Secondary Logo

Journal Logo

Basic Science Aspects

Antibiotics Improve Survival and Alter the Inflammatory Profile in a Murine Model of Sepsis From Pseudomonas aeruginosa Pneumonia

Coopersmith, Craig M.*; Amiot , Daniel M. II*; Stromberg, Paul E.*; Dunne, W. Michael; Davis, Christopher G.; Osborne, Dale F.§; Husain, Kareem D.*; Turnbull, Isaiah R.*; Karl, Irene E.§; Hotchkiss, Richard S.; Buchman, Timothy G.*

Author Information
doi: 10.1097/
  • Free



Sepsis is the leading cause of death in intensive care units, with over 210,000 people succumbing to overwhelming infection in the United States annually (1). A recent epidemiological analysis estimates that 750,000 people develop sepsis each year, at a hospital cost of $16.7 billion (1), and mortality from the disease has increased greater than 90% over the last two decades (2).

Pseudomonas aeruginosa is the most common cause of gram-negative nosocomial pneumonia, and it is associated with a mortality rate of 40%–68% (3–5). Because this represents a common, potentially devastating, clinical problem in critically ill patients, P. aeruginosa pneumonia is an increasingly used animal model of sepsis (3,6–13). Morbidity and mortality vary in this model, depending on the dose and strain of bacteria used, although multiple publications indicate that the histologic features seen in pneumonia in patients are reproduced in the murine model (3,7,11,13). However, no published report examines the mortality effect of the clinically relevant scenario of giving commonly used antibiotics to treat the bacterial infection in mice. This is important because treatment with antibiotics with a gram-negative spectrum of action is standard of care for P. aeruginosa pneumonia, and both the presence and the type of antimicrobial treatment have been demonstrated to effect outcome in other murine models of sepsis (14).

No profile of the systemic cytokine response to murine pneumonia-induced sepsis is currently available. Animal models of sepsis have marked differences in both the timing of appearance as well as levels of cytokines (15,16). Using animal models with different cytokine profiles than those seen in human sepsis may have played a major role in the failure of all but one clinical trial in the treatment of overwhelming infection (15,17,18). Because a model's utility is based, in part, on its ability to replicate the pro- and anti-inflammatory state seen in human sepsis, the lack of knowledge of how pneumonia from P. aeruginosa impacts cytokine levels diminishes its usefulness as an animal model of sepsis.

Regardless of the origins of sepsis, antibiotics have been shown to impact cytokine levels. Antimicrobial agents can cause increases in lipopolysaccharide (LPS) in humans and animals with sepsis (19,20). Increased LPS, in turn, can lead to the production of multiple cytokines, including TNF-α and IL-6, both of which have been associated with increased mortality in sepsis (14,21–23). An example of the impact of antimicrobial agents on cytokines is the Jarisch-Herxheimer reaction, where fevers, rigors, and hypotension after treatment with antibiotics for louse-borne relapsing fever are associated with increased plasma levels of TNF-α, IL-6, and IL-8 (24).

In this study, we examined the effect of two common antibiotics—gentamicin and imipenem—on survival from P. aeruginosa pneumonia-induced sepsis as well as levels of common pro- and anti-inflammatory cytokines at different time points after onset of sepsis with and without antimicrobial treatment. To verify that P. aeruginosa pneumonia is a valid model of monomicrobial gram-negative sepsis as opposed to a model of endotoxemia caused by a large intratracheal bacterial bolus, we also compared cytokine levels in the pneumonia model with levels seen in mice that received intraperitoneal LPS. We also examined whether IL-6 levels would be predictive of mortality as has been shown in other animal models of sepsis (23).


Sepsis model

A transcervical incision was made on 6- to 10-week-old FVB/N mice under halothane anesthesia as previously described (6,10). Mice received an intratracheal injection of a solution containing the ATCC 27853 strain of P. aeruginosa diluted in 40μL of 0.9% NaCl to a final density of between 5 × 108 and 1 × 109 colony-forming units (CFU)/mL as determined by serial dilution and colony counts, corresponding to a dose of 2 × 107 and 4 × 107 CFU/injection. After injection, animals were held vertically for 10 s to increase bacterial delivery into the lung, and the incision was subsequently closed in two layers. Previous studies demonstrated that all animals given this dose of bacteria via intratracheal injection have P. aeruginosa detectable in their bloodstream 16 h postoperatively (6). All studies were approved by the Washington University Animal Studies Committee and were in accordance with the National Institutes of Health guidelines.

