CD46 Contributes to the Severity of Group A Streptococcal Infection

ABSTRACT Streptococcus pyogenes (group A Streptococcus) is a human pathogen that causes a wide variety of diseases ranging from uncomplicated superficial infections to severe infections such as streptococcal toxic shock syndrome and necrotizing fasciitis. These bacteria interact with several host cell receptors, one of which is the cell surface complement regulator CD46. In this study, we demonstrate that infection of epithelial cells with S. pyogenes leads to the shedding of CD46 at the same time as the bacteria induce apoptosis and cell death. Soluble CD46 attached to the streptococcal surface, suggesting that bacteria might bind available extracellular CD46 as a strategy to survive and avoid host defenses. The protective role of human CD46 was demonstrated in ex vivo whole-blood assays showing that the growth of S. pyogenes was enhanced in blood from mice expressing human CD46. Finally, in vivo experimental infection showed that bacteremia levels, arthritis frequency, and mortality were higher in CD46 transgenic mice than in nontransgenic mice. Taken together, these results argue that bacterial exploitation of human CD46 enhances bacterial survival and represents a novel pathogenic mechanism that contributes to the severity of group A streptococcal disease.

The gram-positive bacterium Streptococcus pyogenes (group A Streptococcus) is a human pathogen that can cause mild to severe infections of the skin and throat including impetigo, erysipelas, and pharyngitis as well as life-threatening toxic shock syndrome (35). In order to cause infection, S. pyogenes must colonize the oropharynx or external skin. These sites of entry are advanced barriers that protect underlying tissues. To maintain the barrier and protect underlying tissue, there is a tight balance between apoptosis and the regeneration of cells. Apoptosis is an essential process in the host defense against pathogens (9). Bacterial exploitation of host cell apoptosis may lead to the destruction of the epithelium, providing colonizing pathogens access to deeper, normally sterile sites. Several studies reported previously that S. pyogenes can induce apoptosis and cell death either by bacterial entry into cells or from an extracellular location (7,31,32,49). It was proposed that the induction of apoptotic cell death is a virulence mechanism that facilitates bacterial dissemination (7).
Bacterial colonization is initiated by interactions between specific virulence factors of the bacteria and defined components of the host cells. An important virulence factor used by S. pyogenes is the M protein, which has been shown to mediate binding to keratinocytes (8,36) and to participate in the invasion of epithelial cells (38). To colonize and cause disease, the bacteria must overcome early defense mechanisms that normally should eliminate and remove bacteria from the mucosal surface. S. pyogenes is capable of immune evasion, mainly by binding to complement regulatory proteins via the M protein (19). The soluble comple-ment regulator factor H binds to the C-terminal conserved region of the M protein, whereas the factor H-like protein binds at the N-terminal hypervariable region (23). It was shown that this may protect the organism from phagocytosis by polymorphonuclear leukocytes in blood (27). Similarly, human C4b-binding protein binds to the hypervariable region of M proteins and interferes with phagocytosis (2).
S. pyogenes strains can be divided into more than a hundred M serotypes or emm types based on their M proteins. It was demonstrated that the conserved C-terminal region of the M6 protein binds to the cell surface glycoprotein CD46 on keratinocytes (14,36). CD46 is an abundant cell surface complement regulator and a receptor for several pathogens (4,29,39). It consists of four complement control protein repeats, a serine/ threonine/proline-rich region, a transmembrane domain, and two types of cytoplasmic tails (4). The protein binds C3b and C4b that are deposited on the host cell membrane and serves as a cofactor for their proteolytic inactivation by plasma serine protease factor I. This process prevents the formation of the membrane attack complex and consequently protects human cells from complement-mediated lysis (30). It was shown that S. pyogenes interacts with CD46 during invasion of epithelial cells (38). Furthermore, the interaction between S. pyogenes and CD46 triggers cell signaling pathways that induce an immunosuppressive/regulatory phenotype in T cells (37).
In this study, we aimed to evaluate the role of CD46 during infection with S. pyogenes. We report that bacterial infection induces apoptosis of host cells as well as the shedding of cell surface CD46 into the extracellular space. S. pyogenes bound soluble CD46 in a growth phase-dependent manner. Furthermore, whole-blood survival assays as well as in vivo experimental infection revealed better bacterial survival in the presence of human CD46. Lethal disease and arthritis were much more frequent in CD46 transgenic mice than in nontransgenic mice, suggesting an important role of CD46 in streptococcal disease outcome.

