Hematopoietic Stem-Progenitor Cells Restore Immunoreactivity and ...

4 downloads 8173 Views 979KB Size Report
Oct 14, 2011 - Emerging data support that persistence of the hypoinflammatory (hyporesponsive) ..... CD34+ cells restore proinflammatory cytokine production.
Hematopoietic Stem-Progenitor Cells Restore Immunoreactivity and Improve Survival in Late Sepsis Laura Brudecki,a Donald A. Ferguson,b Deling Yin,a Gene D. Lesage,a Charles E. McCall,c and Mohamed El Gazzara Departments of Internal Medicinea and Microbiology,b East Tennessee State University College of Medicine, Johnson City, Tennessee, USA, and Department of Internal Medicine, Section of Molecular Medicine, and Translational Science Institute, Wake Forest University School of Medicine, Winston-Salem, North Carolina, USAc

Sepsis progresses from an early/acute hyperinflammatory to a late/chronic hypoinflammatory phase with immunosuppression. As a result of this phenotypic switch, mortality in late sepsis from persistent primary infection or opportunistic new infection often exceeds that in acute sepsis. Emerging data support that persistence of the hypoinflammatory (hyporesponsive) effector immune cells during late sepsis might involve alterations in myeloid differentiation/maturation that generate circulating repressor macrophages that do not readily clear active infection. Here, we used a cecal ligation and puncture (CLP) murine model of prolonged sepsis to show that adoptive transfer of CD34ⴙ hematopoietic stem-progenitor cells after CLP improves long-term survival by 65%. CD34ⴙ cell transfer corrected the immunosuppression of late sepsis by (i) producing significantly higher levels of proinflammatory mediators upon ex vivo stimulation with the Toll-like receptor 4 (TLR4) agonist lipopolysaccharide, (ii) enhancing phagocytic activity of peritoneal macrophages, and (iii) clearing bacterial peritonitis. Improved immunity by CD34ⴙ cell transfer decreased inflammatory peritoneal exudate of surviving late-sepsis mice. Cell tracking experiments showed that the transferred CD34ⴙ cells first appeared in the bone marrow and then homed to the spleen and peritoneum. Because CD34ⴙ cells did not affect the early-phase hyperinflammatory response, it is likely that the newly incorporated pluripotent CD34ⴙ cells differentiated into competent immune cells in blood and tissue, thereby reversing or replacing the hyporesponsive endotoxintolerant cells that occur and persist after the initiation of early sepsis.

S

epsis is a major clinical problem (9, 52), with more than a 40% mortality rate, and is the leading cause of death in intensive care units (5, 17). Evidence supports that the pathophysiology of sepsis varies as it moves from an initiating early/acute hyperinflammatory phase to a late/chronic hypoinflammatory and immunosuppressive phase (31, 47, 51, 67). The early phase of sepsis is typified by a systemic inflammatory response syndrome (SIRS) characterized by excessive production of proinflammatory mediators by neutrophils and macrophages (53), increased generation of reactive oxygen species, and leukocyte-induced microvascular injury and organ failure (35). These destructive inflammatory responses occur in human (28) and animal (46, 51) sepsis, producing multiorgan dysfunction. While the early systemic inflammatory reaction of sepsis often spans several days (47, 61) and is considered a normal defense, the transition to a compensatory anti-inflammatory response syndrome (sometimes called CARS) to limit damage generates immunosuppression and promotes chronic infection (6, 12). CARS is characterized by downregulation in the ability of leukocytes to express proinflammatory mediators, impaired phagocytic capacity of neutrophils and macrophages (33, 40, 50), and significant apoptosis of lymphocytes and dendritic cells (16, 29). Previous studies have shown that monocytes/macrophages isolated from humans and mice during sepsis response do not produce inflammatory mediators in response to bacterial stimuli, thus producing the persistent phenomenon of endotoxin tolerance (11, 14, 20, 22). This hyporesponsive state predicts a poor outcome of sepsis (39). Mortality rates in late sepsis are high in humans (1, 27) and mice (46, 67) and often exceed mortality rates in the early phase of sepsis, which is defined as the first 5 days after cecal ligation and puncture (CLP) (67). While mortality during early sepsis correlates with hyperinflammation caused by the excessive systemic

602

iai.asm.org

production of inflammatory mediators (28, 46, 60), immunoincompetency (hyporesponsiveness) with persistent primary or secondary infection is often the cause of mortality in late sepsis (32, 50, 55). Anti-inflammatory treatment modalities targeting inflammatory mediators and bacterial toxins during the acute phase of sepsis were often effective in animal models of sepsis (44, 57) but failed in human clinical trials (26, 27, 49). This may be attributed to a delay between the onset of sepsis and the delivery of anti-inflammatory therapy when most patients entered the late hypoinflammatory (immunosuppressive) phase. There are no current effective treatments that target the late phase of sepsis, except the use of antibiotics and stabilizing organ functions, which improve survival by ⬃10% only (56). A high percentage of patients surviving sepsis and also systemic inflammation triggered by noninfection causes like trauma or major surgery develop prolonged systemic immunosuppression marked by monocyte/macrophage hyporesponsiveness (20, 23). Recovery of monocyte function results in clearance of sepsis in patients (20). Madonna and Vogel (36) used a murine model of endotoxemia to show that hyporesponsiveness to bacterial endotoxin is associated with alterations in the bone marrow-derived macrophage precursor pools. They demonstrated that acquisition and maintenance of the macrophage hyporesponsive state coin-

Received 15 September 2011 Returned for modification 14 October 2011 Accepted 23 November 2011 Published ahead of print 5 December 2011 Editor: B. A. McCormick Address correspondence to Mohamed El Gazzar, [email protected]. Copyright © 2012, American Society for Microbiology. All Rights Reserved. doi:10.1128/IAI.05480-11

0019-9567/12/$12.00

Infection and Immunity p. 602– 611

CD34⫹ Cells in Late Sepsis

cided with an increase in the number of macrophage progenitor cells in the bone marrow. Although the pathophysiologic phenotype of the inflammatory response induced by endotoxemia differs from that induced by sepsis, the hyporesponsiveness of innate cells is observed in both models (11, 12, 66, 67). These observations suggest that the prolonged hyporesponsive state that follows the hyperinflammatory (early) phase of sepsis is caused by persistent growth of microorganisms during acquired immunosuppression. Later studies reported rapid expansion of the immature myeloid cell population known as myeloid-derived suppressor cells in murine sepsis induced by CLP (18). These studies indicate alterations in innate cell differentiation and/or maturation program that may sustain the chronic-sepsis phenotype. Emerging evidence suggests that stem cell therapy can improve the clinical outcome in a variety of surgical pathologies, including cardiovascular and neurodegenerative diseases (63). The beneficial effects of stem cells have been attributed to their capacity to differentiate into mature immunocompetent cells (41) and their ability to home to sites of tissue injury and inflammation. We hypothesized that the early hyperinflammatory phase of sepsis may induce reprogramming of innate cell differentiation and/or maturation, thus providing a pool of dysregulated innate cells that sustain immunosuppression. Reconstituting the innate immune system with new myeloid progenitors may provide a source of fully functional innate cells required for sepsis clearance. Here, we used the cecal ligation and puncture (CLP) murine model of polymicrobial inflammation, a clinically relevant model of human sepsis (11, 46, 51), to determine whether adoptive transfer of CD34⫹ hematopoietic stem-progenitor cells (HSCs) to mice after the onset of sepsis restores the immunoinflammatory response and improves survival in late sepsis. Our results support that CD34⫹ cell transfer enhances immune clearance of late (chronic) bacterial infection and improves the outcome of late sepsis. MATERIALS AND METHODS Animals. Male BALB/c mice, 8 weeks old (Harlan Sprague Dawley, Indianapolis, IN), were used for all experiments. The mice were housed in a pathogen-free facility and were acclimated to the new environment for a week before surgery. All experiments were conducted in accordance with National Institutes of Health guidelines and were approved by the East State Tennessee University Animal Care and Use Committee. Sepsis model. Polymicrobial sepsis was induced using the cecal ligation and puncture (CLP) model, which closely emulates polymicrobial intraabdominal sepsis occurring in humans and has been used widely to study the immunopathological changes in sepsis (11, 46, 51). Most murine models of sepsis, which use cecal ligation and puncture to focus on the early/acute hyperinflammatory phase, are designed to produce increased early mortality (usually within 48 to 96 h) and then attempt to attenuate this phase therapeutically to improve outcomes. Because we were interested in the late/chronic phase of sepsis and the processes that may underlie the persistence of chronic infections and inflammation, we developed a less lethal cecal ligation and puncture model. To do this, we tested different CLP injuries, with or without administration of antibiotics. With the use of a 21-gauge needle, ⬃70% of the mice died within 3 days even with the antibiotic treatment. A 25-gauge needle produced only 10% mortality during the first 5 days, and mice who survived after day 5 did not develop chronic infections even when antibiotic treatment was withheld. With a 23-gauge needle, the mortality rate during the first 5 days was ⬃50% without or 20% with antibiotic treatment. Under both conditions, mice developed chronic infections after day 5. Therefore, we used a 23-gauge needle to produce ⬃20 to 30% mortality during early sepsis

