Secondary Logo

Journal Logo

Basic Science Aspects

DEPLETION OF NK CELLS IN A MURINE POLYTRAUMA MODEL IS ASSOCIATED WITH IMPROVED OUTCOME AND A MODULATION OF THE INFLAMMATORY RESPONSE

Barkhausen, Tanja*; Frerker, Christian*†; Pütz, Claudia*; Pape, Hans-Christoph*‡; Krettek, Christian*; van Griensven, Martijn

Author Information
doi: 10.1097/SHK.0b013e31816e2cda
  • Free

Abstract

INTRODUCTION

Sepsis and associated diseases such as systemic inflammatory response syndrome and multiple organ dysfunction syndrome are common posttraumatic complications on intensive care units. Currently, it is well known that multiple body defense mechanisms such as the immune system and the coagulation and complement systems are involved in the development of those posttraumatic complications. However, the origin of this overwhelming inflammation is not yet clearly defined. Moreover, there is a large demand for the development of new treatment strategies because there is still no effective therapy available.

Natural killer (NK) cells are part of the innate immune system. They represent the first line of immune defense and are therefore crucial for immunosurveillance. Natural killer cells are mainly activated by interferons (IFNs) or IL-12. The latter mediator is found to be exclusively efficient in costimulation with IL-2 or IL-15 in some experimental settings (1). Primarily known as effector cells able to kill tumor and virus-infected cells with no need for previous priming reaction, the name "natural killer cell" was chosen. Besides their ability to deal with viral infections and tumorigenesis, they also play a role in tissue trauma and sepsis (2-8).

Studies revealed that NK cells play a critical role in the development of a systemic inflammatory reaction by producing IFN-γ and subsequent activation of macrophages after induction of further inflammatory events such as production of TNF-α, IL-1β, and IL-6 (6, 9). Moreover, activated NK cells induce the shift of T helper cells to the TH1 subgroup (10) and secrete an array of different cytokines such as transforming growth factor β, IL-3, IL-5, IL-10, and IL-13 (11) that contribute to modulation of the immune response.

Depletion of NK cells in different models concerning bacterial infection revealed that mortality is highly reduced by this treatment (2, 5). Furthermore, inflammatory events, represented by macrophage activation and the secretion of cytokines such as IL-6, IL-12 (2, 12), TNF-α, IL-1β, and IFN-γ (5), are diminished. In addition, bacterial counts in the peritoneum are elevated (12). In contrast to these results, higher neutrophil infiltration in several organs can be detected (12).

Patients suffering from multiple traumas often develop septic complications as a secondary insult. The reactions and complications are often different from those detected in "normal" patients who "only" have sepsis. For this reason, we established an animal model combining a traumatic event, consisting of femur fracture and hemorrhage, with subsequent sepsis (two-hit model). Moreover, previous studies in a cecal ligation and puncture (CLP) model using dehydroepiandrosterone (DHEA) as a therapeutic agent revealed an improved outcome combined with modulated immune parameters in DHEA-treated animals (13, 14). As an interesting observation, we determined a less sustained increase in NK cells in animals treated with DHEA (13). Based on this observation and in accordance to studies in infection models, we designed a study to investigate the effects of NK cells in this model of combined trauma, hemorrhage, and sepsis. Thereby, it is of our interest to evaluate the impact of NK cell depletion on the single traumatic hit, with additional focus on the effect of NK cell depletion on the combination of insults.

In this study, NK cells were depleted by administration of antiasialo-GM1 antibody in C57BL/6J mice. Data are collected after the traumatic impact (one hit) and after the combination of traumatic and septic events (two hits). In the following, we determined mortality rate, white blood cell phenotype, organ histology, and cytokine expression levels in comparison to vehicle-treated mice.

MATERIALS AND METHODS

Animal care

The study was approved by the animal welfare committee of the state of lower Saxony (Germany). 36 male C57BL/6J-mice (Charles River, Sulzfeld, Germany) weighing 20 ± 3 g were used for the study. All animals were handled at room temperature for 14 days before treatment. Throughout the study period, pelleted mouse chow and water were available ad libitum. The lighting was maintained on a 12-h cycle. Analgetic treatment was performed in all animals (200 mg/kg body weight metamizol-sodium; Novalgin, Hoechst, Unterschleißheim, Germany) throughout the study.

All surgical procedures were performed after deeply anesthetizing the animals with 100 mg/kg body weight ketamine (Ketanest; Bayer, Leverkusen, North Rhine-Westphalia, Germany) and 16 mg/kg body weight xylazine (Rompun; Bayer). The mice were warmed to 36°C using infrared warming lamps after having finished the surgical procedures. All mice received subcutaneous injections of 1 mL 0.9% sterile saline for fluid replacement once daily.

Group distribution and experimental procedure

Four different groups were included in the experimental design (Table 1). This consisted of a two-hit-model. The first hit was reflected by a closed femur fracture, followed by volume-controlled hemorrhage. A standardized femur fracture was induced in all groups using a blunt guillotine device with a weight of 500 g. This resulted in an A-type femoral fracture combined with a moderate soft tissue injury. A hemorrhagic shock was induced by withdrawing 60% of the total blood volume (calculated through the body weight of the animals) via orbital puncture 2 h later. Resuscitation using sterile Ringer lactate was performed with four times the shed blood volume in the tail vein after 1 h. This means that every animal received an individual resuscitation regimen.