MATERIALS AND METHODS
Bacterial strains and cell culture. S. pyogenes strain S165 of serotype T6 emm6 isolated from the blood of a patient suffering from severe invasive streptococcal disease was kindly provided by Birgitta Henriques Normark, Swedish Institute for Infectious Disease Control. The bacteria were grown in Todd-Hewitt broth (Difco Laboratories) supplemented with 1.5% yeast extract (Oxoid) at 37°C in a 5% CO 2 atmosphere. The human pharyngeal cell line FaDu (ATCC HTB-43) was maintained in Dulbecco's modified Eagle's medium (Sigma) supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 0.1 mM nonessential amino acids, and 1.0 mM sodium pyruvate. Unless stated otherwise, all experiments were performed using 100% confluent cells maintained in medium supplemented with heat-inactivated serum.
Flow cytometry of FaDu cells. For analysis of CD46, HLA, or ␤1 integrin expression, cells were collected by trypsinization and washed with phosphatebuffered saline (PBS). The cells were incubated in 3% bovine serum albumin (BSA) on ice for 1 h and washed with PBS. Staining for CD46 was conducted by incubation with 30 l of a primary antibody against CD46 (MAb-2, 2 g/ml; NeoMarkers) in 0.3% BSA on ice for 1 h. Cells were washed and incubated with 50 l of a secondary anti-mouse immunoglobulin G (IgG) antibody (R-phycoerythrin [R-PE], 4 g/ml; Jackson ImmunoResearch) in 0.3% BSA for an additional 40 min on ice. HLA was detected by the incubation of cells with an R-PE-conjugated HLA-A, -B, and -C monoclonal antibody (diluted according to the manufacturer's recommendations; BD Pharmingen). For ␤1 integrin (CD29) expression, a monoclonal ␤1 integrin antibody (10 g/ml; Chemicon) was used, followed by incubation with the secondary antibody anti-mouse IgG-Alexa 488 at 4 g/ml (Molecular Probes). The cells were washed, resuspended in PBS, and analyzed for fluorescence intensity by FACScan and Cellquest Pro (Becton Dickinson).
Analysis of CD46, HLA, and ␤1 integrin expression after infection. FaDu cells were grown to 100% confluence, and the cell medium was changed 24 h prior to infection. Bacteria were grown overnight to the stationary phase and added at a multiplicity of infection (MOI) of 100 in cell medium (Dulbecco's modified Eagle's medium with fetal bovine serum) by replacing 100 l cell medium with 100 l of the concentrated bacterial suspension. As a control, cells were treated with 150 g/ml gentamicin (Sigma) to prevent the overgrowth of extracellular bacteria, which gave results similar to those of untreated cells. At 3 and 24 h postinfection, cells were collected by trypsinization, stained for CD46 expression, and analyzed by flow cytometry. Staining for HLA and ␤1 integrin was performed 24 h postinfection.
Transcriptional analysis. Total RNA from FaDu cells infected with S. pyogenes was isolated at 6 and 24 h. RNA (1 g) was reverse transcribed with 0.5 g/l oligo(dT) using Superscript II reverse transcriptase (Invitrogen). For realtime PCR amplification, Power Sybr green PCR master mix (ABI) was used with the ABI Prism 7300 sequence detector system (Applied Biosystems) according to the manufacturer's guidelines. The cd46 gene was amplified with primers CD46F (5Ј-GTGGTCAAATGTCGATTTCCAGTAGTCG-3Ј) and CD46R (5Ј-CAAG CCACATTGCAATATTAGCTAAGCCACA-3Ј). The ␤-actin gene was chosen as an internal control and amplified simultaneously with primers ␤-actinF (5Ј-GGCACCACACCTTCTACAATGAG-3Ј) and ␤-actinR (5Ј-CGTCATACTCC TGCTTGCTGATC-3Ј). cd46 gene transcription was analyzed using the standard curve method, and data were presented as an expression ratio compared to the uninfected control, which was set as 100%.
