S100A8/A9 mRNA induction in an ex vivo model of endotoxin tolerance: roles of IL-10 and IFNγ

Abstract

Objectives: Septic syndromes are the leading cause of death in intensive care units. They are characterized by the development of immune dysfunctions such as endotoxin tolerance (ET), whose intensity and duration are associated with increased risk of nosocomial infections and mortality. Alarmins S100A8 and S100A9 have been shown to be increased after septic shock. Importantly, a delayed S100A9 mRNA increase predicts hospital-acquired infection in patients. The aim of this study was to investigate the regulation of S100A8 and S100A9 mRNA expression in an ex vivo model of ET.

Subjects and measurements: ET was reproduced ex vivo by priming healthy peripheral blood mononuclear cells (number of donors = 9 to 10) with low-dose endotoxin (2 ng/ml) before stimulation with high dose endotoxin (100 ng/ml). S100A8 and S100A9 mRNA levels were measured by quantitative real-time polymerase chain reactions.

Main results: ET was established by observing decreased TNFα and increased IL-10 transcriptomic responses to two subsequent endotoxin challenges. Interestingly, ET was associated with increased S100A8 and S100A9 mRNA expression ex vivo. We showed that IL-10 played a role in this process, since S100A8 and S100A9 mRNA increases were significantly abrogated by IL-10 blockade in the model. Conversely, treatment with rIFN-γ, a pro-inflammatory and immunostimulating molecule known to block ET induction, was able to restore normal S100A8 and S100A9 mRNA in this model.

Conclusions: In this ex vivo model, we observed that S100A8 and S100A9 mRNA expression was significantly increased during ET. This reproduced ex vivo the observations we had previously made in septic shock patients. Interestingly, IL-10 blockade and rIFN-γ treatment partially abrogated S100A8/A9 mRNA increases in this model. Pending confirmation in larger, independent clinical studies, these preliminary results suggest that S100A8 and S100A9 mRNA levels might be used as surrogate markers of ET and as stratification tools for personalized immunotherapy in septic shock patients.

Figure 1. Double hit model with LPS.

Figure 1A. Diagram of the model of endotoxin tolerance used in our study. Peripheral blood mononuclear cells (PBMCs) were obtained from fresh whole blood from healthy volunteers (HV) by Ficoll-Paque density gradient centrifugation. Endotoxin tolerance was reproduced ex vivo by priming PBMCs with low-dose endotoxin (2 ng/ml) before stimulation with high dose endotoxin (100 ng/ml). S100A8 and S100A9 mRNA levels were measured by quantitative real-time polymerase chain reactions. Figure 1B. Diagram of immune-stimulating therapy model during endotoxin tolerance.

Figure 2. Endotoxin tolerance is associated with decreased TNFα and increased IL-10, S100A8 and S100A9 mRNA expressions.

Figures 2A, B, C and D. Messenger RNA (mRNA) level of tolerizable gene (TNFα) and non-tolerizable genes (IL-10, S100A8 and S100A9) in an ex vivo model of endotoxin tolerance. The mRNA level was normalized to that of the reference gene peptidylpropylisomerase B (PPIB) and then compared to the control group. Black columns represent controls (cells without any lipopolysaccharide (LPS)), white columns represent LPS-unprimed cells (only stimulated once with 100 ng/ml LPS), light grey columns represent LPS-primed cells (stimulated twice: 2 ng/ml followed by 100 ng/ml) and dark grey columns represent cells stimulated once with 2 ng/ml LPS without any second stimulation. **<0.01, *<0.05, Wilcoxon signed-rank test. Median (+/− interquartile range) data from 10 independent experiments are given. Figure 2E. Correlation between S100A8 mRNA expression (x-axis) and S100A9 mRNA expression (y-axis) in the endotoxin tolerance model. S100A8 and S100A9 mRNA levels were normalized to that of PPIB. Data were obtained from the 10 independent experiments described above.

Figure 3. Endotoxin tolerance is associated with increased SOCS3 mRNA expression.

Messenger RNA (mRNA) level of SOCS3 in an ex vivo model of endotoxin tolerance. The mRNA level was normalized to that of the reference gene peptidylpropylisomerase B (PPIB) and then compared to the control group. Black columns represent controls (cells without any lipopolysaccharide (LPS)), white columns represent LPS-unprimed cells (only stimulated once with 100 ng/ml LPS) and grey columns represent LPS-primed cells (stimulated twice: 2 ng/ml followed by 100 ng/ml). **<0.01, *<0.05, Wilcoxon signed rank test. Median (+/− interquartile range) data from 9 independent experiments are given.

Figure 4. S100A8 and S100A9 mRNA expression increases were partially abrogated by IL-10 blockade.

Messenger RNA (mRNA) level of TNFα, IL-10, S100A8 and S100A9 in an ex vivo model of endotoxin tolerance in the presence or absence of anti-IL-10 antibodies (Ab). Anti IL-10 Ab was used at 100 ng/ml. The mRNA level was normalized to that of the reference gene peptidylpropylisomerase B (PPIB) and then compared to the control group. Black columns represent controls (cells without any lipopolysaccharide (LPS)), white columns represent LPS-unprimed cells (only stimulated once with 100 ng/ml LPS) and grey columns represent LPS-primed cells (stimulated twice: 2 ng/ml followed by 100 ng/ml). †p<0.01, Wilcoxon paired test. Median (+/− interquartile range) data from 9 independent experiments are given.

Figure 5. S100A8 and S100A9 mRNA level increases were significantly abrogated by rIFN-γ in endotoxin tolerance model.

Messenger RNA (mRNA) level of TNFα, IL-10, S100A8 and S100A9 in an ex vivo model of endotoxin tolerance. The mRNA level was normalized to that of the reference gene peptidylpropylisomerase B and then compared to the control group. Black columns represent controls (cells without any lipopolysaccharide (LPS)), white columns represent LPS-unprimed cells (only stimulated once with 100 ng/ml LPS) and grey columns represent LPS-primed cells (stimulated three times: 2 ng/ml LPS followed by vehicle or IFN-γ followed by 100 ng/ml LPS). †p<0.01, ‡p<0.05, Wilcoxon paired test. Median (+/− interquartile range) data from 10 independent experiments are given.