February 2012 Volume 80 Number 2

(days 1 to 5). We define late sepsis as the period after day 5. Briefly, mice were anesthetized via inhalation with 2.5% isoflurane (Abbott Laboratories, Abbott Park, IL). A midline abdominal incision was made and the cecum was exteriorized, ligated distal to the ileocecal valve, and then punctured twice. A small amount of fecal material was extruded into the abdominal cavity. The abdominal wall and skin were sutured in layers with 3-0 silk. Sham-operated mice were treated identically except that the cecum was not ligated or punctured. For fluid resuscitation, mice received 1 ml lactated Ringers plus 5% dextrose intraperitoneally (i.p.). Mice were subcutaneously (s.c.) administered imipenem (25 mg/kg body weight) and 1 ml lactated Ringers plus 5% dextrose or an equivalent volume of 0.9% saline. To establish intraabdominal infection and create late sepsis, we delivered antibiotic therapy (37). We injected imipenem s.c. at 8 h after CLP and repeated the injection once, for a total of two doses over a 24-h period. Sampling. The sepsis response was divided into two phases: early/ acute phase (days 1 to 5) and late/chronic phase (after day 6). These distinctions were based on previously published studies (11, 46, 47, 67). Mice were monitored twice daily for 28 days. Dying mice were defined as extremely ill or moribund and thereafter sacrificed. We used the following criteria to identify moribund mice: (i) loss of righting reflex, (ii) hypothermia (⬍34°C), and (iii) loss of body weight (⬎30% of the baseline weight recorded before the surgery). These criteria were based on the Duke University guidelines for judging endpoints for morbidity in rodents and published studies showing that mice with significant weight loss, hypothermia, and loss of physical activity would die within 1 to 3 days (43). For each sacrificed (moribund) mouse from the CLP group, a mouse from the treatment group (i.e., receiving CD34⫹ cells) that appeared healthy was also sacrificed. To increase the statistical power of analysis, 6 to 10 mice (from repeated experiments) that were sacrificed on days 2 to 4 (representing early sepsis) or days 14 to 16 (representing late sepsis) were pooled as one group representing one data point. Thus, we present two data points, i.e., days 2 to 4 and days 14 to 16, representing early and late sepsis, respectively. Mice that died spontaneously (i.e., found dead) were not included in the analyses. These spontaneous deaths occurred mostly during the first 12 h after CLP and involved ⬃4% of CLP mice. Isolation and adoptive transfer of CD34ⴙ HSCs. The bone marrow cells were flushed out of the femurs and tibias with RPMI 1640 medium (without fetal bovine serum [FBS]) under aseptic conditions (15). A single cell suspension was made by repeated pipetting and filtering through a 70-␮m nylon strainer, followed by erythrocyte lysis. Next, we used magnetic-assisted cell sorting to isolate CD34⫹ hematopoietic stemprogenitor cells (HSCs). Bone marrow cells (2 ⫻ 107 cells) were incubated with biotinylated mouse anti-CD34 antibody (eBioscience, San Diego, CA) for 15 min at 4°C. After being washed, cells were incubated with antibiotin magnetic beads (Miltenyi Biotech, Auburn, CA) for 20 min at 4°C and subsequently applied to a Miltney selection (MS) column for positive selection of CD34⫹ cells according to the manufacturer’s instructions (Miltneyi). The purity of the CD34⫹ cells, as determined by flow cytometry, was higher than 90%. For adoptive transfer, isolated CD34⫹ cells were washed, resuspended in sterile 0.9% saline, and injected (5 ⫻ 105 cells in 100 ␮l volume) via the tail vein into mice within 1 day after CLP surgery. Control septic mice received heat-killed CD34⫹ cells (vehicle). Blood and peritoneal cell collection. At the time of sacrifice, mice were subjected to deep anesthesia, and heparinized whole blood was collected via the vena cava (or, in some cases, via the submandibular vein) and aliquoted to determine systemic bacteremia and circulating (plasma) cytokine levels. For cytokine analysis, blood was centrifuged for 5 min at 1,000 rpm and 4°C, and the plasma was removed and stored at ⫺20°C until analysis. After blood was collected, mice were sacrificed by an overdose of isoflurane, and the abdominal cavity was opened and flushed with 5 ml of cold phosphate-buffered saline (PBS). The peritoneal lavage fluid was

iai.asm.org 603

Brudecki et al.

