Abstract
Bacterial sepsis is a serious life-threatening condition caused by an excessive immune response to infection. B-1 cells differ from conventional B-2 cells by their distinct phenotype and function. A subset of B-1 cells expressing CD5, known as B-1a cells, exhibits innate immune activity. Here we report that B-1a cells play a beneficial role in sepsis by mitigating exaggerated inflammation through a novel mechanism. Using a mouse model of bacterial sepsis, we found that the numbers of B-1a cells in various anatomical locations were significantly decreased. Adoptive transfer of B-1a cells into septic mice significantly attenuated systemic inflammation and improved survival, whereas B-1a cell–deficient CD19−/− mice were more susceptible to infectious inflammation and mortality. We also demonstrated B-1a cells produced ample amounts of IL-10 which controlled excessive inflammation and the mice treated with IL-10–deficient B-1a cells were not protected against sepsis. Moreover, we identified a novel intracellular signaling molecule, cAMP-response element binding protein (CREB), which serves as a pivotal transcription factor for upregulating IL-10 production by B-1a cells in sepsis through its nuclear translocation and binding to putative responsive elements on IL-10 promoter. Thus, the benefit of B-1a cells in bacterial sepsis is mediated by CREB and the identification of CREB in B-1a cells reveals a potential avenue for treatment in bacterial sepsis.
Introduction
Sepsis occurs due to an exaggerated immune response to invading pathogens (1). In the United States, hospitalization rates for sepsis have increased since the start of this century (2–4); hence there is an urgent unmet medical need for the development of novel treatments for sepsis. Host immune cells recognize bacterial pathogen-associated molecular patterns via their TLRs to rapidly produce a broad array of cytokines, chemokines, and other proteins to counteract invading pathogens. However, these molecules can trigger inflammation and impair tissue function, a form of “friendly fire” (1, 5). This early onset of “cytokine storm,” often contributing to the lethality of sepsis, can be opposed by the anti-inflammatory cytokine IL-10 (6).
Innate like B cell function in sepsis is a new area of research (7–9). The role of B cells in inducing an early innate immune response has been demonstrated in T and B cell–deficient Rag-1−/− mice, which displayed an inadequate early innate immune response and reduced survival during sepsis (8). The treatment of these mice with B cells improved sepsis survival, suggesting beneficial roles for B cells in the outcome of sepsis (8). Murine B cells consist of several subpopulations, which include follicular cells, marginal zone B cells, and B-1 cells (9). Although the role of follicular and marginal zone B cells (B-2 cells) as key participants in the early inflammatory cytokine response during sepsis has recently been demonstrated (8, 10), the role of B-1 cells in sepsis has been implicated but still remains to be fully explored. The phenotype of murine B-1 cells is CD45 (B220)lo, surface IgM (sIgM)hi, sIgDlo, CD23lo/−, CD19hi, CD43+, and CD11b+ in the peritoneal cavity (PerC) (9, 11, 12). B-1 cells can be further subdivided into either B-1a (CD5+) or B-1b (CD5) cells (11, 12). B-1a cells present in the PerC, spleen, and bone marrow (BM) are the major source of natural IgM at steady state, which provides an early defense against infections. In contrast, B-1b cells produce adaptive Abs in response to T cell–independent type 2 Ags (9, 13–15). It has also been reported that mouse peritoneal B-1a cells produce a large amount of IL-10 even in the absence of stimulation (16–19). Among other subsets of B cells with immunoregulatory properties, regulatory B cells, also known as B10 cells, are capable of generating IL-10 to attenuate various inflammatory diseases (9, 20–22). Recently, a novel type of B cell, innate response activator (IRA), has been described. These B cells originate from peritoneal B-1a cells and relocate to the spleen, producing GM-CSF upon LPS stimulation to protect the host against sepsis (7). IRA B cells are a differentiated form of B-1a cells that differ from B-1a cells, as well as B10 cells, in terms of GM-CSF and IL-3, but not IL-10 production, after LPS stimulation (7, 23). IRA B cells exhibit both protective and deleterious roles in sepsis depending on GM-CSF and proinflammatory cytokine IL-3 production, respectively (23).
Recent progress has been made for elucidation of the role of B-1a cells in protecting against various inflammatory and autoimmune diseases (9). Natural IgM secreted from B-1a cells not only neutralizes invading pathogens but also recognizes and speeds removal of dying cells leading to suppression of uncontrolled inflammation and autoimmunity (9, 16). Beyond natural IgM, several immunoregulatory molecules secreted by B-1 cells are also shown to attenuate acute and chronic inflammatory diseases (9). Our current study dealt with the efficacy of B-1a cells to protect animals from polymicrobial sepsis by controlling hyperimmune responses. Here, we adopted a clinically relevant model and assessed parameters for monitoring sepsis pathophysiology after treatment with B-1a cells isolated from normal animals. Recruiting B-1a cell–deficient animals, we further confirmed the importance of B-1a cells in protecting against sepsis. To our knowledge, our study also for the first time revealed the involvement of cAMP-response element binding protein (CREB) in B-1a cells, which was upregulated to increase the production of IL-10 for attenuating inflammation and organ injury, and finally improving survival during sepsis. Thus, B-1a cells might serve as a novel area of translational research during sepsis.
Materials and Methods
Animals
Wild-type (WT) C57BL/6 mice were purchased from Taconic (Albany, NY). B6.129P2(C)Cd19tm1(cre)cgn/J (CD19−/−), B6.129P2-Il10tm1cgn/J (IL-10−/−), and GFP transgenic (Tg) [C57BL/6-Tg(UBC-GFP)30Scha/J] mice were purchased from The Jackson Laboratory. TLR4 knockout (TLR4−/−) mice were obtained as a gift from K. J. Tracey (The Feinstein Institute for Medical Research, Manhasset, NY). For all of the experiments 8–10 wk-old male mice with 21–28 g body weight were used. The animals were housed in a temperature-controlled room on a 12h light/dark cycle and fed a standard laboratory diet. Animals were randomly assigned to the sham, vehicle control, and treatment groups. The number of animals estimated in each group was based on our previous publications on animal models of sepsis (24, 25). Animal studies were conducted in an unblinded fashion. All age-matched healthy animals were included, whereas those which died during the surgical operation or had an uneven body weight were excluded from the analysis. All experiments were approved by the Institutional Animal Care and Use Committee at the Feinstein Institute for Medical Research.
Sepsis induction
Mice were anesthetized with isoflurane and underwent cecal ligation and puncture (CLP) (24). Briefly, a 1.5-cm incision was made to the abdominal wall, and the cecum was exposed and ligated 0.5 cm from the tip with 4-0 silk suture. A 22-gauge needle was used to make one puncture through and through to the distal cecum, extruding a small amount of fecal material. The cecum was replaced into the abdominal cavity, and the abdomen was closed in two layers with running 6-0 silk suture. The sham mice underwent the same procedure with the exception that their cecum was neither ligated nor punctured. Animals were resuscitated with 1 ml of normal saline s.c.
