Enhancement of Dendritic Cell Production by Fms-Like Tyrosine Kinase-3 Ligand Increases the Resistance of Mice to a Burn Wound Infection¹

Severe burn injury imposes a high risk for the development of opportunistic infections and associated complications that can delay wound healing, prolong hospitalization, lead to sepsis, and increase mortality. Risk factors for developing nosocomial infections after burn injury include large burn size (>30% of the total body surface area), full-thickness (third-degree) burns, and concurrent inhalation injury (1). Advances in patient care have increased early survival in severely burned patients, but infections remain a leading cause of mortality in these patients (1, 2). Loss of the protective barrier provided by the skin is primarily responsible for susceptibility to infections after burns. Furthermore, numerous immunological alterations induced by burn injury impair innate and acquired immunities and decrease the ability of burn patients to control and eliminate infections. In both burned patients and animal models, NK cell, neutrophil, macrophage, and T cell functions are impaired (3, 4, 5, 6, 7, 8, 9, 10). Alterations in pathogen-elicited production of cytokines associated with Th1, Th2, and proinflammatory responses likely contribute to these global impairments (6, 11, 12, 13, 14, 15). Topical and systemic antibiotics are frequently used and have decreased, but not eliminated, the incidence of infections (16). Aside from prophylactic antibiotic administration, there are no accepted pharmacological interventions for preventing infections in burn patients. Recent studies have focused on the therapeutic potential of targeting specific inflammatory mediators or metabolic pathways. This strategy may be limited due to the large number of immunological alterations induced by burns and an incomplete understanding of the contributions of specific alterations to postburn immunosuppression.

An approach that may increase resistance to infections after burn injury is the stimulated production of new immune cells through treatments with various hemopoietic growth factors. The supposition underlying this approach is that immune cells generated after the initial burn shock may outnumber and compensate for previously damaged cells. Stimulated production of macrophages and neutrophils after burns has been achieved using G-CSF, a neutrophil growth factor, and GM-CSF, a growth factor for neutrophils and macrophages. Both G-CSF and GM-CSF improved immune function in animal models of burn-associated sepsis (17, 18). Treatment of burn patients with GM-CSF improved neutrophil activities in vitro, but effects on the incidence of infections or mortality were not reported (19). It is likely that the most widely effective therapies will be those that impact numerous effector cells, which may require combinatorial therapies. Therefore, it may be useful to examine different immune cell growth factors for their potential to increase resistance to infections.

This study evaluated the hemopoietic cytokine Fms-like tyrosine kinase-3 ligand (Flt3L),³ a growth factor that enhances the production of dendritic cells and, to a lesser degree, NK cells. Flt3L is a hemopoietic factor that interacts with the Flt3R expressed on common myeloid and common lymphoid progenitor cells (20). Engagement of the Flt3R stimulates the expansion of progenitor cells and their differentiation into dendritic cells and NK cells. Transgenic mice lacking Flt3L have reduced numbers of bone marrow progenitor cells, dendritic cells, and NK cells (21). Treatment of mice with exogenous Flt3L causes a dramatic expansion in the numbers of dendritic cells and, to a lesser degree, NK cells in the blood, spleen, and liver (22, 23). In humans, administration of Flt3L causes a dramatic increase in the number of dendritic cells in peripheral blood, with no overt toxicity (24). We evaluated the ability of Flt3L to protect against burn wound infection and report that Flt3L treatment enhances production of dendritic cells and increases the resistance of mice to a burn wound infection with Pseudomonas aeruginosa.

All animal procedures were approved by the Institutional Animal Care and Use Committee at the University of Texas Medical Branch and met the National Institutes of Health Guidelines for the Care and Use of Experimental Animals. A widely used technique for induction of full-thickness scald burns (25, 26, 27) was used in these studies. Male BALB/c mice, 6–8 wk old and weighing 20–25 g, were housed in an institutional animal care facility and allowed to acclimate for 1 wk after arrival. Mice were anesthetized during all procedures with 2.5% isoflurane delivered in air through a circuit, and the dorsal and lateral surfaces of the mice were shaved with clippers. A small amount of saline (1 ml) was injected s.c. along the spinal column to provide a protective cushion for the spinal cord during the burn procedure. Mice were placed on their backs and secured in a protective template with an opening corresponding to 30% of the total body surface area, and the exposed skin was immersed in 97°C water for 10 s. Lactated Ringers solution (LR; 2 ml) was administered i.p. for fluid resuscitation and buprenorphine (2 mg/kg) was given for analgesia. Sham-injured mice were subjected to all of the procedures except immersion in water.