Table 1
Table 1:
Overview of group distribution, treatment procedure, and group size

In the sepsis groups, the second hit was performed by inducing sepsis 2 days after the first hit. Two groups (vehicle, n = 17; asialo-GM1, n = 6) underwent CLP to induce a polymicrobial sepsis. As a control, a sham operation with only a laparotomy was performed in two other groups (vehicle, n = 6; asialo-GM1, n = 7). The surgical procedure was as follows: the cecum was exposed after a midline laparotomy and unilaterally punctured two times using a 21-G needle. The model also includes the protrusion of the contents of the cecum to assure the presence of bacteria in the peritoneum. The closure of the abdomen was performed with double-layer sutures. All data were obtained 96 h after surgery. All animals were clinically observed 96 h after instrumentation.

Administration of an NK cell-depleting antibody (antiasialo-GM1) or vehicle was performed 1 h after trauma/hemorrhage.

Administration of NK cell-depleting antibody

A rabbit antimouse/rat asialo-GM1 polyclonal antibody (Cedarlane, Hornby, Canada) was used for the specific depletion of NK cells. The lyophilized antibody was reconstituted in 1 mL of distilled water. Twenty microliters of this dilution was used for intravenous injection (recommended by the manufacturer). As a vehicle, 20 μL of sterile phosphate-buffered saline (PBS) mimicking the salt content in the antibody dilution was administered intravenously.

Collection of organ samples

After a newly performed laparotomy, mice were exsanguinated by cardiac puncture. For polymerase chain reaction (PCR) analysis of cytokine expression, samples of liver and lung were collected immediately after exsanguination. One lobe of each organ was excised and put into a microfuge tube. These specimens were immediately snap-frozen in liquid nitrogen and stored at −80°C until further processing. Remaining lung and liver specimens as well as kidney and half a spleen were immediately stored in 4% formalin for histological assessment.

Lymphocyte phenotyping by flow cytometry

Blood derived from cardiac puncture was used for the immunological assessment. To prepare single-cell solutions from the second half of the spleen, it was minced by grinding through a metal wire with a pore size of 0.5 mm. In all samples, erythrocytes were lysed for 15 min using a hypotonic solution containing 0.16 mM NH4Cl, 0.27 mM EDTA, and 11.9 mM KHCO3. Cells were washed with PBS and dissolved in a specific staining buffer (BenderMed Systems, Vienna, Austria) afterward. Staining of 100-μL cell suspension was performed with 5 μL of each antibody staining agent for 30 min. Staining was performed in three different tubes.

To solely select lymphocytes for analysis, cells were gated for lymphocytes during the measurements using forward and side scatter. This gate had been confirmed using a CD3 antibody in previous experiments and in the NK1.1 staining samples.

Total proportion of NK cells was determined by fluorescence-activated cell sorter analysis with antibodies NK1.1-fluorescein isothiocyanate (FITC) and CD3-PE (BD Biosciences, Heidelberg, Germany). NK1.1+ CD3 cells were defined as NK cells. Mouse-specific Ly49-FITC (BD Biosciences) was used in a single staining to depict activated NK cells and to define the degree of NK cell depletion.

For determination of CD4-CD8 ratios, cells were incubated with mouse-specific CD4-FITC and CD8-PE (BD Biosciences). White blood cell count was performed by Pappenheim staining.

Furthermore, annexin-FITC and PI (BenderMed Systems, Vienna, Austria) were double-stained to discriminate living, apoptotic, and necrotic cells in the lymphocyte population. After finishing the staining procedure, cells were washed again with PBS and were dissolved in 200 μL PBS for flow cytometric measurement.

RNA purification and cytokine quantification by PCR

For semiquantifying RNA message, the frozen organ samples were homogenized in TRIZOL reagent (Invitrogen, Carlsbad, Calif) using an ultraturrax (IKA Labortechnik, Staufen, Germany). The purification was performed as recommended by the manufacturer. For each sample, 2 μg of purified RNA was reversely transcribed into cDNA by moloney murine leukemia virus reverse transcriptase (Invitrogen) using oligo(dT)12-18 primer (Invitrogen). Cytokine transcription was detected by semiquantitative PCR using specific primer pairs for murine TNF-α, IL-1β, IL-6, and IL-10 (Table 2). The amount of cycles was chosen to be in the exponential phase of the PCR. The amount of the specific PCR product was quantified densitometrically using UV radiation and specific software (Bio2, Biometra, Göttingen, Germany). The values were normalized by calculating the quotient of amount of cytokine mRNA against the amount of the housekeeping gene GAPDH.

Table 2
Table 2:
Primer sequences and length of PCR products in base pairs (bp) for murine TNF-α, IL-1β, IL-6, IL-10, and GAPDH

Histological assessment

Lung and liver tissue was stained with hematoxylin and eosin as described previously (15). A scoring system was used to quantify immune cell accumulation. For lung tissue, the degree of interstitial thickening was additionally determined. A score of 0 reflects no or just marginal; a score of 1, medium; and a score of 2, strong neutrophil accumulation and interstitial thickening.

