Objectives: Our understanding of septic acute kidney injury is limited. We therefore assessed renal histopathological changes induced by septic acute kidney injury and their evolution during recovery.
Design: Prospective experimental study.
Setting: Physiology Research Institute.
Subjects: Twenty-two Merino sheep.
Intervention: We induced septic acute kidney injury by continuous i.v. infusion of Escherichia coli. We studied histology, immunohistochemistry, markers of apoptosis, and expression of nitric oxide synthase isoforms and hypoxia-inducible factor-1α. Analysis was performed on kidneys from normal sheep, sheep with septic acute kidney injury, and sheep after recovery from septic acute kidney injury.
Measurements and main results: In normal, septic, and recovery sheep, respectively, serum creatinine was (median) 82 (interquartile range, 70-85), 289 (171-477), and 70 (51-91) μmol/L and renal blood flow was 270 ± 42, 653 ± 210, and 250 ± 49 mL/min. There were no histological differences between baseline, acute kidney injury, and recovery sheep. There was no evidence of macrophage or myofibroblast infiltration, no evidence of caspase-3 cleavage to suggest activation of apoptotic pathways, and no increase in neutrophil gelatinase-associated lipocalin to suggest tubular injury. Similarly, quantification of apoptosis revealed no differences between the normal and septic groups (normal: median, 3; interquartile range, 0-5 cells per visual field and septic acute kidney injury: median, 3.5; interquartile range, 0-8 cells per visual field; p = 0.618), but in the recovery group, there was increased apoptosis (median, 14; interquartile range, 4-34 cells per visual field; p = 0.002). Expression of all nitric oxide synthase subtypes increased significantly in the renal cortex during septic acute kidney injury but tended to decrease in the medulla. Medullary hypoxia-inducible factor gene expression decreased from 1.00 (95% CI, 0.74-1.36) to 0.26 (95% CI, 0.09-0.76) in recovery (p = 0.0106). Both inducible nitric oxide synthase and neuronal nitric oxide synthase expressions correlated with renal blood flow.
Conclusion: The lack of any tubular injury or increased apoptosis, the increased expression of all cortical nitric oxide synthase isoforms, and the link between inducible nitric oxide synthase and neuronal nitric oxide synthase with renal blood flow suggest in this experimental model that severe sepsis acute kidney injury can develop in the absence of histological or immunohistological changes and may be functional in nature.