Survival studies

Two hours after intratracheal instillation of pathogenic bacteria, mice were given a subcutaneous injection of either gentamicin 0.2 mg (Boehringer Ingelheim Vetmedica, St. Joseph, MO) or imipenem 0.5 mg (Merck, West Point, PA). Both antibiotics were diluted in 1 mL of 0.9% NaCl. These antibiotics were chosen because they both have a gram-negative bacterial spectrum of action, and each are effective against this strain of P. aeruginosa in vitro (unpublished observations by C. Coopersmith, D. Amiot, P. Stromberg, W. Dunne, C. Davis, D. Osborne, K. Husain, I. Turnbull, I. Karl, R. Hotchkiss, and T. Buchman). The dose of each antibiotic was based upon the similar weights of animals (approximately 20 g in these age-controlled littermates) and was intended to reflect doses of 10 mg/kg for gentamicin and 25 mg/kg for imipenem, both of which have been frequently used in various strains of mice using differing disease models (14,16,23,25). Control mice were given an injection of 1 mL of 0.9% NaCl at the same time. Six hours after the operation, approximately 80 μL of blood was obtained from each mouse via tail vein for IL-6 levels (analysis described below). Antibiotics were repeated every 12 h for 3 days, and mice were followed 7 days postoperatively for survival. Mortality curves were generated in three separate experiments, and the procedure was performed by a single investigator (D.A.) blinded to whether gentamicin, imipenem, or 0.9% NaCl was injected.

Cytokine determination

A separate group of animals was euthanized and had blood drawn via cardiac puncture at 3, 6, 12, or 24 h after intratracheal injection with P. aeruginosa treated with either gentamicin, imipenem, or 0.9% NaCl (n = 8 at each time for each group, except n = 6 for 24 h treated with 0.9% NaCl due to 80% mortality in this group). An additional group of FVB/N mice received an intraperitoneal injection of 250 μg of LPS instead of intratracheal P. aeruginosa. These animals also had blood drawn via cardiac puncture at the same time points after injection (n = 8 mice at 3, 6, 12, and 24 h). Blood was centrifuged at 6000 g for 5 min to separate plasma. TNF-α, IL-6, IL-10, and IL-12 levels were measured using commercially available enzyme-linked immunosorbent assay (ELISA) according to manufacturer specifications (R&D Systems, Minneapolis, MN). All sample concentrations were determined by comparison with a standard curve run in the same ELISA.


Differences in group survival were analyzed using log-rank analysis. Comparisons of cytokine levels at each time point between animals treated with 0.9% NaCl, gentamicin, or imipenem were analyzed using analysis of the variance (ANOVA) followed by Tukey test if the P value was significant. Comparisons of cytokine levels at each time point between animals treated with 0.9% NaCl or LPS were analyzed using the t test. Data analysis was performed using Prism 3.0 (GraphPad Software, San Diego, CA).


Antibiotics improve survival in pneumonia-induced sepsis

Mice subjected to intratracheal injection of P. aeruginosa were randomized to receive gentamicin, imipenem, or 0.9% NaCl for six doses and were followed for 7 days for survival (Fig. 1). Survival was 100% for mice treated with gentamicin (24/24 alive), 88% for mice treated with imipenem (21/24 alive), and 8% for sham mice (2/24 alive, P < 0.0001 between either antibiotic and 0.9% NaCl, p = NS between antibiotics). Mortality began as early as 12 h after injection of bacteria in mice that did not receive antibiotics with an 80% mortality at 24 h. In contrast, the earliest death seen in a mouse that received imipenem was 60 h postoperatively.

Fig. 1:
Effect of antimicrobial therapy in P. aeruginosa pneumonia. Both gentamicin and imipenem improve survival compared with 0.9% NaCl (saline). Differences in mortality are readily apparent by 24 h postoperatively.

IL-6 levels predict survival

All 72 mice followed for survival had blood taken for IL-6 levels 6 h after surgery to see if this would be predictive of survival (Fig. 2). Average plasma levels were 2937 ± 299 pg/mL for sham mice, 1021 ± 96 pg/mL for mice receiving gentamicin, and 1822 ± 183 pg/mL for mice receiving imipenem (P < 0.001 between either antibiotic and 0.9% NaCl, P < 0.05 between antibiotics). Each treatment was associated with a unique scatter plot of levels (Fig. 2A), with the widest range occurring in animals that received 0.9% NaCl (867 to 5536 pg/mL) and the narrowest range in animals that received gentamicin (284 to 1972 pg/mL). Of note, the only two survivors in the sham group had the lowest IL-6 levels of all 24 mice that received 0.9% NaCl.

