Rationale: The extent, timing, and significance of mitochondrial injury and recovery in bacterial sepsis are poorly characterized, although oxidative and nitrosative mitochondrial damage have been implicated in the development of organ failure. Objectives: To define the relationships between mitochondrial biogenesis, oxidative metabolism, and recovery from Staphylococcus aureus sepsis. Methods: We developed a murine model of fibrin clot peritonitis, using S. aureus. The model yielded dose-dependent decreases in survival and resting energy expenditure, allowing us to study recovery from sublethal sepsis. Measurements and Main Results: Peritonitis caused by 10 6 colonyforming units of S. aureus induced a low tumor necrosis factor-a state and minimal hepatic cell death, but activated prosurvival protein kinase A, B, and C sequentially over 3 days. Basal metabolism by indirect calorimetry was depressed because of selective mitochondrial oxidative stress and subsequent loss of mitochondrial DNA copy number. During recovery, mitochondrial biogenesis was strongly activated by regulated expression of the requisite nuclear respiratory factors 1 and 2 and the coactivator peroxisome proliferatoractivated receptor g coactivator-1a, as well as by repression of the biogenesis suppressor nuclear receptor interacting protein-140. Biogenesis reconstituted mitochondrial DNA copy number and transcription, and restored basal metabolism without significant hepatocellular proliferation. These events dramatically increased hepatic mitochondrial density in transgenic mice expressing mitochondrially targeted green fluorescent protein.Conclusions: This is the first demonstration that mitochondrial biogenesis restores oxidative metabolism in bacterial sepsis and is therefore an early and important prosurvival factor.

Background and objectives: Patients who have systemic lupus erythematosus (SLE) and manifest antiphospholipid antibodies (APA) are at increased risk for thrombosis; however, it is difficult to predict who will clot. This study tested the hypothesis that peak D-dimer level measured routinely during follow-up identifies whether a hypercoagulable state is developing and, therefore, the patient is at increased risk for thrombosis.Design, setting, participants, & measurements: One hundred consecutive patients who had SLE with recurrent activity (71% renal SLE) and were evaluated for or enrolled in the Ohio SLE Study were studied. D-dimer testing was done annually and usually at SLE flare or other serious illness. When D-dimer was elevated, evaluation for thrombosis (large vessel, small vessel, or Libman-Sacks) was undertaken. Mean follow-up was 37.5 ؎ 15 SD months.Results: Of those with peak D-dimer <0.5 g/ml (n ‫؍‬ 46), 0% thrombosed, 33% had APA. Of those with peak D-dimer 0.5 to 2.0 g/ml (n ‫؍‬ 19), 6% thrombosed, 44% had APA. Of those with peak D-dimer >2.0 g/ml (n ‫؍‬ 36), 42% thrombosed, 76% had APA. The most common causes of elevated D-dimer in the absence of demonstrable thrombosis were SLE flare and systemic infection. D-dimer levels were usually elevated for several months before thrombosis.Conclusions: Patients with SLE and normal D-dimer levels are at low risk for thrombosis, irrespective of APA status. Those with persistent unexplained elevated D-dimer levels, particularly when >2.0 g/ml, are at high risk for thrombosis.