that PavA expres- sion is important for pneumococci to diminish significantly phago- cytosis by DCs but cannot significantly improve the intracellular fate of pneumococci within phagosomes of DCs.

Pneumococcal uptake in membrane-bound phagosomes of DCs relies on actin cytoskeleton dynamics The influence of the actin cytoskeleton dynamics on pneumococcal phagocytosis by DCs was investigated in the presence of the phar- by guest on September 27, 2021 macological inhibitor cytochalasin D, which inhibits actin poly- merization. In the presence of cytochalasin D, phagocytosis of pneumococci was diminished as determined by enumeration of recovered intracellular pneumococci (Fig. 3A). Additionally, illus- trations by CLSM indicated that adherence was not affected by cytochalasin D treatment of DCs, while this treatment resulted in loss of intracellular pneumococci as indicated by CLSM (Fig. 3B). The uptake and localization of pneumococci was further analyzed by FESEM and TEM (Fig. 3C). Illustrations by FESEM showed the formation of membranous structures that engulf the bacteria during uptake, indicative for the essential role of the host cell actin cytoskeleton dynamics during pneumococcal uptake by DCs (Fig. 3C). Intracellular localization studies by TEM 30 min FIGURE 3. Phagocytosis of pneumococci by human DCs and intracel- lular fate of bacteria as determined by electron microscopy and immuno- postinfection of DCs revealed that the major proportion of fluorescence microscopy. A, DCs were pretreated with cytochalasin D pneumococci is located in membrane-bound phagosomes (Fig. (0.05 mmol), and the number of intracellular pneumococci were deter- 3C). Strikingly, both pneumolysin-deficient strains S.p. 35A⌬ply and p Ͻ 0.05. B, Immunofluores- S.p. 35A⌬ply⌬pavA were also localized in the cytoplasm, and ,ء .mined by the antibiotic protection assay cence microscopy of adherent (yellow) and intracellular (red) pneumococci electron micrographs suggest that pneumococci exit membrane- after infection of cytochalasin D pretreated (CytD) or untreated DCs bound phagosomes independently from pneumolysin produc- (none). C, Electron microscopic visualization of pneumococcal adherence tion (Fig. 3C). CLSM and the use of polyclonal Abs against to and invasion into DCs and intracellular trafficking inside DCs. a and b, Wild-type S.p. 35A (NCTC 10319) and S.p. 35A⌬ply trigger cytoskeletal rearrangements resulting in uptake by membrane ruffles; c and d, ultrathin sections depict S.p.35A⌬ply (c) and the isogenic pavA-mutant S.p. (a and b), 2 ␮m(c and d), 0.5 ␮m(e and f), and 0.1 ␮m in the inset of f. 35A⌬ply⌬pavA (d) inside membrane-bound compartments (indicated by D, Immunofluorescence detection of pneumococci in late endosomes/ arrows), most probably inside phagosomes, after 30 min of infection. As phagosomes. The intracellular Ag Lamp1 was performed after infecting has been demonstrated by confocal microscopy, ultrathin sections also re- human DCs with pneumococci. Lamp1 was used a marker protein for late veal that the isogenic pavA mutant is more invasive compared with S.p. endosomes/lysosomes and labeled with specific anti-Lamp1 Abs (BD 35A⌬ply. e and f and inset of f, After longer infection times both strains are Pharmingen) and secondary Alex Fluor 488-conjugated Abs (green). Pneu- able to exit the membrane-bound compartments (phagosome, phagolyso- mococci were visualized with the Alexa Fluor 568-conjugated Ab (red). some) to reside free inside the cytoplasm of DCs. Bars represent 1 ␮m Arrows indicate pneumococci in late endosomes/phagolysosomes. The Journal of Immunology 1957

FIGURE 4. Profiles of surface marker expression by DCs stimulated with pneumococci. The profile of sur- face marker expression of HLA-DR (MHC class II), CD25, CD83, and CD86 indicative of maturation of CD11cϩ DCs infected with S.p. 35A⌬ply or S.p. 35A⌬ply⌬pavA was measured by flow cytometry. The to- tal fluorescence values were moni- tored after gating the main population of events, substracting the corre- sponding isotype background of the Downloaded from used Ab used to detect the surface marker, and after selecting a thresh- old for the control of untreated human DCs. For each flow cytometric anal- ysis, values were recorded for 10,000 events and the means Ϯ SD of at least three independent experiments are http://www.jimmunol.org/ shown. To demonstrate the expres- sion levels of the markers, the total fluorescence quantity value (geomet- ric mean fluorescence intensity mul- tiplied with the percentage of positive events) was used in A and B, respec- tively. The representative histograms in C show the log fluorescence inten-

sity on the x-axis, and the y-axis by guest on September 27, 2021 shows the numbers of events after 18 h of stimulation. A, DCs were in- fected for 30 min with pneumococci (S.p. 35A⌬ply). After killing the ex- tracellular bacteria by antibiotic treat- ment, the stimulation was continued for the indicated time points (2, 8, and 18 h). B and C, Maturation profiles of DCs infected for 30 min with S.p. 35A⌬ply or its pavA mutant S.p. 35A⌬ply⌬pavA. Bacteria were killed by antibiotic treatment, and analysis of surface marker expression was per- formed after 18 h by flow cytometry.

