G-CSF drives a posttraumatic immune program that protects the host from infection.

Traumatic injury is generally considered to have a suppressive effect on the immune system, resulting in increased susceptibility to infection. Paradoxically, we found that thermal injury to the skin induced a robust time-dependent protection of mice from a lethal Klebsiella pneumoniae pulmonary challenge. The protective response was neutrophil dependent and temporally associated with a systemic increase in neutrophils resulting from a reprioritization of hematopoiesis toward myeloid lineages. A prominent and specific activation of STAT3 in the bone marrow preceded the myeloid shift in that compartment, in association with durable increases in STAT3 activating serum cytokines G-CSF and IL-6. Neutralization of the postburn increase in serum G-CSF largely blocked STAT3 activation in marrow cells, reversing the hematopoietic changes and systemic neutrophilia. Daily administration of rG-CSF was sufficient to recapitulate the changes induced by injury including hematopoietic reprioritization and protection from pulmonary challenge with K. pneumoniae. Analysis of posttraumatic gene expression patterns in humans reveals that they are also consistent with a role for G-CSF as a switch that activates innate immune responses and suppresses adaptive immune responses. Our findings suggest that the G-CSF STAT3 axis constitutes a key protective mechanism induced by injury to reduce the risk for posttraumatic infection.

PU.1 promotes cell cycle exit in the murine myeloid lineage associated with downregulation of E2F1.

Acute myeloid leukemia (AML) is characterized by increased proliferation and reduced differentiation of myeloid lineage cells. AML is frequently associated with mutations or chromosomal rearrangements involving transcription factors. PU.1 (encoded by Sfpi1) is an E26 transformation-specific family transcription factor that is required for myeloid differentiation. Reduced PU.1 levels, caused by either mutation or repression, are associated with human AML and are sufficient to cause AML in mice. The objective of this study was to determine whether reduced PU.1 expression induces deregulation of the cell cycle in the myeloid lineage. Our results showed that immature myeloid cells expressing reduced PU.1 levels (Sfpi1(BN/BN) myeloid cells) proliferated indefinitely in cell culture and expanded in vivo. Transplantation of Sfpi1(BN/BN) cells induced AML in recipient mice. Cultured Sfpi1(BN/BN) cells expressed elevated messenger RNA transcript and protein levels of E2F1, an important regulator of cell cycle entry. Restoration of PU.1 expression in Sfpi1(BN/BN) myeloid cells blocked proliferation, induced differentiation, and reduced E2F1 expression. Taken together, these data show that PU.1 controls cell cycle exit in the myeloid lineage associated with downregulation of E2F1 expression.

Optimization and application of a flow cytometric PU.1 assay for murine immune cells.

PU.1 is a master transcription factor whose levels directly influence hematopoiesis, leukemia, susceptibility to sepsis, and macrophage function. Though measurement of PU.1 levels is important to health and disease, most studies have relied on PCR or western blots to measure the expression of this transcription factor. An accessible, validated assay that could measure PU.1 protein in subpopulations of cells is needed. In this work we present an optimized flow cytometric assay to detect PU.1 in subpopulations of immune cells. Murine myeloid cells were fixed in paraformaldehyde, permeabilized, and then stained with anti PU.1 in the presence and absence of a blocking peptide containing the binding site of the antibody. The bound anti PU.1 was then visualized with a labeled second antibody. Methanol and ethanol were tested for their relative ability to permeabilize cells and detect PU.1. The effect of the procedure upon the ability to detect cellular subpopulations was examined. Relative PU.1 1evels in normal T cells, B cells, monocytes, macrophages, dendritic cells, neutrophils, and progenitors from the spleen and/or bone marrow were determined. Finally, PU.1 levels in proliferating myeloid cells from burn mice were determined. There was a dose dependent increase in the amount of PU.1 detected with increasing amounts of PU.1 antibody that was not seen when blocking peptide was used. Methanol or ethanol gave equivalent results as permeabilization agents, but the latter allowed easier detection of surface antigens when surface staining was performed prior to permeabilization. T cells had little if any PU.1, while B cells had intermediate levels of PU.1, and myeloid cells had high levels of PU.1. Monocytes had higher levels of PU.1 than did neutrophils or spleen macrophages. Plasmacytoid dendritic cells had lower levels of PU.1 than did conventional dendritic cells. Immature myeloid cells had higher levels of PU.1 than did mature myeloid cells. In addition, PU.1 levels were higher in proliferating cells than the corresponding non proliferating cells. Myeloid cells derived from burn mice tended to have higher levels of PU.1 than did unburned, but proliferating cells from burn or sham mice showed no difference in their levels of PU.1. This assay should be a useful addition to the tools used to study the function of PU.1 in health and disease.

