Laser-induced endothelial cell activation supports fibrin formation

Abstract

Laser-induced vessel wall injury leads to rapid thrombus formation in an animal thrombosis model. The target of laser injury is the endothelium. We monitored calcium mobilization to assess activation of the laser-targeted cells. Infusion of Fluo-4 AM, a calcium-sensitive fluorochrome, into the mouse circulation resulted in dye uptake in the endothelium and circulating hematopoietic cells. Laser injury in mice treated with eptifibatide to inhibit platelet accumulation resulted in rapid calcium mobilization within the endothelium. Calcium mobilization correlated with the secretion of lysosomal-associated membrane protein 1, a marker of endothelium activation. In the absence of eptifibatide, endothelium activation preceded platelet accumulation. Laser activation of human umbilical vein endothelial cells loaded with Fluo-4 resulted in a rapid increase in calcium mobilization associated cell fluorescence similar to that induced by adenosine diphosphate (10 μM) or thrombin (1 U/mL). Laser activation of human umbilical vein endothelial cells in the presence of corn trypsin inhibitor treated human plasma devoid of platelets and cell microparticles led to fibrin formation that was inhibited by an inhibitory monoclonal anti-tissue factor antibody. Thus laser injury leads to rapid endothelial cell activation. The laser activated endothelial cells can support formation of tenase and prothrombinase and may be a source of activated tissue factor as well.

Figure 1

Activation of arteriolar endothelium in vivo by laser-induced injury. The endothelium of the cremaster microcirculation was loaded with either Fluo-4 AM or DIOC₆ by systemic infusion of Fluo-4 AM/Cremophore or DIOC₆ via the femoral artery. A period of 20 minutes was allowed after infusion for uptake and de-esterification of Fluo-4 AM or uptake of DIOC₆ within the endothelium. Concurrently platelet accumulation was inhibited by infusion of eptifibatide. After laser-induced vessel wall injury, changes in endothelial Ca²⁺ levels were observed by excitation at 488 nm. (A) Representative composite fluorescence and brightfield images after vessel injury show Ca²⁺ elevation in the endothelium in the absence of platelet accumulation. The fluorescence signal is presented as a pseudocolor intensity map where blue represents the least intense and red represents most intense fluorescence signal. (B) Calcium elevation after vessel injury as determined by the median integrated fluorescence intensity (y-axis) from Fluo-4–loaded endothelium (top black curve, 27 thrombi from 3 mice) in comparison to the median integrated fluorescence of the inert dye DiOC₆ (bottom gray curve, 15 thrombi from 3 mice). (C) Propagation of Ca²⁺ elevation and endothelial activation along the vessel after injury is presented as a pseudocolor intensity map as in panel A. Image is representative of peak endothelial activation, the site of injury is marked (X) and one line demarking the vessel wall on the same side as the injury (purple) and a second demarking the opposing side (gray). (D) Representative trace from a single experiment showing quantitation of the Fluo-4 fluorescence signal longitudinally along each vessel wall. A line was drawn along each wall and the intensity of the pixels at each step along this line was determined and plotted. The start of the line (0 pixels) is bottom right and the end (∼ 300 pixels) is top left.

Figure 2

In vivo imaging of platelet accumulation concomitantly with endothelial calcium elevation after laser-induced injury. Platelets were labeled with anti–mouse CD41 Fab fragments conjugated to Alexa 647 infused via the jugular vein, and the endothelium of the cremaster microcirculation was loaded with Fluo-4 AM. After laser-induced vessel wall injury activation of the endothelium and thrombus formation were observed and recorded over time. (A) Composite fluorescence and brightfield images after vessel injury show Ca²⁺ elevation (green) in the endothelium in conjunction with and preceding platelet accumulation (red) or presence of both signals (yellow). The fluorescence signal is shown binarized for ease of visual interpretation. (B) Kinetic curves displaying the median integrated Fluo-4 fluorescence (gray curve on left y-axis) and median integrated platelet fluorescence (black curve on right y-axis) for 34 thrombi in 3 wild-type mice. Fluo-4 fluorescence originating from platelets and not the endothelium was eliminated by subtracting any Fluo-4 fluorescence in pixels where Alexa 647 fluorescence was observed.