Lamp1 (lysosomal-associated membrane protein 1), a mem- Immature human DCs, positive for CD11c, were infected for 30 brane marker protein of late endosomes and phagolysosomes, min with live pneumococci (S.p. 35A⌬ply). The changes of mat- confirmed that intracellular pneumococci were mainly located uration-associated surface marker expression of CD25 (IL-2 re- in phagolysosomes (Fig. 3D). ceptor, ␣-chain), DC-specific CD83, costimulatory molecule CD86, and MHC class II (subunit HLA-DR) were measured by Pneumococci induce maturation of DCs independently of PavA flow cytometry. Directly after killing extracellular bacteria by an- expression tibiotic treatment and after further cultivation of treated DCs for 2, To investigate the host cell response after pneumococcal infection, 8, and 18 h in RPMI 1640 medium supplemented with 1% FBS we assessed the expression of surface marker proteins on DCs. and IL-4/GM-CSF, all four surface marker proteins examined were 1958 PavA INFLUENCES PHAGOCYTOSIS AND HOST IMMUNE RESPONSE

FIGURE 5. Quantitative cytokine mRNA levels of DCs infected with pneumococci. DCs were infected for 30 min with S.p. 35A⌬ply or the pavA-deficient iso- genic strain and pulsed in total for 18 h after killing extracellular and viable pneumococci 30 min postinfec- tion. Quantitative fold differences for each transcript were determined using the comparative CT calculation method. A, Cytokine mRNA expression of DCs infected with S.p. 35A⌬ply (filled bars) and unstimulated DCs p Ͻ 0.05. B, Cytokine mRNA ,ء .open bars) is shown) levels of DCs infected with S.p. 35A⌬ply⌬pavA relative to DCs pulsed with S.p. 35A⌬ply. RPS9 served as

control. Downloaded from http://www.jimmunol.org/

significantly up-regulated in a time-dependent manner (Fig. 4A). pneumoniae induced higher levels of immune regulatory cyto- The up-regulation of the MHC class II complex (HLA-DR) was kine mRNA expression in DCs compared with unstimulated prominent, as indicated by the highest total fluorescence intensity, DCs (Fig. 5A). Additionally, the levels of mRNA expression followed by CD86 and CD25, and CD83 (Fig. 4A). To elucidate were significantly reduced for DCs infected with PavA-defi- whether the sole association of pneumococci is sufficient for the cient pneumococci compared with DCs infected with PavA- up-regulation of DC surface markers, DC infections were con- positive S.p. 35A⌬ply bacteria (Fig. 5B). by guest on September 27, 2021 ducted in the presence of cytochalasin D. The results revealed no The up-regulation of the cytokine gene products was further up-regulation of DC-specific marker proteins, suggesting that in- analyzed using the Luminex technology (BioSource Europe). ternalization of pneumococci is a prerequisite for induction of DC The amounts of cytokines in cell culture supernatants of in- maturation (data not shown). Remarkably, no significant differ- fected DCs were measured at indicated time points postinfec- ences in DC maturation were observed when surface marker ex- tion (Fig. 6A). The supernatant of LPS-pulsed DCs (48 h) was pression of DCs infected with pavA mutant was compared after used as a control (data not shown). The concentrations of the ⌬ 18 h to infections with the PavA-producing strain S.p. 35A ply immune regulatory proteins increased in a time-dependent man- (Fig. 4, B and C). In conclusion, uptake of pneumococci induced ner, and after 18 and 48 h postinfection, we measured also the maturation of DC, and this process resulted in slightly higher ex- release of the antiinflammatory IL-10 (Fig. 6A). Highest effects pression of CD86 and MHC class II compared with control DCs on cytokines/chemokines secreted by DCs pulsed with viable pulsed with LPS (0.5 ␮g/ml) (Fig. 4C). These data show that mat- pneumococci were shown for IL-8, followed IL-6, IL-12, uration of DCs is induced by internalized pneumococci but does TNF-␣, IL-1␤, and IL-10. No production or release of IFN-␥ not rely on PavA expression. was measured, which confirms the purity of the population of PavA expression by pneumococci is required to induce a DCs used (34). Compared with the supernatant of LPS-pulsed maximum host inflammatory response DCs, the relative amounts of the cytokines measured 18 h postinfection were higher in S.p. 35A⌬ply-infected DC cul- The activation process of DCs involves, in addition to the up- regulation of surface markers, the expression and release of tures, with the only exception of IL-8 (data not shown). These cytokines. These host cell immune responses are essential to data show a pneumococcal-induced proinflammatory host cell stimulate and prime the adaptive . To explore response, which is followed by an antiinflammatory response as whether infections with pneumococci induce or modulate re- indicated by IL-10 release. However, compared with the LPS- lease of IL-1␤, IL-6, IL-8, IL- 10, IL-12, IFN-␥, and TNF-␣ by induced cytokine release, the proportion of IL-12 and IL-10 DCs, the APCs were infected for 30 min with a MOI of 50 of shifted to a reduced proinflammatory overbalance for both live S.p. 35A⌬ply. Similar to our previous experiments, extra- pneumococcal strains (Fig. 6C). Similar to the PavA-positive cellular bacteria were killed by applying the antibiotics for 1 h S.p. 35A⌬ply strain, the corresponding pavA mutant induced and, in a first approach, cytokine mRNA expression was quan- release of IL-1␤, IL-6, IL-8, IL-12, TNF-␣, and, to a minor tified after 18 h for IL-1␤, IL-6, IL-8, IL-10, IL-12p35, IL- degree, IL-10. Again, IFN-␥ release was not measured (data not 12p40, IFN-␥, and TNF-␣ by employing the real-time reverse- shown). The deficiency of PavA in pneumococci resulted in a transcription PCR technique (qPCR). RBS9 served as the significantly decreased cytokine release compared with S.p. housekeeping gene. The qPCR revealed that phagocytosis of S. 35A⌬ply as indicated after 18 h postinfection of DCs (Fig. 6B). The Journal of Immunology 1959