then collected, and a small portion was used for bacterial cultures to determine the local bacterial load. The remainder was centrifuged at 1,500 rpm for 5 min. The pellets were resuspended in cold PBS and aliquoted. One aliquot was used for total and differential cell analysis on cytospin slides stained using a Diff-Stain kit (IMEB, Inc., San Marcos, CA). The remaining aliquot was used for phagocytosis assay and ex vivo stimulation as described below. Bacterial culture. To determine systemic bacteremia, whole blood was immediately diluted 4-fold. To determine local bacteremia, peritoneal lavage fluid was diluted 6-fold. Blood and peritoneal lavage fluid were then plated on 5% sheep blood agar plates with a Trypticase soy agar base (BD Biosciences, Sparks, MD). The plates were incubated for 24 h at 37°C under aerobic conditions. The plates were read by a microbiologist, and the CFU were determined and multiplied by dilution factor. Phagocytosis. Phagocytic activity of peritoneal neutrophils and macrophages was analyzed using fluorescein-conjugated bacteria followed by flow cytometry. Briefly, heat-killed pHrodo Escherichia coli (catalog no. MP10025; Invitrogen, Eugene, OR) was opsonized for 1 h at 37°C with polyclonal IgG antibody specific for E. coli plus purified bovine serum albumin (to block nonspecific binding) according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). Peritoneal lavage fluid cells (⬃0.5 ⫻ 106) were then incubated with the opsonized bacteria at 37°C at a 1:30 ratio, and the phagocytosis was stopped after 1 h according to the manufacturer’s instructions. To discriminate between the phagocytic activities of macrophages and neutrophils, cells were washed following the phagocytosis incubation and stained for 20 min at 4°C for macrophage marker allophycocyanin (APC)-conjugated anti-F4/80 and granulocyte marker fluorescein isothiocyanate (FITC)-conjugated anti-Gr1 antibody (eBioscience). Cells were then analyzed by flow cytometry. Cell culture. Unless otherwise stated, cells were cultured in complete RPMI 1640 medium that was supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 1⫻ penicillin-streptomycin (HyClone, Logan, UT). The cultures were incubated at 37°C in 5% CO2-humidified atmosphere. Function of peritoneal macrophages assessed ex vivo. To determine immune function of peritoneal macrophages, harvested peritoneal lavage fluid cells (see above) were washed with sterile PBS, resuspended in RPMI 1640 medium supplemented with 10% FBS, and then cultured for 2 h in a 24-well plastic plate. Nonadherent cells were removed by a wash with PBS. Adherent cells (macrophages) were incubated in complete RPMI 1640 medium and stimulated with 100 ng/ml of the Toll-like receptor 4 (TLR4) agonist lipopolysaccharide (LPS) (serotype 0111:B4; Sigma-Aldrich, St. Louis, MO) for 24 h. Supernatants were collected and stored at ⫺20°C for future cytokine analysis. Flow cytometry. Cells were labeled by incubation for 30 min on ice in staining buffer (PBS plus 2% FBS) with appropriate fluorochromeconjugated antibodies. After being washed, the samples were analyzed by a FACSCaliber flow cytometer (BD Biosciences, Sparks, MD). Data were acquired and analyzed using CellQuest Pro software (BD Biosciences). The following antibodies were used for flow cytometry: anti-CD34 – eFluor 660 and anti-CD38 –FITC (eBioscience). An appropriate isotypematched control was used for each antibody. Cell tracking. CD34⫹ cells were isolated from normal mice and fluorescently labeled using a Vybrant CFDA (carboxyfluorescein diacetate) cell tracer kit (Invitrogen, Eugene, OR). After a 10-min incubation at room temperature with 20 ␮M dye, cells were washed, resuspended in PBS, and injected intravenously (i.v.) into mice that underwent CLP surgery. The dye-protein adducts that form in labeled cells are retained throughout development and cell division and have been detected up to a few weeks after injection into mice (64). At days 2 and 5 after CLP, mice were sacrificed and cytospin slides were prepared from the peritoneal lavage fluid cells and single cell suspensions of bone marrow and spleen and then visualized by fluorescence microscopy. Cytokine measurement. Cytokine concentrations were measured by enzyme-linked immunosorbent assay (ELISA). Tumor necrosis factor al-

604

iai.asm.org

FIG 1 Adoptive transfer of normal CD34⫹ cells improves survival in late sepsis. Mice were subjected to cecal ligation and puncture (CLP) using a 23gauge needle and given antibiotics (imipenem) with fluid resuscitation. Bone marrow CD34⫹ hematopoietic progenitor cells were isolated from normal mice and injected (5 ⫻ 105 cells in 100 ␮l saline) intravenously 1 day after CLP. Mortality was monitored for 28 days, and death (moribundity) was separated into two categories: early-phase deaths (those occurring on days 1 to 5) and late-phase deaths (those occurring on days 6 to 28). All dying (moribund) mice suffered significant weight loss (⬎30%), hypothermia (⬍34°C), and loss of righting reflex. CLP mice without CD34⫹ cell transfer showed 8% survival by day 28. In contrast, 72.7% of CLP mice with CD34⫹ cell transfer survived 28 days. The data from 30 mice per group were pooled from four independent experiments. Vehicle, heat-killed CD34⫹ cells.

pha (TNF-␣), interleukin-6 (IL-6), and IL-10 levels in the supernatants of LPS-stimulated cells and their circulating (plasma) levels were determined using specific ELISA kits (eBioscience) according to the manufacturer’s instructions. Each sample was run in duplicate. Statistical analysis. To increase the statistical power of analysis, each data point (days 2 to 4 and days 14 to 16) was represented by 6 to 8 mice pooled from four repeated experiments that were sacrificed on the same days. The Kaplan-Meier survival curve was plotted using GraphPad prism 5.0 (GraphPad Software), and survival significance was determined by the log rank test. Other data were analyzed using Microsoft Excel 3.0, and statistical significance was determined by Student’s t test for two-group comparison or one-way analysis of variance (ANOVA) for multiple comparisons. All values are expressed as means ⫾ standard deviations (SD). P values of ⱕ0.05 were considered to be statistically significant.

RESULTS

Model of CLP-induced late sepsis. Because mortality and inflammatory profiles change over the time course of sepsis (8, 47), sepsis response can be divided into early and late phases, reflecting SIRS and CARS, respectively (see introduction). We used a 23-gauge needle and administered antibiotics (imipenem) with fluid resuscitation to produce low mortality during the early phase (defined as days 1 to 5 after CLP), thereby establishing a late phase (defined as day 6 and onward after CLP) (11, 47). This regimen established a chronic infection and allowed for a longer time window to investigate the late-sepsis response. Our results with this model show that administration of antibiotics with fluid resuscitation produced about 20 to 30% mortality during the early phase and about 70% during the late phase (Fig. 1 and data not shown). Our preliminary experiment indicated that early-phase mortality reached 60% in mice that did not receive antibiotics (data not shown). Thus, the short-term antibiotic treatment not only was important in reducing early mortality

Infection and Immunity

CD34⫹ Cells in Late Sepsis

but also simulated anti-inflammatory therapy for patients with sepsis (12, 60). This provided a sensitive model to assess effects of CD34⫹ cell transfer. CD34ⴙ transfer improves survival during late sepsis from peritonitis. To evaluate the effect of CD34⫹ cell transfer on the late-sepsis response, we monitored mortality over the course of 4 weeks after CLP. We used weight loss (⬎30%), hypothermia (⬍34°C), and loss of righting reflex as criteria to identify dying/ moribund mice. Previous studies reported a significant weight loss in mice destined to die after CLP, as those mice would die within 1 to 3 days after experiencing significant weight loss (47, 67). In our preliminary experiments, we found that mice experiencing deep hypothermia and loss of righting reflex died within 1 day if they were left to die (data not shown). The weight loss in ⬃10% of dying mice ranged from 7 to 25% only, but mice still experienced deep hypothermia and loss of the righting reflex as well. This large fluctuation in weight loss was most notable during the early phase (days 1 to 5). To control for the transferred cell, a group of mice received heat-killed CD34⫹ cells (vehicle), and a second group were injected with LLC-PK1 epithelial cells as a differentiated control (data not shown). As shown in Fig. 1, CLP mice showed an 8% survival rate by the end of the experiment. Transfer of CD34⫹ cells increased survival by ⬃65% (72.7% survived, versus 8% of CLP control mice) (P ⬍ 0.05). We did not observe any significant difference in survival up to day 6 after CLP (i.e., during the early phase) between CLP mice with and without CD34⫹ cell transfer. These results demonstrate that CD34⫹ cells have a protective effect against sepsis, especially during the late phase. Mice surviving late sepsis after CD34ⴙ cell transfer mount a normal inflammatory response. In order to simplify our comparisons, we pooled the results such that each phase of sepsis is represented by one data point. The data point covering days 2 to 4 represents the early phase and includes mice dying on days 2, 3, and 4 after CLP. The data point covering days 14 to 16 represents the late phase and includes mice dying and surviving (i.e., with CD34⫹ cell transfer) on days 14, 15, and 16. Days 2 to 4 come 28 h after cessation of the antibiotic and fluid resuscitation therapy and therefore reflect a typical systemic proinflammatory response (or SIRS), whereas days 14 to 16 reflect a compensatory antiinflammatory (i.e., late) response. Previous studies have shown that the early phase of CLPinduced sepsis in mice is associated with elevated levels of proinflammatory cytokines, whereas the chronic phase is associated with upregulation of anti-inflammatory cytokines (11, 25, 46, 67). To determine the effect of CD34⫹ cells on the sepsis cytokine response, we measured the circulating (plasma) levels of the proinflammatory cytokines TNF-␣ and IL-6 and the antiinflammatory cytokine IL-10 by using ELISA. All mice dying (moribund) on days 2 to 4 had significantly larger amounts of TNF-␣ and IL-6 than sham-operated mice (Fig. 2, top). Moribund mice also produced larger amounts of IL-10 (data not shown). These changes were not modified in the mice receiving CD34⫹ cells. All dying (moribund) mice sacrificed before day 2 had significantly (3- to 7-fold) higher levels of all three cytokines than mice dying on days 2 to 4 (data not shown). During the late phase (days 14 to 16), levels of proinflammatory TNF-␣ and IL-6 were reduced in mice with and without CD34⫹ cell transfer (Fig. 2, bottom). IL-10 levels were also reduced in mice with CD34⫹ cell transfer (data not shown).