Detection of apoptosis and cell permeability. FaDu cells were infected with bacteria at an MOI of 100 for 3 or 24 h. For chemical induction of apoptosis, FaDu cells were incubated with 10 M camptothesin (Sigma) overnight. For microscopic analysis, cells were grown on coverslips in a 24-well plate. After infection or chemical treatment, the cells were harvested, washed, and stained by annexin V for phosphatidylserine exposure or by propidium iodide (PI) for cell permeabilization. Annexin V and PI (diluted according to the manufacturer's recommendations; BD Pharmingen) were added to the cells, followed by flow cytometry analysis. Also, FaDu cells were simultaneously stained for CD46 and phosphatidylserine using primary antibody against CD46 (MAb-2, 2 g/ml; Neo-Markers) and annexin V. For nuclear degeneration analysis, Vectashield mounting medium containing DAPI (4Ј,6Ј-diamidino-2-phenylindole) (Vector Laboratories Inc.) was added to slides before mounting and microscopy analysis with a Leica microscope connected to a charge-coupled-device camera.
Analysis of CD46 binding to S. pyogenes. Recombinant CD46 was purified as previously described (24,46). Briefly, a fusion protein composed of the maltose binding protein (MBP) and the extracellular region of CD46 was purified over an amylose column according to the manual provided by the manufacturer (New England BioLabs). MBP was purchased from New England BioLabs. S. pyogenes (lag, early log, mid-log, and stationary phases) was preincubated with recombinant MBP-CD46 or MBP for 2 h at 37°C. Unbound CD46 was removed by washing, and bacteria were analyzed for CD46 binding by staining with CD46 polyclonal antibody (H294, 2 g/ml; Santa Cruz Technologies), followed by incubation with an anti-rabbit IgG antibody (Alexa 488, 4 g/ml; Molecular Probes) and flow cytometry analysis.
Animal infection assay. The pathogenicity of S. pyogenes was analyzed in transgenic mice expressing human CD46. These mice were previously described and express CD46 in a human-like pattern (4,21,22). In two independent experiments, 5-to 8-week-old CD46 transgenic mice (n ϭ 4 to 12) and corresponding nontransgenic mice (n ϭ 4 to 12) were infected intravenously (i.v.) with 3 ϫ 10 8 CFU. Blood samples were taken from the tail daily for 6 days after infection, diluted in PBS, and spread onto Todd-Hewitt agar plates to determine the bacterial load in the blood. The criteria for mice positive for arthritis were significant joint swelling in one leg. Joint swelling was graded significant joint swelling or severe joint swelling. The CD46 transgenic mice more often had severe joint swelling. Animal care and experiments were performed in accordance with institutional guidelines and have been approved by national ethical committees.
Blood survival assay. Blood from CD46 transgenic mice and appropriate nontransgenic control mice (B6/C3HeN) was collected by retro-orbital bleeding and heparin treated. Whole blood was diluted 1:1 with bacterial suspension in FaDu cell medium (10 4 CFU). Bacterial survival was determined by serial dilutions and plating of the blood-bacterium mixture at 2, 4, and 6 h postinfection.
Statistical analysis. A Student's t test was used to assess significance in ex vivo blood survival assays. In vivo animal infection assays were assessed using Fisher's exact test, and bacteremia was monitored with a nonparametric Mann-Whitney test. All in vitro and ex vivo experiments were performed in triplicates in sets of at least two independent experiments.

S. pyogenes infection leads to the loss of CD46 from epithelial cells.
In order to establish whether infection with S. pyogenes could affect CD46 expression in host cells, the human pharyngeal epithelial FaDu cell line was incubated with S. pyogenes strain S165 (emm6) and analyzed by flow cytometry using CD46 antibodies. As shown in Fig. 1A, bacterial infection induced the loss of CD46 after 24 h, whereas the CD46 level remained unchanged in control cells and in cells infected for 3 h. Furthermore, the infection of FaDu cells did not affect the expression of the human major histocompatibility complex (HLA), indicating that CD46 downregulation is specific and not part of a general release of cell surface proteins (Fig. 1B). Interestingly, the expression of ␤1 integrin was slightly reduced at 24 h (Fig. 1C), supporting that CD46 is linked to ␤1 integrin, as was previously reported (38). Control experiments showed that bacteria pretreated with ethanol or heat did not mediate the loss of CD46 in FaDu cells, suggesting that the process required viable bacteria (data not shown). Furthermore, Escherichia coli did not induce the shedding of CD46 (data not shown). These data demonstrate that the infection of epithelial cells with S. pyogenes reduces cell surface-expressed CD46.