Isolation of PerC macrophages and coculture with B-1a cells
Mice were anesthetized by isoflurane inhalation, and then peritoneal cells were isolated by washing with ice-cold PBS, with 5% FBS. Collected peritoneal cells were washed once with cold PBS by centrifugation at 400 × g for 10 min at 4°C and the resulting pellet was suspended in culture medium consisting of RPMI 1640 (Invitrogen) supplemented with 25 mM HEPES, 2 mM glutamine, 10% FBS (Solon, OH), penicillin (100 IU/ml), and streptomycin (100 IU/ml). Peritoneal macrophages were then allowed to adhere in 10-cm culture plates for 2 h at 37°C in 5% CO2. Nonadherent cells were removed by washing with prewarmed culture medium. Adhered PerC macrophages were then mechanically detached from the plate using a rubber scraper and counted. In a 48-well flat-bottom cell culture plate, a total of 1.5 × 105 PerC macrophages and an equal number of B-1a cells in 300 μl of RPMI medium with 10% FBS were cocultured. The cocultured cells were treated with either isotype control Ab (20 μg/ml) or anti–IL-10 neutralizing Ab (20 μg/ml) and then stimulated by PBS as vehicle or LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination. After 20 h, culture supernatants were removed and analyzed for TNF-α and IL-1β production by ELISA.
Flow cytometry
B-1a cells were identified based on their surface phenotype as described previously (26, 27). Cells present in the PerC, spleen, and BM of C57BL/6 mice were stained with PE-B220 (clone RA3-6B2), PE-Cy7-CD23 (clone B3B4), PerCP-Cy5.5-CD5 (clone 53-7.3), APC-IgM (clone RMM-1), and Pacific Blue-IgD (clone 11-26c-2a), all purchased from BD Biosciences (San Jose, CA). Stained cells were analyzed on a BD LSRFortessa cell analyzer (BD Biosciences) and at least 3 × 104 cells were collected and were analyzed with FlowJo software (TreeStar). Compensation was adjusted using unstained and single color stained controls for each flow experiment. Fluorescent-labeled isotype Abs were used as Ab control.
Cell sorting and adoptive transfer
B-1a cells in the peritoneal washouts with phenotype B220loCD23−CD5int were sort purified using a BD Biosciences Influx instrument (26). As a non–B-1a cell control for subsequent in vivo and in vitro experiments splenic B-2 cells with surface phenotype B220hiCD5−CD23hi were sorted. Postsort analysis of the PerC B-1a and splenic B-2 cell populations showed each to be ≥98% pure. Sort-purified B-1a cells or B-2 cells were washed with PBS twice and then suspended in PBS for adoptive transfer. At the time of CLP operation, 5 × 105 B-1a cells suspended in 150 μl of PBS were delivered into the PerC and the abdominal wound was closed with running 4-0 silk suture. As vehicle negative control, 150 μl of PBS was injected into the abdomen of CLP-operated mice. The equal amounts of B-2 cells in 150 μl of PBS were also delivered into the CLP-operated mice i.p., which served as a cellular control group. The animals were allowed food and water ad libitum, and at 20 h after CLP operation and cell transfer the animals were euthanized. Blood and peritoneal washout samples were collected for various ex vivo analyses.
Analysis of organ injury markers, cytokines, and chemokines
Blood was drawn from mice by cardiac puncture using 1-ml syringes rinsed with an anticoagulant heparin solution. Blood samples were centrifuged at 2000 × g for 15 min to collect plasma and then either analyzed for injury parameters immediately, or stored at −80°C. Alanine aminotransferase (ALT), aspartate aminotransferase (AST), and lactate dehydrogenase (LDH) levels were measured using assay kits from Pointe Scientific (Canton, MI). IL-6, IL-1β, IL-10, TNF-α, and MIP-2 levels in the plasma and PerC washouts were quantified using mouse ELISA kits (BD Biosciences, Franklin Lakes, NJ).
Bacterial cultures
For bacterial culturing, whole blood without anticoagulant was diluted in sterile PBS at a range of 1:10–1:100, and 100 μl of diluted blood was cultured on 5% sheep blood agar plates (BD Diagnostic Systems, Sparks, MD). Peritoneal fluid was collected after washing the cavity with 5 ml of sterile PBS. Peritoneal fluid was diluted 1:10–1:100 in sterile PBS, and 100 μl of each dilution was cultured on 5% sheep blood agar plates. Plates were incubated at 37°C for 24 h, and the numbers of colonies were counted and expressed as CFUs per milliliter of blood or CFUs per cavity.
Survival study
Mice were subjected to cecal ligation at 0.5 cm from the tip with 4-0 silk suture and a single puncture to the distal part of the ligated cecum using a 22-gauge needle. Immediately after the CLP operation, 5 × 105 B-1a or B-2 cells in 150 μl of PBS or equal volume of PBS alone were delivered i.p. After wound closure, the mice were injected with 500 μl of the antibiotic imipenem (0.5 μg/kg), and 500 μl of normal saline s.c. The animals were then followed daily for 10 d.
Detection of phosphorylated CREB in peritoneal B-1a cells
PerC washouts were collected from sham or CLP-operated animals. A total of 5 × 106 peritoneal cells were first surface stained by anti-mouse PE-B220, PE-Cy7-CD23, PerCP-Cy5.5-CD5, and Pacific Blue-CD19 Abs (BD Biosciences) to identify B-1a cells. Peritoneal cells were then fixed and permeabilized using BD Cytofix/Cytoperm Fixation/Permeabilization kit (BD Biosciences), followed by intracellular staining using APC–IL-10 (BD Biosciences) and rabbit anti-mouse monoclonal phosphorylated CREB (pCREB) (Ser133) Ab (Cell Signaling Technology, Beverly, MA). After incubating the cells with FITC-conjugated anti-rabbit secondary Ab, the samples were subjected to flow cytometric data acquisition and then analysis using FlowJo software.
Assessment of IL-10 in B-1a cells treated with CREB inhibitor in vitro
B-1a cells (1.5 × 105 cells/ml) were pretreated with DMSO as vehicle or CREB inhibitor (catalog number 217505; EMD Millipore, Darmstadt, Germany) at 20 μM for 30 min, and then the cells were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by the assessment of IL-10 production by ELISA.
Detection of nuclear translocation of pCREB by imaging flow cytometry
A total of 2 × 106 cells isolated from the PerC were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by surface staining with FITC-labeled anti-B220 and PE-Cy5–labeled anti-CD5 and intracellular staining with anti-pCREB primary Abs. The cells were reacted with a secondary Ab recognizing rabbit anti-pCREB (BV480 polyclonal goat anti-rabbit IgG; BD Biosciences) and stained the nucleus with 3 μM of DAPI (BioLegend). The samples were then subjected to acquisition in ImageStreamX Mark II Imaging Flow Cytometer (EMD Millipore, Billerica, MA) and analysis by INSPIRE and IDEAS software (EMD Millipore).
Chromatin immunoprecipitation and quantitative real-time PCR assay
In order to determine the binding of pCREB in IL-10 promoter of murine B-1a cells, a total of 1 × 106 sorted B-1a cells in 1 ml of RPMI medium with 10% FBS were plated in a 12-well cell culture plate and stimulated with either PBS as vehicle or LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h. Sonicated and cross-linked chromatin extracted from the treated B-1a cells using a chromatin immunoprecipitation (ChIP) assay kit (Agarose ChIP assay kit, catalog number 26156; Pierce Biotechnology, Rockford, IL) was immunoprecipitated with anti-mouse monoclonal pCREB (Ser133) Ab (Cell Signaling Technology). DNA was recovered by reversing chromatin cross-linking and extracted using phenol-chloroform extraction and DNA clean-up column and reagents (Agarose ChIP assay kit; Pierce Biotechnology). DNA was eluted from the column in 50 μl PCR-grade water and used for a conventional as well as quantitative real-time PCR using the primer forward: 5′-GTCTACCCGACAGCACAGAG-3′ and reverse: 5′-CTAGGAGCATGTGGCTCTGG-3′ of murine IL-10 promoter flanking CREB.