February 2012 Volume 80 Number 2

FIG 2 CD34⫹ cell transfer affects circulating levels of TNF-␣, IL-6, and IL-10 only in late sepsis. Mice were treated and categorized as described in the legend for Fig. 1. The plasma concentrations of TNF-␣, IL-6, and IL-10 were measured by ELISA. The results shown represent two data points. The data point covering days 2 to 4 represents the early-phase response and is the average concentration taken from mice dying (moribund) on days 2, 3, and 4. The data point covering days 14 to 16 represents the late-phase response and is the average taken from mice without CD34⫹ cell transfer dying on days 14, 15, and 16. These data points were compared to the corresponding numbers of surviving mice (i.e., receiving CD34⫹ cells) sampled on the same days. All dying (moribund) mice that were sacrificed before day 2 had significantly (3- to 7-fold) higher levels of all three cytokines (not presented). Similar levels of cytokines TNF-␣, IL-6, and IL-10 were seen on days 2 to 4. Levels of IL-10 were reduced in late sepsis following transfer of CD34⫹ cells, while TNF-␣ and IL-6 were diminished in all animals. The data from 7 or 8 mice per experimental group for each data point were pooled from four independent experiments and are expressed as means ⫾ SD. ⴱ, P ⬍ 0.05, comparing CLP to shamoperated mice.

CD34ⴙ cells restore proinflammatory cytokine production in peritoneal macrophages. To test whether CD34⫹ cell transfer results in reconstitution of innate cell immunocompetency, peritoneal macrophages were isolated and stimulated ex vivo with LPS, and cytokine levels were determined after 24 h. Macrophages isolated during the early phase (days 2 to 4) in CLP mice with and without CD34⫹ cell transfer responded to LPS by producing low levels of the proinflammatory cytokines TNF-␣ and IL-6 and, thus, were hyporesponsive/tolerant (Fig. 3, top). Macrophages isolated at 24 h after CLP produced larger amounts of both cytokines (data not shown). In contrast, macrophages isolated from mice with CD34⫹ cell transfer during the late phase (days 14 to 16), but not mice without transfer, were able to respond to LPS with increased TNF-␣ and IL-6 production (Fig. 3, bottom), supporting that CD34⫹ cell transfer generates a peritoneal cell population that is immunoresponsive. The sustained state of immunosuppression (i.e., anti-inflammatory response) in mice without CD34⫹ cell transfer is further supported by elevated IL-10 production by macrophages from mice without CD34⫹ cell transfer (data not shown). Together, the results presented suggest that CD34⫹ cell transfer generated competent innate immune responses in the peritoneum. CD34ⴙ cells improve bacterial clearance in the peritoneum during late sepsis. Mortality from late sepsis is associated with bacterial overgrowth due to an impaired phagocytic capacity of

iai.asm.org 605

Brudecki et al.

FIG 3 CD34⫹ cell transfer restores LPS reactivity of peritoneal macrophages in late sepsis. Peritoneal lavage fluid cells were harvested from mice representing the two data points. Cells were cultured for 2 h. Nonadherent cells were removed, adherent macrophages were stimulated with 100 ng/ml LPS for 24 h, and supernatants were analyzed for TNF-␣, IL-6, and IL-10 proteins by ELISA. The data from 8 mice per experimental group for each data point were pooled from four independent experiments and are expressed as means ⫾ SD. ⴱ, P ⬍ 0.05, compared to CLP mice without CD34⫹ cell transfer.

macrophages and neutrophils (11, 12, 67). To test whether transfer of CD34⫹ cells might improve survival in late sepsis by increasing bacterial clearance, we first determined systemic (blood) and local (peritoneal) bacterial growth. Blood and peritoneal lavage fluid were collected from dying and surviving mice at the time of sacrifice and cultured. Bacteria were detected in blood and peritoneal lavage fluid from mice with and without CD34⫹ cell transfer in the early phase (Fig. 4, top), without a significant difference between the two groups. The level of blood bacteremia was higher in all dying mice that were sacrificed 12 h before day 2 (data not shown). All cultures were positive for Escherichia coli, group D Enterococcus, Staphylococcus aureus, viridans Streptococcus, and Lactobacillus sp. Some cultures were grown under anaerobic conditions and tested positive for Campylobacter gracilis. We observed that blood bacteremia was mostly cleared in mice with and without CD34⫹ cell transfer in the late phase (Fig. 4, bottom right). We expected that this might reflect decreased bacterial dissemination, since bacterial growth becomes more localized in late sepsis (11, 12). To test this, we determined bacterial growth in the peritoneal lavage fluid from mice with and without CD34⫹ cell transfer that died during the late phase (days 14 to 16). Mice without CD34⫹ cell transfer had substantial bacteria in the peritoneum (Fig. 4, bottom left). In contrast, peritoneal bacteria were significantly diminished in animals with CD34⫹ cell transfer. Cells derived from transferred CD34ⴙ cells home to the peritoneum. The enhanced bacterial clearance observed in late-septic mice that received CD34⫹ cells suggested that the replaced peritoneal cells were immunocompetent and that mice without CD34⫹ cell transfer did not mount a competent immunoinflammatory response (Fig. 3). We predicted that immunocompetent macrophages would be derived from the transferred CD34⫹ cells. To test this, CD34⫹ cells isolated from normal mice were fluorescently labeled and injected into CLP mice. Bone marrow, spleen, and peritoneal lavage fluid cells were harvested at days 2 and 5

606

iai.asm.org

FIG 4 CD34⫹ cell transfer enhances local bacterial clearance in late sepsis. Blood and peritoneal lavage fluid were collected from mice representing the two data points. Systemic (blood) and local (peritoneal) bacteria were determined by bacterial culture. The levels of blood bacteremia were higher in all dying mice that were sacrificed before day 2 (not presented). Levels of bacteria were similar in early sepsis (days 2 to 4) in mice with and without CD34⫹ cell transfer. Both groups cleared blood bacteria in late sepsis (days 14 to 16). In contrast, levels of peritoneal bacteria were significantly reduced in late sepsis only in mice with CD34⫹ cell transfer, supporting enhanced local immunity. Each symbol represents an individual mouse. Note the difference in scale. The data from 8 mice per experimental group for each data point were pooled from four independent experiments and are expressed as means ⫾ SD.

after transfer, cytospun, and visualized by fluorescence microscopy. Labeled cells were detected in the bone marrow but not in the spleen or peritoneum at day 2 (Fig. 5). By day 5, cells were detected also in the spleen and peritoneum. In addition, flow cytometry showed fluorescent CD34⫹ cell-derived macrophages (F4/80positive) in the peritoneum at day 5 (data not shown). We also found that purified normal CD34⫹ cells differentiated into macrophages when cultured in the presence of macrophage colony-stimulating factor (M-CSF) for 4 days (data not shown). These results suggest that the adoptively transferred multipotent CD34⫹ cells home to the bone marrow first and that their myeloid derivatives populate the peritoneum upon differentiation. Peritoneal macrophages and neutrophils from mice receiving CD34ⴙ cells have enhanced phagocytic function. We next determined the bacterial phagocytosis capacity of macrophages and granulocytes. Peritoneal lavage fluid cells were isolated, and their capacity to phagocytize bacteria was measured by incubation with fluorescein-conjugated E. coli. After the phagocytosis assay, numbers of phagocytosed bacteria were measured by flow cytometry analysis after staining for the macrophage marker F4/80 and the neutrophil marker Gr1. Our preliminary experiments showed that macrophages and neutrophils from naïve mice (i.e., without CLP or CD34⫹ cell transfer) were highly phagocytic (data not shown). Phagocytic activities of macrophages (Gr1low F4/80positive) and neutrophils (Gr1high F4/80negative) isolated from

Infection and Immunity

CD34⫹ Cells in Late Sepsis

FIG 5 Cells derived from CD34⫹ cells home to bone marrow, spleen, and peritoneum. Purified CD34⫹ cells were fluorescently labeled, washed, and injected i.v. into mice 1 day after CLP. Mice were sacrificed on days 2 and 5, and cytospin slides were prepared from bone marrow, spleen, and peritoneal lavage fluid cells. Representative photomicrographs (⫻200 magnification) are shown.

