ABSTRACT
CD38, adenosine-5′-diphosphate-ribosyl cyclase 1, is a multifunctional enzyme, expressed on a wide variety of cell types. CD38 has been assigned diverse functions, including generation of calcium-mobilizing metabolites, cell activation, and chemotaxis. Using a murine Listeria monocytogenes infection model, we found that CD38 knockout (KO) mice were highly susceptible to infection. Enhanced susceptibility was already evident within 3 days of infection, suggesting a function of CD38 in the innate immune response. CD38 was expressed on neutrophils and inflammatory monocytes, and especially inflammatory monocytes further upregulated CD38 during infection. Absence of CD38 caused alterations of the migration pattern of both cell types to sites of infection. We observed impaired accumulation of cells in the spleen but surprisingly similar or even higher accumulation of cells in the liver. CD38 KO and wild-type mice showed similar changes in the composition of neutrophils and inflammatory monocytes in blood and bone marrow, indicating that mobilization of these cells from the bone marrow was CD38 independent. In vitro, macrophages of CD38 KO mice were less efficient in uptake of listeria but still able to kill the bacteria. Dendritic cells also displayed enhanced CD38 expression following infection. However, absence of CD38 did not impair the capacity of mice to prime CD8+ T cells against L. monocytogenes, and CD38 KO mice could efficiently control secondary listeria infection. In conclusion, our results demonstrate an essential role for CD38 in the innate immune response against L. monocytogenes.
INTRODUCTION
The ADP-ribosyl cyclase 1 CD38 is expressed on a wide variety of hematopoietic and nonhematopoietic cell types. In the mouse, hematopoietic cells with CD38 expression include B cells, neutrophils, monocytes, NK cells, and certain T cell subsets (1, 2). CD38 is present on the cell surface, but it can also be found in various intracellular organelles (3, 4). A recent report indicates that a fraction of CD38 molecules may also display an inside-out orientation in cellular membranes with the catalytic domain facing the cytosol (5).
At neutral or alkaline pH, CD38 is a nucleotide-metabolizing enzyme that hydrolizes NAD (NAD+) or cyclic ADP-ribose (cADPR) to adenosine diphosphoribose (ADPR) (6). Additionally, it can generate cADPR from NAD+ (7). cADPR is a calcium-mobilizing metabolite that acts on ryanodine receptors (RyR) inducing calcium release from intracellular stores distinct from those controlled by inositol triphosphate receptors (IP3R) (8, 9). In contrast, ADPR is a calcium-mobilizing metabolite that activates the transient receptor potential cation channel subfamily M member 2 (TRPM2) ion channel (10). CD38 can also generate nicotinic acid adenine dinucleotide phosphate (NAADP+) from NAD(P)+ in the presence of nicotinic acid (NA) at acidic pH (11). NAADP+ is a calcium messenger targeting acidic organelles like lysosomes (12–14). Eventually, CD38 can hydrolyze NAADP+ to ADP-ribose 2′-phosphate (ADPRP) (15).
CD38 also functions as a plasma membrane signaling receptor in leukocytes. On B cells, CD38 acts as a coreceptor and modulates B cell receptor (BCR) signals (16). On neutrophils and monocytes, CD38 cooperates with several chemotactic receptors, such as CXCR4, CCR7, and N-formyl-methionylleucyl-phenylalanine (fMLF) sensors like N-formyl peptide receptor 1 (FPR1) (17). Via generation of cADPR from NAD+, CD38 causes intracellular calcium release, which synergizes with signals derived from chemotactic receptors in inducing migration of neutrophils and monocytes toward sites of inflammation (17, 18). Cooperation of CD38 with chemokine receptors has also been demonstrated for dendritic cells (DC). Absence of CD38 causes impaired migration of DC and can result in inefficient T cell priming (19, 20).
CD38 is also the main hydrolase of extracellular NAD+. Stressed or damaged tissue cells can release NAD+, which can act as a potential damage-associated molecular pattern (DAMP). Extracellular NAD+ can further induce apoptosis in T cells via ART2-mediated ADP-ribosylation of P2X7 (1, 21, 22). Consequently, CD38 can limit these processes by reducing extracellular NAD+ levels (21).