Immunohistochemistry

Immunohistological staining of infiltrating neutrophils was conducted using antimouse Ly-6G (BD Biosciences) on paraffin-embedded 3-μm-thick tissue sections. Epitopes were unmasked by microwave treatment in citrate buffer (10 mM citric acid; pH 6.0). Subsequently, endogenous peroxidase activity in the sections was inhibited using 3% hydrogen peroxidase (Merck, Darmstadt, Germany) in methanol (JTBaker, Deventer, the Netherlands). They were incubated with the primary antibody for 1 h. After washing, an incubation with the secondary antibody (rabbit-antirat immunoglobulin G; horseradish-peroxidase conjugated, Dako, Hamburg, Germany) was performed for 1 h. Reactions were visualized by diaminobenzidine (Dako). For evaluation of number of infiltrating neutrophils, sections were blinded, and positive cells were counted in four representative high-power fields per section at ×400 magnification by two investigators (T.B. and C.F.).

Statistics

Statistical analysis was performed using a standard software application (SPSS). Comparisons between groups were performed using one-way ANOVAs and a post hoc Tukey test. Survival was compared using a chi-square test. To calculate significances in cytokine mRNA expression and secretion, a t test was used. Probability values less then 0.05 were considered statistically significant. The data are expressed as mean ± SEM.

RESULTS

Mortality

In our murine polytrauma model, we determined a mortality rate of 47% in the CLP group with preceding femur fracture/hemorrhage receiving vehicle (two-hit vehicle). In contrast to this elevated death rate, depletion of NK cells in the CLP group (two-hit asialo-GM1) resulted in a complete abrogation of mortality (0%; P ≤ 0.05; Fig. 1). Furthermore, all animals of the one-hit groups survived the procedure (mortality rate in one-hit vehicle and one-hit asialo-GM1, 0%).

Fig. 1
Fig. 1:
Percentages of survival in the trial groups. *Two-hit vehicle is significantly reduced compared with two-hit asialo-GM1; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

NK cell depletion

Blood compartment

The proportion of activated NK cells is significantly reduced in the groups receiving NK cell-depleting antibody in comparison to the corresponding vehicle-treated groups as detected using flow cytometry (P ≤ 0.01). We detected 11.71% ± 2.85% positive cells in the two-hit vehicle group and 15.17% ± 5.75% positive cells in the one-hit vehicle group. In comparison, we found 2.21% ± 0.76% in the two-hit asialo-GM1 group and 1.53% ± 0.82% in the one-hit asialo-GM1 group 96 h after administration of the depleting antibody (Fig. 2). In contrast, we did not find significant differences in proportions of total NK cells between corresponding groups. However, the proportion of activated NK cells is significantly reduced compared with the proportion of total NK cells within both antibody-depleted groups (P ≤ 0.01).

Fig. 2
Fig. 2:
Proportion of NK cell populations (%) in blood compartment. Symbol, percentage of total NK cells; Symbol, percentage of activated NK cells. *Significant difference between the marked groups; †NK cell count in the depleted group is significantly changed compared with the corresponding vehicle group; one-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.
Symbol
Symbol
Symbol
Symbol

Spleen compartment

The percentage of activated NK cells is significantly reduced in the antibody-receiving groups in comparison to the vehicle-treated groups (P ≤ 0.01). We detected 3.16% ± 0.29% vs. 2.86% ± 0.18% positive cells using flow cytometry in the two-hit asialo-GM1 group and in the one-hit asialo-GM1 group, respectively. Furthermore, we determined values of 7.3% ± 2.02% for two-hit vehicle and 17.84% ± 2.45% for one-hit vehicle. The proportion of activated NK cells is significantly diminished in the two-hit vehicle group compared with the corresponding one-hit vehicle group (P ≤ 0.01; Fig. 3). Comparing proportions of total NK cells versus activated NK cells within one group, activation of NK cells is suppressed in the antibody-depleted groups and in between the two-hit vehicle group (P ≤ 0.01). As in the blood compartment, the proportion of total NK cells is not significantly changed between corresponding groups.

Fig. 3
Fig. 3:
Proportion of NK cell populations (%) in spleen compartment. Symbol, percentage of total NK cells; Symbol, percentage of activated NK cells. *Significant difference between the marked groups; †NK cell count in the depleted group is significantly changed compared with the corresponding vehicle group; one-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

Tissue cytokine expression

Liver

Expression levels of IL-6 mRNA were significantly diminished in both NK cell-depleted groups in comparison to the corresponding vehicle-receiving groups (P ≤ 0.01). Levels ranged from 0.03 ± 0.02 in the two-hit asialo-GM1 group and 0.03 ± 0.01 in the one-hit asialo-GM1 group to 0.28 ± 0.08 in the two-hit vehicle and 0.23 ± 0.03 in the one-hit vehicle group (Fig. 4). mRNA expression levels of TNF-α, IL-1β, and IL-10 were not altered by any treatment.

Fig. 4
Fig. 4:
Relative expression levels of IL-6 in liver tissue determined by semiquantitative PCR. *Expression is significantly changed compared with the corresponding vehicle group. One-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

Lung

We could not detect changes in cytokine mRNA expression patterns of lung tissue for any cytokine.