Fig. 2:
IL-6 levels 6 h after intratracheal injection of bacteria. (A) Scatter plot of IL-6 concentrations in 72 mice subsequently followed for survival. Levels greater than 1200 pg/mL (24 mice) were associated with 100% survival, whereas levels greater than 3600 pg/mL (10 mice) were associated with 100% mortality. (B) Bar graph of data from same animals to show average IL-6 levels with each treatment. Data shown are mean ± SEM.

IL-6 levels below 1200 pg/mL were associated with a 100% survival (24/24 alive), whereas IL-6 levels above 3600 pg/mL were associated with a 100% mortality (10/10 animals dead). Results between 1200 and 3600 pg/mL were not predictive of mortality, with 23/38 animals having IL-6 levels in this range surviving their septic insult.

Impact of antibiotics on cytokine levels

Levels of the proinflammatory cytokine IL-12 peak at 3 h in mice treated without antibiotics, whereas the proinflammatory cytokines TNF-α and IL-6 as well as the anti-inflammatory cytokine IL-10 all rise more slowly, showing an initial peak at 6 h postoperatively (Figs. 3–6). After this, all cytokines measured show a decrease of varying size between 6 and 12 h and a second peak at 24 h. This is most pronounced in TNF-α and IL-10.

Fig. 3:
Plasma TNF-α levels 3, 6, 12, or 24 h after P. aeruginosa pneumonia. Mice were treated with gentamicin, imipenem, or 0.9% NaCl (saline) beginning 2 h after injection of pathogenic bacteria and given a second dose 12 h later. TNF-α levels were decreased at all time points after 3 h by antimicrobial treatment (* P < 0.05). Data shown are mean ± SEM.
Fig. 4:
Plasma IL-6 levels 3, 6, 12, or 24 h after P. aeruginosa pneumonia. Animals were treated with gentamicin, imipenem, or 0.9% NaCl (saline). IL-6 levels were decreased at all time points after 3 h by antimicrobial treatment (* P < 0.05). Although this represents a different experiment from the one in which mice were subsequently followed for survival, levels of IL-6 with either antibiotic or with saline are similar in both sets of animals (compare Fig. 2B to this figure). Data shown are mean ± SEM.
Fig. 5:
Plasma IL-12 levels 3, 6, 12, or 24 h after P. aeruginosa pneumonia. No statistical differences in IL-12 levels were present at any time point. Mice were treated with gentamicin, imipenem, or 0.9% NaCl (saline) 2 h after injection of pathogenic bacteria and were given a second dose 12 h later. Data shown are mean ± SEM.
Fig. 6:
Plasma IL-10 levels 3, 6, 12, or 24 h after P. aeruginosa pneumonia. Animals were treated with gentamicin, imipenem, or 0.9% NaCl (saline) beginning 2 h after injection of P. aeruginosa and were given a second dose 12 h later. Systemic IL-10 levels were decreased at 6 h (* P < 0.05). Although IL-10 was not detectable in animals treated with either antibiotic at 24 h, this was not statistically significant. Data shown are mean ± SEM.

Antibiotics have widely disparate effects on cytokine levels. Although no differences were seen between animals that received 0.9% NaCl, gentamicin, or imipenem at 3 h, marked differences were noted at all other time points. TNF-α levels are markedly depressed at all later time points by either gentamicin or imipenem (Fig. 3). Six-hour levels fell from 179 ± 29 pg/mL in mice that did not receive antibiotics to 13 ± 4 pg/mL with gentamicin (P < 0.0001) or 43.5 ± 13 pg/mL with imipenem (P < 0.0001). As with all results in this group of experiments, there were no statistically significant differences between cytokine levels in animals that received gentamicin and those that received imipenem. Although levels of TNF-α were much lower by 12 h in mice that received 0.9% NaCl (51.5 ± 13 pg/mL), levels continued to fall with antibiotics to undetectable in all samples with gentamicin (P < 0.01, samples from all groups run on same ELISA) and 25 ± 10 pg/mL with imipenem (p = NS). A substantial second spike of TNF-α levels was present at 24 h in animals that did not receive antibiotics (187 ± 100 pg/mL), but there was no accompanying increase in animals treated with antibiotics: 14 ± 7 pg/mL with gentamicin (p = NS) or 2 ± 2 pg/mL with imipenem (P < 0.05).