FIGURE 7. T cell proliferation is promoted by human DCs infected with S.p. 35A⌬ply or its isogenic mutant S.p. 35A⌬ply⌬pavA. DCs were infected for 30 min with viable pneumococci, and after extracellular bac- teria were eliminated and killed by antibiotic treatment, stimulation was continued for 18 h. Proliferation of T cells was measured using the non- radioactive OxMi T cell proliferation tests. iDCs and LPS-pulsed DCs (0.5 Downloaded from p Ͻ 0.05 for ,ء .␮g/ml E. coli LPS; Sigma-Aldrich) were used as controls S.p. 35A⌬ply relative to its isogenic pavA mutant (S.p. 35A⌬ply⌬pavA).

of T cells, albeit to a lesser extent as LPS at all DC:T cell ratios. However, the pavA mutant and PavA-producing strain S.p. ⌬ply

differed in their potential to activate T cells via pulsed DCs. Con- http://www.jimmunol.org/ sistent with the cytokine profiles of infected DCs, proliferation of T cells was significantly higher with S.p. 35A⌬ply-infected DCs compared with T cell proliferation pulsed with PavA-deficient S.p. 35A⌬ply⌬pavA-infected DCs (Fig. 7). In conclusion, the lack of a functionally active pneumococcal PavA protein resulted in en- hanced pneumococcal internalization into DCs (Fig. 2) and a sig- nificantly decreased inflammatory host immune response of pulsed DCs (Fig. 6B). Consequently, this reduced immune response was less efficient to activate the adaptive immune response as measured by guest on September 27, 2021 FIGURE 6. Pneumococci-induced cytokine production by DCs is mod- by induction of T cell proliferation. ulated by PavA expression. A, DCs were infected with the pneumococcal strain S.p. 35A⌬ply for 30 min, and extracellular bacteria were eliminated PavA protein reassociates to pneumococci and complements the by antibiotic treatment. Cytokines were measured with the Luminex tech- defects of the pavA mutant nology using a multiplex bead-based cytokine assay. IFN-␥ was not de- tected. Results of a representative experiment are shown. B, Cytokine re- Binding studies were performed to investigate whether recombi- lease of DCs infected with the pavA mutant strain S.p. 35A⌬ply⌬pavA at nant PavA reassociates to the pneumococcal cell wall, as has been 18 h postinfection relative to DCs infected with the PavA-positive strain shown for the pneumococcal enolase, and/or binds to DCs. The S.p. 35A⌬ply. Results represent means Ϯ SD of four independent exper- PavA protein was produced as a MBP fusion protein (MBP-PavA) iments performed with DCs isolated from healthy donors. C, Ratios be- and purified under nondenaturing conditions. The functional ac- tween IL-12 and IL-10 of pneumococci-infected DCs or LPS-stimulated tivity of the recombinant PavA was confirmed by its ability to bind p Ͻ 0.05. to immobilized fibronectin (data not shown). First, binding of ,ء .DCs FITC-labeled pneumococci was investigated for immobilized MBP-PavA. The results showed a dose-dependent binding of wild- Induction of T cell proliferation by DCs pulsed with type pneumococci (data not shown) and pneumococci deficient in pneumococci is influenced by PavA expression PavA (Fig. 8A) to immobilized PavA protein. Specific binding of Stimulated and Ag-presenting DCs activate proliferation of naive pneumococci to immobilized MBP was not detected (Fig. 8A). T cells, thereby linking the innate immune response with the adap- Additionally, we assessed binding of soluble PavA protein (MBP- tive immune response of the host. To assess the impact of the PavA) to viable pneumococci by flow cytometry. Strikingly, sol- pneumococcal PavA protein on subsequent T cell proliferation, uble PavA binds to viable pneumococci as shown for pavA mutant activation of T cells was measured after pulsing human DCs with S.p. 35A⌬ply⌬pavA and similarly to S.p. 35A⌬ply (Fig. 8B). Bind- viable S.p. 35A⌬ply or its isogenic pavA mutant. Human DCs were ing of MBP alone was not detected (Fig. 8B). These results suggest infected with pneumococci and 30 min postinfection extracellular that PavA binds directly to the pneumococcal cell surface. More- bacteria were killed. The infections were continued for 18 h, rep- over, flow cytometric analysis shows that PavA is not recognized resenting a time point with significant levels of DC-released cy- by iDCs (Fig. 8B). In infection experiments with immature human tokines. Proliferation of T cells stimulated with LPS-matured DCs, DCs we investigated whether exogenously added recombinant iDCs, or infected DCs was measured using a nonradioactive OxMi PavA protein has the capability to inhibit phagocytosis of the pavA assay. In general, the rate of T cell activation was dependent on the mutant, so that the level of internalized pneumococci resembles numbers of employed DCs per assay. Both the DCs infected with that of S.p. 35A⌬ply. The pavA mutant was preincubated with the pneumolysin-deficient strain S.p. 35A⌬ply and DCs infected recombinant PavA protein and DCs were infected for 30 min. Re- with the double mutant S.p. 35A⌬ply⌬pavA induced proliferation markably, the internalization rates of pneumococci deficient for 1960 PavA INFLUENCES PHAGOCYTOSIS AND HOST IMMUNE RESPONSE

