The Receptor for the Complement C3a Anaphylatoxin (C3aR) Provides Host Protection against Listeria monocytogenes Induced Apoptosis

Abstract

Listeria monocytogenes (LM) is a Gram-positive intracellular bacterium that is acquired through tainted food and may lead to systemic infection and possible death. Despite the importance of the innate immune system in fighting LM infection, little is known about the role of complement and its activation products, including the potent C3a anaphylatoxin. In a model of systemic LM infection, we show here that mice lacking the receptor for C3a (C3aR-/-) are significantly more sensitive to infection compared to WT mice as demonstrated by decreased survival, increased bacterial burden, and increased damage to their livers and spleens. The inability of the C3aR-/- mice to clear the bacterial infection was not caused by defective macrophages or by reduction of cytokines/chemokines known to be critical in the host response to LM, including IFN-γ and TNF-α. Instead, TUNEL staining together with Fas, active caspase-3, and Bcl-2 expression data indicate that the increased susceptibility of C3aR-/- mice to LM infection was largely caused by increased LM-induced apoptosis of myeloid and lymphoid cells in the spleen that are required for ultimate clearance of LM, including neutrophils, macrophages, dendritic cells, and T cells. These findings reveal an unexpected function of C3a/C3aR signaling during the host immune response that suppresses Fas expression and caspase-3 activity while increasing Bcl-2 expression, thereby providing protection to both myeloid and lymphoid cells against LM-induced apoptosis.

Introduction

LM is a Gram-positive facultative intracellular pathogen that is transmitted via contaminated food, and infection with this pathogen can lead to sepsis and meningitis. This bacterium infects mostly older adults, persons with weakened immune systems, pregnant women, and newborns. Pregnant women account for approximately 25% of the cases of listeriosis, which can lead to miscarriage, stillbirth, and death of the newborn soon after birth (1). Outbreaks of LM that result in invasive disease have mortality rates of 20-30%, which is considerably higher than the mortality rates of other foodborne bacteria such as Salmonella and Shigella (1). A recent outbreak of LM was the result of contaminated canteloupes from a farm in Colorado (2). In this outbreak, 147 cases were reported across 28 states, with 33 reported deaths, resulting in a 22% mortality rate (2)(final update information on this outbreak was published in 2012 on the CDC website, www.cdc.gov).

Following infection with LM, innate immune responses are rapidly triggered and are essential for host survival (3). Early resistance to infection is attributed to the production of IFN-γ (4-6) and TNF-α (7-9) and the recruitment and activation of monocytes, macrophages, and neutrophils (3), but ultimate clearance of LM is dependent on CD4⁺ and CD8⁺ lymphocytes (10). The proinflammatory state initiated in the host upon LM infection promotes Th1 lymphocyte development (11). Functionally, these CD4⁺ Th1 cells and their secreted products are important for efficient dendritic cell activation and subsequent maintenance of memory CD8⁺ T cells (12). During the initial stages of infection, LM causes extensive apoptosis of lymphocytes, which serves to impair the host response and to create a more permissive microenvironment to support bacterial growth (13).

Despite the importance of the innate immune system in fighting LM infection, little is known about the role of the complement system. Early studies showed that LM is able to activate the alternative pathway of complement activation, which results in opsonization of LM by C3-derived fragments (14-17) and subsequent phagocytosis by macrophages. Phagocytosis by macrophages was dependent on CR3 binding to the C3 fragments deposited on the bacterial surface (18). A recent study using LM-infected C3-/- mice showed that C3 opsonization is not only important for bacterial clearance by macrophages but is also critical for platelet-binding and subsequent transport and targeting of LM to splenic CD8α⁺ dendritic cells (19). Another recent study reported that C3 is essential for optimal activation of antigen-specific T cells during LM infection of mice (20). Although these most recent studies demonstrated that complement component C3 is important in bacterial transport and T cell activation during LM infection, they did not address the importance or biological effects of a major C3 activation product, C3a.

C3a is a 77 amino acid peptide that is generated when complement C3 is cleaved during the activation of the complement cascade; it is traditionally known as an anaphylatoxin that causes smooth muscle contraction, histamine release from mast cells, and vasodilation. During the past several years published investigations have shown that in addition to its anaphylatoxin properties, C3a is a potent mediator of numerous other biological responses (both inflammatory and anti-inflammatory) (21). C3a triggers these biological responses by binding to a specific G protein-coupled receptor, C3aR, that is expressed on both bone marrow-derived myeloid and lymphoid cells, including monocytes/macrophages, neutrophils, dendritic cells, basophils, eosinophils, mast cells, platelets, T lymphocytes, and B lymphocytes (22-28). In addition, C3aR is found on parenchymal cells of the central nervous system (29, 30), lungs (31), and kidney (32). The generation of C3aR-/- mice by targeted gene deletion, have facilitated numerous in vivo studies that have revealed unexpected biological functions of C3a in disease pathogenesis. For example, the use of C3aR-/- mice in allergic models of the lung and skin by our laboratory as well as others have shown that C3a is an important molecular regulator of CD4⁺ T cell effector functions, primarily via binding and signaling through its receptor expressed on antigen presenting cells (33-37). Other recent investigations have also shown that C3a regulates the production of regulatory T cells (38, 39), as well as augmenting alloreactive CD8⁺ T cell responses (40).

Although it is now accepted that C3a has a significant role in innate and adaptive immunity, there has been little attention paid to the impact that C3a has on the host cellular immune response to intracellular pathogens, such as LM. Accordingly, in the current investigation, we have subjected WT and C3aR-/- mice to a model of systemic LM infection. The results demonstrate that the absence of C3aR greatly increases sensitivity to LM infection and reveal a previously unknown function of C3a in promoting protection to both myeloid and lymphoid cells against LM-induced apoptosis.

