Experimental studies in a rat model of fecal peritonitis conducted by Michael Bauer and colleagues show that in this model, changes in liver function occur early in the development of sepsis, with potential implications for prognosis and development of new therapeutic approaches.

Community-acquired pneumonia presents a spectrum of clinical phenotypes, from lobar pneumonia to septic shock, while mechanisms underlying progression are incompletely understood. In a transcriptomic and metabolomic study across tissues, we examined serotype-specific regulation of signaling and metabolic pathways in C57BL/6 mice intratracheally instilled with either serotype 19F Streptococcus pneumoniae (S19; causing lobar pneumonia), or serotype 2 S. pneumoniae (S2; causing septic pneumococcal disease,) or vehicle (Todd-Hewitt broth). Samples of lung, liver, and blood were collected at 6 and 24 h postinfection and subjected to microarray analysis and mass spectrometry. Results comprise a preferential induction of cholesterol biosynthesis in lobar pneumonia at low-infection doses (10(5) colony forming units/mouse) leading to increased plasma cholesterol (vehicle: 1.8±0.12 mM, S2: 2.3±0.10 mM, S19: 2.9±0.15 mM; P<0.05, comparing S19 to vehicle and S2). This induction was pneumolysin dependent, as a pneumolysin-deficient strain of serotype 19F failed to induce cholesterol biosynthesis (S19ΔPLY: 1.9±0.03 mM). Preincubation of pneumolysin with purified cholesterol or plasma from hypercholesterolemic mice prior to intratracheal instillation protected against lung barrier dysfunction and alveolar macrophage necrosis. Cholesterol may attenuate disease severity by neutralizing pneumolysin in the alveolar compartment and thus prevent septic disease progression.

Chronic sequelae of sepsis represent a major, yet underappreciated clinical problem, contributing to long-term mortality and quality-of-life impairment. In chronic liver disease, inflammation perpetuates fibrogenesis, but development of fibrosis in the post-acute phase of systemic inflammation has not been studied. Therefore, a mouse model of post-acute sequelae of sepsis was established based on polymicrobial peritonitis under antibiotic protection. Survival decreased to approximately 40% within 7 days and remained constant until day 28 (post-acute phase). In survivors, clinical recovery was observed within 1 week, whereas white blood cell and platelet count, as well as markers of liver injury, remained elevated until day 28. Macroscopically, inflammation and abscess formation were detected in the peritoneal space and on/in the liver. Microscopically, acute-chronic inflammation with ductular proliferation, focal granuloma formation in the parenchyma, and substantial hepatic fibrosis were observed. Increased numbers of potentially pathogenetic macrophages and α-smooth muscle actin-positive cells, presumably activated hepatic stellate cells, were detected in the vicinity of fibrotic areas. Fibrosis was associated with the presence of elastin and an augmented production/deposition of collagen types I and III. Microarray analyses revealed early activation of canonical and noncanonical pathways of hepatic stellate cell transdifferentiation. Thus, chronic sequelae of experimental sepsis were characterized by abscess formation, persistent inflammation, and substantial liver injury and fibrosis, the latter associated with increased numbers of macrophages/α-smooth muscle actin-positive cells and deposition of collagen types I and III. This suggests persistent activation of stellate cells, with consecutive fibrosis-a hallmark of chronic liver disease-as a result of acute life-threatening infection.

Experimental models, mimicking physiology, and molecular dynamics of diseases in human, harbor the possibility to study the effect of interventions and transfer results from bench to bedside. Recent advances in high-throughput technologies, standardized protocols, and integration of knowledge from databases yielded rising consistency and usability of results for inter-species comparisons. Here, we explored similarities and dissimilarities in gene expression from blood samples of a murine sepsis model (peritoneal contamination and infection, PCI) and patients from the pediatric intensive care unit (PICU) measured by microarrays. Applying a consistent pre-processing and analysis workflow, differentially expressed genes (DEG) from PCI and PICU data significantly overlapped. A major fraction of DEG was commonly expressed and mapped to adaptive and innate immune response related pathways, whereas the minor fraction, including the chemokine (C–C motif) ligand 4, exhibited constant inter-species disparities. Reproducibility of transcriptomic observations was validated experimentally in PCI. These data underline, that inter-species comparison can obtain commonly expressed transcriptomic features despite missing homologs and different protocols. Our findings point toward a high suitability of an animal sepsis model and further experimental efforts in order to transfer results from animal experiments to the bedside.

No abstract