All animal experiments were performed according to German law for animal protection and approved by the local government authorities (Regierung von „Oberbayern“). The use of human autopsy specimen was approved by the ethics committee of the TUM (Ref. 225/20S).

Male C57BL/6J mice were purchased from Charles River (Sulzfeld, Germany). Male VN^-/- mice were generated as described previously (32) and backcrossed to the C57BL/6J background for 6 – 10 generations. All experiments were performed using mice at the age of 10-12 weeks. Animals were housed under conventional conditions with free access to food and water.

Lipopolysaccharide (LPS, Escherichia coli O111:B4; 1 mg in 600 μl NaCl intraperitoneally (i.p.); Sigma-Aldrich, Schnelldorf, Germany) was used to induce a systemic inflammation driven by a robust release of pro-inflammatory cytokines ultimately leading to the development of systemic inflammatory response syndrome (SIRS) associated with subsequent organ failure such as acute respiratory distress syndrome (ARDS) (33, 34). After 6 hours, mice were sacrificed and flushed with PBS. Subsequently, organs were harvested.

For analyzing the formation of microvascular immunothrombi, liver, kidneys, and lungs of endotoxemic mice were embedded in TissueTek (Sakura Finetek) and frozen using dry ice. Afterwards, the organs were cut into 50 µm slices, put on glass slides, and fixed in 4% paraformaldehyde. Tissues were then washed with PBS and blocked with 2% Bovine Serum Albumin (BSA; Sigma Aldrich) in PBS for 60 minutes at 4° C. After another washing step with PBS, samples were incubated at 4° C for 24 hours with fluorescence-labeled monoclonal antibodies (mAb) directed against GPIbβ (Dylight649; emfret, Eibelstadt, Germany) and GR-1 (PE; Invitrogen, Waltham, Massachusetts, USA). In additional experiments, primary rabbit mAb directed against vitronectin (invitrogen), fibrin(ogen) (invitrogen), or fibronectin (invitrogen) were employed, followed by incubation with secondary AF-488-conjugated anti-rabbit antibodies for 120 minutes at room temperature. In further experiments, rabbit AF-555-conjugated anti-vWF (Bioss Inc., Woburn, Massachusetts, USA), rat Dylight649-conjugated anti-GPIbβ (emfret), and rat BV421-conjugated anti-GR-1 (biolegend, San Diego, California, USA) mABs were used. Immunostained tissues were mounted in PermaFluor (Beckman Coulter, Fullerton, CA). Z-stacks typically covering 30 μm (z-spacing 0.2 μm; objective magnification 20 x) were acquired using a Zeiss confocal laser-scanning microscope (LSM 900 with Airyscan 2; Carl Zeiss, Oberkochen, Deutschland). Acquired images were analyzed with Fiji ImageJ software (National Institutes of Health, Bethesda, Maryland, USA).

For the analysis of adhesion and signaling molecules in microvascular thrombi, excised mouse cremaster muscles were stained after induction of inflammation and photochemical injury as described below. Mouse cremaster muscles were then fixed in 4% paraformaldehyde. Subsequently, tissues were washed, blocked, and permeabilized in PBS, supplemented with 2% BSA (Sigma Aldrich) and 0.5% Triton X-100 (Sigma Aldrich) for 2 hours. After incubation at 4° C for 24 hours with rat mAb directed against GP1bβ (Dylight 488; emfret), GR-1 (brilliant violet 421; biolegend), CD11a (PE; biolegend), CD11b (PE; biolegend), CD54 (PE; biolegend), CD51 (PE; biolegend), or CD41/CD61 (PE, emfret), tissues were analyzed by confocal microscopy as described above.

To study the effect of VN-deficiency on immune cell infiltration of liver, kidneys, and lungs of mice with severe systemic inflammation, organs were harvested 6 h after intraperitoneal injection of LPS (see above) and homogenized. The platelet count as well as the total white blood count of each organ homogenate was determined using a ProCyte Dx cell counter (IDEXX Laboratories, Westbrook, ME, USA). After incubation with anti-CD45-APC/Cy7 mAb (clone 30-F11; BD Bioscience), anti-CD11b-FITC mAb (clone M1/70; eBioscience, Waltham, Massachusetts, USA), anti-GR-1-PE mAb (clone RB6-8c5; eBioscience), anti-F4/80-eFluor 450 mAb (clone BM8; eBioscience) for 30 min on ice, lysis of erythrocytes was performed using lysing solution (BD Biosciences) for 10 min at room temperature. After washing, immunostained cells were analyzed on a flow cytometer (Gallios flow cytometer; Beckman Coulter Inc, Brea, California, USA). Approximately 20,000 gated events were collected in each analysis and isotype-matched controls were used in all experiments. Data analysis was performed using FlowJo software (Becton, Dickinson and Company). Absolute immune cell numbers were calculated using the white blood count and the relative share of the respective immune cell subset as assessed by flow cytometry.