Materials and Methods

Mice

The C3aR^-/- mice used in these studies were generated in our laboratory and have been described previously (41). The C3aR^-/- mice were backcrossed for over ten generations onto the C57BL/6 background. Age-matched C57BL/6 mice from our own inbred C57BL/6 colony served as wild-type controls. All mice were housed in HEPA-filtered Tecniplast cages in a barrier facility. Male mice that were eleven to fourteen weeks of age were used for these studies. All mouse protocols followed institutional guidelines for animal care and welfare.

Bacterial infection

Listeria monocytogenes ATCC strain 13932 (serotype 4b) (MicroBioLogics, Inc.) was used for all infection studies. Bacteria were cultured in Bacto brain heart infusion (BHI) broth at 37°C to mid-logarithmic phase, harvested by centrifugation, washed once in sterile PBS, and resuspended in sterile PBS. Mice were infected i.v. with 1 × 10⁵ bacteria in a volume of 100 μl. Control mice in all experiments received 100 μl of sterile PBS i.v. The number of bacteria present in the inoculum was verified by culturing serial dilutions of the inoculum on BHI agar plates. In some experiments, C3aR-/- mice were pre-treated with 10 mg/kg Z-VAD-FMK (prepared per manufacturer’s instructions) (R&D Systems) or vehicle (10% DMSO in PBS) given i.p. two hours prior to LM infection.

Survival study

Mice were infected i.v. with 1 × 10⁵ LM and were observed for survival every 6 hours for twelve days. Survival curves were generated using GraphPad Prism (San Diego, CA) software, and statistical significance was assessed using the Log-rank test.

Bacterial burden in the liver and spleen

Liver and spleen were aseptically removed from the mice either 1 day or 3 days post-infection, rinsed in PBS, and then placed in HBSS in 5 ml tubes. The organs were homogenized using a PRO200 homogenizer (ProScientific) on medium speed and were then placed on ice. Bacterial counts were obtained by plating serial dilutions of each homogenate on BHI agar plates. Data are expressed as mean CFU per organ (Log₁₀) ± SEM.

In vitro LM killing assay

WT and C3aR-/- mice were injected i.p. with 1 ml of 3% Proteose Peptone (Oxoid). Three days later the peritoneal cavity was lavaged with 5 ml of HBSS to collect the cells. Using a multiplicity of infection (MOI) of 1, 1 × 10⁶ peritoneal exudate cells were incubated with 1 × 10⁶ LM in a volume of 1 ml of HBSS containing 5% autologous normal mouse serum at 37°C with gentle shaking. Aliquots were removed immediately upon infection (0 h) and also at 2 h, 4 h, and 6 h, and serial dilutions were made and plated on BHI agar plates. The data is presented as mean CFU/ml (Log₁₀) ± SEM.

Cytokine and chemokine analysis

Cytokines and chemokines were measured from sera taken on days 1 and 3 post-infection using the Milliplex mouse cytokine/chemokine 22-plex kit (Millipore #MPXMCYTO70KPMX22) with the Luminex 200 system. G-CSF (R&D Systems), IL-10 (BD Biosciences), IL-6, TNF-α, IFN-γ, and MCP-1 (Biolegend) were measured in the sera of the Z-VAD-FMK and vehicle pre-treated mice by ELISA assay. Data are expressed as mean pg/ml ± SEM.

Liver and spleen histology

The large lobe of the liver and the entire spleen were removed from the mice either 1 day or 3 days post-infection, rinsed in PBS, and placed in 10% buffered formalin. The livers and spleens were dehydrated with increasing concentrations of ethanol, embedded in paraffin, cut into 5-μm sections, and stained with either H&E (Fisher) or the DeadEnd™ Colorimetric TUNEL System (Promega) for visualization of apoptotic cells. The brightfield images were acquired using SPOT Advanced software (Diagnostic Instruments, Inc.) and a Zeiss Axioskop microscope (Carl Zeiss, Inc.) equipped with a SPOT-RT digital camera (Diagnostic Instruments, Inc.). The abscess area in the livers was quantitated using Amira software (FEI Visualization Sciences Group) on H&E images taken at 50X magnification. The outer edge of each abscess was outlined using the drawing tool, and then the outlined area was filled in to highlight each abscess. The software determined the number of pixels within each highlighted abscess per image. The abscess area was then calculated by multiplying the number of pixels in the highlighted abscesses times 100 and then dividing this number by the total number of pixels in the image. Three images of different sections of the liver were used to obtain a mean liver abscess area per mouse. The data are expressed as mean liver abscess area ± SEM. The percentage of TUNEL⁺ staining in the spleen sections was also calculated using Amira software on TUNEL images taken at 100X magnification. The pixels that were above the background darkness threshold were counted as TUNEL⁺ pixels. The percentage of TUNEL⁺ staining was then calculated by multiplying the number of TUNEL⁺ pixels times 100 and then dividing this number by the total number of pixels in the image. Three images of different sections of the spleen were used to obtain a mean value per mouse. The data is presented as mean percentage of TUNEL⁺ staining ± SEM.

Serum AST and ALT analysis

Sera samples from PBS-treated and LM-infected mice were submitted to the University of Texas M.D. Anderson Cancer Center Veterinary Medicine and Surgery Department. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were quantitated using an automated COBAS Integra 400 Analyzer (Roche). Data are expressed as mean units per liter ± SEM.