FIGURE 8. Binding of recombinant PavA to pneumococci affects internalization of pneu- mococci deficient for PavA and modulates cy- tokine expression of DCs. A, Binding of the FITC-labeled pneumococci S.p. 35A⌬ply⌬pavA to immobilized recombinant MBP or MBP- PavA as measured at 485/538 (excitation/ emission). Values are the means Ϯ SD from at least three independent experiments, each per- formed in triplicate. B, Binding of soluble MBP and MBP-PavA, respectively, to viable S. pneumoniae and human DCs was deter- mined by flow cytometry. Bound proteins were detected using anti-MBP or anti-PavA antiserum and anti-rabbit FITC conjugate. The dot plots show the percentage of positive events, and the log fluorescence intensity is shown on the x-axis. C, Number of intracellu- lar surviving pneumococci in the absence or presence of bacterial-bound MBP-PavA pro- Downloaded from tein. Pneumococci were pretreated for 30 min with the recombinant proteins (20 ␮g/108 bac- teria). DCs seeded in wells of a 96-well cell culture plate were infected for 30 min with a MOI of 50 bacteria (5 ϫ 106). The numbers of intracellular pneumococci were determined us- ing the antibiotic protection assay and quanti- http://www.jimmunol.org/ tative plating. Results represent the means Ϯ SD of at least three independent experiments, -p Ͻ 0.005. D, Impact of bacterial PavA pre ,ء treatment on pneumococci-induced cytokine release by DCs. Human iDCs were infected for 30 min with S.p. 35A⌬ply or its isogenic pavA mutant with a MOI of 50. The pavA mutant was employed without further treatment or af-

ter preincubation with MBP and MBP-PavA. by guest on September 27, 2021 Cytokines were determined after 18 h of stim- ulation with the Luminex technology using a multiplex bead-based cytokine assay. The graphs show the results of representative ex- periments (done in triplicates) and the amount of cytokines that are mostly affected. Values (automatically processed by the software with- out showing the SD) are the means of three DC charges as calculated by the MasterPlexQT software of the Luminex instrumentation.

pavA gene expression but complemented with MBP-PavA pro- gest that recombinant PavA protein initiates a phenotypic con- tein were similar to S.p. 35A⌬ply (Fig. 8C). These results dem- version of the pavA mutant, resulting in a phenotype similar to onstrated that exogenously added PavA protein has the ability pneumococci that produce endogenous PavA. to modulate the pavA mutant in a way that the level of bacterial ingestion by DCs is similar to the levels measured for PavA- Discussion expressing pneumococci. MBP alone did not prevent the in- Pneumococcal infections are relatively common, however, creased uptake of the pavA mutant by DCs (Fig. 8C). Strikingly, compared with high nasopharyngeal carrier rates in healthy immature human DCs pulsed for 30 min with PavA protein- adults and in children infections are less common than ex- pretreated pavA mutant bacteria showed higher amounts of cy- pected. Nevertheless, pneumococci are the most common bac- tokines IL-6, IL-8, and TNF-␣ 18 h postinfection compared teria associated with community-acquired pneumonia, which is with DCs infected with untreated S.p. 35A⌬ply⌬pavA (Fig. often accompanied by bacteraemia (2). Binding to epithelial 8D). PavA treatment of S.p. 35A⌬ply⌬pavA resulted in cyto- cells of host mucosal surfaces and immune evasion mechanisms kine levels that were similar to cytokine levels measured for are essential steps for pneumococci to initiate nasopharyngeal DCs pulsed with PavA-producing S.p. 35A⌬ply bacteria. No colonization and enable bacterial dissemination into submuco- changes in cytokine release were measured for IL-1␤, IL-10, sal host niches or the bloodstream. Encapsulation protects and IL-12 after PavA protein reassociation to the pavA mutant pathogens against phagocytosis by professional phagocytes and (data not shown). MBP treatment of bacteria did not influence allows pneumococci to survive and multiply during nasopha- the cytokine profiles (Fig. 8D). In conclusion, these data sug- ryngeal colonization or in the bloodstream (35). The Journal of Immunology 1961

