Differential angiogenesis function of CCM2 and CCM3 in cerebral cavernous malformations

Abstract

Object: Loss-of-function mutations in CCM genes are frequently detected in familial cerebral cavernous malformations (CCMs). However, the current functional studies of the CCM genes in vitro have been performed mostly in commercially purchased normal cell lines and the results appeared discrepant. The fact that the cerebral vascular defects are rarely observed in CCM gene-deficient animals suggests the requirement of additional pathological background for the formation of vascular lesions. Consistent with these data, the authors assumed that silencing CCM genes in the endothelium derived from CCMs (CCM-ECs) serves as a unique and valuable model for investigating the function of the CCM genes in the pathogenesis of CCMs. To this end, the authors investigated the role and signaling of CCM2 and CCM3 in the key steps of angiogenesis using CCM-ECs.

Methods: Endothelial cells (ECs) derived from CCMs were isolated, purified, and cultured from the fresh operative specimens of sporadic CCMs (31 cases). The CCM2 and CCM3 genes were silenced by the specific short interfering RNAs in CCM-ECs and in control cultures (human brain microvascular ECs and human umbilical vein ECs). The efficiency of gene silencing was proven by real-time reverse transcriptase polymerase chain reaction. Cell proliferation and apoptosis, migration, tube formation, and the expression of phosphor-p38, phosphor-Akt, and phosphor-extracellular signal-regulated kinase-1 and 2 (ERK1/2) were analyzed under CCM2 and CCM3 silenced conditions in CCM-ECs.

Results: The CCM3 silencing significantly promoted proliferation and reduced apoptosis in all 3 types of endothelium, but accelerated cell migration exclusively in CCM-ECs. Interestingly, CCM2 siRNA influenced neither cell proliferation nor migration. Silencing of CCM3, and to a lesser extent CCM2, stimulated the growth and extension of sprouts selectively in CCM-ECs. Loss of CCM2 or CCM3 did not significantly influence the formation of the tubelike structure. However, the maintenance of tube stability was largely impaired by CCM2, but not CCM3, silencing. Western blot analysis revealed that CCM2 and CCM3 silencing commonly activated p38, Akt, and ERK1/2 in CCM-ECs.

Conclusions: The unique response of CCM-ECs to CCM2 or CCM3 siRNA indicates that silencing CCM genes in CCM-ECs is valuable for further studies on the pathogenesis of CCMs. Using this model system, the authors demonstrate a distinct role of CCM2 and CCM3 in modulating the different processes of angiogenesis. The stimulation of endothelial proliferation, migration, and massively growing and branching angiogenic sprouts after CCM3 silencing may potentially contribute to the formation of enriched capillary-like immature vessels in CCM lesions. The severe impairment of the tube integrity by CCM2, but not CCM3, silencing is associated with the different intracranial hemorrhage rate observed from CCM2 and CCM3 mutation carriers. The activation of p38, ERK1/2, and Akt signal proteins in CCM2- or CCM3-silenced CCM-ECs suggests a possible involvement of these common pathways in the pathogenesis of CCMs. However, the specific signaling mediating the distinct function of CCM genes in the pathogenesis of CCMs needs to be further elucidated.