Immunophenotyping

Leukocytes were harvested from spleens for FACS analysis on days 1 and 3 post-infection using the gentleMACS Dissociator (Miltenyi Biotec) with the pre-set spleen setting. Prior to staining, erythrocytes were lysed using ACK lysing buffer (Lonza BioWhittaker). Total live cells were then counted with a hemacytometer using trypan blue exclusion. Non-specific antibody binding was blocked using mouse Fc block (BD Pharmingen). The following markers (Biolegend), with the clone listed in parentheses, were used to characterize the cell populations: CD4 (GK1.5), CD8a (53-6.7), CD11b (M1/70), Ly-6G (1A8), F4/80 (BM8), CD11c (N418), MHC class II (M5/114.15.2), NK1.1 (PK136), and CD19 (6D5). Fas/CD95 expression was analyzed using a PE-conjugated hamster anti-mouse CD95/Fas (Jo2) antibody (BD Biosciences). 7-AAD (BD Pharmingen) was added in the final 10 min of staining for dead cell exclusion. The cells were washed one time with PBS before analysis. A minimum of 20,000 events were collected on a FACSCalibur (BD Biosciences) flow cytometer, and the data was analyzed using Kaluza software (Beckman Coulter, Inc.).

Caspase-3 activity

Caspase-3 activity was measured in splenocytes harvested from spleens of infected WT and C3aR-/- mice or from PBS-treated mice using the CaspACE Colorimetric Assay System (Promega). The splenocytes were prepared from the spleens as described above. An equal density of splenocytes were lysed in the lysis buffer contained in the kit according to the manufacturer’s protocol. The splenocyte lysates were centrifuged at 15,000 × g for 20 min at 4°C to obtain the supernatant fractions, and then the supernatant fractions were assayed for caspase-3 activity according to the manufacturer’s protocol. The data is expressed as mean absorbance (405 nm) ± SEM.

Bcl-2 Expression

Leukocytes were harvested from spleens for Bcl-2 staining on days 1 and 3 post-infection or from PBS-treated mice using the gentleMACS Dissociator (Miltenyi Biotec) with the pre-set spleen setting. Prior to staining, erythrocytes were lysed using ACK lysing buffer (Lonza BioWhittaker). Cells were fixed and permeabilized using the Nuclear Factor Fixation and Permeabilization Buffer Set (Biolegend) following the manufacturer’s protocol. After fixation and permeabilization, the cells were stained with FITC-conjugated anti-Bcl-2 antibody (Biolegend) or FITC conjugated Mouse IgG1 isotype control (Biolegend) for 30 min in the dark. The cells were washed two times and then were analyzed on a FACSCalibur (BD Biosciences) flow cytometer. A minimum of 10,000 events were collected, and the data was analyzed using Kaluza software.

Inhibition of apoptosis

Splenocytes were harvested from spleens of uninfected WT and C3aR-/- mice using the gentleMACS Dissociator, and RBCs were lysed as described above. The cells were resuspended in IMDM with GlutaMax (Gibco) containing 1X non-essential amino acids (Lonza) and 1X HL-1 serum-free supplement (Lonza). 2 × 10⁶ cells were pre-incubated with 20 μM Z-VAD-FMK (prepared in DMSO per manufacturer’s instructions) (R&D Systems) or vehicle for 45 min. After 45 min, 2 × 10⁴ LM was added to the cells. The cells were cultured in a volume of 1 ml in sterile 24 well polypropylene plates for 20-44 hours at 37°C with 5% CO₂. At the indicated times, the cells were stained with trypan blue, and live cells were counted using a hemacytometer. The data is presented as mean cells/ml ± SEM.

Statistical analyses

All statistical analyses were done using GraphPad Prism software. For most experiments, comparisons between WT and C3aR-/- mice were done with the unpaired two-tailed t test, with P values < 0.05 considered significant. Survival curves were generated using GraphPad Prism software, and the Log-rank test was used to assess statistical significance among groups of mice, with P values < 0.05 considered significant. One-way ANOVA, with the Tukey post-test, was used to measure significance in the Z-VAD-FMK experiments, with P values < 0.05 considered significant.

Results

Lack of C3aR leads to increased susceptibility to LM infection

To assess the overall impact of the C3aR during the innate immune response to systemic LM infection, we injected WT and C3aR-/- mice with 10⁵ CFUs of LM and monitored their survival for 12 days. The C3aR-/- mice began to die on day 3 with only 14% survival (1 of 7 survived) by day 8, compared to 66% survival (6 of 9 survived) in the WT mice (P = 0.017) (Fig. 1).

Figure 1. C3aR-/- mice have decreased survival following LM infection.

WT and C3aR-/- mice were infected i.v. with 1 × 10⁵ LM and were monitored for survival for 12 days. The data are pooled from two independent experiments. n = 9 for WT and n = 7 for C3aR-/- mice. * P = 0.017 by the Log-rank test.

To determine if the increased mortality seen in the C3aR-/- mice correlated with an increase in bacterial numbers in the spleen and liver, we injected WT and C3aR-/- mice with LM and harvested spleens and livers from the mice 1 day and 3 days post-infection. On day 1, only a small increase in LM bacterial burden was observed in the tissues of the C3aR-/- mice compared to WT mice, with the increase in CFUs reaching significance only in the livers but not in the spleens of the C3aR-/- mice (Fig. 2A and 2B). In contrast, on day 3 post-infection, there was a dramatic difference in LM bacterial burden between the WT and C3aR-/- mice, with both the spleens (7-fold) and livers (3-fold) of the C3aR-/- mice containing significantly increased CFUs compared to WT mice. Moreover, from day 1 to day 3 post-infection the bacterial burdens in the spleens and livers of the C3aR-/- mice increased significantly (3-fold and 9-fold, respectively); however, no increase in infection was observed in the WT mice from day 1 to day 3. Instead, on day 3 post-infection, the numbers of bacteria in the WT tissues either decreased (spleens) or were not statistically significantly different (livers) from the CFUs on day 1 (Fig. 2A and 2B). To determine if macrophages from C3aR-/- mice have an innate defect in their ability to kill LM, proteose-peptone elicited macrophages from WT and C3aR-/- mice were incubated with LM in vitro at a MOI of 1 for 2 h, 4 h and 6 h. As shown in Figure 2C, the amount of LM recovered at the various time points was similar amongst the WT and C3aR-/- macrophages, indicating no inherent defect in killing by C3aR-/- macrophages.

