Effects of Local and Systemic Endotoxin in Humans

Clinical Center

Investigators

Abstract

Integrating human physiology with inflammatory mechanisms discovered in vitro is an essential step to define the relevance of these mechanisms in health and disease. While animal models often provide insight into their significance, there are multiple differences in human host immunity that require further refinement with clinical investigation. The administration of endotoxin to humans by different routes has been used over the last century as a tool to understand human inflammatory responses to microbes. When administered intravenously in small doses to humans, endotoxin elicits a series of events that are similar to the initial phases of a true infection (i.e. fever, an increase in the heart rate, respiratory rate, leukocyte count, and a decrease in blood pressure as well as the induction of a wide spectrum of host inflammatory genes and proteins). When directly instilled or inhaled into the lungs, endotoxin elicits brisk local inflammatory responses in the pulmonary tissue. These pulmonary inflammatory responses are thought to qualitatively reproduce some of the events that might occur in the context of pneumonia due to Gram-negative organisms. Prior studies pioneered at the Clinical Center, have shown that the instillation of endotoxin (E. coli O:113, 2-4 ng/kg) into a lung segment can be safely performed. It results in a localized inflammatory response with a slightly more limited systemic responses, although it does lead to leukocytosis, bandemia, slight lymphopenia, and an increase in some inflammatory markers, all of these most prominent at early time points (NIH protocols, 92-CC-0141, 05-CC-0012). It is also characterized by an increase in inflammatory mediators in the bronchoalveolar lavage (BAL) accompanied by an early influx of neutrophils followed later by a mononuclear cell influx into the lung segment. This human model provides a unique tool to study the mechanisms associated with the initiation and resolution of acute lung inflammation and will allow comparison of lung cellular subsets. Specific Aim 1: Characterize the immunophenotypes of blood and lung resident as well as recruited cells and to assess the changes of inflammatory cells (i.e., neutrophils, monocytes, macrophages, lymphocytes populations and subpopulations) during the initiation and resolution of acute human lung inflammation after segmental lung challenge with endotoxin as compared to non-challenged segments. Current status: Protocol 15-CC-0091 is currently active, however recruitment is currently on hold due to the pandemic. Subjects receive a local instillation of endotoxin in a lung segment and normal saline in a contralateral segment. At 6, 24, or 48h post challenge, the same two segments are sampled. Lavage samples and blood samples at this time are assayed by comprehensive flow cytometry, cytokine measurements, and gene expression analysis. Six control subjects have been enrolled and fully evaluated. Additionally, 6 subjects have been challenged with endotoxin and sampled 6h post-challenge. At 6h following challenge, there is significant influx of neutrophils in the challenged segment, and this is not present in the contralateral side. Circulating lymphocytes tend to have a central memory phenotype and nave CD4+ and CD8+ lymphocytes are fewer. Regulatory T cell numbers are also decreased in circulation following local endotoxin challenge. This trend is also evident in the lavage fluid. We have recruited 6 patients for the 6h timepoint, 6 patients for the 24h timepoint, and 5 patients for the 48h timepoint. The data have all been acquired but have not been analyzed yet. Specific Aim 2: Investigate the temporal evolution of the inflammatory response to systemic endotoxin by interrogating the plasma proteome. Current status: In this protocol, subjects received an intravenous injection of a control solution or of high doses of endotoxin (2 or 4ng/kg) at Duke Medical Center. Samples were collected at 0, 1, 2, 3, 4, 6, and 8h following the start of the infusion. Plasma samples from 4 subjects who received a carrier and 8 subjects who received high doses of intravenous endotoxin have been obtained by our lab and fully analyzed on the Somalogic platform at the CHI. Principal component analysis demonstrates a strong effect of the endotoxin challenge. Further analysis indicates that 426 distinct proteins are differentially measured at any timepoint after endotoxin challenge. There are distinct patterns of change of these proteins. Some proteins increase rapidly early after endotoxin challenge and either remain elevated over the following timepoints or return quickly to baseline. Some proteins show rapid depletion in plasma. Although some of these proteins have been associated with inflammatory responses following endotoxin challenge, many have not and suggest novel pathways and cellular populations potentially involved in this inflammatory response. We are currently working to identify putative cellular and tissue sources of detected mediators by using known protein databases. Specific Aim 3: To investigate the evolution of the metabolic response to systemic endotoxin administration by interrogating the plasma metabolome. Current status: 23-subjects were challenged with a single intravenous dose of either low (0.6 or 1 ng/kg, n = 8) or high (2 or 4 ng/kg, n = 8) dose E.coli O:113 LPS or normal saline-placebo (n = 7). Blood samples were obtained at 0, 1, 2, 3, 4, 6, and 8h. The samples were analyzed with ultrahigh performance liquid chromatography-tandem mass spectroscopy by Metabolon. Molecules were identified by comparison to spectral features present in a reference library. 815 metabolites of known identity were found to be differentially affected by LPS at any given time point. Maximal changes were observed at 4h in the low dose group (363 analytes), and 6h in the high dose group (501 analytes) when compared to baseline. PL infusion led to changes for 253 analytes at 6h. Compared to PL, there was an increase in lactate and pyruvate production in response to high dose LPS, with changes noted as early as 2h, indicative of increased aerobic glycolysis. Similarly, metabolites involved in oxidative phosphorylation (aconitate, succinylcarnitine, fumarate, and malate) were also elevated in response to high dose LPS, despite preferential lactate production via aerobic glycolysis, suggesting carbon input from other sources, likely lipid oxidation. Interestingly, metabolites associated with aerobic glycolysis were noted to increase at earlier time points whereas TCA cycle metabolites rose later, suggesting a shift from aerobic glycolysis typical of innate immunity to oxidative phosphorylation, the preferred energy source of adaptive immune cells. Exposure to both low dose and high dose LPS led to significant decreases in amino acids (serine, asparagine, glutamine, glycine) (0.48-0.86 fold) compared to placebo. Long chain fatty acids were found to accumulate in response to high dose LPS, suggesting changes in lipid oxidation, and trended towards decreasing in response to low dose endotoxin. These trends were most apparent at 6h. Finally, steroid hormones (cortisol, pregnenolone and progestin steroids), were more abundant following low dose and high dose challenge (1.62-20.62 fold). In addition to these samples, samples from patients with infectious pneumonias were also assessed on this platform. Samples were analyzed and raw data sent back to us but analysis is pending at this time. Several manuscript are currently in preparation.