The Gram-positive bacterium Listeria monocytogenes can cause severe disease in immunosuppressed individuals, and in pregnant women, infection of the fetus can lead to abortion or to high fatality rates in neonates (23). In mice, infection with L. monocytogenes provokes a rapid activation of the innate immune system, which is essential for the restriction of bacterial replication. In particular, production of the proinflammatory cytokines tumor necrosis factor alpha (TNF-α) and gamma interferon (IFN-γ) and recruitment of inflammatory monocytes to sites of infection are crucial for the early control of L. monocytogenes (24, 25). As an immune escape mechanism, L. monocytogenes leaves the phagosome of macrophages and is then able to replicate in the cytoplasm. Due to this intracellular localization, L. monocytogenes induces strong CD8+ T cell responses, which are central for clearance of the bacteria from the organism and for effective protection after reinfection (26).
In this study, we characterized the role of CD38 in the innate phase of immune responses against L. monocytogenes. We found that CD38 is essential for the control of infection. Closer analyses revealed that neutrophils, inflammatory monocytes, and DC expressed CD38 and that expression was upregulated during infection. Absence of CD38 resulted in altered recruitment of neutrophils and inflammatory monocytes to sites of listeria infection. In vitro, the capability of peritoneal macrophages from CD38 knockout (KO) mice to take up listeria was reduced compared to that of wild-type (WT) macrophages, but cells were still able to kill bacteria. Despite an impaired innate response, CD38 deficiency did not prevent priming of CD8+ T cells and generation of immunological memory. In summary, our results assign CD38 a crucial function in the innate immune response against L. monocytogenes.
MATERIALS AND METHODS
Mice.CD38 KO (19) and OT-I (27) mice were backcrossed for 12 generations to BALB/c and/or C57BL/6 mice. CD90.1 congenic C57BL/6 mice (B6.PL-Thy1a/CyJ) were obtained from Jackson, Bar Harbor, ME. All mice were bred under specific-pathogen-free conditions. Experiments were performed according to state guidelines.
Infections, antibiotic treatment, and titer determination.Mice were infected intravenously (i.v.) with 5 × 103, 5 × 104, or 1 × 105 Listeria monocytogenes wild-type strain EGD bacteria as described previously (28) or 1 × 105 Listeria monocytogenes bacteria expressing ovalbumin (LmOVA) (29). Bacterial inocula were controlled by plating serial dilutions on tryptic soy broth (TSB) agar plates. For analysis of T cell responses, infected mice were treated with 2 mg of ampicillin in 200 μl of phosphate-buffered saline (PBS) intraperitoneally (i.p.) on days 2 and 3 or with 2 mg of ampicillin i.p. on day 2 and 1 g/liter of ampicillin in drinking water from day 2 until day 9 postinfection to clear the listeria infection (30). For determination of bacterial burdens in spleens and livers, mice were killed, organs were homogenized in double-distilled water (ddH2O), serial dilutions of homogenates were plated on TSB agar plates, and colonies were counted after 24 h of incubation at 37°C.
Cell isolations.Cells from spleens were obtained by mashing the disintegrated organs through cell sieves into PBS followed by erythrocyte lysis with ACK lysing buffer (155 mM NH4Cl, 10 mM KHCO3, 100 μM EDTA [pH ∼7.2]). For liver lymphocyte isolation, the liver cell suspension was additionally separated using a Percoll gradient (Biochrom AG, Berlin, Germany) before erythrocyte lysis. All cells were counted using a hemocytometer.
Cell culture.Splenocytes from C57BL/6 WT and CD38 KO mice were restimulated with ovalbumin peptide (OVA257–264; SIINFEKL) (JPT Peptide Technologies GmbH, Berlin, Germany) at 2 × 106 cells/ml in fully supplemented medium (RPMI 1640, 5% fetal calf serum [FCS], glutamine, pyruvate, nonessential amino acids, gentamicin, 2-mercaptoethanol) for 4.5 h at 37°C. Brefeldin A (BFA; Sigma) was added at 10 μg/ml for the last 3.5 h of culture to prevent intracellular protein transport.
Macrophage function assay.BALB/c WT and CD38 KO mice were injected i.p. with 2 ml of thioglycolate medium (BD Biosciences). Five days later, peritoneal macrophages were obtained by peritoneal lavage with RPMI 1640 medium. The frequency of CD11b+ CD45+ macrophages was determined by flow cytometry. Cells were seeded in 24-well plates with 1 × 105 cells per well in antibiotic-free fully supplemented RPMI 1640 medium and were rested overnight at 37°C. On the next day, cells were washed and infected with 3 × 106 EGD bacteria/well for 1 h at 37°C (multiplicity of infection [MOI] = 30). Subsequently, peritoneal macrophages were washed and incubated in fully supplemented RPMI 1640 medium with 200 μg/ml of gentamicin at 37°C to kill remaining extracellular listeria. After 30 min, either the culture medium was diluted to reach a gentamicin concentration of 50 μg/ml or cells were lysed to determine the efficacy of infection (0 h). Cells were further analyzed after 2 and 4 h. For determination of bacterial titers, cells were washed three times with antibiotic-free medium, macrophages were lysed with 0.1% Triton X-100 in ddH2O, and serial dilutions of lysates were plated on TSB agar. All values were determined in triplicate wells.
Flow cytometric analysis.For surface staining, cells were first incubated with 10 μg/ml of 2.4G2 (anti-FcγRII/III; BioXCell, West Lebanon, NH) and 1:100 normal rat serum (NRS; Jackson Laboratories, Bar Harbor, ME) in PBS–0.5% bovine serum albumin (BSA)–0.05% sodium azide (NaN3) to minimize unspecific antibody binding. Staining was performed on ice with the following fluorochrome-conjugated monoclonal antibodies (MAbs) according to standard methods: anti-CD3ε clone 145-2C11, anti-CD11c clone HL3, anti-Ly6C clone AL-21 (all BD Biosciences, San Jose, CA), anti-CD4 clone RM4-5, anti-CD8α clone 53-6.7, anti-CD38 clone 90, anti-CD45 clone 30-F11, anti-CD90.1 clone His51, anti-Ly6GC clone RB6-8C5, anti-major histocompatibility complex class II (anti-MHC-II) clone M5/114.15.2 (all eBioscience, San Diego, CA), anti-CD11b clone M1/70, anti-CD44 clone IM7, anti-CD62L clone MEL-14, anti-CD90.2 clone 53-2.1, anti-CD127 clone A7R34, and anti-KLRG1 clone 2F1 (all BioLegend, San Diego, CA). Cells (1 × 106 to 2 × 106) were measured on a CantoII flow cytometer (BD Biosciences), and data were analyzed with the FlowJo software (Treestar, Ashland, OR). Debris, doublets, and 4′,6-diamidino-2-phenylindole-positive (DAPI+) dead cells were excluded from analysis.
Intracellular staining.After surface staining, cells were stained using a fixable dead cell stain (Pacific orange succinimidyl ester; Invitrogen Life Technologies) to exclude dead cells from analysis.
For cytokine staining, cells were washed with PBS and fixed for 20 min with PBS–2% paraformaldehyde at room temperature. Thereafter, cells were washed with PBS–0.2% BSA, permeabilized with PBS–0.1% BSA–0.3% saponin (Sigma), and incubated in this buffer with 1:100 NRS. After 5 min, fluorochrome-conjugated anti-IFN-γ MAb clone XMG1.2 (BioLegend) was added. After a further 20 min on ice, cells were washed with PBS and measured by flow cytometry.
For Ki-67 staining, cells were fixed with Foxp3 fixation buffer (eBioscience) for 1 h at 4°C and permeabilized with Foxp3 perm buffer (eBioscience) with 1:100 NRS. After 5 min, fluorochrome-conjugated anti-Ki-67 MAb clone SolA15 (eBioscience) was added. After a further 20 min on ice, cells were washed with PBS and measured by flow cytometry.