Figure 2. C3aR-/- mice have higher bacterial burdens in spleen and liver following LM infection.

WT and C3aR-/- mice were infected i.v. with 1 × 10⁵ LM, and 1 day and 3 days post-infection, the spleens (A) and livers (B) were removed and homogenized under sterile conditions. The data are pooled from two independent experiments and is presented as mean CFU per organ (Log₁₀) ± SEM, with n = 8-11 mice per group per time point. * P ≤ 0.035; ** P = 0.001 by t test. (C) Proteose-peptone elicited peritoneal exudate cells were infected in vitro with LM at a MOI of 1. At the indicated times, the number of live LM was determined as CFU / ml, with the data presented as mean CFU/ml (Log₁₀) ± SEM. The experiment was performed in triplicate and is representative of two independent experiments.

Cytokine and chemokine concentrations in LM-infected WT and C3aR-/- mice

Early resistance to LM infection has been attributed to the production of IFN-γ and TNF-α and the recruitment and activation of monocytes/macrophages and neutrophils. To investigate if the absence of C3aR alters the host’s cytokine response to LM infection, IFN-γ, TNF-α, and other relevant cytokine and chemokine concentrations were determined from the sera of WT and C3aR-/- mice on days 1 and 3 post LM infection. On day 1 post-infection, IFN-γ serum concentrations were 38% higher in the C3aR-/- mice compared to the WT mice, while the TNF-α serum concentrations were comparable in both strains of infected mice (Fig. 3). Also on day 1, C3aR-/- sera concentrations of the cytokines and chemokines involved in the generation and chemoattraction of monocytes/macrophages and neutrophils were either significantly increased (G-CSF and IL-17) or were comparable (MIP-1α, IP-10, and MCP-1) to that in the WT mice (Fig. 3). On day 3 the sera levels of IFN-γ had decreased dramatically in both groups of mice. The TNF-α levels in the sera of the WT mice remained unchanged on day 3 compared to day 1, but TNF-α increased 3-fold in the sera of the C3aR-/- mice from day 1 to day 3. Except for IL-17, on day 3 post-infection the sera concentrations of the cytokines and chemokines involved in the generation and chemoattraction of monocytes/macrophages and neutrophils were significantly higher in the C3aR-/- mice compared to the WT mice (46% more IP-10, 3-fold increase of G-CSF, 7-fold increase of MCP-1, and 10-fold increase of MIP-1a) (Fig. 3). In addition, serum levels of IL-6 were 3-fold higher in the C3aR-/- mice on day 3 compared to WT mice, but there was no significant difference on day 1. On day 1, the infected WT and C3aR-/- mice expressed similar low levels of the anti-inflammatory cytokine IL-10 in their sera. On day 3 the IL-10 concentration in WT mice remained essentially unchanged from day 1. In contrast, the IL-10 in the sera of the infected C3aR-/- mice increased dramatically from day 1 to day 3 and was 32-fold higher than that of the WT mice on day 3. Collectively, these data reveal no reduction of critical interleukins, cytokines, or chemokines that would cause the increased LM infection observed in the C3aR-/- mice.

Figure 3. Cytokine and chemokine production following LM infection.

WT and C3aR-/- mice were infected i.v. with 1 × 10⁵ LM or PBS, and serum was isolated from the mice 1 day and 3 days post-infection. WT and C3aR-/- mice injected with PBS had little to no detectable levels of these cytokines and chemokines, with no differences between groups (data not shown). The data are pooled from two independent experiments. The data for the LM-infected mice is presented as mean pg/ml ± SEM, with n = 8-11 mice per group per time point. ** P ≤ 0.0096; *** P ≤ 0.0003 by t test.

C3aR-/- mice have more severe liver and spleen pathology

Microabscess formation is a histological hallmark of LM-infected liver (42). To evaluate microabscess formation in the livers of LM-infected WT and C3aR-/- mice, livers at 1 and 3 days post-infection were stained with H&E. In both of the WT and C3aR-/- mice, LM infection resulted in the development of organized microabscesses (Fig. 4A). Morphologically, the microabscesses were not remarkably different in the WT and C3aR-/- mice at either time point. Moreover, there was no significant difference in the number of microabscesses in the WT and C3aR-/- infected mice 1 day post-infection (as determined by liver abscess area Fig. 4B), but by day 3 post-infection, the number of microabscesses was significantly higher in the C3aR-/- mice compared to the WT mice (Fig. 4B). In addition, serum levels of ALT and AST, which are markers of liver inflammation and injury, were elevated in both strains of mice 1 and 3 days post-infection (compared to their PBS controls) (Fig. 4C and 4D). In correlation with microabscess formation, there was no significant difference in ALT and AST levels in the WT and C3aR-/- mice 1 day post-infection; however, the C3aR-/- mice had 1.7-fold higher levels of ALT (P = 0.006) and 2-fold higher levels of AST (P = 0.0004) in their sera on day 3 post-infection compared to the WT mice.

Figure 4. C3aR-/- mice have more liver damage following LM infection.