CD4-CD8 ratios

Blood lymphocytes

CD4-CD8 ratios were significantly lower after the second hit in comparison to the corresponding one-hit group as detected using flow cytometric analysis. The following ratios were found in between the blood lymphocyte population: one-hit vehicle, 1.58 ± 0.07; two-hit vehicle, 1.07 ± 0.37 (one-hit vehicle versus two-hit vehicle, P < 0.05); one-hit asialo-GM1, 1.48 ± 0.04; two-hit asialo-GM1, 0.86 ± 0.19 (one-hit asialo-GM1 versus two-hit asialo-GM1, P = 0.004; Fig. 5).

Fig. 5
Fig. 5:
Ratio of CD4-CD8+ cells determined in circulation. Measurements of CD4 and CD8 were performed by flow cytometry. *Difference in CD4-CD8 ratio is significant between the marked groups. One-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

Splenic lymphocytes

CD4-CD8 ratios of spleen lymphocytes exhibited no differences between single groups as analyzed by flow cytometry.

Apoptosis rates

Blood lymphocytes

Apoptosis rate of the vehicle-treated groups has a higher tendency than in antibody-depleted groups. This was observed using annexin V staining in a flow cytometric setting. Nevertheless, this difference is not statistically significant in our setting (Fig. 6).

Fig. 6
Fig. 6:
Apoptosis rate of lymphocytes in blood and spleen. Symbol, determination of apoptosis rate in blood compartment; Symbol, determination of apoptosis rate in spleen compartment. *Apoptosis rate is significantly increased compared with the other groups in spleen. One-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

Splenic lymphocytes

In contrast to the results examined for blood lymphocytes, we measured a significantly increased apoptosis rate in spleen lymphocytes of the one-hit vehicle group (9.24% ± 2.82%; Figs. 6 and 7A). All other groups exhibited only moderate apoptosis rates (two-hit vehicle, 0.61% ± 0.2%; Figs. 6 and 7B), one-hit asialo-GM1, 0.69% ± 0.26%; two-hit asialo-GM1, 0.73% ± 0.09%; one-hit vehicle versus two-hit vehicle, P = 0.003; one-hit vehicle versus one-hit asialo-GM1, P = 0.005; Fig. 6).

Fig. 7
Fig. 7:
Representative dot plots of Annexin V-FITC measurement (apoptosis). Data show representative measurements of group one-hit vehicle (A) and two-hit vehicle (B) lymphocytes derived from spleen according to statistical data. One-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group.

Histology

Liver

Assessment of histological specimens derived from liver tissue revealed significantly less infiltration of immune cells after antibody in comparison to vehicle treatment (two-hit vehicle, 1.31 ± 0.18; two-hit asialo-GM1, 0.67 ± 0.21; P < 0.05). The evaluation of infiltration in the one-hit groups revealed similar values for both groups (one-hit vehicle, 1.00 ± 0.26; one-hit asialo-GM1, 1.00 ± 0.00; Fig. 8).

Fig. 8
Fig. 8:
Determination of the degree of infiltrating immune cells in lung and liver tissue by hematoxylin and eosin staining. Evaluation was performed using a score from 0 (no or just marginal immune cell accumulation) to 2 (strong immune cell accumulation). Symbol, Evaluation of the degree of infiltration in liver tissue; Symbol, evaluation of the degree of infiltration in lung tissue. *Infiltration is significantly changed compared with the corresponding vehicle group. One-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

We could not detect significant differences of neutrophil infiltration in liver specimens between single groups. We determined values of infiltrating cells of 1.5 ± 0.5 for one-hit vehicle, 1.1 ± 0.3 for two-hit-vehicle, 1.5 ± 0.3 for one-hit asialo-GM1, and 0.6 ± 0.6 for two-hit asialo-GM1 (Fig. 9).

Fig. 9
Fig. 9:
Degree of neutrophil infiltration in lung and liver specimens. Degree of infiltration was determined by the count of Ly-6G+ cells in four representative high-power fields. Symbol, Neutrophil count in liver tissue; Symbol, neutrophil count in lung tissue. *Level of infiltration is significantly changed compared with the corresponding vehicle group. One-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

Lung

As found in liver tissue, we evaluated a significant decrease of infiltrating immune cells in lung tissue after medication with the NK cell-depleting antibody in sepsis animals (two-hit vehicle, 2.00 ± 0.00; two-hit asialo-GM1, 0.5 ± 0.22; P < 0.001; Fig. 8).

Determination of neutrophil infiltration revealed similar results. Invasion of neutrophils was significantly diminished in antibody-treated groups in comparison to the concerned vehicle groups (Fig. 9).

We determined a neutrophil count of 7.43 ± 0.76 for one-hit asialo-GM1 compared with 29.35 ± 0.97 for one-hit vehicle (P = 0.002). Evaluation of the two-hit groups revealed values of 6.46 ± 1.4 for two-hit asialo-GM1 and 35.94 ± 1.4 for two-hit vehicle (P = 0.003; Figs. 9 and 10).