The surgical preparation of the mouse cremaster muscle was performed as originally described by Baez with minor modifications. Briefly, mice were anesthetized using a ketamine/xylazine mixture (100 mg/kg ketamine and 10 mg/kg xylazine), administered by i.p. injection. The left femoral artery was cannulated in a retrograde manner for administration of antibodies. The right cremaster muscle was exposed through a ventral incision of the scrotum. The muscle was opened ventrally in a relatively avascular zone, using careful electrocautery to stop any bleeding, and spread over the pedestal of a specialized custom-made microscopy stage. Epididymis and testicle were detached from the cremaster muscle and placed into the abdominal cavity. Throughout the procedure as well as after surgical preparation during in vivo microscopy, the muscle was superfused with warm buffered saline.

The setup for in vivo microscopy was centered around an AxioTech-Vario 100 Microscope (Zeiss MicroImaging GmbH, Goettingen, Germany), equipped with a Colibiri LED light source (Zeiss MicroImaging GmbH) for fluorescence epi-illumination microscopy. Light was directed onto the specimen via the filter set 62 HE (Zeiss MicroImaging GmbH) fitted with dichroic and emission filters [TFT 495 + 610 (HE); TBP 527 + LP615 (HE)]. Microscopy images were obtained with an AxioCam Hsm digital camera using a 20x water immersion lens (0.5 NA, Zeiss MicroImaging Gmbh). The images were processed with AxioVision 4.6 software (Zeiss MicroImaging GmbH).

Lipopolysaccharide (LPS, Escherichia coli O111:B4; 10 ng intrascrotally (i.s.); Sigma-Aldrich, Schnelldorf, Germany) was used to induce inflammation of the cremaster muscle. After 6 hours, microvascular thrombus formation was induced by photochemical injury as described before. Briefly, animals received an i.a. injection of a FITC-dextran solution (150 kDa, 2.5%, 6 ml/kg body weight, Sigma-Aldrich). Five minutes later, the vessel of interest was exposed to continuous epi-illumination using the FITC filter cube and appropriate illumination (λ = 488 nm). The field of illumination covered the entire field of view under investigation. To assure intergroup comparability, the mean fluorescence intensity was determined in each of the analyzed vessels immediately after onset of light exposure. The onset of the thrombus was defined as the first observation of endothelially adherent platelets during thrombus formation and the cessation of thrombus formation was defined as the complete occlusion of the vessel segment with flow cessation for over 60 s.

To determine the functional relevance of distinct adhesion and signaling molecules for microvascular thrombus formation in inflamed tissue, experiments were performed upon i.a. administration of blocking anti-CD11a mAbs (clone M17/4; 50 μg in 100 μl NaCl; biolegend, San Diego, CA, USA), anti-CD11b mAbs (clone M1/70; 50 μg in 100 μl NaCl; biolegend), anti-CD54 mAbs (clone YN1/1.7.4; 50 μg in 100 μl NaCl; biolegend), anti-CD51 mAbs (clone RMV-7; 50µg in 100µl NaCl; biolegend), or anti-CD61/41 mAbs (clone Leo.H4; 50 μg in 100 μl NaCl; emfret). In additional experiments, heparin (500 IU intraarterially (i.a.); ratiopharm GmbH, Ulm, Germany), a non-peptic GPIIb/IIIa inhibitor (GR144053 trihydrochloride; 10 mg/kg i.a.; 10 min prior to induction of thrombosis; R&D Systems, Lille, France), or drug vehicle were applied 10 min prior to induction of microvascular thrombus formation. Batroxobin (3 BU/kg; Defibrase^®; DSM Nutritional Products Ltd, Branch Pentapharm, Switzerland) was applied i.p. 12 h prior to induction of microvascular thrombus formation.

To assess interactions of platelets, neutrophils, and endothelial cells during microvascular thrombogenesis, neutrophils were identified by anti-Ly6G mAbs (clone 1A8; 10 μg in 100 μl NaCl i.a.; BD Biosciences), platelets by anti-GPIbβ mABs (clone X-649; 10 μg in 100 μl NaCl i.a.; emfret), and endothelial cells by anti-PECAM-1/CD31 (clone 390; 3 μg in 100 μl NaCl i.a; eBiosciences).