In this study we have for the first time characterized the inter- rophages and DCs (44). Despite the important contribution of action of viable pneumococci with human monocyte-derived DCs phagocytes to the initiation of innate immune responses against and the induced inflammatory response reliant on bacterial PavA pneumococci, little is known about the DC-pneumococci interac- expression. We have demonstrated that pneumococcal phagocyto- tion per se and the role of bacterial virulence factors for uptake, sis by DCs is independent of pneumolysin production but is sig- intracellular survival, or immunomodulation. Colino and Snapper nificantly influenced by the capsular polysaccharide and PavA ex- (24) suggested that apoptosis is triggered by TLRs and in a pression. Pneumolysin is a cell-modulatory at sublytic MyD88-dependent manner. Immune recognition of the related concentrations and induces cytokine synthesis and CD4ϩ T cell bacterial species Streptococcus pyogenes by DCs occurs via dif- activation (17). Interestingly, a pneumolysin variant lacking he- ferent TLRs and MyD88 (45). Similar to pneumococci, S. pyo- molytic and complement-activating activity showed enhanced vir- genes induce maturation of DCs and the IL-12 response is critical ulence compared with pneumococci lacking pneumolysin. It has to combat infections with S. pyogenes (46, 47). Whether the sig- been suggested that pneumolysin is recognized by TLR4 and that naling through TLRs and MyD88 is the principal pathway by that the relatively rare occurrence of invasive disease after asymp- which DCs also sense pneumococci has to be proven. Using dif- tomatic pneumococcal colonization is due to pneumolysin TLR4- ferent mice infection models, the impact of MyD88 on host sus- induced robust inflammatory host cell responses (16). After expo- ceptibility to pneumococci has already been demonstrated (48– sure of Ͼ3 h to BMDCs, heat-killed pneumococci elicit apoptosis 50), suggesting that MyD88 also contributes to immune where the rapid caspase-induced apoptosis depends on pneumoly- recognition of pneumococci by DCs. sin expression (24). In our experiments DCs were pulsed at a max- After phagocytosis, pneumococci are mainly found, similar to imum for 1 h with viable wild-type pneumococci or isogenic mu- (28), within membrane-bound phagosomes, preferentially Downloaded from tants. The stimulation and infection of DCs for up to 2 h with lysosomes, and they become degraded in these DC compartments viable pneumococci did not induce necrotic or apoptotic death in as shown in killing assays. However, electron micrographs sug- the major proportion of DCs. However, to determine cytokine pro- gested that a minor proportion of intracellular pneumococci es- files or adaptive immune responses infections were continued for capes from the vacuole and resides in the cytosol. One might spec- up to 18 h. Flow cytometric analysis with annexin V and pro- ulate that these bacteria survive phagocytosis and contribute to the

