The balanced translocation of 5′ segment of the breakpoint cluster region (BCR) gene on chromosome 22 to 3′ segment of the Abelson leukemia virus (ABL) gene on chromosome 9 in chronic myeloid leukemia (CML), results in the formation of the BCR-ABL hybrid fusion gene, which has constitutively upregulated tyrosine kinase activity. Imatinib mesylate, a small molecule developed specifically to inhibit BCR-ABL kinase activity, has revolutionized the treatment of CML. Despite the overwhelming success, about 20% of patients develop resistance to imatinib due to point mutations in the BCR-ABL kinase domain. Dasatinib is one of the second generation tyrosine kinase inhibitors (TKI) that was developed to overcome the resistance to imatinib. However, dasatinib inhibits, in addition to BCR-ABL kinase, other tyrosine kinases such as Src family kinases (SFKs), leading to its unique side effects. Clinically, one of the most common non-hematologic side effects of dasatinib is peripheral edema, pleural effusion and pulmonary hypertension and it is seen 20% of patients treated with dasatinib. We investigated the effect of dasatinib on the barrier function of human microvascular endothelial cells in vitro and in vivo.


In vitro endothelial permeability assays were performed with endothelial cells grown on transwell polycarbonate membranes inserts. FITC-Dextran was added to the upper chamber followed by dasatinib (0-100 ng/ml) or vehicle (DMSO) and at various time intervals, the medium in the lower chamber was collected and the fluorescence intensity was measured. In some experiments, the monolayers were incubated with ROCK1 inhibitor y27632 for 10 minutes before the addition of dasatinib. RhoA activity assay was performed and quantified using the RhoA activation assay kit. Spreading of control and dasatinib treated endothelial cells were quantified by Image J software. Actin cytoskeleton was visualized by Immunocytochemistry. Phosphorylation of cytoskeletal proteins were assessed by Western blot with phospho specific antibodies. In vivo endothelial permeability was measured by quantifying the extravasation of intravenously administered Evans Blue to lung parenchyma in mice.


The permeability of human microvascular endothelial cells (HMEC1) to FITC-dextran increases in Transwell chambers within 5 minutes following the addition of therapeutic concentrations of dasatinib. These changes in permeability are associated with increased activation of RhoA GTPase and its effector Rho associated coiled-coil kinase I (ROCKI). RhoA inhibitor C3 almost completely inhibits dasatinib-induced increase in permeability. Under similar conditions, imatinib had no effect on permeability and it does not activate RhoA. Since integrin-induced cell spreading suppresses RhoA activation, we examined the effect of dasatinib on cell spreading on fibronectin substrate. Dasatinib impairs endothelial cell spreading in a concentration-dependent manner and induces disorganization of actin fibers. Tyrosine kinases play an essential role in transmitting signals from integrins to RhoA and we examined tyrosine phosphorylation of several cytoskeletal proteins. Dasatinib markedly inhibits tyrosine phosphorylation of p130 Crk associated substrate (p130cas), paxillin and vinculin. These results suggest inhibition of tyrosine phosphorylation of the focal adhesion plaque components by dasatinib may alter the assembly of actin fibers resulting in the activation of RhoA/ROCK pathway. In vitro, consistent with these findings, dasatinib-induced increase in the permeability is blocked by ROCK inhibitor y-23632. In vivo administration of y-27632 significantly inhibits the dasatinib-induced extravasation of Evans Blue to lung parenchyma in mice. Furthermore, dasatinib-induced microvascular permeability is attenuated in ROCK1-deficient mice.


Our findings suggest that dasatinib inhibits actin fiber reorganization and promotes endothelial permeability through RhoA-ROCK pathway and ROCK inhibitors could serve as a therapeutic modality to ameliorate the dasatinib-induced pulmonary changes.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution