Adamts-18 Is a Novel Candidate Gene Of Vascular Development That Is Related To Aggravated Thrombosis: Evidence From a Adamts-18 Knock Out Mice

HIV-ITP patients have a unique antibody (Ab) against platelet GPIIIa49-66 which induces oxidative platelet fragmentation in the absence of complement (Cell 106: 551, 2001; JCI 113: 973, 2004). The search for a physiologic ligand that could induce this reaction was undertaken by panning the GPIIIa49-66 peptide with a phage surface display 7-mer peptide library. From 20 positive clones, 1 had 70% identity with a C-terminal region of ADAMTS-18 (a disintegrin and metalloproteinase with thrombospondin (TSR)-like motif 18), which is secreted by endothelial cell (EC). The recombinant C-terminal fragment of ADAMTS-18 can completely dissolve platelet aggregates formed in vitro. Moreover, this fragment lyses thrombi formed in the cerebral artery of mice and reduces infarction and neurologic impairment in murine ischemic stroke model (Blood 113: 6051, 2009). However, whether ADAMTS-18 represents the dominant physiologic mechanism controlling thrombus dissolution in vivo remains to be clarified.

Here, we used ADAMTS-18-deficient (ADAMTS-18^-/-) mice to study the contributions of ADAMTS-18 to thrombus formation in vivo. To investigate possible functional differences between WT and ADAMTS-18^-/- platelets, we tested WT and ADAMTS-18^-/- platelets in a model of pulmonary thromboembolism induced by infusion of a mixture of platelet agonist collagen (25 µg per mouse) and epinephrine (1 µg per mouse). In lung tissue Hematoxylin and eosin-stained (HE) slides, the mean number of thrombi per lung was same in the ADAMTS-18^-/- group compared with WT group (163.7 ±14.38 vs 174.9 ±11.73, n=30/group, P=0.5480). In vitro, there is no difference between WT and ADAMTS-18^-/- platelet aggregation trace and activation initiated by various platelet agonists ADP (10 µM) or collagen (2 µg/mL). No difference was noted on WT and ADAMTS-18^-/- platelet adhesion on immobilized ligand (fibrinogen). These results indicated ADAMTS-18 had no effect on platelet function.

We next evaluate the effect of ADAMTS-18 on thrombus formation in a second well-established carotid artery thrombosis model, which is induced by 10% FeCl3 patch. In the process of surgical operation, we unexpectedly observed that all ADAMTS-18^-/- mice have premature common carotid artery bifurcation compared with WT mice. A Doppler flow monitor showed ADAMTS-18^-/- mice exhibited significantly reduced carotid artery blood flow than WT mice (ADAMTS-18^-/- vs WT, 0.5 ± 0.11 vs 0.75 ± 0.21 mL/min, n=7/group, P=0.0298), which results in shortened time of thrombus formation (ADAMTS-18^-/- vs WT, 452.17 ± 68.88 vs 611.43 ± 92.02 sec, n=7/group, P=0.0005 ). Immunohistochemistry staining showed that the common carotid artery of ADAMTS-18^-/- mice had increased adventitial collagen deposition compared with WT mice. In vivo matrigel plug assay demonstrated that ADAMTS-18^-/- mice had significantly lower density of blood vessels compared to the WT mice. Since the middle cerebral artery arises from the internal carotid artery, we conjecture that ADAMTS-18^-/- mice would have aggravated brain infarction for the less cerebral blood flow supplying. This proved to be true. In transient middle cerebral artery occlusion (tMCAO) model, the infarction size in ADAMTS-18^-/- mice was significantly larger than in WT mice (mean infarction %, 25.68 ± 4.13 vs 17.41 ± 3.24, n=8, P=0.0012). Taken together, these observations suggest vasculature is the potential site of action of ADAMTS-18. To our knowledge, this is the first validation study of linkage and association of ADAMTS-18 as a pro-vasculature gene that is related to aggravated thrombosis.