To analyze the effect of VN and vWF on hemostasis, recombinant mouse VN and vWF was added in different doses (1 ng/ml, 10 ng/ml, 100 ng/ml) to whole blood samples harvested from C57BL/6J mice. Rotational thromboelastometry (ROTEM, Pentapharm, Munich, Germany) was then performed (35). In separate experiments, the effect of a blocking anti-CD61/41 mAb upon stimulation with VN and vWF, of heparin or heparin together with a blocking anti-CD61/41 mAb upon stimulation with ADP, as well as of the effect of heteromerization of VN with recombinant murine PAI-1 (CPAI; 50 μg in 100 μl PBS; Molecular Innovations, Novi, MI) were evaluated.

To evaluate the effect of VN on platelet activation, platelets from the whole blood of C57BL/6J mice were incubated with PBS, recombinant mouse VN (100 ng/ml; sinobiological, Beijing, China), or ADP (10 ng/ml; Sigma-Aldrich, St. Louis, Missouri, USA). As a measure of platelet activation, surface binding of the high-affinity conformation-specific anti-integrin αIIbβ3 mAb conjugated with PE (clone JON/A; emfret) was assessed by flow cytometry (see above). Platelets were identified as CD45^- and CD41⁺ cells using anti-CD45 mAb labeled with APC/Cy7 (clone 30-F11; BD Bioscience) and anti-CD41 mAb conjugated with APC (clone MWReg30, biolegend). The cells were analyzed using the BD FACSCanto II (Becton, Dickinson and Company, Franklin Lakes, New Jersey, USA).

Spreading of platelets from the whole blood of C57BL/6J mice was analyzed in blood smears mounted in PermaFluor by confocal microscopy upon incubation with PBS, recombinant murine VN (100 ng/ml; sinobiological, Beijing, China), or ADP (10 ng/ml; Sigma-Aldrich, St. Louis, Missouri, USA) for 60 minutes at 37°C on glass, collagen I, or fibrin(ogen). The spreading area of single platelets was measured using Fiji ImageJ.

Platelet adhesion under varying shear rates was assessed in ibidi µ-slides VI0.1 (Ibidi, Gräfelfing, Germany). Slides were coated either with fibrin(ogen) (100 µg/ml in TBS, Sigma-Aldrich, Taufkirchen, Germany) or Vitronectin XF (25 µg/ml in TBS, STEMCELL Technologies, Köln, Germany) together with collagen-related peptide (CRP, 0.5 µg/ml, kindly provided by Prof. Siess, LMU, Munich) at 4°C overnight. Blood collected from wild-type mice was mixed with 1/2 volume of heparin (20 U/ml in TBS) and perfused through fibrinogen/CRP and vitronectin/CRP-coated flow chambers at defined flow rates resulting in 1, 5, 10 or 50 dyne/cm2 wall shear stress for 4 min using a PHD ULTRA pump (Harvard Apparatus, Holliston, MA, USA). Subsequently, flow chambers were washed by perfusing Tyrode’s buffer (136 mM NaCl, 0.43 mM NaH2PO4, 2.7 mM KCl, 12mM NaHCO3, 2 µM CaCl2, 1 µM MgCl2, 5 mM HEPES, 0.1% glucose, 0.35% BSA; pH 7.35) at the respective flow rate for 20 min and adherent platelets were stained with CellTrace CFSE dye (Thermo Fisher Scientific, Darmstadt, Germany). Images were acquired with an Evos M7000 life cell microscope (Thermo Fisher Scientific) and surface coverage was quantified using ImageJ software.

3 different groups with 10 individuals each were analyzed: i) lungs of patients with severe systemic infectious inflammation (COVID-19; individuals with COVID-19 confirmed by PCR for SARS-CoV-2 were included; autopsies were performed at the Institute of Pathology of the Technical University of Munich (TUM) between June 2020 and December 2020), ii) lungs of patients with severe systemic non-infectious inflammation (pancreatitis-associated SIRS/ARDS; individuals with clinically and histopathologically confirmed diagnosis were included; autopsies were performed at the Institute of Pathology of the Technical University of Munich (TUM) between 2001 and 2016), and iii) lungs of control patients (surgical tissue samples harvested from patients with a primary lung carcinoma; samples have been obtained from lung tissue far from the cancer region (> 3 cm) and have histopathologically been classified as “healthy/non-pathological”, particularly without signs of inflammation or infection; autopsies were performed at the Institute of Pathology of the Technical University of Munich (TUM) between 2008 and 2015).