Figure 3

Exposure of activation and secretion marker LAMP-1 in comparison with platelet accumulation in vivo after laser injury. Anti–mouse CD41 Fab fragments conjugated to Alexa 647 and anti–mouse LAMP-1 antibodies conjugated to Alexa 488 were infused to label platelets and LAMP-1, respectively. Injuries were induced by laser pulses to cremaster arteriole vessel walls and subsequent LAMP-1 accumulation and thrombus formation recorded over time. (A) Images from a representative experiment showing fluorescence over time overlaying brightfield data before and after vessel injury. LAMP-1 accumulated rapidly at the vessel wall (green) followed by platelet accumulation (red) or presence of both signals (yellow). The fluorescence signal is shown binarized for ease of visual interpretation. (B) Kinetic curves displaying the median integrated platelet fluorescence (black curve on right y-axis) and median integrated LAMP-1 fluorescence (gray curve on left y-axis) for 18 thrombi in 3 wild-type mice. LAMP-1 fluorescence originating from platelets and not the endothelium was eliminated by subtracting any LAMP-1 fluorescence in pixels where Alexa 647 fluorescence was observed.

Figure 4

Rapid endothelial cell activation in vitro follows targeted laser pulse. HUVECs were loaded with Fluo-4 AM (3μM) and observed using fluorescence microscopy. (A) Representative images of a cell before and after a direct laser pulse to the point the indicated (X). An increase in Fluo-4 fluorescence (green) reflects a rise in intracellular Ca²⁺. (B) Quantification of this signal is plotted against time showing 1 representative trace (solid line) and the mean trace of laser-induced activation of 41 cells (dotted line). (C) Similarly prepared HUVECs loaded with Fluo-4 AM were stimulated with ADP (10μM), thrombin (1 U/mL), or histamine (10μM) as a comparison to the laser-induced activation. Agonists or vehicle were added after 10 seconds of image acquisition. The graph shows median curves as a comparison of the kinetics of these modes of activation. ADP, solid line; thrombin, – – –; histamine, -.-.; laser, ---; vehicle ⋯. (D) The peak cell activation was extracted from the kinetic data for each agonist and the laser. The mean ± SEM is plotted for each group; ADP stimulation, n = 82 cells from 7 experiments, histamine n = 45 cells from 3 experiments; thrombin n = 85 cells from 6 experiments; vehicle n = 10 cells from 2 experiments.

Figure 5

Rapid propagation of endothelial cell activation within a confluent cell population follows laser-induced activation in vitro. HUVECs were loaded with Fluo-4 AM. Cells were observed for 1 minute prior to activation to confirm a stable baseline then subjected to a single pulse from a nitrogen laser and continuous imaging. (A) Images from different time points of a representative cell population. The first frame shows the cell monolayer in its basal state just prior to activation and the X marks the point upon which the laser is focused. The subsequent time points show activation of the target cell (within 1 second of laser firing), closely followed by activation of surrounding cells and then steady return toward a basal cytoplasmic Ca²⁺ concentration. (B) The mean pixel intensity of the 4 cells is plotted versus time to yield the corresponding 4 kinetic traces shown on the right. Cell 1, solid line; Cell 2, -.-.; Cell 3, - - -; Cell 4, ⋯.

Figure 6

Laser activated endothelial cells can induce thrombin generation. HUVECs plated on photo-etched coverslips were loaded with Fluo-4 AM and incubated with plasma in presence of calcium and corn trypsin inhibitor for 15 minutes. Cells were either stimulated with laser or left unstimulated in the presence or absence of various antibodies. After incubation with plasma, cells were fixed and immunostained for fibrin (red), phalloidin (green) and DAPI (blue). (A) Representative images of cells not stimulated by laser and incubated with recalcified plasma show minimal fibrin specific Alexa 647 signal (red). (B) Detection of fibrin-specific signal (red) after laser induced injury of a single cell in a field in presence of recalcified plasma. (C) Representative image showing lack of fibrin formation after laser injury when the cells were incubated with recalcified plasma containing 100 μg/mL function blocking tissue factor monoclonal antibody cH36. (D) No signal was detected when laser stimulated cells were immunostained with an isotype matched control IgG instead of fibrin antibody in the presence of plasma. (E) Fibrin meshwork was detected on cells activated with laser and incubated with recalcified plasma in the presence of an isotype matched control human IgG instead of the monoclonal antibody cH36. (F) Mean integrated fluorescence signal intensity of fibrin (n = 29). Data are from 2 independently performed experiments. The mean ± SEM is plotted for each group. A background mask was created for all images from panel D. Mean of the maximum signal intensities from 29 images in panel D was used as a constant background number to create a threshold segment mask for all other conditions and integrated fluorescence intensity was calculated. The means of the integrated fluorescence intensity show a significant decrease in fibrin generation when laser stimulated endothelial cells are incubated with recalcified plasma in presence of function blocking tissue factor antibody.