T cell adoptive-transfer assay.Spleen cells from (OT-I × B6.PL-Thy1a/CyJ)F1 mice were isolated as described above and depleted of CD4+, CD11b+, CD19+, F4/80+, and MHC-II+ cells by magnetically activated cell sorting (MACS) using anti-CD4-fluorescein isothiocyanate (FITC) clone RM4-5 (BioLegend), anti-CD11b-FITC clone M1/70 (BD Biosciences), anti-CD19-FITC clone 1D3 (eBioscience), anti-F4/80-FITC clone BM8 (BioLegend), anti-MHC-II-FITC clone OX-6 (Serotec, Kidlington, United Kingdom) MAb, and anti-FITC microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's protocol. Relative frequencies of Vα2+ (clone B20.1; BD Biosciences) CD8α+ OT-I T cells were determined by flow cytometry. Cells were washed twice with PBS and counted, and 5 × 104 CD8α+ OT-I T cells were adoptively transferred via i.v. injection into the lateral tail vein.
Statistical analysis.All statistical analysis was performed with Prism software (GraphPad Software Inc., La Jolla, CA). In the graphs, each mouse is presented by a single symbol. A line indicates the median value for all animals within one experimental group. Differences between groups of mice were analyzed by Mann-Whitney U test. A P value of <0.05 was considered significant (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant).
RESULTS
Increased susceptibility of CD38-deficient mice to listeria infection.To test the relevance of CD38 during the innate phase of infection, we intravenously (i.v.) infected BALB/c WT and CD38 KO mice with 5 × 104 Listeria monocytogenes strain EGD bacteria, a dose approximately 5-fold above the 50% lethal dose (LD50) for WT BALB/c mice (T. Lischke and H.-W. Mittrücker, unpublished data). Sixty hours postinfection, CD38 KO mice showed signs of severe illness (rough fur, reduced motility, and encrusted eyes), while WT mice still had a healthy appearance. Determination of listeria titer revealed approximately 10-fold-elevated EGD titers in spleens and 100-fold-higher titers in livers of CD38 KO mice (Fig. 1A). CD38 KO mice on the C57BL/6 background also showed significantly enhanced susceptibility, although differences in listeria titers were less pronounced (Fig. 1B). Finally, even with a low infection dose (5 × 103 bacteria), CD38 KO mice on a BALB/c background harbored higher listeria titers in spleen and liver 60 h postinfection (Fig. 1C). Thus, CD38 KO mice had an increased susceptibility to listeria infection.
Listeria titers in spleen and liver 60 h postinfection. On day 0, WT and CD38 KO mice were infected i.v. with EGD bacteria. Sixty hours postinfection, EGD titers in spleen and liver were determined. (A) BALB/c WT and CD38 KO mice infected i.v. with 5 × 104 EGD bacteria. (B) C57BL/6 WT and CD38 KO mice infected i.v. with 5 × 104 EGD bacteria. (C) BALB/c WT and CD38 KO mice infected i.v. with 5 × 103 EGD bacteria. There were 5 to 10 mice per group; median bars are shown. One representative experiment out of two each is shown.
CD38 is expressed by neutrophils, inflammatory monocytes, and dendritic cells.We then analyzed the expression levels of CD38 on the surfaces of various immune cell types (flow cytometric gating strategies in Fig. S1 in the supplemental material). When comparing BALB/c WT and CD38 KO mice, we found that CD38 was expressed in the steady state by CD11b+ Ly6GChi Ly6Clo neutrophils, CD11b+ Ly6GClo Ly6Chi inflammatory monocytes, and CD11c+ MHC-II+ DC in the spleen (Fig. 2). CD38 was expressed by all B cells but only by subpopulations of CD4+ and CD8+ T cells, i.e., ∼50% of effector (CD44hi CD62Llo) T cells (data not shown) (31). Upon EGD infection, CD38 expression even increased on neutrophils, inflammatory monocytes, and DC in the spleen. Neutrophils and inflammatory monocytes in the liver also expressed CD38, which increased after listeria infection (Fig. 2). Thus, CD38 was expressed on cell types responsible for early control of listeria infection, and expression even increased in response to infection.