CLP mice were not significantly different during early sepsis between mice with and without CD34⫹ cell transfer (Fig. 6A, top, and B, top). In contrast, during late sepsis, we observed a significant increase in the phagocytic activity of macrophages from mice with CD34⫹ cell transfer compared with that of macrophages from mice without CD34⫹ cell transfer (Fig. 6B, bottom left). The same pattern was also observed for neutrophils, although the difference between the two treatment groups was not as dramatic (Fig. 6B, bottom right). We also performed phagocytosis assays using sorted macrophages and neutrophils. Although the flow cytometry cell counts were similar, the fluorescent signals from the phagocytosed bacteria were also significantly reduced in both macrophages and neutrophils from late-septic mice without CD34⫹ cell transfer (data not shown). These results support that the phagocytic capacity of macrophages and neutrophils is impaired in late sepsis and that CD34⫹ cells enhance bacterial phagocytic capacity during late but not early sepsis, thus promoting clearance of the local infection in the peritoneum. Finally, we determined local cell recruitment during the early (day 2 to 4)- and late (day 14 to 16)-sepsis response. A mixed inflammatory cell infiltrate of macrophages, neutrophils, and lymphocytes was observed in the peritoneum of all mice with and without CD34⫹ cell transfer (Table 1). However, we observed relative changes in the numbers of different cell types. In early sepsis, mice with and without CD34⫹ cell transfer exhibited no significant difference in the total or differential cell counts. In late sepsis, we observed a significant decrease in macrophage and neutrophil recruitment in mice with CD34⫹ cell transfer compared with that in mice not receiving CD34⫹ cells. In addition, the numbers of macrophages and neutrophils in mice without CD34⫹ cell

February 2012 Volume 80 Number 2

transfer were higher during late sepsis than during early sepsis. Also, CLP mice without CD34⫹ cell transfer that were sacrificed on day 8 had even greater numbers of macrophages (4.6 ⫻ 106 ⫾ 0.3 ⫻ 106) and neutrophils (7.2 ⫻ 106 ⫾ 0.8 ⫻ 106) (data not shown) than mice sacrificed later (days 14 to 16). Interestingly, the number of lymphocytes was significantly increased in the peritoneum of late-septic mice that received CD34⫹ cells. These results demonstrate that CLP mice dying from early or late sepsis have modest changes in inflammatory cell recruitment despite their reduced capacity for bacterial clearance. DISCUSSION

The acute phase of sepsis is characterized by a robust innate immune response resulting in the release of proinflammatory mediators, including cytokines, and subsequent metabolic deregulation. The persistently elevated levels of proinflammatory cytokines lead to priming of immune cells and vascular endothelium, activation of the coagulation cascade, and induction of hypotension and septic shock (10, 19, 45). Despite the damaging effects of excessive inflammation, most septic patients survive this acute phase but may have late organ dysfunction and metabolic deregulation (14, 63). The mortality rate in late sepsis is higher than that in early sepsis (20, 46, 67) and is characterized by attenuated immunosuppression with enhanced bacterial overgrowth and effector immune cell apoptosis (12, 47, 63). Here we have shown, for the first time, that adoptive transfer of CD34⫹ hematopoietic progenitor cells restores the immunocompetence in late-septic mice with peritonitis. Our evidence for this is as follows: (i) improved survival, (ii) enhanced clearance of local

iai.asm.org 607

Brudecki et al.

FIG 6 CD34⫹ cell transfer enhances the phagocytic capacity of peritoneal macrophages from late-septic mice. Peritoneal lavage fluid cells were collected and incubated with fluorescein-conjugated pH-sensitive E. coli. After phagocytosis was stopped, cells were washed and stained for macrophage marker anti-F4/80APC and granulocyte marker anti-Gr1-FITC, in order to discriminate between phagocytosis activities of macrophages and neutrophils. (A) Histogram of gated macrophages and neutrophils showing the phagocytic capacity of each cell population. Gated macrophages and neutrophils were identified as Gr1low F4/80positive and Gr1high F4/80 negative, respectively. According to this assay, fluorogenic E. coli becomes highly fluorescent once endocytosed. (B) Quantitative analysis of the phagocytic capacity. The data from 6 mice per experimental group for each data point were pooled from three independent experiments and are expressed as means ⫾ SD. ⴱ, P ⬍ 0.05, comparing CLP to CD34⫹ cell-administered CLP mice.

bacteria, (iii) a reduced repressive state, and (iv) increased phagocytosis by macrophages and neutrophils. Since death in early sepsis is due mainly to both heightened inflammation and acute infection and late sepsis is characterized by persistent bacterial growth and hyporesponsive inflammation (immunosuppression), our results have clinical relevance. CD34⫹ cells did not protect the early phase of sepsis, as the survival rates as well as in vitro and in vivo cytokine responses were similar in mice receiving and not receiving CD34⫹ cells. Five to 6 days after CLP (i.e., beginning of the chronic phase), no further deaths were observed in mice receiving CD34⫹ cells concurrent with immunosuppression. Hematopoietic CD34⫹ cells are stem cell-derived progenitors with combined myeloid and lymphoid differentiation and self-renewal potential (2, 4). Recently, Akashi et al. (4) trans-

608

iai.asm.org

planted CD34⫹ cells into lethally irradiated mice and detected the donor-derived granulocytes and monocytes 6 days after transplantation. It is likely that the time between our cell transfer and the mounting protective response in mice was required for CD34⫹ cells to differentiate into immunocompetent macrophages and neutrophils. In our study, CD34⫹ cell derivatives were detected in the spleen and peritoneum approximately 5 days after cell transfer. Recently, Unsinger et al. (59) demonstrated that adoptive transfer of human CD34⫹ hematopoietic cord blood cells into immunodeficient NOD-scid mice produced a complete lineage of human cells of the innate and adaptive immune systems. These mice later mounted a typical acute sepsis response upon CLP, as demonstrated by the high production of pro- and anti-inflammatory cytokines and enhanced lymphocyte apopto-

Infection and Immunity

CD34⫹ Cells in Late Sepsis

TABLE 1 Mice surviving CLP due to CD34⫹ cell transfer have fewer inflammatory cells in the peritoneuma No. of cells Time point

Mouse group

Total cells

Macrophages

Neutrophils

Lymphocytes

Days 2-4 (n ⫽ 7)

CLP CLP ⫹ CD34⫹

7.8 ⫻ 106 ⫾ 1.0 ⫻ 106 8.0 ⫻ 106 ⫾ 1.2 ⫻ 106

2.6 ⫻ 106 ⫾ 0.2 ⫻ 106 2.4 ⫻ 106 ⫾ 0.4 ⫻ 106

4.5 ⫻ 106 ⫾ 0.4 ⫻ 106 4.7 ⫻ 106 ⫾ 0.2 ⫻ 106

0.7 ⫻ 105 ⫾ 0.2 ⫻ 105 0.9 ⫻ 105 ⫾ 0.3 ⫻ 105

Days 14–16 (n ⫽ 6)

