Biomarkers Of Acute Graft-Versus-Host Disease: Surface Antigens And Micro Rnas In Extracellular Vesicles

IntroductionBiomarkers could be crucial to identify patients at high-risk of acute Graft-vs.-Host Disease (aGVHD). Given their involvement in inflammation, Extracellular Vesicles (EVs) may become attractive biomarkers. Moreover, EVs are non-invasively extracted from body fluids. In a preliminary study, we significantly correlated CD146, CD31 and CD140a expression on EVs membranes with the onset of aGVHD (Lia G. Leukemia 2017).ObjectivesWe designed a prospective study to further characterize EVs by their surface antigens and by their content in MicroRNAs.MethodsEVs are extracted from serum samples at given time-points (pre-transplant, on day 0, 3, 7, 14, 21, 28, 35, 45 and then monthly up to 1 year) by a protamine-based precipitation method and analyzed by flow-cytometry (Guava EasyCyte Flow Cytometer) for the expression of 13 membrane proteins (CD44, CD138, CD146, KRT18, CD120a, CD8, CD30, CD106, CD25, CD31, CD144, CD86, and CD140a). MicroRNAs (miR100, miR92b, miR155, miR194) are extracted from EVs at pre- transplant and on day 0, 7, 14, 28, and quantified by real time PCR as relative quantification compared to healthy donors after cDNA Reverse Transcription. Logistic Regression Analysis is performed for each marker.ResultsThirty-five transplant patients with hematological diseases have so far been enrolled. Seventeen/35 patients (49%) developed grade II-IV aGVHD. Our preliminary findings show that CD146 (melanoma cell adhesion molecule, MCAM-1) and CD44 (homing-associated cell adhesion molecule, H-CAM) were associated with an increased risk of aGVHD (Odds Ratio (OR) 4.3, p=0.008; OR 2.1, p=0.039), whereas CD31 (platelet endothelial cell adhesion molecule, PECAM-1) level was associated with a decreased risk of aGVHD (OR 0.31, p=0.001). Moreover, increased risk of aGVHD was significantly correlated with levels of miR100 (OR 4.66, p=0.004), miR194 (OR 2.2, p=0.01) and miR155 (OR 3.56, p=0.035). Of note, biomarkers associated with aGVHD showed a constant consensual change in signal levels before aGVHD onset (Figure 1).ConclusionsAn association of 3 EVs membrane antigens and onset of aGVHD was observed. Of note, CD146, CD44 and CD31 belong to the Cell Adhesion Molecule Family and are critical for endothelium and immune cells interactions. The functional role of miR-194 in GVHD pathogenesis remains to be determined while JAK/STAT and TGFβ pathways were shown to be involved in other studies (Gimondi S Exp Hematol, 2016). MiR-100 has been reported to regulate inflammatory neovascularization during GvHD (Leonhardt F, Blood 2013) while miR-155 drives donor T cell expansion and tissue infiltration (Zitzer N, J Immunol 2018, Ranganathan P Blood 2012). Biomarkers could be crucial to identify patients at high-risk of acute Graft-vs.-Host Disease (aGVHD). Given their involvement in inflammation, Extracellular Vesicles (EVs) may become attractive biomarkers. Moreover, EVs are non-invasively extracted from body fluids. In a preliminary study, we significantly correlated CD146, CD31 and CD140a expression on EVs membranes with the onset of aGVHD (Lia G. Leukemia 2017). We designed a prospective study to further characterize EVs by their surface antigens and by their content in MicroRNAs. EVs are extracted from serum samples at given time-points (pre-transplant, on day 0, 3, 7, 14, 21, 28, 35, 45 and then monthly up to 1 year) by a protamine-based precipitation method and analyzed by flow-cytometry (Guava EasyCyte Flow Cytometer) for the expression of 13 membrane proteins (CD44, CD138, CD146, KRT18, CD120a, CD8, CD30, CD106, CD25, CD31, CD144, CD86, and CD140a). MicroRNAs (miR100, miR92b, miR155, miR194) are extracted from EVs at pre- transplant and on day 0, 7, 14, 28, and quantified by real time PCR as relative quantification compared to healthy donors after cDNA Reverse Transcription. Logistic Regression Analysis is performed for each marker. Thirty-five transplant patients with hematological diseases have so far been enrolled. Seventeen/35 patients (49%) developed grade II-IV aGVHD. Our preliminary findings show that CD146 (melanoma cell adhesion molecule, MCAM-1) and CD44 (homing-associated cell adhesion molecule, H-CAM) were associated with an increased risk of aGVHD (Odds Ratio (OR) 4.3, p=0.008; OR 2.1, p=0.039), whereas CD31 (platelet endothelial cell adhesion molecule, PECAM-1) level was associated with a decreased risk of aGVHD (OR 0.31, p=0.001). Moreover, increased risk of aGVHD was significantly correlated with levels of miR100 (OR 4.66, p=0.004), miR194 (OR 2.2, p=0.01) and miR155 (OR 3.56, p=0.035). Of note, biomarkers associated with aGVHD showed a constant consensual change in signal levels before aGVHD onset (Figure 1). An association of 3 EVs membrane antigens and onset of aGVHD was observed. Of note, CD146, CD44 and CD31 belong to the Cell Adhesion Molecule Family and are critical for endothelium and immune cells interactions. The functional role of miR-194 in GVHD pathogenesis remains to be determined while JAK/STAT and TGFβ pathways were shown to be involved in other studies (Gimondi S Exp Hematol, 2016). MiR-100 has been reported to regulate inflammatory neovascularization during GvHD (Leonhardt F, Blood 2013) while miR-155 drives donor T cell expansion and tissue infiltration (Zitzer N, J Immunol 2018, Ranganathan P Blood 2012).