Similar results were also observed with IL-6 (Fig. 4). In a separate experiment than the one described in the survival curves above (using the same 24 animals used to examine TNF-α, IL-10, and IL-12 levels), IL-6 levels decreased at 6 h from 3474 ± 535 pg/mL with 0.9% NaCl to 1613 ± 334 pg/mL with gentamicin or 1987 ± 396 pg/mL with imipenem. Although these animals were not followed for survival, the trend in the three groups was similar to those in the 72 animals whose IL-6 levels were determined as part of the survival experiment (compare Fig. 2B with Fig. 4). Significant decreases were also noted in IL-6 at later time points as well: 2917 ± 439 pg/mL in 0.9% NaCl animals down to 212 ± 17 pg/mL with gentamicin (P < 0.001) and 189 ± 22 pg/mL with imipenem (P < 0.001) at 12 h and 3652 ± 1711 pg/mL in sham animals at 24 h, which decreased to 113 ± 48 pg/mL with gentamicin (P < 0.05) and 132 ± 41 pg/mL with imipenem (P < 0.05).

In contrast to TNF-α and IL-6, antibiotics had no significant impact on IL-12 levels at any time point (Fig. 5). The appearance of this cytokine occurred earlier than all others measured and relative levels were similar regardless of whether an animal received antimicrobial therapy.

Levels of the anti-inflammatory cytokine IL-10 fell from 140 ± 48 pg/mL in mice that received 0.9% NaCl at 6 h to undetectable levels in all samples with gentamicin (P < 0.05) and 24 ± 15 pg/mL with imipenem (P < 0.05). Levels were similar between all groups at 12 h. Although IL-10 levels fell from 131 ± 84 pg/mL at 24 h to undetectable in both antibiotic groups, these were not statistically significant secondary to the large standard error in the 0.9% NaCl data.

Cytokine levels in mice with P. aeruginosa pneumonia-induced sepsis or endotoxemia

To verify that the P. aeruginosa pneumonia model used in the above experiments is not simply an endotoxemia model caused by a large bolus of pathogenic bacteria, cytokines were also measured in mice given intraperitoneal LPS. Marked differences were seen between the groups at all time points with all cytokines except IL-10 (Fig. 7, A–D). These differences were most pronounced in IL-6 (Fig. 7A). Initially, IL-6 levels were much higher in LPS versus P. aeruginosa pneumonia: 3633 ± 73 pg/mL at 3 h compared with 935 ± 341 pg/mL (P < 0.0001). By 6 h, levels of the cytokine were similar in both groups, and at later time points, measurements showed elevations in pneumonia animals but substantial reductions in LPS animals (P = 0.05 at 12 h, P < 0.03 at 24 h).

Fig. 7:
Plasma cytokine levels 3, 6, 12, or 24 h after P. aeruginosa pneumonia or LPS injection. Animals were given either an intratracheal injection of P. aeruginosa (black columns) or an intraperitoneal injection of LPS (white columns). IL-6 (A), TNF-α (B), IL-12 (C), and IL-10 (D) levels at all time points are shown. * P < 0.05. Data shown are mean ± SEM.

A similar trend of disproportionate early elevations in cytokine levels in LPS and later elevations in P. aeruginosa pneumonia was seen with TNF-α (Fig. 7B). At 3 h, animals given LPS had TNF-α levels of 341 ± 152 pg/mL compared with 31 ± 5 pg/mL for mice given intratracheal bacterial injection (P = 0.06). By 24 h, TNF-α levels were nearly undetectable in mice that received LPS at 13 ± 7 pg/mL, whereas levels had increased to 187 ± 100 pg/mL that had pneumonia (P = 0.06).

Higher levels of IL-12 were also noted at 3 h in mice that received LPS compared with those that received bacteria: 247 ± 35 pg/mL vs. 68 ± 26 pg/mL (P < 0.02;Fig. 7C). Like IL-6 and TNF-α, cytokine levels rapidly trailed off in mice given LPS; however, unlike these other proinflammatory cytokines, levels stayed relatively low in pneumonia animals as well after an initial peak.


This study demonstrates that both gentamicin and imipenem markedly increase survival from P. aeruginosa pneumonia-induced sepsis. High and low IL-6 levels drawn 6 h after intratracheal injection of pathogenic bacteria are predictive of survival, although intermediate levels do not correlate with mortality. Antibiotics decrease cytokine levels at all time points examined except 3 h in the proinflammatory mediators TNF-α and IL-6, but decrease levels of the anti-inflammatory mediator IL-10 only at 6 h while having no statistically significant impact on IL-12 levels.