P. aeruginosa was purchased from the American Type Culture Collection (ATCC no. 19660) and is the same strain used by others investigating P. aeruginosa burn wound infections (26). P. aeruginosa was used, because it is a common source of wound infections and pneumonia in burn patients (1, 28). Cultures were grown in tryptic soy broth, and serial dilutions were plated on tryptic soy agar for determination of CFU. Wounds were dried and immediately inoculated after burn injury by the topical application of 500 CFU of P. aeruginosa to the burn wound. To examine horizontal spread of bacteria within the wound, a modification of a previously described technique was used (29). Briefly, the burned area was divided into three regions (see Fig. 3), designated as region 1 (site of inoculation, caudal one-third of wound), region 2 (middle one-third of wound, adjacent to inoculation site), and region 3 (cervical one-third of wound). Sections of burn wound were aseptically harvested from the center of each region 24 or 48 h after inoculation and were weighed, minced, and homogenized in sterile saline. The muscle tissue underlying the inoculation region was also collected, homogenized, and cultured for examination of vertical spread of bacteria. To examine systemic dissemination of bacteria, the spleen, lungs, heart, and liver were aseptically harvested. Tissues were weighed and homogenized in sterile saline. Serial dilutions of tissue homogenates were grown overnight on tryptic soy agar for determination of CFU/gram of tissue, wet weight. For survival studies, mice were monitored daily for up to 2 wk following wound inoculation.

Recombinant human Flt3L was provided by Amgen. Flt3L (10 μg in 0.1 ml of LR) was administered by i.p. injection once daily for 5 days before the burn and wound infection. Control mice received injections of LR alone. For adoptive transfer experiments, cells were harvested from spleens on day 6 from donor mice that had been treated with or without Flt3L for 5 days. Recipient mice received 10⁷ dendritic cells, 7.5 × 10⁶ NK cells, or 1 × 10⁸ DX5⁻CD11c⁻ splenocytes by i.p. injection 24 h before burn injury and wound infection. The numbers of cells transferred were based on the average numbers of these cells obtained from spleens of a group of donor mice that had been treated with Flt3L for 5 days. Control recipient mice received an equivalent injection of RPMI 1640 (1 ml) 24 h before burn and wound inoculation. In some experiments, dendritic cells and NK cells were labeled with the cell tracer CFSE (Molecular Probes) before adoptive transfer. Specifically, isolated cells were incubated with 5 μM CFSE for 15 min at 37°C, and then washed twice in RPMI 1640 before injection. For detection of CFSE-positive cells, fluorescein fluorescence was examined in PE-labeled DX5⁺ or CD11c⁺ cells by FACS analysis.

Splenocytes were manually dispersed as described (22) and the RBC were lysed (erythrocyte lysis kit; R&D Systems). Dendritic cells were positively selected based on CD11c expression, and NK cell selection was based on DX5 expression. Specifically, splenic leukocytes were incubated with either anti-CD11c or anti-DX5 Ab-coupled magnetic microbeads (Miltenyi Biotec) for 15 min at 4°C. The Ab-bound cells were then isolated by passage of the splenocytes through MACS columns fitted with 30-μm preseparation filters (Miltenyi Biotec). The highly enriched CD11c^high and DX5⁺ fractions were used for dendritic cell and NK cell adoptive transfer experiments, respectively.

Splenic mononuclear cells and enriched cell populations (1 × 10⁶) were incubated with 0.1 mg/ml Fc block (BD Pharmingen) for 5 min at 4°C before addition of 0.5 μg of Abs of interest. Cells were incubated with Abs for 30 min at 4°C, washed in PBS, and collected by centrifugation at 300 × g for 10 min at 4°C. Cells were reconstituted in 250 μl of 1% paraformaldehyde and analyzed with a FACSort flow cytometer (BD Biosciences). Specific staining was determined by comparison with appropriate Ab isotype controls. Fluorescence-conjugated Abs were purchased from Caltag Laboratories.

