Early label-free analysis of mitochondrial redox states by Raman spectroscopy predicts septic outcomes

Background: Sepsis remains an unacceptably high mortality due to the lack of biomarkers for predicting septic outcomes in the early period. Mitochondrial redox states play a pivotal role in this condition and are disturbed early in the development of sepsis. Here, we hypothesized that visualizing mitochondrial redox states via resonance Raman spectroscopy (RRS) could identify septic outcomes at an early time point. Sepsis was induced by cecal ligation and puncture (CLP). We applied RRS analysis at baseline and 30 min, 1 h, 2 h, 4 h, and 6 h after CLP, and the mitochondrial redox states were identified. The levels of blood lactate as a predictor in sepsis were assessed. Our study is the first to reveal the possibility of in vivo detection of the mitochondrial redox state by using RRS in septic mice. The peak area for the Raman reduced mitochondrial fraction, the indicator of mitochondrial redox states, fluctuated significantly at 2 h after CLP. This fluctuation occurred earlier than the change in lactate level. Moreover, this fluctuation had higher prognostic accuracy for mortality than the lactate level during sepsis and could be a novel diagnostic marker for predicting septic outcomes according to the cutoff value of 1.059, which had a sensitivity of 80% and a specificity of 90%.

Objectives: To explore an effective indicator concerning mitochondrial redox states in the early stage of sepsis and to predict septic outcomes accurately in vivo using non-invasive and label-free Resonance Raman spectroscopy (RRS) analysis.

Methods: Mitochondria, primary skeletal muscle cells andex-vivo muscles harvested from gastrocnemius were detected mitochondrial redox states respectively by using RRS. Sepsis was induced by cecal ligation and puncture (CLP). We applied RRS analysis at baseline and 30 min, 1 h, 2 h, 4 h, and 6 h after CLP, and the mitochondrial redox states were identified. The levels of blood lactate as a predictor in sepsis were assessed. The predictive correlation of mitochondrial redox states on mortality, inflammation and organ dysfunction was further assessed.

Results: Mitochondrial redox states were clearly recognized in ex-vivo gastrocnemius muscles as well as purified mitochondria and primary skeletal muscle cells by using RRS. The peak area for the Raman reduced mitochondrial fraction, the indicator of mitochondrial redox states, fluctuated significantly at 2 h after CLP. This fluctuation occurred earlier than the change in lactate level. Moreover, this fluctuation had higher prognostic accuracy for mortality than the lactate level during sepsis and could be a novel diagnostic marker for predicting septic outcomes according to the cutoff value of 1.059, which had a sensitivity of 80% and a specificity of 90%.

Conclusions: This study demonstrated that monitoring mitochondrial redox states using RRS as early as 2 h could indicate outcomes in septic mice. These data may contribute to developing a non-invasive clinical device concerning mitochondrial redox states by using bedside-RRS.