NF-κΒ inhibition is ineffective in blocking cytokine-induced IL-8 production but P38 and STAT1 inhibitors are effective.

OBJECTIVE: In vitro but not in vivo evidence indicates that blockade of NF-κB is effective in reducing inflammation and production of IL-8. We hypothesized that the failure of in vitro experiments to predict in vivo outcome was due to the use of short time periods of observation and the use of single cytokines to stimulate NF-κB. METHODS: HEK cells with a NF-κB reporter gene or CaCo-2 cells were stimulated with CM (IL-1-ß; TNF-α, and IFN-γ) or individual cytokines in the presence and absence of NF-κB inhibitors, a STAT1 inhibitor, and/or a p38 MAPK inhibitor for periods up to 24 h. NF-κB activation, IL-8 production, and nitric oxide production were measured. RESULTS: CM-induced IL-8 production in HEK cells was additive to synergistic. CM enhanced production of IL-8 at 24 h but not 4 h was independent of NF-κB. The p38 inhibitor SB203580 and the STAT1 inhibitor EGCG blocked CM-induced IL-8 production at both early and late time periods. The NF-κB inhibitors PDTC and BAY11-7082 were found to increase CM-stimulated IL-8 production in Caco-2 cells at 24 h. CONCLUSIONS: Our data suggest an effective strategy to reduce IL-8 production is to block p38 or STAT1 rather than NF-κB.

Prior thermal injury accelerates endotoxin-induced inflammatory cytokine production and intestinal nuclear factor-κB activation in mice.

The objective of this study was to increase the understanding of the "second-hit" response in thermal injury. The authors hypothesized that prior thermal injury increases the endotoxin-induced inflammatory response of intestinal mucosa. Mice underwent sham or 25% TBSA scald injury. Seven days after injury, mice were injected with lipopolysaccharide. Blood, jejunum, and colon specimens were obtained at intervals. Serum, jejunal, and colon inflammatory cytokine levels were measured by enzyme-linked immunosorbent assay. Jejunal and colon nuclear factor (NF)-κB activation was measured by electrophoretic mobility shift assay. After remote thermal injury, lipopolysaccharide exposure led to an acute increase in serum interleukin (IL)-6, IL-10, and chemokine keratinocyte-derived chemokine (KC) levels. This correlated with lipopolysaccharide-induced increased IL-6 in colon and chemokine KC in the jejunum and colon in burned mice when compared with sham-injured mice. Lipopolysaccharide-induced NF-κB activation occurred more rapidly in jejunum and colon from burned mice compared with sham-injured mice. Prior thermal injury accelerates lipopolysaccharide-induced inflammatory cytokine production systemically in jejunum and colon. The "second hit" of lipopolysaccharide led to earlier intestinal NF-κB activation in burned mice compared with sham-injured mice. These results indicate that there is a heightened inflammatory response by jejunum and colon in response to a "second hit" of lipopolysaccharide after burn injury.

Neutrophils, not monocyte/macrophages, are the major splenic source of postburn IL-10.