GraphPad Prism 4.0 for Windows (GraphPad Software) was used for all statistical analyses. Survival curves were compared by the log-rank test. Comparisons of group means were performed using an unpaired, two-tailed Student’s t test.

Treatment of mice with Flt3L for 5 days increases the number of dendritic cells in the spleen (22). To determine whether stimulated expansion of dendritic cells is associated with an increased resistance to burn wound infections, we used a previously established model of P. aeruginosa burn wound infection (25, 30). The burn injury alone did not cause any mortality during the experimental period. However, inoculation of the burn wounds with 500 CFU of P. aeruginosa induced 86% mortality typically between 3 and 6 days after wound inoculation (Fig. 1). Mortality after wound infection was associated with systemic dissemination of bacteria and colonization of major organs. Specifically, 80% of the control mice had large numbers of bacteria in cultures of spleens, lungs, hearts, and livers 3 days after inoculation of burn wounds (Fig. 2). Systemic dissemination was examined 3 days after wound inoculation, because mortality began to occur at this time. Bacterial cultures of organs harvested from Flt3L-treated mice were essentially negative. Specifically, bacteria were detected in organ homogenates from only two mice, and these colony counts (<300 CFU/gram of tissue, wet weight) were negligible (Fig. 2). Decreased bacterial burden in mice that had been treated with Flt3L was associated with significantly greater survival. Specifically, mortality occurred in only 10% of the Flt3L-treated mice compared with 86% of the control-treated mice (Fig. 1).

To determine whether the decrease in systemic dissemination of bacteria in mice treated with Flt3L was due to improved bacterial clearance at the wound site, quantitative cultures were performed with different regions of the burn wound and the underlying muscle harvested 24 and 48 h following burn wound inoculation. At 24 h after burn wound inoculation, the percentages of mice with positive bacterial cultures in each wound region were similar in the Flt3L- and LR-treated groups. Numbers of bacteria in the inoculated region and the adjacent wound (region 2) were also similar (Fig. 3,A). However, the numbers of bacteria (CFU/gram) in region 3 and in the underlying muscle were lower in the Flt3L-treated mice than in the control-treated mice. By 48 h after wound inoculation, the percentages of Flt3L-treated mice that had positive cultures in each region were lower than in the control treatment group (Fig. 3 B). Additionally, mean bacterial counts were lower in the underlying muscle and in regions 2 and 3 of the Flt3L-treated mice than in the same regions of the control-treated mice.

To determine whether the effect of Flt3L on increased resistance to wound infection was due specifically to the expansion of the dendritic cell population, adoptive transfer experiments were performed using dendritic cells that had been isolated from Flt3L- or LR-treated mice. Following CD11c enrichment, ∼88% of the isolated cells both expressed high levels of CD11c and had high forward light scatter properties (data not shown), two distinguishing characteristics of dendritic cells (22, 31). Recipient mice received either 1 × 10⁷ dendritic cells, or RPMI 1640 as a vehicle control, by i.p. injection 24 h before burn and wound infection. We and others have used i.p. injection successfully to transfer lymphocytes to recipient mice (32, 33, 34). To confirm successful transfer of dendritic cells after i.p. injection, adoptive transfer experiments were performed using CFSE-labeled cells. Highly fluorescent CFSE-positive dendritic cells were detected in spleens (Fig. 4) and in blood (not shown) 48 h after i.p. injection. Fig. 5 shows survival in recipient mice after burn wound inoculation. In this experiment, mortality occurred in 52% of the control mice (no adoptive transfer), but was significantly attenuated in the recipient mice that had received dendritic cells from Flt3L-treated mice. Specifically, 95% of the mice that received dendritic cells from Flt3L-treated mice survived after burn wound inoculation. Interestingly, adoptive transfer of the same number of dendritic cells from LR-treated donor mice did not confer resistance to infection. Only 50% of the LR-dendritic cell recipient mice survived after burn wound inoculation (Fig. 5).