CD38 expression on neutrophils, inflammatory monocytes, and dendritic cells at sites of listeria infection. On day 0, BALB/c WT and CD38 KO mice were infected i.v. with 5 × 104 EGD bacteria. Sixty hours postinfection, expression of CD38 on neutrophils, inflammatory monocytes, and dendritic cells in spleen and liver was determined. There were 3 to 5 mice per group; one representative histogram overlay is shown. One representative experiment out of two is shown.
Recruitment of neutrophils and inflammatory monocytes to sites of infection is altered in CD38-deficient mice.It has been demonstrated that CD38 is required for migration of neutrophils and macrophages to sites of infection (31). Thus, we analyzed the recruitment of neutrophils and inflammatory monocytes into spleen and liver, the main sites of listeria replication, as well as the frequencies of these cells in the bone marrow and the blood (Fig. 3A). Under steady-state conditions, relative frequencies of neutrophils and inflammatory monocytes in all organs were comparable in BALB/c WT and CD38 KO mice. Upon EGD infection, we observed a drastic increase of neutrophils and inflammatory monocytes in spleens and livers of BALB/c WT mice (60 h postinfection). This increase was significantly lower in the spleens of CD38 KO mice. However, in the livers of CD38 KO mice, the relative frequency of neutrophils was comparable to that in WT mice, and the relative frequency of inflammatory monocytes was even higher than in WT mice. No differences in relative frequencies of neutrophils and inflammatory monocytes between WT and CD38 KO mice were observed in the bone marrow and the blood (Fig. 3A). The total numbers of neutrophils and inflammatory monocytes calculated per spleen and liver confirmed the frequency data (Fig. 3B). Relative frequencies and total numbers of DC were comparable in the spleens of WT and CD38 KO mice 60 h postinfection (Fig. 3C). Thus, recruitment of neutrophils and inflammatory monocytes, but not of DC, to sites of listeria infection was altered in CD38-deficient mice.
Recruitment of neutrophils and inflammatory monocytes to sites of listeria infection. On day 0, BALB/c WT and CD38 KO mice were infected i.v. with 5 × 104 EGD bacteria. Sixty hours postinfection, relative frequencies (A) and absolute numbers (B) of neutrophils and inflammatory monocytes in spleen and liver were determined. For bone marrow and blood, only relative frequencies are shown. (C) Relative frequencies and absolute numbers of dendritic cells in the spleen. There were 3 to 5 mice per group; median bars are shown. One representative experiment out of two is shown.
CD38-deficient macrophages are less efficient in uptake of listeria.To determine whether CD38 can directly influence the antimicrobial functions of macrophages, we conducted in vitro infection assays. Thioglycolate-elicited macrophages were isolated from the peritoneum of BALB/c WT and CD38 KO mice. A total of 1 × 105 peritoneal cells (approximately 50% CD11b+ CD45+ macrophages) were infected with 3 × 106 EGD bacteria for 1 h at 37°C (MOI = 30). Subsequently, macrophages were washed and medium containing gentamicin was added to kill extracellular bacteria. After 30 min at 37°C (0 h), bacterial titers in the macrophages were determined as an estimate for bacterial uptake. To analyze the listeriocidal activity of macrophages, bacterial titers were determined after 2 and 4 h. We observed a reduced ability of CD38 KO macrophages compared to WT macrophages to take up listeria (Fig. 4). However, CD38 KO macrophages could efficiently kill listeria.
