Unbiased MIR screening and confirmation in the endothelium of mice with a systemic inflammatory response. Mice were either given LPS (17.5 mg/kg BW i.p.) or vehicle (0.9% NaCl) and killed after 24 h or at indicated time-points. a A smallRNA sequencing was conducted using isolated CD146 + pulmonary endothelial cells (ECs) from endotoxemic compared to healthy mice. Shown is a heat map that highlights the gene distribution with the arrow pointing to the upregulation of MIR155. b Volcano plot showing a log scaled distribution of 318 analyzed endothelial MIRs. 5 MIRs were down- and 31 upregulated; MIR155 (arrow) showed the strongest upregulation (c). Bar graphs showing normalized MIR155 expression in lung lysates via RT-PCR after 4, 12 and 24 h compared to control (CTR) (CTR: n = 8; LPS: n = 4–5). d Bar graphs showing normalized MIR155/Rnu6b expression in CD146 + pulmonary ECs (n = 5). e In vitro confirmation of time-dependent MIR155 upregulation in human umbilical endothelial cells (HUVEC) (n = 4–6) after stimulation with the proinflammatory cytokine TNFa (50 ng/mL) compared to vehicle (CTR); Bar graphs show mean ± SD of normalized MIR155 with a maximum of 15 × fold increase at 24 h. (c–e, all **p < 0.01)

Functional consequences of MIR155 overexpression and inhibition in zebrafish. a–f Transgenic zebrafish larvae (Tg(flk1:mCherry)/Tg(I-fabp:eGFP-DBP) were used to assess the role of MIR155 in vivo. Uninjected larvae (n = 26), scrambled MIR injected (25 µM, n = 16) and MIR155 injected larvae (25 µM, n = 26) were analyzed at 96 h post-fertilization. a Phenotype analysis shows the eGFP-DBP fusion protein (green) predominantly within the mcherry-flk1 positive vasculature (red) in the uninjected larva. b Equivalent eGFP-DBP expression is being detected in the scrambled MIR injection group. c MIR155 overexpression leads to loss of eGFP-DBP (green) from the mcherry-flk1 positive vasculature (#) (red) and shows accumulation in the pericardium and yolk sac (*). d To quantify protein leakage, the eGFP-DBP content was measured by detection of eGFP-DBP in the retinal vasculature in scrambled and in e the MIR155 injected fish. In the retinal vasculature of the MIR155 zebrafish, the fluorescence is diminished as eGFP-DBP is not detectable (arrow). f Box and whisker plots showing a quantification of protein leakage that was performed by measurement of the maximum fluorescence intensity of eGFP-DBP in the retinal vasculature indicating a significant loss in the MIR155 injected fish (***p < 0.0001)

Functional consequences of MIR155 overexpression and inhibition in mice. B6.Cg-Mir155tm1.1Rsky/J knockout mice and wildtype litters were challenged with LPS (17.5 mg/kg BW i.p.) or vehicle (0.9% NaCl) and killed after 16 h. a Bar graphs showing Evans blue extravasation in the lung of healthy control mice (CTR, n = 5) and LPS challenged B6.Cg-Mir155tm1.1Rsky/J knockout mice (n = 3–12) b–d Perivascular fluid cuffs (lung cuffing) were evaluated. b PAS staining indicating the bronchi (b), its corresponding arterial vasa vasorum (a) and cuffing (black arrows). c Bar graph showing the percentage of lung cuffing. Indicating a significant difference between the LPS challenged MIR155^+/+ (n = 8) and MIR155^+/− group (n = 9; p = 0.001). d Bar graph showing the lung cuffing severity score of single bronchi (n = 41–263 arteriolae per group) from a total of 27 mice. (*p < 0.05, **p < 0.01, ***p < 0.0001)

Functional effects of MIR155 on permeability and target candidates. a Control, Anti-MIR155 and MIR155^high human umbilical vein endothelial cells (HUVECs) were challenged with thrombin or vehicle following staining with a tight-junction protein (ZO-1, green), the cytoskeleton (F-actin, red) and DAPI (blue). Thrombin challenged ECs showed visible gap formation between adjacent cells (white arrows). Blockade of MIR155 had visually lesser gap formations upon thrombin stimulation (lower middle panel). Whereas the MIR155^high transfected HUVECs showed mild spontaneous (upper right) and severe gap formation after thrombin stimulation (lower right). b MIR155 transfected HUVECS (MIR155^high) were grown under constant detection of the transendothelial resistance (TER) until confluency was reached. Challenge with thrombin showed a stronger decrease and slower recovery (endpoint p < 0.05) in MIR155^high compared to naïve ECs. c Continuous TER between vehicle and Anti-MIR155 challenged with thrombin showed an amelioration of maximal response and faster re-bound recovery than the corresponding control group (deepest drop point p < 0.05) d Densitometry from C57BL/6J mouse lungs challenged with either LPS (17.5 mg/kg BW, n = 5) or vehicle (NaCl 0.9%, n = 7). After 16 h they were killed and Claudin-1 (CLDN-1) and b-Actin were detected by immunoblotting (p < 0.05). e Bar graphs showing normalized mRNA of Claudin-1/bActin in MIR155^high and naïve (CTR) HUVECs (n = 6–7; p < 0.01). f Bar graphs showing densitometry results of Claudin-1/bActin immunoblots in MIR155^high versus naïve (CTR) HUVECs (n = 6–7; p < 0.01). g Bar graphs showing immunoprecipitation of the Claudin-onefold enrichment normalized to GAPDH (n = 4, p < 0.05), after the overexpression of MIR155 in HUVECs (n = 4, p < 0.05)

Effect of MIR155 depletion on clinical outcomes in murine sepsis. a Box and whiskers showing a descending activity score (explained in Additional file 1: Table S2.) that is attenuated in the heterozygote background with a significant difference at all time points calculated by Kruskal–Wallis test (*p < 0.05, **p < 0.001). b Kaplan–Meier-survival analysis of mice suspected to cecal ligature & puncture (CLP) showing improved outcome in heterozygous knockout mice (p = 0.0135 with Mantel–Cox test)

Translational evidence of MIR155 regulation in human disease. a Box and whisker plots showing normalized circulating MIR155 levels in serum of healthy controls (CTR, n = 5) and patients with acute respiratory distress syndrome (ARDS, n = 16) b Box and whisker plots showing normalized MIR155 in bronchoalveolar lavage fluid (BALf) of the same patients as in a (**p < 0.001), c Box and whisker plots showing normalized MIR155 in immediate postmortem kidney biopsies from healthy control biopsies (CTR, n = 8) and septic acute kidney injuries (sep AKI, n = 12) (*p = 0.02)