CD46 is released into the extracellular environment after S. pyogenes infection.
To determine if the loss of CD46 membrane staining was due to decreased transcription, we used real-time reverse transcription-PCR and showed that the CD46 mRNA level was not significantly affected after infection of cells with S. pyogenes ( Fig. 2A). Since the transcription of CD46 was unaffected by bacteria, we next checked if CD46 was released from cells into the extracellular environment. Cell supernatants collected before and after infection were analyzed by Western blotting using CD46 antibodies. High levels of extracellular CD46 were detected at 24 h postinfection, suggesting that bacterial infection triggered the shedding of CD46 from the host cell surface (Fig. 2B).
CD46 is released from cells in response to infection and apoptotic cell death. Recently, Elward et al. (11) demonstrated that the chemical stimulation of apoptosis results in the loss of CD46 from various human cell lines, whereas other cell surface markers remain stable on the cell surface (11). It has also independently been shown that S. pyogenes induces apoptosis and cell death in host epithelial cells (31,49). Therefore, we assessed apoptotic cell death in FaDu cells after S. pyogenes infection by staining with annexin V and PI at different time points. Annexin V staining to detect phosphatidylserine exposure in the outer leaflet of the cytoplasmic membrane was positive for one-third of the FaDu cells at 6 h and for all cells at 24 h, whereas uninfected cells remained annexin V negative (Fig. 3A). Annexin V did not bind bacteria by itself (data not shown). Staining with PI increased with time, and at 24 h postinfection, all cells were positive (Fig. 3B). Furthermore, nuclear staining with DAPI and microscopic analysis of cells at 24 h postinfection demonstrated nuclear shrinkage (data not shown). These data argue that the infection of epithelial cells with S. pyogenes leads to apoptotic cell death as well as the specific release of cell surface CD46 (Fig. 1).
To further confirm that the apoptosis of FaDu epithelial cells caused the shedding of CD46, we induced the apoptosis of cells by treatment with camptothesin and then analyzed cell surface markers by flow cytometry. As shown in Fig. 3C, camptothesin induced both apoptosis and the loss of CD46 as detected by staining with annexin V and monoclonal antibodies against CD46. To verify that CD46 was released extracellularly after the induction of apoptosis, the cell supernatants were subjected to immunoblotting with CD46 antibodies. As shown in Fig. 3D, CD46 was clearly accumulated in the apoptotic supernatants. Strains expressing or not expressing SpeB, a cysteine protease secreted from S. pyogenes, all induced the shedding of CD46 (data not shown). Taken together, these data argue that the shedding of CD46 into the extracellular environment occurs after infection with S. pyogenes as a response to bacterium-induced apoptosis.
S. pyogenes binds soluble CD46. Since CD46 is released from human cells after S. pyogenes infection, we hypothesized that it would be beneficial for the bacteria to bind soluble CD46 as a strategy to evade early immune responses. Therefore, in a first step, we wanted to evaluate if CD46 could bind to the bacterial surface. Consequently, we generated and purified recombinant CD46 as an MBP-CD46 fusion protein. The CD46 recombinant protein contained the surface-exposed domains but not the transmembrane region and the cytoplasmic tail. Since the M protein expres- sion of S. pyogenes may vary depending on the growth phase, we incubated bacteria from lag, early log, mid-log, or stationary phase with recombinant purified CD46 for 2 h, washed away unbound protein, and measured bacteriumbound CD46 by flow cytometry. As shown in Fig. 4, S. pyogenes in mid-log phase bound well to recombinant CD46 but not to the control MBP. S. pyogenes from lag, early log, or stationary phase showed impaired and low levels of binding to CD46, clearly demonstrating the importance of the growth phase in S. pyogenes interactions with the environment. Furthermore, CD46 from cell supernatants bound to S. pyogenes, i.e., shed native CD46, can bind bacteria (data not shown). These data argue that S. pyogenes in mid-log phase binds soluble recombinant CD46 protein, whereas S. pyogenes in other growth phases fail to bind CD46 in a strong and consistent manner. S. pyogenes shows enhanced survival in the presence of human CD46. We hypothesized that the presence of CD46 might be beneficial for bacterial survival. To evaluate the importance of CD46 during early steps of