pidium iodide showed that infections with viable and pneumoly- severe infections. The induced cytokine levels of DCs pulsed with http://www.jimmunol.org/ sin-expressing wild-type pneumococci induced significant necro- pneumococci deficient for PavA were significantly decreased com- sis, while infections with ply mutants did not (data not shown). pared with PavA-producing pneumococci, while maturation of Hence, the use of wild-type bacteria precludes the analysis of DCs did not depend on PavA. This effect was measured despite the pneumococci-induced cytokines and T cell responses. Similar to in higher number of initially phagocytosed pavA-deficient pneumo- vitro studies with nonprofessional and professional phagocytes cocci. Compared with LPS-induced DC activation, S.p. ⌬ply and such as (25, 29), ingestion of encapsulated pneumo- PavA-deficient S.p. ⌬ply⌬pavA-infected DCs exhibited a signifi- cocci by DCs was massively impaired. These results are in accor- cantly weaker T cell stimulatory capacity at a physiological pro- dance with Neisseria meningitidis whose adherence to and uptake portion of 30 T cells per DC. This is in accordance with a by DCs are significantly impaired for encapsulated bacteria (36). weaker proinflammatory dominance of the cytokine profile in- by guest on September 27, 2021 In contrast to strain D39, which is a high-encapsulated and mice- duced by pneumococci compared with LPS-activated cells. In- virulent strain, nonencapsulated pneumococci or serotypes produc- terestingly, PavA-producing pneumococci induced a higher T ing lower amounts of CPS including serotype 35A (25) were ef- cell proliferation than did PavA-deficient pneumococci, being ficiently phagocytosed by DCs. Interestingly, uptake experiments in concordance with the higher, albeit not significant, IL-12/ with FITC-labeled pneumococcal CPS from types 9N and 14 dem- IL-10 ratio (51). This effect was also observed under in vivo con- onstrated phagocytosis of CPS material by immature human DCs ditions in an experimental mouse meningitis model. Here, menin- into lysosomal compartments. No DC maturation or IL-10 or IL-12 geal inflammation was significantly reduced when animals were expression was induced by these CPS (37). A recent study dem- infected with a pneumolysin-producing PavA-deficient strain onstrated that the pneumococcal CPS of a serotype 1 activates (D39⌬pavA) compared with its isogenic wild-type D39 (22). Thus, CD4ϩ T cells via its presentation by MHC class II-positive tubules the results of the OxMi confirm the finding that PavA is a relevant of murine DCs (38). This observation also suggests that the CPS is virulent factor for the induction of a proper immune answer. recognized by DC receptor(s) and is taken up into lysosomal com- In agreement with earlier studies, these data support the idea that partments. Pneumococcal serotypes 3 and 14 are specifically rec- PavA is pivotal for the functional activity of other unidentified ognized by the C-type lectin DC-SIGN (DC-specific intercellular pneumococcal virulence determinants (22). These unknown viru- adhesion molecule-grabbing nonintegrin) and also by SIGN-R1, a lence determinants contribute presumably to protection against C-type lectin expressed on macrophages within the marginal zone phagocytosis by DCs and improve intracellular survival of the bac- of the spleen (39, 40). The recognition of pneumococcal serotypes teria. Moreover, comparisons of the induced cytokine levels sup- by SIGN-R1 mediates clearance of the bacteria under in vivo con- port the hypothesis that PavA modulates especially immunogenic ditions (41, 42). Although CPS from serotype 14 interacts with bacterial proteins. Therefore, the unidentified proteins will repre- DC-SIGN, immunomodulatory effects were not observed (40). sent promising vaccine candidates or target structures for novel Some pathogens such as HIV or Mycobacterium tuberculosis ex- therapeutic interventions. In the PavA ho- ploit the DC-SIGN to escape by manipulating DC func- molog Fbp54 decreases the protein levels of the adhesin internalin tion (43). This may also enhance the possibility that DCs are mis- B and toxin listeriolysin O and has been suggested to act as a used as a Trojan horse for dissemination in the host. Once taken up chaperone (52). Listeriolysin O is a highly immunogenic virulence by iDCs, the survival rate of intracellular pneumococci decreases determinant of Listeria and mediates the escape of the bacteria rapidly, suggesting that this scenario is negligible for pneumococ- from the vacuole into the cytosol (53). In pneumococci PavA does cal dissemination within the host, especially when PavA is lacking. not affect levels of pneumolysin or other known virulence factors Another receptor mediating clearance of pneumococci in a murine (21, 22). Here, we show for the first time that exogenously added pneumonia model is the scavenger receptor MARCO ( recombinant PavA is able to complement the deficiency of PavA in receptor with collagenous structure), which is produced by mac- PavA-deficient pneumococci. PavA has been identified on the 1962 PavA INFLUENCES PHAGOCYTOSIS AND HOST IMMUNE RESPONSE pneumococcal surface despite lacking a signal peptide and mem- 2003. Recognition of pneumolysin by Toll-like receptor 4 confers resistance to brane anchor domain (22). Our binding assays indicated a reasso- pneumococcal infection. Proc. Natl. Acad. Sci. USA 100: 1966–1971. 17. Kadioglu, A., J. N. Weiser, J. C. Paton, and P. W. Andrew. 2008. The role of ciation of purified PavA to the pneumococcal cell surface while Streptococcus pneumoniae virulence factors in host respiratory colonization and binding to DCs was not observed. Remarkably, the exogenously disease. Nat. Rev. Microbiol. 6: 288–301. added and reassociated PavA protein prevents phagocytosis of 18. Matthias, K. A., A. M. Roche, A. J. Standish, M. Shchepetov, and J. N. Weiser. 2008. Neutrophil-toxin interactions promote delivery and mucosal clear- pavA-deficient pneumococci, which are otherwise taken up in high ance of Streptococcus pneumoniae. J. Immunol. 180: 6246–6254. numbers compared with PavA-expressing bacteria. Additionally, 19. Bergmann, S., and S. Hammerschmidt. 2006. Versatility of pneumococcal sur- the immune response of DCs resembles that of DCs infected with face proteins. Microbiology 152: 295–303. 20. Hammerschmidt, S. 2006. Adherence molecules of pathogenic pneumococci. PavA-positive pneumococci. It is noteworthy that purified PavA Curr. Opin. Microbiol. 9: 12–20. protein does not induce maturation of DCs. Again, these data dem- 21. Holmes, A. R., R. McNab, K. W. Millsap, M. Rohde, S. Hammerschmidt, J. L. Mawdsley, and H. F. Jenkinson. 2001. The pavA gene of Streptococcus onstrate the important role of PavA for full virulence of pneumo- pneumoniae encodes a fibronectin-binding protein that is essential for virulence. cocci and that it is sufficient when PavA exerts its function on the Mol. Microbiol. 41: 1395–1408. pneumococcal cell surface. 22. Pracht, D., C. Elm, J. Gerber, S. Bergmann, M. Rohde, M. Seiler, K. S. Kim, H. F. Jenkinson, R. Nau, and S. Hammerschmidt. 2005. PavA of Streptococcus In summary, this study suggests that PavA is essential for pneu- pneumoniae modulates adherence, invasion, and meningeal inflammation. Infect. mococci to escape phagocytosis and killing by innate immune Immun. 73: 2680–2689. cells. Moreover, the lack of a functional PavA causes less severe 23. Colino, J., Y. Shen, and C. M. Snapper. 2002. Dendritic cells pulsed with intact Streptococcus pneumoniae elicit both protein- and polysaccharide-specific im- immune responses in the host, and the virulence factors modulated munoglobulin isotype responses in vivo through distinct mechanisms. J. Exp. by a functional PavA represent immunogenic factors that may Med. 195: 1–13. have a high potential as vaccine candidates. As a consequence, the 24. Colino, J., and C. M. Snapper. 2003. Two distinct mechanisms for induction of Downloaded from dendritic cell apoptosis in response to intact Streptococcus pneumoniae. J. Im- loss of function of PavA in S. pneumoniae improves survival of the munol. 171: 2354–2365. host and results in less severe immune responses to pneumococcal 25. Hammerschmidt, S., S. Wolff, A. Hocke, S. Rosseau, E. Muller, and M. Rohde. infections. 2005. Illustration of pneumococcal polysaccharide capsule during adherence and invasion of epithelial cells. Infect. Immun. 73: 4653–4667. 26. Rennemeier, C., S. Hammerschmidt, S. Niemann, S. Inamura, U. Zahringer, and Acknowledgments B. E. Kehrel. 2007. Thrombospondin-1 promotes cellular adherence of Gram- positive pathogens via recognition of peptidoglycan. FASEB J. 21: 3118–3132. We thank Ina Schleicher (Helmholtz Center for Infection Research, Braun- http://www.jimmunol.org/ 27. Romani, N., S. Gruner, D. Brang, E. Kampgen, A. Lenz, B. Trockenbacher, schweig, Germany) for excellent technical assistant in the preparation of G. Konwalinka, P. O. Fritsch, R. M. Steinman, and G. Schuler. 1994. Prolifer- the FESEM and TEM. We are especially grateful to Andreas Opitz (Trans- ating dendritic cell progenitors in human blood. J. Exp. Med. 180: 83–93. fusion Medicine, University of Wu¨rzburg, Germany) for providing human 28. Kolb-Maurer, A., I. Gentschev, H. W. Fries, F. Fiedler, E. B. Brocker, buffy coats of healthy donors, and we thank also Ana Maria Waaga-Gasser E. Kampgen, and W. Goebel. 2000. Listeria monocytogenes-infected human den- dritic cells: uptake and host cell response. Infect. Immun. 68: 3680–3688. (University of Wu¨rzburg, Wu¨rzburg, Germany) for her support with the 29. Hermans, P. W., P. V. Adrian, C. Albert, S. Estevao, T. Hoogenboezem, Luminex technology. I. H. Luijendijk, T. Kamphausen, and S. Hammerschmidt. 2006. The streptococ- cal lipoprotein rotamase A (SlrA) is a functional peptidyl-prolyl isomerase in- Disclosures volved in pneumococcal colonization. J. Biol. Chem. 281: 968–976. 30. Hammerschmidt, S., V. Agarwal, A. Kunert, S. Haelbich, C. Skerka, and

