A-115 Direct single cell-type transcript abundance (DIRECT LS-TA) assay of Monocyte Interferon (IFN) gene expression as biomarker of viral infection: better performance than IFN score

Abstract

Blood transcriptome has been explored to differentiate bacterial and viral infections. However, gene expression in blood represents a composite score of gene expression of all the component cell-types present in the sample. In most proposed signatures, numerous genes need to be measured. Here, we propose to develop a rapid ratio-based biomarker (RBB) that reflect gene expressions of a specified single cell type in peripheral blood (e.g., monocytes or granulocytes) and it is called Direct Leukocyte Single cell-type transcript abundance (TA) assay (DIRECT LS-TA) as it works directly in a peripheral whole blood (WB) sample without the need of cell sorting. Based on our algorithm, IFI27, IFI44L and SIGLEC1 are target genes and serve as numerator genes of DIRECT LS-TA of monocytes. PSAP is the denominator gene. For example, the ratio of IFI44L / PSAP in WB correlated with IFI44L / PSAP in purified monocytes (R2=>0.9). So ratios of these 3 genes in WB reveal the activation of these 3 interferon (IFN) stimulated genes in monocyte and may be useful as biomarker to differentiate viral from bacterial infections. Two public transcriptome datasets of WB samples from viral infection quantified by microarray (GSE111368) and RNA-sequencing (GSE171110) were analysed. After log transformation, DIRECT LS-TA values of the 3 genes were obtained by subtraction of log PSAP. In addition, the conventional IFN score was obtained by the average of IFIT1, IFITM3, SIGLEC1, IFI44L divided by average of UBC, HPRT and YWHAZ. ROC curve analysis was performed to compare DIRECT LS-TA monocyte : IFI27 / PSAP vs IFN score. In both datasets, DIRECT LS-TA IFI27 and IFN score were significantly different between control and viral infection (p<1e-9). Direct IFI44L and SIGLEC1 were also significantly increased in viral infection (p<1e-5). IFN score was also significantly increased. To compare the diagnostic performance for these markers, the average sensitivity and specificity were evaluated. sensitivity specificity AUC IFN score 0.81 0.92 0.89 DIRECT LS-TA IFI27 0.87 0.92 0.93 DIRECT LS-TA IFI44L 0.7 0.87 0.78 DIRECT LS-TA SIGLEC1 0.58 0.87 0.67 Results show that DIRECT LS-TA for IFI27 (equal to the ratio of IFI27 to PSAP) in WB was the best indicator of viral infection. This 2-gene ratio based biomarker performed even better than the conventional IFN score which required quantification 8 genes. This finding is consistent with our previous study of IFN score in COVID patients. There were a small percentage of patients with "normal" IFN score but only IFI27 remained elevated to a small extent while other IFN-stimulated genes returned to normal level (DOI: 10.3389/fimmu.2023.1315602). The results suggested that DIRECT LS-TA of IFI27 is a biomarker which is easy to determine and better in performance to indicate an ongoing viral infection. Disclosure: Patent pending. Monocytes Direct LS-TA Patent: https://patents.google.com/patent/GB2629711A/en Granulocytes Direct LS-TA Patent: https://patents.google.com/patent/GB202400180D0