streptococcal systemic infection, we investigated bacterial survival and growth in blood collected from transgenic mice expressing human CD46 and nontransgenic control mice lacking CD46. Bacteria were mixed with blood for 2, 4, or 6 h; incubated at 37°C; diluted; and spread onto plates for viable counts. In the ex vivo blood assay, S. pyogenes showed significantly better growth in blood from CD46 transgenic mice than in blood from nontransgenic mice after 6 h of incubation (Fig. 5), whereas there was no growth difference in serum from these mice (data not shown), indicating that human CD46 facilitates bacterial survival and suggesting that the binding of CD46 might help the bacteria to survive host defenses. It was shown that soluble CD46 is present in normal human plasma (18) and that CD46 levels are elevated under autoimmune conditions (12,25). To investigate the importance of CD46 in human blood, we pre- incubated S. pyogenes with recombinant CD46 (MBP-CD46) before adding the bacteria to human blood. Bacteria preincubated with MBP-CD46 survived significantly threefold better in human blood than did bacteria incubated with buffer or control protein (data not shown). Human CD46 increases disease severity in mice. The finding that S. pyogenes survived better in blood from CD46 ϩ/ϩ mice than in blood from CD46 Ϫ/Ϫ mice prompted us to evaluate the in vivo infection of mice. The capacity of S. pyogenes to cause arthritis as well as lethal disease has been studied in mouse model systems using i.v. inoculation doses of 2 ϫ 10 8 to 4 ϫ 10 8 cells per mouse (40). To study the in vivo impact of CD46, we challenged CD46 transgenic mice and nontransgenic mice with 3 ϫ 10 8 S. pyogenes cells i.v. The majority (80%) of the nontransgenic mice survived, but only 40% of the CD46 transgenic mice survived (Fig. 6A), indicating that CD46 augments disease severity. To analyze bacteremia, blood samples from the tail vein were collected at day 3 postinfection, serially diluted, and spread onto bacterial growth plates for viable count. The bacterial blood counts in CD46 transgenic mice were significantly higher than those in nontransgenic mice (Fig. 6B). Furthermore, survival correlated with bacteremia levels (data not shown). As a control of the murine model system, we previously showed that CD46 transgenic mice are not more sensitive than nontransgenic mice to infections with bacteria unable to interact with CD46 (21).
S. pyogenes is frequently isolated from patients with arthritis and accounts for 8% to 16% of cases of septic arthritis (17,26). Arthritic joint swelling of the S. pyogenes-infected mice was visually examined three times a day during 6 days. About 70% of the CD46 transgenic mice developed arthritis, whereas only 15% of the nontransgenic mice showed arthritic joint swelling ( Fig. 6C and D). There was no correlation between bacteremia and arthritis. Taken together, these data support that human CD46 may play an important role in protecting bacteria during invasive S. pyogenes disease.

DISCUSSION
S. pyogenes can cause a range of diseases that vary in manifestations and outcome depending on the bacterial strain and host susceptibility (1,28,34,47,51). In this study, we analyzed host responses after infection with S. pyogenes strain S165 of type emm6 isolated from a patient with severe invasive disease. We demonstrated that infection induced the apoptosis and shedding of CD46 from host epithelial cells. Recombinant soluble CD46 containing the extracellular protein domains bound to S. pyogenes in mid-log phase. Furthermore, bacteria survived better in blood of CD46 transgenic mice than in blood of nontransgenic mice lacking CD46, and in vivo infection showed that the presence of human CD46 increased the severity of disease. These data suggest that human CD46 plays an important role during S. pyogenes infection. CD46 serves as a receptor for several pathogens including strains of measles virus (10), human herpesvirus type 6 (42), group A streptococci (38), adenovirus (13,45), and Neisseria (24). Among these pathogens, infection by certain strains of measles virus (33,43), human herpesvirus type 6 (42), serogroup B adenovirus (41), and Neisseria gonorrhoeae (15) has been shown to cause CD46 downregulation from the cell surface. The detailed mechanisms of surface CD46 downregulation upon infection by these pathogens remain to be elucidated; however, the decrease in the surface density of CD46 renders the cells more susceptible to lysis by complement, as demonstrated in vitro (44), and may contribute to the attenuation of these pathogens by the rapid clearing of infected cells.
Elward et al. (11) and Cole et al. (6) demonstrated that the chemical induction of apoptosis results in the specific release of CD46. These investigators revealed that apoptotic cells shed CD46 in the form of apoptotic membrane vesicles or blebs and that necrotic cells release soluble CD46 protein (6,11). We show that the chemical induction of apoptosis also triggers apoptosis in FaDu epithelial cells and that this results in the shedding of soluble CD46 from the cell surface. After infection of the cell with S. pyogenes, apoptotic cell death takes place, and CD46 is released into the extracellular environment. CD46 may be released by a specific cleavage at the cell surface or as part of membrane apoptotic blebs. It is possible that a combination of these release processes occurs at different time points during infection. The downregulation of CD46 expression on epithelial cells after Neisseria gonorrhoeae infection results in the release of membrane vesicles (16). Future studies will reveal if S. pyogenes infection triggers the release of CD46 membrane vesicles, a free CD46 cleavage product, or a combination of both. The finding that HLA protein expression remained stable when ␤1 integrin expression was reduced supports that CD46 is closely linked to ␤1 integrin, as was previously shown (38).
It is well known that bacterial protein or gene expression may vary dramatically between different growth phases or growth conditions (3,5). Several virulence factors of S. pyogenes have been shown to be growth phase dependent (50). Expression of the M protein, the C5a peptidase, and capsule are maximal at the exponential growth phase, whereas streptococcal pyogenic exotoxins A and B and mitotic factor are maximally expressed in later phases of growth. We found that the best binding of CD46 to S. pyogenes occurred at mid-log phase. Several lines of evidence support that CD46 can interact with the M protein (14, 36-38); however, it cannot be excluded that another surface component of S. pyogenes binds to CD46 as well.
As shown in the ex vivo blood assay, bacteria were able to survive better in blood from transgenic mice expressing human CD46 than in blood from nontransgenic mice. One possibility is that S. pyogenes uses extracellularly released CD46 to mask and hide from the immune system. Since S. pyogenes is a human-specific pathogen, there must be specific factors of the bacteria and the host that cause this specificity and prevent other species from being affected by severe S. pyogenes disease. The data in this work suggest that human CD46 is one such host factor. However, the in vivo experiments were performed in a murine model system, and it is yet to be determined whether CD46 also has similar properties in humans.
To further evaluate the importance of CD46, we infected transgenic mice expressing CD46. These mice express CD46 in a human-like manner and elicit immune responses similar to that of humans as a consequence of bacterial attack (21,22), suggesting that the mouse model is an adequate experimental in vivo system. Experimental infection revealed that CD46 transgenic mice developed a higher level of bacteremia than nontransgenic mice, supporting that CD46 enhances blood survival. In addition, bacterial numbers increased over time and reached the highest level at day 3 postinfection, showing that bacteria were able to grow in blood. It is well known that S. pyogenes avoids the complement alternative pathway attack by binding soluble complement regulators such as factor H or C4b binding protein on their surface (20). The data in this work argue that soluble and shed CD46 could also play a role in protecting bacteria against complement alternative pathway attack. The development of arthritis occurred more often in CD46 transgenic mice than in nontransgenic mice but was not always linked to high bacterial blood counts (data not shown), indicating that inflammatory factors are likely involved in this process. Finally, the higher mortality of CD46 transgenic mice than nontransgenic mice argues that CD46 facilitates severe lethal disease. The findings that bacteria survived better in blood from CD46 transgenic mice than in blood from nontransgenic mice and that bacteria preincubated with recombinant CD46 survived better in human blood support that CD46 increases bacterial survival and that soluble CD46 is important for bacterial survival. However, we cannot exclude that the increased virulence in CD46 transgenic mice is caused by enhanced levels of CD46-mediated adherence or invasion of certain cell populations. In future studies, we will further evaluate the role of CD46 in bacterium-host cell interactions.
This study demonstrates that S. pyogenes induces apoptosis and the release of CD46 and that bacteria bind soluble CD46. The interaction with CD46 might help to overcome innate immune defenses since bacteria survived better in blood and were more lethal in the presence of human CD46. However, it cannot be excluded that bacterial interaction with CD46 mediates host cell signal transduction that in turn affects innate immune defenses against the disease. Taken together, the data in this work support that S. pyogenes interacts with soluble free CD46 and that CD46 plays an important role during S. pyogenes infection by contributing to bacterial survival. 3956 LÖ VKVIST ET AL. INFECT. IMMUN.