Gene expression profiling of whole blood in patients with severe leptospirosis

Background: Leptospirosis is a global zoonotic infectious disease with various clinical manifestations ranging from mild self-limiting illness to life-threatening infection with multi-organ damage. This study was aimed at investigating transcription profiles from whole blood samples of patients with leptospirosis and identifying genes as novel biomarkers for predicting severe leptospirosis. Methods: In a discovery cohort, 12 serum samples of patients with severe and non-severe leptospirosis at initial clinical presentation were selected for gene expression profiling using the NanoString nCounter PanCancer IO 360 gene expression panel. In a validated cohort of 99 samples, top candidate genes were selected and further tested by qRT-PCR in whole blood samples of 30 and 69 individuals with severe and non-severe leptospirosis, respectively. Results: The discovery set identified 20 differentially expressed genes (DEGs) among the two groups. The top three down-regulated candidate genes including programmed cell death 1 (PDCD1), interleukin 4 (IL4), and nitric oxide synthase 2 (NOS2) were selected and further validated. In the validated cohort, PDCD1 levels were significantly lower in the severe group compared with the non-severe group(Stats?). Based on the ROC analysis, PDCD1 levels could discriminate between the severe and non-severe groups. Additionally , PDCD1 levels displayed good predictive power of subsequent pulmonary hemorrhage with an AUROC of 0.86 (95% CI;0.76-0.96, p = 0.007). PDCD1 also emerged as an independent prognostic factor of severe leptospirosis in the multivariate regression analysis.. Conclusion: Our data indicated that whole blood gene expression profiles were significantly different between the severe and non-severe leptospirosis groups. PDCD1 expression levels at presentation could potentially serve as a biomarker for predicting severe leptospirosis. A total of 12 serum samples (8 and 4 samples from the severe and non-severe groups, respectively) were randomly selected from the stored samples to investigate the gene expression profile using the nCounter PanCancer IO 360 panel (NanoString Technologies, Seattle, WA). The samples were tested according to the manufacturer’s protocol.