The authors have no financial conflicts of interest. P. F. Zipfel. 2007. The host immune regulator factor H interacts via two contact by guest on September 27, 2021 sites with the PspC protein of Streptococcus pneumoniae and mediates adhesion References to host epithelial cells. J. Immunol. 178: 5848–5858. 31. von Kockritz-Blickwede, M., O. Goldmann, P. Thulin, K. Heinemann, 1. Bogaert, D., R. De Groot, and P. W. Hermans. 2004. Streptococcus pneumoniae A. Norrby-Teglund, M. Rohde, and E. Medina. 2008. Phagocytosis-independent colonisation: the key to pneumococcal disease. Lancet Infect. Dis. 4: 144–154. antimicrobial activity of mast cells by means of extracellular trap formation. 2. Cartwright, K. 2002. Pneumococcal disease in Western Europe: burden of dis- Blood 111: 3070–3080. ease, antibiotic resistance and management. Eur. J. Pediatr. 161: 188–195. 32. Hill, M. E., D. J. Ferguson, J. M. Austyn, J. Newsom-Davis, and H. N. Willcox. 3. Brogden, K. A. 2005. Antimicrobial peptides: pore formers or metabolic inhib- 1999. Potent immunostimulatory dendritic cells can be cultured in bulk from itors in bacteria? Nat. Rev. Microbiol. 3: 238–250. progenitors in normal infant and adult myasthenic human thymus. Immunology 4. Jenssen, H., P. Hamill, and R. E. Hancock. 2006. Peptide antimicrobial agents. 97: 325–332. Clin. Microbiol. Rev. 19: 491–511. 33. Hammerschmidt, S., S. R. Talay, P. Brandtzaeg, and G. S. Chhatwal. 1997. SpsA, Annu. 5. Janeway, C. A., Jr., and R. Medzhitov. 2002. Innate immune recognition. a novel pneumococcal surface protein with specific binding to secretory immu- Rev. Immunol. 20: 197–216. noglobulin A and secretory component. Mol. Microbiol. 25: 1113–1124. 6. Reis e Sousa, C. 2001. Dendritic cells as sensors of infection. Immunity 14: 34. Schleicher, U., A. Hesse, and C. Bogdan. 2005. Minute numbers of contaminant 495–498. ϩ ϩ ϩ CD8 T cells or CD11b CD11c NK cells are the source of IFN-␥ in IL-12/ 7. Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. J. Liu, IL-18-stimulated mouse macrophage populations. Blood 105: 1319–1328. B. Pulendran, and K. Palucka. 2000. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18: 767–811. 35. Magee, A. D., and J. Yother. 2001. Requirement for capsule in colonization by 8. Cavrois, M., J. Neidleman, and W. C. Greene. 2008. The Achilles heel of the Streptococcus pneumoniae. Infect. Immun. 69: 3755–3761. Trojan horse model of HIV-1 trans-infection. PLoS Pathog. 4: e1000051. 36. Unkmeir, A., U. Kammerer, A. Stade, C. Hubner, S. Haller, A. Kolb-Maurer, 9. Rey-Ladino, J., X. Jiang, B. R. Gabel, C. Shen, and R. C. Brunham. 2007. Sur- M. Frosch, and G. Dietrich. 2002. Lipooligosaccharide and polysaccharide cap- vival of Chlamydia muridarum within dendritic cells. Infect. Immun. 75: sule: virulence factors of Neisseria meningitidis that determine meningococcal 3707–3714. interaction with human dendritic cells. Infect. Immun. 70: 2454–2462. 10. Chain, B. M., and J. Marcinkiewicz. 2003. The dendritc cell in bacterial infection: 37. Meltzer, U., and D. Goldblatt. 2006. Pneumococcal polysaccharides interact with sentinel or Trojan horse. In Bacterial Evasion of Host Immune Responses. Cam- human dendritic cells. Infect. Immun. 74: 1890–1895. bridge Univ. Press, Cambridge, U.K., pp. 3–19. 38. Stephen, T. L., M. Fabri, L. Groneck, T. A. Rohn, H. Hafke, N. Robinson, 11. Austrian, R. 1981. Some observations on the pneumococcus and on the current J. Rietdorf, D. Schrama, J. C. Becker, G. Plum, et al. 2007. Transport of Strep- status of pneumococcal disease and its prevention. Rev. Infect. Dis. 3(Suppl.): tococcus pneumoniae capsular polysaccharide in MHC class II tubules. PLoS S1–S17. Pathog. 3: e32. 12. Abeyta, M., G. G. Hardy, and J. Yother. 2003. Genetic alteration of capsule type 39. Kang, Y. S., J. Y. Kim, S. A. Bruening, M. Pack, A. Charalambous, A. Pritsker, but not PspA type affects accessibility of surface-bound complement and surface T. M. Moran, J. M. Loeffler, R. M. Steinman, and C. G. Park. 2004. The C-type of Streptococcus pneumoniae. Infect. Immun. 71: 218–225. lectin SIGN-R1 mediates uptake of the capsular polysaccharide of Streptococcus 13. AlonsoDeVelasco, E., A. F. Verheul, J. Verhoef, and H. Snippe. 1995. Strepto- pneumoniae in the marginal zone of mouse spleen. Proc. Natl. Acad. Sci. USA coccus pneumoniae: virulence factors, pathogenesis, and vaccines. Microbiol. 101: 215–220. Rev. 59: 591–603. 40. Koppel, E. A., E. Saeland, D. J. de Cooker, Y. van Kooyk, and T. B. Geijtenbeek. 14. Hirst, R. A., A. Kadioglu, C. O’Callaghan, and P. W. Andrew. 2004. The role of 2005. DC-SIGN specifically recognizes Streptococcus pneumoniae serotypes 3 pneumolysin in pneumococcal pneumonia and meningitis. Clin. Exp. Immunol. and 14. Immunobiology 210: 203–210. 138: 195–201. 41. Koppel, E. A., C. W. Wieland, V. C. van den Berg, M. Litjens, S. Florquin, 15. Johnson, M. K., C. Geoffroy, and J. E. Alouf. 1980. Binding of by Y. van Kooyk, T. van der Poll, and T. B. Geijtenbeek. 2005. Specific ICAM-3 sulfhydryl-activated cytolysins. Infect. Immun. 27: 97–101. grabbing nonintegrin-related 1 (SIGNR1) expressed by marginal zone macro- 16. Malley, R., P. Henneke, S. C. Morse, M. J. Cieslewicz, M. Lipsitch, phages is essential for defense against pulmonary Streptococcus pneumoniae in- C. M. Thompson, E. Kurt-Jones, J. C. Paton, M. R. Wessels, and D. T. Golenbock. fection. Eur. J. Immunol. 35: 2962–2969. The Journal of Immunology 1963

