Background: Tyrosine kinase inhibitors (TKIs) have revolutionized treatment of chronic myeloid leukemia (CML) resulting in long term survival with chronic therapy. Vascular toxicities have emerged as complications of the second generation CML TKIs; dasatinib, ponatinib and nilotinib. Dasatinib is associated with increased risk of pulmonary hypertension, ponatinib with arterial ischemic events and nilotinib with peripheral arterial occlusive events and cardiovascular ischemia by unknown mechanisms. The first generation CML TKI imatinib is not associated with vascular toxicity.

Hypothesis: We hypothesize that the CML TKIs- ponatinib, dasatinib and nilotinib, but not imatinib, cause vascular toxicity by affecting endothelial cell (EC) function in association with alterations in phospho-signaling and histone modifications.

Methods To test this hypothesis, we studied the effect of CML TKIs imatinib, nilotinib, dasatinib and ponatinib on EC function using three in-vitro assays measuring EC survival, EC-leucocyte adhesion (a measure of vascular inflammation) and scratch wound closure (a measure of wound healing capacity). Cultured human umbilical vein ECs (HUVECs) were treated with vehicle, imatinib (4uM), dasatinib (200nM), nilotinib (3 uM) or ponatinib (100nM). Cell viability was measured at 18 hours using a Cell TiterGlo Luminescent Cell Viability assay (Promega). For the leucocyte adhesion assay, fluorescent labeled leucocytes (U937) were incubated with the ECs and adherent fluorescent cells were quantified. For the wound healing assay, a scratch "wound" was made with a plastic P1000 pipette tip and cells that migrated into the wound area were counted at 18 hours. To assess signaling, two novel, global, unbiased, high throughput proteomic assays developed at the Broad Institute were used. The P100 assay is a mass spectrometry-based targeted proteomics assay that quantifies 100 representative phosphopeptides. The global chromatin profiling (GCP) assay uses mass spectrometry to simultaneously quantify over 70 bulk histone modifications including phosphorylation, ubiquitination, methylation and acetylation. HUVECs were treated in triplicate with each drug for 3 hours for the P100 assay and for24 hours for the GCP assay. For each assay, the peptides were mixed with isotopically labeled internal standards and mass spectrometry was performed to quantify each phospho-peptide or modified histone peptide relative to the universal standard.

Results: The CML TKIs that are associated with vascular toxicity (nilotinib, ponatinib and dasatinib) negatively impacted all the EC functions measured. Specifically, nilotinib, ponatinib and dasatinib significantly decreased endothelial cell viability and endothelial cell migration in response to a scratch wound and increased leucocyte adhesion to ECs. Imatinib did not have a negative effect on ECs in any of those assays. Using the mass spectrometry based P100 and GCP assays, we found that nilotinib affects histone epigenetics and protein phosphorylation in HUVECs. In particular, nilotinib treatment modulates ABL1, MAP4, BRD4 and JUND phosphorylation and also increases histone H3K9 trimethylation with H3S10 phosphorylation in HUVECS. We also compared the nilotinib profiles obtained in HUVECs to that of primary human aortic smooth muscle cells (HAoSMC) and several cancer cell lines including MCF7 (breast), A549 (lung), A375 (skin) and PC3 (prostate) (Figures 1a and 1b). While some changes are unique to ECs, others are similar to the effects in cancer cells and smooth muscle cells. GCP assay and P100 assays comparing the effects of all the CML TKIs in endothelial cells are in process.

Conclusions: CML TKIs associated with vascular toxicity affect human EC function in vitro resulting in impaired endothelial survival and wound healing and increase inflammation. These changes are associated with alterations in signaling and epigenetic pathways in ECs in response to treatment with nilotinib. Increased vascular inflammation and impaired vascular healing are associated with cardiovascular thrombosis and ischemia. Further studies comparing the effects of all CML TKIs on ECs will be used to elucidate molecular mechanisms of vascular toxicity and help devise strategies to overcome these toxicities while maintaining anti-cancer efficacy.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution