Role of interleukin-6 in mortality from and physiologic response to sepsis

Abstract

Previous studies have suggested that interleukin-6 (IL-6) serves as both a marker and a mediator for the severity of sepsis. We tested whether interleukin 6 knockout (IL-6KO) mice were more susceptible to sepsis mortality induced by cecal ligation and puncture. IL-6KO and wild-type (WT) mice were subjected to increasing degrees of sepsis severity. Physiologic support was given with fluids and appropriate antibiotics. Plasma IL-6 levels were determined 6 h after the onset of sepsis, and a complete hematologic profile was performed on day 2. As expected, increasing sepsis severity resulted in greater and more rapid mortality. However, the mortality was nearly identical in the IL-6KO and WT mice. All WT septic mice had high plasma levels of IL-6 6 h after the onset of sepsis, while IL-6KO were near or below the lower limit of detection. Among the WT mice, mortality was significantly higher in mice with plasma IL-6 >3,000 pg/ml. Both IL-6KO and WT mice destined to die in the early stages of sepsis had substantial and nearly identical weight gain in the first 24 h. However, at later stages the WT mice had significantly greater weight loss than the KO mice. The KO mice failed to develop the characteristic hypothermia within the first 24 h of severe sepsis routinely observed in the WT mice. These data demonstrate that IL-6 serves as a marker of disease severity in sepsis and does modulate some physiologic responses, but complete lack of IL-6 does not does not alter mortality due to sepsis.

FIG. 1.

Mortality following cecal ligation and puncture. Both wild-type and IL-6KO mice were subjected to CLP of increasing severity. There was no difference in mortality between the IL-6KO and wild-type mice when compared with log rank survival analysis in any of the models shown.

FIG. 2.

Plasma IL-6 levels of wild-type and knockout mice and correlation with mortality. (A) Plasma levels of IL-6 were determined 6 h after the onset of sepsis induced with a 21-gauge needle. All the wild-type mice had significant levels of IL-6 within the plasma, while in the knockout mice levels were nearly undetectable. Each value is the mean ± the standard error of the mean for nine mice, * = P <0.05. (B) Mice were divided into those with IL-6 levels greater than 3,000 pg/ml (n = 18) or less than 3,000 pg/ml (n = 33). Mice with higher levels of plasma IL-6 had significantly greater mortality, particularly in early phases of sepsis. P = 0.014 comparing the curves by log rank survival analysis.

FIG. 3.

Change in body weight 24 h prior to death. Mice were weighed daily, and the data are displayed as the change in body weight on the day prior to death. Since statistical analysis showed that a group effect (wild type versus knockout) was not present, the data were combined. There is a significant increase in the body weight of the mice that would die within the next 24 h compared to those mice that would live in the first 2 days. This difference became less apparent as the sepsis evolved. The value beneath each bar indicates the number of animals in that group. * = P <0.05 comparing mice that lived versus those that died.

FIG. 4.

Daily body weight following nonlethal cecal ligation and puncture. Mice were subjected to nonlethal 25-gauge single-puncture CLP. The IL-6KO mice had significantly less loss of body weight compared to the wild-type mice. There was a significant effect based on the day, as well as differences between the groups. Each value is the mean ± the standard error of the mean for 18 to 31 mice. * = P <0.05 (wild type compared to KO with Bonferroni correction).

FIG. 5.

Daily food consumption. The amount of food consumed each day was determined in 25-gauge single-puncture mice to match the data in Fig. 4. There is virtually no difference in the consumption of food between the wild-type and knockout mice, with the single exception of the day 3, when the knockout mice consumed more food. Each value is the mean ± the standard error of the mean for 18 to 31 mice.

FIG. 6.

Gross motor activity. Gross motor activity was determined using implanted radio transmitters. There is a rapid return to the normal diurnal variation in both the sham-operated and the 25-gauge single-puncture septic animals. In contrast, 25-gauge double-puncture animals essentially never regain normal diurnal variation. The light-dark bars indicate when the lights were on or off in the room. Each line is the mean of two to four mice.

FIG. 7.

Body temperature change. Mice were subjected to 25-gauge double-puncture CLP, and the body temperature was determined with implanted radio transmitters. Panel A is the temperature trace for each individual mouse, and it is apparent that the knockout mice failed to develop hypothermia during the first 24 h. Panel B represents the mean ± the standard error of the mean for six wild-type and six IL-6KO mice. WT mice had significantly lower body temperatures at each of the indicated time points. * = P <0.05 (WT versus IL-6KO).