42. Lanoue, A., M. R. Clatworthy, P. Smith, S. Green, M. J. Townsend, H. E. Jolin, 88-dependent signalling controls bacterial growth during colonization and sys- K. G. Smith, P. G. Fallon, and A. N. McKenzie. 2004. SIGN-R1 contributes to temic pneumococcal disease in mice. Cell. Microbiol. 7: 1603–1615. protection against lethal pneumococcal infection in mice. J. Exp. Med. 200: 49. Khan, A. Q., Q. Chen, Z. Q. Wu, J. C. Paton, and C. M. Snapper. 2005. Both 1383–1393. innate immunity and type 1 to Streptococcus pneumoniae are 43. van Kooyk, Y., and T. B. Geijtenbeek. 2003. DC-SIGN: escape mechanism for mediated by MyD88 but differ in their relative levels of dependence on Toll-like pathogens. Nat. Rev. Immunol. 3: 697–709. receptor 2. Infect. Immun. 73: 298–307. 44. Arredouani, M., Z. Yang, Y. Ning, G. Qin, R. Soininen, K. Tryggvason, and 50. Koedel, U., T. Rupprecht, B. Angele, J. Heesemann, H. Wagner, H. W. Pfister, L. Kobzik. 2004. The scavenger receptor MARCO is required for lung defense and C. J. Kirschning. 2004. MyD88 is required for mounting a robust host im- against pneumococcal pneumonia and inhaled particles. J. Exp. Med. 200: mune response to Streptococcus pneumoniae in the CNS. Brain 127: 1437–1445. 267–272. 51. Stax, A. M., C. Crul, S. W. Kamerling, N. Schlagwein, R. N. van der Geest, 45. Loof, T. G., O. Goldmann, and E. Medina. 2008. Immune recognition of Strep- A. M. Woltman, and C. van Kooten. 2008. CD40L stimulation of rat dendritic tococcus pyogenes by dendritic cells. Infect. Immun. 76: 2785–2792. cells specifically favors the IL-12/IL-10 ratio resulting in a strong T cell stimu- 46. Loof, T. G., M. Rohde, G. S. Chhatwal, S. Jung, and E. Medina. 2007. The latory capacity. Mol. Immunol. 45: 2641–2650. contribution of dendritic cells to host defenses against Streptococcus pyogenes. 52. Dramsi, S., F. Bourdichon, D. Cabanes, M. Lecuit, H. Fsihi, and P. Cossart. 2004. J. Infect. Dis. 196: 1794–1803. FbpA, a novel multifunctional Listeria monocytogenes virulence factor. Mol. Mi- 47. Veckman, V., and I. Julkunen. 2008. Streptococcus pyogenes activates human crobiol. 53: 639–649. plasmacytoid and myeloid dendritic cells. J. Leukocyte Biol. 83: 296–304. 53. Dietrich, G., J. Hess, I. Gentschev, B. Knapp, S. H. Kaufmann, and W. Goebel. 48. Albiger, B., A. Sandgren, H. Katsuragi, U. Meyer-Hoffert, K. Beiter, F. Wartha, 2001. From evil to good: a cytolysin in vaccine development. Trends Microbiol. M. Hornef, S. Normark, and B. H. Normark. 2005. Myeloid differentiation factor 9: 23–28. Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021