Fig. 10
Fig. 10:
Immunohistochemical staining of murine neutrophils in paraffin-embedded lung tissue specimens with antibody Ly-6G. Staining was visualized using diaminobenzidine. A, Stained lung specimen derived from group two-hit vehicle. B, Stained lung specimen derived from group two-hit asialo-GM1. Black arrows, examples of stained neutrophils. Two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

In addition, a significant decrease in interstitial thickening can be recognized after antibody administration. This effect can be observed in both one-hit and two-hit animals (two-hit vehicle, 1.88 ± 0.13; two-hit asialo-GM1, 0.50 ± 0.22; one-hit vehicle, 1.83 ± 0.17; P = 0.01; one-hit asialo-GM1, 0.57 ± 0.30; P < 0.05; Fig. 11).

Fig. 11
Fig. 11:
Degree of interstitial thickening in lung tissue. Evaluation was performed in hematoxylin and eosin-stained specimens using a score from 0 (no or just marginal interstitial thickening) to 2 (strong interstitial thickening). *Degree of interstitial thickening is significantly changed compared with the corresponding vehicle group. One-hit vehicle, femur fracture/hemorrhage, followed 2 days later by laparotomy in a vehicle-treated group; two-hit vehicle, femur fracture/hemorrhage, followed 2 days later by CLP in a vehicle-treated group; one-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by laparotomy in a group receiving NK cell-depleting antibody; two-hit asialo-GM1, femur fracture/hemorrhage, followed 2 days later by CLP in a group receiving NK cell-depleting antibody.

DISCUSSION

In the present study, we were able to demonstrate that depletion of NK cell results in resistance to injury caused by a combination of trauma, hemorrhage, and sepsis. The most important and new finding in this study is that all animals survived after depletion of their NK cells in comparison to death rates of approximately 50% in animals without NK cell depletion in this model. Data from several infection trials are in part congruent with the results derived from our multihit study. In these models, depletion of NK cells by antibody treatment or deficiency of NK cells induces protection against bacterial challenges such as Streptococcus pyogenes or Escherichia coli (2, 5). However, in contrast to our results, protection is only partial in many of these trials because they use more severe animal models with higher bacterial loads. Furthermore, those animals were observed for an extended period, which is not allowed in our institution.

Besides demonstrating the depletion of NK cells by antibody treatment, we determined rates of activated NK cells. In spleen, we found a reduction of activated NK cells in the two-hit compared with the one-hit group without NK cell depletion. Similar but not as distinct data can be obtained for blood NK cells. This was surprising because a previous study of the group, performed with a single septic insult, revealed opposite data. In that study, the amount of activated NK cells determined in blood increased nearly 10-fold after sepsis induction compared with a control group (16). The difference in outcome might be reasoned in the preceding insults to sepsis induction in our setting of femur fracture/hemorrhage compared with trauma/thermal injury that contributes to a shift in the time course or even may contribute to an opposite effect in NK cell activation (7, 8). Moreover, contrary results concerning NK cell activation deactivation after sepsis onset in different settings can be obtained from literature (17, 18). Tsujimoto et al. (19) found out that NK cell activation in lymph nodes is induced by flagellin, whereas it is decreased by different concentrations of LPS. Thus, it can strongly be assumed that the detected NK cell-specific reaction after infection is different and dependent on several parameters such as time, kind of insult, body compartment, and maybe type of infection. In addition, the results may vary depending on the receptor specificity of the antibody used in the experimental setting.

The powerful improvement in outcome in the present study is associated with a reduction of IL-6 mRNA expression in liver tissue of NK cell-depleted animals. IL-6 is one of the key cytokines in inflammatory conditions, with proven proinflammatory and anti-inflammatory properties (20-22). Up to now, several studies brought evidence that high IL-6 levels in serum indicate worsened outcome (23, 24). However, IL-6 alone as a prognostic marker for sepsis could not be established in the clinical routine yet.

Natural killer cells are originally known to defend viral infections and tumor cells (25, 26). However, newer findings suggest that NK cells play also a role in bacterial infections and trauma (1, 4, 5, 9), a theory that is supported by our results.

As a target for the modulated IL-6 response by NK cell depletion in liver tissue, Kupffer cells can strongly be assumed. Kupffer cells, which represent a liver-resident type of macrophage, are potent producers of IL-6. Recent findings revealed that Kupffer cells are primed by traumatic/hemorrhagic insults to produce IL-6 and monocyte chemoattractant protein 1 (27). In contrast, macrophages of different cell types such as lung and spleen did not disclose an elevated cytokine response in that study (27). This is congruent with the fact that we could not detect a modulation of IL-6 mRNA expression in lung tissue by NK cell depletion. However, Sherwood et al. (3) found a reduction of IL-6 expression in spleen and heart in mice deficient of NK and CD8+ cells after CLP treatment. This might point to a differential regulation of spleen-specific IL-6 expression after trauma/hemorrhage or sepsis and/or a specific role of T-cell subpopulations. In addition, the down-regulating effect in tendency on IL-6 levels by NK cell depletion can be observed systemically. However, the repercussion on serum IL-6 levels is not as strongly pronounced as in liver specimens (data not shown).

In that view, our results show that NK cells are involved in the inflammatory response after trauma and sepsis because we could first demonstrate that depletion results in a reduction of liver IL-6 expression. It can be suggested that NK cells are involved in the priming of Kupffer cells after trauma/hemorrhage and sepsis and are thereby involved in the detrimental outcome after such impacts.