Burn induces myeloid-derived suppressor cells (MDSCs), a heterogeneous population of immature polymorphonuclear neutrophils (PMNs) and monocytes, which protect against infection. Previous work from our laboratory demonstrated that inflammatory monocytes (iMos) were the major MDSC source of TNF-α in the postburn spleen, and we hypothesized that they were also the major source of postburn IL-10. To test this hypothesis, we examined cytokine production by postburn CCR2 knockout (KO) mice, which have fewer iMos than burn wild-type (WT) splenocytes, but equal numbers of PMNs and F4/80 macrophages. Using cell sorting and/or intracellular cytokine techniques, we examined IL-10 production by postburn PMNs and iMos. Finally, we compared IL-10 production by postburn PMNs and iMos with culture-derived MDSCs. Splenocytes from postburn CCR2 KO mice produced less IL-6 and TNF-α than WT burn splenocytes in response to LPS, but KO and WT burn splenocytes produced equal amounts of IL-10 in response to peptidoglycan. Depletion of PMNs from postburn splenocytes led to reductions in IL-10 and increases in IL-6 and TNF-α in response to peptidoglycan, but not in response to LPS. Sorting or intracellular cytokine techniques gave consistent results: Burn PMNs made more IL-10 than sham PMNs and also more IL-10 than burn or sham iMos. Polymorphonuclear neutrophil and iMos subpopulations from culture-derived MDSCs produced the same cytokine profiles in response to LPS and peptidoglycan as did the PMNs and iMos from postburn spleens: PMNs made IL-10, whereas iMos made IL-6. Finally, LPS-induced mortality of burn mice was made worse by anti-Gr-1 depletion of all PMNs and 66% of iMos from burn mice. This suggests that PMNs play a primarily anti-inflammatory role in vitro and in vivo.

Remote thermal injury increases LPS-induced intestinal IL-6 production.

BACKGROUND: Patients suffering from burn injury are at high risk for subsequent infection. Thermal injury followed by endotoxemia may result in a "second hit," causing an exaggerated inflammatory response with increased morbidity and mortality. The role of the intestine in this "second hit" response is unknown. We hypothesized that remote thermal injury increases the inflammatory response of intestinal mucosa to subsequent treatment with lipopolysaccharide (LPS). METHODS: Mice underwent sham or scald injury. Seven days after injury, mice were treated with LPS. Blood and bowel specimens were obtained. Serum and intestinal inflammatory cytokines were measured by enzyme-linked immunosorbent assay (ELISA). Changes in TLR-4 pathway components in intestine were measured by reverse transcription-polymerase chain reaction (RT-PCR), Western blot, and electrophoretic mobility shift assay (EMSA). Intestinal leukocyte infiltration was analyzed by myeloperoxidase assay. RESULTS: A "second hit" of injected LPS resulted in increased IL-6 in intestine of burned mice compared with sham. Similarly, jejunal IL-6 mRNA levels increased in mice with prior thermal injury, suggesting a transcriptional mechanism. Of transcription factors known to drive IL-6 expression, only AP-1 activation was significantly elevated by a "second hit" of LPS. CONCLUSION: Prior thermal injury potentiates LPS-induced IL-6 cytokine production in intestine. These results indicate a heightened inflammatory response to a second hit by intestine after burn injury.

Differential immunological phenotypes are exhibited after scald and flame burns.

A dysfunctional immune system is known to be part of the pathophysiology after burn trauma. However, reports that support this have used a variety of methods, with numerous variables, to induce thermal injury. We hypothesized that, all other parameters being equal, an injury infliction by a scald would yield different immunological responses than one inflicted by a flame. Here, we demonstrated that both burn methods produced a full-thickness burn, yet there was more of an increase in subdermal temperature, hematocrit, mortality, and serum IL-6 concentrations associated with the scald burn. On postinjury day 1, the scald-burned mice showed diminished lymphocyte numbers, interferon gamma production, and lymphocyte T-bet expression as compared with sham- and flame-burned mice. On postburn day 8, spleens from both sets of thermally injured animals showed an increase in proinflammatory myeloid cells as compared with sham-burned mice. Furthermore, the T-cell numbers, T-bet expression, and phenotype were changed such that interferon gamma production was higher in scald-burned mice than in sham- and flame-burned mice. Altogether, the data show that differential immunological phenotypes were observed depending on the thermal injury method used.