Statistical analysis
Figures and data analyses were carried out using SigmaPlot 12.5 graphing and statistical software (Systat Software, San Jose, CA). Data represented in the figures are expressed as mean ± SE. All in vitro and in vivo studies were repeated at least three times. ANOVA was used for one-way comparison among multiple groups and the significance was determined by the Student–Newman–Keuls (SNK) test. The paired two-tailed Student t test was applied for two-group comparisons. For analyzing survival assay data, the Kaplan–Meier log-rank statistical tool was used. Significance was considered for p ≤ 0.05 between experimental groups.
Results
Alterations in the status of B-1a cells at different compartments during sepsis
Sepsis-induced alterations in immune cell frequencies and numbers have been demonstrated in several studies (24, 28). However, the status of B-1a cell frequency during sepsis is unknown. In the current study, B-1a cells within various compartments of sham and CLP-operated animals were identified based on their surface phenotype CD23−B220loCD5intIgM+IgD− (Fig. 1A). Here, we chose samples from 20-h CLP mice to represent our gating strategy for identification of B-1a cells, because at this time point the decrease in the B-1a cell frequencies and numbers at various organs compared with sham-operated mice were clearly noticeable. To study the dynamics of alterations in the frequencies and numbers of B-1a cells, we isolated cells from PerC, spleen, and BM at 5, 10, and 20 h after CLP. Sepsis resulted in a significant reduction in the frequencies of B-1a cells as a proportion of total lymphocytes in the PerC by mean values of 43.8, 58.2, and 75.4% at 5, 10, and 20 h after CLP, respectively, whereas their frequency was not noticeably altered in the spleen or BM as compared with sham animals (Fig. 1B–D). We found significant reductions in the numbers of B-1a cells in PerC by mean values of 38.7, 50.8, and 55.3% at 5, 10, and 20 h after CLP, respectively (Fig. 1B). Interestingly, our results demonstrated a trend in the decrease of B-1a cell numbers at 5 and 10 h after CLP in the spleen and BM, whereas significant reductions in their numbers were noticed at 20 h after CLP as compared with sham-operated mice (Fig. 1C, 1D).
Status of B-1a cell frequencies and numbers in sepsis. (A) At 20 h after CLP, cells were isolated from PerC, spleen, and BM. The cells were then stained with anti-mouse PE-B220, PE-Cy7-CD23, PerCP/Cy5.5-CD5, APC-IgM, and Pacific Blue-IgD Abs, and then subjected to flow cytometric detection of B-1a cells frequencies. (B–D) Dynamics of B-1a cell frequencies and numbers in PerC, spleen, and BM at different time points after CLP operation are shown. Data are expressed as means ± SE (n = 6–9 mice per group) and compared by one-way ANOVA and SNK method obtained from three independent experiments (*p < 0.05 versus sham-operated animals).
Status of B-1a cell frequencies and numbers in sepsis. (A) At 20 h after CLP, cells were isolated from PerC, spleen, and BM. The cells were then stained with anti-mouse PE-B220, PE-Cy7-CD23, PerCP/Cy5.5-CD5, APC-IgM, and Pacific Blue-IgD Abs, and then subjected to flow cytometric detection of B-1a cells frequencies. (B–D) Dynamics of B-1a cell frequencies and numbers in PerC, spleen, and BM at different time points after CLP operation are shown. Data are expressed as means ± SE (n = 6–9 mice per group) and compared by one-way ANOVA and SNK method obtained from three independent experiments (*p < 0.05 versus sham-operated animals).
To study cellular translocation, B-1a cells were sort purified from the PerC of GFP Tg mice and then transferred immediately into the WT mice via i.p. at the time of CLP operation. At 20 h after CLP, we noticed the adoptively transferred GFP+ B-1a cells were translocated into the spleen and mesenteric lymph nodes (MLNs) (Supplemental Fig. 1). Thus, the decrease in the numbers of peritoneal B-1a cells during sepsis could be due to their translocation into adjacent lymphoid organs.
Replenishment of B-1a cell deficiency attenuates systemic inflammation and improves survival in sepsis
Sorted PerC B-1a cells were adoptively transferred into syngeneic mice at the time of CLP induction. Levels of systemic injury and proinflammatory markers such as ALT, AST, LDH, IL-6, IL-1β, and MIP-2 were significantly downregulated by mean values of 27, 30, 46, 36, 87, and 39%, respectively in B-1a cell–treated mice compared with PBS-treated mice at 20 h after CLP (Fig. 2A–F). Levels of the proinflammatory cytokines IL-6 and IL-1β in peritoneal cavities were also reduced in B-1a cell–treated septic mice compared with the PBS-treated group (Fig. 2G, 2H). These results correlate with significant inhibition of the bacterial burden in blood and PerC by mean values of 60 and 64%, respectively in B-1a cell–treated mice compared with PBS-treated mice (Fig. 2I, 2J). At 20 h after CLP, mice treated with B-1a cells showed dramatic improvement in their physical appearances and behaviors (activity) as compared with the PBS-treated group (Supplemental Video). Inhibition of injury and proinflammatory markers as well as reduced bacterial loads in blood and PerC ultimately led to significant improvement of the survival outcome in B-1a cell–treated mice (75% survival at 10 d) over that of PBS-treated mice (40% survival) (Fig. 2K). These results indicate a potential beneficial role of B-1a cells during sepsis. On the other hand, mice treated with conventional B-2 cells did not show protection in terms of survival outcome (35% survival) (Fig. 2K), as well as plasma injury and inflammatory markers (data not shown).
Therapeutic potential of adoptive transfer of B-1a cells during sepsis. At the time of CLP, mice were treated with either PBS as vehicle or 5 × 105 PerC B-1a cells in 150 μl of PBS by i.p. injection. After 20 h, blood was drawn to assess plasma levels of injury markers (A) ALT, (B) AST, and (C) LDH and proinflammatory cytokines (D) IL-6, (E) IL-1β, and chemokine (F) MIP-2. Aside from blood, at 20 h after CLP PerC washouts were collected and spun down, and the supernatants were used to assess (G) IL-6 and (H) IL-1β by ELISA. Data are expressed as means ± SE (n = 6–9 mice per group) and compared by one-way ANOVA and SNK method obtained from five independent experiments (*p < 0.05 versus sham-operated group, #p < 0.05 versus PBS-treated CLP mice). Bacterial counts as expressed by CFUs in (I) blood and (J) peritoneal washouts are presented. Data are expressed as means ± SE (n = 12 mice per group for blood and n = 18 mice per group for peritoneal washouts) and compared by Student t test (*p < 0.05 versus PBS-treated CLP mice) acquired from five independent experiments. (K) A 10-d period Kaplan–Meier survival curve generated from CLP mice treated with either PBS (n = 20 mice), or B-1a cells (5 × 105 cells per mouse) isolated from peritoneal cavities (n = 20 mice), or B-2 cells (5 × 105 cells per mouse) isolated from spleens of syngeneic mice (n = 14 mice) in 150 μl of sterile PBS via i.p. is shown. *p < 0.05 versus PBS-injected mice, #p < 0.05 versus B-2 cell–treated mice determined by the log-rank test.