These results extend our understanding of the P. aeruginosa pneumonia model of sepsis in a number of ways. First, antibiotics have a profound impact on survival in this model. Survival in animals with sepsis that received 0.9% NaCl postoperatively was nearly zero, whereas antibiotics improved survival to 100% (gentamicin) or nearly 90% (imipenem). Although the concept that antibiotics decrease morbidity and mortality in infection has existed since the development of penicillin, the actual effect of antibiotics can vary widely in different forms of sepsis, even when an appropriate spectrum antimicrobial agent is chosen. In intra-abdominal sepsis, for example, clinical trials have led to an understanding that differing regimens of antibiotics can clearly impact survival in patients (26). In the cecal ligation and puncture (CLP) model of polymicrobial intra-abdominal sepsis, survival is significantly higher in mice treated with imipenem compared with those treated with gentamicin, ciprofloxacin, and clindamycin, which have similar survival to sham-treated mice (14). The survival advantage conferred by imipenem in this CLP study by Newcomb et al. (14) is substantial (45% alive to 72% alive), but much less than the greater than 10-fold advantage conferred by antibiotics in this P. aeruginosa pneumonia study. Although part of this difference is simply due to the higher mortality in the pneumonia model, this may also be due to the monomicrobial nature of this model and the fact that there is a much lower likelihood of abscess formation with pneumonia than with fecal peritonitis. The fact that infection is due to a single organism may also help explain why a single antimicrobial agent was so effective in our results.

The mortality results presented in this manuscript must also be interpreted in light of the fact that the experiments were performed in a murine model. Although multiple studies have examined the effect of antibiotics survival in P. aeruginosa pneumonia in other species, we could not find evidence of mortality studies involving P. aeruginosa pneumonia and FDA approved antibiotics in mice, although one study examined the effect of adding a cationic antimicrobial peptide to aztreonam therapy (27). The species-specific effects are highlighted by multiple studies in guinea pigs demonstrating that survival rates using aminoglycosides are much lower than seen in this study (28–30). For instance, guinea pigs that received gentamicin (which conferred 100% survival in mice) had only a 39% survival in a study by Pennington et al. (28), whereas those that received tobramycin had a 67%–69% survival (29).

Our results also broaden the utility of using IL-6 as a predictor of mortality in murine models of sepsis. Recently published work by Remick et al. (23) demonstrates that IL-6 levels can be used to predict 3-day mortality in a CLP model in BALB/c mice. Similar to this work and other data from this group (31), our results demonstrate that animals with higher levels of IL-6 have higher mortalities. Furthermore, the data also demonstrate that IL-6 levels can be predictive of mortality in a different sepsis model and a different murine strain of mice than previously published. Our finding that levels below 1200 pg/mL were associated with a 100% survival, whereas levels above 3600 pg/mL were associated with a 100% mortality may be useful in designing future experiments to evaluate differences in animals given the same insult, which can be prospectively determined to be destined toward life or death.

The effect of antibiotics on systemic cytokine levels is intriguing. The improvement in survival caused by antimicrobial agents in CLP is not accompanied by alteration in cytokine levels (14), nor is the 10-fold improvement in survival after P. aeruginosa pneumonia caused by intestinal overexpression of the anti-apoptotic protein Bcl-2 (6). Nonetheless, the fact that alterations in cytokine levels are associated with a survival improvement from 8% (no antibiotics) to between 88% and 100% (with antibiotics) is not surprising. The disparate response between the proinflammatory cytokines TNF-α and IL-6 with decreasing levels at all time points (outside of 3 h where all levels are similar) and the anti-inflammatory cytokine IL-10 with statistically significant decreases only at 6 h was somewhat unexpected and merits additional future study. Although IL-12 levels were not altered by antibiotics in this study, IL-12 has been shown to be associated with both human and animal response to infection (35,36), and it is possible that the systemic levels of this cytokine do not reflect its importance in pneumonia-induced sepsis.

The cytokine profile of mice subjected to P. aeruginosa pneumonia but not given antibiotics is also useful toward understanding its clinical relevance as a murine model of sepsis. Together with data using the same model at different time points in an earlier publication from our group (6), we now have a comprehensive assessment of plasma levels of TNF-α, IL-6, and IL-10 at 2, 3, 6, 12, 16, and 24 h after pneumonia-induced sepsis. This allows comparison of this model with other murine models of overwhelming infection (16). Because it has been proposed that one reason many clinical trials of immunomodulators have failed is because they relied on animal studies that did not mimic cytokines seen in human sepsis, understanding how P. aeruginosa pneumonia compares with other models of critical illness is important toward its acceptance as an appropriate murine model of sepsis.