It is reported that CD11c is also expressed at low levels on NK cells (35). To exclude the possibility that low numbers of CD11c^low NK cells could be mediating the protective effects induced by Flt3L, we repeated the experiment in Fig. 4, using splenocytes that were first depleted of DX5⁺ NK cells before CD11c enrichment. Depletion of DX5⁺ cells from splenocytes of Flt3L-treated mice before dendritic cell isolation did not significantly alter the survival benefit conferred by CD11c⁺ cells (Fig. 6). Furthermore, we also performed adoptive transfer with the DX5⁺ population and the remaining DX5⁻CD11c⁻ splenocytes from Flt3L-treated mice. Successful transfer of DX5⁺ cells after i.p. injection was confirmed using CFSE-labeled cells (Fig. 4). CFSE-positive NK cells were detected in spleens (Fig. 4) and in blood (not shown) 48 h after i.p. injection. In survival studies, mortality occurred in 60% of the control (no adoptive transfer) mice. The same level of mortality (60%) occurred in the DX5⁻CD11c⁻ splenocytes, which were composed of ∼72% CD3 (T cell marker)-positive, 11% CD19 (B cell marker)-positive, 4% F4/80 (macrophage marker)-positive, and 8% 7/4 Ag neutrophil marker (36)-positive cells (data not shown). Mortality occurred in ∼44% of the mice that received DX5⁺ cells (Fig. 6). However, the percentage of mortalities in the group of mice that had received CD11c⁺DX5⁻ dendritic cells was only 10%. Survival in the dendritic cell recipient mice was significantly greater than in the control (no adoptive transfer) mice and in the DX5⁻CD11c⁻ recipients. Survival in the NK cell recipient mice was not significantly improved.

Dendritic cells are typically classified as either lymphoid-related or DC1 (CD8α⁺CD11b⁻) or myeloid-related or DC2 (CD8α⁻CD11b⁺). FACS analysis was used to determine whether the relative proportions of the major classes of dendritic cells are altered after Flt3L treatment (Fig. 7). Approximately 37% and 32% of the splenic CD11c⁺ dendritic cells were positive for CD8α surface expression in the Flt3L-treated and LR-treated mice, respectively. CD11b expression was detected on ∼57% of the Flt3L DC and 44% of the LR DC. There was also a group of CD8α⁻CD11b⁻ double-negative CD11c⁺ dendritic cells, as reported by others (35). Approximately 20% of splenic dendritic cells were double negative (CD8α⁻CD11b⁻) in the FL DC, whereas 41% of splenic dendritic cells were double negative in the LR DC group.

We previously demonstrated that treatment of severely burned mice with Flt3L expands the dendritic and NK cell populations and restores normal production of IFN-γ and IL-12 in response to a challenge with heat-killed P. aeruginosa (22). In the present study, we used an established model of burn wound infection with P. aeruginosa to evaluate the immunomodulatory properties of Flt3L in burned mice. Treatment of mice with Flt3L significantly increased the survival of mice subjected to a burn wound infection with P. aeruginosa. This increase in survival was associated with an increase in splenic dendritic cell numbers, a decrease in bacterial growth and spread within the wound, and a decrease in systemic dissemination. These effects were specific to Flt3L-expanded dendritic cells, because adoptive transfer of DX5⁺ NK cells or DX5⁻CD11c⁻ splenic leukocytes from Flt3L-treated mice did not confer the same resistance to burn wound infection. Furthermore, adoptive transfer of an equivalent number of dendritic cells from normal (LR-treated) donor mice did not increase the resistance of recipient mice to a burn wound infection, indicating that the Flt3L-induced resistance to wound infection was not simply due to increased dendritic cell numbers. Rather, Flt3L appeared to also modulate the dendritic cell properties such that a greater ability to control infection was conferred after Flt3L treatments. Two major classes of dendritic cells have been described. Lymphoid-related or DC1 dendritic cells are typically characterized as CD8α⁺CD11b⁻, whereas myeloid-related or DC2 are described as CD8α⁻CD11b⁺. Although the phylogenetic origin of these two classes has been debated, it has been demonstrated that both classes of dendritic cells can be generated from both the common myeloid and common lymphoid progenitor cells, which express high levels of the Flt3R. Dendritic cells generated directly from these progenitors maintain a low level of Flt3R expression (20, 37). Therefore, administration of Flt3L acts predominantly on progenitor cells to generate both classes of dendritic cells, but may also have some effects on previously differentiated dendritic cells through interactions with the Flt3R expressed at low levels on dendritic cells. The specific mechanism by which Flt3L enhances dendritic cell functions in this model is not known. However, other studies have reported that Flt3L-generated dendritic cells express higher levels of MHC class II, CD86, and CD40, produce higher levels of IL-12, and are more efficient at stimulating T cell proliferation than normal dendritic cells from mice that did not receive Flt3L treatments (38, 39). Future studies will be performed to fully characterize and compare dendritic cells isolated from normal and Flt3L-treated mice. An important issue to consider is whether or not mobilization of immune cells by Flt3L increases the risk for an exacerbated proinflammatory response after burn injury. Serum levels of IL-6, IL-1β, MIP-1α, and TNF-α were similar or lower in the Flt3L group, compared with LR group, 24 and 48 h after burn wound inoculation (data not shown), perhaps due to the fact that bacterial growth and spread were limited within the wounds of these mice.