Gammadelta T cells mitigate the organ injury and mortality of sepsis.

Sepsis is a difficult condition to treat and is associated with a high mortality rate. Sepsis is known to cause a marked depletion of lymphocytes, although the function of different lymphocyte subsets in the response to sepsis is unclear. gammadelta T cells are found largely in epithelial-rich tissues, and previous studies of gammadelta T cells in models of sepsis have yielded divergent results. In the present study, we examined the function of gammadelta T cells during sepsis in mice using cecal ligation and puncture (CLP). Mice deficient in gammadelta T cells had decreased survival times and increased tissue damage after CLP compared with wild-type mice. Furthermore, bacterial load was increased in gammadelta T cell-deficient mice, yet antibiotic treatment did not change mortality. Additionally, we found that recruitment of neutrophils and myeloid suppressor cells to the site of infection was diminished in gammadelta T cell-deficient mice. Finally, we found that circulating levels of IFN-gamma were increased, and systemic levels of IL-10 were decreased in gammadelta T cell-deficient mice after CLP compared with wild-type mice. gammadelta T cell-deficient mice also had increased intestinal permeability after CLP compared with wild-type mice. Neutralization of IFN-gamma abrogated the increase in intestinal permeability in gammadelta T cell-deficient mice. The intestines taken from gammadelta T cell-deficient mice had decreased myeloperoxidase yet had increased tissue damage as compared with wild-type mice. Collectively, our data suggest that gammadelta T cells modulate the response to sepsis and may be a potential therapeutic target.

Thermal injury elevates the inflammatory monocyte subpopulation in multiple compartments.

Recent publications have demonstrated that human resident and inflammatory monocyte (IM) subpopulations have equivalents in rodents. The effect of thermal injury upon these subpopulations has not been studied. Mice were given a scald burn and killed on postburn days (PBDs) 2, 4, and 8. Bone marrow, blood, and spleen white cells were isolated, and the percentage of resident monocytes (CD11b LY6C), IMs (CD11b LY6C), and monocyte progenitors (macrophage-colony-forming unit [M-CFU]) were determined. The ability of each monocyte population to make TNF-alpha was determined by intracellular cytokine staining. Finally, the ability of sorted fractions from PBD 8 spleen to inhibit lymphocyte proliferation was performed. We noted that there was an increase in M-CFU in the blood and spleen at PBD 8, but the marrow only had a nonsignificant increase in M-CFU. All compartments showed a significant increase in the number of IMs by PBD 8, but no significant changes in resident monocytes were seen. In all compartments, IMs were a major source of TNF-alpha. The postburn increase in IMs and monocyte progenitors in the spleen was accompanied by an increase in the monocyte chemokine monocyte chemoattractant protein 1 and constitutively high levels of the progenitor chemokine stromal-derived factor 1alpha. After burn injury, mice deficient in the receptor for soluble TNF-alpha had equal levels of splenic M-CFU and monocytes, as did wild-type mice, suggesting that this cytokine is not essential for this effect. We conclude that in this model, IMs are a significant source of in vivo TNF-alpha.

Psychogenic stress prior to burn injury has differential effects on bone marrow and cytokine responses.