Peritoneal macrophages of CD38 KO mice are less efficient in uptake of listeria but can efficiently kill the bacteria. Thioglycolate-elicited peritoneal macrophages were rested overnight, and 1 × 105 peritoneal cells (∼50% CD45+ CD11b+ macrophages) were then infected with 3 × 106 EGD bacteria (MOI = 30) for 1 h at 37°C. Cells were washed and incubated with medium containing gentamicin to kill remaining extracellular bacteria. After 30 min, macrophages were lysed, and listeria titers in lysates were determined to estimate the uptake of bacteria (0 h). Bacterial titers in macrophages were further determined after 2 and 4 h. All titers were determined in triplicates. Mean bars are shown. One representative experiment out of two is shown.
Unimpaired CD8+ T cell priming in CD38-deficient mice upon listeria infection.The high expression level of CD38 on dendritic cells and the further upregulation following infection could indicate a role for CD38 on DC in T cell priming. Such a function for CD38 on DC has been proposed for the priming of CD4+ T cells following hapten and protein immunization (20). To study the capacity of CD38 KO mice to induce an adaptive CD8+ T cell response against L. monocytogenes, we used an adoptive transfer system. A total of 5 × 104 OVA-specific OT-I CD90.1+ CD8α+ T cells were adoptively transferred into CD90.1− C57BL/6 WT or CD38 KO recipients on day −1, which were i.v. infected with 1 × 105 Listeria monocytogenes bacteria expressing ovalbumin (LmOVA) the next day. To ensure survival of mice and enable the analysis of the adaptive OVA-specific CD8α+ T cell response, infected mice were treated with 2 mg of ampicillin in 200 μl of PBS i.p. on days 2 and 3 postinfection to resolve the listeria infection. At this time point, priming of CD8+ T cells has sufficiently progressed to allow for a regular CD8+ T cell response to L. monocytogenes (30). The CD8+ T cell response was analyzed on day 7 postinfection (Fig. 5A). Surprisingly, the relative frequencies of OT-I CD8α+ T cells in spleens and livers of CD38 KO mice were slightly elevated compared to those in WT mice, and the total numbers of OT-I CD8α+ T cells in spleens and livers were significantly increased in CD38 KO mice compared to WT mice (due to increased organ sizes [data not shown]) (Fig. 5B). We also analyzed the differentiation status of OT-I CD8+ T cells in the spleens of infected mice (Fig. 5C). OT-I CD8+ T cells from WT and CD38 KO mice showed similar frequencies of CD44hi CD62Llo effector CD8+ T cells and of KLRG1+ CD127− CD44hi CD62Llo short-lived effector CD8+ T cells, as well as of Ki-67+ proliferating cells. After antigen-specific in vitro stimulation, a comparable frequency of CD8+ T cells produced the effector cytokine IFN-γ. Thus, priming of OT-I CD8+ T cell was not impaired in CD38 KO mice compared to WT mice. CD8+ T cells showed normal differentiation and even enhanced accumulation in CD38 mice.
Expansion and differentiation of CD8+ T cells at sites of listeria infection. (A) Experimental setup. On day −1, C57BL/6 WT and CD38 KO mice were adoptively transferred with 5 × 104 OT-I CD8+ T cells. On day 0, all mice were infected i.v. with 1 × 105 LmOVA bacteria. On days 2 and 3, all mice were injected i.p. with 2 mg of ampicillin (in PBS). (B) Seven days postinfection, relative frequencies and absolute numbers of OT-I CD8+ T cells in spleen and liver were determined. (C) Seven days postinfection, the differentiation status of OT-I CD8+ T cells in the spleen was determined by means of CD44, CD62L, CD127, and KLRG1 expression. Proliferating OT-I CD8+ T cells were detected by staining of Ki-67. IFN-γ producing OT-I CD8+ T cells were detected after 4.5 h in vitro restimulation with pOVA. There were 6 or 7 mice per group; median bars are shown. One representative experiment out of two is shown.
Effective control of secondary listeria infection in CD38 KO mice.As CD8+ T cell priming was not impaired in CD38 KO mice, we wanted to test the development of functional endogenous CD8+ T cell memory in CD38 KO mice. Therefore, we infected BALB/c WT and CD38 KO mice with 5 × 103 EGD bacteria. To ensure survival of mice and enable the analysis of the memory response, listeriae were cleared by ampicillin treatment on day 2 postinfection. At this time point, priming of CD8+ T cells has sufficiently progressed to allow for the generation of CD8+ T cell memory to L. monocytogenes (30). After 6 weeks, the primary infected mice as well as naive controls were i.v. infected with 1 × 105 EGD bacteria, 10-fold the LD50 for WT BALB/c mice (Lischke and Mittrücker, unpublished). Bacterial titers were determined 48 h later (Fig. 6). In mice without prior infection, we observed high listeria titers in spleen and liver. As expected, organs of CD38 KO mice harbored significantly higher numbers of bacteria than organs from WT mice. Following secondary infection, both WT and CD38 KO mice showed profoundly reduced bacterial titers in spleen and liver and in contrast to the primary infection, there was no difference in titers between organs from WT and CD38 KO mice. Thus, when protected during primary listeria infection by antibiotics application, CD38 KO mice showed a normal control of secondary listeria infection, indicating an unimpaired generation of endogenous T cell memory.
Listeria titers in spleen and liver 48 h after primary and secondary infection. On day 0, BALB/c WT and CD38 KO mice were infected i.v. with 5 × 103 EGD bacteria. Forty-eight hours postinfection, mice were treated with a single i.p. injection of 2 mg of ampicillin and received 1 g/liter of ampicillin in drinking water for 1 week. After 6 weeks, primary infection of previously untreated mice or secondary infection of previously infected mice was carried out i.v. with 1 × 105 EGD bacteria. Forty-eight hours after primary or secondary infection, EGD titers in spleen and liver were determined. There were 7 or 8 mice per group; median bars are shown.
DISCUSSION
Here we demonstrate that CD38-deficient mice are highly susceptible to L. monocytogenes infection. Enhanced bacterial loads in spleen and liver were already evident at 2 to 3 days postinfection, suggesting that CD38 is essential for the early innate phase of the immune response. Our observations are in accordance with results from a Streptococcus pneumoniae infection model, where CD38 KO mice also show enhanced susceptibility (17, 18). CD38 is expressed on neutrophils and macrophages in the steady state (1, 2) and further upregulated on these cells upon infection (18). Our results confirm the results for neutrophils in the context of listeria infection. Inflammatory monocytes were also CD38+ and showed a strong upregulation following infection. The high expression levels suggest an important function of CD38 for these cells. Since granulocytes and particularly inflammatory monocytes are critical for the innate control of L. monocytogenes (32, 33), an absence of CD38 expression on these cells might be responsible for the enhanced susceptibility of CD38 KO mice.
Lack of CD38 resulted in altered migration patterns of neutrophils and inflammatory monocytes. There was reduced accumulation of neutrophils and inflammatory monocytes in the spleen but surprisingly normal or even stronger accumulation of these cells in the liver. In contrast, we observed no differences in bone marrow and blood. Thus, the mobilization of neutrophils and inflammatory monocytes from the bone marrow as well as the accumulation of cells in the liver appeared to be CD38 independent, whereas the recruitment to the spleen was CD38 dependent. In this regard, the L. monocytogenes infection model differs from other infection models in which CD38 KO mice display a more general impairment of neutrophil and macrophage recruitment to sites of infection (18, 34). The discrepancy might be due to differences in the mechanisms of recruitment of inflammatory cells, such as the dependency on different chemotactic factors, in the infection models. It has also been shown that absence of CD38 does not generally impair chemotaxis but rather affects only a subset of chemotactic receptors and that the requirement of these receptors for CD38 even differs between different cell types (2). Therefore, differences in the mechanisms controlling mobilization of cells from the bone marrow and regulating accumulation of cells in listeria-infected tissues could be responsible for the differential CD38 requirement. Recently, Shi et al. demonstrated that accumulation of inflammatory monocytes to the listeria-infected liver is independent of chemokines and mainly relies on adhesion molecules expressed on endothelial cells of the infected liver (35). This observation could explain the CD38-independent recruitment of inflammatory monocytes and probably also of neutrophils to the infected livers of CD38 mice.