WT and C3aR-/- mice were infected i.v. with 1 × 10⁵ LM or PBS, and 1 day and 3 days post-infection, (A) the large lobe of the liver was removed and embedded in paraffin. Sections were stained with hematoxylin and eosin. Sections are representative of at least 3 mice per group. Magnification is 100x. (B) The liver abscess area was quantitated from H&E images taken at 50x as described in Materials and Methods and is presented as mean liver abscess area ± SEM. The data was quantitated from 3-5 mice per group per time point. The data shown is from one experiment but is representative of two experiments. Sera levels of ALT (C) and AST (D) were also measured. The data in (C, D) are pooled from two independent experiments and is presented as mean Units/L ± SEM, with n = 8-11 mice per group per time point. ** P = 0.006; *** P ≤ 0.0005 by t test.

The spleens of the WT and C3aR-/- infected mice also exhibited increased pathology compared to their non-infected PBS controls (Fig. 5A). As visualized by H&E staining, the spleens of the WT and C3aR-/- mice 1 day post-infection exhibited similar levels of white pulp damage, which was primarily localized to the area surrounding the central arteries known as the periarteriolar lymphoid sheaths. By day 3 post-infection, the cellular destruction observed in the white pulp of the spleens of both strains of mice had increased significantly; however, at this time point, the damage was much greater in the C3aR-/- spleens compared to the WT spleens (Fig. 5A). A 200x magnification view of day 3 spleen H&E images are shown in the supplemental data (Supplemental Fig. 1). TUNEL staining was performed to determine if the cellular destruction observed in the white pulp of the spleens of the WT and C3aR-/- infected mice was due to apoptosis. Indeed, and in accord with the H&E staining, on day 1 post-infection the WT and C3aR-/- mice both showed similar levels of TUNEL⁺ staining (Fig 5B and 5C). Also in accord with the H&E staining, the C3aR-/- mice had 2-fold more TUNEL⁺ staining compared to the WT mice on day 3 post-infection (Fig. 5B and 5C), suggesting that the C3aR-/- splenic cells are more susceptible to LM-induced apoptosis than are WT splenic cells.

Figure 5. Increased spleen pathology and apoptosis in the C3aR-/- mice.

WT and C3aR-/- mice were infected i.v. with 1 × 10⁵ LM or PBS, and 1 day and 3 days post-infection, the spleens were removed and embedded in paraffin. Sections were stained with H&E (A) or with the DeadEnd™ Colorimetric TUNEL System (B). Sections are representative of at least 3 mice per group. Magnification is 100x. (C) The TUNEL staining shown in (B) was quantitated as described in Materials and Methods and is presented as mean percent TUNEL⁺ staining ± SEM. The data was quantitated from 3-4 mice per group per time point. The data shown is from one experiment but is representative of two experiments. * P = 0.027 by t test. RP = red pulp; WP = white pulp; CA = central artery.

To quantify the cell populations in the spleens and to determine which cells are lost in the C3aR-/- mice, splenocytes harvested on days 1 and 3 post-infection were counted and stained for cell surface markers. There was no significant difference in the total cell numbers in uninfected C3aR-/- and WT spleens (Fig. 6A). There were also no differences in the number of cells of the different cell populations in the spleens of the untreated WT and C3aR-/- mice (data not shown). On day 1 post-infection, there was a 27% decrease in total live splenocytes in the C3aR-/- mice compared to the WT mice (P = 0.005) (Fig. 6A), which resulted in 38% fewer CD19⁺ cells (P = 0.002), 31% fewer CD11c⁺ cells (P = 0.019), and 27% fewer CD4⁺ cells (P = 0.030) (Fig. 6B). Other cell populations such as CD8⁺, NK1.1⁺, and F4/80⁺ cells were also reduced in the C3aR-/- mice compared to the WT mice, but the differences were not statistically significant (Fig. 6B). On day 3 post-infection, the C3aR-/- mice had 42% fewer live cells compared to the WT mice (P = 0.002) (Fig. 6A), which resulted in the following reductions: 41% fewer Ly-6G⁺ cells (P = 0.007), 39% fewer F4/80⁺ cells (P = 0.016), 37% fewer CD11c⁺ cells (P = 0.026), 43% fewer CD19⁺ cells (P = 0.002), 38% fewer NK1.1⁺ cells, 45% fewer CD4⁺ cells (P = 0.0002), and 45% fewer CD8⁺ cells (P = 0.0001) (Fig.6C). The F4/80⁺ cells in the WT mice and the C3aR-/- mice were equally activated as determined by similar CD11b and MHC class II expression on both days 1 and 3 post-infection (data not shown).

Figure 6. C3aR-/- mice have fewer cells in their spleens following LM infection.

WT and C3aR-/- mice were infected i.v. with 1 × 10⁵ LM, and 1 day and 3 days post-infection, the spleens were removed and total viable cells were counted (A). Spleens from untreated mice were also harvested as controls. Splenocytes from day 1 (B) and day 3 (C) LM-infected mice were stained with various cell surface markers to determine cell populations. The gating strategies are included in the supplemental data (Supplemental Fig. 2). The data are pooled from two independent experiments and is presented as mean cells per spleen ± SEM. n = 13 for the untreated mice, n = 4-6 for the infected mice. * P ≤ 0.030; ** P ≤ 0.007; *** P ≤ 0.0002 by t test.