CLP CLP ⫹ CD34⫹

9.4 ⫻ 106 ⫾ 0.8 ⫻ 106 7.0 ⫻ 106 ⫾ 1.0 ⫻ 106

3.8 ⫻ 106 ⫾ 0.7 ⫻ 106 2.1 ⫻ 106 ⫾ 0.3 ⫻ 106*

5.3 ⫻ 106 ⫾ 0.3 ⫻ 106 3.4 ⫻ 106 ⫾ 0.3 ⫻ 106*

0.3 ⫻ 105 ⫾ 0.1 ⫻ 105 1.5 ⫻ 105 ⫾ 0.5 ⫻ 105*

Peritoneal lavage fluid was collected from dying (CLP) and surviving (CLP ⫹ CD34⫹) mice, and the number of inflammatory cells was quantified. Data are expressed as the means ⫾ SD from 6 or 7 mice/group/data point (pooled from three independent experiments). The numbers shown are total cells/mouse. *, P ⬍ 0.05, comparing CLP to CD34⫹ cell-administered CLP mice. a

sis. These studies indicate that CD34⫹ cells can reconstitute an otherwise defective innate immune repertoire. Our finding that CD34⫹ cells increased bacterial clearance during late sepsis is clinically significant. Increased bacterial growth and dissemination are common during early and late sepsis (11, 12). We detected systemic and local bacterial growth during the early phase in mice with and without CD34⫹ cell transfer. However, we found increased local bacterial overgrowth only in control septic mice. Bacterial growth persisted in some mice for the entire course of the experiment and was detected in all dying mice in both early and late phases. The profile of bacterial growth corresponded with the functional state of macrophages and neutrophils. As confirmed in this study, other reports have demonstrated that macrophage function is suppressed in mice with sepsis (6, 11), which supports continued infection. Many patients die during the late/chronic phase of sepsis because of opportunistic (new) infections that accompany hyporesponsiveness/tolerance of circulating neutrophils and monocyte/macrophages (20). Our results showed that late-septic mice had an attenuated cytokine response concurrent with increased local (peritoneal) bacterial growth. Interestingly, mice surviving late sepsis after CD34⫹ cell transfer had significantly fewer peritoneal bacteria and a normal cytokine response in macrophages upon ex vivo stimulation with LPS, which supports retrieval of immunocompetency. This pattern coincided with CD34⫹ cell derivatives homing to the peritoneum. Thus, in our model, mortality in late sepsis is caused largely by bacterial overgrowth due to the acquired immunosuppression associated with the sepsis inflammatory adaptive reaction. Macrophages from mice that survived late sepsis due to CD34⫹ cell transfer released more proinflammatory cytokines after ex vivo stimulation than did those from dying mice. These results support the notion that downregulation of proinflammatory cytokines after the initial early phase plays a role in the establishment of the hyporesponsive (late) phase of sepsis (34, 38). Cytokines such as TNF-␣ and IL-6 are important in activating macrophages and neutrophils and regulating apoptotic cell death. Spontaneous apoptosis occurs in both macrophages and neutrophils (16, 29, 65). The process of apoptosis is accelerated during the late phase of sepsis (63). We detected a significant increase in the number of lymphocytes in surviving mice compared to that in dying mice during the late phase, suggesting an attenuated lymphocyte apoptosis. Alternatively, CD34⫹ cell-derived lymphocytes may have contributed to the increase in lymphocytes and supported improved survival. In this regard, Hotchkiss et al. (30) demonstrated that the adoptive transfer of T cells overexpressing the antiapoptotic protein Bcl-2 into mice 2 h after CLP significantly improves

February 2012 Volume 80 Number 2

survival in early/acute sepsis. Several studies reported that TNF-␣ and IL-6 repress neutrophil apoptosis in vitro (13, 29). In this study, we found higher levels of macrophages and neutrophils in the peritoneal cavity of CLP mice without CD34⫹ cell transfer, simultaneously with low levels of circulating TNF-␣ and IL-6. This profile was reversed in surviving mice that received CD34⫹ cells. The significantly lower numbers of macrophages and neutrophils observed in mice that received CD34⫹ cells, although not dramatic, reflected a state of successful inflammation resolution and a restoration of immunocompetence and cellular homeostasis. In human and murine sepsis, the immunosuppressive phase that follows initial hyperinflammation is typified by hyporesponsiveness or tolerance to ex vivo stimulation by bacterial endotoxin (7, 54, 65). We have observed this endotoxin tolerance in a promonocytic cell model of sepsis and human sepsis (21). The attenuated cytokine production in these tolerant monocyte/macrophages generates impaired defense against infection (48, 68). It is caused by reprogramming of a chromatin structure that supports gene-specific epigenetic modifications. Our results showed that peritoneal macrophages isolated from mice that survived late sepsis released significantly larger amounts of TNF-␣ and IL-6 upon ex vivo stimulation with LPS than macrophages isolated from mice that did not survive (i.e., did not receive CD34⫹ cells). This indicates that macrophages from mice surviving late sepsis have a normal/functional immunoinflammatory response. Our cell tracking studies showed that transferred CD34⫹ cells were detected in the peritoneum around day 5 after transfer but not at day 2. The majority of peritoneal macrophages recovered after day 5 from mice that received CD34⫹ cells may represent entry of CD34⫹ cell-derived macrophages into various tissues. Ayala et al. (7) reported that peritoneal macrophages isolated from mice 24 h after CLP-induced sepsis had a diminished capacity to release IL-6 upon ex vivo stimulation. However, these cells retained a normal capacity to produce IL-10. This supports our finding of elevated IL-10 levels in the plasma and ex vivo culture of peritoneal macrophages from septic mice without CD34⫹ cell transfer. Others have shown that cells from septic patients fail to release proinflammatory cytokines following ex vivo stimulation (24). Our finding that TNF-␣, IL-6, and IL-10 were detected at very low circulating levels in mice surviving late sepsis indicates a state of immune recovery after inflammation resolution and is not due to tolerance, as observed in mice without CD34⫹ cell transfer (Fig. 2, bottom). The circulating levels of TNF-␣ and IL-6 in mice sampled between days 1 and 2 were higher than the levels on days 2 to 4 (data not shown). Given the important role of macrophages and neutrophils in

iai.asm.org 609

Brudecki et al.