Therapeutic potential of adoptive transfer of B-1a cells during sepsis. At the time of CLP, mice were treated with either PBS as vehicle or 5 × 105 PerC B-1a cells in 150 μl of PBS by i.p. injection. After 20 h, blood was drawn to assess plasma levels of injury markers (A) ALT, (B) AST, and (C) LDH and proinflammatory cytokines (D) IL-6, (E) IL-1β, and chemokine (F) MIP-2. Aside from blood, at 20 h after CLP PerC washouts were collected and spun down, and the supernatants were used to assess (G) IL-6 and (H) IL-1β by ELISA. Data are expressed as means ± SE (n = 6–9 mice per group) and compared by one-way ANOVA and SNK method obtained from five independent experiments (*p < 0.05 versus sham-operated group, #p < 0.05 versus PBS-treated CLP mice). Bacterial counts as expressed by CFUs in (I) blood and (J) peritoneal washouts are presented. Data are expressed as means ± SE (n = 12 mice per group for blood and n = 18 mice per group for peritoneal washouts) and compared by Student t test (*p < 0.05 versus PBS-treated CLP mice) acquired from five independent experiments. (K) A 10-d period Kaplan–Meier survival curve generated from CLP mice treated with either PBS (n = 20 mice), or B-1a cells (5 × 105 cells per mouse) isolated from peritoneal cavities (n = 20 mice), or B-2 cells (5 × 105 cells per mouse) isolated from spleens of syngeneic mice (n = 14 mice) in 150 μl of sterile PBS via i.p. is shown. *p < 0.05 versus PBS-injected mice, #p < 0.05 versus B-2 cell–treated mice determined by the log-rank test.
Deficiency of B-1a cells in CD19−/− mice exacerbates sepsis severity and mortality
The knockout mice strains that are commonly being used to demonstrate the effects of B-1a cell–deficient/altered condition during inflammation include CD19−/− mice, Bruton’s tyrosine kinase gene (Btk) knockout or XiD mice, and sIgM−/− mice (reviewed in Ref. 9). These mice were mainly created by targeting the signaling pathways required for B cell development and function, thus might not preclude additional defects other than B-1 cell deficiency. In the current study, we recruited CD19−/− mice as one of the B-1a cell–deficient mice strains to grossly represent their observational outcomes during sepsis. CD19−/− mice were previously shown to have diminished numbers of B-1a cells in the PerC as well as spleen as compared with their basal levels in the WT counterparts (15). Sepsis in CD19−/− mice showed significantly higher levels of plasma ALT, AST, LDH, and IL-6 by mean values of 30, 30, 44, and 35%, respectively, as compared with WT mice (Fig. 3A–D). Similarly, the blood and peritoneal bacterial CFUs after CLP were noticeably higher in CD19−/− mice by mean values of 77 and 74%, respectively, as compared with WT mice (Fig. 3E, 3F). These results correlate with survival outcomes in which CD19−/− mice had a significantly reduced rate of survival (16%) as compared with WT mice, which showed a survival rate of 48% (Fig. 3G).
Status of B-1a cell–deficient CD19−/− mice during sepsis. At 20 h after CLP induced in WT and CD19−/− mice, blood was collected and plasma was subjected to assay for (A) ALT, (B) AST, (C) LDH, and (D) IL-6. Data are expressed as means ± SE (n = 6–9 mice per group) and compared by one-way ANOVA and SNK method obtained from five independent experiments (*p < 0.05 versus sham-operated group, #p < 0.05 versus WT CLP mice). Bacterial counts in terms of CFU in (E) whole blood and (F) peritoneal lavage fluid are shown. Data are expressed as means ± SE (n = 6–12 mice per group) and compared by Student t test obtained from five independent experiments (*p < 0.05 versus WT CLP mice). (G) Kaplan–Meier survival curve generated from WT and CD19−/− CLP mice during the 10-d monitoring period is shown. n = 21 mice in each group, *p < 0.05 versus WT CLP mice, determined by the log-rank test.
Status of B-1a cell–deficient CD19−/− mice during sepsis. At 20 h after CLP induced in WT and CD19−/− mice, blood was collected and plasma was subjected to assay for (A) ALT, (B) AST, (C) LDH, and (D) IL-6. Data are expressed as means ± SE (n = 6–9 mice per group) and compared by one-way ANOVA and SNK method obtained from five independent experiments (*p < 0.05 versus sham-operated group, #p < 0.05 versus WT CLP mice). Bacterial counts in terms of CFU in (E) whole blood and (F) peritoneal lavage fluid are shown. Data are expressed as means ± SE (n = 6–12 mice per group) and compared by Student t test obtained from five independent experiments (*p < 0.05 versus WT CLP mice). (G) Kaplan–Meier survival curve generated from WT and CD19−/− CLP mice during the 10-d monitoring period is shown. n = 21 mice in each group, *p < 0.05 versus WT CLP mice, determined by the log-rank test.
B-1a cells produce IL-10 during in vivo and in vitro conditions to downregulate proinflammatory cytokines
A growing body of evidence has demonstrated the protective effect of IL-10 produced by macrophages, B10, and induced regulatory T cells to control acute and chronic inflammatory diseases (9, 20, 29, 30). Although B-1a cells are one of the major cell types that produce ample amount of IL-10, the immunomodulatory function of B-1a cell–derived IL-10 in sepsis was not clearly defined. CLP mice treated with B-1a cells showed significant upregulation of IL-10 in the plasma by a mean value of 61% compared with PBS-treated mice at 20 h (Fig. 4A). Next, to determine whether the adoptively transferred B-1a cells were capable of producing IL-10 by the mice directly under septic condition, we isolated B-1a cells from WT mice and transferred them into the GFP Tg mice at the time of CLP operation. We used GFP Tg mice to distinguish the donor’s (adoptively transferred) versus recipient’s (endogenous) cells. We found that the WT B-1a cells were responsive to septic insults and capable of producing considerable amounts of IL-10 (Fig. 4B). We validated this finding by sorting B-1a cells from TLR4−/− mice, considering the fact that during sepsis TLR4-mediated pathway plays a pivotal role in initiating innate immune response to induce cytokine production. As expected, we noticed that the B-1a cells from TLR4−/− mice following their transfer into the GFP Tg septic mice were not capable of producing IL-10 (Fig. 4B), suggesting B-1a cells produced IL-10 directly through TLR4-mediated pathway in septic condition. Similarly, in vitro stimulation of B-1a cells isolated from the PerC with LPS or LPS together with PMA and ionomycin significantly upregulated IL-10 expression at the protein level (Fig. 4C). To address whether B-1a cells downregulate proinflammatory cytokine production by macrophages, an in vitro experiment was performed using PerC macrophages treated with LPS, which significantly upregulated TNF-α and IL-1β production. When macrophages treated with LPS were cocultured with B-1a cells isolated from WT mice, LPS-induced TNF-α and IL-1β productions were significantly downregulated (Fig. 4D, 4E). On the other hand, macrophages cocultured with B-1a cells obtained from IL-10−/− mice showed remarkable upregulation of TNF-α and IL-1β productions as compared with macrophages cocultured with WT B-1a cells (Fig. 4D, 4E). The role of IL-10 was further demonstrated in a coculture study with PerC B-1a cells and macrophages. When neutralizing Ab targeting IL-10 was added to PerC B-1a cell and macrophage cocultures stimulated with LPS or LPS with PMA and ionomycin in combination, production of TNF-α and IL-1β was significantly higher than with macrophages cultured without B-1a cells and higher than macrophages and PerC B-1a cells stimulated by LPS or LPS plus PMA and ionomycin plus isotype control Ab (Supplemental Fig. 2). These results clearly suggest the B-1a cell secreted IL-10 plays a pivotal immunomodulatory role during inflammation.