Fas expression, Caspase-3 activity, and Bcl-2 expression in WT and C3aR-/- splenocytes

To examine in more depth why the C3aR-/- splenocytes were more susceptible to LM-induced cell death, we looked at surface Fas expression, caspase-3 activity, and intracellular Bcl-2 expression. Fas (CD95) is a death receptor that is expressed on both myeloid and lymphoid cells, and ligation of this receptor leads to apoptosis (43). As shown in Figure 7A, the C3aR-/- mice had 10% more Fas⁺ cells (P = 0.0004) in their spleens compared to the WT mice on day 3 post-infection. In addition, the C3aR-/- splenocytes had a higher intensity of Fas⁺ staining, as evidenced by a higher geometric mean fluorescence intensity (P = 0.032), compared to the WT splenocytes on day 3 post-infection (Fig. 7B). In accord with the TUNEL staining data, there was no difference in Fas expression between WT and C3aR-/- splenocytes on day 1 post-infection or in untreated mice (data not shown). Ligation of Fas ultimately leads to activation of caspase-3, which is an effector caspase involved in apoptosis (43). We measured caspase-3 activity in WT and C3aR-/- cell-free splenocyte lysates. As shown in Figure 7C, on day 3 post-infection, the C3aR-/- splenocyte lysates had 2-fold higher caspase-3 activity compared to the WT splenocyte lysates (P = 0.034). There was no difference in caspase-3 activity in PBS-treated WT and C3aR-/- splenocyte lysates (data not shown). Bcl-2, which is an anti-apoptotic molecule, has been shown to inhibit activation of caspase-3 (44). Since the C3aR-/- mice had higher caspase-3 activity in their splenocytes on day 3, we suspected that they had lower Bcl-2 expression. As shown in Figure 7D, on day 3 post-infection the C3aR-/- mice had 10% fewer Bcl-2⁺ cells in their spleens compared to the WT mice (P = 0.004). In addition, the C3aR-/- splenocytes had a lower intensity of Bcl-2⁺ staining, as evidenced by a lower geometric mean fluorescence intensity (P = 0.011), compared to the WT splenocytes on day 3 post-infection (Fig. 7E). There was no difference in Bcl-2 expression between WT and C3aR-/- splenocytes on day 1 post-infection or in untreated mice (data not shown). These data, together with the TUNEL staining results, indicate that the increased destruction of the lymphoid and myeloid cells in the spleens of the C3aR-/- infected mice was in large part due to LM-induced apoptosis attributed to higher expression of the Fas death receptor, greater caspase-3 activity, and reduced Bcl-2 expression.

FIGURE 7. C3aR-/- mice have higher Fas expression and caspase-3 activity and lower Bcl-2 expression in their spleens following LM infection.

WT and C3aR-/- mice were infected i.v. with 1 × 10⁵ LM, and 3 days post-infection, the spleens were removed and processed for flow cytometry (A-B) and (D-E) or the isolated splenocytes were lysed and assayed for caspase-3 activity (C). (A, B) Splenocytes were stained with an antibody to Fas. The data are pooled from two independent experiments, with n = 5-6 mice per group, and is presented as (A) mean percent Fas⁺ cells ± SEM and (B) geometric mean fluorescence intensity ± SEM. (C) Splenocytes were lysed and the cell-free homogenates were assayed for caspase-3 activity. The data are pooled from two independent experiments, with n = 6-7 mice per group, and is presented as mean absorbance (405 nm) ± SEM. (D, E) Splenocytes were fixed and permeabilized and stained for intracellular Bcl-2 expression. The data are pooled from two independent experiments, with n = 6-8 mice per group, and is presented as (D) mean percent Bcl-2⁺ cells ± SEM and (E) geometric mean fluorescence intensity ± SEM. * P ≤ 0.034; ** P = 0.004; *** P = 0.0004 by t test.

Caspase inhibition is protective in the C3aR-/- mice both in vitro and in vivo

Given the significantly higher caspase-3 activity in the splenocytes of the infected C3aR-/- mice (Fig. 7C), we wanted to determine if the use of a broad caspase inhibitor such as Z-VAD-FMK would protect these C3aR-/- cells from LM-induced cell death. Splenocytes harvested from naïve C3aR-/- mice were pre-treated with 20 μM Z-VAD-FMK or vehicle for 45 min prior to the addition of LM (MOI = 0.01). In the vehicle-treated group at 20 hours post-infection, there were 30% fewer live C3aR-/- splenocytes compared to the WT splenocytes (P = 0.001)(Fig. 8). The use of Z-VAD-FMK increased the number of live C3aR-/- splenocytes by 35% (P = 0.004) compared to the vehicle-treated C3aR-/- cells (Fig. 8). In the vehicle-treated group at 44 hours post-infection, there were 40% fewer live C3aR-/- splenocytes compared to the WT splenocytes (P = 0.003)(Fig. 8). The use of Z-VAD-FMK increased the number of live C3aR-/- splenocytes by 44% (P = 0.0003) compared to the vehicle-treated C3aR-/- cells (Fig. 8). We observed no statistically significant difference in cell survival between WT and C3aR-/- splenocytes that were not infected with LM (data not shown).

Figure 8. C3aR-/- splenocytes are more susceptible to LM-mediated killing in vitro.

WT and C3aR-/- splenocytes from naïve mice were pre-treated with either vehicle or 20 μM Z-VAD-FMK and then were infected with LM (MOI = 0.01) for 20 h or 44 h. The number of viable cells were enumerated using a hemacytometer and trypan blue exclusion. The data is presented as mean cells/ml ± SEM. ** P ≤ 0.004; *** P = 0.0003 by t test. The experiment was performed in triplicate and is representative of three independent experiments.