It is well known that many burn patients experience psychopathological disorders prior to burn injury. However, it is not known whether individuals that have been exposed to chronic psychological stresses will respond differently than unstressed individuals when challenged by a burn injury. In this study, we assessed whether chronic psychogenic stress prior to burn injury had any significant impact on burn injury-induced alterations in the myeloid compartment in the bone marrow and serum cytokine levels utilizing a well-controlled purely psychogenic stress model (predator exposure). Mice were individually caged and exposed to a Long Evans rat for 1 hr a day on 3 consecutive days prior to a 15% total body surface area flame burn. Four days after burn injury, bone marrow and serum were collected to assess myeloid cells and cytokine levels, respectively. Bone marrow cells were cultured in granulocyte-macrophage colony-stimulating factor (GM-CSF) to assess clonogenic ability. Flow cytometry was also used to characterize the populations of myeloid cells based on Gr-1 and CD11b staining intensity and to determine the expression of the macrophage colony-stimulating factor receptor (M-CSFR). Serum was assayed for IL-6, IL-12p70, MCP-1, and IFN-gamma by multiplexed sandwich enzyme-linked immunoabsorbent assay (ELISA). We found that predator exposure prior to burn injury ablated the burn-induced increase in myeloid colony formation and attenuated the burn-induced increases in immature monocytes and immature neutrophils in the bone marrow, as well as MCP-1 levels in the serum. Conversely, psychogenic stress exaggerated the burn-induced increase in the number of M-CSFR-positive cells. This study is the first to show the effects of a pure psychogenic stressor (predator exposure) on burn-induced alterations of the immune system. The clinical ramifications of our findings remain to be elucidated.

Stress and prolactin effects on bone marrow myeloid cells, serum chemokine and serum glucocorticoid levels in mice.

OBJECTIVE: Current evidence supports the conclusion that prolactin (PRL) is not an obligate immunoregulatory hormone and influences the immune system predominantly during stress conditions. In this study, we examined the impact of PRL on the psychogenic stress-induced responses of myeloid cells. METHODS: Seven-week-old PRL+/- (normal) and PRL-/- (deficient) mice were exposed to a predator for 1 h/day on 3 consecutive days. Another group of PRL-deficient mice received either 1 pituitary graft (hyperprolactinemic) or sham surgery at 5 weeks of age, while PRL-normal mice only received sham surgery. Two weeks later, these mice were also subjected to predator exposure. One day after the last predator exposure session, all mice were killed and the bone marrow and blood harvested. RESULTS: Significant differences in the myeloid cells between PRL-normal and PRL-deficient mice only occurred in stressed conditions. The median serum corticosterone levels were consistently higher in PRL-deficient mice. The implantation of a pituitary graft lowered the corticosterone levels to those observed in PRL-normal mice. The absolute number of immature neutrophils as well as the numbers of granulocyte macrophage, monocyte/macrophage and granulocyte colonies were significantly higher in the stressed PRL-deficient mice; however, only the increased number of immature neutrophils was reversed by pituitary grafting. CONCLUSIONS: Our findings support previous observations that PRL influences myeloid cells of the bone marrow most profoundly in stressed conditions. However, the mechanism by which PRL influences bone marrow myeloid cells during stress cannot be explained solely by its effect on serum corticosterone.

Effect of thermal injury on splenic myelopoiesis.

Thermal injury increases the number of macrophage progenitors in the bone marrow but leads to a decrease in the number of granulocyte progenitors. In the spleen, thermal injury increases the numbers of myeloid progenitors, but the lineage commitment of these cells is unknown. In this study mice were given a scald burn, and the number of splenic myeloid progenitors as well as their progeny was determined. BrdU uptake was used to monitor the de novo production of splenocytes for 8 days after the burn. Burn injury increased the numbers of splenic granulocyte-macrophage (GM), granulocyte (G), and macrophage (M) progenitors at postburn day 8 by 12-, 11-, and 18-fold, respectively. Scald injury increased the number of mature PMN (CD11b GR1(bright)) in the spleen and increased the number of white pulp monocyte/macrophages. Increased numbers of BrdU-positive PMN and monocyte/macrophages were seen after injury. Burn macrophages produced increased levels of the anti-inflammatory hematopoietic cytokine G-CSF. Our work clearly shows that the increased myelopoiesis observed postinjury leads to the production of mature myeloid cells. However, the effects of thermal injury on progenitors in the spleen and marrow are not equivalent.

Effects of prolactin level on burn-induced aberrations in myelopoiesis.