The LPS results presented in this manuscript confirm that P. aeruginosa pneumonia is fully distinct from endotoxemia and does not simply represent a large endotoxin bolus. Cytokine levels are generally low at the earliest time points in animals given P. aeruginosa and rise later. In contrast, FVB/N mice injected with LPS have levels of IL-6, TNF-α, IL-10, and IL-12 that all peak early (3–6 h) and decrease by 12 to 24 h. Of note, although the absolute cytokines concentrations are lower in FVB/N mice than in BALB/c mice with endotoxemia (16), the trend as to when cytokines peak and then decrease is similar between strains.

Although we cannot directly compare P. aeruginosa to CLP, the timing trend of appearance and disappearance of cytokines is similar in FVB/N mice subjected to P. aeruginosa pneumonia and BALB/C mice subjected to CLP. In both models of gram-negative sepsis, animals have markedly lower levels of cytokines at early time points than seen with endotoxemia in the same strain, but concentrations of inflammatory mediators continue to rise at 8 h after the initial insult as opposed to the falling levels seen with endotoxemia. These similarities in cytokine profiles are potentially significant because CLP is generally felt to be a representative model of human sepsis (15), and the trend of cytokine levels—both in terms appearance and disappearance as well as the rate of rising and falling—is much more similar between P. aeruginosa pneumonia and CLP than it is between P. aeruginosa pneumonia and endotoxemia (15,32–34). This similarity in the trend of an animal's early proinflammatory response suggests that like CLP, P. aeruginosa pneumonia may be considered to be a valid model of untreated sepsis of pulmonary origin.

Although this study gives new insights into the P. aeruginosa pneumonia model of sepsis, it has a number of limitations. The results are using only a single strain of animal and might not be applicable to other inbred or outbred mice. In addition, although antibiotics improve survival in humans, the near 100% mortality in this model is probably higher than exists in untreated pneumonia in patients, and the near 100% survival is certainly higher than exists in treated pneumonia, demonstrating this is not a perfect model for human pneumonia-induced sepsis. The dosage of antibiotics given were constant as opposed to the exact weight-based dosing used in many published animal models. Although the littermate mice had similar ages and weights, we cannot rule out a disproportionately high dose of an antibiotic in an animal affected cytokine levels.

The predictive value of IL-6 is limited by the fact that nearly all mice died in one group (no antibiotics) and nearly all mice lived in one group (antibiotics). Although IL-6 levels were able to accurately predict the only two sham mice that lived, the test would be much more useful if applied to a model with a lower mortality where it cannot be predicted in advance if an animal is likely to live or die. Comparing cytokine levels of mice treated with and without antibiotics might also be limited by the relatively small number of animals (n = 8) in each group. For instance, although no differences in cytokine levels were noted between gentamicin- or imipenem-treated animals in experiments done specifically for cytokine levels at 3, 6, 12, or 24 h, there was a small but statistically significant difference between IL-6 levels between the two antibiotic groups in the 48 animals followed for survival despite the fact the IL-6 levels were similar in both groups of experiments (i.e., those done as part of the survival curve in Fig. 2 and those done specifically for cytokines in Fig. 4). In addition, only limited conclusions can be drawn from the 24-h cytokine data in mice with sepsis that received 0.9% NaCl instead of antibiotics because the 80% mortality at this time point ensures that the mice studied represent only a small subset of all animals randomized to this treatment. Although comparing cytokine levels between murine models of sepsis is valuable toward determining their utility to act as a surrogate for human disease, the fact that cytokine levels discussed were obtained in FVB/N mice for P. aeruginosa pneumonia and BALB/c mice for CLP limits the interpretations that can be drawn from these comparisons. Although trends in cytokine levels between strains may be useful, comparing actual concentrations may be misleading, as is demonstrated by differing IL-6 and TNF-α levels in FVB/N mice and BALB/c mice given identical concentrations of intraperitoneal LPS (compare results in this manuscript to Reference 16). Thus, although the results contained herein clearly demonstrate that P. aeruginosa pneumonia and endotoxemia have different cytokine profiles in FVB/N mice and thus represent different models, analysis comparing P. aeruginosa pneumonia to CLP in other strains beyond showing the models have similar trends in the appearance and disappearance of cytokines must be interpreted with caution.