Assessment of IL-10 production by B-1a cells during sepsis and under in vitro condition. (A) At 20 h after CLP, the IL-10 level in plasma was assessed. Data are expressed as means ± SE (n = 9 mice per group) and compared by one-way ANOVA and SNK method obtained from three independent experiments (*p < 0.05 versus sham-operated group, #p < 0.05 versus PBS-treated CLP mice). (B) B-1a cells from PerC of WT and TLR4−/− mice were sort purified by staining the cells with PE-Cy7-B220, PE-CD23, PerCP-Cy5.5-CD5, and Pacific Blue-CD19 Abs. A total of 0.5 × 106 sorted B-1a cells were adoptively transferred into GFP Tg mice via i.p. during CLP operation. At 20 h after CLP, cells from the PerC of septic GFP Tg mice were isolated to perform intracellular staining using anti-mouse APC–IL-10 Ab. The levels of intracellular IL-10 at 20 h after CLP within GFP− B-1a cells are shown. Representative blots obtained from three independent experiments using five mice per group are presented. (C) In an in vitro experiment, a total of 1.5 × 105 PerC B-1a cells/ml were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by the assessment of IL-10 at the protein level. Data are expressed as means ± SE. The experiment was performed three independent times with n = 3–4 wells per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated group, #p < 0.05 versus LPS-treated group). In an in vitro coculture experiment, 1.5 × 105 PerC macrophages were cocultured with 1.5 × 105 B-1a cells (1:1) isolated from either WT or IL-10−/− mice, and then after stimulation with PBS or LPS (20 μg/ml) for 20 h (D) TNF-α and (E) IL-1β levels in the culture supernatant were assessed. Data are expressed as means ± SE obtained from three independent experiment with n = 3–4 per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated group, #p < 0.05 versus LPS-treated macrophages, §p < 0.05 versus LPS-treated macrophages and WT B-1a cell cocultures).
Assessment of IL-10 production by B-1a cells during sepsis and under in vitro condition. (A) At 20 h after CLP, the IL-10 level in plasma was assessed. Data are expressed as means ± SE (n = 9 mice per group) and compared by one-way ANOVA and SNK method obtained from three independent experiments (*p < 0.05 versus sham-operated group, #p < 0.05 versus PBS-treated CLP mice). (B) B-1a cells from PerC of WT and TLR4−/− mice were sort purified by staining the cells with PE-Cy7-B220, PE-CD23, PerCP-Cy5.5-CD5, and Pacific Blue-CD19 Abs. A total of 0.5 × 106 sorted B-1a cells were adoptively transferred into GFP Tg mice via i.p. during CLP operation. At 20 h after CLP, cells from the PerC of septic GFP Tg mice were isolated to perform intracellular staining using anti-mouse APC–IL-10 Ab. The levels of intracellular IL-10 at 20 h after CLP within GFP− B-1a cells are shown. Representative blots obtained from three independent experiments using five mice per group are presented. (C) In an in vitro experiment, a total of 1.5 × 105 PerC B-1a cells/ml were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by the assessment of IL-10 at the protein level. Data are expressed as means ± SE. The experiment was performed three independent times with n = 3–4 wells per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated group, #p < 0.05 versus LPS-treated group). In an in vitro coculture experiment, 1.5 × 105 PerC macrophages were cocultured with 1.5 × 105 B-1a cells (1:1) isolated from either WT or IL-10−/− mice, and then after stimulation with PBS or LPS (20 μg/ml) for 20 h (D) TNF-α and (E) IL-1β levels in the culture supernatant were assessed. Data are expressed as means ± SE obtained from three independent experiment with n = 3–4 per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated group, #p < 0.05 versus LPS-treated macrophages, §p < 0.05 versus LPS-treated macrophages and WT B-1a cell cocultures).
IL-10 produced from B-1a cells protects mice from sepsis
Next, to evaluate the in vivo effects of B-1a cells as mediated through IL-10 production, B-1a cells were sorted from the PerC of WT and IL-10−/− mice and then adoptively transferred into WT CLP mice. According to B-1a cell gating strategy while sorting them by a flow sorter, we noticed that the PerC B-1a cells isolated from WT and IL-10−/− mice showed a similar phenotypic profile (Fig. 5A), which helped us to exclude the concern of sorting different cell populations other than B-1a cells that do not resemble those of WT mice. Plasma inflammatory markers ALT, AST, LDH, and IL-6 were significantly reduced by mean values of 46, 37, 30, and 42%, respectively in mice treated with WT B-1a cells as compared with PBS-treated mice (Fig. 5B–E). However, CLP mice treated with B-1a cells isolated from IL-10−/− mice did not show any significant protection in terms of downregulating ALT, AST, LDH, and IL-6 as compared with PBS-treated mice, whereas these parameters were higher when compared with WT B-1a cell–treated CLP mice (Fig. 5B–E). Finally, the survival study showed that CLP mice treated with IL-10−/− B-1a cells exhibited significantly less survival (32% at 10 d) as compared with the WT B-1a cell–treated animals (79% survival) (Fig. 5F), which strongly suggests that the beneficial effect of B-1a cells in sepsis is mediated through IL-10 production. With LPS stimulation, peritoneal B-1a cells migrate into the spleen, where they differentiate into GM-CSF–expressing IRA B cells to protect mice from sepsis (7). We therefore intended to determine whether the IRA B cells were lacking from IL-10−/− mice, which could inflict therapeutic ineffectiveness when treating septic animals with B-1a cells derived from IL-10−/− mice. Nonetheless, we noticed IL-10−/− mice were not deficient in IRA B cells in the spleen, and their frequencies were similar to those from WT animals during sepsis (Supplemental Fig. 3).
Protective effect of murine B-1a cell–secreted IL-10 during sepsis. (A) Sorting of PerC B-1a cells from the WT and IL-10−/− mice by staining them with FITC-B220, PE-CD23, PerCP-Cy5.5-CD5, and APC-CD19 Abs. (B–E) At the time of CLP, mice were injected with PBS or B-1a cells (5 × 105 cells) isolated from either WT or IL-10−/− mice in 150 μl of sterile PBS. At 20 h after CLP, blood was drawn and the plasma was assessed for ALT, AST, LDH, and IL-6. Data are expressed as means ± SE (n = 5 mice per group) and compared by one-way ANOVA and SNK method obtained from three independent experiments (*p < 0.05 versus sham mice, #p < 0.05 versus PBS-treated CLP mice). (F) A 10-d period Kaplan–Meier survival curve generated from CLP mice treated with either PBS, or PerC B-1a cells of WT or IL-10−/− mice (5 × 105 cells) in 150 μl of sterile PBS via i.p. is shown. n = 19 mice per group, (*p < 0.05 versus WT B-1a cell–injected mice, #p < 0.05 versus IL-10−/− mice B-1a cells determined by the log-rank test).
Protective effect of murine B-1a cell–secreted IL-10 during sepsis. (A) Sorting of PerC B-1a cells from the WT and IL-10−/− mice by staining them with FITC-B220, PE-CD23, PerCP-Cy5.5-CD5, and APC-CD19 Abs. (B–E) At the time of CLP, mice were injected with PBS or B-1a cells (5 × 105 cells) isolated from either WT or IL-10−/− mice in 150 μl of sterile PBS. At 20 h after CLP, blood was drawn and the plasma was assessed for ALT, AST, LDH, and IL-6. Data are expressed as means ± SE (n = 5 mice per group) and compared by one-way ANOVA and SNK method obtained from three independent experiments (*p < 0.05 versus sham mice, #p < 0.05 versus PBS-treated CLP mice). (F) A 10-d period Kaplan–Meier survival curve generated from CLP mice treated with either PBS, or PerC B-1a cells of WT or IL-10−/− mice (5 × 105 cells) in 150 μl of sterile PBS via i.p. is shown. n = 19 mice per group, (*p < 0.05 versus WT B-1a cell–injected mice, #p < 0.05 versus IL-10−/− mice B-1a cells determined by the log-rank test).