The effect on inflammation, reflected by IL-6 levels, can be underlined by a further novel finding of this study. Evaluating the degree of infiltrating immune cells in several organs, we detected a pronounced reduction of infiltrating cells in liver and lung after depletion of NK cells. An increase in neutrophil extravasation in lungs can be observed after both trauma/hemorrhage alone and after a combination of that hit with subsequent sepsis. This is not surprising because neutrophil activation in the lung is well known to occur in response to infection (28-30) just as to tissue trauma/hemorrhage (27, 31). This detected increase in neutrophil infiltration can be blocked to a major extent by depleting NK cells. Thus, it can be concluded that NK cells may play a major role in neutrophil activation in lung tissue after traumatic/hemorrhagic events. In addition, we could define only a slight increase after two-hit compared with one-hit insults in vehicle groups. This indicates for the first time that the infective/septic challenge exhibits no further challenge concerning recruitment of neutrophils if trauma and hemorrhage had already occurred. A further point of interest is that we did not observe changes in neutrophil counts in liver tissue after antibody treatment. This does not seem to be a sign of an NK cell-independent neutrophil influx in liver but rather be traced back on very low counts of infiltrating neutrophils at all (approximately one per high-power field). This is exceptional because many studies document higher rates of liver-specific neutrophil influx (27, 31, 32). Again, this specificity might be associated with the late measurement point and indicates varying kinetics in lung and liver.

Neutrophils are activated and attracted by inflammatory stimuli such as TNF-α, IL-1β, granulocyte-macrophage colony-stimulating factor and IFN-γ and chemokines such as IL-8 and growth-regulated oncogene-α (33-36). In this study, we were not able to detect differences in TNF-α and IL-1β levels between antibody and vehicle-treated groups. Both cytokines are induced early in the inflammatory cascade and are known to exhibit mainly proinflammatory properties. Our measurements were conducted a few days after the first hit in these experiments, and values were low in all trial groups. As induction of the first proinflammatory cytokines takes place in between hours after an insult, changes in expression levels might not be detected at the end of observation. However, we suggest that the lack of activated NK cells in a setting of both trauma and sepsis results in an overall diminishment of the inflammatory cascade, which results in a decrease in effector cells in remote organs.

Another finding in our study was that CD4-CD8 ratios of circulating leukocytes are depressed after two hits compared with the corresponding one-hit group. It is known from literature that ratios of CD4-CD8 display a shift after a CLP-induced sepsis on the basis of an increased level of circulating CD8+ cells (16). Similar CD4-CD8 ratio depression can be obtained after soft tissue trauma and hemorrhage; moreover, this change seems to be dependent on male sex and age (37). Nevertheless, NK cells are not involved in this specific shift.

Lymphocyte apoptosis is discussed as an important factor contributing to immunosuppression after severe trauma and sepsis. It is widely known that different leukocyte cell types exhibit changes in their apoptosis pattern after sepsis and shock (38-40). In the present study, we evaluated lymphocyte apoptosis rates in spleen and blood compartment. An intensely up-regulated apoptosis rate was only found in animals undergoing trauma/hemorrhage with functional NK cells (one-hit vehicle). All other groups displayed moderate apoptosis rates. It can be reasoned that a second hit induces a shift in the apoptotic time course or counteracts the effect of trauma/hemorrhage. This is suggested itself because it is known that sepsis alone results in an increase in lymphocyte apoptosis (38), and therefore, opposite results could be expected after a second hit. Furthermore, NK cells seem to play a role in the induction of apoptosis because the corresponding NK cell-depleted group (one-hit asialo-GM1) exhibited only moderate apoptosis rates after trauma/hemorrhage. This goes along with previous findings that suggest that NK cells may work by the release of granzymes and perforins and may contribute to the apoptosis of target cells (41, 42).

To summarize, depletion of NK cells in this model of fracture, hemorrhage, and sepsis shows a highly protective effect on the overall outcome. Therefore, inflammatory events are diminished, which results in a reduced infiltration of effector cells in remote organs. Furthermore, lymphocyte apoptosis, a factor contributing to immunosuppression as a late complication, is also decreased. Macrophages seem to be an important target for NK cell-specific-modulating events in this setting. Thus, we conclude that NK cells play an important role in the development of immune disturbances after trauma and sepsis. In that line, they might represent an important target for new treatment strategies in future trials.

Finally, depletion of NK cells occurred transiently by administrating a single bolus of antibody in this model. Thus, NK cells are diminished in the early phase of the experiment but reappear in the later phase. Measurements 96 h after sepsis induction revealed approximately 2% of activated NK cells in the lymphocyte population. This complies with approximately 15% of the amount determined in an untreated animal. It might be assumed that the depletion in the early phase of the immune reaction induces protective effects by preventing an overwhelming immune reaction. Otherwise, NK cell-specific functions that are regained in the later phase might also be beneficial and contribute to an improved outcome.