In all individuals, an ultrasound-guided minimally invasive autopsy was performed, which included retrieval of biopsies (14-Gauge) of morphologically visible peripheral lung lesions or of morphologically healthy lung tissue. Biopsies were fixed in 4% formaldehyde for 72h, paraffin-embedded (FFPE) and 2µm slides were stained with hematoxylin & eosin (HE) and LADEWIG’s solution (12086.01000, Morphisto, Offenbach, Germany) according to standard procedures. IHC was performed using a vWF antibody (clone VWF/1859R, NSJ Bioreagents, V3491) diluted 1:100 in antibody diluent or a VN antibody (MA5-32157, Invitrogen) diluted 1:1000 in antibody diluent and stained on a Bond Rxm (Leica, Wetzlar, Germany) system after heat mediated antigen retrieval (ER1 for 30 min) using a Polymer Refine Detection Kit without post-primary antibody. Stained slides were scanned using an Aperio AT2 system (Leica, Wetzlar, Germany) and visualized using Aperio ImageScope (version 12.3.1.5011). Histopathological analyses of 2 different biopsy samples of 10 different individuals per group were analyzed in this study. For assessment of fibrin (bright red in LADEWIG staining), vWF (brown in immunostaining), and VN (brown in immunostaining) deposition in the microvasculature, four vessels with a diameter of >100 µm and four vessels with a diameter =< 100 µm were randomly selected for analysis in each biopsy sample. Presence of specific staining regardless of staining intensity was evaluated as positive.

Data analysis was performed with the statistical software SigmaPlot for Windows; Jandel Scientific, Erkrath, Germany). After confirming normality and equal variance of data (using the Shapiro-Wilk and Brown-Forsythe tests), the One-way ANOVA test followed by the Dunnett test (>2 groups) or the t test (2 groups) was used for the estimation of stochastic probability in intergroup comparisons. If normality and/or equal variance testing failed, the Kruskal-Wallis One-way ANOVA of Ranks test followed by the Dunnett test (>2 groups) or the Mann-Whitney rank sum test (2 groups) was used. For survival estimation, the log rank test was employed. Mean values and SEM are given. P values <0.05 were considered significant.

In initial experiments, we sought to characterize the functional relevance of VN for microvascular immunothrombosis in systemic inflammation. Employing a mouse model of severe systemic inflammation, confocal laser scanning microscopy analyses of immunostained tissue samples detected various microvascular immunothrombi in liver, kidneys, and lungs of diseased animals, but not in healthy controls ( Figure 1A ). In these blood clots, significantly higher levels of VN and – to a lesser degree – of vWF were identified than in non-thrombotic microvessels ( Figure 1B ). Here, dense networks of VN tightly framed platelets and immune cells in the thrombotic microvasculature ( Figure 2A ). In contrast, fibrin(ogen) was only enriched in the thrombotic lower-shear liver sinusoids, but not in the thrombotic higher-shear pulmonary and renal microvessels, whereas FN did not preferentially accumulate in the thrombotic microvasculature of these organs ( Figure 1B ). Flow cytometry additionally revealed that numbers of neutrophils and – to a much lesser extent – numbers of classical monocytes and platelets in homogenates of inflamed tissues harvested from VN-deficient animals were lower than in wild-type controls ( Figure 2A ).

To directly evaluate the functional relevance of VN for microvascular clot formation, we used a model of photochemical injury-elicited microvascular thrombogenesis in the inflamed mouse cremaster muscle. Multi-channel in vivo microcopy demonstrated that thrombus formation in venules (but not in arterioles; Figure 2B ) was significantly prolonged in VN-deficient mice as compared to WT controls. At baseline conditions prior to induction of thrombus formation, the number of neutrophils adherent to endothelial cells in postcapillary venules was significantly lower in VN-deficient animals than in WT controls, whereas VN-deficiency did not significantly alter adherence of platelets to endothelial cells or neutrophils. During thrombus formation, however, interactions between all these cell types were significantly reduced by VN-deficiency ( Figure 2C ). Moreover, flow chamber experiments revealed that VN – in contrast to fibrin(ogen) – facilitates platelet adhesion primarily under lower shear ( Figure S1 ). Hence, our experimental data suggest that VN promotes microvascular immunothrombosis in inflamed tissue by supporting the multi-cellular interplay required for proper clot formation.