Phosphorylation of CREB in B-1a cells is upregulated during an in vivo model of sepsis and during in vitro culture conditions
In macrophages, phosphorylation and activation of CREB following LPS stimulation results in transcriptional activation of the IL-10 promoter to increase IL-10 expression (31). In the context of B-1a cells, the association of CREB phosphorylation and upregulation of IL-10 expression has never been examined. Here, we show ∼3–4% of B-1a cells spontaneously produced IL-10 in sham mice (Fig. 6A, 6B). However, sepsis resulted in robust upregulation of IL-10 expressing B-1a cells (∼45%) as compared with sham mice (Fig. 6A, 6B). In addition, sepsis caused a significant upregulation of CREB phosphorylation in B-1a cells as compared with sham mice (Fig. 6C). Interestingly, the phosphorylation status of CREB in B-1a cells expressing IL-10, whether it was from sham or CLP-operated mice, was significantly higher as compared with the IL-10 nonexpressing B-1a cells (Fig. 6A, 6D), suggesting pCREB plays a pivotal role in IL-10 production by B-1a cells. Similarly, under in vitro conditions B-1a cells treated with LPS or LPS with PMA and ionomycin experienced significantly upregulated phosphorylation of CREB compared with PBS-treated B-1a cells (Fig. 6E, 6F). Furthermore, the critical involvement of CREB in B-1a cells for IL-10 production was confirmed by using CREB inhibitor, which inhibits CREB’s interaction to its binding protein prior to its binding to the IL-10 promoter. B-1a cells pretreated with CREB inhibitor significantly inhibited LPS or LPS plus PMA and ionomycin induced IL-10 production by mean values of 64.5 and 65.4%, respectively, compared with DMSO-treated conditions (Fig. 6G). Interestingly, B-1a cells were also found to produce a minimal level of TNF-α after stimulation with LPS or LPS plus PMA and ionomycin; however, the inhibition of CREB did not alter TNF-α production by these cells (Fig. 6H), indicating a distinct role of CREB for regulating IL-10 production by B-1a cells.
Phosphorylation of CREB in IL-10 expressing B-1a cells during CLP and under in vitro stimulation. At 20 h after CLP, PerC washouts were collected and the cells were stained by anti-mouse PE-B220, PE-Cy7-CD23, PerCP-Cy5.5-CD5, Pacific Blue-CD19, APC–IL-10, and FITC-pCREB Abs and subjected to flow cytometry to detect IL-10 and CREB phosphorylation in B-1a cells and compared with that of sham mice. An isotype control Ab for pCREB Ab (goat FITC-IgG) was included while staining the cells with IL-10 Ab to get rid of background staining. (A) Dot blots and histograms representing IL-10 and pCREB expression in B-1a cells in sham and CLP-operated animals, respectively, are shown. (B) Quantitative bar diagrammatic presentation of IL-10 expression in B-1a cells in sham and CLP-operated mice. (C) pCREB expression as measured by mean fluorescence intensity (MFI) in B-1a cells in sham and CLP-operated mice is shown. (D) pCREB expression as measured by MFI in IL-10 expressing or nonexpressing B-1a cells in sham and CLP-operated mice is shown. Data are expressed as means ± SE obtained from three independent experiments with n = 6 mice per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus sham mice or IL-10 nonexpressing B-1a cells from sham mice, #p < 0.05 versus IL-10 nonexpressing B-1a cells from CLP mice). (E) A total of 1 × 106 cells isolated from PerC were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by the surface staining with specific fluorochrome-labeled anti-B220, anti-CD23, anti-CD5, and intracellular staining with anti-pCREB Abs. The samples were then subjected to flow cytometric acquisition and analysis by FlowJo software. Representative histograms of pCREB expression in B-1a cells as indicated by the MFI of isotype control, PBS, LPS, and LPS+PMA and ionomycin-treated samples are shown. (F) A quantitative bar diagram of MFI of pCREB expression is presented. Data are expressed as means ± SE obtained from three independent experiments with n = 3–4 wells per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated cells). (G and H) B-1a cells (1.5 × 106 cells/ml) were plated in 48-well cell culture plate and pretreated with DMSO or CREB inhibitor (20 μM) for 30 min, afterward stimulated with either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h, followed by the assessment of IL-10 and TNF-α in culture supernatants at protein level. Data are expressed as means ± SE obtained from three independent experiments (n = 6 wells per group). The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated cells, #p < 0.05 versus DMSO-treated cells).
Phosphorylation of CREB in IL-10 expressing B-1a cells during CLP and under in vitro stimulation. At 20 h after CLP, PerC washouts were collected and the cells were stained by anti-mouse PE-B220, PE-Cy7-CD23, PerCP-Cy5.5-CD5, Pacific Blue-CD19, APC–IL-10, and FITC-pCREB Abs and subjected to flow cytometry to detect IL-10 and CREB phosphorylation in B-1a cells and compared with that of sham mice. An isotype control Ab for pCREB Ab (goat FITC-IgG) was included while staining the cells with IL-10 Ab to get rid of background staining. (A) Dot blots and histograms representing IL-10 and pCREB expression in B-1a cells in sham and CLP-operated animals, respectively, are shown. (B) Quantitative bar diagrammatic presentation of IL-10 expression in B-1a cells in sham and CLP-operated mice. (C) pCREB expression as measured by mean fluorescence intensity (MFI) in B-1a cells in sham and CLP-operated mice is shown. (D) pCREB expression as measured by MFI in IL-10 expressing or nonexpressing B-1a cells in sham and CLP-operated mice is shown. Data are expressed as means ± SE obtained from three independent experiments with n = 6 mice per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus sham mice or IL-10 nonexpressing B-1a cells from sham mice, #p < 0.05 versus IL-10 nonexpressing B-1a cells from CLP mice). (E) A total of 1 × 106 cells isolated from PerC were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by the surface staining with specific fluorochrome-labeled anti-B220, anti-CD23, anti-CD5, and intracellular staining with anti-pCREB Abs. The samples were then subjected to flow cytometric acquisition and analysis by FlowJo software. Representative histograms of pCREB expression in B-1a cells as indicated by the MFI of isotype control, PBS, LPS, and LPS+PMA and ionomycin-treated samples are shown. (F) A quantitative bar diagram of MFI of pCREB expression is presented. Data are expressed as means ± SE obtained from three independent experiments with n = 3–4 wells per group. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated cells). (G and H) B-1a cells (1.5 × 106 cells/ml) were plated in 48-well cell culture plate and pretreated with DMSO or CREB inhibitor (20 μM) for 30 min, afterward stimulated with either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h, followed by the assessment of IL-10 and TNF-α in culture supernatants at protein level. Data are expressed as means ± SE obtained from three independent experiments (n = 6 wells per group). The groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated cells, #p < 0.05 versus DMSO-treated cells).
pCREB is translocated into nucleus to recognize its responsive elements on IL-10 promoter in B-1a cells under in vitro stimulation
It has been shown that CREB is phosphorylated by TLR agonists and participates in the initiation of IL-10 expression in macrophages (32). By using a flow cytometric imaging tool, we qualitatively identified the location of pCREB in the cytoplasm and nucleus of the B-1a cells. Under the PBS-treated condition, pCREB was largely found in the cytoplasm, whereas upon stimulation with either LPS or LPS and PMA and ionomycin, the B-1a cells showed a remarkable translocation of cytoplasmic pCREB into the nucleus (Fig. 7A). By using a ChIP assay, we further identified potential binding sites for pCREB present in the IL-10 promoter (Fig. 7B). Chromatin precipitated with pCREB Ab showed significantly increased signal in B-1a cells stimulated with LPS or LPS plus PMA and ionomycin compared with PBS-treated control, confirming increased amounts of pCREB was recruited for increased expression of IL-10 by B-1a cells (Fig. 7B). Collectively, these findings clearly demonstrate that B-1a cells upon recognizing LPS or LPS and PMA and ionomycin upregulated intracellular pCREB, which in turn induced IL-10 production for diminishing proinflammatory responses during sepsis (Fig. 7C).