REFERENCES

1. Carson WE, Yu H, Dierksheide J, Pfeffer K, Bouchard P, Clark R, Durbin J, Baldwin AS, Peschon J, Johnson PR, et al.: A fatal cytokine-induced systemic inflammatory response reveals a critical role for NK cells. J Immunol 162:4943-4951, 1999.
2. Goldmann O, Chhatwal GS, Medina E: Contribution of natural killer cells to the pathogenesis of septic shock induced by Streptococcus pyogenes in mice. J Infect Dis 191:1280-1286, 2005.
3. Sherwood ER, Enoh VT, Murphey ED, Lin CY: Mice depleted of CD8+ T and NK cells are resistant to injury caused by cecal ligation and puncture. Lab Invest 84:1655-1665, 2004.
4. Emoto M, Miyamoto M, Yoshizawa I, Emoto Y, Schaible UE, Kita E, Kaufmann SH: Critical role of NK cells rather than V alpha 14(+)NKT cells in lipopolysaccharide-induced lethal shock in mice. J Immunol 169:1426-1432, 2002.
5. Badgwell B, Parihar R, Magro C, Dierksheide J, Russo T, Carson WE, III: Natural killer cells contribute to the lethality of a murine model of Escherichia coli infection. Surgery 132:205-212, 2002.
6. Scott MJ, Hoth JJ, Gardner SA, Peyton JC, Cheadle WG: Natural killer cell activation primes macrophages to clear bacterial infection. Am Surg 69:679-686, 2003.
7. Joshi P, Hauser CJ, Jones Q, Kennedy R, Thomae KR, Zhou X: Mechanism of suppression of natural killer cell activity in trauma patients. Res Commun Mol Pathol Pharmacol 101:241-248, 1998.
8. Blazar BA, Rodrick ML, O'Mahony JB, Wood JJ, Bessey PQ, Wilmore DW, Mannick JA: Suppression of natural killer-cell function in humans following thermal and traumatic injury. J Clin Immunol 6:26-36, 1986.
9. Heremans H, Dillen C, Van Damme J, Billiau A: Essential role for natural killer cells in the lethal lipopolysaccharide-induced Shwartzman-like reaction in mice. Eur J Immunol 24:1155-1160, 1994.
10. Billiau A, Heremans H, Vermeire K, Matthys P: Immunomodulatory properties of interferon-gamma. An update. Ann N Y Acad Sci 856:22-32, 1998.
11. Seki S, Osada S, Ono S, Aosasa S, Habu Y, Nishikage T, Mochizuki H, Hiraide H: Role of liver NK cells and peritoneal macrophages in gamma interferon and interleukin-10 production in experimental bacterial peritonitis in mice. Infect Immun 66:5286-5294, 1998.
12. Godshall CJ, Scott MJ, Burch PT, Peyton JC, Cheadle WG: Natural killer cells participate in bacterial clearance during septic peritonitis through interactions with macrophages. Shock 19:144-149, 2003.
13. van Griensven M, Dahlweid FM, Giannoudis PV, Wittwer T, Bottcher F, Breddin M, Pape HC: Dehydroepiandrosterone (DHEA) modulates the activity and the expression of lymphocyte subpopulations induced by cecal ligation and puncture. Shock 18:445-449, 2002.
14. Hildebrand F, Pape HC, Hoevel P, Krettek C, van Griensven M: The importance of systemic cytokines in the pathogenesis of polymicrobial sepsis and dehydroepiandrosterone treatment in a rodent model. Shock 20:338-346, 2003.
15. Seekamp A, van Griensven M, Breyhan K, Pohlmann TH: The effect of fucoidin on the remote pulmonary injury in a two-hit trauma model in mice. Eur J Trauma 6:317-326, 2001.
16. Hildebrand F, Pape HC, Harwood P, Wittwer T, Krettek C, van Griensven M: Are alterations of lymphocyte subpopulations in polymicrobial sepsis and DHEA treatment mediated by the tumour necrosis factor (TNF)-alpha receptor (TNF-RI)? A study in TNF-RI (TNF-RI(−/−)) knock-out rodents. Clin Exp Immunol 138:221-229, 2004.
17. Giamarellos-Bourboulis EJ, Tsaganos T, Spyridaki E, Mouktaroudi M, Plachouras D, Vaki I, Karagianni V, Antonopoulou A, Veloni V, Giamarellou H: Early changes of CD4-positive lymphocytes and NK cells in patients with severe Gram-negative sepsis. Crit Care 10:R166, 2006.
18. el Sameea ER, Metwally SS, Mashhour E, el Bendary A, Hassan AM, el Sharkawy H, el Shennawy FA: Evaluation of natural killer cells as diagnostic markers of early onset neonatal sepsis: comparison with C-reactive protein and interleukin-8. Egypt J Immunol 11:91-102, 2004.
19. Tsujimoto H, Uchida T, Efron PA, Scumpia PO, Verma A, Matsumoto T, Tschoeke SK, Ungaro RF, Ono S, Seki S, et al.: Flagellin enhances NK cell proliferation and activation directly and through dendritic cell-NK cell interactions. J Leukoc Biol 78:888-897, 2005.
20. Hildebrand F, Pape HC, Krettek C: [The importance of cytokines in the posttraumatic inflammatory reaction]. Unfallchirurg 108:793-803, 2005.