In summary, both gentamicin and imipenem markedly improve survival in a P. aeruginosa pneumonia-induced sepsis. The cytokine profile of these animals is consistent with that seen in other widely used animal models of sepsis and is clearly distinct from endotoxemia in this strain of mice, and P. aeruginosa pneumonia represents a valid murine model of pulmonary sepsis. Because antibiotics appear to have a greater impact on systemic levels of proinflammatory cytokines than anti-inflammatory cytokines, further studies are needed to identify the mechanisms underlying the immunopathologic response to this form of sepsis.


The authors thank Katherine C. Chang and Kevin W. Tinsley for technical assistance.


1. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR: Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 29:1303–1310, 2001.
2. Murphy SL: Deaths: final data for 1998. NatlVital Stat Rep 48:1–105, 2000.
3. Sawa T, Corry DB, Gropper MA, Ohara M, Kurahashi K, Wiener-Kronish JP: IL-10 improves lung injury and survival in Pseudomonas aeruginosa pneumonia. J Immunol 159:2858–2866, 1997.
4. Bodey GP, Bolivar R, Fainstein V: Jadeja: infections caused by Pseudomonas aeruginosa. Rev Infect Dis 5:279–313, 1983.
5. Rouby JJ: Nosocomial infection in the critically ill: the lung as a target organ. Anesthesiology 84:757–759, 1996.
6. Coopersmith CM, Stromberg PE, Dunne WM, Davis CG, Amiot DM, Buchman TG, Karl IE, Hotchkiss RS: Inhibition of intestinal epithelial apoptosis and survival in a murine model of pneumonia-induced sepsis. J Am Med Assoc 287:1716–1721, 2002.
7. Hotchkiss RS, Dunne WM, Swanson PE, Davis CG, Tinsley KW, Chang KC, Buchman TG, Karl IE: Role of apoptosis in Pseudomonas aeruginosa pneumonia. Science 294:1783, 2001.
8. Grassme H, Kirschnek S, Riethmueller J, Riehle A, von Kurthy G, Lang F, Weller M, Gulbins E: CD95/CD95 ligand interactions on epithelial cells in host defense to Pseudomonas aeruginosa. Science 290:527–530, 2000.
9. Starke JR, Edwards MS, Langston C, Baker CJ: A mouse model of chronic pulmonary infection with Pseudomonas aeruginosa and Pseudomonas cepacia. Pediatr Res 22:698–702, 1987.
10. Yu P, Martin CM: Increased gut permeability and bacterial translocation in Pseudomonas pneumonia-induced sepsis. Crit Care Med 28:2573–2577, 2000.
11. Schultz MJ, Rijneveld AW, Florquin S, Edwards CK, Dinarello CA, van der Poll T: Role of interleukin-1 in the pulmonary immune response during Pseudomonas aeruginosa pneumonia. Am J Physiol Lung Cell Mol Physiol 282:L285–L290, 2002.
12. Skerrett SJ, Martin TR, Chi EY, Peschon JJ, Mohler KM, Wilson CB: Role of the type 1 TNF receptor in lung inflammation after inhalation of endotoxin or Pseudomonas aeruginosa. Am J Physiol 276:L715–L727, 1999.
13. Kaneko Y, Yanagihara K, Kuroki M, Ohi H, Kakeya H, Miyazaki Y, Higashiyama Y, Hirakata Y, Tomono K, Kadota JI, Kohno S: Effects of parenterally administered ciprofloxacin in a murine model of pulmonary Pseudomonas aeruginosa infection mimicking ventilator-associated pneumonia. Chemotherapy 47:421–429, 2001.
14. Newcomb D, Bolgos G, Green L, Remick DG: Antibiotic treatment influences outcome in murine sepsis: mediators of increased morbidity. Shock 10:110–117, 1998.
15. Deitch EA: Animal models of sepsis and shock: a review and lessons learned. Shock 9:1–11, 1998.
16. Remick DG, Newcomb DE, Bolgos GL, Call DR: Comparison of the mortality and inflammatory response of two models of sepsis: lipopolysaccharide vs. cecal ligation and puncture. Shock 13:110–116, 2000.
17. Bernard GR, Vincent JL, Laterre PF, LaRosa SP, Dhainaut JF, Lopez-Rodriguez A, Steingrub JS, Garber GE, Helterbrand JD, Ely EW, Fisher CJ: Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 344:699–709, 2001.
18. Abraham E, Marshall JC: Sepsis- and mediator-directed therapy: rethinking the target populations. Mediator-directed therapy in sepsis: rethinking the target populations. Mol Med Today 5:56–58, 1999.
19. Munford RS, Hall CL, Grimm L: Detection of free endotoxin in cerebrospinal fluid by the limulus lysate test. Infect Immun 45:531–533, 1984.
20. Shenep JL, Mogan KA: Kinetics of endotoxin release during antibiotic therapy for experimental gram-negative bacterial sepsis. J Infect Dis 150:380–388, 1984.
21. Hack CE, De Groot ER, Felt-Bersma RJ, Nuijens JH, Strack Van Schijndel RJ, Eerenberg-Belmer AJ, Thijs LG, Aarden LA: Increased plasma levels of interleukin-6 in sepsis. Blood 74:1704–1710, 1989.
22. Waage A, Brandtzaeg P, Halstensen, Kierulf P, Espevik T: The complex pattern of cytokines in serum from patients with meningococcal septic shock. Association between interleukin 6, interleukin 1, and fatal outcome. J Exp Med 169:333–338, 1989.
23. Remick DG, Bolgos GR, Siddiqui J, Shin J, Nemzek JA: Six at six: Interleukin-6 measured 6 h after the initiation of sepsis predicts mortality over 3 days. Shock 17:463–467, 2002.
24. Negussie Y, Remick DG, DeForge LE, Kunkel SL, Eynon A, Griffin GE: Detection of plasma tumor necrosis factor, and interleukins 6 and 8 during the Jarisch-Herxheimer reaction of relapsing fever. J Exp Med 175:1207–1212, 1992.
25. Tateda K, Matsumoto T, Miyazaki S, Yamaguchi K: Efficacy of β-lactam antibiotics combined with gentamicin against penicillin-resistant pneumococcal pneumonia in CBA/J mice. J Antimicrob Chemother 43:367–371, 1999.
26. Bohnen JM, Solomkin JS, Dellinger EP, Bjornson HS, Page CP: Guidelines for clinical care: Anti-infective agents for intra-abdominal infection. A Surgical Infection Society policy statement. Arch Surg 127:83–89, 1992.
27. Sawa T, Kurahashi K, Ohara M, Gropper MA, Doshi V, Larrick JW, Wiener-Kronish JP: Evaluation of antimicrobial and lipopolysaccharide-neutralizing effects of a synthetic CAP18 fragment against Pseudomonas aeruginosa in a mouse model. Antimicrob Agents Chemother 42:3269–3275, 1998.
28. Pennington JE, Stone RM: Comparison of antibiotic regimens for treatment of experimental pneumonia due to Pseudomonas. J Infect Dis 140:881–889, 1979.
29. Schiff JB, Small GJ, Pennington JE: Comparative activities of ciprofloxacin, ticarcillin, and tobramycin against experimental Pseudomonas aeruginosa pneumonia. Antimicrob Agents Chemother 26:1–4, 1984.
30. Gordin FM, Rusnak MG, Sande MA: Evaluation of combination chemotherapy in a lightly anesthetized animal model of Pseudomonas pneumonia. Antimicrob Agents Chemother 31:398–403, 1987.
31. Ebong S, Call D, Nemzek J, Bolgos G, Newcomb D, Remick D: Immunopathologic alterations in murine models of sepsis of increasing severity. Infect Immun 67:6603–6610, 1999.
32. Baker CC, Chaudry IH, Gaines HO, Baue AE: Evaluation of factors affecting mortality rate after sepsis in a murine cecal ligation and puncture model. Surgery 94:331–335, 1983.
33. Chaudry IH, Wichterman KA, Baue AE: Effect of sepsis on tissue adenine nucleotide levels. Surgery 85:205–211, 1979.
34. Coopersmith CM, Chang KC, Swanson PE, Tinsley KW, Stromberg PE, Buchman TG, Karl IE, Hotchkiss RS: Overexpression of Bcl-2 in the intestinal epithelium improves survival in septic mice. Crit Care Med 30:195–201, 2002.
35. Goebel A, Kavanagh E, Lyons A, Saporoschetz IB, Soberg C, Lederer JA, Mannick JA, Rodrick ML: Injury induces deficient interleukin-12 production, but interleukin-12 therapy after injury restores resistance to infection. Ann Surg 231:253–261, 2000.
36. O'Sullivan ST, Lederer JA, Horgan AF, Chin DH, Mannick JA, Rodrick ML: Major injury leads to predominance of the T helper-2 lymphocyte phenotype and diminished interleukin-12 production associated with decreased resistance to infection. Ann.Surg 222:482–490, 1995.

Cytokines; interleukin 6; tumor necrosis factor; gentamicin; imipenem; interleukin 10; interleukin 12; infection

©2003The Shock Society