local bacterial clearance in sepsis (12), bacterial overgrowth in late sepsis may reflect a defective phagocytic capacity as a component of septic dysfunction. Our results showed that the phagocytic capacity of macrophages and neutrophils was greatly reduced in mice dying during late sepsis. Peritoneal macrophages from mice that survived late sepsis due to CD34⫹ cells showed an enhanced capacity to phagocytize bacteria. Because the difference in the phagocytic capacities of neutrophils observed during late sepsis between dying and surviving mice was not as dramatic as that seen in macrophages, these findings support that defects in macrophage phagocytic capacity during late sepsis are the major contributor to the bacterial overgrowth. These results also suggest that restoring the phagocytic capacity of effector immune cells and increasing bacterial clearance are a main mechanism by which CD34⫹ cells improve survival in late sepsis. Recently, a number of studies reported that bone marrow stromal cells (also known as mesenchymal stem cells) can modify the inflammatory response of sepsis (42, 62). For example, Nemeth et al. (42) demonstrated that transfer of bone marrow stromal cells into mice 24 h before and 1 h after CLP reduces mortality and improves organ function in early sepsis. The protective mechanism was attributed to the ability of stromal cells to increase IL-10 production in septic macrophages, thus attenuating the hyperinflammatory phase. Unlike hematopoietic stem-progenitor cells, mesenchymal stem cells cannot be mobilized from the bone marrow into the peripheral blood and also cannot terminally differentiate ex vivo (58), implying that they cannot sustain a long-term innate immunocompetency. In addition, Weil et al. (62) reported that mesenchymal stem cells improved survival in a mouse model of endotoxemia by attenuating systemic and myocardial inflammation, which was attributed to a reduction in the circulating levels of the proinflammatory cytokines TNF-␣, IL-1␤, and IL-6. Thus, in these previous studies the protective effect of mesenchymal stem cells appeared indirect, i.e., through releasing factors that target effector immune cells and blunting an already heightened proinflammatory response. These studies were aimed at attenuating the hyperinflammatory (early) phase of sepsis so that it may not proceed to a late (chronic) response. Our study, however, targets the hyporesponsive late phase, where mortality from bacterial overgrowth is dominant. In our model, the mechanism of CD34⫹ cell-mediated protection against sepsis was different. CD34⫹ cells did not attenuate the early hyperinflammatory, potentially lethal phase induced after CLP but rather reversed the immunosuppression of the late phase. The time gap between administration of CD34⫹ cells and the mounted protection supports that these cells repopulate the host’s innate immune system and that the immunocompetent macrophages and neutrophils recovered from mice surviving late sepsis are derived from CD34⫹ myeloid progenitors. Our cell tracking studies showed that CD34⫹ cells and their derivatives were detected in the bone marrow shortly after transfer (at day 2) but homed to the peritoneum at a later time (around day 5), suggesting that these progenitors required time for differentiation/maturation and homing to these tissues. Recently, Wysocka et al. (66) reported that in vivo administration of the Flt3 ligand accelerates recovery from immunosuppression in a murine model of endotoxin tolerance involving alterations in the responsiveness of macrophages and dendritic cells by stimulating the generation and homing of immunocompetent dendritic cells into the spleen. Interestingly, Flt3 is a tyrosine kinase receptor expressed on mul-

610

iai.asm.org

tipotent stem-progenitor cells, including CD34⫹ cells, and is required for sustaining their lymphoid and myeloid (granulocytemonocyte) differentiation potential (3). Upon investigating lineage commitment and myeloid differentiation potential, Adolfsson et al. (3) found that transplantation of Flt3⫹ cells into irradiated mice reconstituted the bone marrow and spleen and that these cells were differentiated into granulocyte-macrophage progenitors ex vivo. These studies support our finding that CD34⫹ cell derivatives reconstituting the lymphoid tissue and peritoneum were critical for immune recovery and resolution of late sepsis. In summary, we find that CD34⫹ hematopoietic myeloid progenitor cells significantly improve survival in late sepsis by restoring the immunocompetency of the innate immune system. Given the clinical sequence of sepsis, CD34⫹ cells may provide an attractive approach for treating late sepsis, for which there are no effective treatment alternatives. ACKNOWLEDGMENT This work was supported by funding from the East Tennessee State University College of Medicine.

REFERENCES 1. Abraham E, et al. 1997. p55 tumor necrosis factor receptor fusion protein in the treatment of patients with severe sepsis and septic shock. A randomized controlled multicenter trial. Ro 45-2081 Study Group. JAMA 277: 1531–1538. 2. Adolfsson J, et al. 2001. Upregulation of Flt3 expression within the bone marrow Lin(-)Sca1(⫹)c-kit(⫹) stem cell compartment is accompanied by loss of self-renewal capacity. Immunity 15:659 – 669. 3. Adolfsson J, et al. 2005. Identification of Flt3⫹ lympho-myeloid stem cells lacking erythro-megakaryocytic potential a revised road map for adult blood lineage commitment. Cell 121:295–306. 4. Akashi K, Traver D, Miyamoto T, Weissman IL. 2000. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature 404:193–197. 5. Angus DC, et al. 2001. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit. Care Med. 29:1303–1310. 6. Ayala A, Chaudry IH. 1996. Immune dysfunction in murine polymicrobial sepsis: mediators, macrophages, lymphocytes and apoptosis. Shock 6(Suppl 1):S27–S38. 7. Ayala A, Urbanich MA, Herdon CD, Chaudry IH. 1996. Is sepsisinduced apoptosis associated with macrophage dysfunction? J. Trauma 40:568 –573. 8. Baker CC, Chaudry IH, Gaines HO, Baue AE. 1983. Evaluation of factors affecting mortality rate after sepsis in a murine cecal ligation and puncture model. Surgery 94:331–335. 9. Baue AE. 1996. MOF/MODS, SIRS: an update. Shock 6(Suppl 1):S1–S5. 10. Beasley D, Schwartz JH, Brenner BM. 1991. Interleukin 1 induces prolonged L-arginine-dependent cyclic guanosine monophosphate and nitrite production in rat vascular smooth muscle cells. J. Clin. Invest. 87: 602– 608. 11. Belikoff B, Hatfield S, Sitkovsky M, Remick DG. 2011. Adenosine negative feedback on A2A adenosine receptors mediates hyporesponsiveness in chronically septic mice. Shock 35:382–387. 12. Belikoff BG, et al. 2011. A2B adenosine receptor blockade enhances macrophage-mediated bacterial phagocytosis and improves polymicrobial sepsis survival in mice. J. Immunol. 186:2444 –2453. 13. Biffl WL, et al. 1996. Interleukin-6 delays neutrophil apoptosis. Arch. Surg. 131:24 –29. 14. Bone RC, et al. 1992. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/ Society of Critical Care Medicine. Chest 101:1644 –1655. 15. Chen CZ, Li L, Lodish HF, Bartel DP. 2004. MicroRNAs modulate hematopoietic lineage differentiation. Science 303:83– 86. 16. Colotta F, Re F, Polentarutti N, Sozzani S, Mantovani A. 1992. Mod-

Infection and Immunity

CD34⫹ Cells in Late Sepsis

17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

27.