Towards a more comprehensive understanding of VN-dependent venular thrombogenesis in inflamed tissue, we sought to define the role of different adhesion and signaling molecules in this process. The leukocyte β1 integrin VLA-4/CD49d and the β2 integrins LFA-1/CD11a and Mac-1/CD11b primarily mediate intravascular adherence of neutrophils to endothelial cells in inflamed tissue, whereas the integrins CD51/CD61 and CD41/CD61 (also: glycoprotein (GP) IIb/IIIa) serve as principal receptors of VN. Employing confocal laser scanning microscopy of immunostained cremasteric tissue whole mounts, all these molecules were detected in venular inflammatory thrombi. Interestingly, only antibody-mediated blockade of GPIIb/IIIa (but not blockade of the other molecules) significantly prolonged thrombus formation in inflamed postcapillary venules ( Figure 3A ), as evidenced in photochemical injury-elicited venular thrombogenesis. Whereas multi-channel in vivo microscopy further documented that blockade of GPIIb/IIIa significantly reduced interactions of platelets with endothelial cells or neutrophils during venular clot formation, neutrophil-endothelial cell interactions were not significantly altered ( Figure 3B ). Consequently, the VN receptor GPIIb/IIIa critically regulates platelet interactions in microvascular immunothrombosis.

In addition to VN and vWF, fibrin(ogen) is able to bind to GPIIb/IIIa. Application of heparin (which inhibits the generation of fibrin polymers) or defibrination with batroxobin (Defibrase^®), however, did not significantly change the formation of immunothrombi in the inflamed cremaster muscle upon photochemical injury ( Figure 3A ; Figure S2B ), thus corroborating the dominant role of VN and vWF in GPIIb/IIIa-dependent venular thrombogenesis. Employing rotational thrombelastometry, recombinant mouse VN dose-dependently increased the aggregation of primary mouse platelets, similar to vWF and fibrin(ogen) ( Figure 4A ). These effects of vWF and VN were severely compromised upon antibody blockade of their receptor GPIIb/IIIa ( Figure 4B ). In line with our in vivo results ( Figure 3A ), blockade of GPIIb/IIIa significantly reduced ADP-elicited platelet aggregation in vitro even after inhibition of fibrin formation by heparin, further highlighting the importance of the vWF/VN-GPIIb/IIIa axis for platelet aggregation. Application of VN-PAI-1 heteromers (which promote fibrinolysis), however, did not significantly alter platelet aggregation ( Figure 4C ). Moreover, VN did not directly activate platelets as indicated by unchanged surface expression of GPIIb/IIIa in high affinity conformation in flow cytometry ( Figure 4D ) and did not induce the spreading of these cellular blood components ( Figure 4E ; Figure S2A ). Collectively, VN primarily mediates the aggregation of platelets by involving its receptor GPIIb/IIIa.

Immunothrombotic dysregulation is a key element in the pathogenesis of severe non-infectious and infectious systemic inflammatory pathologies including COVID-19 (36). Mechanistically, SARS-CoV-2 is thought to bind to its endothelial receptor angiotensin converting enzyme-2 (ACE2) which induces a pro-inflammatory and pro-coagulative phenotype in the infected host cells (37, 38). Microvascular thrombi are likewise found in lungs of patients with severe cases of non-infectious pancreatitis-associated SIRS/ARDS promoting organ tissue damage and functional failure. Here, the pancreatitis-induced systemic inflammatory response promotes direct (e.g., complement-based ‘membrane attack complex’) as well as indirect (e.g., elevated intravascular levels of pro-inflammatory cytokines) injury to the microvascular endothelium, ultimately resulting in a pro-inflammatory and pro-coagulative phenotype (39). These processes lead to a massive accumulation of immune cells and platelets in the affected microvasculature, eventually resulting in life-threatening organ dysfunction and failure (40–42). As a translational perspective of our study, we therefore sought to evaluate the prevalence of VN in the lung vasculature of patients with severe non-infectious (pancreatitis-associated SIRS/ARDS) and infectious (COVID-19-associated) systemic inflammation. Immunohistochemical analyses of autopsy specimens detected a massive accumulation of VN intravascularly – but not in the underlying subendothelial matrix – particularly in microvessels of the pulmonary vasculature of COVID-19 and pancreatitis-associated SIRS/ARDS patients. Conversely, deposition of fibrin(ogen) and vWF was found to be elevated both in micro- and macrovessels and was additionally less pronounced in microvessels as compared to VN under systemic inflammatory conditions ( Figure 4F ). Noteworthy, microvascular pulmonary VN deposition was slightly lower in pancreatitis-associated SIRS/ARDS patients as compared to COVID-19 patients.