Assessment of nuclear translocation of pCREB and its binding to the IL-10 promoter in B-1a cells under in vitro stimulation. (A) A total of 2 × 106 cells isolated from PerC were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by the surface staining with FITC-labeled anti-B220, PE-Cy5–labeled anti-CD5, intracellular staining with BV480-labeled anti-pCREB Abs and nuclear staining with DAPI. The samples were then subjected to acquisition in ImageStreamX Mark II Imaging Flow Cytometer and analysis by INSPIRE and IDEAS software. Representative images of nuclear translocation pCREB in B-1a cells treated with PBS, LPS, and LPS+PMA and ionomycin are shown. (B) A total of 1 × 106 B-1a cells sorted from PerC of C57BL/6 mice were stimulated by either PBS as control or LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h. Chromatin was extracted for ChIP assay by sonication, followed by cross-linking and immunoprecipitation using anti-pCREB Ab. Conventional qualitative, as well as quantitative real-time, PCR was performed and data are presented as gel-electrophoresis blots and fold induction of amplified CREB response element containing DNA in IL-10 promoter normalized against respective input controls. Data are expressed as means ± SE obtained from three independent experiments (n = 5 samples per group). Multiple groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated cells). (C) Hypothesis scheme. In B-1a cells, LPS through TLR4-mediated pathway and PMA and Ca2+ ionophore ionomycin upregulated the phosphorylation of CREB, which in turn served as a pivotal transcription factor to induce IL-10 expression. Increased expression of IL-10 from the B-1a cells finally led to a protection against sepsis.
Assessment of nuclear translocation of pCREB and its binding to the IL-10 promoter in B-1a cells under in vitro stimulation. (A) A total of 2 × 106 cells isolated from PerC were stimulated by either LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h followed by the surface staining with FITC-labeled anti-B220, PE-Cy5–labeled anti-CD5, intracellular staining with BV480-labeled anti-pCREB Abs and nuclear staining with DAPI. The samples were then subjected to acquisition in ImageStreamX Mark II Imaging Flow Cytometer and analysis by INSPIRE and IDEAS software. Representative images of nuclear translocation pCREB in B-1a cells treated with PBS, LPS, and LPS+PMA and ionomycin are shown. (B) A total of 1 × 106 B-1a cells sorted from PerC of C57BL/6 mice were stimulated by either PBS as control or LPS (20 μg/ml) or LPS (20 μg/ml) and PMA (50 ng/ml) and ionomycin (100 ng/ml) in combination for 20 h. Chromatin was extracted for ChIP assay by sonication, followed by cross-linking and immunoprecipitation using anti-pCREB Ab. Conventional qualitative, as well as quantitative real-time, PCR was performed and data are presented as gel-electrophoresis blots and fold induction of amplified CREB response element containing DNA in IL-10 promoter normalized against respective input controls. Data are expressed as means ± SE obtained from three independent experiments (n = 5 samples per group). Multiple groups were compared by one-way ANOVA and SNK method (*p < 0.05 versus PBS-treated cells). (C) Hypothesis scheme. In B-1a cells, LPS through TLR4-mediated pathway and PMA and Ca2+ ionophore ionomycin upregulated the phosphorylation of CREB, which in turn served as a pivotal transcription factor to induce IL-10 expression. Increased expression of IL-10 from the B-1a cells finally led to a protection against sepsis.
Discussion
Normal immune responses eradicate pathogens and quickly return the host to homeostasis, whereas the septic response accelerates due to inappropriate regulation of activated immune cells (33, 34). Here, we focused on the role of B-1a cells in regulating the acute hyperinflammatory response during sepsis. Lymphocytes are known to either stimulate or temper proinflammatory responses of macrophages (35). The role of regulatory T and B cells in controlling inflammatory diseases has been demonstrated in several studies (9, 30). By contrast, the role in sepsis of B-1a cells that are known to produce ample amounts of anti-inflammatory cytokine IL-10 remains to be defined.
Our current study showed a dramatic reduction of B-1a cell numbers within various compartments during sepsis, which might render infected hosts susceptible to hyperinflammation as these cells possess immunomodulatory properties. B-1a cells are predominantly localized in the PerC, accounting for a major portion of the total B cells of this serosal cavity (9). Although cellular apoptosis is one of the hallmarks of decreased cell numbers in lymphoid organs during sepsis (28), B-1a cells are resistant to apoptosis (36), thus the decrease of B-1a cells after sepsis might instead involve translocation and differentiation. In a recent study injection of bacteria or LPS into the PerC promoted PerC B-1a cells to migrate toward the spleen (37). Interestingly, although the percentage of B-1a cells in spleen and BM did not alter during sepsis, their absolute numbers were significantly reduced, which could be due to the overall loss of total splenocytes and BM cells and/or their differentiation into IgM-secreting cells, followed by their return into the PerC as memory B-1a cells (26, 38).
In the current study we tracked adoptively transferred GFP+ B-1a cells in WT mice during sepsis. We found after CLP a portion of the GFP+ B-1a cells initially transferred into the PerC in WT mice migrated into the spleen and MLNs. Although in frequencies the translocated GFP+ cells in spleen and MLNs following CLP looked few, if expressed in absolute cell numbers they might represent notable amounts. Despite the translocation of GFP+ PerC B-1a cells into the spleen and MLNs following CLP, we noticed the presence of adoptively transferred GFP+ cells in PerC. If this was the case, the question then arises as to why the endogenous B-1a cells were decreased in the PerC after CLP induction. While gating the endogenous cells by following conventional B-1a cell phenotype B220loCD23−CD5intIgMhiIgDlo, the differentiated cells with altered surface phenotype were precisely excluded, leading to decreased numbers of B-1a cells in the PerC following sepsis. By contrast, in the GFP+ B-1a cell transfer study, similar to endogenous B-1a cells the GFP+ B-1a cells might also be differentiated to express altered surface phenotype, while maintaining their GFP expression unchanged. We identified cells at various organs based on the presence of GFP, which might contain differentiated cells as well. Thus, our current study using GFP+ B-1a cell transfer barely depicts the event of B-1a cell translocation from PerC to the adjacent organs, rather than reveals their differentiation status.