To determine if the use of Z-VAD-FMK would also provide protection to LM infection in vivo, C3aR-/- mice were pre-treated with 10 mg/kg Z-VAD-FMK or vehicle i.p. two hours prior to LM infection. On day 3 post-infection, the C3aR-/- mice pre-treated with Z-VAD-FMK had 6.8-fold fewer bacteria in their livers compared to the C3aR-/- mice pre-treated with vehicle, and the C3aR-/- mice pre-treated with Z-VAD-FMK had similar liver CFUs as WT mice (Fig. 9A). The liver pathology of the C3aR-/- mice was significantly improved by pre-treating the mice with Z-VAD-FMK, with nearly 2-fold lower microabscess area in the livers compared to the vehicle-treated mice, and the C3aR-/- mice pre-treated with Z-VAD-FMK had similar liver pathology and microabscess area as WT mice (Fig. 9B and 9C). Pre-treatment of the C3aR-/- mice with Z-VAD-FMK had varying effects on serum cytokine production on day 3 post-infection. IFN-γ, TNF-α, and IL-6 production was not significantly impacted by Z-VAD-FMK pre-treatment (data not shown). However, G-CSF, MCP-1, and IL-10 levels were significantly reduced in the infected C3aR-/- mice pre-treated with Z-VAD-FMK, bringing the levels down to those found in the infected WT mice (Fig. 9D). Pre-treatment of the C3aR-/- mice with Z-VAD-FMK resulted in less destruction of the splenic white pulp compared to the vehicle-treated mice (Fig. 10A). In addition, pre-treatment of the C3aR-/- mice with Z-VAD-FMK also significantly reduced the number of TUNEL⁺ cells in the spleens (Fig. 10B and 10C). In conclusion, pre-treatment of the C3aR-/- mice with a pan caspase inhibitor provided significant protection during LM infection, which indicates that excessive caspase activation in the C3aR-/- mice renders these mice more susceptible than WT mice to LM infection.

Figure 9. Pre-treatment with Z-VAD-FMK decreases liver bacterial burden, liver pathology, and serum cytokine production.

C3aR-/- mice were pre-treated with Z-VAD-FMK or vehicle i.p. and then were infected i.v. with 1 × 10⁵ LM. Three days post-infection the livers were removed and homogenized under sterile conditions (A). The data is presented as mean CFU per liver (Log₁₀) ± SEM, with n = 3-4 mice per group. *** P = 0.0006 by ANOVA with the Tukey post-test (** between WT and C3aR-/- + Vehicle; *** between C3aR-/- + Vehicle and C3aR-/- + Z-VAD-FMK.) (B) The large lobe of the liver was removed and embedded in paraffin. Sections were stained with H&E. Sections are representative of 3 mice per group. Magnification is 100x. (C) The liver abscess area was quantitated from H&E images taken at 50x as described in Materials and Methods and is presented as mean liver abscess area ± SEM. The data was quantitated from 3 mice per group. * P = 0.013 by ANOVA with the Tukey post-test (* between WT and C3aR-/- + Vehicle and between C3aR-/- + Vehicle and C3aR-/- + Z-VAD-FMK). (D) G-CSF, MCP-1, and IL-10 were measured by ELISA in the sera of the mice on day 3 post-infection. The data is presented as mean pg/ml ± SEM, with n = 3-4 mice per group. For G-CSF *** P < 0.0001 by ANOVA with the Tukey post-test (*** between WT and C3aR-/- + Vehicle and between C3aR-/- + Vehicle and C3aR-/- + Z-VAD-FMK); For MCP-1 ** P = 0.004 by ANOVA with the Tukey post-test (** between WT and C3aR-/- + Vehicle and between C3aR-/- + Vehicle and C3aR-/- + Z-VAD-FMK); For IL-10 ** P = 0.006 by ANOVA with the Tukey post-test (* between WT and C3aR-/- + Vehicle and between C3aR-/- + Vehicle and C3aR-/- + Z-VAD-FMK). In all cases (A, C, and D), there was no significant difference between WT and C3aR-/- + Z-VAD-FMK by the Tukey post-test.

Figure 10. Pre-treatment with Z-VAD-FMK decreases spleen pathology and apoptosis.

C3aR-/- mice were pre-treated with Z-VAD-FMK or vehicle i.p. and then were infected i.v. with 1 × 10⁵ LM. Three days post-infection, the spleens were removed and embedded in paraffin. Sections were stained with H&E (A) or with the DeadEnd™ Colorimetric TUNEL System (B). Sections are representative of 3-4 mice per group. Magnification is 100x. (C) The TUNEL staining shown in (B) was quantitated as described in Materials and Methods and is presented as mean percent TUNEL⁺ staining ± SEM. The data was quantitated from 3-4 mice per group. ** P = 0.006 by ANOVA with the Tukey post-test (* between WT and C3aR-/- + Vehicle; ** between C3aR-/- + Vehicle and C3aR-/- + Z-VAD-FMK; no significant difference between WT and C3aR-/- + Z-VAD-FMK).

Discussion

The results of this study show for the first time a critical role for C3aR in providing host protection to systemic LM infection. Absence of C3aR in infected mice resulted in increased mortality, increased bacterial burden, increased liver damage, and elevated destruction of immune cells of the spleen important in LM clearance, including neutrophils, macrophages, dendritic cells, and T cells.

LM has been known for some time to cause extensive depletion of lymphocytes in the periarteriolar lymphoid sheaths located in the spleens of infected mice (45), with TUNEL⁺ cells appearing as early as 24 hours after infection (46). Only live LM, and not heat-killed LM, is able to induce apoptosis in infected mice (46), and listeriolysin O, a secreted virulence factor of LM, is the apoptogenic molecule (47). Lymphocyte apoptosis is so detrimental to early clearance of LM that SCID mice (48, 49) and RAG2-/- mice (13), which lack lymphocytes, have reduced bacterial burdens in their spleens and livers during early infection, but because CD4⁺ and CD8⁺ cells are necessary for ultimate clearance of LM, these mice become chronically infected (48, 49). In addition to lymphocytes, LM also causes apoptosis of macrophages (50), neutrophils (51), and dendritic cells (52), which are all important for clearance of LM infection.

