MicroRNA fingerprints identify miR-150 as a plasma prognostic marker in patients with sepsis

Abstract

Background: The physiopathology of sepsis continues to be poorly understood, and despite recent advances in its management, sepsis is still a life-threatening condition with a poor outcome. If new diagnostic markers related to sepsis pathogenesis will be identified, new specific therapies might be developed and mortality reduced. Small regulatory non-coding RNAs, microRNAs (miRNAs), were recently linked to various diseases; the aim of our prospective study was to identify miRNAs that can differentiate patients with early-stage sepsis from healthy controls and to determine if miRNA levels correlate with the severity assessed by the Sequential Organ Failure Assessment (SOFA) score.

Methodology/principal findings: By using genome-wide miRNA profiling by microarray in peripheral blood leukocytes, we found that miR-150, miR-182, miR-342-5p, and miR-486 expression profiles differentiated sepsis patients from healthy controls. We also proved by quantitative reverse transcription-polymerase chain reaction that miR-150 levels were significantly reduced in plasma samples of sepsis patients and correlated with the level of disease severity measured by the SOFA score, but were independent of the white blood counts (WBC). We found that plasma levels of tumor necrosis factor alpha, interleukin-10, and interleukin-18, all genes with sequence complementarity to miR-150, were negatively correlated with the plasma levels of this miRNA. Furthermore, we identified that the plasma levels ratio for miR-150/interleukin-18 can be used for assessing the severity of the sepsis.

Conclusions/significance: We propose that miR-150 levels in both leukocytes and plasma correlate with the aggressiveness of sepsis and can be used as a marker of early sepsis. Furthermore, we envision miR-150 restoration as a future therapeutic option in sepsis patients.

Figure 1. Leukocyte miRNA Signature Differentiates Early Sepsis Patients from Healthy Controls.

(A) The cluster shows a perfect separation between the two classes of samples (sepsis vs. controls). The miRNA expression (log₂) cluster shows differentially expressed genes as determined by t-test analysis. Yellow indicates high expression, and blue indicates low expression, relative to the median. (B) Fold change by qRT-PCR and array is shown for the leukocyte signature miRNAs. The most conspicuous differentially expressed miRNAs (sepsis vs. control) identified by both BRB array tools and GeneSpring GX software and also by qRT-PCR in leukocyte cells are shown.

Figure 2. Plasma Level of miR-150 is Significantly Lower in Sepsis Patients as in Controls.

(A) Microarray analysis of miR-192 levels in leukocytes found no variance in expression between patients and controls; qRT-PCR confirmed the same finding in plasma. Therefore, we used this miRNA as reference and normalized miR-150 levels in respect to miR-192 and found significant downregulation at both day 1 and day 7. Mean Ct plus standard deviation have been reported. The lower panel represents the combination of data from the upper two. (B) The miR-150 and miR-192 levels of 16 sepsis patients are plotted by their mean Ct values at day 1 after ICU admission. The round grey dots represent the patients with lower expression levels (2^-ΔCt) and lower SOFA scores than the patients identified by square black dots. The grey shaded region indicates the threshold to classify miR-150, showing a clear division between samples with high and low expression levels based on miR-192 as the normalizer. (C) The miR-150 and miR-192 levels of 16 sepsis patients are plotted by their miR-150/miR-192 relative expression values at day 1 after ICU admission related to the respective SOFA score. Interpolation line was reported.

Figure 3. Negative Correlation Between Plasma Levels of miR-150 and Cytokines.

(A) The complementary sequence between miR-150 and the mRNAs of TNF-alpha, IL-10, and IL-18 is shown. Complementary sequences are reported, as well as the relative folding energy between miR-150 and mRNAs by using RNA22. (B) ELISA determination for plasma markers of sepsis and correlation with miR-150 expression in plasma are depicted in the graphs. The left panels show IL-10, IL-18 and TNF-alpha measurements by ELISA in plasma, mean Ct +/− standard deviation have been reported. The right panels show the negative correlation between miR-150 and IL-10, IL-18 and TNF-alpha, respectively. Values, on a patient by patient basis, have been reported for each cytokine studied.

Figure 4. Ratio Between IL-18 and miR-150/miR-192 in Plasma of Sepsis Patients.

A ratio between IL-18 ELISA assay expression and miR-150 expression based on miR-192 normalization was computed. Patients were separated into two groups related to the miR-150 expression before the threshold was established (8.5). Patients with miR-150 fold difference less than 8.5 provided an IL-18/miR-150 ratio statistically significantly higher than the group of patients with miR-150 expression of more than 8.5 (P<0.05 by two-side t-test). Mean Ct +/− standard deviation have been reported.