In this study, we sought to determine if prolactin (PRL) had any influence on burn-induced alterations in myelopoiesis and serum IL-6, IL-10, IL-12, IFN-gamma, TNF-alpha, and MCP-1 levels. To do this, we used mice that were PRL normal, PRL deficient, or hyperprolactinemic and had received a 15% total body surface area burn, sham treatment, or no treatment. We performed clonogenic assays of bone marrow cells, and we found that sham treatment significantly decreased monocyte/macrophage (M) colony formation relative to the control group in the PRL-deficient and PRL-normal mice (P < 0.01). Hyperprolactinemia attenuated the sham-induced decrease in M colony formation. Burn injury significantly increased M colony formation relative to the sham group with an equal significance in the PRL-deficient and PRL-normal mice (P < 0.05). We also showed that burn led to a significant increase in GM colony formation relative to the sham group. This burn-induced increase was significant in the PRL-normal (P < 0.05) and the PRL-deficient (P < 0.01) mice. In the PRL-normal mice, burn injury caused a 2.1-fold increase in the GM colony number, whereas in the PRL-deficient mice burn led to a 2.6-fold increase in GM colony number. When comparing the effects of burn injury on colony formation to the control groups, there were no significant differences seen, irrespective of the PRL level. We observed that all of the cytokines studied, with the exception of IL-10, were influenced by either sham treatment, burn injury, or both forms of stress. This stress-induced response occurred most often in animals that were either hypo- or hyperprolactinemic. We conclude that the PRL level was able to influence the sham-induced and burn-induced alterations in GM and M colony formation. Under euprolactinemic conditions, mice exhibited less often with stress-induced serum cytokine level alterations. We did not find any significant correlations with any of the serum cytokine levels and the ability to form colonies. Importantly, the sham treatment led to immune alterations independent of, and sometimes opposite of burn-induced effects.

Influence of glutamine-supplemented Caco-2 cells on cytokine production of mononuclear cells.

BACKGROUND: Caco-2 cells, cultured with mononuclear cells, were used as an in vitro model of human intestinal cell function. This study shows the effect of glutamine (Gln) supplementation on the production of tumor necrosis factor alpha, interleukin-10 (IL-10), and interleukin-6 (IL-6). METHODS: Confluent Caco-2 cells were cultured in media with Gln at 0 mmol/L, 4 mmol/L, or 10 mmol/L +/- 1 microg/mL lipopolysaccharide (LPS), treated with fluorescein isothiocyanate- (FITC-) conjugated intercellular adhesion molecule-1 (ICAM-1) mononuclear antibody, and assessed for ICAM-1 expression levels via flow cytometry. Confluent Caco-2 cells alone in apical inserts, or mononuclear cells (MNCs) alone in basal chambers of transwells, were cultured in media with 0 mmol/L, 4 mmol/L, or 10 mmol/L Gln. Supernatants were taken to assess cytokine and endotoxin levels. Confluent Caco-2 cells in apical inserts of transwells were cultured in media containing Gln at 0 mmol/L, 4 mmol/L, or 10 mmol/L, whereas MNCs were cultured in the basal chamber in media containing Gln at 4 mmol/L +/- LPS. Supernatants were collected to determine cytokine levels in each chamber. RESULTS: With Gln supplementation of the media at 10 mmol/L, enterocytes displayed a decrease in ICAM-1 expression. MNCs showed a decrease in tumor necrosis factor alpha and IL-6 production and an increase in IL-10 production when incubated with Caco-2 cells in media supplemented with Gln at 10 mmol/L. CONCLUSIONS: Although cytokine production by Caco-2 or mononuclear cells incubated alone was not influenced by the Gln concentration of the media, cultured together, Gln levels affected cytokine production by mononuclear cells, which suggests that Caco-2 cells produce mediators in Gln-rich conditions that can influence mononuclear cell cytokine production.

Thermal injury-induced increases of hepatocyte SOCS3 lead to decreases in STAT3.