This study demonstrated that local wound immunity was improved in Flt3L-treated mice, which likely played a role in the decreased systemic dissemination of bacteria in that group. Flt3L treatment of mice causes an increase in the numbers of dendritic cells in the skin (40). Dendritic cells are primary responders to pathogens and play a central role in the activation of both innate and acquired immune responses (41). Recognition of pathogens by dendritic cells triggers the production of cytokines such as IL-12, IL-2, IL-15, and IFN-γ that activate macrophages, neutrophils, and NK cells during the innate response to infection (42, 43, 44, 45). These cytokines also activate and direct the acquired immune response that is induced when dendritic cells present Ag to T cells. In addition to Ag presentation, dendritic cells also provide costimulatory signals (through CD40 and B7 proteins) that are required for efficient activation of acquired immunity (46, 47). Expansion of dendritic cells should therefore have an impact on the activation of multiple immune functions. Indeed, in normal mice, exogenous Flt3L promotes the induction of IL-12, IFN-γ, and Th1 responses (48, 49), and enhances acquired immunity by promoting the expansion of Ag-specific T cells and increasing Ag-specific Ab production (50, 51). Treatment of mice with Flt3L has been shown to increase resistance to infections with HSV and Listeria monocytogenes (40, 52). These properties of Flt3L and the data presented in this study suggest that stimulation of dendritic cell production by Flt3L may have potential for increasing resistance to infections after burn injury.

In this study, Flt3L was administered before burn and wound infection. We are currently evaluating the efficacy of Flt3L when administered to burned mice both as a prophylactic agent after burns but before onset of infection and as a treatment in the presence of an existing infection. In earlier studies, Flt3L administration after burn injury prevented burn-associated impairments in IFN-γ and IL-12 production (22), suggesting that Flt3L will be effective when administered after injury for increasing the resistance of burned mice to a wound infection. This study suggests that enhancement of dendritic cell production can compensate for the immunological deficits induced by burn injury. It is interesting to note that the Flt3L-expanded dendritic cells were present at the time of severe burn injury, which is known to affect the functions of nearly every type of immune cell. It is not known whether severe burn injury directly impairs dendritic cell functions. Therefore, it cannot be determined whether Flt3L is directly counteracting a potential impairment in dendritic cell functions after burn injury or whether Flt3L is enhancing overall immunity to compensate for burn-induced impairments in multiple cell types. The latter possibility is supported by the findings that in normal, uninjured mice, Flt3L enhances overall immune responses and increases resistance to HSV-1, Listeria monocytogenes, and progressive cutaneous leishmaniasis infections (40, 52, 53). However, a recent study of trauma patients that included some burn patients has suggested that impaired T cell responses after trauma may be due to a decreased promotion of T cell activation by dendritic cells, subsequent to an impairment in dendritic cell differentiation from monocyte precursors (54). Additionally, a significant depletion of dendritic cells in spleens of patients that died from sepsis-associated complications has been reported (55). In an animal model of sepsis, there is significant apoptosis-associated depletion of dendritic cells in the spleen (56). Although the relevance of these findings to burn injury remains to be determined, it is possible that Flt3L-induced enhancement of dendritic cell production and functions after burn trauma can compensate for a loss in numerous immune cell functions.