Treatment with B-1a cells represents a potential therapy in sepsis. Our data revealed that adoptive transfer of B-1a cells into septic animals reduced the proinflammatory milieu and the bacterial load in the blood and PerC, which finally led to an overall increase in survival from sepsis. On the other hand, transfer of B-2 cells did not show protection against sepsis, thus pinpointing the importance of B-1a cells during sepsis over that of conventional B-2 cells. Using B-1a cell–deficient CD19−/− mice, we further validated our findings of B-1a cell adoptive transfers by showing CD19−/− mice fared poorly in CLP as compared with WT mice in terms of worse inflammation and tissue damage, and diminished survival. Next, the question becomes how do the B-1a cells improve sepsis? B-1a cells are known to perform several innate immune functions by releasing mediators and facilitating other cells’ functions in a contact-dependent manner (9). B-1a cells are known to secrete natural IgM, which provides protection from infection (16). It is therefore reasonable to consider B-1a cell–mediated natural IgM may be responsible for ameliorating sepsis. In our study, we noticed a significant decrease in plasma IgM levels during sepsis compared with its basal levels in sham animals, yet treatment with B-1a cells only marginally increased its level to restore the basal condition (Supplemental Fig. 4). To evaluate the role of natural IgM in the immediate response against microbial infection, a relevant study revealed that mice deficient in secreted (s)IgM in the CLP model of sepsis showed significantly higher mortality than their WT counterparts and the susceptibility to CLP by sIgM-deficient mice was reversed by reconstitution with polyclonal IgM from normal mouse serum (39). In humans, serum IgM levels were also found to be diminished in sepsis (40), although treatment with IgM-enriched Ig preparations in patients with severe sepsis did not demonstrate any improvement in mortality rate (41). Thus, considering clinical trials, therapy with IgM may not be an effective tool for sepsis treatment or it may be that a higher dose of specific Ab is required to provide protection against sepsis.
Aside from natural IgM, B-1a cells and the related B cell population, B10 cells, spontaneously secrete IL-10 and, after stimulation with LPS, secrete GM-CSF and IL-3 (7, 9, 23). In sepsis, the role of IRA B cells has been demonstrated previously (7, 23). The IRA B cells are developed in the spleen from the translocated PerC B-1a cells following LPS stimulation (7). IRA B cells produce ample amounts of GM-CSF and IL-3 but not IL-1β, IL-6, TNF-α, and even IL-10 (7, 23). Thus, the IRA B cells are distinct from other B cells, including the IL-10–producing B10 B cells, as well as the B-1a cells. However, considering their developmental features, it is likely that a portion of adoptively transferred PerC B-1a cells could become IRA B cells and might play an additive role to protect against sepsis. Moreover, as both B-1a cells and IRA B cells express several common phenotypic markers, functional overlaps between these two types of B cells might be possible.
In order to confirm the notion that B-1a cell–secreted IL-10 provides protection against sepsis, we found that the protective role of B-1a cells was only observed in septic mice treated with B-1a cells isolated from WT mice, but not from IL-10−/− mice. Considering the fact that IRA B cells are developed from the PerC B-1a cells, it is reasonable to think that, like WT mice, the IRA B cells can also be developed normally from IL-10−/− PerC B-1a cells. Interestingly, the IRA B cells were not deficient in the spleen of IL-10−/− mice after sepsis, representing the same frequency as WT mice (Supplemental Fig. 3). Our current study with B-1a cell–mediated IL-10 that provides therapeutic benefit in a clinically relevant model of sepsis not only identified the source of IL-10 from a rare population of B cells during sepsis but also further focuses on enhancing its production by B-1a cells through the upregulation of a relevant intracellular signaling pathway. Here we connect the beneficial role of IL-10 with B-1a cells and CREB.
A fine-tuned balance between pro- and anti-inflammatory cytokines is essential for proper immune reactions against pathogens and prevention of excessive damage to the host. There are several potent and commonly encountered transcription factors, such as NF-κB and the STAT3 that are well reported to act on the transcription of both pro- and anti-inflammatory cytokines during inflammation (42). In B-1a cells, constitutive activation of the intracellular STAT3 signaling pathway has already been demonstrated (13, 14). CREB is a transcription factor that is known for its role in cell proliferation, differentiation, and survival (43). However, emerging evidence has recently indicated a novel function of CREB for upregulation of IL-10 production, which in turn inhibits TLR-induced inflammation and prevents tissue damage (43, 44). A recent study utilizing macrophages demonstrated phosphorylation of CREB promotes IL-10 induction and TNF-α suppression (31); however, the role of CREB to induce IL-10 production by B-1a cells has never been studied. Global analysis of CREB occupancy in the promoter region of a wide range of genes present in various cell types has been demonstrated (45); however, information regarding B-1a cells is lacking. Elucidation of the importance of this transcription factor in B-1a cells for IL-10 production may further implicate a novel area regulating hyperimmune responses during sepsis.
Multiple sepsis clinical trials have been performed aiming to counteract many inflammatory mediators, such as TNF-α and IL-1β, or bacteria-derived pathogen-associated molecular patterns (33, 34). Unfortunately, all of these trials failed to demonstrate any benefit from treatment. Massive, instead of isolated, removal of excess quantities of inflammatory molecules through high-volume hemofiltration systems also failed to show any beneficial outcome in sepsis, because besides removing deleterious molecules this device also removes valuable serum anti-inflammatory molecules (46). Therefore, instead of targeting a particular inflammatory mediator or removing a bulk amount of mediators from the body either in early or late phase of sepsis, administration of immunomodulatory compounds which can regulate a wide range of proinflammatory molecules could be effective in sepsis treatment. It has previously been shown that treatment with recombinant IL-10 in mice improved the survival and delayed the onset of mortality and irreversible shock (47); conversely in the absence of IL-10 the mortality was more rapid after lethal CLP (47). In our study, we have identified that treatment with IL-10–producing B-1a cells in septic animals improved their survival, whereas IL-10−/− B-1a cells were unable to protect septic mice from inflammation and mortality. Thus, our work confirms that IL-10 could be an important immunomodulatory molecule secreted from B-1a cells to protect against sepsis. Under in vitro conditions, we have provided further evidence that IL-10 was produced by B-1a cells autonomously, although its production was dramatically increased by LPS alone or by LPS and PMA and ionomycin together. Our work shows the importance of B-1a cells as a source of immunoregulatory IL-10 in sepsis and suggests the value of further research to identify other potential inducers of IL-10 expression by B-1a cells through the involvement of novel intracellular–signaling molecules.
The phenotype of human B-1a cells has been recently redefined to identify a population that expresses functional features that match with murine B-1a cells, including spontaneous secretion of natural Ig, constitutive intracellular signaling, and efficient T cell stimulation (48). Our study demonstrating the role of mouse B-1a cells in sepsis further focuses on identifying valuable insights that may be applicable to human B-1a cells in sepsis patients to further move B-1a cell–oriented clinical research to the next level.
Acknowledgements
We thank Christopher Colon and Herb Borrero at the flow cytometry core facility of the Feinstein Institute for Medical Research for technical assistance and Dr. Mahendar Ochani at the Center for Immunology and Inflammation at the Feinstein Institute for Medical Research for technical support in animal works.
Footnotes
This work was supported by National Institutes of Health Grants R35GM118337, R01GM053008, and R01GM057468 to P.W. and R01AI029690 to T.L.R.
The online version of this article contains supplemental material.
Abbreviations used in this article:
- ALT
alanine aminotransferase
- AST
aspartate aminotransferase
- BM
bone marrow
- ChIP
chromatin immunoprecipitation
- CLP
cecal ligation and puncture
- CREB
cAMP-response element binding protein
- IRA
innate response activator
- LDH
lactate dehydrogenase
- MLN
mesenteric lymph node
- pCREB
phosphorylated CREB
- PerC
peritoneal cavity
- sIgM
surface IgM
- SNK
Student–Newman–Keuls
- Tg
transgenic
- WT
wild-type.
References
Disclosures
The authors have no financial conflicts of interest.