Previous work in this laboratory has shown an increase of both mRNA and protein for suppressor of cytokine signaling 3 (SOCS3) in rat liver after thermal injury. This study identifies which liver cell type (parenchymal or non-parenchymal) is responsible for the postburn increase in SOCS3 and how this increase is connected to the signal transducer and activator of transcription (STAT) pathway. Parenchymal (hepatocytes) and non-parenchymal cells were isolated by Liberase digestion from postburn day 1 (PBD1) rats (including sham controls) and were analyzed for the expression of SOCS3 mRNA and protein and STAT3 and p-STAT3 protein. Reverse transcriptase (RT)-PCR performed on the isolated cells showed a significant increase of SOCS3 in the hepatocytes, but not in the non-parenchymal cells. When isolated hepatocytes from rats and the human hepatocyte cell line, HepG2, were cultured in the presence of IL-6, both showed an increase in SOCS3 mRNA expression. Anti-SOCS3, anti-STAT3, and anti-phosphorylated STAT3 labeling in both postburn rat liver and isolated hepatocyte cells that were cultured in the presence of IL-6 revealed that an increase in SOCS3 protein was accompanied by decrease in STAT3 protein. We propose that thermal injury stimulates non-parenchymal cells to produce cytokines, including IL-6, which in tum stimulate the Jak/STAT pathway in hepatocytes. The signal transduction pathway triggered by non-parenchymal cells causes an increase in SOCS3 production, which in turn induces the reduction of STAT3 protein in the hepatocytes.

Effects of prolactin deficiency on myelopoiesis and splenic T lymphocyte proliferation in thermally injured mice.

The importance of prolactin (PRL) in mammopoiesis and milk production is undisputed. However, previous studies investigating the role of PRL in immune function have yielded inconsistencies. These inconsistencies have led to our hypothesis that the immunomodulatory effects of PRL are only manifest under conditions in which the organism is subjected to stress. Thermal injury is a well-known stressor. The goal of this study was to determine whether the lack of PRL enhanced the negative effects of thermal injury-induced immune alterations utilizing a mouse model in which the PRL gene had been disrupted. Mice received either sham or burn treatment, and were sacrificed 4 days later. The immune parameters studied were the capacity of bone marrow cells to form granulocyte-macrophage colony forming units (GM-CFU) in the presence of granulocyte-macrophage colony stimulating factor, and the ability of the splenic T lymphocytes to proliferate in response to phytohemagglutin (PHA). As shown by others, our results reveal that burn increased the number of GM-CFU compared to sham controls; however, this elevation was only significant in the PRL-/- mice. Thermal injury increased PHA-stimulated proliferation of splenic T lymphocytes, however this increase was only significant in the PRL+/- group. We conclude that under conditions of a controlled stress event (thermal injury) [a] the increase in the GM-CFU is exaggerated in the absence of PRL, and [b] the enhancement of PHA-induced proliferation of splenic lymphocytes required PRL. This study supports the hypothesis that the immunomodulatory effects of PRL are manifest when the organism is subjected to stress.

The ability of endotoxin-stimulated enterocytes to produce bactericidal factors.