28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

ulation of granulocyte survival and programmed cell death by cytokines and bacterial products. Blood 80:2012–2020. Cunha BA. 1995. Antibiotic treatment of sepsis. Med. Clin. North Am. 79:551–558. Delano MJ, et al. 2007. MyD88-dependent expansion of an immature GR-1(⫹)CD11b(⫹) population induces T cell suppression and Th2 polarization in sepsis. J. Exp. Med. 204:1463–1474. Dinarello CA. 1997. Proinflammatory and anti-inflammatory cytokines as mediators in the pathogenesis of septic shock. Chest 112:321S–329S. Docke WD, et al. 1997. Monocyte deactivation in septic patients: restoration by IFN-gamma treatment. Nat. Med. 3:678 – 681. El Gazzar M, Yoza BK, Hu JY, Cousart SL, McCall CE. 2007. Epigenetic silencing of tumor necrosis factor alpha during endotoxin tolerance. J. Biol. Chem. 282:26857–26864. Ellaban E, Bolgos G, Remick D. 2004. Selective macrophage suppression during sepsis. Cell. Immunol. 231:103–111. Ertel W, et al. 1997. Inhibition of the defense system stimulating interleukin-12 interferon-gamma pathway during critical illness. Blood 89:1612–1620. Ertel W, et al. 1995. Downregulation of proinflammatory cytokine release in whole blood from septic patients. Blood 85:1341–1347. Ertel W, et al. 1991. The complex pattern of cytokines in sepsis. Association between prostaglandins, cachectin, and interleukins. Ann. Surg. 214: 141–148. Fisher CJ, Jr, et al. 1994. Recombinant human interleukin 1 receptor antagonist in the treatment of patients with sepsis syndrome. Results from a randomized, double-blind, placebo-controlled trial. Phase III rhIL-1ra Sepsis Syndrome Study Group. JAMA 271:1836 –1843. Fisher CJ, Jr, et al. 1994. Initial evaluation of human recombinant interleukin-1 receptor antagonist in the treatment of sepsis syndrome: a randomized, open-label, placebo-controlled multicenter trial. Crit. Care Med. 22:12–21. Groeneveld AB, Tacx AN, Bossink AW, van Mierlo GJ, Hack CE. 2003. Circulating inflammatory mediators predict shock and mortality in febrile patients with microbial infection. Clin. Immunol. 106:106 –115. Haslett C. 1992. Resolution of acute inflammation and the role of apoptosis in the tissue fate of granulocytes. Clin. Sci. (Lond.) 83:639 – 648. Hotchkiss RS, et al. 2000. Caspase inhibitors improve survival in sepsis: a critical role of the lymphocyte. Nat. Immunol. 1:496 –501. Hotchkiss RS, Karl IE. 2003. The pathophysiology and treatment of sepsis. N. Engl. J. Med. 348:138 –150. Hotchkiss RS, Nicholson DW. 2006. Apoptosis and caspases regulate death and inflammation in sepsis. Nat. Rev. Immunol. 6:813– 822. Huang X, et al. 2009. PD-1 expression by macrophages plays a pathologic role in altering microbial clearance and the innate inflammatory response to sepsis. Proc. Natl. Acad. Sci. U. S. A. 106:6303– 6308. Lederer JA, Rodrick ML, Mannick JA. 1999. The effects of injury on the adaptive immune response. Shock 11:153–159. Lukashev D, Ohta A, Apasov S, Chen JF, Sitkovsky M. 2004. Cutting edge: physiologic attenuation of proinflammatory transcription by the Gs protein-coupled A2A adenosine receptor in vivo. J. Immunol. 173:21–24. Madonna GS, Vogel SN. 1985. Early endotoxin tolerance is associated with alterations in bone marrow-derived macrophage precursor pools. J. Immunol. 135:3763–3771. Mazuski JE, et al. 2002. The Surgical Infection Society guidelines on antimicrobial therapy for intra-abdominal infections: an executive summary. Surg. Infect. (Larchmt.) 3:161–173. McCall CE, Yoza B, Liu T, El Gazzar M. 2010. Gene-specific epigenetic regulation in serious infections with systemic inflammation. J. Innate Immun. 2:395– 405. Meakins JL, et al. 1977. Delayed hypersensitivity: indicator of acquired failure of host defenses in sepsis and trauma. Ann. Surg. 186:241–250. Munoz C, et al. 1991. Dysregulation of in vitro cytokine production by monocytes during sepsis. J. Clin. Invest. 88:1747–1754. Nagy RD, et al. 2005. Stem cell transplantation as a therapeutic approach to organ failure. J. Surg. Res. 129:152–160. Nemeth K, et al. 2009. Bone marrow stromal cells attenuate sepsis via prostaglandin E(2)-dependent reprogramming of host macrophages to increase their interleukin-10 production. Nat. Med. 15:42– 49.

February 2012 Volume 80 Number 2

43. Nemzek JA, Xiao HY, Minard AE, Bolgos GL, Remick DG. 2004. Humane endpoints in shock research. Shock 21:17–25. 44. Ohlsson K, Bjork P, Bergenfeldt M, Hageman R, Thompson RC. 1990. Interleukin-1 receptor antagonist reduces mortality from endotoxin shock. Nature 348:550 –552. 45. Okusawa S, Gelfand JA, Ikejima T, Connolly RJ, Dinarello CA. 1988. Interleukin 1 induces a shock-like state in rabbits. Synergism with tumor necrosis factor and the effect of cyclooxygenase inhibition. J. Clin. Invest. 81:1162–1172. 46. Osuchowski MF, Welch K, Siddiqui J, Remick DG. 2006. Circulating cytokine/inhibitor profiles reshape the understanding of the SIRS/CARS continuum in sepsis and predict mortality. J. Immunol. 177:1967–1974. 47. Osuchowski MF, Welch K, Yang H, Siddiqui J, Remick DG. 2007. Chronic sepsis mortality characterized by an individualized inflammatory response. J. Immunol. 179:623– 630. 48. Randow F, et al. 1995. Mechanism of endotoxin desensitization: involvement of interleukin 10 and transforming growth factor beta. J. Exp. Med. 181:1887–1892. 49. Reinhart K, Karzai W. 2001. Anti-tumor necrosis factor therapy in sepsis: update on clinical trials and lessons learned. Crit. Care Med. 29:S121– S125. 50. Remick DG. 2007. Pathophysiology of sepsis. Am. J. Pathol. 170:1435– 1444. 51. Remick DG, Bolgos GR, Siddiqui J, Shin J, Nemzek JA. 2002. Six at six: interleukin-6 measured 6 h after the initiation of sepsis predicts mortality over 3 days. Shock 17:463– 467. 52. Rice TW, Bernard GR. 2005. Therapeutic intervention and targets for sepsis. Annu. Rev. Med. 56:225–248. 53. Sands KE, et al. 1997. Epidemiology of sepsis syndrome in 8 academic medical centers. JAMA 278:234 –240. 54. Seatter SC, Li MH, Bubrick MP, West MA. 1995. Endotoxin pretreatment of human monocytes alters subsequent endotoxin-triggered release of inflammatory mediators. Shock 3:252–258. 55. Shelley O, Murphy T, Paterson H, Mannick JA, Lederer JA. 2003. Interaction between the innate and adaptive immune systems is required to survive sepsis and control inflammation after injury. Shock 20:123–129. 56. Szabo G, Romics L Jr, Frendl G. 2002. Liver in sepsis and systemic inflammatory response syndrome. Clin. Liver Dis. 6:1045–1066. 57. Tracey KJ, et al. 1987. Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteraemia. Nature 330:662– 664. 58. Tyndall A, Pistoia V. 2009. Mesenchymal stem cells combat sepsis. Nat. Med. 15:18 –20. 59. Unsinger J, McDonough JS, Shultz LD, Ferguson TA, Hotchkiss RS. 2009. Sepsis-induced human lymphocyte apoptosis and cytokine production in “humanized” mice. J. Leukoc. Biol. 86:219 –227. 60. Vianna RC, et al. 2004. Antibiotic treatment in a murine model of sepsis: impact on cytokines and endotoxin release. Shock 21:115–120. 61. Wang H, et al. 1999. HMG-1 as a late mediator of endotoxin lethality in mice. Science 285:248 –242. 62. Weil BR, et al. 2010. Mesenchymal stem cells attenuate myocardial functional depression and reduce systemic and myocardial inflammation during endotoxemia. Surgery 148:444 – 452. 63. Weil BR, et al. 2009. Stem cells in sepsis. Ann. Surg. 250:19 –27. 64. Weston SA, Parish CR. 1990. New fluorescent dyes for lymphocyte migration studies. Analysis by flow cytometry and fluorescence microscopy. J. Immunol. Methods 133:87–97. 65. Williams MA, Withington S, Newland AC, Kelsey SM. 1998. Monocyte anergy in septic shock is associated with a predilection to apoptosis and is reversed by granulocyte-macrophage colony-stimulating factor ex vivo. J. Infect. Dis. 178:1421–1433. 66. Wysocka M, Montaner LJ, Karp CL. 2005. Flt3 ligand treatment reverses endotoxin tolerance-related immunoparalysis. J. Immunol. 174:7398 – 7402. 67. Xiao H, Siddiqui J, Remick DG. 2006. Mechanisms of mortality in early and late sepsis. Infect. Immun. 74:5227–5235. 68. Ziegler-Heitbrock HW, Frankenberger M, Wedel A. 1995. Tolerance to lipopolysaccharide in human blood monocytes. Immunobiology 193: 217–223.

iai.asm.org 611