Fas is a death receptor that belongs to the TNF superfamily of membrane receptors and is expressed on multiple cell types including monocytes, macrophages, dendritic cells, neutrophils, T cells, and B cells (53). Ligation of Fas results in cleavage of procaspase-8 to active caspase-8, which leads to cleavage and activation of effector caspases, such as caspase-3, resulting in apoptosis (53). In our investigations reported here, C3aR-/- mice had greater expression of Fas on their splenocytes on day 3 post-LM infection compared to WT mice (see Fig. 7). In addition, splenocytes from the C3aR-/- mice had 2-fold more caspase-3 activity than splenocytes from WT mice on day 3 post-infection (see Fig. 7), which correlated with the increase in TUNEL⁺ staining in the spleens of the C3aR-/- mice. We also showed that splenocytes from the C3aR-/- mice had decreased expression of Bcl-2 (see Fig. 7), an anti-apoptotic molecule which is known to inhibit activation of caspase-3 (44).

Collectively, these findings lead us to conclude that the increased susceptibility to LM systemic infection in the absence of C3aR was due largely to increased apoptosis of the important lymphoid and myeloid cells of the C3aR-/- infected mice, which was caused by increased Fas expression and decreased Bcl-2 expression leading to excessive activation of the caspase cascade. In support of this conclusion, C3aR-/- mice pre-treated with the pan caspase inhibitor Z-VAD-FMK had significant protection against LM infection compared to vehicle-treated C3aR-/- mice, as demonstrated by reduced bacterial burden in their livers, improved liver and spleen pathologies, and decreased TUNEL⁺ staining in their spleens. Furthermore, the pan caspase inhibitor reduced the bacterial burden, tissue pathologies, and TUNEL⁺ staining in the infected C3aR-/- mice to levels at or below that observed in the WT infected mice (see Figures 9 and 10).

Interestingly, there have been several recent reports indicating that C3a is of critical importance in tissue regeneration by providing pro-survival/anti-apoptotic signals to numerous cell types in various tissues and in different biological micro-environments (54). For example, when C3aR-deficient mice are subjected to liver injury, they exhibit severe hepatic apoptosis, preventing normal liver regeneration (55-57). C3a has also been reported to induce the complete regeneration of the embryonic chick retina from stem/progenitor cells present in the eye (58), and in a mouse model of retinal injury, C3aR-/- mice showed increased retinal degeneration compared to WT mice (59). In mouse models C3a has been shown to promote both basal and ischemia-induced neurogenesis (60) and provide neuronal protection during neonatal hypoxic-ischemic brain injury (61). In the examples discussed above, the mechanism(s) by which C3a provides pro-survival and regenerative signals remains poorly understood, although recent publications have begun to reveal important insights. In the liver, the general theme proposed from current data is that C3a protects regenerating hepatocytes from apoptotic death indirectly by acting as an upstream mediator that increases STAT-3 dependent IL-6 and TNF-α gene expression (likely by activation of C3aR expressing Kupffer cells) (55). Both TNF-α and IL-6 are crucial regulators of the priming phase of liver regeneration, and IL-6 in particular is a major pro-survival factor for regenerating hepatocytes via the PI3K/AKT/mTOR pathway (62). A similar upstream impact on STAT-3 dependent IL-6 and TNF-α expression and protection of stem/progenitor cells was also proposed for C3a-mediated regeneration of embryonic chick retinas (58). Other recent investigations have reported that locally produced C3a (possibly by intracellular activation of C3 in T cells or in antigen presenting cells) provides pro-survival signals to both activated and naïve T cells (63, 64). Similar to regenerating hepatocytes, the pro-survival effect of C3a on T cells is thought to occur through the PI3K/AKT/mTOR pathway (64).

At first, it may be tempting to speculate that C3a provides anti-apoptotic signals during LM infection via the mechanisms discussed above. However, LM-induced apoptosis is dependent on bacterial expression of the pore-forming toxin listeriolysin O, and thus the mechanisms by which C3a provides anti-apoptotic signals during LM infection may be quite different than those during cellular regeneration. For example, compared to infected WT mice, the C3aR-/- LM-infected mice contained significantly increased levels of IL-6 and TNF-α. Despite these increases in IL-6 and TNF-α, splenocytes of the C3aR-/- mice were highly sensitive to LM-induced apoptosis, indicating that the protection provided by C3a against LM-induced apoptosis was not dependent on upstream regulation of IL-6/TNF-α expression. Our studies do demonstrate that the impact of C3a-mediated inhibition of LM-induced apoptosis involves modulation of Fas and Bcl-2 expression and caspase-3 activation. Since numerous different cell types from both lymphoid and myeloid lineages were highly susceptible to cell death during LM infection in the absence of C3aR, it is unlikely that C3a inhibits LM-induced apoptosis through direct C3a/C3aR signaling (or at least not on all cells). Instead, it is more likely due to upstream C3a regulation of other molecular factors that modulate Fas and Bcl-2 expression and caspase-3 activation. Extensive future investigations will be required to identify these molecular factors and to elucidate the mechanisms by which C3a modulates their expression.

In summary, this study reveals a previously unknown, yet important, function of C3aR in providing host defense against LM systemic infection. The inability of C3aR-/- mice to control and clear LM was not caused by defective macrophage activation or reduction in expression of IFN-γ or TNF-α. Instead, TUNEL staining together with Fas, active caspase-3, and Bcl-2 expression data indicate that the increased susceptibility of C3aR-/- mice to LM infection was largely caused by increased LM-induced apoptosis of cells that are required for ultimate clearance of LM, including neutrophils, macrophages, dendritic cells, and T cells. These findings reveal an unexpected function of C3a/C3aR signaling during the host immune response that suppresses Fas expression and caspase-3 activity while increasing Bcl-2 expression, thereby providing protection to both myeloid and lymphoid cells against LM-induced apoptosis.

Abbreviations

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BHI

brain heart infusion

C3aR

C3a receptor

LM

Listeria monocytogenes

MOI

multiplicity of infection