OBJECTIVE: Bactericidal peptides, specifically defensins, are produced by polymorphonuclear cells. Intestinal epithelial cells also produce bactericidal peptides, perhaps as part of their barrier function, to the greatest load of endogenous bacteria present in the body. We sought to determine whether and under what conditions intestinal cell lines could produce bactericidal compounds. DESIGN: Laboratory investigation. SETTING: Children's burn hospital. SUBJECTS: Caco-2, IEC-6, and HT-29 cell lines. INTERVENTIONS: Three different enterocyte lines were cultured for 1 day +/- lipopolysaccharide (1 or 10 microg/mL), and their supernatants were tested for bactericidal activity. Also, reverse transcription-polymerase chain reaction of Caco-2 cells was performed to assess the expression of defensin-6 mRNA. MEASUREMENTS AND MAIN RESULTS: After culture, enterocytes all were found to release one or more soluble factors with bactericidal activity (as measured fluorometrically by using a metabolizable dye) when stimulated by lipopolysaccharide (1 microg/mL). The bactericidal activity of these culture supernatants was saturated by increased bacterial load, additive to the effects of normal human peripheral blood polymorphonuclear cells, and was reduced by serial supernatant dilution. Enterocyte stimulation with larger amounts of lipopolysaccharide (10 microg/mL) resulted in greater bactericidal activity. After supernatant fractionation based on molecular weight, the bactericidal effect was best retained in the <10-kDa fraction. In addition, the expression of mRNA for defensin-6, a bactericidal peptide produced by neutrophils, was seen in Caco-2 cells. CONCLUSION: Enterocytes are shown to produce a soluble, low molecular weight, bactericidal compound in response to endotoxin stimulation. The expression of defensin-6 mRNA in Caco-2 cells suggests that intestinal cells may release defensins as bactericidal peptides. This experimental system provides an in vitro model to study the activity and production of bactericidal factors by enterocytes.

Changes in bone marrow-derived myeloid cells from thermally injured rats reflect changes in the progenitor cell population.

Bone marrow progenitor cells develop into mature tissue myeloid cells under the influence of colony-stimulating factors. Cytokines that are elevated post-thermal injury have been shown to influence this process. We hypothesize that thermal injury alters myelopoiesis at the level of the progenitor cell. These differences should be visible after in vitro cultures that include colony-stimulating factors. Prior to culture, bone marrow at postburn day 1 (PBD1) was assessed for cell surface markers and the levels of myeloid progenitors. After culture in granulocyte/macrophage-stimulating colony-stimulating factor, the cell surface markers of the cultured cells were determined. PBD1 marrow from thermally injured rats had more progenitor cells responsive to granulocyte/macrophage-stimulating colony-stimulating factor than did sham. Cultured PBD1 marrow produced more CD90(br) MY(br) CD45(dim) CD4(-) MHCII(-) CD11b(dim) eosinophils than did sham. Cultured bone marrow from thermally injured animals produces myeloid cells with an altered phenotype. Similar changes in myelopoiesis may take place in vivo.

Effect of in vivo infusion of granulocyte colony-stimulating factor on immune function.

As the applications of hematopoietic growth factors increase, their complex impact on host defense and immune responses continues to unfold. The effect of the administration of granulocyte colony-stimulating factor (G-CSF) on bacterial defense, proliferation of lymphocytes, and cytokine production by lymphocytes and peripheral blood mononuclear cells (PBMC) was studied. The effect of G-CSF administration on the phenotype of the cells in the major hematopoietic organs was studied as well. ACI rats were given 10 mg/kg/day G-CSF or vehicle daily for 4 days. Isolated bone marrow neutrophils and enterocytes from treated animals showed a greater bactericidal activity than controls. Proliferation of mitogen-stimulated lymphocytes and PBMC was reduced in G-CSF-treated animals. The production of proinflammatory cytokines, tumor necrosis factor (TNF), and interleukin 6 (IL-6) by lymphocytes and PBMC was reduced by G-CSF pretreatment. G-CSF administration caused an increase in IL-4 (Th2 cytokine) release and a decrease in interferon-gamma (IFNgamma, Th1 cytokine) release by mitogen-stimulated lymphocytes. Cytometric analysis of cells in the progenitor cell region indicated a large increase in immature cells in the bone marrow of G-CSF-treated animals compared with sham along with an increase in B cells and a decrease in polymorphonuclear leukocytes (PMNs). In addition, cytometric analysis showed a large increase in PMNs in blood and splenocytes of the treated animals compared with sham. This study confirms and extends previous observations that G-CSF administration has a number of effects that might simultaneously enhance host defense while reducing the risk of developing uncontrolled systemic inflammation. This may also be efficacious in prolonging graft survival and reducing graft vs. host disease.