Abstract 650

Laboratory mice are used in hemostasis and thrombosis research in experiments ranging from the investigation of basic mechanisms of thrombus formation to the generation of preclinical data regarding potential therapeutic anti-thrombotic agents. With the common ferric chloride (FeCl3) model of vascular thrombosis, FeCl3 is applied directly to outside of the carotid artery or mesenteric vessels to induce thrombus formation. However, despite its widespread use, surprisingly little is understood regarding the mechanisms by which FeCl3 induces endothelial injury and subsequent thrombus formation, although it is believed that endothelial denudation and collagen exposure occur, and trigger the onset of thrombosis. Similarly, very little is known regarding the nature of the thrombi that are generated by this method, and importantly, whether these experimental thrombi resemble bona-fide thrombi that occur as a result of “true” vascular injury. To address these issues, we developed a scanning electron microscopy (SEM) protocol to visualize endothelial damage and thrombus formation that occur in situ. Briefly, thrombus formation is initiated by direct application of FeCl3 (or vehicle control) to the carotid artery, and subsequently at defined time points, the circulation is flushed, and then internally aldehyde-fixed. The affected section of carotid artery is removed, externally fixed, sectioned, processed for SEM, and visualized. With this procedure, we have obtained high-quality and high-magnification images of FeCl3-induced endothelial damage and thrombus formation through approximately 5 minutes of injury, in multiple samples and uninjured controls (N>100). Interestingly, we found that through these time points, FeCl3 induces little, if any, endothelial damage or subendothelial exposure. Rather, the endothelium rapidly assumes an “activated” phenotype, clearly changed from baseline, but non-denuded and intact. For verification, we found that mechanical injury of the murine carotid results in endothelial denudation which is easily identified by SEM. Perhaps more surprisingly, we also found that the first cells to adhere to the endothelium following FeCl3 application actually are red blood cells (RBCs), and not platelets. Following binding to the vascular surface, RBCs become misshapen and elongated in the direction of blood flow, or break off, leaving fragments associated with the endothelial surface. Subsequently, surface-bound RBCs and fragments bind additional RBCs and platelets from the circulation, forming large and characteristic-appearing complexes. From this point forward, large numbers of additional platelets accumulate on the RBC and platelet/RBC complexes, and the thrombus grows inward rapidly until the vessel becomes occluded. Next, to investigate formation of these potential RBC and platelet/RBC complexes with an independent system, we turned to intravital fluorescent microscopy in mesenteric venules. We found that FeCl3 rapidly induces structures consistent in size and shape with the “early” platelet/RBC complexes observed by SEM. With further time points, these small/early lesions subsequently bound more labeled platelets, and increased rapidly in size until occlusion of the vessel. As these platelet lesions grew, they also resembled in size and shape the “later” platelet/RBC complexes observed by SEM, including possession of a characteristic trailing “tail” of platelets. It is interesting that similar structures appeared to form in both the carotid artery and mesenteric venules, given the marked difference in shear stress between the two (1000-1500 s-1 vs. 100–200 s-1, respectively). To our knowledge, the existence of such platelet/RBC complexes has not been documented in any system. In summary, we have developed a technique to investigate in situ endothelial damage and thrombus formation in mice by SEM, and demonstrated that thrombus formation in the commonly-used FeCl3 murine model occurs in the absence of endothelial denudation, and appears to involve the active participation of RBCs. Further work will be necessary to confirm whether or not RBCs are necessary for thrombus formation in this model, and whether RBCs participate in any type of naturally-occurring hemostasis or thrombosis. Additionally, these results will likely have strong impact on future interpretation of experiments resulting from the use of FeCl3 in mice.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.