Despite normal or even enhanced accumulation of neutrophils and inflammatory cells to the liver, we still observed profoundly elevated bacterial levels in the liver of CD38 KO mice. Thus, impaired control of infection appears to be not simply a consequence of reduced recruitment of these cells to sites of infection. CD38 could be necessary for the correct positioning of granulocytes and inflammatory monocytes within the liver, and an absence of CD38 could result in impaired abscess formation and compromised containment of bacteria at initial sites of replication. Absence of CD38 could also affect mechanisms independent of cell migration. Although neutrophils of CD38 KO mice showed normal phagocytosis and superoxide production (17), CD38 could be required for other bactericidal mechanisms of granulocytes and inflammatory monocytes. In fact, we were able to show a reduced ability for CD38-deficient macrophages in uptake of listeria in vitro. However, CD38-deficient macrophages could kill the ingested bacteria. Finally, CD38 might be necessary for the correct function of other cells involved in the innate response to L. monocytogenes, such as NK cells or innate T cells (γδ-T cells or NK T cells) not characterized in our study.
Interestingly, TRPM2 KO mice infected with L. monocytogenes display several features observed in our study in CD38 KO mice (36, 37). These mice are highly susceptible to L. monocytogenes infection due to an impaired innate immune response. TRPM2 KO mice show reduced accumulation of inducible nitric oxide synthase (iNOS)-positive monocytes, which likely resemble inflammatory monocytes in our study. However, macrophages are able to kill listeria when infected in vitro. TRPM2 is a Ca2+-permeable membrane channel which binds the NAD+-metabolite ADP-ribose (ADPR). As ADPR formation can also be elicited via CD38, TRPM2 could be a main target of CD38 function in the innate immune response to L. monocytogenes.
It has been suggested that CD38 expression on DC is necessary for their migration and consequently the priming of CD4+ T cells (20). We also detected expression of CD38 on DC and profound upregulation of the molecule on these cells during listeria infection. However, we did not observe major changes in the number of these cells in spleens of either WT or CD38 KO mice upon infection. To study T cell priming, we transferred CD38-sufficient CD8+ T cells into WT or CD38 KO mice and monitored their response to listeria infection. This transfer model should exclude altered responses due to the absence of CD38 expression on CD8+ T cells and restrict our analysis to the role of CD38 on antigen-presenting cells. In contrast to other studies (20), we did not observe impaired CD8+ T cell priming in CD38 KO mice. We even detected higher numbers of CD8+ T cells in CD38 KO mice, which might be a consequence of higher listeria titers at the time point of antibiotic treatment (38). All other analyzed attributes, including proliferation, differentiation markers, and cytokine production, were identical for CD8+ T cells activated in WT and CD38 KO mice. In line with these results, there was no difference between WT and CD38 KO mice in the ability to control a secondary listeria infection. Our results indicate that deficiency of CD38 on antigen-presenting cells does not interfere with CD8+ T cell priming or the generation of T cell memory. Features that might determine the dependency of effective antipathogen immune responses on CD38 include the type and application route of the antigen but also the localization and environment of antigen presentation.
In conclusion, our results identify CD38 as central element of the innate immune response against L. monocytogenes in the mouse. Considering the multiple functions of CD38, it might be interesting to target either CD38 or its downstream signaling pathways in inflammatory diseases to modulate (auto)immune responses in humans as well.
ACKNOWLEDGMENTS
This work was supported by DFG (Deutsche Forschungsgemeinschaft) grant MI476/3-1 and SFB841 to H.-W.M., as well as SFB877 to F. K.-N.
We thank Frances Lund for providing the CD38 KO mice, Hao Shen for making LmOVA available for us, and the animal facility staff for help with the experiments.
Disclosures: The authors have no financial conflict of interest.
FOOTNOTES
- Received 15 March 2013.
- Returned for modification 13 April 2013.
- Accepted 31 July 2013.
- Accepted manuscript posted online 26 August 2013.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.00340-13.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