21. Xing Z, Gauldie J, Cox G, Baumann H, Jordana M, Lei XF, Achong MK: IL-6 is an antiinflammatory cytokine required for controlling local or systemic acute inflammatory responses. J Clin Invest 101:311-320, 1998.
22. Lin E, Calvano SE, Lowry SF: Inflammatory cytokines and cell response in surgery. Surgery 127:117-126, 2000.
23. Patel RT, Deen KI, Youngs D, Warwick J, Keighley MR: Interleukin 6 is a prognostic indicator of outcome in severe intra-abdominal sepsis. Br J Surg 81:1306-1308, 1994.
24. Casey LC, Balk RA, Bone RC: Plasma cytokine and endotoxin levels correlate with survival in patients with the sepsis syndrome. Ann Intern Med 119:771-778, 1993.
25. Santoli D, Koprowski H: Mechanisms of activation of human natural killer cells against tumor and virus-infected cells. Immunol Rev 44:125-163, 1979.
26. Cerwenka A, Lanier LL: Natural killer cells, viruses and cancer. Nat Rev Immunol 1:41-49, 2001.
27. Hildebrand F, Hubbard WJ, Choudhry MA, Frink M, Pape HC, Kunkel SL, Chaudry IH: Kupffer cells and their mediators: the culprits in producing distant organ damage after trauma-hemorrhage. Am J Pathol 169:784-794, 2006.
28. Hildebrand F, Pape HC, Harwood P, Muller K, Hoevel P, Putz C, Siemann A, Krettek C, van Griensven M: Role of adhesion molecule ICAM in the pathogenesis of polymicrobial sepsis. Exp Toxicol Pathol 56:281-290, 2005.
29. Miyashita M, Onda M, Matsuda T, Moriyama Y, Takizawa T, Kim DY, Matsuda N, Kawanami O: [Endotoxin-induced lung injury-the role of leukocytes and oxidants]. Kokyu To Junkan 37:65-70, 1989.
30. van Griensven M, Kuzu M, Breddin M, Bottcher F, Krettek C, Pape HC, Tschernig T: Polymicrobial sepsis induces organ changes due to granulocyte adhesion in a murine two hit model of trauma. Exp Toxicol Pathol 54:203-209, 2002.
31. Frink M, Lu A, Thobe BM, Hsieh YC, Choudhry MA, Schwacha MG, Kunkel SL, Chaudry IH: Monocyte chemoattractant protein-1 influences trauma-hemorrhage-induced distal organ damage via regulation of keratinocyte-derived chemokine production. Am J Physiol Regul Integr Comp Physiol 292:R1110-R1116, 2007.
32. Jaeschke H, Farhood A, Fisher MA, Smith CW: Sequestration of neutrophils in the hepatic vasculature during endotoxemia is independent of beta 2 integrins and intercellular adhesion molecule-1. Shock 6:351-356, 1996.
33. Jablonska E, Kiluk M, Markiewicz W, Jablonski J: Priming effects of GM-CSF, IFN-gamma and TNF-alpha on human neutrophil inflammatory cytokine production. Melanoma Res 12:123-128, 2002.
34. Bratt J, Palmblad J: Cytokine-induced neutrophil-mediated injury of human endothelial cells. J Immunol 159:912-918, 1997.
35. Harada A, Sekido N, Akahoshi T, Wada T, Mukaida N, Matsushima K: Essential involvement of interleukin-8 (IL-8) in acute inflammation. J Leukoc Biol 56:559-564, 1994.
36. Geiser T, Dewald B, Ehrengruber MU, Clark-Lewis I, Baggiolini M: The interleukin-8-related chemotactic cytokines GRO alpha, GRO beta, and GRO gamma activate human neutrophil and basophil leukocytes. J Biol Chem 268:15419-15424, 1993.
37. Schneider CP, Schwacha MG, Chaudry IH: Influence of gender and age on T-cell responses in a murine model of trauma-hemorrhage: differences between circulating and tissue-fixed cells. J Appl Physiol 100:826-833, 2006.
38. Hotchkiss RS, Swanson PE, Cobb JP, Jacobson A, Buchman TG, Karl IE: Apoptosis in lymphoid and parenchymal cells during sepsis: findings in normal and T- and B-cell-deficient mice. Crit Care Med 25:1298-1307, 1997.
39. Jimenez MF, Watson RW, Parodo J, Evans D, Foster D, Steinberg M, Rotstein OD, Marshall JC: Dysregulated expression of neutrophil apoptosis in the systemic inflammatory response syndrome. Arch Surg 132:1263-1269, 1997.
40. Matute-Bello G, Liles WC, Radella F, Steinberg KP, Ruzinski JT, Jonas M, Chi EY, Hudson LD, Martin TR: Neutrophil apoptosis in the acute respiratory distress syndrome. Am J Respir Crit Care Med 156:1969-1977, 1997.
41. Kagi D, Ledermann B, Burki K, Zinkernagel RM, Hengartner H: Molecular mechanisms of lymphocyte-mediated cytotoxicity and their role in immunological protection and pathogenesis in vivo. Annu Rev Immunol 14:207-232, 1996.
42. Berke G: Killing mechanisms of cytotoxic lymphocytes. Curr Opin Hematol 4:32-40, 1997.
Keywords:

Sepsis; trauma; leukocytes; cytokines; apoptosis

©2008The Shock Society